{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 436.28808864265926,
  "global_step": 630000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.35,
      "learning_rate": 1.999307479224377e-05,
      "loss": 5.6258,
      "step": 500
    },
    {
      "epoch": 0.69,
      "learning_rate": 1.9986149584487535e-05,
      "loss": 5.4807,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "eval_loss": 5.224731922149658,
      "eval_runtime": 75.7111,
      "eval_samples_per_second": 6503.99,
      "eval_steps_per_second": 2.127,
      "step": 1444
    },
    {
      "epoch": 1.04,
      "learning_rate": 1.9979224376731302e-05,
      "loss": 5.3786,
      "step": 1500
    },
    {
      "epoch": 1.39,
      "learning_rate": 1.9972299168975073e-05,
      "loss": 5.2907,
      "step": 2000
    },
    {
      "epoch": 1.73,
      "learning_rate": 1.996537396121884e-05,
      "loss": 5.2208,
      "step": 2500
    },
    {
      "epoch": 2.0,
      "eval_loss": 5.035708427429199,
      "eval_runtime": 74.9448,
      "eval_samples_per_second": 6570.491,
      "eval_steps_per_second": 2.148,
      "step": 2888
    },
    {
      "epoch": 2.08,
      "learning_rate": 1.9958448753462603e-05,
      "loss": 5.163,
      "step": 3000
    },
    {
      "epoch": 2.42,
      "learning_rate": 1.9951523545706373e-05,
      "loss": 5.1119,
      "step": 3500
    },
    {
      "epoch": 2.77,
      "learning_rate": 1.994459833795014e-05,
      "loss": 5.0685,
      "step": 4000
    },
    {
      "epoch": 3.0,
      "eval_loss": 4.928272724151611,
      "eval_runtime": 74.5205,
      "eval_samples_per_second": 6607.9,
      "eval_steps_per_second": 2.16,
      "step": 4332
    },
    {
      "epoch": 3.12,
      "learning_rate": 1.9937673130193907e-05,
      "loss": 5.034,
      "step": 4500
    },
    {
      "epoch": 3.46,
      "learning_rate": 1.9930747922437674e-05,
      "loss": 5.0038,
      "step": 5000
    },
    {
      "epoch": 3.81,
      "learning_rate": 1.992382271468144e-05,
      "loss": 4.9748,
      "step": 5500
    },
    {
      "epoch": 4.0,
      "eval_loss": 4.848881721496582,
      "eval_runtime": 73.1546,
      "eval_samples_per_second": 6731.277,
      "eval_steps_per_second": 2.201,
      "step": 5776
    },
    {
      "epoch": 4.16,
      "learning_rate": 1.991689750692521e-05,
      "loss": 4.9513,
      "step": 6000
    },
    {
      "epoch": 4.5,
      "learning_rate": 1.9909972299168975e-05,
      "loss": 4.9265,
      "step": 6500
    },
    {
      "epoch": 4.85,
      "learning_rate": 1.9903047091412745e-05,
      "loss": 4.9046,
      "step": 7000
    },
    {
      "epoch": 5.0,
      "eval_loss": 4.779133319854736,
      "eval_runtime": 74.0757,
      "eval_samples_per_second": 6647.578,
      "eval_steps_per_second": 2.173,
      "step": 7220
    },
    {
      "epoch": 5.19,
      "learning_rate": 1.9896121883656512e-05,
      "loss": 4.8871,
      "step": 7500
    },
    {
      "epoch": 5.54,
      "learning_rate": 1.988919667590028e-05,
      "loss": 4.8665,
      "step": 8000
    },
    {
      "epoch": 5.89,
      "learning_rate": 1.9882271468144046e-05,
      "loss": 4.8483,
      "step": 8500
    },
    {
      "epoch": 6.0,
      "eval_loss": 4.739624977111816,
      "eval_runtime": 74.1752,
      "eval_samples_per_second": 6638.657,
      "eval_steps_per_second": 2.171,
      "step": 8664
    },
    {
      "epoch": 6.23,
      "learning_rate": 1.9875346260387813e-05,
      "loss": 4.8328,
      "step": 9000
    },
    {
      "epoch": 6.58,
      "learning_rate": 1.986842105263158e-05,
      "loss": 4.8205,
      "step": 9500
    },
    {
      "epoch": 6.93,
      "learning_rate": 1.9861495844875347e-05,
      "loss": 4.8032,
      "step": 10000
    },
    {
      "epoch": 7.0,
      "eval_loss": 4.699394702911377,
      "eval_runtime": 74.6581,
      "eval_samples_per_second": 6595.725,
      "eval_steps_per_second": 2.156,
      "step": 10108
    },
    {
      "epoch": 7.27,
      "learning_rate": 1.9854570637119114e-05,
      "loss": 4.7918,
      "step": 10500
    },
    {
      "epoch": 7.62,
      "learning_rate": 1.9847645429362884e-05,
      "loss": 4.7758,
      "step": 11000
    },
    {
      "epoch": 7.96,
      "learning_rate": 1.984072022160665e-05,
      "loss": 4.7675,
      "step": 11500
    },
    {
      "epoch": 8.0,
      "eval_loss": 4.659648895263672,
      "eval_runtime": 77.484,
      "eval_samples_per_second": 6355.173,
      "eval_steps_per_second": 2.078,
      "step": 11552
    },
    {
      "epoch": 8.31,
      "learning_rate": 1.9833795013850418e-05,
      "loss": 4.7551,
      "step": 12000
    },
    {
      "epoch": 8.66,
      "learning_rate": 1.9826869806094185e-05,
      "loss": 4.7441,
      "step": 12500
    },
    {
      "epoch": 9.0,
      "eval_loss": 4.621845722198486,
      "eval_runtime": 74.377,
      "eval_samples_per_second": 6620.653,
      "eval_steps_per_second": 2.165,
      "step": 12996
    },
    {
      "epoch": 9.0,
      "learning_rate": 1.9819944598337952e-05,
      "loss": 4.7288,
      "step": 13000
    },
    {
      "epoch": 9.35,
      "learning_rate": 1.981301939058172e-05,
      "loss": 4.7199,
      "step": 13500
    },
    {
      "epoch": 9.7,
      "learning_rate": 1.9806094182825486e-05,
      "loss": 4.7093,
      "step": 14000
    },
    {
      "epoch": 10.0,
      "eval_loss": 4.591496467590332,
      "eval_runtime": 74.6789,
      "eval_samples_per_second": 6593.887,
      "eval_steps_per_second": 2.156,
      "step": 14440
    },
    {
      "epoch": 10.04,
      "learning_rate": 1.9799168975069252e-05,
      "loss": 4.7023,
      "step": 14500
    },
    {
      "epoch": 10.39,
      "learning_rate": 1.9792243767313023e-05,
      "loss": 4.6882,
      "step": 15000
    },
    {
      "epoch": 10.73,
      "learning_rate": 1.9785318559556786e-05,
      "loss": 4.6787,
      "step": 15500
    },
    {
      "epoch": 11.0,
      "eval_loss": 4.566598415374756,
      "eval_runtime": 74.19,
      "eval_samples_per_second": 6637.336,
      "eval_steps_per_second": 2.17,
      "step": 15884
    },
    {
      "epoch": 11.08,
      "learning_rate": 1.9778393351800557e-05,
      "loss": 4.668,
      "step": 16000
    },
    {
      "epoch": 11.43,
      "learning_rate": 1.9771468144044324e-05,
      "loss": 4.6627,
      "step": 16500
    },
    {
      "epoch": 11.77,
      "learning_rate": 1.976454293628809e-05,
      "loss": 4.6523,
      "step": 17000
    },
    {
      "epoch": 12.0,
      "eval_loss": 4.539627552032471,
      "eval_runtime": 74.3111,
      "eval_samples_per_second": 6626.517,
      "eval_steps_per_second": 2.167,
      "step": 17328
    },
    {
      "epoch": 12.12,
      "learning_rate": 1.9757617728531857e-05,
      "loss": 4.6439,
      "step": 17500
    },
    {
      "epoch": 12.47,
      "learning_rate": 1.9750692520775624e-05,
      "loss": 4.6371,
      "step": 18000
    },
    {
      "epoch": 12.81,
      "learning_rate": 1.9743767313019395e-05,
      "loss": 4.6281,
      "step": 18500
    },
    {
      "epoch": 13.0,
      "eval_loss": 4.52562141418457,
      "eval_runtime": 75.4167,
      "eval_samples_per_second": 6529.372,
      "eval_steps_per_second": 2.135,
      "step": 18772
    },
    {
      "epoch": 13.16,
      "learning_rate": 1.9736842105263158e-05,
      "loss": 4.6176,
      "step": 19000
    },
    {
      "epoch": 13.5,
      "learning_rate": 1.9729916897506925e-05,
      "loss": 4.6111,
      "step": 19500
    },
    {
      "epoch": 13.85,
      "learning_rate": 1.9722991689750695e-05,
      "loss": 4.6068,
      "step": 20000
    },
    {
      "epoch": 14.0,
      "eval_loss": 4.500814437866211,
      "eval_runtime": 74.0488,
      "eval_samples_per_second": 6649.993,
      "eval_steps_per_second": 2.174,
      "step": 20216
    },
    {
      "epoch": 14.2,
      "learning_rate": 1.9716066481994462e-05,
      "loss": 4.5989,
      "step": 20500
    },
    {
      "epoch": 14.54,
      "learning_rate": 1.970914127423823e-05,
      "loss": 4.5929,
      "step": 21000
    },
    {
      "epoch": 14.89,
      "learning_rate": 1.9702216066481996e-05,
      "loss": 4.5839,
      "step": 21500
    },
    {
      "epoch": 15.0,
      "eval_loss": 4.482580661773682,
      "eval_runtime": 74.9815,
      "eval_samples_per_second": 6567.273,
      "eval_steps_per_second": 2.147,
      "step": 21660
    },
    {
      "epoch": 15.24,
      "learning_rate": 1.9695290858725763e-05,
      "loss": 4.5769,
      "step": 22000
    },
    {
      "epoch": 15.58,
      "learning_rate": 1.968836565096953e-05,
      "loss": 4.5733,
      "step": 22500
    },
    {
      "epoch": 15.93,
      "learning_rate": 1.9681440443213297e-05,
      "loss": 4.5638,
      "step": 23000
    },
    {
      "epoch": 16.0,
      "eval_loss": 4.467035293579102,
      "eval_runtime": 74.2416,
      "eval_samples_per_second": 6632.726,
      "eval_steps_per_second": 2.169,
      "step": 23104
    },
    {
      "epoch": 16.27,
      "learning_rate": 1.9674515235457067e-05,
      "loss": 4.5584,
      "step": 23500
    },
    {
      "epoch": 16.62,
      "learning_rate": 1.9667590027700834e-05,
      "loss": 4.5497,
      "step": 24000
    },
    {
      "epoch": 16.97,
      "learning_rate": 1.9660664819944598e-05,
      "loss": 4.5469,
      "step": 24500
    },
    {
      "epoch": 17.0,
      "eval_loss": 4.447102069854736,
      "eval_runtime": 74.6073,
      "eval_samples_per_second": 6600.208,
      "eval_steps_per_second": 2.158,
      "step": 24548
    },
    {
      "epoch": 17.31,
      "learning_rate": 1.9653739612188368e-05,
      "loss": 4.5406,
      "step": 25000
    },
    {
      "epoch": 17.66,
      "learning_rate": 1.9646814404432135e-05,
      "loss": 4.5333,
      "step": 25500
    },
    {
      "epoch": 18.0,
      "eval_loss": 4.4333930015563965,
      "eval_runtime": 74.4269,
      "eval_samples_per_second": 6616.211,
      "eval_steps_per_second": 2.163,
      "step": 25992
    },
    {
      "epoch": 18.01,
      "learning_rate": 1.9639889196675902e-05,
      "loss": 4.5288,
      "step": 26000
    },
    {
      "epoch": 18.35,
      "learning_rate": 1.963296398891967e-05,
      "loss": 4.5174,
      "step": 26500
    },
    {
      "epoch": 18.7,
      "learning_rate": 1.9626038781163436e-05,
      "loss": 4.5178,
      "step": 27000
    },
    {
      "epoch": 19.0,
      "eval_loss": 4.415244102478027,
      "eval_runtime": 74.8709,
      "eval_samples_per_second": 6576.976,
      "eval_steps_per_second": 2.15,
      "step": 27436
    },
    {
      "epoch": 19.04,
      "learning_rate": 1.9619113573407206e-05,
      "loss": 4.5116,
      "step": 27500
    },
    {
      "epoch": 19.39,
      "learning_rate": 1.961218836565097e-05,
      "loss": 4.5031,
      "step": 28000
    },
    {
      "epoch": 19.74,
      "learning_rate": 1.960526315789474e-05,
      "loss": 4.501,
      "step": 28500
    },
    {
      "epoch": 20.0,
      "eval_loss": 4.401387691497803,
      "eval_runtime": 74.9583,
      "eval_samples_per_second": 6569.303,
      "eval_steps_per_second": 2.148,
      "step": 28880
    },
    {
      "epoch": 20.08,
      "learning_rate": 1.9598337950138507e-05,
      "loss": 4.4967,
      "step": 29000
    },
    {
      "epoch": 20.43,
      "learning_rate": 1.9591412742382274e-05,
      "loss": 4.4896,
      "step": 29500
    },
    {
      "epoch": 20.78,
      "learning_rate": 1.958448753462604e-05,
      "loss": 4.4853,
      "step": 30000
    },
    {
      "epoch": 21.0,
      "eval_loss": 4.387256622314453,
      "eval_runtime": 75.2035,
      "eval_samples_per_second": 6547.89,
      "eval_steps_per_second": 2.141,
      "step": 30324
    },
    {
      "epoch": 21.12,
      "learning_rate": 1.9577562326869807e-05,
      "loss": 4.4815,
      "step": 30500
    },
    {
      "epoch": 21.47,
      "learning_rate": 1.9570637119113574e-05,
      "loss": 4.4741,
      "step": 31000
    },
    {
      "epoch": 21.81,
      "learning_rate": 1.956371191135734e-05,
      "loss": 4.4705,
      "step": 31500
    },
    {
      "epoch": 22.0,
      "eval_loss": 4.372128963470459,
      "eval_runtime": 75.7644,
      "eval_samples_per_second": 6499.411,
      "eval_steps_per_second": 2.125,
      "step": 31768
    },
    {
      "epoch": 22.16,
      "learning_rate": 1.9556786703601108e-05,
      "loss": 4.465,
      "step": 32000
    },
    {
      "epoch": 22.51,
      "learning_rate": 1.954986149584488e-05,
      "loss": 4.4612,
      "step": 32500
    },
    {
      "epoch": 22.85,
      "learning_rate": 1.9542936288088645e-05,
      "loss": 4.4548,
      "step": 33000
    },
    {
      "epoch": 23.0,
      "eval_loss": 4.359036445617676,
      "eval_runtime": 75.0614,
      "eval_samples_per_second": 6560.284,
      "eval_steps_per_second": 2.145,
      "step": 33212
    },
    {
      "epoch": 23.2,
      "learning_rate": 1.9536011080332412e-05,
      "loss": 4.452,
      "step": 33500
    },
    {
      "epoch": 23.55,
      "learning_rate": 1.952908587257618e-05,
      "loss": 4.4467,
      "step": 34000
    },
    {
      "epoch": 23.89,
      "learning_rate": 1.9522160664819946e-05,
      "loss": 4.4435,
      "step": 34500
    },
    {
      "epoch": 24.0,
      "eval_loss": 4.351931571960449,
      "eval_runtime": 73.9885,
      "eval_samples_per_second": 6655.409,
      "eval_steps_per_second": 2.176,
      "step": 34656
    },
    {
      "epoch": 24.24,
      "learning_rate": 1.9515235457063713e-05,
      "loss": 4.4384,
      "step": 35000
    },
    {
      "epoch": 24.58,
      "learning_rate": 1.950831024930748e-05,
      "loss": 4.4346,
      "step": 35500
    },
    {
      "epoch": 24.93,
      "learning_rate": 1.9501385041551247e-05,
      "loss": 4.4276,
      "step": 36000
    },
    {
      "epoch": 25.0,
      "eval_loss": 4.33967924118042,
      "eval_runtime": 73.6152,
      "eval_samples_per_second": 6689.162,
      "eval_steps_per_second": 2.187,
      "step": 36100
    },
    {
      "epoch": 25.28,
      "learning_rate": 1.9494459833795017e-05,
      "loss": 4.4249,
      "step": 36500
    },
    {
      "epoch": 25.62,
      "learning_rate": 1.948753462603878e-05,
      "loss": 4.4245,
      "step": 37000
    },
    {
      "epoch": 25.97,
      "learning_rate": 1.948060941828255e-05,
      "loss": 4.4166,
      "step": 37500
    },
    {
      "epoch": 26.0,
      "eval_loss": 4.3328375816345215,
      "eval_runtime": 74.8225,
      "eval_samples_per_second": 6581.225,
      "eval_steps_per_second": 2.152,
      "step": 37544
    },
    {
      "epoch": 26.32,
      "learning_rate": 1.9473684210526318e-05,
      "loss": 4.4133,
      "step": 38000
    },
    {
      "epoch": 26.66,
      "learning_rate": 1.9466759002770085e-05,
      "loss": 4.41,
      "step": 38500
    },
    {
      "epoch": 27.0,
      "eval_loss": 4.31511926651001,
      "eval_runtime": 75.2713,
      "eval_samples_per_second": 6541.986,
      "eval_steps_per_second": 2.139,
      "step": 38988
    },
    {
      "epoch": 27.01,
      "learning_rate": 1.9459833795013852e-05,
      "loss": 4.4064,
      "step": 39000
    },
    {
      "epoch": 27.35,
      "learning_rate": 1.945290858725762e-05,
      "loss": 4.403,
      "step": 39500
    },
    {
      "epoch": 27.7,
      "learning_rate": 1.944598337950139e-05,
      "loss": 4.3996,
      "step": 40000
    },
    {
      "epoch": 28.0,
      "eval_loss": 4.304774761199951,
      "eval_runtime": 73.8005,
      "eval_samples_per_second": 6672.362,
      "eval_steps_per_second": 2.182,
      "step": 40432
    },
    {
      "epoch": 28.05,
      "learning_rate": 1.9439058171745153e-05,
      "loss": 4.3957,
      "step": 40500
    },
    {
      "epoch": 28.39,
      "learning_rate": 1.943213296398892e-05,
      "loss": 4.3921,
      "step": 41000
    },
    {
      "epoch": 28.74,
      "learning_rate": 1.942520775623269e-05,
      "loss": 4.3886,
      "step": 41500
    },
    {
      "epoch": 29.0,
      "eval_loss": 4.294938564300537,
      "eval_runtime": 76.8852,
      "eval_samples_per_second": 6404.662,
      "eval_steps_per_second": 2.094,
      "step": 41876
    },
    {
      "epoch": 29.09,
      "learning_rate": 1.9418282548476453e-05,
      "loss": 4.3851,
      "step": 42000
    },
    {
      "epoch": 29.43,
      "learning_rate": 1.9411357340720224e-05,
      "loss": 4.3808,
      "step": 42500
    },
    {
      "epoch": 29.78,
      "learning_rate": 1.940443213296399e-05,
      "loss": 4.3776,
      "step": 43000
    },
    {
      "epoch": 30.0,
      "eval_loss": 4.287764549255371,
      "eval_runtime": 74.0431,
      "eval_samples_per_second": 6650.502,
      "eval_steps_per_second": 2.174,
      "step": 43320
    },
    {
      "epoch": 30.12,
      "learning_rate": 1.9397506925207758e-05,
      "loss": 4.3721,
      "step": 43500
    },
    {
      "epoch": 30.47,
      "learning_rate": 1.9390581717451524e-05,
      "loss": 4.3705,
      "step": 44000
    },
    {
      "epoch": 30.82,
      "learning_rate": 1.938365650969529e-05,
      "loss": 4.3686,
      "step": 44500
    },
    {
      "epoch": 31.0,
      "eval_loss": 4.277154922485352,
      "eval_runtime": 74.74,
      "eval_samples_per_second": 6588.493,
      "eval_steps_per_second": 2.154,
      "step": 44764
    },
    {
      "epoch": 31.16,
      "learning_rate": 1.937673130193906e-05,
      "loss": 4.3631,
      "step": 45000
    },
    {
      "epoch": 31.51,
      "learning_rate": 1.936980609418283e-05,
      "loss": 4.363,
      "step": 45500
    },
    {
      "epoch": 31.86,
      "learning_rate": 1.9362880886426595e-05,
      "loss": 4.3572,
      "step": 46000
    },
    {
      "epoch": 32.0,
      "eval_loss": 4.265904903411865,
      "eval_runtime": 74.1213,
      "eval_samples_per_second": 6643.49,
      "eval_steps_per_second": 2.172,
      "step": 46208
    },
    {
      "epoch": 32.2,
      "learning_rate": 1.9355955678670362e-05,
      "loss": 4.3549,
      "step": 46500
    },
    {
      "epoch": 32.55,
      "learning_rate": 1.934903047091413e-05,
      "loss": 4.352,
      "step": 47000
    },
    {
      "epoch": 32.89,
      "learning_rate": 1.9342105263157896e-05,
      "loss": 4.3518,
      "step": 47500
    },
    {
      "epoch": 33.0,
      "eval_loss": 4.2521071434021,
      "eval_runtime": 73.7943,
      "eval_samples_per_second": 6672.929,
      "eval_steps_per_second": 2.182,
      "step": 47652
    },
    {
      "epoch": 33.24,
      "learning_rate": 1.9335180055401663e-05,
      "loss": 4.345,
      "step": 48000
    },
    {
      "epoch": 33.59,
      "learning_rate": 1.932825484764543e-05,
      "loss": 4.344,
      "step": 48500
    },
    {
      "epoch": 33.93,
      "learning_rate": 1.93213296398892e-05,
      "loss": 4.3377,
      "step": 49000
    },
    {
      "epoch": 34.0,
      "eval_loss": 4.250110626220703,
      "eval_runtime": 76.5664,
      "eval_samples_per_second": 6431.332,
      "eval_steps_per_second": 2.103,
      "step": 49096
    },
    {
      "epoch": 34.28,
      "learning_rate": 1.9314404432132964e-05,
      "loss": 4.333,
      "step": 49500
    },
    {
      "epoch": 34.63,
      "learning_rate": 1.9307479224376734e-05,
      "loss": 4.3307,
      "step": 50000
    },
    {
      "epoch": 34.97,
      "learning_rate": 1.93005540166205e-05,
      "loss": 4.3292,
      "step": 50500
    },
    {
      "epoch": 35.0,
      "eval_loss": 4.243873596191406,
      "eval_runtime": 75.8252,
      "eval_samples_per_second": 6494.199,
      "eval_steps_per_second": 2.123,
      "step": 50540
    },
    {
      "epoch": 35.32,
      "learning_rate": 1.9293628808864268e-05,
      "loss": 4.3287,
      "step": 51000
    },
    {
      "epoch": 35.66,
      "learning_rate": 1.9286703601108035e-05,
      "loss": 4.3236,
      "step": 51500
    },
    {
      "epoch": 36.0,
      "eval_loss": 4.227120399475098,
      "eval_runtime": 75.2969,
      "eval_samples_per_second": 6539.764,
      "eval_steps_per_second": 2.138,
      "step": 51984
    },
    {
      "epoch": 36.01,
      "learning_rate": 1.9279778393351802e-05,
      "loss": 4.3221,
      "step": 52000
    },
    {
      "epoch": 36.36,
      "learning_rate": 1.927285318559557e-05,
      "loss": 4.3191,
      "step": 52500
    },
    {
      "epoch": 36.7,
      "learning_rate": 1.9265927977839336e-05,
      "loss": 4.3147,
      "step": 53000
    },
    {
      "epoch": 37.0,
      "eval_loss": 4.228574752807617,
      "eval_runtime": 73.7741,
      "eval_samples_per_second": 6674.754,
      "eval_steps_per_second": 2.182,
      "step": 53428
    },
    {
      "epoch": 37.05,
      "learning_rate": 1.9259002770083103e-05,
      "loss": 4.3139,
      "step": 53500
    },
    {
      "epoch": 37.4,
      "learning_rate": 1.9252077562326873e-05,
      "loss": 4.3107,
      "step": 54000
    },
    {
      "epoch": 37.74,
      "learning_rate": 1.924515235457064e-05,
      "loss": 4.3054,
      "step": 54500
    },
    {
      "epoch": 38.0,
      "eval_loss": 4.218676567077637,
      "eval_runtime": 73.9641,
      "eval_samples_per_second": 6657.61,
      "eval_steps_per_second": 2.177,
      "step": 54872
    },
    {
      "epoch": 38.09,
      "learning_rate": 1.9238227146814407e-05,
      "loss": 4.303,
      "step": 55000
    },
    {
      "epoch": 38.43,
      "learning_rate": 1.9231301939058174e-05,
      "loss": 4.3031,
      "step": 55500
    },
    {
      "epoch": 38.78,
      "learning_rate": 1.922437673130194e-05,
      "loss": 4.3004,
      "step": 56000
    },
    {
      "epoch": 39.0,
      "eval_loss": 4.211672306060791,
      "eval_runtime": 73.1433,
      "eval_samples_per_second": 6732.321,
      "eval_steps_per_second": 2.201,
      "step": 56316
    },
    {
      "epoch": 39.13,
      "learning_rate": 1.9217451523545708e-05,
      "loss": 4.2952,
      "step": 56500
    },
    {
      "epoch": 39.47,
      "learning_rate": 1.9210526315789474e-05,
      "loss": 4.291,
      "step": 57000
    },
    {
      "epoch": 39.82,
      "learning_rate": 1.920360110803324e-05,
      "loss": 4.2922,
      "step": 57500
    },
    {
      "epoch": 40.0,
      "eval_loss": 4.2032246589660645,
      "eval_runtime": 74.4105,
      "eval_samples_per_second": 6617.665,
      "eval_steps_per_second": 2.164,
      "step": 57760
    },
    {
      "epoch": 40.17,
      "learning_rate": 1.9196675900277012e-05,
      "loss": 4.2891,
      "step": 58000
    },
    {
      "epoch": 40.51,
      "learning_rate": 1.9189750692520775e-05,
      "loss": 4.2848,
      "step": 58500
    },
    {
      "epoch": 40.86,
      "learning_rate": 1.9182825484764546e-05,
      "loss": 4.2841,
      "step": 59000
    },
    {
      "epoch": 41.0,
      "eval_loss": 4.192102432250977,
      "eval_runtime": 75.6355,
      "eval_samples_per_second": 6510.485,
      "eval_steps_per_second": 2.129,
      "step": 59204
    },
    {
      "epoch": 41.2,
      "learning_rate": 1.9175900277008312e-05,
      "loss": 4.2807,
      "step": 59500
    },
    {
      "epoch": 41.55,
      "learning_rate": 1.916897506925208e-05,
      "loss": 4.2781,
      "step": 60000
    },
    {
      "epoch": 41.9,
      "learning_rate": 1.9162049861495846e-05,
      "loss": 4.2765,
      "step": 60500
    },
    {
      "epoch": 42.0,
      "eval_loss": 4.186478614807129,
      "eval_runtime": 74.8014,
      "eval_samples_per_second": 6583.085,
      "eval_steps_per_second": 2.152,
      "step": 60648
    },
    {
      "epoch": 42.24,
      "learning_rate": 1.9155124653739613e-05,
      "loss": 4.2787,
      "step": 61000
    },
    {
      "epoch": 42.59,
      "learning_rate": 1.9148199445983384e-05,
      "loss": 4.2747,
      "step": 61500
    },
    {
      "epoch": 42.94,
      "learning_rate": 1.9141274238227147e-05,
      "loss": 4.2699,
      "step": 62000
    },
    {
      "epoch": 43.0,
      "eval_loss": 4.183748245239258,
      "eval_runtime": 74.0325,
      "eval_samples_per_second": 6651.456,
      "eval_steps_per_second": 2.175,
      "step": 62092
    },
    {
      "epoch": 43.28,
      "learning_rate": 1.9134349030470914e-05,
      "loss": 4.2688,
      "step": 62500
    },
    {
      "epoch": 43.63,
      "learning_rate": 1.9127423822714684e-05,
      "loss": 4.2662,
      "step": 63000
    },
    {
      "epoch": 43.98,
      "learning_rate": 1.9120498614958448e-05,
      "loss": 4.2634,
      "step": 63500
    },
    {
      "epoch": 44.0,
      "eval_loss": 4.1764092445373535,
      "eval_runtime": 74.4165,
      "eval_samples_per_second": 6617.138,
      "eval_steps_per_second": 2.163,
      "step": 63536
    },
    {
      "epoch": 44.32,
      "learning_rate": 1.9113573407202218e-05,
      "loss": 4.2594,
      "step": 64000
    },
    {
      "epoch": 44.67,
      "learning_rate": 1.9106648199445985e-05,
      "loss": 4.2598,
      "step": 64500
    },
    {
      "epoch": 45.0,
      "eval_loss": 4.171951770782471,
      "eval_runtime": 73.3834,
      "eval_samples_per_second": 6710.29,
      "eval_steps_per_second": 2.194,
      "step": 64980
    },
    {
      "epoch": 45.01,
      "learning_rate": 1.9099722991689752e-05,
      "loss": 4.256,
      "step": 65000
    },
    {
      "epoch": 45.36,
      "learning_rate": 1.909279778393352e-05,
      "loss": 4.2541,
      "step": 65500
    },
    {
      "epoch": 45.71,
      "learning_rate": 1.9085872576177286e-05,
      "loss": 4.2499,
      "step": 66000
    },
    {
      "epoch": 46.0,
      "eval_loss": 4.162701606750488,
      "eval_runtime": 73.3897,
      "eval_samples_per_second": 6709.716,
      "eval_steps_per_second": 2.194,
      "step": 66424
    },
    {
      "epoch": 46.05,
      "learning_rate": 1.9078947368421056e-05,
      "loss": 4.2489,
      "step": 66500
    },
    {
      "epoch": 46.4,
      "learning_rate": 1.9072022160664823e-05,
      "loss": 4.2488,
      "step": 67000
    },
    {
      "epoch": 46.75,
      "learning_rate": 1.906509695290859e-05,
      "loss": 4.2437,
      "step": 67500
    },
    {
      "epoch": 47.0,
      "eval_loss": 4.157140254974365,
      "eval_runtime": 73.7335,
      "eval_samples_per_second": 6678.427,
      "eval_steps_per_second": 2.184,
      "step": 67868
    },
    {
      "epoch": 47.09,
      "learning_rate": 1.9058171745152357e-05,
      "loss": 4.2438,
      "step": 68000
    },
    {
      "epoch": 47.44,
      "learning_rate": 1.9051246537396124e-05,
      "loss": 4.2436,
      "step": 68500
    },
    {
      "epoch": 47.78,
      "learning_rate": 1.904432132963989e-05,
      "loss": 4.24,
      "step": 69000
    },
    {
      "epoch": 48.0,
      "eval_loss": 4.155510902404785,
      "eval_runtime": 73.7665,
      "eval_samples_per_second": 6675.438,
      "eval_steps_per_second": 2.183,
      "step": 69312
    },
    {
      "epoch": 48.13,
      "learning_rate": 1.9037396121883658e-05,
      "loss": 4.2359,
      "step": 69500
    },
    {
      "epoch": 48.48,
      "learning_rate": 1.9030470914127425e-05,
      "loss": 4.2356,
      "step": 70000
    },
    {
      "epoch": 48.82,
      "learning_rate": 1.9023545706371195e-05,
      "loss": 4.2334,
      "step": 70500
    },
    {
      "epoch": 49.0,
      "eval_loss": 4.148888111114502,
      "eval_runtime": 76.009,
      "eval_samples_per_second": 6478.494,
      "eval_steps_per_second": 2.118,
      "step": 70756
    },
    {
      "epoch": 49.17,
      "learning_rate": 1.901662049861496e-05,
      "loss": 4.2319,
      "step": 71000
    },
    {
      "epoch": 49.52,
      "learning_rate": 1.900969529085873e-05,
      "loss": 4.2288,
      "step": 71500
    },
    {
      "epoch": 49.86,
      "learning_rate": 1.9002770083102496e-05,
      "loss": 4.2286,
      "step": 72000
    },
    {
      "epoch": 50.0,
      "eval_loss": 4.136005401611328,
      "eval_runtime": 74.0583,
      "eval_samples_per_second": 6649.141,
      "eval_steps_per_second": 2.174,
      "step": 72200
    },
    {
      "epoch": 50.21,
      "learning_rate": 1.8995844875346263e-05,
      "loss": 4.2234,
      "step": 72500
    },
    {
      "epoch": 50.55,
      "learning_rate": 1.898891966759003e-05,
      "loss": 4.2226,
      "step": 73000
    },
    {
      "epoch": 50.9,
      "learning_rate": 1.8981994459833796e-05,
      "loss": 4.2213,
      "step": 73500
    },
    {
      "epoch": 51.0,
      "eval_loss": 4.137174606323242,
      "eval_runtime": 74.5156,
      "eval_samples_per_second": 6608.338,
      "eval_steps_per_second": 2.161,
      "step": 73644
    },
    {
      "epoch": 51.25,
      "learning_rate": 1.8975069252077563e-05,
      "loss": 4.2186,
      "step": 74000
    },
    {
      "epoch": 51.59,
      "learning_rate": 1.896814404432133e-05,
      "loss": 4.2182,
      "step": 74500
    },
    {
      "epoch": 51.94,
      "learning_rate": 1.8961218836565097e-05,
      "loss": 4.2152,
      "step": 75000
    },
    {
      "epoch": 52.0,
      "eval_loss": 4.127338886260986,
      "eval_runtime": 76.201,
      "eval_samples_per_second": 6462.173,
      "eval_steps_per_second": 2.113,
      "step": 75088
    },
    {
      "epoch": 52.29,
      "learning_rate": 1.8954293628808867e-05,
      "loss": 4.2122,
      "step": 75500
    },
    {
      "epoch": 52.63,
      "learning_rate": 1.894736842105263e-05,
      "loss": 4.2135,
      "step": 76000
    },
    {
      "epoch": 52.98,
      "learning_rate": 1.89404432132964e-05,
      "loss": 4.211,
      "step": 76500
    },
    {
      "epoch": 53.0,
      "eval_loss": 4.125185489654541,
      "eval_runtime": 74.874,
      "eval_samples_per_second": 6576.7,
      "eval_steps_per_second": 2.15,
      "step": 76532
    },
    {
      "epoch": 53.32,
      "learning_rate": 1.8933518005540168e-05,
      "loss": 4.2079,
      "step": 77000
    },
    {
      "epoch": 53.67,
      "learning_rate": 1.8926592797783935e-05,
      "loss": 4.2036,
      "step": 77500
    },
    {
      "epoch": 54.0,
      "eval_loss": 4.116921901702881,
      "eval_runtime": 74.7043,
      "eval_samples_per_second": 6591.639,
      "eval_steps_per_second": 2.155,
      "step": 77976
    },
    {
      "epoch": 54.02,
      "learning_rate": 1.8919667590027702e-05,
      "loss": 4.2046,
      "step": 78000
    },
    {
      "epoch": 54.36,
      "learning_rate": 1.891274238227147e-05,
      "loss": 4.201,
      "step": 78500
    },
    {
      "epoch": 54.71,
      "learning_rate": 1.8905817174515236e-05,
      "loss": 4.1969,
      "step": 79000
    },
    {
      "epoch": 55.0,
      "eval_loss": 4.117580890655518,
      "eval_runtime": 74.4844,
      "eval_samples_per_second": 6611.1,
      "eval_steps_per_second": 2.162,
      "step": 79420
    },
    {
      "epoch": 55.06,
      "learning_rate": 1.8898891966759006e-05,
      "loss": 4.1987,
      "step": 79500
    },
    {
      "epoch": 55.4,
      "learning_rate": 1.889196675900277e-05,
      "loss": 4.1951,
      "step": 80000
    },
    {
      "epoch": 55.75,
      "learning_rate": 1.888504155124654e-05,
      "loss": 4.1944,
      "step": 80500
    },
    {
      "epoch": 56.0,
      "eval_loss": 4.110030651092529,
      "eval_runtime": 73.4183,
      "eval_samples_per_second": 6707.104,
      "eval_steps_per_second": 2.193,
      "step": 80864
    },
    {
      "epoch": 56.09,
      "learning_rate": 1.8878116343490307e-05,
      "loss": 4.1924,
      "step": 81000
    },
    {
      "epoch": 56.44,
      "learning_rate": 1.8871191135734074e-05,
      "loss": 4.1913,
      "step": 81500
    },
    {
      "epoch": 56.79,
      "learning_rate": 1.886426592797784e-05,
      "loss": 4.1923,
      "step": 82000
    },
    {
      "epoch": 57.0,
      "eval_loss": 4.108407974243164,
      "eval_runtime": 74.3146,
      "eval_samples_per_second": 6626.209,
      "eval_steps_per_second": 2.166,
      "step": 82308
    },
    {
      "epoch": 57.13,
      "learning_rate": 1.8857340720221608e-05,
      "loss": 4.1889,
      "step": 82500
    },
    {
      "epoch": 57.48,
      "learning_rate": 1.8850415512465378e-05,
      "loss": 4.1856,
      "step": 83000
    },
    {
      "epoch": 57.83,
      "learning_rate": 1.884349030470914e-05,
      "loss": 4.1869,
      "step": 83500
    },
    {
      "epoch": 58.0,
      "eval_loss": 4.103403568267822,
      "eval_runtime": 75.7605,
      "eval_samples_per_second": 6499.745,
      "eval_steps_per_second": 2.125,
      "step": 83752
    },
    {
      "epoch": 58.17,
      "learning_rate": 1.883656509695291e-05,
      "loss": 4.181,
      "step": 84000
    },
    {
      "epoch": 58.52,
      "learning_rate": 1.882963988919668e-05,
      "loss": 4.1801,
      "step": 84500
    },
    {
      "epoch": 58.86,
      "learning_rate": 1.8822714681440442e-05,
      "loss": 4.1802,
      "step": 85000
    },
    {
      "epoch": 59.0,
      "eval_loss": 4.101166248321533,
      "eval_runtime": 73.6678,
      "eval_samples_per_second": 6684.382,
      "eval_steps_per_second": 2.185,
      "step": 85196
    },
    {
      "epoch": 59.21,
      "learning_rate": 1.8815789473684213e-05,
      "loss": 4.1776,
      "step": 85500
    },
    {
      "epoch": 59.56,
      "learning_rate": 1.880886426592798e-05,
      "loss": 4.1765,
      "step": 86000
    },
    {
      "epoch": 59.9,
      "learning_rate": 1.8801939058171746e-05,
      "loss": 4.1764,
      "step": 86500
    },
    {
      "epoch": 60.0,
      "eval_loss": 4.089655876159668,
      "eval_runtime": 74.6724,
      "eval_samples_per_second": 6594.462,
      "eval_steps_per_second": 2.156,
      "step": 86640
    },
    {
      "epoch": 60.25,
      "learning_rate": 1.8795013850415513e-05,
      "loss": 4.1752,
      "step": 87000
    },
    {
      "epoch": 60.6,
      "learning_rate": 1.878808864265928e-05,
      "loss": 4.1747,
      "step": 87500
    },
    {
      "epoch": 60.94,
      "learning_rate": 1.878116343490305e-05,
      "loss": 4.1668,
      "step": 88000
    },
    {
      "epoch": 61.0,
      "eval_loss": 4.0857367515563965,
      "eval_runtime": 73.6262,
      "eval_samples_per_second": 6688.167,
      "eval_steps_per_second": 2.187,
      "step": 88084
    },
    {
      "epoch": 61.29,
      "learning_rate": 1.8774238227146814e-05,
      "loss": 4.1695,
      "step": 88500
    },
    {
      "epoch": 61.63,
      "learning_rate": 1.8767313019390584e-05,
      "loss": 4.1645,
      "step": 89000
    },
    {
      "epoch": 61.98,
      "learning_rate": 1.876038781163435e-05,
      "loss": 4.1633,
      "step": 89500
    },
    {
      "epoch": 62.0,
      "eval_loss": 4.080641269683838,
      "eval_runtime": 74.2817,
      "eval_samples_per_second": 6629.141,
      "eval_steps_per_second": 2.167,
      "step": 89528
    },
    {
      "epoch": 62.33,
      "learning_rate": 1.8753462603878118e-05,
      "loss": 4.1629,
      "step": 90000
    },
    {
      "epoch": 62.67,
      "learning_rate": 1.8746537396121885e-05,
      "loss": 4.1631,
      "step": 90500
    },
    {
      "epoch": 63.0,
      "eval_loss": 4.07755708694458,
      "eval_runtime": 73.9609,
      "eval_samples_per_second": 6657.9,
      "eval_steps_per_second": 2.177,
      "step": 90972
    },
    {
      "epoch": 63.02,
      "learning_rate": 1.8739612188365652e-05,
      "loss": 4.1607,
      "step": 91000
    },
    {
      "epoch": 63.37,
      "learning_rate": 1.873268698060942e-05,
      "loss": 4.1596,
      "step": 91500
    },
    {
      "epoch": 63.71,
      "learning_rate": 1.872576177285319e-05,
      "loss": 4.1559,
      "step": 92000
    },
    {
      "epoch": 64.0,
      "eval_loss": 4.075015068054199,
      "eval_runtime": 73.801,
      "eval_samples_per_second": 6672.317,
      "eval_steps_per_second": 2.182,
      "step": 92416
    },
    {
      "epoch": 64.06,
      "learning_rate": 1.8718836565096953e-05,
      "loss": 4.155,
      "step": 92500
    },
    {
      "epoch": 64.4,
      "learning_rate": 1.8711911357340723e-05,
      "loss": 4.1567,
      "step": 93000
    },
    {
      "epoch": 64.75,
      "learning_rate": 1.870498614958449e-05,
      "loss": 4.1529,
      "step": 93500
    },
    {
      "epoch": 65.0,
      "eval_loss": 4.0660247802734375,
      "eval_runtime": 73.35,
      "eval_samples_per_second": 6713.346,
      "eval_steps_per_second": 2.195,
      "step": 93860
    },
    {
      "epoch": 65.1,
      "learning_rate": 1.8698060941828257e-05,
      "loss": 4.1521,
      "step": 94000
    },
    {
      "epoch": 65.44,
      "learning_rate": 1.8691135734072024e-05,
      "loss": 4.1485,
      "step": 94500
    },
    {
      "epoch": 65.79,
      "learning_rate": 1.868421052631579e-05,
      "loss": 4.1485,
      "step": 95000
    },
    {
      "epoch": 66.0,
      "eval_loss": 4.064459800720215,
      "eval_runtime": 73.345,
      "eval_samples_per_second": 6713.802,
      "eval_steps_per_second": 2.195,
      "step": 95304
    },
    {
      "epoch": 66.14,
      "learning_rate": 1.8677285318559558e-05,
      "loss": 4.1469,
      "step": 95500
    },
    {
      "epoch": 66.48,
      "learning_rate": 1.8670360110803325e-05,
      "loss": 4.1459,
      "step": 96000
    },
    {
      "epoch": 66.83,
      "learning_rate": 1.866343490304709e-05,
      "loss": 4.1431,
      "step": 96500
    },
    {
      "epoch": 67.0,
      "eval_loss": 4.058371067047119,
      "eval_runtime": 72.3552,
      "eval_samples_per_second": 6805.646,
      "eval_steps_per_second": 2.225,
      "step": 96748
    },
    {
      "epoch": 67.17,
      "learning_rate": 1.8656509695290862e-05,
      "loss": 4.143,
      "step": 97000
    },
    {
      "epoch": 67.52,
      "learning_rate": 1.8649584487534625e-05,
      "loss": 4.1405,
      "step": 97500
    },
    {
      "epoch": 67.87,
      "learning_rate": 1.8642659279778396e-05,
      "loss": 4.1404,
      "step": 98000
    },
    {
      "epoch": 68.0,
      "eval_loss": 4.054144382476807,
      "eval_runtime": 73.8095,
      "eval_samples_per_second": 6671.555,
      "eval_steps_per_second": 2.181,
      "step": 98192
    },
    {
      "epoch": 68.21,
      "learning_rate": 1.8635734072022163e-05,
      "loss": 4.1381,
      "step": 98500
    },
    {
      "epoch": 68.56,
      "learning_rate": 1.862880886426593e-05,
      "loss": 4.1343,
      "step": 99000
    },
    {
      "epoch": 68.91,
      "learning_rate": 1.8621883656509697e-05,
      "loss": 4.1338,
      "step": 99500
    },
    {
      "epoch": 69.0,
      "eval_loss": 4.050791263580322,
      "eval_runtime": 75.7075,
      "eval_samples_per_second": 6504.296,
      "eval_steps_per_second": 2.127,
      "step": 99636
    },
    {
      "epoch": 69.25,
      "learning_rate": 1.8614958448753463e-05,
      "loss": 4.1313,
      "step": 100000
    },
    {
      "epoch": 69.6,
      "learning_rate": 1.860803324099723e-05,
      "loss": 4.132,
      "step": 100500
    },
    {
      "epoch": 69.94,
      "learning_rate": 1.8601108033240997e-05,
      "loss": 4.1352,
      "step": 101000
    },
    {
      "epoch": 70.0,
      "eval_loss": 4.043735980987549,
      "eval_runtime": 74.0516,
      "eval_samples_per_second": 6649.741,
      "eval_steps_per_second": 2.174,
      "step": 101080
    },
    {
      "epoch": 70.29,
      "learning_rate": 1.8594182825484764e-05,
      "loss": 4.1307,
      "step": 101500
    },
    {
      "epoch": 70.64,
      "learning_rate": 1.8587257617728535e-05,
      "loss": 4.1268,
      "step": 102000
    },
    {
      "epoch": 70.98,
      "learning_rate": 1.85803324099723e-05,
      "loss": 4.1307,
      "step": 102500
    },
    {
      "epoch": 71.0,
      "eval_loss": 4.04426908493042,
      "eval_runtime": 72.7883,
      "eval_samples_per_second": 6765.157,
      "eval_steps_per_second": 2.212,
      "step": 102524
    },
    {
      "epoch": 71.33,
      "learning_rate": 1.857340720221607e-05,
      "loss": 4.1277,
      "step": 103000
    },
    {
      "epoch": 71.68,
      "learning_rate": 1.8566481994459835e-05,
      "loss": 4.1241,
      "step": 103500
    },
    {
      "epoch": 72.0,
      "eval_loss": 4.041337966918945,
      "eval_runtime": 75.6211,
      "eval_samples_per_second": 6511.731,
      "eval_steps_per_second": 2.129,
      "step": 103968
    },
    {
      "epoch": 72.02,
      "learning_rate": 1.8559556786703602e-05,
      "loss": 4.1256,
      "step": 104000
    },
    {
      "epoch": 72.37,
      "learning_rate": 1.8552631578947373e-05,
      "loss": 4.1215,
      "step": 104500
    },
    {
      "epoch": 72.71,
      "learning_rate": 1.8545706371191136e-05,
      "loss": 4.1227,
      "step": 105000
    },
    {
      "epoch": 73.0,
      "eval_loss": 4.042394161224365,
      "eval_runtime": 75.0493,
      "eval_samples_per_second": 6561.341,
      "eval_steps_per_second": 2.145,
      "step": 105412
    },
    {
      "epoch": 73.06,
      "learning_rate": 1.8538781163434903e-05,
      "loss": 4.118,
      "step": 105500
    },
    {
      "epoch": 73.41,
      "learning_rate": 1.8531855955678673e-05,
      "loss": 4.1166,
      "step": 106000
    },
    {
      "epoch": 73.75,
      "learning_rate": 1.8524930747922437e-05,
      "loss": 4.1186,
      "step": 106500
    },
    {
      "epoch": 74.0,
      "eval_loss": 4.030467510223389,
      "eval_runtime": 73.3813,
      "eval_samples_per_second": 6710.483,
      "eval_steps_per_second": 2.194,
      "step": 106856
    },
    {
      "epoch": 74.1,
      "learning_rate": 1.8518005540166207e-05,
      "loss": 4.1165,
      "step": 107000
    },
    {
      "epoch": 74.45,
      "learning_rate": 1.8511080332409974e-05,
      "loss": 4.113,
      "step": 107500
    },
    {
      "epoch": 74.79,
      "learning_rate": 1.850415512465374e-05,
      "loss": 4.1118,
      "step": 108000
    },
    {
      "epoch": 75.0,
      "eval_loss": 4.033474922180176,
      "eval_runtime": 72.6909,
      "eval_samples_per_second": 6774.219,
      "eval_steps_per_second": 2.215,
      "step": 108300
    },
    {
      "epoch": 75.14,
      "learning_rate": 1.8497229916897508e-05,
      "loss": 4.1129,
      "step": 108500
    },
    {
      "epoch": 75.48,
      "learning_rate": 1.8490304709141275e-05,
      "loss": 4.1103,
      "step": 109000
    },
    {
      "epoch": 75.83,
      "learning_rate": 1.8483379501385045e-05,
      "loss": 4.1123,
      "step": 109500
    },
    {
      "epoch": 76.0,
      "eval_loss": 4.0271382331848145,
      "eval_runtime": 73.3585,
      "eval_samples_per_second": 6712.566,
      "eval_steps_per_second": 2.195,
      "step": 109744
    },
    {
      "epoch": 76.18,
      "learning_rate": 1.847645429362881e-05,
      "loss": 4.1101,
      "step": 110000
    },
    {
      "epoch": 76.52,
      "learning_rate": 1.846952908587258e-05,
      "loss": 4.1042,
      "step": 110500
    },
    {
      "epoch": 76.87,
      "learning_rate": 1.8462603878116346e-05,
      "loss": 4.1071,
      "step": 111000
    },
    {
      "epoch": 77.0,
      "eval_loss": 4.022892951965332,
      "eval_runtime": 74.3109,
      "eval_samples_per_second": 6626.539,
      "eval_steps_per_second": 2.167,
      "step": 111188
    },
    {
      "epoch": 77.22,
      "learning_rate": 1.8455678670360113e-05,
      "loss": 4.1012,
      "step": 111500
    },
    {
      "epoch": 77.56,
      "learning_rate": 1.844875346260388e-05,
      "loss": 4.101,
      "step": 112000
    },
    {
      "epoch": 77.91,
      "learning_rate": 1.8441828254847647e-05,
      "loss": 4.1028,
      "step": 112500
    },
    {
      "epoch": 78.0,
      "eval_loss": 4.022953987121582,
      "eval_runtime": 75.1521,
      "eval_samples_per_second": 6552.363,
      "eval_steps_per_second": 2.142,
      "step": 112632
    },
    {
      "epoch": 78.25,
      "learning_rate": 1.8434903047091414e-05,
      "loss": 4.1002,
      "step": 113000
    },
    {
      "epoch": 78.6,
      "learning_rate": 1.842797783933518e-05,
      "loss": 4.1001,
      "step": 113500
    },
    {
      "epoch": 78.95,
      "learning_rate": 1.8421052631578947e-05,
      "loss": 4.1002,
      "step": 114000
    },
    {
      "epoch": 79.0,
      "eval_loss": 4.016762733459473,
      "eval_runtime": 74.5194,
      "eval_samples_per_second": 6607.994,
      "eval_steps_per_second": 2.161,
      "step": 114076
    },
    {
      "epoch": 79.29,
      "learning_rate": 1.8414127423822718e-05,
      "loss": 4.0983,
      "step": 114500
    },
    {
      "epoch": 79.64,
      "learning_rate": 1.8407202216066485e-05,
      "loss": 4.0972,
      "step": 115000
    },
    {
      "epoch": 79.99,
      "learning_rate": 1.840027700831025e-05,
      "loss": 4.0958,
      "step": 115500
    },
    {
      "epoch": 80.0,
      "eval_loss": 4.014660835266113,
      "eval_runtime": 74.3599,
      "eval_samples_per_second": 6622.175,
      "eval_steps_per_second": 2.165,
      "step": 115520
    },
    {
      "epoch": 80.33,
      "learning_rate": 1.839335180055402e-05,
      "loss": 4.0944,
      "step": 116000
    },
    {
      "epoch": 80.68,
      "learning_rate": 1.8386426592797785e-05,
      "loss": 4.0919,
      "step": 116500
    },
    {
      "epoch": 81.0,
      "eval_loss": 4.004740238189697,
      "eval_runtime": 73.8401,
      "eval_samples_per_second": 6668.784,
      "eval_steps_per_second": 2.18,
      "step": 116964
    },
    {
      "epoch": 81.02,
      "learning_rate": 1.8379501385041552e-05,
      "loss": 4.091,
      "step": 117000
    },
    {
      "epoch": 81.37,
      "learning_rate": 1.837257617728532e-05,
      "loss": 4.0913,
      "step": 117500
    },
    {
      "epoch": 81.72,
      "learning_rate": 1.8365650969529086e-05,
      "loss": 4.0891,
      "step": 118000
    },
    {
      "epoch": 82.0,
      "eval_loss": 4.007106781005859,
      "eval_runtime": 73.8477,
      "eval_samples_per_second": 6668.102,
      "eval_steps_per_second": 2.18,
      "step": 118408
    },
    {
      "epoch": 82.06,
      "learning_rate": 1.8358725761772856e-05,
      "loss": 4.0869,
      "step": 118500
    },
    {
      "epoch": 82.41,
      "learning_rate": 1.835180055401662e-05,
      "loss": 4.0896,
      "step": 119000
    },
    {
      "epoch": 82.76,
      "learning_rate": 1.834487534626039e-05,
      "loss": 4.0902,
      "step": 119500
    },
    {
      "epoch": 83.0,
      "eval_loss": 4.002437591552734,
      "eval_runtime": 73.8311,
      "eval_samples_per_second": 6669.604,
      "eval_steps_per_second": 2.181,
      "step": 119852
    },
    {
      "epoch": 83.1,
      "learning_rate": 1.8337950138504157e-05,
      "loss": 4.0872,
      "step": 120000
    },
    {
      "epoch": 83.45,
      "learning_rate": 1.8331024930747924e-05,
      "loss": 4.084,
      "step": 120500
    },
    {
      "epoch": 83.8,
      "learning_rate": 1.832409972299169e-05,
      "loss": 4.0846,
      "step": 121000
    },
    {
      "epoch": 84.0,
      "eval_loss": 3.9989962577819824,
      "eval_runtime": 74.1606,
      "eval_samples_per_second": 6639.967,
      "eval_steps_per_second": 2.171,
      "step": 121296
    },
    {
      "epoch": 84.14,
      "learning_rate": 1.8317174515235458e-05,
      "loss": 4.0813,
      "step": 121500
    },
    {
      "epoch": 84.49,
      "learning_rate": 1.8310249307479225e-05,
      "loss": 4.0794,
      "step": 122000
    },
    {
      "epoch": 84.83,
      "learning_rate": 1.8303324099722992e-05,
      "loss": 4.0785,
      "step": 122500
    },
    {
      "epoch": 85.0,
      "eval_loss": 3.9976210594177246,
      "eval_runtime": 74.1372,
      "eval_samples_per_second": 6642.06,
      "eval_steps_per_second": 2.172,
      "step": 122740
    },
    {
      "epoch": 85.18,
      "learning_rate": 1.829639889196676e-05,
      "loss": 4.0776,
      "step": 123000
    },
    {
      "epoch": 85.53,
      "learning_rate": 1.828947368421053e-05,
      "loss": 4.079,
      "step": 123500
    },
    {
      "epoch": 85.87,
      "learning_rate": 1.8282548476454296e-05,
      "loss": 4.0788,
      "step": 124000
    },
    {
      "epoch": 86.0,
      "eval_loss": 3.9947969913482666,
      "eval_runtime": 74.7164,
      "eval_samples_per_second": 6590.579,
      "eval_steps_per_second": 2.155,
      "step": 124184
    },
    {
      "epoch": 86.22,
      "learning_rate": 1.8275623268698063e-05,
      "loss": 4.0777,
      "step": 124500
    },
    {
      "epoch": 86.57,
      "learning_rate": 1.826869806094183e-05,
      "loss": 4.0768,
      "step": 125000
    },
    {
      "epoch": 86.91,
      "learning_rate": 1.8261772853185597e-05,
      "loss": 4.0759,
      "step": 125500
    },
    {
      "epoch": 87.0,
      "eval_loss": 3.9896225929260254,
      "eval_runtime": 73.8648,
      "eval_samples_per_second": 6666.556,
      "eval_steps_per_second": 2.18,
      "step": 125628
    },
    {
      "epoch": 87.26,
      "learning_rate": 1.8254847645429364e-05,
      "loss": 4.0709,
      "step": 126000
    },
    {
      "epoch": 87.6,
      "learning_rate": 1.824792243767313e-05,
      "loss": 4.0711,
      "step": 126500
    },
    {
      "epoch": 87.95,
      "learning_rate": 1.82409972299169e-05,
      "loss": 4.0728,
      "step": 127000
    },
    {
      "epoch": 88.0,
      "eval_loss": 3.988676071166992,
      "eval_runtime": 73.3696,
      "eval_samples_per_second": 6711.549,
      "eval_steps_per_second": 2.194,
      "step": 127072
    },
    {
      "epoch": 88.3,
      "learning_rate": 1.8234072022160668e-05,
      "loss": 4.0693,
      "step": 127500
    },
    {
      "epoch": 88.64,
      "learning_rate": 1.822714681440443e-05,
      "loss": 4.0658,
      "step": 128000
    },
    {
      "epoch": 88.99,
      "learning_rate": 1.82202216066482e-05,
      "loss": 4.0668,
      "step": 128500
    },
    {
      "epoch": 89.0,
      "eval_loss": 3.9832513332366943,
      "eval_runtime": 73.9518,
      "eval_samples_per_second": 6658.718,
      "eval_steps_per_second": 2.177,
      "step": 128516
    },
    {
      "epoch": 89.34,
      "learning_rate": 1.821329639889197e-05,
      "loss": 4.0669,
      "step": 129000
    },
    {
      "epoch": 89.68,
      "learning_rate": 1.8206371191135735e-05,
      "loss": 4.0669,
      "step": 129500
    },
    {
      "epoch": 90.0,
      "eval_loss": 3.9873828887939453,
      "eval_runtime": 75.3146,
      "eval_samples_per_second": 6538.224,
      "eval_steps_per_second": 2.138,
      "step": 129960
    },
    {
      "epoch": 90.03,
      "learning_rate": 1.8199445983379502e-05,
      "loss": 4.0619,
      "step": 130000
    },
    {
      "epoch": 90.37,
      "learning_rate": 1.819252077562327e-05,
      "loss": 4.0611,
      "step": 130500
    },
    {
      "epoch": 90.72,
      "learning_rate": 1.818559556786704e-05,
      "loss": 4.0625,
      "step": 131000
    },
    {
      "epoch": 91.0,
      "eval_loss": 3.9783737659454346,
      "eval_runtime": 75.884,
      "eval_samples_per_second": 6489.166,
      "eval_steps_per_second": 2.122,
      "step": 131404
    },
    {
      "epoch": 91.07,
      "learning_rate": 1.8178670360110803e-05,
      "loss": 4.0609,
      "step": 131500
    },
    {
      "epoch": 91.41,
      "learning_rate": 1.8171745152354573e-05,
      "loss": 4.0605,
      "step": 132000
    },
    {
      "epoch": 91.76,
      "learning_rate": 1.816481994459834e-05,
      "loss": 4.0571,
      "step": 132500
    },
    {
      "epoch": 92.0,
      "eval_loss": 3.976148843765259,
      "eval_runtime": 74.7074,
      "eval_samples_per_second": 6591.369,
      "eval_steps_per_second": 2.155,
      "step": 132848
    },
    {
      "epoch": 92.11,
      "learning_rate": 1.8157894736842107e-05,
      "loss": 4.0565,
      "step": 133000
    },
    {
      "epoch": 92.45,
      "learning_rate": 1.8150969529085874e-05,
      "loss": 4.057,
      "step": 133500
    },
    {
      "epoch": 92.8,
      "learning_rate": 1.814404432132964e-05,
      "loss": 4.0572,
      "step": 134000
    },
    {
      "epoch": 93.0,
      "eval_loss": 3.9746015071868896,
      "eval_runtime": 73.6659,
      "eval_samples_per_second": 6684.561,
      "eval_steps_per_second": 2.186,
      "step": 134292
    },
    {
      "epoch": 93.14,
      "learning_rate": 1.8137119113573408e-05,
      "loss": 4.0551,
      "step": 134500
    },
    {
      "epoch": 93.49,
      "learning_rate": 1.8130193905817175e-05,
      "loss": 4.0562,
      "step": 135000
    },
    {
      "epoch": 93.84,
      "learning_rate": 1.8123268698060942e-05,
      "loss": 4.0533,
      "step": 135500
    },
    {
      "epoch": 94.0,
      "eval_loss": 3.971306562423706,
      "eval_runtime": 74.5978,
      "eval_samples_per_second": 6601.056,
      "eval_steps_per_second": 2.158,
      "step": 135736
    },
    {
      "epoch": 94.18,
      "learning_rate": 1.8116343490304712e-05,
      "loss": 4.0537,
      "step": 136000
    },
    {
      "epoch": 94.53,
      "learning_rate": 1.810941828254848e-05,
      "loss": 4.0532,
      "step": 136500
    },
    {
      "epoch": 94.88,
      "learning_rate": 1.8102493074792246e-05,
      "loss": 4.0508,
      "step": 137000
    },
    {
      "epoch": 95.0,
      "eval_loss": 3.971055507659912,
      "eval_runtime": 75.83,
      "eval_samples_per_second": 6493.789,
      "eval_steps_per_second": 2.123,
      "step": 137180
    },
    {
      "epoch": 95.22,
      "learning_rate": 1.8095567867036013e-05,
      "loss": 4.0504,
      "step": 137500
    },
    {
      "epoch": 95.57,
      "learning_rate": 1.808864265927978e-05,
      "loss": 4.0466,
      "step": 138000
    },
    {
      "epoch": 95.91,
      "learning_rate": 1.8081717451523547e-05,
      "loss": 4.0487,
      "step": 138500
    },
    {
      "epoch": 96.0,
      "eval_loss": 3.964989185333252,
      "eval_runtime": 75.3545,
      "eval_samples_per_second": 6534.761,
      "eval_steps_per_second": 2.137,
      "step": 138624
    },
    {
      "epoch": 96.26,
      "learning_rate": 1.8074792243767314e-05,
      "loss": 4.0443,
      "step": 139000
    },
    {
      "epoch": 96.61,
      "learning_rate": 1.806786703601108e-05,
      "loss": 4.044,
      "step": 139500
    },
    {
      "epoch": 96.95,
      "learning_rate": 1.806094182825485e-05,
      "loss": 4.0446,
      "step": 140000
    },
    {
      "epoch": 97.0,
      "eval_loss": 3.9668915271759033,
      "eval_runtime": 74.5714,
      "eval_samples_per_second": 6603.39,
      "eval_steps_per_second": 2.159,
      "step": 140068
    },
    {
      "epoch": 97.3,
      "learning_rate": 1.8054016620498614e-05,
      "loss": 4.046,
      "step": 140500
    },
    {
      "epoch": 97.65,
      "learning_rate": 1.8047091412742385e-05,
      "loss": 4.0446,
      "step": 141000
    },
    {
      "epoch": 97.99,
      "learning_rate": 1.804016620498615e-05,
      "loss": 4.0432,
      "step": 141500
    },
    {
      "epoch": 98.0,
      "eval_loss": 3.958078145980835,
      "eval_runtime": 76.6585,
      "eval_samples_per_second": 6423.607,
      "eval_steps_per_second": 2.1,
      "step": 141512
    },
    {
      "epoch": 98.34,
      "learning_rate": 1.803324099722992e-05,
      "loss": 4.043,
      "step": 142000
    },
    {
      "epoch": 98.68,
      "learning_rate": 1.8026315789473685e-05,
      "loss": 4.0413,
      "step": 142500
    },
    {
      "epoch": 99.0,
      "eval_loss": 3.9584155082702637,
      "eval_runtime": 74.9791,
      "eval_samples_per_second": 6567.486,
      "eval_steps_per_second": 2.147,
      "step": 142956
    },
    {
      "epoch": 99.03,
      "learning_rate": 1.8019390581717452e-05,
      "loss": 4.0386,
      "step": 143000
    },
    {
      "epoch": 99.38,
      "learning_rate": 1.801246537396122e-05,
      "loss": 4.0369,
      "step": 143500
    },
    {
      "epoch": 99.72,
      "learning_rate": 1.8005540166204986e-05,
      "loss": 4.0354,
      "step": 144000
    },
    {
      "epoch": 100.0,
      "eval_loss": 3.9502811431884766,
      "eval_runtime": 75.7506,
      "eval_samples_per_second": 6500.598,
      "eval_steps_per_second": 2.125,
      "step": 144400
    },
    {
      "epoch": 100.07,
      "learning_rate": 1.7998614958448753e-05,
      "loss": 4.0367,
      "step": 144500
    },
    {
      "epoch": 100.42,
      "learning_rate": 1.7991689750692523e-05,
      "loss": 4.0353,
      "step": 145000
    },
    {
      "epoch": 100.76,
      "learning_rate": 1.798476454293629e-05,
      "loss": 4.0357,
      "step": 145500
    },
    {
      "epoch": 101.0,
      "eval_loss": 3.948287010192871,
      "eval_runtime": 75.3101,
      "eval_samples_per_second": 6538.622,
      "eval_steps_per_second": 2.138,
      "step": 145844
    },
    {
      "epoch": 101.11,
      "learning_rate": 1.7977839335180057e-05,
      "loss": 4.0344,
      "step": 146000
    },
    {
      "epoch": 101.45,
      "learning_rate": 1.7970914127423824e-05,
      "loss": 4.0308,
      "step": 146500
    },
    {
      "epoch": 101.8,
      "learning_rate": 1.796398891966759e-05,
      "loss": 4.0321,
      "step": 147000
    },
    {
      "epoch": 102.0,
      "eval_loss": 3.9529218673706055,
      "eval_runtime": 74.5513,
      "eval_samples_per_second": 6605.174,
      "eval_steps_per_second": 2.16,
      "step": 147288
    },
    {
      "epoch": 102.15,
      "learning_rate": 1.7957063711911358e-05,
      "loss": 4.0305,
      "step": 147500
    },
    {
      "epoch": 102.49,
      "learning_rate": 1.7950138504155125e-05,
      "loss": 4.0291,
      "step": 148000
    },
    {
      "epoch": 102.84,
      "learning_rate": 1.7943213296398895e-05,
      "loss": 4.0292,
      "step": 148500
    },
    {
      "epoch": 103.0,
      "eval_loss": 3.9466750621795654,
      "eval_runtime": 76.3415,
      "eval_samples_per_second": 6450.282,
      "eval_steps_per_second": 2.109,
      "step": 148732
    },
    {
      "epoch": 103.19,
      "learning_rate": 1.7936288088642662e-05,
      "loss": 4.0281,
      "step": 149000
    },
    {
      "epoch": 103.53,
      "learning_rate": 1.7929362880886426e-05,
      "loss": 4.0281,
      "step": 149500
    },
    {
      "epoch": 103.88,
      "learning_rate": 1.7922437673130196e-05,
      "loss": 4.0271,
      "step": 150000
    },
    {
      "epoch": 104.0,
      "eval_loss": 3.9457340240478516,
      "eval_runtime": 77.413,
      "eval_samples_per_second": 6361.001,
      "eval_steps_per_second": 2.08,
      "step": 150176
    },
    {
      "epoch": 104.22,
      "learning_rate": 1.7915512465373963e-05,
      "loss": 4.0259,
      "step": 150500
    },
    {
      "epoch": 104.57,
      "learning_rate": 1.790858725761773e-05,
      "loss": 4.0247,
      "step": 151000
    },
    {
      "epoch": 104.92,
      "learning_rate": 1.7901662049861497e-05,
      "loss": 4.0245,
      "step": 151500
    },
    {
      "epoch": 105.0,
      "eval_loss": 3.944822311401367,
      "eval_runtime": 79.2928,
      "eval_samples_per_second": 6210.2,
      "eval_steps_per_second": 2.03,
      "step": 151620
    },
    {
      "epoch": 105.26,
      "learning_rate": 1.7894736842105264e-05,
      "loss": 4.0233,
      "step": 152000
    },
    {
      "epoch": 105.61,
      "learning_rate": 1.7887811634349034e-05,
      "loss": 4.0232,
      "step": 152500
    },
    {
      "epoch": 105.96,
      "learning_rate": 1.7880886426592798e-05,
      "loss": 4.0204,
      "step": 153000
    },
    {
      "epoch": 106.0,
      "eval_loss": 3.9429657459259033,
      "eval_runtime": 76.6917,
      "eval_samples_per_second": 6420.824,
      "eval_steps_per_second": 2.099,
      "step": 153064
    },
    {
      "epoch": 106.3,
      "learning_rate": 1.7873961218836568e-05,
      "loss": 4.0201,
      "step": 153500
    },
    {
      "epoch": 106.65,
      "learning_rate": 1.7867036011080335e-05,
      "loss": 4.0191,
      "step": 154000
    },
    {
      "epoch": 106.99,
      "learning_rate": 1.7860110803324102e-05,
      "loss": 4.0218,
      "step": 154500
    },
    {
      "epoch": 107.0,
      "eval_loss": 3.9408297538757324,
      "eval_runtime": 75.1357,
      "eval_samples_per_second": 6553.796,
      "eval_steps_per_second": 2.143,
      "step": 154508
    },
    {
      "epoch": 107.34,
      "learning_rate": 1.785318559556787e-05,
      "loss": 4.0232,
      "step": 155000
    },
    {
      "epoch": 107.69,
      "learning_rate": 1.7846260387811636e-05,
      "loss": 4.018,
      "step": 155500
    },
    {
      "epoch": 108.0,
      "eval_loss": 3.9399232864379883,
      "eval_runtime": 74.43,
      "eval_samples_per_second": 6615.938,
      "eval_steps_per_second": 2.163,
      "step": 155952
    },
    {
      "epoch": 108.03,
      "learning_rate": 1.7839335180055402e-05,
      "loss": 4.0175,
      "step": 156000
    },
    {
      "epoch": 108.38,
      "learning_rate": 1.783240997229917e-05,
      "loss": 4.0165,
      "step": 156500
    },
    {
      "epoch": 108.73,
      "learning_rate": 1.7825484764542936e-05,
      "loss": 4.0137,
      "step": 157000
    },
    {
      "epoch": 109.0,
      "eval_loss": 3.935981035232544,
      "eval_runtime": 75.1558,
      "eval_samples_per_second": 6552.043,
      "eval_steps_per_second": 2.142,
      "step": 157396
    },
    {
      "epoch": 109.07,
      "learning_rate": 1.7818559556786707e-05,
      "loss": 4.0159,
      "step": 157500
    },
    {
      "epoch": 109.42,
      "learning_rate": 1.7811634349030474e-05,
      "loss": 4.0177,
      "step": 158000
    },
    {
      "epoch": 109.76,
      "learning_rate": 1.780470914127424e-05,
      "loss": 4.0114,
      "step": 158500
    },
    {
      "epoch": 110.0,
      "eval_loss": 3.934030771255493,
      "eval_runtime": 75.3274,
      "eval_samples_per_second": 6537.118,
      "eval_steps_per_second": 2.137,
      "step": 158840
    },
    {
      "epoch": 110.11,
      "learning_rate": 1.7797783933518007e-05,
      "loss": 4.011,
      "step": 159000
    },
    {
      "epoch": 110.46,
      "learning_rate": 1.7790858725761774e-05,
      "loss": 4.0116,
      "step": 159500
    },
    {
      "epoch": 110.8,
      "learning_rate": 1.778393351800554e-05,
      "loss": 4.0122,
      "step": 160000
    },
    {
      "epoch": 111.0,
      "eval_loss": 3.9282586574554443,
      "eval_runtime": 75.7551,
      "eval_samples_per_second": 6500.211,
      "eval_steps_per_second": 2.125,
      "step": 160284
    },
    {
      "epoch": 111.15,
      "learning_rate": 1.7777008310249308e-05,
      "loss": 4.0088,
      "step": 160500
    },
    {
      "epoch": 111.5,
      "learning_rate": 1.7770083102493075e-05,
      "loss": 4.0097,
      "step": 161000
    },
    {
      "epoch": 111.84,
      "learning_rate": 1.7763157894736845e-05,
      "loss": 4.0093,
      "step": 161500
    },
    {
      "epoch": 112.0,
      "eval_loss": 3.9263861179351807,
      "eval_runtime": 77.7083,
      "eval_samples_per_second": 6336.824,
      "eval_steps_per_second": 2.072,
      "step": 161728
    },
    {
      "epoch": 112.19,
      "learning_rate": 1.775623268698061e-05,
      "loss": 4.0096,
      "step": 162000
    },
    {
      "epoch": 112.53,
      "learning_rate": 1.774930747922438e-05,
      "loss": 4.0042,
      "step": 162500
    },
    {
      "epoch": 112.88,
      "learning_rate": 1.7742382271468146e-05,
      "loss": 4.0068,
      "step": 163000
    },
    {
      "epoch": 113.0,
      "eval_loss": 3.9213218688964844,
      "eval_runtime": 76.9094,
      "eval_samples_per_second": 6402.654,
      "eval_steps_per_second": 2.093,
      "step": 163172
    },
    {
      "epoch": 113.23,
      "learning_rate": 1.7735457063711913e-05,
      "loss": 4.005,
      "step": 163500
    },
    {
      "epoch": 113.57,
      "learning_rate": 1.772853185595568e-05,
      "loss": 4.0043,
      "step": 164000
    },
    {
      "epoch": 113.92,
      "learning_rate": 1.7721606648199447e-05,
      "loss": 4.002,
      "step": 164500
    },
    {
      "epoch": 114.0,
      "eval_loss": 3.9210433959960938,
      "eval_runtime": 73.884,
      "eval_samples_per_second": 6664.824,
      "eval_steps_per_second": 2.179,
      "step": 164616
    },
    {
      "epoch": 114.27,
      "learning_rate": 1.7714681440443214e-05,
      "loss": 4.0039,
      "step": 165000
    },
    {
      "epoch": 114.61,
      "learning_rate": 1.770775623268698e-05,
      "loss": 4.0012,
      "step": 165500
    },
    {
      "epoch": 114.96,
      "learning_rate": 1.7700831024930748e-05,
      "loss": 4.0013,
      "step": 166000
    },
    {
      "epoch": 115.0,
      "eval_loss": 3.9188640117645264,
      "eval_runtime": 74.0122,
      "eval_samples_per_second": 6653.285,
      "eval_steps_per_second": 2.175,
      "step": 166060
    },
    {
      "epoch": 115.3,
      "learning_rate": 1.7693905817174518e-05,
      "loss": 4.0011,
      "step": 166500
    },
    {
      "epoch": 115.65,
      "learning_rate": 1.7686980609418285e-05,
      "loss": 4.0031,
      "step": 167000
    },
    {
      "epoch": 116.0,
      "learning_rate": 1.7680055401662052e-05,
      "loss": 3.9978,
      "step": 167500
    },
    {
      "epoch": 116.0,
      "eval_loss": 3.920518398284912,
      "eval_runtime": 74.6861,
      "eval_samples_per_second": 6593.245,
      "eval_steps_per_second": 2.156,
      "step": 167504
    },
    {
      "epoch": 116.34,
      "learning_rate": 1.767313019390582e-05,
      "loss": 3.9936,
      "step": 168000
    },
    {
      "epoch": 116.69,
      "learning_rate": 1.7666204986149586e-05,
      "loss": 3.9978,
      "step": 168500
    },
    {
      "epoch": 117.0,
      "eval_loss": 3.918830394744873,
      "eval_runtime": 74.6329,
      "eval_samples_per_second": 6597.945,
      "eval_steps_per_second": 2.157,
      "step": 168948
    },
    {
      "epoch": 117.04,
      "learning_rate": 1.7659279778393353e-05,
      "loss": 3.9979,
      "step": 169000
    },
    {
      "epoch": 117.38,
      "learning_rate": 1.765235457063712e-05,
      "loss": 3.9972,
      "step": 169500
    },
    {
      "epoch": 117.73,
      "learning_rate": 1.764542936288089e-05,
      "loss": 3.9966,
      "step": 170000
    },
    {
      "epoch": 118.0,
      "eval_loss": 3.914811611175537,
      "eval_runtime": 74.0729,
      "eval_samples_per_second": 6647.825,
      "eval_steps_per_second": 2.174,
      "step": 170392
    },
    {
      "epoch": 118.07,
      "learning_rate": 1.7638504155124657e-05,
      "loss": 3.9952,
      "step": 170500
    },
    {
      "epoch": 118.42,
      "learning_rate": 1.763157894736842e-05,
      "loss": 3.9942,
      "step": 171000
    },
    {
      "epoch": 118.77,
      "learning_rate": 1.762465373961219e-05,
      "loss": 3.9923,
      "step": 171500
    },
    {
      "epoch": 119.0,
      "eval_loss": 3.915539503097534,
      "eval_runtime": 77.3659,
      "eval_samples_per_second": 6364.87,
      "eval_steps_per_second": 2.081,
      "step": 171836
    },
    {
      "epoch": 119.11,
      "learning_rate": 1.7617728531855957e-05,
      "loss": 3.9941,
      "step": 172000
    },
    {
      "epoch": 119.46,
      "learning_rate": 1.7610803324099724e-05,
      "loss": 3.9925,
      "step": 172500
    },
    {
      "epoch": 119.81,
      "learning_rate": 1.760387811634349e-05,
      "loss": 3.9901,
      "step": 173000
    },
    {
      "epoch": 120.0,
      "eval_loss": 3.9078481197357178,
      "eval_runtime": 73.0002,
      "eval_samples_per_second": 6745.514,
      "eval_steps_per_second": 2.205,
      "step": 173280
    },
    {
      "epoch": 120.15,
      "learning_rate": 1.7596952908587258e-05,
      "loss": 3.9908,
      "step": 173500
    },
    {
      "epoch": 120.5,
      "learning_rate": 1.759002770083103e-05,
      "loss": 3.9875,
      "step": 174000
    },
    {
      "epoch": 120.84,
      "learning_rate": 1.7583102493074792e-05,
      "loss": 3.989,
      "step": 174500
    },
    {
      "epoch": 121.0,
      "eval_loss": 3.9121310710906982,
      "eval_runtime": 73.6203,
      "eval_samples_per_second": 6688.695,
      "eval_steps_per_second": 2.187,
      "step": 174724
    },
    {
      "epoch": 121.19,
      "learning_rate": 1.7576177285318562e-05,
      "loss": 3.987,
      "step": 175000
    },
    {
      "epoch": 121.54,
      "learning_rate": 1.756925207756233e-05,
      "loss": 3.991,
      "step": 175500
    },
    {
      "epoch": 121.88,
      "learning_rate": 1.7562326869806096e-05,
      "loss": 3.9876,
      "step": 176000
    },
    {
      "epoch": 122.0,
      "eval_loss": 3.9086544513702393,
      "eval_runtime": 74.7783,
      "eval_samples_per_second": 6585.118,
      "eval_steps_per_second": 2.153,
      "step": 176168
    },
    {
      "epoch": 122.23,
      "learning_rate": 1.7555401662049863e-05,
      "loss": 3.9863,
      "step": 176500
    },
    {
      "epoch": 122.58,
      "learning_rate": 1.754847645429363e-05,
      "loss": 3.9866,
      "step": 177000
    },
    {
      "epoch": 122.92,
      "learning_rate": 1.7541551246537397e-05,
      "loss": 3.9856,
      "step": 177500
    },
    {
      "epoch": 123.0,
      "eval_loss": 3.9068329334259033,
      "eval_runtime": 74.5275,
      "eval_samples_per_second": 6607.276,
      "eval_steps_per_second": 2.16,
      "step": 177612
    },
    {
      "epoch": 123.27,
      "learning_rate": 1.7534626038781164e-05,
      "loss": 3.9857,
      "step": 178000
    },
    {
      "epoch": 123.61,
      "learning_rate": 1.752770083102493e-05,
      "loss": 3.9853,
      "step": 178500
    },
    {
      "epoch": 123.96,
      "learning_rate": 1.75207756232687e-05,
      "loss": 3.9834,
      "step": 179000
    },
    {
      "epoch": 124.0,
      "eval_loss": 3.90472412109375,
      "eval_runtime": 74.8567,
      "eval_samples_per_second": 6578.224,
      "eval_steps_per_second": 2.151,
      "step": 179056
    },
    {
      "epoch": 124.31,
      "learning_rate": 1.7513850415512468e-05,
      "loss": 3.9833,
      "step": 179500
    },
    {
      "epoch": 124.65,
      "learning_rate": 1.7506925207756235e-05,
      "loss": 3.9812,
      "step": 180000
    },
    {
      "epoch": 125.0,
      "learning_rate": 1.7500000000000002e-05,
      "loss": 3.9786,
      "step": 180500
    },
    {
      "epoch": 125.0,
      "eval_loss": 3.89973783493042,
      "eval_runtime": 73.8136,
      "eval_samples_per_second": 6671.181,
      "eval_steps_per_second": 2.181,
      "step": 180500
    },
    {
      "epoch": 125.35,
      "learning_rate": 1.749307479224377e-05,
      "loss": 3.9793,
      "step": 181000
    },
    {
      "epoch": 125.69,
      "learning_rate": 1.7486149584487536e-05,
      "loss": 3.9788,
      "step": 181500
    },
    {
      "epoch": 126.0,
      "eval_loss": 3.900991201400757,
      "eval_runtime": 76.4593,
      "eval_samples_per_second": 6440.341,
      "eval_steps_per_second": 2.106,
      "step": 181944
    },
    {
      "epoch": 126.04,
      "learning_rate": 1.7479224376731303e-05,
      "loss": 3.98,
      "step": 182000
    },
    {
      "epoch": 126.39,
      "learning_rate": 1.747229916897507e-05,
      "loss": 3.979,
      "step": 182500
    },
    {
      "epoch": 126.73,
      "learning_rate": 1.746537396121884e-05,
      "loss": 3.9782,
      "step": 183000
    },
    {
      "epoch": 127.0,
      "eval_loss": 3.898531436920166,
      "eval_runtime": 73.137,
      "eval_samples_per_second": 6732.899,
      "eval_steps_per_second": 2.201,
      "step": 183388
    },
    {
      "epoch": 127.08,
      "learning_rate": 1.7458448753462603e-05,
      "loss": 3.9772,
      "step": 183500
    },
    {
      "epoch": 127.42,
      "learning_rate": 1.7451523545706374e-05,
      "loss": 3.9759,
      "step": 184000
    },
    {
      "epoch": 127.77,
      "learning_rate": 1.744459833795014e-05,
      "loss": 3.9733,
      "step": 184500
    },
    {
      "epoch": 128.0,
      "eval_loss": 3.892979145050049,
      "eval_runtime": 74.6554,
      "eval_samples_per_second": 6595.957,
      "eval_steps_per_second": 2.157,
      "step": 184832
    },
    {
      "epoch": 128.12,
      "learning_rate": 1.7437673130193908e-05,
      "loss": 3.9761,
      "step": 185000
    },
    {
      "epoch": 128.46,
      "learning_rate": 1.7430747922437674e-05,
      "loss": 3.9733,
      "step": 185500
    },
    {
      "epoch": 128.81,
      "learning_rate": 1.742382271468144e-05,
      "loss": 3.9765,
      "step": 186000
    },
    {
      "epoch": 129.0,
      "eval_loss": 3.8959548473358154,
      "eval_runtime": 72.9991,
      "eval_samples_per_second": 6745.614,
      "eval_steps_per_second": 2.206,
      "step": 186276
    },
    {
      "epoch": 129.16,
      "learning_rate": 1.741689750692521e-05,
      "loss": 3.9713,
      "step": 186500
    },
    {
      "epoch": 129.5,
      "learning_rate": 1.7409972299168975e-05,
      "loss": 3.9714,
      "step": 187000
    },
    {
      "epoch": 129.85,
      "learning_rate": 1.7403047091412742e-05,
      "loss": 3.9733,
      "step": 187500
    },
    {
      "epoch": 130.0,
      "eval_loss": 3.8933827877044678,
      "eval_runtime": 75.3723,
      "eval_samples_per_second": 6533.221,
      "eval_steps_per_second": 2.136,
      "step": 187720
    },
    {
      "epoch": 130.19,
      "learning_rate": 1.7396121883656512e-05,
      "loss": 3.9725,
      "step": 188000
    },
    {
      "epoch": 130.54,
      "learning_rate": 1.738919667590028e-05,
      "loss": 3.9698,
      "step": 188500
    },
    {
      "epoch": 130.89,
      "learning_rate": 1.7382271468144046e-05,
      "loss": 3.9695,
      "step": 189000
    },
    {
      "epoch": 131.0,
      "eval_loss": 3.888564348220825,
      "eval_runtime": 76.8687,
      "eval_samples_per_second": 6406.041,
      "eval_steps_per_second": 2.094,
      "step": 189164
    },
    {
      "epoch": 131.23,
      "learning_rate": 1.7375346260387813e-05,
      "loss": 3.9677,
      "step": 189500
    },
    {
      "epoch": 131.58,
      "learning_rate": 1.736842105263158e-05,
      "loss": 3.9709,
      "step": 190000
    },
    {
      "epoch": 131.93,
      "learning_rate": 1.7361495844875347e-05,
      "loss": 3.9708,
      "step": 190500
    },
    {
      "epoch": 132.0,
      "eval_loss": 3.889085054397583,
      "eval_runtime": 73.3155,
      "eval_samples_per_second": 6716.506,
      "eval_steps_per_second": 2.196,
      "step": 190608
    },
    {
      "epoch": 132.27,
      "learning_rate": 1.7354570637119114e-05,
      "loss": 3.9635,
      "step": 191000
    },
    {
      "epoch": 132.62,
      "learning_rate": 1.7347645429362884e-05,
      "loss": 3.9658,
      "step": 191500
    },
    {
      "epoch": 132.96,
      "learning_rate": 1.734072022160665e-05,
      "loss": 3.9673,
      "step": 192000
    },
    {
      "epoch": 133.0,
      "eval_loss": 3.885761022567749,
      "eval_runtime": 72.7481,
      "eval_samples_per_second": 6768.893,
      "eval_steps_per_second": 2.213,
      "step": 192052
    },
    {
      "epoch": 133.31,
      "learning_rate": 1.7333795013850415e-05,
      "loss": 3.9664,
      "step": 192500
    },
    {
      "epoch": 133.66,
      "learning_rate": 1.7326869806094185e-05,
      "loss": 3.9635,
      "step": 193000
    },
    {
      "epoch": 134.0,
      "eval_loss": 3.8835644721984863,
      "eval_runtime": 72.9182,
      "eval_samples_per_second": 6753.106,
      "eval_steps_per_second": 2.208,
      "step": 193496
    },
    {
      "epoch": 134.0,
      "learning_rate": 1.7319944598337952e-05,
      "loss": 3.9667,
      "step": 193500
    },
    {
      "epoch": 134.35,
      "learning_rate": 1.731301939058172e-05,
      "loss": 3.9627,
      "step": 194000
    },
    {
      "epoch": 134.7,
      "learning_rate": 1.7306094182825486e-05,
      "loss": 3.962,
      "step": 194500
    },
    {
      "epoch": 135.0,
      "eval_loss": 3.8817522525787354,
      "eval_runtime": 74.832,
      "eval_samples_per_second": 6580.394,
      "eval_steps_per_second": 2.151,
      "step": 194940
    },
    {
      "epoch": 135.04,
      "learning_rate": 1.7299168975069253e-05,
      "loss": 3.9648,
      "step": 195000
    },
    {
      "epoch": 135.39,
      "learning_rate": 1.7292243767313023e-05,
      "loss": 3.9619,
      "step": 195500
    },
    {
      "epoch": 135.73,
      "learning_rate": 1.7285318559556787e-05,
      "loss": 3.9635,
      "step": 196000
    },
    {
      "epoch": 136.0,
      "eval_loss": 3.8839354515075684,
      "eval_runtime": 75.8115,
      "eval_samples_per_second": 6495.378,
      "eval_steps_per_second": 2.124,
      "step": 196384
    },
    {
      "epoch": 136.08,
      "learning_rate": 1.7278393351800557e-05,
      "loss": 3.9585,
      "step": 196500
    },
    {
      "epoch": 136.43,
      "learning_rate": 1.7271468144044324e-05,
      "loss": 3.9597,
      "step": 197000
    },
    {
      "epoch": 136.77,
      "learning_rate": 1.7264542936288087e-05,
      "loss": 3.9601,
      "step": 197500
    },
    {
      "epoch": 137.0,
      "eval_loss": 3.8804266452789307,
      "eval_runtime": 75.6839,
      "eval_samples_per_second": 6506.321,
      "eval_steps_per_second": 2.127,
      "step": 197828
    },
    {
      "epoch": 137.12,
      "learning_rate": 1.7257617728531858e-05,
      "loss": 3.9609,
      "step": 198000
    },
    {
      "epoch": 137.47,
      "learning_rate": 1.7250692520775625e-05,
      "loss": 3.956,
      "step": 198500
    },
    {
      "epoch": 137.81,
      "learning_rate": 1.724376731301939e-05,
      "loss": 3.9593,
      "step": 199000
    },
    {
      "epoch": 138.0,
      "eval_loss": 3.879089117050171,
      "eval_runtime": 74.4504,
      "eval_samples_per_second": 6614.122,
      "eval_steps_per_second": 2.163,
      "step": 199272
    },
    {
      "epoch": 138.16,
      "learning_rate": 1.723684210526316e-05,
      "loss": 3.9569,
      "step": 199500
    },
    {
      "epoch": 138.5,
      "learning_rate": 1.7229916897506925e-05,
      "loss": 3.9541,
      "step": 200000
    },
    {
      "epoch": 138.85,
      "learning_rate": 1.7222991689750696e-05,
      "loss": 3.9566,
      "step": 200500
    },
    {
      "epoch": 139.0,
      "eval_loss": 3.8803153038024902,
      "eval_runtime": 77.8029,
      "eval_samples_per_second": 6329.123,
      "eval_steps_per_second": 2.069,
      "step": 200716
    },
    {
      "epoch": 139.2,
      "learning_rate": 1.7216066481994462e-05,
      "loss": 3.956,
      "step": 201000
    },
    {
      "epoch": 139.54,
      "learning_rate": 1.720914127423823e-05,
      "loss": 3.9517,
      "step": 201500
    },
    {
      "epoch": 139.89,
      "learning_rate": 1.7202216066481996e-05,
      "loss": 3.9535,
      "step": 202000
    },
    {
      "epoch": 140.0,
      "eval_loss": 3.872063159942627,
      "eval_runtime": 75.6675,
      "eval_samples_per_second": 6507.738,
      "eval_steps_per_second": 2.128,
      "step": 202160
    },
    {
      "epoch": 140.24,
      "learning_rate": 1.7195290858725763e-05,
      "loss": 3.9534,
      "step": 202500
    },
    {
      "epoch": 140.58,
      "learning_rate": 1.718836565096953e-05,
      "loss": 3.9534,
      "step": 203000
    },
    {
      "epoch": 140.93,
      "learning_rate": 1.7181440443213297e-05,
      "loss": 3.9525,
      "step": 203500
    },
    {
      "epoch": 141.0,
      "eval_loss": 3.8730010986328125,
      "eval_runtime": 73.7321,
      "eval_samples_per_second": 6678.556,
      "eval_steps_per_second": 2.184,
      "step": 203604
    },
    {
      "epoch": 141.27,
      "learning_rate": 1.7174515235457064e-05,
      "loss": 3.9519,
      "step": 204000
    },
    {
      "epoch": 141.62,
      "learning_rate": 1.7167590027700834e-05,
      "loss": 3.9488,
      "step": 204500
    },
    {
      "epoch": 141.97,
      "learning_rate": 1.7160664819944598e-05,
      "loss": 3.9515,
      "step": 205000
    },
    {
      "epoch": 142.0,
      "eval_loss": 3.870518207550049,
      "eval_runtime": 74.6993,
      "eval_samples_per_second": 6592.087,
      "eval_steps_per_second": 2.155,
      "step": 205048
    },
    {
      "epoch": 142.31,
      "learning_rate": 1.7153739612188368e-05,
      "loss": 3.9499,
      "step": 205500
    },
    {
      "epoch": 142.66,
      "learning_rate": 1.7146814404432135e-05,
      "loss": 3.9488,
      "step": 206000
    },
    {
      "epoch": 143.0,
      "eval_loss": 3.868440628051758,
      "eval_runtime": 73.8028,
      "eval_samples_per_second": 6672.162,
      "eval_steps_per_second": 2.181,
      "step": 206492
    },
    {
      "epoch": 143.01,
      "learning_rate": 1.7139889196675902e-05,
      "loss": 3.9525,
      "step": 206500
    },
    {
      "epoch": 143.35,
      "learning_rate": 1.713296398891967e-05,
      "loss": 3.9484,
      "step": 207000
    },
    {
      "epoch": 143.7,
      "learning_rate": 1.7126038781163436e-05,
      "loss": 3.944,
      "step": 207500
    },
    {
      "epoch": 144.0,
      "eval_loss": 3.8690621852874756,
      "eval_runtime": 74.7082,
      "eval_samples_per_second": 6591.3,
      "eval_steps_per_second": 2.155,
      "step": 207936
    },
    {
      "epoch": 144.04,
      "learning_rate": 1.7119113573407206e-05,
      "loss": 3.9512,
      "step": 208000
    },
    {
      "epoch": 144.39,
      "learning_rate": 1.711218836565097e-05,
      "loss": 3.9497,
      "step": 208500
    },
    {
      "epoch": 144.74,
      "learning_rate": 1.7105263157894737e-05,
      "loss": 3.9455,
      "step": 209000
    },
    {
      "epoch": 145.0,
      "eval_loss": 3.8670220375061035,
      "eval_runtime": 74.3832,
      "eval_samples_per_second": 6620.095,
      "eval_steps_per_second": 2.164,
      "step": 209380
    },
    {
      "epoch": 145.08,
      "learning_rate": 1.7098337950138507e-05,
      "loss": 3.9445,
      "step": 209500
    },
    {
      "epoch": 145.43,
      "learning_rate": 1.709141274238227e-05,
      "loss": 3.945,
      "step": 210000
    },
    {
      "epoch": 145.78,
      "learning_rate": 1.708448753462604e-05,
      "loss": 3.9439,
      "step": 210500
    },
    {
      "epoch": 146.0,
      "eval_loss": 3.865227699279785,
      "eval_runtime": 76.1017,
      "eval_samples_per_second": 6470.6,
      "eval_steps_per_second": 2.116,
      "step": 210824
    },
    {
      "epoch": 146.12,
      "learning_rate": 1.7077562326869808e-05,
      "loss": 3.9444,
      "step": 211000
    },
    {
      "epoch": 146.47,
      "learning_rate": 1.7070637119113575e-05,
      "loss": 3.9402,
      "step": 211500
    },
    {
      "epoch": 146.81,
      "learning_rate": 1.706371191135734e-05,
      "loss": 3.9431,
      "step": 212000
    },
    {
      "epoch": 147.0,
      "eval_loss": 3.8667104244232178,
      "eval_runtime": 74.0634,
      "eval_samples_per_second": 6648.686,
      "eval_steps_per_second": 2.174,
      "step": 212268
    },
    {
      "epoch": 147.16,
      "learning_rate": 1.705678670360111e-05,
      "loss": 3.9421,
      "step": 212500
    },
    {
      "epoch": 147.51,
      "learning_rate": 1.704986149584488e-05,
      "loss": 3.9406,
      "step": 213000
    },
    {
      "epoch": 147.85,
      "learning_rate": 1.7042936288088646e-05,
      "loss": 3.9422,
      "step": 213500
    },
    {
      "epoch": 148.0,
      "eval_loss": 3.8651583194732666,
      "eval_runtime": 74.2586,
      "eval_samples_per_second": 6631.208,
      "eval_steps_per_second": 2.168,
      "step": 213712
    },
    {
      "epoch": 148.2,
      "learning_rate": 1.703601108033241e-05,
      "loss": 3.9379,
      "step": 214000
    },
    {
      "epoch": 148.55,
      "learning_rate": 1.702908587257618e-05,
      "loss": 3.9402,
      "step": 214500
    },
    {
      "epoch": 148.89,
      "learning_rate": 1.7022160664819946e-05,
      "loss": 3.9389,
      "step": 215000
    },
    {
      "epoch": 149.0,
      "eval_loss": 3.8652150630950928,
      "eval_runtime": 74.5415,
      "eval_samples_per_second": 6606.034,
      "eval_steps_per_second": 2.16,
      "step": 215156
    },
    {
      "epoch": 149.24,
      "learning_rate": 1.7015235457063713e-05,
      "loss": 3.9387,
      "step": 215500
    },
    {
      "epoch": 149.58,
      "learning_rate": 1.700831024930748e-05,
      "loss": 3.9394,
      "step": 216000
    },
    {
      "epoch": 149.93,
      "learning_rate": 1.7001385041551247e-05,
      "loss": 3.9387,
      "step": 216500
    },
    {
      "epoch": 150.0,
      "eval_loss": 3.8585476875305176,
      "eval_runtime": 75.2231,
      "eval_samples_per_second": 6546.184,
      "eval_steps_per_second": 2.14,
      "step": 216600
    },
    {
      "epoch": 150.28,
      "learning_rate": 1.6994459833795017e-05,
      "loss": 3.9363,
      "step": 217000
    },
    {
      "epoch": 150.62,
      "learning_rate": 1.698753462603878e-05,
      "loss": 3.9406,
      "step": 217500
    },
    {
      "epoch": 150.97,
      "learning_rate": 1.698060941828255e-05,
      "loss": 3.9375,
      "step": 218000
    },
    {
      "epoch": 151.0,
      "eval_loss": 3.8584671020507812,
      "eval_runtime": 75.4049,
      "eval_samples_per_second": 6530.4,
      "eval_steps_per_second": 2.135,
      "step": 218044
    },
    {
      "epoch": 151.32,
      "learning_rate": 1.6973684210526318e-05,
      "loss": 3.9365,
      "step": 218500
    },
    {
      "epoch": 151.66,
      "learning_rate": 1.6966759002770085e-05,
      "loss": 3.9365,
      "step": 219000
    },
    {
      "epoch": 152.0,
      "eval_loss": 3.8540356159210205,
      "eval_runtime": 72.8254,
      "eval_samples_per_second": 6761.71,
      "eval_steps_per_second": 2.211,
      "step": 219488
    },
    {
      "epoch": 152.01,
      "learning_rate": 1.6959833795013852e-05,
      "loss": 3.9354,
      "step": 219500
    },
    {
      "epoch": 152.35,
      "learning_rate": 1.695290858725762e-05,
      "loss": 3.9347,
      "step": 220000
    },
    {
      "epoch": 152.7,
      "learning_rate": 1.6945983379501386e-05,
      "loss": 3.9313,
      "step": 220500
    },
    {
      "epoch": 153.0,
      "eval_loss": 3.853323459625244,
      "eval_runtime": 74.3001,
      "eval_samples_per_second": 6627.5,
      "eval_steps_per_second": 2.167,
      "step": 220932
    },
    {
      "epoch": 153.05,
      "learning_rate": 1.6939058171745153e-05,
      "loss": 3.9335,
      "step": 221000
    },
    {
      "epoch": 153.39,
      "learning_rate": 1.693213296398892e-05,
      "loss": 3.932,
      "step": 221500
    },
    {
      "epoch": 153.74,
      "learning_rate": 1.692520775623269e-05,
      "loss": 3.9287,
      "step": 222000
    },
    {
      "epoch": 154.0,
      "eval_loss": 3.853574275970459,
      "eval_runtime": 74.3871,
      "eval_samples_per_second": 6619.754,
      "eval_steps_per_second": 2.164,
      "step": 222376
    },
    {
      "epoch": 154.09,
      "learning_rate": 1.6918282548476454e-05,
      "loss": 3.935,
      "step": 222500
    },
    {
      "epoch": 154.43,
      "learning_rate": 1.6911357340720224e-05,
      "loss": 3.9323,
      "step": 223000
    },
    {
      "epoch": 154.78,
      "learning_rate": 1.690443213296399e-05,
      "loss": 3.9304,
      "step": 223500
    },
    {
      "epoch": 155.0,
      "eval_loss": 3.854647636413574,
      "eval_runtime": 75.0693,
      "eval_samples_per_second": 6559.592,
      "eval_steps_per_second": 2.145,
      "step": 223820
    },
    {
      "epoch": 155.12,
      "learning_rate": 1.6897506925207758e-05,
      "loss": 3.9285,
      "step": 224000
    },
    {
      "epoch": 155.47,
      "learning_rate": 1.6890581717451525e-05,
      "loss": 3.927,
      "step": 224500
    },
    {
      "epoch": 155.82,
      "learning_rate": 1.688365650969529e-05,
      "loss": 3.9304,
      "step": 225000
    },
    {
      "epoch": 156.0,
      "eval_loss": 3.8501617908477783,
      "eval_runtime": 75.8197,
      "eval_samples_per_second": 6494.673,
      "eval_steps_per_second": 2.123,
      "step": 225264
    },
    {
      "epoch": 156.16,
      "learning_rate": 1.687673130193906e-05,
      "loss": 3.9299,
      "step": 225500
    },
    {
      "epoch": 156.51,
      "learning_rate": 1.686980609418283e-05,
      "loss": 3.9282,
      "step": 226000
    },
    {
      "epoch": 156.86,
      "learning_rate": 1.6862880886426592e-05,
      "loss": 3.9257,
      "step": 226500
    },
    {
      "epoch": 157.0,
      "eval_loss": 3.8476805686950684,
      "eval_runtime": 73.4445,
      "eval_samples_per_second": 6704.713,
      "eval_steps_per_second": 2.192,
      "step": 226708
    },
    {
      "epoch": 157.2,
      "learning_rate": 1.6855955678670363e-05,
      "loss": 3.9263,
      "step": 227000
    },
    {
      "epoch": 157.55,
      "learning_rate": 1.684903047091413e-05,
      "loss": 3.9247,
      "step": 227500
    },
    {
      "epoch": 157.89,
      "learning_rate": 1.6842105263157896e-05,
      "loss": 3.9242,
      "step": 228000
    },
    {
      "epoch": 158.0,
      "eval_loss": 3.8495914936065674,
      "eval_runtime": 74.4187,
      "eval_samples_per_second": 6616.942,
      "eval_steps_per_second": 2.163,
      "step": 228152
    },
    {
      "epoch": 158.24,
      "learning_rate": 1.6835180055401663e-05,
      "loss": 3.9235,
      "step": 228500
    },
    {
      "epoch": 158.59,
      "learning_rate": 1.682825484764543e-05,
      "loss": 3.9275,
      "step": 229000
    },
    {
      "epoch": 158.93,
      "learning_rate": 1.68213296398892e-05,
      "loss": 3.9219,
      "step": 229500
    },
    {
      "epoch": 159.0,
      "eval_loss": 3.850391387939453,
      "eval_runtime": 74.2125,
      "eval_samples_per_second": 6635.322,
      "eval_steps_per_second": 2.169,
      "step": 229596
    },
    {
      "epoch": 159.28,
      "learning_rate": 1.6814404432132964e-05,
      "loss": 3.9203,
      "step": 230000
    },
    {
      "epoch": 159.63,
      "learning_rate": 1.680747922437673e-05,
      "loss": 3.9239,
      "step": 230500
    },
    {
      "epoch": 159.97,
      "learning_rate": 1.68005540166205e-05,
      "loss": 3.9236,
      "step": 231000
    },
    {
      "epoch": 160.0,
      "eval_loss": 3.8478267192840576,
      "eval_runtime": 74.9917,
      "eval_samples_per_second": 6566.379,
      "eval_steps_per_second": 2.147,
      "step": 231040
    },
    {
      "epoch": 160.32,
      "learning_rate": 1.6793628808864265e-05,
      "loss": 3.9238,
      "step": 231500
    },
    {
      "epoch": 160.66,
      "learning_rate": 1.6786703601108035e-05,
      "loss": 3.9222,
      "step": 232000
    },
    {
      "epoch": 161.0,
      "eval_loss": 3.8385064601898193,
      "eval_runtime": 74.3809,
      "eval_samples_per_second": 6620.298,
      "eval_steps_per_second": 2.165,
      "step": 232484
    },
    {
      "epoch": 161.01,
      "learning_rate": 1.6779778393351802e-05,
      "loss": 3.92,
      "step": 232500
    },
    {
      "epoch": 161.36,
      "learning_rate": 1.677285318559557e-05,
      "loss": 3.9184,
      "step": 233000
    },
    {
      "epoch": 161.7,
      "learning_rate": 1.6765927977839336e-05,
      "loss": 3.9231,
      "step": 233500
    },
    {
      "epoch": 162.0,
      "eval_loss": 3.844747543334961,
      "eval_runtime": 73.4727,
      "eval_samples_per_second": 6702.138,
      "eval_steps_per_second": 2.191,
      "step": 233928
    },
    {
      "epoch": 162.05,
      "learning_rate": 1.6759002770083103e-05,
      "loss": 3.9186,
      "step": 234000
    },
    {
      "epoch": 162.4,
      "learning_rate": 1.6752077562326873e-05,
      "loss": 3.9207,
      "step": 234500
    },
    {
      "epoch": 162.74,
      "learning_rate": 1.674515235457064e-05,
      "loss": 3.9161,
      "step": 235000
    },
    {
      "epoch": 163.0,
      "eval_loss": 3.839341640472412,
      "eval_runtime": 74.6904,
      "eval_samples_per_second": 6592.87,
      "eval_steps_per_second": 2.156,
      "step": 235372
    },
    {
      "epoch": 163.09,
      "learning_rate": 1.6738227146814404e-05,
      "loss": 3.9223,
      "step": 235500
    },
    {
      "epoch": 163.43,
      "learning_rate": 1.6731301939058174e-05,
      "loss": 3.9151,
      "step": 236000
    },
    {
      "epoch": 163.78,
      "learning_rate": 1.672437673130194e-05,
      "loss": 3.918,
      "step": 236500
    },
    {
      "epoch": 164.0,
      "eval_loss": 3.8450632095336914,
      "eval_runtime": 74.8435,
      "eval_samples_per_second": 6579.385,
      "eval_steps_per_second": 2.151,
      "step": 236816
    },
    {
      "epoch": 164.13,
      "learning_rate": 1.6717451523545708e-05,
      "loss": 3.9184,
      "step": 237000
    },
    {
      "epoch": 164.47,
      "learning_rate": 1.6710526315789475e-05,
      "loss": 3.9175,
      "step": 237500
    },
    {
      "epoch": 164.82,
      "learning_rate": 1.670360110803324e-05,
      "loss": 3.9153,
      "step": 238000
    },
    {
      "epoch": 165.0,
      "eval_loss": 3.8364455699920654,
      "eval_runtime": 75.2179,
      "eval_samples_per_second": 6546.636,
      "eval_steps_per_second": 2.14,
      "step": 238260
    },
    {
      "epoch": 165.17,
      "learning_rate": 1.6696675900277012e-05,
      "loss": 3.9151,
      "step": 238500
    },
    {
      "epoch": 165.51,
      "learning_rate": 1.6689750692520775e-05,
      "loss": 3.9139,
      "step": 239000
    },
    {
      "epoch": 165.86,
      "learning_rate": 1.6682825484764546e-05,
      "loss": 3.9158,
      "step": 239500
    },
    {
      "epoch": 166.0,
      "eval_loss": 3.8352112770080566,
      "eval_runtime": 74.8807,
      "eval_samples_per_second": 6576.117,
      "eval_steps_per_second": 2.15,
      "step": 239704
    },
    {
      "epoch": 166.2,
      "learning_rate": 1.6675900277008313e-05,
      "loss": 3.9151,
      "step": 240000
    },
    {
      "epoch": 166.55,
      "learning_rate": 1.666897506925208e-05,
      "loss": 3.9146,
      "step": 240500
    },
    {
      "epoch": 166.9,
      "learning_rate": 1.6662049861495847e-05,
      "loss": 3.9138,
      "step": 241000
    },
    {
      "epoch": 167.0,
      "eval_loss": 3.835671901702881,
      "eval_runtime": 74.7206,
      "eval_samples_per_second": 6590.203,
      "eval_steps_per_second": 2.155,
      "step": 241148
    },
    {
      "epoch": 167.24,
      "learning_rate": 1.6655124653739613e-05,
      "loss": 3.9147,
      "step": 241500
    },
    {
      "epoch": 167.59,
      "learning_rate": 1.664819944598338e-05,
      "loss": 3.9128,
      "step": 242000
    },
    {
      "epoch": 167.94,
      "learning_rate": 1.6641274238227147e-05,
      "loss": 3.9135,
      "step": 242500
    },
    {
      "epoch": 168.0,
      "eval_loss": 3.829854726791382,
      "eval_runtime": 72.8907,
      "eval_samples_per_second": 6755.652,
      "eval_steps_per_second": 2.209,
      "step": 242592
    },
    {
      "epoch": 168.28,
      "learning_rate": 1.6634349030470914e-05,
      "loss": 3.9116,
      "step": 243000
    },
    {
      "epoch": 168.63,
      "learning_rate": 1.6627423822714685e-05,
      "loss": 3.9104,
      "step": 243500
    },
    {
      "epoch": 168.98,
      "learning_rate": 1.6620498614958448e-05,
      "loss": 3.9086,
      "step": 244000
    },
    {
      "epoch": 169.0,
      "eval_loss": 3.83186411857605,
      "eval_runtime": 73.8836,
      "eval_samples_per_second": 6664.86,
      "eval_steps_per_second": 2.179,
      "step": 244036
    },
    {
      "epoch": 169.32,
      "learning_rate": 1.661357340720222e-05,
      "loss": 3.9102,
      "step": 244500
    },
    {
      "epoch": 169.67,
      "learning_rate": 1.6606648199445985e-05,
      "loss": 3.9093,
      "step": 245000
    },
    {
      "epoch": 170.0,
      "eval_loss": 3.831186532974243,
      "eval_runtime": 75.6023,
      "eval_samples_per_second": 6513.346,
      "eval_steps_per_second": 2.13,
      "step": 245480
    },
    {
      "epoch": 170.01,
      "learning_rate": 1.6599722991689752e-05,
      "loss": 3.9092,
      "step": 245500
    },
    {
      "epoch": 170.36,
      "learning_rate": 1.659279778393352e-05,
      "loss": 3.9083,
      "step": 246000
    },
    {
      "epoch": 170.71,
      "learning_rate": 1.6585872576177286e-05,
      "loss": 3.9087,
      "step": 246500
    },
    {
      "epoch": 171.0,
      "eval_loss": 3.8296992778778076,
      "eval_runtime": 74.0581,
      "eval_samples_per_second": 6649.158,
      "eval_steps_per_second": 2.174,
      "step": 246924
    },
    {
      "epoch": 171.05,
      "learning_rate": 1.6578947368421053e-05,
      "loss": 3.9097,
      "step": 247000
    },
    {
      "epoch": 171.4,
      "learning_rate": 1.6572022160664823e-05,
      "loss": 3.9103,
      "step": 247500
    },
    {
      "epoch": 171.75,
      "learning_rate": 1.6565096952908587e-05,
      "loss": 3.9078,
      "step": 248000
    },
    {
      "epoch": 172.0,
      "eval_loss": 3.8280537128448486,
      "eval_runtime": 74.6178,
      "eval_samples_per_second": 6599.284,
      "eval_steps_per_second": 2.158,
      "step": 248368
    },
    {
      "epoch": 172.09,
      "learning_rate": 1.6558171745152357e-05,
      "loss": 3.9065,
      "step": 248500
    },
    {
      "epoch": 172.44,
      "learning_rate": 1.6551246537396124e-05,
      "loss": 3.9052,
      "step": 249000
    },
    {
      "epoch": 172.78,
      "learning_rate": 1.654432132963989e-05,
      "loss": 3.9052,
      "step": 249500
    },
    {
      "epoch": 173.0,
      "eval_loss": 3.828386068344116,
      "eval_runtime": 74.5947,
      "eval_samples_per_second": 6601.323,
      "eval_steps_per_second": 2.158,
      "step": 249812
    },
    {
      "epoch": 173.13,
      "learning_rate": 1.6537396121883658e-05,
      "loss": 3.9037,
      "step": 250000
    },
    {
      "epoch": 173.48,
      "learning_rate": 1.6530470914127425e-05,
      "loss": 3.9046,
      "step": 250500
    },
    {
      "epoch": 173.82,
      "learning_rate": 1.6523545706371195e-05,
      "loss": 3.9059,
      "step": 251000
    },
    {
      "epoch": 174.0,
      "eval_loss": 3.82259464263916,
      "eval_runtime": 75.1329,
      "eval_samples_per_second": 6554.041,
      "eval_steps_per_second": 2.143,
      "step": 251256
    },
    {
      "epoch": 174.17,
      "learning_rate": 1.651662049861496e-05,
      "loss": 3.9044,
      "step": 251500
    },
    {
      "epoch": 174.52,
      "learning_rate": 1.6509695290858726e-05,
      "loss": 3.9041,
      "step": 252000
    },
    {
      "epoch": 174.86,
      "learning_rate": 1.6502770083102496e-05,
      "loss": 3.9006,
      "step": 252500
    },
    {
      "epoch": 175.0,
      "eval_loss": 3.8257510662078857,
      "eval_runtime": 75.0788,
      "eval_samples_per_second": 6558.759,
      "eval_steps_per_second": 2.144,
      "step": 252700
    },
    {
      "epoch": 175.21,
      "learning_rate": 1.649584487534626e-05,
      "loss": 3.9025,
      "step": 253000
    },
    {
      "epoch": 175.55,
      "learning_rate": 1.648891966759003e-05,
      "loss": 3.9015,
      "step": 253500
    },
    {
      "epoch": 175.9,
      "learning_rate": 1.6481994459833797e-05,
      "loss": 3.9046,
      "step": 254000
    },
    {
      "epoch": 176.0,
      "eval_loss": 3.8246407508850098,
      "eval_runtime": 76.1781,
      "eval_samples_per_second": 6464.119,
      "eval_steps_per_second": 2.113,
      "step": 254144
    },
    {
      "epoch": 176.25,
      "learning_rate": 1.6475069252077564e-05,
      "loss": 3.9034,
      "step": 254500
    },
    {
      "epoch": 176.59,
      "learning_rate": 1.646814404432133e-05,
      "loss": 3.8996,
      "step": 255000
    },
    {
      "epoch": 176.94,
      "learning_rate": 1.6461218836565097e-05,
      "loss": 3.9027,
      "step": 255500
    },
    {
      "epoch": 177.0,
      "eval_loss": 3.81931471824646,
      "eval_runtime": 74.6152,
      "eval_samples_per_second": 6599.512,
      "eval_steps_per_second": 2.158,
      "step": 255588
    },
    {
      "epoch": 177.29,
      "learning_rate": 1.6454293628808868e-05,
      "loss": 3.9021,
      "step": 256000
    },
    {
      "epoch": 177.63,
      "learning_rate": 1.644736842105263e-05,
      "loss": 3.8989,
      "step": 256500
    },
    {
      "epoch": 177.98,
      "learning_rate": 1.6440443213296398e-05,
      "loss": 3.9024,
      "step": 257000
    },
    {
      "epoch": 178.0,
      "eval_loss": 3.819705009460449,
      "eval_runtime": 75.4024,
      "eval_samples_per_second": 6530.618,
      "eval_steps_per_second": 2.135,
      "step": 257032
    },
    {
      "epoch": 178.32,
      "learning_rate": 1.643351800554017e-05,
      "loss": 3.9012,
      "step": 257500
    },
    {
      "epoch": 178.67,
      "learning_rate": 1.6426592797783935e-05,
      "loss": 3.9007,
      "step": 258000
    },
    {
      "epoch": 179.0,
      "eval_loss": 3.817988634109497,
      "eval_runtime": 74.2631,
      "eval_samples_per_second": 6630.806,
      "eval_steps_per_second": 2.168,
      "step": 258476
    },
    {
      "epoch": 179.02,
      "learning_rate": 1.6419667590027702e-05,
      "loss": 3.8962,
      "step": 258500
    },
    {
      "epoch": 179.36,
      "learning_rate": 1.641274238227147e-05,
      "loss": 3.9005,
      "step": 259000
    },
    {
      "epoch": 179.71,
      "learning_rate": 1.6405817174515236e-05,
      "loss": 3.897,
      "step": 259500
    },
    {
      "epoch": 180.0,
      "eval_loss": 3.8208701610565186,
      "eval_runtime": 75.7163,
      "eval_samples_per_second": 6503.537,
      "eval_steps_per_second": 2.126,
      "step": 259920
    },
    {
      "epoch": 180.06,
      "learning_rate": 1.6398891966759006e-05,
      "loss": 3.9019,
      "step": 260000
    },
    {
      "epoch": 180.4,
      "learning_rate": 1.639196675900277e-05,
      "loss": 3.8984,
      "step": 260500
    },
    {
      "epoch": 180.75,
      "learning_rate": 1.638504155124654e-05,
      "loss": 3.8967,
      "step": 261000
    },
    {
      "epoch": 181.0,
      "eval_loss": 3.814295768737793,
      "eval_runtime": 74.2326,
      "eval_samples_per_second": 6633.527,
      "eval_steps_per_second": 2.169,
      "step": 261364
    },
    {
      "epoch": 181.09,
      "learning_rate": 1.6378116343490307e-05,
      "loss": 3.8996,
      "step": 261500
    },
    {
      "epoch": 181.44,
      "learning_rate": 1.6371191135734074e-05,
      "loss": 3.8951,
      "step": 262000
    },
    {
      "epoch": 181.79,
      "learning_rate": 1.636426592797784e-05,
      "loss": 3.8978,
      "step": 262500
    },
    {
      "epoch": 182.0,
      "eval_loss": 3.816028118133545,
      "eval_runtime": 74.7425,
      "eval_samples_per_second": 6588.271,
      "eval_steps_per_second": 2.154,
      "step": 262808
    },
    {
      "epoch": 182.13,
      "learning_rate": 1.6357340720221608e-05,
      "loss": 3.896,
      "step": 263000
    },
    {
      "epoch": 182.48,
      "learning_rate": 1.6350415512465375e-05,
      "loss": 3.8943,
      "step": 263500
    },
    {
      "epoch": 182.83,
      "learning_rate": 1.6343490304709142e-05,
      "loss": 3.894,
      "step": 264000
    },
    {
      "epoch": 183.0,
      "eval_loss": 3.8133652210235596,
      "eval_runtime": 76.198,
      "eval_samples_per_second": 6462.426,
      "eval_steps_per_second": 2.113,
      "step": 264252
    },
    {
      "epoch": 183.17,
      "learning_rate": 1.633656509695291e-05,
      "loss": 3.8947,
      "step": 264500
    },
    {
      "epoch": 183.52,
      "learning_rate": 1.632963988919668e-05,
      "loss": 3.8949,
      "step": 265000
    },
    {
      "epoch": 183.86,
      "learning_rate": 1.6322714681440443e-05,
      "loss": 3.8919,
      "step": 265500
    },
    {
      "epoch": 184.0,
      "eval_loss": 3.8209118843078613,
      "eval_runtime": 74.4007,
      "eval_samples_per_second": 6618.538,
      "eval_steps_per_second": 2.164,
      "step": 265696
    },
    {
      "epoch": 184.21,
      "learning_rate": 1.6315789473684213e-05,
      "loss": 3.8918,
      "step": 266000
    },
    {
      "epoch": 184.56,
      "learning_rate": 1.630886426592798e-05,
      "loss": 3.893,
      "step": 266500
    },
    {
      "epoch": 184.9,
      "learning_rate": 1.6301939058171747e-05,
      "loss": 3.8915,
      "step": 267000
    },
    {
      "epoch": 185.0,
      "eval_loss": 3.812711477279663,
      "eval_runtime": 74.8088,
      "eval_samples_per_second": 6582.437,
      "eval_steps_per_second": 2.152,
      "step": 267140
    },
    {
      "epoch": 185.25,
      "learning_rate": 1.6295013850415514e-05,
      "loss": 3.8911,
      "step": 267500
    },
    {
      "epoch": 185.6,
      "learning_rate": 1.628808864265928e-05,
      "loss": 3.8878,
      "step": 268000
    },
    {
      "epoch": 185.94,
      "learning_rate": 1.6281163434903047e-05,
      "loss": 3.8909,
      "step": 268500
    },
    {
      "epoch": 186.0,
      "eval_loss": 3.8139772415161133,
      "eval_runtime": 73.8157,
      "eval_samples_per_second": 6670.991,
      "eval_steps_per_second": 2.181,
      "step": 268584
    },
    {
      "epoch": 186.29,
      "learning_rate": 1.6274238227146814e-05,
      "loss": 3.8906,
      "step": 269000
    },
    {
      "epoch": 186.63,
      "learning_rate": 1.626731301939058e-05,
      "loss": 3.8891,
      "step": 269500
    },
    {
      "epoch": 186.98,
      "learning_rate": 1.626038781163435e-05,
      "loss": 3.8894,
      "step": 270000
    },
    {
      "epoch": 187.0,
      "eval_loss": 3.811793327331543,
      "eval_runtime": 76.5559,
      "eval_samples_per_second": 6432.215,
      "eval_steps_per_second": 2.103,
      "step": 270028
    },
    {
      "epoch": 187.33,
      "learning_rate": 1.625346260387812e-05,
      "loss": 3.8857,
      "step": 270500
    },
    {
      "epoch": 187.67,
      "learning_rate": 1.6246537396121885e-05,
      "loss": 3.886,
      "step": 271000
    },
    {
      "epoch": 188.0,
      "eval_loss": 3.812798500061035,
      "eval_runtime": 73.9286,
      "eval_samples_per_second": 6660.807,
      "eval_steps_per_second": 2.178,
      "step": 271472
    },
    {
      "epoch": 188.02,
      "learning_rate": 1.6239612188365652e-05,
      "loss": 3.8875,
      "step": 271500
    },
    {
      "epoch": 188.37,
      "learning_rate": 1.623268698060942e-05,
      "loss": 3.8906,
      "step": 272000
    },
    {
      "epoch": 188.71,
      "learning_rate": 1.622576177285319e-05,
      "loss": 3.8868,
      "step": 272500
    },
    {
      "epoch": 189.0,
      "eval_loss": 3.808310031890869,
      "eval_runtime": 74.6,
      "eval_samples_per_second": 6600.861,
      "eval_steps_per_second": 2.158,
      "step": 272916
    },
    {
      "epoch": 189.06,
      "learning_rate": 1.6218836565096953e-05,
      "loss": 3.8846,
      "step": 273000
    },
    {
      "epoch": 189.4,
      "learning_rate": 1.621191135734072e-05,
      "loss": 3.8873,
      "step": 273500
    },
    {
      "epoch": 189.75,
      "learning_rate": 1.620498614958449e-05,
      "loss": 3.8827,
      "step": 274000
    },
    {
      "epoch": 190.0,
      "eval_loss": 3.8040313720703125,
      "eval_runtime": 73.574,
      "eval_samples_per_second": 6692.91,
      "eval_steps_per_second": 2.188,
      "step": 274360
    },
    {
      "epoch": 190.1,
      "learning_rate": 1.6198060941828254e-05,
      "loss": 3.888,
      "step": 274500
    },
    {
      "epoch": 190.44,
      "learning_rate": 1.6191135734072024e-05,
      "loss": 3.8879,
      "step": 275000
    },
    {
      "epoch": 190.79,
      "learning_rate": 1.618421052631579e-05,
      "loss": 3.8854,
      "step": 275500
    },
    {
      "epoch": 191.0,
      "eval_loss": 3.8076326847076416,
      "eval_runtime": 73.5075,
      "eval_samples_per_second": 6698.963,
      "eval_steps_per_second": 2.19,
      "step": 275804
    },
    {
      "epoch": 191.14,
      "learning_rate": 1.6177285318559558e-05,
      "loss": 3.8866,
      "step": 276000
    },
    {
      "epoch": 191.48,
      "learning_rate": 1.6170360110803325e-05,
      "loss": 3.8838,
      "step": 276500
    },
    {
      "epoch": 191.83,
      "learning_rate": 1.6163434903047092e-05,
      "loss": 3.8814,
      "step": 277000
    },
    {
      "epoch": 192.0,
      "eval_loss": 3.8077051639556885,
      "eval_runtime": 74.0447,
      "eval_samples_per_second": 6650.364,
      "eval_steps_per_second": 2.174,
      "step": 277248
    },
    {
      "epoch": 192.17,
      "learning_rate": 1.6156509695290862e-05,
      "loss": 3.8826,
      "step": 277500
    },
    {
      "epoch": 192.52,
      "learning_rate": 1.6149584487534626e-05,
      "loss": 3.8837,
      "step": 278000
    },
    {
      "epoch": 192.87,
      "learning_rate": 1.6142659279778396e-05,
      "loss": 3.8825,
      "step": 278500
    },
    {
      "epoch": 193.0,
      "eval_loss": 3.8101003170013428,
      "eval_runtime": 74.4578,
      "eval_samples_per_second": 6613.463,
      "eval_steps_per_second": 2.162,
      "step": 278692
    },
    {
      "epoch": 193.21,
      "learning_rate": 1.6135734072022163e-05,
      "loss": 3.8808,
      "step": 279000
    },
    {
      "epoch": 193.56,
      "learning_rate": 1.612880886426593e-05,
      "loss": 3.8832,
      "step": 279500
    },
    {
      "epoch": 193.91,
      "learning_rate": 1.6121883656509697e-05,
      "loss": 3.8824,
      "step": 280000
    },
    {
      "epoch": 194.0,
      "eval_loss": 3.8035974502563477,
      "eval_runtime": 73.2562,
      "eval_samples_per_second": 6721.947,
      "eval_steps_per_second": 2.198,
      "step": 280136
    },
    {
      "epoch": 194.25,
      "learning_rate": 1.6114958448753464e-05,
      "loss": 3.8806,
      "step": 280500
    },
    {
      "epoch": 194.6,
      "learning_rate": 1.610803324099723e-05,
      "loss": 3.8783,
      "step": 281000
    },
    {
      "epoch": 194.94,
      "learning_rate": 1.6101108033240998e-05,
      "loss": 3.8784,
      "step": 281500
    },
    {
      "epoch": 195.0,
      "eval_loss": 3.8016698360443115,
      "eval_runtime": 75.1162,
      "eval_samples_per_second": 6555.495,
      "eval_steps_per_second": 2.143,
      "step": 281580
    },
    {
      "epoch": 195.29,
      "learning_rate": 1.6094182825484764e-05,
      "loss": 3.8792,
      "step": 282000
    },
    {
      "epoch": 195.64,
      "learning_rate": 1.6087257617728535e-05,
      "loss": 3.8828,
      "step": 282500
    },
    {
      "epoch": 195.98,
      "learning_rate": 1.60803324099723e-05,
      "loss": 3.8796,
      "step": 283000
    },
    {
      "epoch": 196.0,
      "eval_loss": 3.8056745529174805,
      "eval_runtime": 74.4209,
      "eval_samples_per_second": 6616.747,
      "eval_steps_per_second": 2.163,
      "step": 283024
    },
    {
      "epoch": 196.33,
      "learning_rate": 1.607340720221607e-05,
      "loss": 3.8769,
      "step": 283500
    },
    {
      "epoch": 196.68,
      "learning_rate": 1.6066481994459835e-05,
      "loss": 3.8801,
      "step": 284000
    },
    {
      "epoch": 197.0,
      "eval_loss": 3.8007116317749023,
      "eval_runtime": 77.2912,
      "eval_samples_per_second": 6371.024,
      "eval_steps_per_second": 2.083,
      "step": 284468
    },
    {
      "epoch": 197.02,
      "learning_rate": 1.6059556786703602e-05,
      "loss": 3.8765,
      "step": 284500
    },
    {
      "epoch": 197.37,
      "learning_rate": 1.605263157894737e-05,
      "loss": 3.8789,
      "step": 285000
    },
    {
      "epoch": 197.71,
      "learning_rate": 1.6045706371191136e-05,
      "loss": 3.8787,
      "step": 285500
    },
    {
      "epoch": 198.0,
      "eval_loss": 3.7981278896331787,
      "eval_runtime": 74.6743,
      "eval_samples_per_second": 6594.29,
      "eval_steps_per_second": 2.156,
      "step": 285912
    },
    {
      "epoch": 198.06,
      "learning_rate": 1.6038781163434903e-05,
      "loss": 3.8782,
      "step": 286000
    },
    {
      "epoch": 198.41,
      "learning_rate": 1.6031855955678673e-05,
      "loss": 3.8791,
      "step": 286500
    },
    {
      "epoch": 198.75,
      "learning_rate": 1.6024930747922437e-05,
      "loss": 3.876,
      "step": 287000
    },
    {
      "epoch": 199.0,
      "eval_loss": 3.800157070159912,
      "eval_runtime": 74.4204,
      "eval_samples_per_second": 6616.785,
      "eval_steps_per_second": 2.163,
      "step": 287356
    },
    {
      "epoch": 199.1,
      "learning_rate": 1.6018005540166207e-05,
      "loss": 3.8771,
      "step": 287500
    },
    {
      "epoch": 199.45,
      "learning_rate": 1.6011080332409974e-05,
      "loss": 3.8753,
      "step": 288000
    },
    {
      "epoch": 199.79,
      "learning_rate": 1.600415512465374e-05,
      "loss": 3.8739,
      "step": 288500
    },
    {
      "epoch": 200.0,
      "eval_loss": 3.801537036895752,
      "eval_runtime": 73.7536,
      "eval_samples_per_second": 6676.611,
      "eval_steps_per_second": 2.183,
      "step": 288800
    },
    {
      "epoch": 200.14,
      "learning_rate": 1.5997229916897508e-05,
      "loss": 3.8727,
      "step": 289000
    },
    {
      "epoch": 200.48,
      "learning_rate": 1.5990304709141275e-05,
      "loss": 3.8757,
      "step": 289500
    },
    {
      "epoch": 200.83,
      "learning_rate": 1.5983379501385042e-05,
      "loss": 3.8718,
      "step": 290000
    },
    {
      "epoch": 201.0,
      "eval_loss": 3.7952911853790283,
      "eval_runtime": 79.0161,
      "eval_samples_per_second": 6231.949,
      "eval_steps_per_second": 2.038,
      "step": 290244
    },
    {
      "epoch": 201.18,
      "learning_rate": 1.597645429362881e-05,
      "loss": 3.8703,
      "step": 290500
    },
    {
      "epoch": 201.52,
      "learning_rate": 1.5969529085872576e-05,
      "loss": 3.8742,
      "step": 291000
    },
    {
      "epoch": 201.87,
      "learning_rate": 1.5962603878116346e-05,
      "loss": 3.8728,
      "step": 291500
    },
    {
      "epoch": 202.0,
      "eval_loss": 3.794914484024048,
      "eval_runtime": 73.3296,
      "eval_samples_per_second": 6715.218,
      "eval_steps_per_second": 2.196,
      "step": 291688
    },
    {
      "epoch": 202.22,
      "learning_rate": 1.5955678670360113e-05,
      "loss": 3.8716,
      "step": 292000
    },
    {
      "epoch": 202.56,
      "learning_rate": 1.594875346260388e-05,
      "loss": 3.8721,
      "step": 292500
    },
    {
      "epoch": 202.91,
      "learning_rate": 1.5941828254847647e-05,
      "loss": 3.871,
      "step": 293000
    },
    {
      "epoch": 203.0,
      "eval_loss": 3.791077136993408,
      "eval_runtime": 75.7759,
      "eval_samples_per_second": 6498.426,
      "eval_steps_per_second": 2.125,
      "step": 293132
    },
    {
      "epoch": 203.25,
      "learning_rate": 1.5934903047091414e-05,
      "loss": 3.8718,
      "step": 293500
    },
    {
      "epoch": 203.6,
      "learning_rate": 1.592797783933518e-05,
      "loss": 3.8714,
      "step": 294000
    },
    {
      "epoch": 203.95,
      "learning_rate": 1.5921052631578948e-05,
      "loss": 3.8686,
      "step": 294500
    },
    {
      "epoch": 204.0,
      "eval_loss": 3.791466474533081,
      "eval_runtime": 76.174,
      "eval_samples_per_second": 6464.462,
      "eval_steps_per_second": 2.114,
      "step": 294576
    },
    {
      "epoch": 204.29,
      "learning_rate": 1.5914127423822714e-05,
      "loss": 3.8699,
      "step": 295000
    },
    {
      "epoch": 204.64,
      "learning_rate": 1.5907202216066485e-05,
      "loss": 3.8704,
      "step": 295500
    },
    {
      "epoch": 204.99,
      "learning_rate": 1.590027700831025e-05,
      "loss": 3.8727,
      "step": 296000
    },
    {
      "epoch": 205.0,
      "eval_loss": 3.7933382987976074,
      "eval_runtime": 75.2857,
      "eval_samples_per_second": 6540.737,
      "eval_steps_per_second": 2.139,
      "step": 296020
    },
    {
      "epoch": 205.33,
      "learning_rate": 1.589335180055402e-05,
      "loss": 3.8698,
      "step": 296500
    },
    {
      "epoch": 205.68,
      "learning_rate": 1.5886426592797786e-05,
      "loss": 3.8697,
      "step": 297000
    },
    {
      "epoch": 206.0,
      "eval_loss": 3.792696952819824,
      "eval_runtime": 77.0496,
      "eval_samples_per_second": 6390.998,
      "eval_steps_per_second": 2.09,
      "step": 297464
    },
    {
      "epoch": 206.02,
      "learning_rate": 1.5879501385041552e-05,
      "loss": 3.8676,
      "step": 297500
    },
    {
      "epoch": 206.37,
      "learning_rate": 1.587257617728532e-05,
      "loss": 3.8685,
      "step": 298000
    },
    {
      "epoch": 206.72,
      "learning_rate": 1.5865650969529086e-05,
      "loss": 3.8684,
      "step": 298500
    },
    {
      "epoch": 207.0,
      "eval_loss": 3.792511224746704,
      "eval_runtime": 72.6673,
      "eval_samples_per_second": 6776.414,
      "eval_steps_per_second": 2.216,
      "step": 298908
    },
    {
      "epoch": 207.06,
      "learning_rate": 1.5858725761772857e-05,
      "loss": 3.8679,
      "step": 299000
    },
    {
      "epoch": 207.41,
      "learning_rate": 1.585180055401662e-05,
      "loss": 3.8687,
      "step": 299500
    },
    {
      "epoch": 207.76,
      "learning_rate": 1.584487534626039e-05,
      "loss": 3.8681,
      "step": 300000
    },
    {
      "epoch": 208.0,
      "eval_loss": 3.790055990219116,
      "eval_runtime": 74.2758,
      "eval_samples_per_second": 6629.67,
      "eval_steps_per_second": 2.168,
      "step": 300352
    },
    {
      "epoch": 208.1,
      "learning_rate": 1.5837950138504157e-05,
      "loss": 3.8691,
      "step": 300500
    },
    {
      "epoch": 208.45,
      "learning_rate": 1.5831024930747924e-05,
      "loss": 3.8646,
      "step": 301000
    },
    {
      "epoch": 208.8,
      "learning_rate": 1.582409972299169e-05,
      "loss": 3.8665,
      "step": 301500
    },
    {
      "epoch": 209.0,
      "eval_loss": 3.7894084453582764,
      "eval_runtime": 74.4843,
      "eval_samples_per_second": 6611.112,
      "eval_steps_per_second": 2.162,
      "step": 301796
    },
    {
      "epoch": 209.14,
      "learning_rate": 1.5817174515235458e-05,
      "loss": 3.8649,
      "step": 302000
    },
    {
      "epoch": 209.49,
      "learning_rate": 1.5810249307479225e-05,
      "loss": 3.8624,
      "step": 302500
    },
    {
      "epoch": 209.83,
      "learning_rate": 1.5803324099722992e-05,
      "loss": 3.868,
      "step": 303000
    },
    {
      "epoch": 210.0,
      "eval_loss": 3.7948224544525146,
      "eval_runtime": 77.5312,
      "eval_samples_per_second": 6351.3,
      "eval_steps_per_second": 2.077,
      "step": 303240
    },
    {
      "epoch": 210.18,
      "learning_rate": 1.579639889196676e-05,
      "loss": 3.8675,
      "step": 303500
    },
    {
      "epoch": 210.53,
      "learning_rate": 1.578947368421053e-05,
      "loss": 3.8625,
      "step": 304000
    },
    {
      "epoch": 210.87,
      "learning_rate": 1.5782548476454296e-05,
      "loss": 3.8626,
      "step": 304500
    },
    {
      "epoch": 211.0,
      "eval_loss": 3.7883081436157227,
      "eval_runtime": 74.8996,
      "eval_samples_per_second": 6574.459,
      "eval_steps_per_second": 2.15,
      "step": 304684
    },
    {
      "epoch": 211.22,
      "learning_rate": 1.5775623268698063e-05,
      "loss": 3.8637,
      "step": 305000
    },
    {
      "epoch": 211.57,
      "learning_rate": 1.576869806094183e-05,
      "loss": 3.864,
      "step": 305500
    },
    {
      "epoch": 211.91,
      "learning_rate": 1.5761772853185597e-05,
      "loss": 3.8627,
      "step": 306000
    },
    {
      "epoch": 212.0,
      "eval_loss": 3.7848432064056396,
      "eval_runtime": 74.92,
      "eval_samples_per_second": 6572.667,
      "eval_steps_per_second": 2.149,
      "step": 306128
    },
    {
      "epoch": 212.26,
      "learning_rate": 1.5754847645429364e-05,
      "loss": 3.8658,
      "step": 306500
    },
    {
      "epoch": 212.6,
      "learning_rate": 1.574792243767313e-05,
      "loss": 3.8599,
      "step": 307000
    },
    {
      "epoch": 212.95,
      "learning_rate": 1.5740997229916898e-05,
      "loss": 3.865,
      "step": 307500
    },
    {
      "epoch": 213.0,
      "eval_loss": 3.787092924118042,
      "eval_runtime": 75.2395,
      "eval_samples_per_second": 6544.757,
      "eval_steps_per_second": 2.14,
      "step": 307572
    },
    {
      "epoch": 213.3,
      "learning_rate": 1.5734072022160668e-05,
      "loss": 3.861,
      "step": 308000
    },
    {
      "epoch": 213.64,
      "learning_rate": 1.572714681440443e-05,
      "loss": 3.8631,
      "step": 308500
    },
    {
      "epoch": 213.99,
      "learning_rate": 1.5720221606648202e-05,
      "loss": 3.8607,
      "step": 309000
    },
    {
      "epoch": 214.0,
      "eval_loss": 3.785153865814209,
      "eval_runtime": 74.7564,
      "eval_samples_per_second": 6587.052,
      "eval_steps_per_second": 2.154,
      "step": 309016
    },
    {
      "epoch": 214.34,
      "learning_rate": 1.571329639889197e-05,
      "loss": 3.8591,
      "step": 309500
    },
    {
      "epoch": 214.68,
      "learning_rate": 1.5706371191135736e-05,
      "loss": 3.8609,
      "step": 310000
    },
    {
      "epoch": 215.0,
      "eval_loss": 3.788590431213379,
      "eval_runtime": 76.5904,
      "eval_samples_per_second": 6429.315,
      "eval_steps_per_second": 2.102,
      "step": 310460
    },
    {
      "epoch": 215.03,
      "learning_rate": 1.5699445983379503e-05,
      "loss": 3.8604,
      "step": 310500
    },
    {
      "epoch": 215.37,
      "learning_rate": 1.569252077562327e-05,
      "loss": 3.8577,
      "step": 311000
    },
    {
      "epoch": 215.72,
      "learning_rate": 1.5685595567867036e-05,
      "loss": 3.8599,
      "step": 311500
    },
    {
      "epoch": 216.0,
      "eval_loss": 3.7840237617492676,
      "eval_runtime": 74.4828,
      "eval_samples_per_second": 6611.243,
      "eval_steps_per_second": 2.162,
      "step": 311904
    },
    {
      "epoch": 216.07,
      "learning_rate": 1.5678670360110803e-05,
      "loss": 3.8602,
      "step": 312000
    },
    {
      "epoch": 216.41,
      "learning_rate": 1.567174515235457e-05,
      "loss": 3.8573,
      "step": 312500
    },
    {
      "epoch": 216.76,
      "learning_rate": 1.566481994459834e-05,
      "loss": 3.8599,
      "step": 313000
    },
    {
      "epoch": 217.0,
      "eval_loss": 3.785320281982422,
      "eval_runtime": 74.9359,
      "eval_samples_per_second": 6571.268,
      "eval_steps_per_second": 2.149,
      "step": 313348
    },
    {
      "epoch": 217.11,
      "learning_rate": 1.5657894736842107e-05,
      "loss": 3.859,
      "step": 313500
    },
    {
      "epoch": 217.45,
      "learning_rate": 1.5650969529085874e-05,
      "loss": 3.8593,
      "step": 314000
    },
    {
      "epoch": 217.8,
      "learning_rate": 1.564404432132964e-05,
      "loss": 3.8573,
      "step": 314500
    },
    {
      "epoch": 218.0,
      "eval_loss": 3.7822437286376953,
      "eval_runtime": 74.2352,
      "eval_samples_per_second": 6633.293,
      "eval_steps_per_second": 2.169,
      "step": 314792
    },
    {
      "epoch": 218.14,
      "learning_rate": 1.5637119113573408e-05,
      "loss": 3.8571,
      "step": 315000
    },
    {
      "epoch": 218.49,
      "learning_rate": 1.5630193905817175e-05,
      "loss": 3.8573,
      "step": 315500
    },
    {
      "epoch": 218.84,
      "learning_rate": 1.5623268698060942e-05,
      "loss": 3.856,
      "step": 316000
    },
    {
      "epoch": 219.0,
      "eval_loss": 3.7809722423553467,
      "eval_runtime": 75.1168,
      "eval_samples_per_second": 6555.443,
      "eval_steps_per_second": 2.143,
      "step": 316236
    },
    {
      "epoch": 219.18,
      "learning_rate": 1.561634349030471e-05,
      "loss": 3.857,
      "step": 316500
    },
    {
      "epoch": 219.53,
      "learning_rate": 1.560941828254848e-05,
      "loss": 3.8534,
      "step": 317000
    },
    {
      "epoch": 219.88,
      "learning_rate": 1.5602493074792243e-05,
      "loss": 3.8546,
      "step": 317500
    },
    {
      "epoch": 220.0,
      "eval_loss": 3.775569200515747,
      "eval_runtime": 74.0328,
      "eval_samples_per_second": 6651.431,
      "eval_steps_per_second": 2.175,
      "step": 317680
    },
    {
      "epoch": 220.22,
      "learning_rate": 1.5595567867036013e-05,
      "loss": 3.8552,
      "step": 318000
    },
    {
      "epoch": 220.57,
      "learning_rate": 1.558864265927978e-05,
      "loss": 3.8551,
      "step": 318500
    },
    {
      "epoch": 220.91,
      "learning_rate": 1.5581717451523547e-05,
      "loss": 3.856,
      "step": 319000
    },
    {
      "epoch": 221.0,
      "eval_loss": 3.7812447547912598,
      "eval_runtime": 77.0452,
      "eval_samples_per_second": 6391.363,
      "eval_steps_per_second": 2.09,
      "step": 319124
    },
    {
      "epoch": 221.26,
      "learning_rate": 1.5574792243767314e-05,
      "loss": 3.8573,
      "step": 319500
    },
    {
      "epoch": 221.61,
      "learning_rate": 1.556786703601108e-05,
      "loss": 3.8548,
      "step": 320000
    },
    {
      "epoch": 221.95,
      "learning_rate": 1.556094182825485e-05,
      "loss": 3.8558,
      "step": 320500
    },
    {
      "epoch": 222.0,
      "eval_loss": 3.781160593032837,
      "eval_runtime": 77.0908,
      "eval_samples_per_second": 6387.583,
      "eval_steps_per_second": 2.088,
      "step": 320568
    },
    {
      "epoch": 222.3,
      "learning_rate": 1.5554016620498615e-05,
      "loss": 3.852,
      "step": 321000
    },
    {
      "epoch": 222.65,
      "learning_rate": 1.5547091412742385e-05,
      "loss": 3.8537,
      "step": 321500
    },
    {
      "epoch": 222.99,
      "learning_rate": 1.5540166204986152e-05,
      "loss": 3.8515,
      "step": 322000
    },
    {
      "epoch": 223.0,
      "eval_loss": 3.7702577114105225,
      "eval_runtime": 75.5669,
      "eval_samples_per_second": 6516.401,
      "eval_steps_per_second": 2.131,
      "step": 322012
    },
    {
      "epoch": 223.34,
      "learning_rate": 1.553324099722992e-05,
      "loss": 3.852,
      "step": 322500
    },
    {
      "epoch": 223.68,
      "learning_rate": 1.5526315789473686e-05,
      "loss": 3.8495,
      "step": 323000
    },
    {
      "epoch": 224.0,
      "eval_loss": 3.772334575653076,
      "eval_runtime": 75.2529,
      "eval_samples_per_second": 6543.584,
      "eval_steps_per_second": 2.139,
      "step": 323456
    },
    {
      "epoch": 224.03,
      "learning_rate": 1.5519390581717453e-05,
      "loss": 3.854,
      "step": 323500
    },
    {
      "epoch": 224.38,
      "learning_rate": 1.551246537396122e-05,
      "loss": 3.851,
      "step": 324000
    },
    {
      "epoch": 224.72,
      "learning_rate": 1.5505540166204986e-05,
      "loss": 3.8516,
      "step": 324500
    },
    {
      "epoch": 225.0,
      "eval_loss": 3.774179458618164,
      "eval_runtime": 75.9135,
      "eval_samples_per_second": 6486.645,
      "eval_steps_per_second": 2.121,
      "step": 324900
    },
    {
      "epoch": 225.07,
      "learning_rate": 1.5498614958448753e-05,
      "loss": 3.8517,
      "step": 325000
    },
    {
      "epoch": 225.42,
      "learning_rate": 1.5491689750692524e-05,
      "loss": 3.8512,
      "step": 325500
    },
    {
      "epoch": 225.76,
      "learning_rate": 1.548476454293629e-05,
      "loss": 3.85,
      "step": 326000
    },
    {
      "epoch": 226.0,
      "eval_loss": 3.775554656982422,
      "eval_runtime": 74.3917,
      "eval_samples_per_second": 6619.337,
      "eval_steps_per_second": 2.164,
      "step": 326344
    },
    {
      "epoch": 226.11,
      "learning_rate": 1.5477839335180058e-05,
      "loss": 3.85,
      "step": 326500
    },
    {
      "epoch": 226.45,
      "learning_rate": 1.5470914127423824e-05,
      "loss": 3.8516,
      "step": 327000
    },
    {
      "epoch": 226.8,
      "learning_rate": 1.546398891966759e-05,
      "loss": 3.8494,
      "step": 327500
    },
    {
      "epoch": 227.0,
      "eval_loss": 3.7784576416015625,
      "eval_runtime": 74.4719,
      "eval_samples_per_second": 6612.215,
      "eval_steps_per_second": 2.162,
      "step": 327788
    },
    {
      "epoch": 227.15,
      "learning_rate": 1.5457063711911358e-05,
      "loss": 3.8482,
      "step": 328000
    },
    {
      "epoch": 227.49,
      "learning_rate": 1.5450138504155125e-05,
      "loss": 3.8501,
      "step": 328500
    },
    {
      "epoch": 227.84,
      "learning_rate": 1.5443213296398892e-05,
      "loss": 3.8519,
      "step": 329000
    },
    {
      "epoch": 228.0,
      "eval_loss": 3.771754741668701,
      "eval_runtime": 74.1616,
      "eval_samples_per_second": 6639.877,
      "eval_steps_per_second": 2.171,
      "step": 329232
    },
    {
      "epoch": 228.19,
      "learning_rate": 1.5436288088642662e-05,
      "loss": 3.8472,
      "step": 329500
    },
    {
      "epoch": 228.53,
      "learning_rate": 1.5429362880886426e-05,
      "loss": 3.8508,
      "step": 330000
    },
    {
      "epoch": 228.88,
      "learning_rate": 1.5422437673130196e-05,
      "loss": 3.8471,
      "step": 330500
    },
    {
      "epoch": 229.0,
      "eval_loss": 3.7720766067504883,
      "eval_runtime": 74.897,
      "eval_samples_per_second": 6574.686,
      "eval_steps_per_second": 2.15,
      "step": 330676
    },
    {
      "epoch": 229.22,
      "learning_rate": 1.5415512465373963e-05,
      "loss": 3.8494,
      "step": 331000
    },
    {
      "epoch": 229.57,
      "learning_rate": 1.540858725761773e-05,
      "loss": 3.8497,
      "step": 331500
    },
    {
      "epoch": 229.92,
      "learning_rate": 1.5401662049861497e-05,
      "loss": 3.8468,
      "step": 332000
    },
    {
      "epoch": 230.0,
      "eval_loss": 3.7739343643188477,
      "eval_runtime": 76.8844,
      "eval_samples_per_second": 6404.731,
      "eval_steps_per_second": 2.094,
      "step": 332120
    },
    {
      "epoch": 230.26,
      "learning_rate": 1.5394736842105264e-05,
      "loss": 3.8449,
      "step": 332500
    },
    {
      "epoch": 230.61,
      "learning_rate": 1.538781163434903e-05,
      "loss": 3.8472,
      "step": 333000
    },
    {
      "epoch": 230.96,
      "learning_rate": 1.5380886426592798e-05,
      "loss": 3.8425,
      "step": 333500
    },
    {
      "epoch": 231.0,
      "eval_loss": 3.7709102630615234,
      "eval_runtime": 74.7014,
      "eval_samples_per_second": 6591.895,
      "eval_steps_per_second": 2.155,
      "step": 333564
    },
    {
      "epoch": 231.3,
      "learning_rate": 1.5373961218836565e-05,
      "loss": 3.8435,
      "step": 334000
    },
    {
      "epoch": 231.65,
      "learning_rate": 1.5367036011080335e-05,
      "loss": 3.8466,
      "step": 334500
    },
    {
      "epoch": 231.99,
      "learning_rate": 1.5360110803324102e-05,
      "loss": 3.8475,
      "step": 335000
    },
    {
      "epoch": 232.0,
      "eval_loss": 3.770082950592041,
      "eval_runtime": 75.48,
      "eval_samples_per_second": 6523.901,
      "eval_steps_per_second": 2.133,
      "step": 335008
    },
    {
      "epoch": 232.34,
      "learning_rate": 1.535318559556787e-05,
      "loss": 3.8447,
      "step": 335500
    },
    {
      "epoch": 232.69,
      "learning_rate": 1.5346260387811636e-05,
      "loss": 3.8462,
      "step": 336000
    },
    {
      "epoch": 233.0,
      "eval_loss": 3.764716148376465,
      "eval_runtime": 74.6162,
      "eval_samples_per_second": 6599.422,
      "eval_steps_per_second": 2.158,
      "step": 336452
    },
    {
      "epoch": 233.03,
      "learning_rate": 1.5339335180055403e-05,
      "loss": 3.843,
      "step": 336500
    },
    {
      "epoch": 233.38,
      "learning_rate": 1.533240997229917e-05,
      "loss": 3.8405,
      "step": 337000
    },
    {
      "epoch": 233.73,
      "learning_rate": 1.5325484764542937e-05,
      "loss": 3.8431,
      "step": 337500
    },
    {
      "epoch": 234.0,
      "eval_loss": 3.7664144039154053,
      "eval_runtime": 74.5316,
      "eval_samples_per_second": 6606.916,
      "eval_steps_per_second": 2.16,
      "step": 337896
    },
    {
      "epoch": 234.07,
      "learning_rate": 1.5318559556786703e-05,
      "loss": 3.842,
      "step": 338000
    },
    {
      "epoch": 234.42,
      "learning_rate": 1.5311634349030474e-05,
      "loss": 3.8452,
      "step": 338500
    },
    {
      "epoch": 234.76,
      "learning_rate": 1.5304709141274237e-05,
      "loss": 3.8427,
      "step": 339000
    },
    {
      "epoch": 235.0,
      "eval_loss": 3.766789197921753,
      "eval_runtime": 75.3868,
      "eval_samples_per_second": 6531.965,
      "eval_steps_per_second": 2.136,
      "step": 339340
    },
    {
      "epoch": 235.11,
      "learning_rate": 1.5297783933518008e-05,
      "loss": 3.8428,
      "step": 339500
    },
    {
      "epoch": 235.46,
      "learning_rate": 1.5290858725761775e-05,
      "loss": 3.8405,
      "step": 340000
    },
    {
      "epoch": 235.8,
      "learning_rate": 1.528393351800554e-05,
      "loss": 3.8406,
      "step": 340500
    },
    {
      "epoch": 236.0,
      "eval_loss": 3.7623260021209717,
      "eval_runtime": 73.9616,
      "eval_samples_per_second": 6657.831,
      "eval_steps_per_second": 2.177,
      "step": 340784
    },
    {
      "epoch": 236.15,
      "learning_rate": 1.527700831024931e-05,
      "loss": 3.8407,
      "step": 341000
    },
    {
      "epoch": 236.5,
      "learning_rate": 1.5270083102493075e-05,
      "loss": 3.8443,
      "step": 341500
    },
    {
      "epoch": 236.84,
      "learning_rate": 1.5263157894736846e-05,
      "loss": 3.8388,
      "step": 342000
    },
    {
      "epoch": 237.0,
      "eval_loss": 3.7657153606414795,
      "eval_runtime": 75.1446,
      "eval_samples_per_second": 6553.015,
      "eval_steps_per_second": 2.143,
      "step": 342228
    },
    {
      "epoch": 237.19,
      "learning_rate": 1.5256232686980609e-05,
      "loss": 3.8371,
      "step": 342500
    },
    {
      "epoch": 237.53,
      "learning_rate": 1.5249307479224378e-05,
      "loss": 3.8424,
      "step": 343000
    },
    {
      "epoch": 237.88,
      "learning_rate": 1.5242382271468146e-05,
      "loss": 3.838,
      "step": 343500
    },
    {
      "epoch": 238.0,
      "eval_loss": 3.764124870300293,
      "eval_runtime": 74.9352,
      "eval_samples_per_second": 6571.333,
      "eval_steps_per_second": 2.149,
      "step": 343672
    },
    {
      "epoch": 238.23,
      "learning_rate": 1.5235457063711913e-05,
      "loss": 3.8418,
      "step": 344000
    },
    {
      "epoch": 238.57,
      "learning_rate": 1.522853185595568e-05,
      "loss": 3.8407,
      "step": 344500
    },
    {
      "epoch": 238.92,
      "learning_rate": 1.5221606648199447e-05,
      "loss": 3.8386,
      "step": 345000
    },
    {
      "epoch": 239.0,
      "eval_loss": 3.763404369354248,
      "eval_runtime": 74.186,
      "eval_samples_per_second": 6637.691,
      "eval_steps_per_second": 2.17,
      "step": 345116
    },
    {
      "epoch": 239.27,
      "learning_rate": 1.5214681440443216e-05,
      "loss": 3.8396,
      "step": 345500
    },
    {
      "epoch": 239.61,
      "learning_rate": 1.5207756232686981e-05,
      "loss": 3.8361,
      "step": 346000
    },
    {
      "epoch": 239.96,
      "learning_rate": 1.520083102493075e-05,
      "loss": 3.837,
      "step": 346500
    },
    {
      "epoch": 240.0,
      "eval_loss": 3.7625999450683594,
      "eval_runtime": 75.4987,
      "eval_samples_per_second": 6522.289,
      "eval_steps_per_second": 2.132,
      "step": 346560
    },
    {
      "epoch": 240.3,
      "learning_rate": 1.5193905817174516e-05,
      "loss": 3.8349,
      "step": 347000
    },
    {
      "epoch": 240.65,
      "learning_rate": 1.5186980609418285e-05,
      "loss": 3.839,
      "step": 347500
    },
    {
      "epoch": 241.0,
      "learning_rate": 1.518005540166205e-05,
      "loss": 3.8383,
      "step": 348000
    },
    {
      "epoch": 241.0,
      "eval_loss": 3.7632668018341064,
      "eval_runtime": 75.4619,
      "eval_samples_per_second": 6525.467,
      "eval_steps_per_second": 2.134,
      "step": 348004
    },
    {
      "epoch": 241.34,
      "learning_rate": 1.5173130193905819e-05,
      "loss": 3.8367,
      "step": 348500
    },
    {
      "epoch": 241.69,
      "learning_rate": 1.5166204986149586e-05,
      "loss": 3.8363,
      "step": 349000
    },
    {
      "epoch": 242.0,
      "eval_loss": 3.763760805130005,
      "eval_runtime": 75.4912,
      "eval_samples_per_second": 6522.929,
      "eval_steps_per_second": 2.133,
      "step": 349448
    },
    {
      "epoch": 242.04,
      "learning_rate": 1.5159279778393353e-05,
      "loss": 3.837,
      "step": 349500
    },
    {
      "epoch": 242.38,
      "learning_rate": 1.515235457063712e-05,
      "loss": 3.8352,
      "step": 350000
    },
    {
      "epoch": 242.73,
      "learning_rate": 1.5145429362880888e-05,
      "loss": 3.8372,
      "step": 350500
    },
    {
      "epoch": 243.0,
      "eval_loss": 3.7610692977905273,
      "eval_runtime": 75.8328,
      "eval_samples_per_second": 6493.549,
      "eval_steps_per_second": 2.123,
      "step": 350892
    },
    {
      "epoch": 243.07,
      "learning_rate": 1.5138504155124655e-05,
      "loss": 3.8372,
      "step": 351000
    },
    {
      "epoch": 243.42,
      "learning_rate": 1.5131578947368422e-05,
      "loss": 3.8364,
      "step": 351500
    },
    {
      "epoch": 243.77,
      "learning_rate": 1.5124653739612189e-05,
      "loss": 3.8382,
      "step": 352000
    },
    {
      "epoch": 244.0,
      "eval_loss": 3.757394313812256,
      "eval_runtime": 75.6891,
      "eval_samples_per_second": 6505.877,
      "eval_steps_per_second": 2.127,
      "step": 352336
    },
    {
      "epoch": 244.11,
      "learning_rate": 1.5117728531855958e-05,
      "loss": 3.8352,
      "step": 352500
    },
    {
      "epoch": 244.46,
      "learning_rate": 1.5110803324099723e-05,
      "loss": 3.837,
      "step": 353000
    },
    {
      "epoch": 244.81,
      "learning_rate": 1.5103878116343492e-05,
      "loss": 3.836,
      "step": 353500
    },
    {
      "epoch": 245.0,
      "eval_loss": 3.7577829360961914,
      "eval_runtime": 75.2986,
      "eval_samples_per_second": 6539.613,
      "eval_steps_per_second": 2.138,
      "step": 353780
    },
    {
      "epoch": 245.15,
      "learning_rate": 1.5096952908587258e-05,
      "loss": 3.8344,
      "step": 354000
    },
    {
      "epoch": 245.5,
      "learning_rate": 1.5090027700831027e-05,
      "loss": 3.8379,
      "step": 354500
    },
    {
      "epoch": 245.84,
      "learning_rate": 1.5083102493074792e-05,
      "loss": 3.8351,
      "step": 355000
    },
    {
      "epoch": 246.0,
      "eval_loss": 3.756089687347412,
      "eval_runtime": 77.0188,
      "eval_samples_per_second": 6393.556,
      "eval_steps_per_second": 2.09,
      "step": 355224
    },
    {
      "epoch": 246.19,
      "learning_rate": 1.5076177285318561e-05,
      "loss": 3.8339,
      "step": 355500
    },
    {
      "epoch": 246.54,
      "learning_rate": 1.5069252077562328e-05,
      "loss": 3.8371,
      "step": 356000
    },
    {
      "epoch": 246.88,
      "learning_rate": 1.5062326869806096e-05,
      "loss": 3.8334,
      "step": 356500
    },
    {
      "epoch": 247.0,
      "eval_loss": 3.754768133163452,
      "eval_runtime": 75.4834,
      "eval_samples_per_second": 6523.608,
      "eval_steps_per_second": 2.133,
      "step": 356668
    },
    {
      "epoch": 247.23,
      "learning_rate": 1.5055401662049862e-05,
      "loss": 3.8314,
      "step": 357000
    },
    {
      "epoch": 247.58,
      "learning_rate": 1.504847645429363e-05,
      "loss": 3.8364,
      "step": 357500
    },
    {
      "epoch": 247.92,
      "learning_rate": 1.5041551246537399e-05,
      "loss": 3.834,
      "step": 358000
    },
    {
      "epoch": 248.0,
      "eval_loss": 3.754401922225952,
      "eval_runtime": 75.9236,
      "eval_samples_per_second": 6485.781,
      "eval_steps_per_second": 2.121,
      "step": 358112
    },
    {
      "epoch": 248.27,
      "learning_rate": 1.5034626038781164e-05,
      "loss": 3.8326,
      "step": 358500
    },
    {
      "epoch": 248.61,
      "learning_rate": 1.5027700831024931e-05,
      "loss": 3.8333,
      "step": 359000
    },
    {
      "epoch": 248.96,
      "learning_rate": 1.50207756232687e-05,
      "loss": 3.8347,
      "step": 359500
    },
    {
      "epoch": 249.0,
      "eval_loss": 3.757708787918091,
      "eval_runtime": 74.9192,
      "eval_samples_per_second": 6572.738,
      "eval_steps_per_second": 2.149,
      "step": 359556
    },
    {
      "epoch": 249.31,
      "learning_rate": 1.5013850415512468e-05,
      "loss": 3.8326,
      "step": 360000
    },
    {
      "epoch": 249.65,
      "learning_rate": 1.5006925207756233e-05,
      "loss": 3.8316,
      "step": 360500
    },
    {
      "epoch": 250.0,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 3.8321,
      "step": 361000
    },
    {
      "epoch": 250.0,
      "eval_loss": 3.751624584197998,
      "eval_runtime": 75.2355,
      "eval_samples_per_second": 6545.106,
      "eval_steps_per_second": 2.14,
      "step": 361000
    },
    {
      "epoch": 250.35,
      "learning_rate": 1.4993074792243769e-05,
      "loss": 3.8329,
      "step": 361500
    },
    {
      "epoch": 250.69,
      "learning_rate": 1.4986149584487534e-05,
      "loss": 3.8303,
      "step": 362000
    },
    {
      "epoch": 251.0,
      "eval_loss": 3.757580280303955,
      "eval_runtime": 76.9563,
      "eval_samples_per_second": 6398.75,
      "eval_steps_per_second": 2.092,
      "step": 362444
    },
    {
      "epoch": 251.04,
      "learning_rate": 1.4979224376731303e-05,
      "loss": 3.8295,
      "step": 362500
    },
    {
      "epoch": 251.39,
      "learning_rate": 1.4972299168975071e-05,
      "loss": 3.8314,
      "step": 363000
    },
    {
      "epoch": 251.73,
      "learning_rate": 1.4965373961218838e-05,
      "loss": 3.8307,
      "step": 363500
    },
    {
      "epoch": 252.0,
      "eval_loss": 3.752763032913208,
      "eval_runtime": 75.1767,
      "eval_samples_per_second": 6550.218,
      "eval_steps_per_second": 2.142,
      "step": 363888
    },
    {
      "epoch": 252.08,
      "learning_rate": 1.4958448753462604e-05,
      "loss": 3.8269,
      "step": 364000
    },
    {
      "epoch": 252.42,
      "learning_rate": 1.4951523545706372e-05,
      "loss": 3.8307,
      "step": 364500
    },
    {
      "epoch": 252.77,
      "learning_rate": 1.494459833795014e-05,
      "loss": 3.8296,
      "step": 365000
    },
    {
      "epoch": 253.0,
      "eval_loss": 3.7549960613250732,
      "eval_runtime": 75.8092,
      "eval_samples_per_second": 6495.567,
      "eval_steps_per_second": 2.124,
      "step": 365332
    },
    {
      "epoch": 253.12,
      "learning_rate": 1.4937673130193906e-05,
      "loss": 3.826,
      "step": 365500
    },
    {
      "epoch": 253.46,
      "learning_rate": 1.4930747922437675e-05,
      "loss": 3.8322,
      "step": 366000
    },
    {
      "epoch": 253.81,
      "learning_rate": 1.4923822714681442e-05,
      "loss": 3.8307,
      "step": 366500
    },
    {
      "epoch": 254.0,
      "eval_loss": 3.753600597381592,
      "eval_runtime": 75.0828,
      "eval_samples_per_second": 6558.417,
      "eval_steps_per_second": 2.144,
      "step": 366776
    },
    {
      "epoch": 254.16,
      "learning_rate": 1.491689750692521e-05,
      "loss": 3.8309,
      "step": 367000
    },
    {
      "epoch": 254.5,
      "learning_rate": 1.4909972299168975e-05,
      "loss": 3.8289,
      "step": 367500
    },
    {
      "epoch": 254.85,
      "learning_rate": 1.4903047091412744e-05,
      "loss": 3.8317,
      "step": 368000
    },
    {
      "epoch": 255.0,
      "eval_loss": 3.7492051124572754,
      "eval_runtime": 74.5037,
      "eval_samples_per_second": 6609.385,
      "eval_steps_per_second": 2.161,
      "step": 368220
    },
    {
      "epoch": 255.19,
      "learning_rate": 1.4896121883656511e-05,
      "loss": 3.8291,
      "step": 368500
    },
    {
      "epoch": 255.54,
      "learning_rate": 1.488919667590028e-05,
      "loss": 3.83,
      "step": 369000
    },
    {
      "epoch": 255.89,
      "learning_rate": 1.4882271468144045e-05,
      "loss": 3.8256,
      "step": 369500
    },
    {
      "epoch": 256.0,
      "eval_loss": 3.752340316772461,
      "eval_runtime": 75.106,
      "eval_samples_per_second": 6556.387,
      "eval_steps_per_second": 2.144,
      "step": 369664
    },
    {
      "epoch": 256.23,
      "learning_rate": 1.4875346260387813e-05,
      "loss": 3.8289,
      "step": 370000
    },
    {
      "epoch": 256.58,
      "learning_rate": 1.486842105263158e-05,
      "loss": 3.8261,
      "step": 370500
    },
    {
      "epoch": 256.93,
      "learning_rate": 1.4861495844875347e-05,
      "loss": 3.8254,
      "step": 371000
    },
    {
      "epoch": 257.0,
      "eval_loss": 3.7502706050872803,
      "eval_runtime": 74.7416,
      "eval_samples_per_second": 6588.351,
      "eval_steps_per_second": 2.154,
      "step": 371108
    },
    {
      "epoch": 257.27,
      "learning_rate": 1.4854570637119114e-05,
      "loss": 3.827,
      "step": 371500
    },
    {
      "epoch": 257.62,
      "learning_rate": 1.4847645429362883e-05,
      "loss": 3.8269,
      "step": 372000
    },
    {
      "epoch": 257.96,
      "learning_rate": 1.484072022160665e-05,
      "loss": 3.8262,
      "step": 372500
    },
    {
      "epoch": 258.0,
      "eval_loss": 3.7514851093292236,
      "eval_runtime": 76.1643,
      "eval_samples_per_second": 6465.283,
      "eval_steps_per_second": 2.114,
      "step": 372552
    },
    {
      "epoch": 258.31,
      "learning_rate": 1.4833795013850417e-05,
      "loss": 3.8283,
      "step": 373000
    },
    {
      "epoch": 258.66,
      "learning_rate": 1.4826869806094184e-05,
      "loss": 3.8252,
      "step": 373500
    },
    {
      "epoch": 259.0,
      "eval_loss": 3.746579647064209,
      "eval_runtime": 73.8659,
      "eval_samples_per_second": 6666.464,
      "eval_steps_per_second": 2.18,
      "step": 373996
    },
    {
      "epoch": 259.0,
      "learning_rate": 1.4819944598337952e-05,
      "loss": 3.8253,
      "step": 374000
    },
    {
      "epoch": 259.35,
      "learning_rate": 1.4813019390581717e-05,
      "loss": 3.823,
      "step": 374500
    },
    {
      "epoch": 259.7,
      "learning_rate": 1.4806094182825486e-05,
      "loss": 3.824,
      "step": 375000
    },
    {
      "epoch": 260.0,
      "eval_loss": 3.7484304904937744,
      "eval_runtime": 75.3537,
      "eval_samples_per_second": 6534.833,
      "eval_steps_per_second": 2.137,
      "step": 375440
    },
    {
      "epoch": 260.04,
      "learning_rate": 1.4799168975069253e-05,
      "loss": 3.8241,
      "step": 375500
    },
    {
      "epoch": 260.39,
      "learning_rate": 1.4792243767313022e-05,
      "loss": 3.8223,
      "step": 376000
    },
    {
      "epoch": 260.73,
      "learning_rate": 1.4785318559556787e-05,
      "loss": 3.8229,
      "step": 376500
    },
    {
      "epoch": 261.0,
      "eval_loss": 3.7492823600769043,
      "eval_runtime": 77.039,
      "eval_samples_per_second": 6391.882,
      "eval_steps_per_second": 2.09,
      "step": 376884
    },
    {
      "epoch": 261.08,
      "learning_rate": 1.4778393351800555e-05,
      "loss": 3.8235,
      "step": 377000
    },
    {
      "epoch": 261.43,
      "learning_rate": 1.4771468144044322e-05,
      "loss": 3.8235,
      "step": 377500
    },
    {
      "epoch": 261.77,
      "learning_rate": 1.476454293628809e-05,
      "loss": 3.8242,
      "step": 378000
    },
    {
      "epoch": 262.0,
      "eval_loss": 3.7440624237060547,
      "eval_runtime": 74.2561,
      "eval_samples_per_second": 6631.428,
      "eval_steps_per_second": 2.168,
      "step": 378328
    },
    {
      "epoch": 262.12,
      "learning_rate": 1.4757617728531856e-05,
      "loss": 3.8232,
      "step": 378500
    },
    {
      "epoch": 262.47,
      "learning_rate": 1.4750692520775625e-05,
      "loss": 3.8234,
      "step": 379000
    },
    {
      "epoch": 262.81,
      "learning_rate": 1.4743767313019393e-05,
      "loss": 3.8215,
      "step": 379500
    },
    {
      "epoch": 263.0,
      "eval_loss": 3.746445894241333,
      "eval_runtime": 74.2073,
      "eval_samples_per_second": 6635.79,
      "eval_steps_per_second": 2.17,
      "step": 379772
    },
    {
      "epoch": 263.16,
      "learning_rate": 1.4736842105263159e-05,
      "loss": 3.8221,
      "step": 380000
    },
    {
      "epoch": 263.5,
      "learning_rate": 1.4729916897506925e-05,
      "loss": 3.823,
      "step": 380500
    },
    {
      "epoch": 263.85,
      "learning_rate": 1.4722991689750694e-05,
      "loss": 3.8222,
      "step": 381000
    },
    {
      "epoch": 264.0,
      "eval_loss": 3.747673749923706,
      "eval_runtime": 76.0947,
      "eval_samples_per_second": 6471.201,
      "eval_steps_per_second": 2.116,
      "step": 381216
    },
    {
      "epoch": 264.2,
      "learning_rate": 1.4716066481994463e-05,
      "loss": 3.8171,
      "step": 381500
    },
    {
      "epoch": 264.54,
      "learning_rate": 1.4709141274238228e-05,
      "loss": 3.8187,
      "step": 382000
    },
    {
      "epoch": 264.89,
      "learning_rate": 1.4702216066481997e-05,
      "loss": 3.823,
      "step": 382500
    },
    {
      "epoch": 265.0,
      "eval_loss": 3.7447397708892822,
      "eval_runtime": 74.1023,
      "eval_samples_per_second": 6645.191,
      "eval_steps_per_second": 2.173,
      "step": 382660
    },
    {
      "epoch": 265.24,
      "learning_rate": 1.4695290858725763e-05,
      "loss": 3.8209,
      "step": 383000
    },
    {
      "epoch": 265.58,
      "learning_rate": 1.4688365650969529e-05,
      "loss": 3.8206,
      "step": 383500
    },
    {
      "epoch": 265.93,
      "learning_rate": 1.4681440443213297e-05,
      "loss": 3.8222,
      "step": 384000
    },
    {
      "epoch": 266.0,
      "eval_loss": 3.7423489093780518,
      "eval_runtime": 75.7006,
      "eval_samples_per_second": 6504.891,
      "eval_steps_per_second": 2.127,
      "step": 384104
    },
    {
      "epoch": 266.27,
      "learning_rate": 1.4674515235457066e-05,
      "loss": 3.8173,
      "step": 384500
    },
    {
      "epoch": 266.62,
      "learning_rate": 1.4667590027700833e-05,
      "loss": 3.8216,
      "step": 385000
    },
    {
      "epoch": 266.97,
      "learning_rate": 1.4660664819944598e-05,
      "loss": 3.8187,
      "step": 385500
    },
    {
      "epoch": 267.0,
      "eval_loss": 3.7439002990722656,
      "eval_runtime": 75.5321,
      "eval_samples_per_second": 6519.399,
      "eval_steps_per_second": 2.132,
      "step": 385548
    },
    {
      "epoch": 267.31,
      "learning_rate": 1.4653739612188367e-05,
      "loss": 3.8174,
      "step": 386000
    },
    {
      "epoch": 267.66,
      "learning_rate": 1.4646814404432135e-05,
      "loss": 3.8213,
      "step": 386500
    },
    {
      "epoch": 268.0,
      "eval_loss": 3.740312099456787,
      "eval_runtime": 73.9502,
      "eval_samples_per_second": 6658.856,
      "eval_steps_per_second": 2.177,
      "step": 386992
    },
    {
      "epoch": 268.01,
      "learning_rate": 1.46398891966759e-05,
      "loss": 3.8172,
      "step": 387000
    },
    {
      "epoch": 268.35,
      "learning_rate": 1.4632963988919669e-05,
      "loss": 3.8195,
      "step": 387500
    },
    {
      "epoch": 268.7,
      "learning_rate": 1.4626038781163436e-05,
      "loss": 3.8181,
      "step": 388000
    },
    {
      "epoch": 269.0,
      "eval_loss": 3.7468435764312744,
      "eval_runtime": 75.0402,
      "eval_samples_per_second": 6562.133,
      "eval_steps_per_second": 2.146,
      "step": 388436
    },
    {
      "epoch": 269.04,
      "learning_rate": 1.4619113573407205e-05,
      "loss": 3.8183,
      "step": 388500
    },
    {
      "epoch": 269.39,
      "learning_rate": 1.461218836565097e-05,
      "loss": 3.8177,
      "step": 389000
    },
    {
      "epoch": 269.74,
      "learning_rate": 1.4605263157894739e-05,
      "loss": 3.819,
      "step": 389500
    },
    {
      "epoch": 270.0,
      "eval_loss": 3.7409017086029053,
      "eval_runtime": 74.5753,
      "eval_samples_per_second": 6603.048,
      "eval_steps_per_second": 2.159,
      "step": 389880
    },
    {
      "epoch": 270.08,
      "learning_rate": 1.4598337950138505e-05,
      "loss": 3.8169,
      "step": 390000
    },
    {
      "epoch": 270.43,
      "learning_rate": 1.4591412742382272e-05,
      "loss": 3.8142,
      "step": 390500
    },
    {
      "epoch": 270.78,
      "learning_rate": 1.458448753462604e-05,
      "loss": 3.819,
      "step": 391000
    },
    {
      "epoch": 271.0,
      "eval_loss": 3.742891550064087,
      "eval_runtime": 74.4741,
      "eval_samples_per_second": 6612.013,
      "eval_steps_per_second": 2.162,
      "step": 391324
    },
    {
      "epoch": 271.12,
      "learning_rate": 1.4577562326869808e-05,
      "loss": 3.8146,
      "step": 391500
    },
    {
      "epoch": 271.47,
      "learning_rate": 1.4570637119113575e-05,
      "loss": 3.8197,
      "step": 392000
    },
    {
      "epoch": 271.81,
      "learning_rate": 1.4563711911357342e-05,
      "loss": 3.8158,
      "step": 392500
    },
    {
      "epoch": 272.0,
      "eval_loss": 3.7434399127960205,
      "eval_runtime": 74.3951,
      "eval_samples_per_second": 6619.035,
      "eval_steps_per_second": 2.164,
      "step": 392768
    },
    {
      "epoch": 272.16,
      "learning_rate": 1.4556786703601109e-05,
      "loss": 3.8153,
      "step": 393000
    },
    {
      "epoch": 272.51,
      "learning_rate": 1.4549861495844877e-05,
      "loss": 3.8146,
      "step": 393500
    },
    {
      "epoch": 272.85,
      "learning_rate": 1.4542936288088644e-05,
      "loss": 3.816,
      "step": 394000
    },
    {
      "epoch": 273.0,
      "eval_loss": 3.738966464996338,
      "eval_runtime": 74.8507,
      "eval_samples_per_second": 6578.747,
      "eval_steps_per_second": 2.151,
      "step": 394212
    },
    {
      "epoch": 273.2,
      "learning_rate": 1.4536011080332411e-05,
      "loss": 3.8154,
      "step": 394500
    },
    {
      "epoch": 273.55,
      "learning_rate": 1.4529085872576178e-05,
      "loss": 3.8152,
      "step": 395000
    },
    {
      "epoch": 273.89,
      "learning_rate": 1.4522160664819947e-05,
      "loss": 3.8133,
      "step": 395500
    },
    {
      "epoch": 274.0,
      "eval_loss": 3.7411460876464844,
      "eval_runtime": 73.8647,
      "eval_samples_per_second": 6666.567,
      "eval_steps_per_second": 2.18,
      "step": 395656
    },
    {
      "epoch": 274.24,
      "learning_rate": 1.4515235457063712e-05,
      "loss": 3.8161,
      "step": 396000
    },
    {
      "epoch": 274.58,
      "learning_rate": 1.450831024930748e-05,
      "loss": 3.8136,
      "step": 396500
    },
    {
      "epoch": 274.93,
      "learning_rate": 1.4501385041551247e-05,
      "loss": 3.8154,
      "step": 397000
    },
    {
      "epoch": 275.0,
      "eval_loss": 3.7406582832336426,
      "eval_runtime": 73.3111,
      "eval_samples_per_second": 6716.909,
      "eval_steps_per_second": 2.196,
      "step": 397100
    },
    {
      "epoch": 275.28,
      "learning_rate": 1.4494459833795016e-05,
      "loss": 3.8141,
      "step": 397500
    },
    {
      "epoch": 275.62,
      "learning_rate": 1.4487534626038781e-05,
      "loss": 3.8111,
      "step": 398000
    },
    {
      "epoch": 275.97,
      "learning_rate": 1.448060941828255e-05,
      "loss": 3.8152,
      "step": 398500
    },
    {
      "epoch": 276.0,
      "eval_loss": 3.741816759109497,
      "eval_runtime": 73.4966,
      "eval_samples_per_second": 6699.952,
      "eval_steps_per_second": 2.191,
      "step": 398544
    },
    {
      "epoch": 276.32,
      "learning_rate": 1.4473684210526317e-05,
      "loss": 3.8147,
      "step": 399000
    },
    {
      "epoch": 276.66,
      "learning_rate": 1.4466759002770084e-05,
      "loss": 3.8135,
      "step": 399500
    },
    {
      "epoch": 277.0,
      "eval_loss": 3.738856315612793,
      "eval_runtime": 75.384,
      "eval_samples_per_second": 6532.21,
      "eval_steps_per_second": 2.136,
      "step": 399988
    },
    {
      "epoch": 277.01,
      "learning_rate": 1.445983379501385e-05,
      "loss": 3.8129,
      "step": 400000
    },
    {
      "epoch": 277.35,
      "learning_rate": 1.445290858725762e-05,
      "loss": 3.8126,
      "step": 400500
    },
    {
      "epoch": 277.7,
      "learning_rate": 1.4445983379501388e-05,
      "loss": 3.8137,
      "step": 401000
    },
    {
      "epoch": 278.0,
      "eval_loss": 3.7304413318634033,
      "eval_runtime": 76.1846,
      "eval_samples_per_second": 6463.564,
      "eval_steps_per_second": 2.113,
      "step": 401432
    },
    {
      "epoch": 278.05,
      "learning_rate": 1.4439058171745153e-05,
      "loss": 3.8111,
      "step": 401500
    },
    {
      "epoch": 278.39,
      "learning_rate": 1.443213296398892e-05,
      "loss": 3.8118,
      "step": 402000
    },
    {
      "epoch": 278.74,
      "learning_rate": 1.4425207756232689e-05,
      "loss": 3.8113,
      "step": 402500
    },
    {
      "epoch": 279.0,
      "eval_loss": 3.735074520111084,
      "eval_runtime": 75.5234,
      "eval_samples_per_second": 6520.156,
      "eval_steps_per_second": 2.132,
      "step": 402876
    },
    {
      "epoch": 279.09,
      "learning_rate": 1.4418282548476454e-05,
      "loss": 3.8114,
      "step": 403000
    },
    {
      "epoch": 279.43,
      "learning_rate": 1.4411357340720222e-05,
      "loss": 3.8118,
      "step": 403500
    },
    {
      "epoch": 279.78,
      "learning_rate": 1.4404432132963991e-05,
      "loss": 3.8099,
      "step": 404000
    },
    {
      "epoch": 280.0,
      "eval_loss": 3.735912799835205,
      "eval_runtime": 75.725,
      "eval_samples_per_second": 6502.791,
      "eval_steps_per_second": 2.126,
      "step": 404320
    },
    {
      "epoch": 280.12,
      "learning_rate": 1.4397506925207758e-05,
      "loss": 3.81,
      "step": 404500
    },
    {
      "epoch": 280.47,
      "learning_rate": 1.4390581717451523e-05,
      "loss": 3.8137,
      "step": 405000
    },
    {
      "epoch": 280.82,
      "learning_rate": 1.4383656509695292e-05,
      "loss": 3.808,
      "step": 405500
    },
    {
      "epoch": 281.0,
      "eval_loss": 3.7392752170562744,
      "eval_runtime": 74.1732,
      "eval_samples_per_second": 6638.844,
      "eval_steps_per_second": 2.171,
      "step": 405764
    },
    {
      "epoch": 281.16,
      "learning_rate": 1.437673130193906e-05,
      "loss": 3.81,
      "step": 406000
    },
    {
      "epoch": 281.51,
      "learning_rate": 1.4369806094182827e-05,
      "loss": 3.8114,
      "step": 406500
    },
    {
      "epoch": 281.86,
      "learning_rate": 1.4362880886426594e-05,
      "loss": 3.8103,
      "step": 407000
    },
    {
      "epoch": 282.0,
      "eval_loss": 3.7333552837371826,
      "eval_runtime": 75.1438,
      "eval_samples_per_second": 6553.09,
      "eval_steps_per_second": 2.143,
      "step": 407208
    },
    {
      "epoch": 282.2,
      "learning_rate": 1.4355955678670361e-05,
      "loss": 3.8101,
      "step": 407500
    },
    {
      "epoch": 282.55,
      "learning_rate": 1.434903047091413e-05,
      "loss": 3.8119,
      "step": 408000
    },
    {
      "epoch": 282.89,
      "learning_rate": 1.4342105263157895e-05,
      "loss": 3.8108,
      "step": 408500
    },
    {
      "epoch": 283.0,
      "eval_loss": 3.735065221786499,
      "eval_runtime": 73.2315,
      "eval_samples_per_second": 6724.206,
      "eval_steps_per_second": 2.199,
      "step": 408652
    },
    {
      "epoch": 283.24,
      "learning_rate": 1.4335180055401664e-05,
      "loss": 3.8071,
      "step": 409000
    },
    {
      "epoch": 283.59,
      "learning_rate": 1.432825484764543e-05,
      "loss": 3.8104,
      "step": 409500
    },
    {
      "epoch": 283.93,
      "learning_rate": 1.4321329639889199e-05,
      "loss": 3.8104,
      "step": 410000
    },
    {
      "epoch": 284.0,
      "eval_loss": 3.7356646060943604,
      "eval_runtime": 75.7247,
      "eval_samples_per_second": 6502.821,
      "eval_steps_per_second": 2.126,
      "step": 410096
    },
    {
      "epoch": 284.28,
      "learning_rate": 1.4314404432132964e-05,
      "loss": 3.8049,
      "step": 410500
    },
    {
      "epoch": 284.63,
      "learning_rate": 1.4307479224376733e-05,
      "loss": 3.8094,
      "step": 411000
    },
    {
      "epoch": 284.97,
      "learning_rate": 1.43005540166205e-05,
      "loss": 3.807,
      "step": 411500
    },
    {
      "epoch": 285.0,
      "eval_loss": 3.7316856384277344,
      "eval_runtime": 75.2405,
      "eval_samples_per_second": 6544.666,
      "eval_steps_per_second": 2.14,
      "step": 411540
    },
    {
      "epoch": 285.32,
      "learning_rate": 1.4293628808864267e-05,
      "loss": 3.8078,
      "step": 412000
    },
    {
      "epoch": 285.66,
      "learning_rate": 1.4286703601108034e-05,
      "loss": 3.806,
      "step": 412500
    },
    {
      "epoch": 286.0,
      "eval_loss": 3.731189250946045,
      "eval_runtime": 73.5011,
      "eval_samples_per_second": 6699.549,
      "eval_steps_per_second": 2.19,
      "step": 412984
    },
    {
      "epoch": 286.01,
      "learning_rate": 1.4279778393351802e-05,
      "loss": 3.8083,
      "step": 413000
    },
    {
      "epoch": 286.36,
      "learning_rate": 1.427285318559557e-05,
      "loss": 3.8082,
      "step": 413500
    },
    {
      "epoch": 286.7,
      "learning_rate": 1.4265927977839336e-05,
      "loss": 3.8085,
      "step": 414000
    },
    {
      "epoch": 287.0,
      "eval_loss": 3.727071523666382,
      "eval_runtime": 75.2005,
      "eval_samples_per_second": 6548.149,
      "eval_steps_per_second": 2.141,
      "step": 414428
    },
    {
      "epoch": 287.05,
      "learning_rate": 1.4259002770083103e-05,
      "loss": 3.8035,
      "step": 414500
    },
    {
      "epoch": 287.4,
      "learning_rate": 1.4252077562326872e-05,
      "loss": 3.8054,
      "step": 415000
    },
    {
      "epoch": 287.74,
      "learning_rate": 1.4245152354570639e-05,
      "loss": 3.8043,
      "step": 415500
    },
    {
      "epoch": 288.0,
      "eval_loss": 3.7320947647094727,
      "eval_runtime": 74.0311,
      "eval_samples_per_second": 6651.583,
      "eval_steps_per_second": 2.175,
      "step": 415872
    },
    {
      "epoch": 288.09,
      "learning_rate": 1.4238227146814406e-05,
      "loss": 3.8056,
      "step": 416000
    },
    {
      "epoch": 288.43,
      "learning_rate": 1.4231301939058172e-05,
      "loss": 3.8044,
      "step": 416500
    },
    {
      "epoch": 288.78,
      "learning_rate": 1.4224376731301941e-05,
      "loss": 3.8051,
      "step": 417000
    },
    {
      "epoch": 289.0,
      "eval_loss": 3.726090431213379,
      "eval_runtime": 75.4663,
      "eval_samples_per_second": 6525.085,
      "eval_steps_per_second": 2.133,
      "step": 417316
    },
    {
      "epoch": 289.13,
      "learning_rate": 1.4217451523545706e-05,
      "loss": 3.8017,
      "step": 417500
    },
    {
      "epoch": 289.47,
      "learning_rate": 1.4210526315789475e-05,
      "loss": 3.8071,
      "step": 418000
    },
    {
      "epoch": 289.82,
      "learning_rate": 1.4203601108033242e-05,
      "loss": 3.8058,
      "step": 418500
    },
    {
      "epoch": 290.0,
      "eval_loss": 3.730849027633667,
      "eval_runtime": 75.3792,
      "eval_samples_per_second": 6532.621,
      "eval_steps_per_second": 2.136,
      "step": 418760
    },
    {
      "epoch": 290.17,
      "learning_rate": 1.419667590027701e-05,
      "loss": 3.8073,
      "step": 419000
    },
    {
      "epoch": 290.51,
      "learning_rate": 1.4189750692520776e-05,
      "loss": 3.8049,
      "step": 419500
    },
    {
      "epoch": 290.86,
      "learning_rate": 1.4182825484764544e-05,
      "loss": 3.8024,
      "step": 420000
    },
    {
      "epoch": 291.0,
      "eval_loss": 3.727689027786255,
      "eval_runtime": 77.757,
      "eval_samples_per_second": 6332.854,
      "eval_steps_per_second": 2.071,
      "step": 420204
    },
    {
      "epoch": 291.2,
      "learning_rate": 1.4175900277008311e-05,
      "loss": 3.803,
      "step": 420500
    },
    {
      "epoch": 291.55,
      "learning_rate": 1.4168975069252078e-05,
      "loss": 3.7999,
      "step": 421000
    },
    {
      "epoch": 291.9,
      "learning_rate": 1.4162049861495845e-05,
      "loss": 3.8047,
      "step": 421500
    },
    {
      "epoch": 292.0,
      "eval_loss": 3.7302093505859375,
      "eval_runtime": 78.9126,
      "eval_samples_per_second": 6240.118,
      "eval_steps_per_second": 2.04,
      "step": 421648
    },
    {
      "epoch": 292.24,
      "learning_rate": 1.4155124653739614e-05,
      "loss": 3.8048,
      "step": 422000
    },
    {
      "epoch": 292.59,
      "learning_rate": 1.4148199445983382e-05,
      "loss": 3.8016,
      "step": 422500
    },
    {
      "epoch": 292.94,
      "learning_rate": 1.4141274238227148e-05,
      "loss": 3.806,
      "step": 423000
    },
    {
      "epoch": 293.0,
      "eval_loss": 3.727792501449585,
      "eval_runtime": 73.701,
      "eval_samples_per_second": 6681.377,
      "eval_steps_per_second": 2.185,
      "step": 423092
    },
    {
      "epoch": 293.28,
      "learning_rate": 1.4134349030470914e-05,
      "loss": 3.8033,
      "step": 423500
    },
    {
      "epoch": 293.63,
      "learning_rate": 1.4127423822714683e-05,
      "loss": 3.8018,
      "step": 424000
    },
    {
      "epoch": 293.98,
      "learning_rate": 1.4120498614958448e-05,
      "loss": 3.8034,
      "step": 424500
    },
    {
      "epoch": 294.0,
      "eval_loss": 3.7282450199127197,
      "eval_runtime": 73.3047,
      "eval_samples_per_second": 6717.497,
      "eval_steps_per_second": 2.196,
      "step": 424536
    },
    {
      "epoch": 294.32,
      "learning_rate": 1.4113573407202217e-05,
      "loss": 3.7997,
      "step": 425000
    },
    {
      "epoch": 294.67,
      "learning_rate": 1.4106648199445986e-05,
      "loss": 3.8002,
      "step": 425500
    },
    {
      "epoch": 295.0,
      "eval_loss": 3.72798490524292,
      "eval_runtime": 75.6916,
      "eval_samples_per_second": 6505.662,
      "eval_steps_per_second": 2.127,
      "step": 425980
    },
    {
      "epoch": 295.01,
      "learning_rate": 1.4099722991689752e-05,
      "loss": 3.8046,
      "step": 426000
    },
    {
      "epoch": 295.36,
      "learning_rate": 1.4092797783933518e-05,
      "loss": 3.7991,
      "step": 426500
    },
    {
      "epoch": 295.71,
      "learning_rate": 1.4085872576177286e-05,
      "loss": 3.8009,
      "step": 427000
    },
    {
      "epoch": 296.0,
      "eval_loss": 3.7270190715789795,
      "eval_runtime": 75.9779,
      "eval_samples_per_second": 6481.15,
      "eval_steps_per_second": 2.119,
      "step": 427424
    },
    {
      "epoch": 296.05,
      "learning_rate": 1.4078947368421055e-05,
      "loss": 3.8023,
      "step": 427500
    },
    {
      "epoch": 296.4,
      "learning_rate": 1.4072022160664822e-05,
      "loss": 3.799,
      "step": 428000
    },
    {
      "epoch": 296.75,
      "learning_rate": 1.4065096952908589e-05,
      "loss": 3.8004,
      "step": 428500
    },
    {
      "epoch": 297.0,
      "eval_loss": 3.725785255432129,
      "eval_runtime": 73.3442,
      "eval_samples_per_second": 6713.882,
      "eval_steps_per_second": 2.195,
      "step": 428868
    },
    {
      "epoch": 297.09,
      "learning_rate": 1.4058171745152356e-05,
      "loss": 3.798,
      "step": 429000
    },
    {
      "epoch": 297.44,
      "learning_rate": 1.4051246537396124e-05,
      "loss": 3.8009,
      "step": 429500
    },
    {
      "epoch": 297.78,
      "learning_rate": 1.404432132963989e-05,
      "loss": 3.8012,
      "step": 430000
    },
    {
      "epoch": 298.0,
      "eval_loss": 3.7256500720977783,
      "eval_runtime": 73.7366,
      "eval_samples_per_second": 6678.15,
      "eval_steps_per_second": 2.183,
      "step": 430312
    },
    {
      "epoch": 298.13,
      "learning_rate": 1.4037396121883658e-05,
      "loss": 3.7996,
      "step": 430500
    },
    {
      "epoch": 298.48,
      "learning_rate": 1.4030470914127425e-05,
      "loss": 3.8001,
      "step": 431000
    },
    {
      "epoch": 298.82,
      "learning_rate": 1.4023545706371194e-05,
      "loss": 3.8005,
      "step": 431500
    },
    {
      "epoch": 299.0,
      "eval_loss": 3.723066568374634,
      "eval_runtime": 77.3225,
      "eval_samples_per_second": 6368.445,
      "eval_steps_per_second": 2.082,
      "step": 431756
    },
    {
      "epoch": 299.17,
      "learning_rate": 1.4016620498614959e-05,
      "loss": 3.7999,
      "step": 432000
    },
    {
      "epoch": 299.52,
      "learning_rate": 1.4009695290858727e-05,
      "loss": 3.7986,
      "step": 432500
    },
    {
      "epoch": 299.86,
      "learning_rate": 1.4002770083102494e-05,
      "loss": 3.7992,
      "step": 433000
    },
    {
      "epoch": 300.0,
      "eval_loss": 3.7242555618286133,
      "eval_runtime": 76.7477,
      "eval_samples_per_second": 6416.137,
      "eval_steps_per_second": 2.098,
      "step": 433200
    },
    {
      "epoch": 300.21,
      "learning_rate": 1.3995844875346261e-05,
      "loss": 3.799,
      "step": 433500
    },
    {
      "epoch": 300.55,
      "learning_rate": 1.3988919667590028e-05,
      "loss": 3.7998,
      "step": 434000
    },
    {
      "epoch": 300.9,
      "learning_rate": 1.3981994459833797e-05,
      "loss": 3.7975,
      "step": 434500
    },
    {
      "epoch": 301.0,
      "eval_loss": 3.7223165035247803,
      "eval_runtime": 73.7218,
      "eval_samples_per_second": 6679.489,
      "eval_steps_per_second": 2.184,
      "step": 434644
    },
    {
      "epoch": 301.25,
      "learning_rate": 1.3975069252077564e-05,
      "loss": 3.8001,
      "step": 435000
    },
    {
      "epoch": 301.59,
      "learning_rate": 1.396814404432133e-05,
      "loss": 3.7979,
      "step": 435500
    },
    {
      "epoch": 301.94,
      "learning_rate": 1.3961218836565098e-05,
      "loss": 3.7973,
      "step": 436000
    },
    {
      "epoch": 302.0,
      "eval_loss": 3.724653959274292,
      "eval_runtime": 77.8047,
      "eval_samples_per_second": 6328.977,
      "eval_steps_per_second": 2.069,
      "step": 436088
    },
    {
      "epoch": 302.29,
      "learning_rate": 1.3954293628808866e-05,
      "loss": 3.7957,
      "step": 436500
    },
    {
      "epoch": 302.63,
      "learning_rate": 1.3947368421052631e-05,
      "loss": 3.7942,
      "step": 437000
    },
    {
      "epoch": 302.98,
      "learning_rate": 1.39404432132964e-05,
      "loss": 3.7987,
      "step": 437500
    },
    {
      "epoch": 303.0,
      "eval_loss": 3.720510244369507,
      "eval_runtime": 76.9622,
      "eval_samples_per_second": 6398.255,
      "eval_steps_per_second": 2.092,
      "step": 437532
    },
    {
      "epoch": 303.32,
      "learning_rate": 1.3933518005540167e-05,
      "loss": 3.795,
      "step": 438000
    },
    {
      "epoch": 303.67,
      "learning_rate": 1.3926592797783936e-05,
      "loss": 3.7965,
      "step": 438500
    },
    {
      "epoch": 304.0,
      "eval_loss": 3.7249743938446045,
      "eval_runtime": 73.4932,
      "eval_samples_per_second": 6700.264,
      "eval_steps_per_second": 2.191,
      "step": 438976
    },
    {
      "epoch": 304.02,
      "learning_rate": 1.39196675900277e-05,
      "loss": 3.796,
      "step": 439000
    },
    {
      "epoch": 304.36,
      "learning_rate": 1.391274238227147e-05,
      "loss": 3.7958,
      "step": 439500
    },
    {
      "epoch": 304.71,
      "learning_rate": 1.3905817174515236e-05,
      "loss": 3.7992,
      "step": 440000
    },
    {
      "epoch": 305.0,
      "eval_loss": 3.721456289291382,
      "eval_runtime": 80.3944,
      "eval_samples_per_second": 6125.106,
      "eval_steps_per_second": 2.003,
      "step": 440420
    },
    {
      "epoch": 305.06,
      "learning_rate": 1.3898891966759005e-05,
      "loss": 3.7968,
      "step": 440500
    },
    {
      "epoch": 305.4,
      "learning_rate": 1.389196675900277e-05,
      "loss": 3.7966,
      "step": 441000
    },
    {
      "epoch": 305.75,
      "learning_rate": 1.3885041551246539e-05,
      "loss": 3.7974,
      "step": 441500
    },
    {
      "epoch": 306.0,
      "eval_loss": 3.7187914848327637,
      "eval_runtime": 73.1181,
      "eval_samples_per_second": 6734.635,
      "eval_steps_per_second": 2.202,
      "step": 441864
    },
    {
      "epoch": 306.09,
      "learning_rate": 1.3878116343490307e-05,
      "loss": 3.7968,
      "step": 442000
    },
    {
      "epoch": 306.44,
      "learning_rate": 1.3871191135734073e-05,
      "loss": 3.7954,
      "step": 442500
    },
    {
      "epoch": 306.79,
      "learning_rate": 1.386426592797784e-05,
      "loss": 3.7955,
      "step": 443000
    },
    {
      "epoch": 307.0,
      "eval_loss": 3.7259676456451416,
      "eval_runtime": 73.6071,
      "eval_samples_per_second": 6689.901,
      "eval_steps_per_second": 2.187,
      "step": 443308
    },
    {
      "epoch": 307.13,
      "learning_rate": 1.3857340720221608e-05,
      "loss": 3.7946,
      "step": 443500
    },
    {
      "epoch": 307.48,
      "learning_rate": 1.3850415512465377e-05,
      "loss": 3.7958,
      "step": 444000
    },
    {
      "epoch": 307.83,
      "learning_rate": 1.3843490304709142e-05,
      "loss": 3.793,
      "step": 444500
    },
    {
      "epoch": 308.0,
      "eval_loss": 3.719923734664917,
      "eval_runtime": 76.694,
      "eval_samples_per_second": 6420.633,
      "eval_steps_per_second": 2.099,
      "step": 444752
    },
    {
      "epoch": 308.17,
      "learning_rate": 1.3836565096952909e-05,
      "loss": 3.7951,
      "step": 445000
    },
    {
      "epoch": 308.52,
      "learning_rate": 1.3829639889196678e-05,
      "loss": 3.7968,
      "step": 445500
    },
    {
      "epoch": 308.86,
      "learning_rate": 1.3822714681440443e-05,
      "loss": 3.7923,
      "step": 446000
    },
    {
      "epoch": 309.0,
      "eval_loss": 3.7208216190338135,
      "eval_runtime": 74.1466,
      "eval_samples_per_second": 6641.218,
      "eval_steps_per_second": 2.171,
      "step": 446196
    },
    {
      "epoch": 309.21,
      "learning_rate": 1.3815789473684211e-05,
      "loss": 3.7895,
      "step": 446500
    },
    {
      "epoch": 309.56,
      "learning_rate": 1.380886426592798e-05,
      "loss": 3.7939,
      "step": 447000
    },
    {
      "epoch": 309.9,
      "learning_rate": 1.3801939058171747e-05,
      "loss": 3.7956,
      "step": 447500
    },
    {
      "epoch": 310.0,
      "eval_loss": 3.7165091037750244,
      "eval_runtime": 75.6044,
      "eval_samples_per_second": 6513.167,
      "eval_steps_per_second": 2.13,
      "step": 447640
    },
    {
      "epoch": 310.25,
      "learning_rate": 1.3795013850415512e-05,
      "loss": 3.7944,
      "step": 448000
    },
    {
      "epoch": 310.6,
      "learning_rate": 1.378808864265928e-05,
      "loss": 3.7903,
      "step": 448500
    },
    {
      "epoch": 310.94,
      "learning_rate": 1.378116343490305e-05,
      "loss": 3.7945,
      "step": 449000
    },
    {
      "epoch": 311.0,
      "eval_loss": 3.715456962585449,
      "eval_runtime": 73.561,
      "eval_samples_per_second": 6694.088,
      "eval_steps_per_second": 2.189,
      "step": 449084
    },
    {
      "epoch": 311.29,
      "learning_rate": 1.3774238227146815e-05,
      "loss": 3.7901,
      "step": 449500
    },
    {
      "epoch": 311.63,
      "learning_rate": 1.3767313019390583e-05,
      "loss": 3.7902,
      "step": 450000
    },
    {
      "epoch": 311.98,
      "learning_rate": 1.376038781163435e-05,
      "loss": 3.7925,
      "step": 450500
    },
    {
      "epoch": 312.0,
      "eval_loss": 3.7176241874694824,
      "eval_runtime": 78.0968,
      "eval_samples_per_second": 6305.307,
      "eval_steps_per_second": 2.062,
      "step": 450528
    },
    {
      "epoch": 312.33,
      "learning_rate": 1.3753462603878119e-05,
      "loss": 3.7932,
      "step": 451000
    },
    {
      "epoch": 312.67,
      "learning_rate": 1.3746537396121884e-05,
      "loss": 3.793,
      "step": 451500
    },
    {
      "epoch": 313.0,
      "eval_loss": 3.7189948558807373,
      "eval_runtime": 74.2572,
      "eval_samples_per_second": 6631.333,
      "eval_steps_per_second": 2.168,
      "step": 451972
    },
    {
      "epoch": 313.02,
      "learning_rate": 1.3739612188365653e-05,
      "loss": 3.7906,
      "step": 452000
    },
    {
      "epoch": 313.37,
      "learning_rate": 1.373268698060942e-05,
      "loss": 3.7898,
      "step": 452500
    },
    {
      "epoch": 313.71,
      "learning_rate": 1.3725761772853188e-05,
      "loss": 3.7905,
      "step": 453000
    },
    {
      "epoch": 314.0,
      "eval_loss": 3.7148730754852295,
      "eval_runtime": 74.5995,
      "eval_samples_per_second": 6600.898,
      "eval_steps_per_second": 2.158,
      "step": 453416
    },
    {
      "epoch": 314.06,
      "learning_rate": 1.3718836565096953e-05,
      "loss": 3.7931,
      "step": 453500
    },
    {
      "epoch": 314.4,
      "learning_rate": 1.3711911357340722e-05,
      "loss": 3.7909,
      "step": 454000
    },
    {
      "epoch": 314.75,
      "learning_rate": 1.3704986149584489e-05,
      "loss": 3.789,
      "step": 454500
    },
    {
      "epoch": 315.0,
      "eval_loss": 3.715317726135254,
      "eval_runtime": 73.3676,
      "eval_samples_per_second": 6711.739,
      "eval_steps_per_second": 2.194,
      "step": 454860
    },
    {
      "epoch": 315.1,
      "learning_rate": 1.3698060941828256e-05,
      "loss": 3.7908,
      "step": 455000
    },
    {
      "epoch": 315.44,
      "learning_rate": 1.3691135734072023e-05,
      "loss": 3.7927,
      "step": 455500
    },
    {
      "epoch": 315.79,
      "learning_rate": 1.3684210526315791e-05,
      "loss": 3.7883,
      "step": 456000
    },
    {
      "epoch": 316.0,
      "eval_loss": 3.7178468704223633,
      "eval_runtime": 77.3962,
      "eval_samples_per_second": 6362.381,
      "eval_steps_per_second": 2.08,
      "step": 456304
    },
    {
      "epoch": 316.14,
      "learning_rate": 1.3677285318559558e-05,
      "loss": 3.7919,
      "step": 456500
    },
    {
      "epoch": 316.48,
      "learning_rate": 1.3670360110803325e-05,
      "loss": 3.7879,
      "step": 457000
    },
    {
      "epoch": 316.83,
      "learning_rate": 1.3663434903047092e-05,
      "loss": 3.7897,
      "step": 457500
    },
    {
      "epoch": 317.0,
      "eval_loss": 3.7118659019470215,
      "eval_runtime": 74.0536,
      "eval_samples_per_second": 6649.566,
      "eval_steps_per_second": 2.174,
      "step": 457748
    },
    {
      "epoch": 317.17,
      "learning_rate": 1.365650969529086e-05,
      "loss": 3.7926,
      "step": 458000
    },
    {
      "epoch": 317.52,
      "learning_rate": 1.3649584487534626e-05,
      "loss": 3.7899,
      "step": 458500
    },
    {
      "epoch": 317.87,
      "learning_rate": 1.3642659279778395e-05,
      "loss": 3.7886,
      "step": 459000
    },
    {
      "epoch": 318.0,
      "eval_loss": 3.7171525955200195,
      "eval_runtime": 73.7675,
      "eval_samples_per_second": 6675.351,
      "eval_steps_per_second": 2.183,
      "step": 459192
    },
    {
      "epoch": 318.21,
      "learning_rate": 1.3635734072022161e-05,
      "loss": 3.7891,
      "step": 459500
    },
    {
      "epoch": 318.56,
      "learning_rate": 1.362880886426593e-05,
      "loss": 3.789,
      "step": 460000
    },
    {
      "epoch": 318.91,
      "learning_rate": 1.3621883656509695e-05,
      "loss": 3.7917,
      "step": 460500
    },
    {
      "epoch": 319.0,
      "eval_loss": 3.713498830795288,
      "eval_runtime": 75.1926,
      "eval_samples_per_second": 6548.836,
      "eval_steps_per_second": 2.141,
      "step": 460636
    },
    {
      "epoch": 319.25,
      "learning_rate": 1.3614958448753464e-05,
      "loss": 3.7887,
      "step": 461000
    },
    {
      "epoch": 319.6,
      "learning_rate": 1.360803324099723e-05,
      "loss": 3.7915,
      "step": 461500
    },
    {
      "epoch": 319.94,
      "learning_rate": 1.3601108033240998e-05,
      "loss": 3.7868,
      "step": 462000
    },
    {
      "epoch": 320.0,
      "eval_loss": 3.7122466564178467,
      "eval_runtime": 73.8863,
      "eval_samples_per_second": 6664.614,
      "eval_steps_per_second": 2.179,
      "step": 462080
    },
    {
      "epoch": 320.29,
      "learning_rate": 1.3594182825484765e-05,
      "loss": 3.7871,
      "step": 462500
    },
    {
      "epoch": 320.64,
      "learning_rate": 1.3587257617728533e-05,
      "loss": 3.7869,
      "step": 463000
    },
    {
      "epoch": 320.98,
      "learning_rate": 1.3580332409972302e-05,
      "loss": 3.7881,
      "step": 463500
    },
    {
      "epoch": 321.0,
      "eval_loss": 3.7132568359375,
      "eval_runtime": 75.8172,
      "eval_samples_per_second": 6494.889,
      "eval_steps_per_second": 2.124,
      "step": 463524
    },
    {
      "epoch": 321.33,
      "learning_rate": 1.3573407202216067e-05,
      "loss": 3.785,
      "step": 464000
    },
    {
      "epoch": 321.68,
      "learning_rate": 1.3566481994459834e-05,
      "loss": 3.7882,
      "step": 464500
    },
    {
      "epoch": 322.0,
      "eval_loss": 3.714015245437622,
      "eval_runtime": 77.194,
      "eval_samples_per_second": 6379.048,
      "eval_steps_per_second": 2.086,
      "step": 464968
    },
    {
      "epoch": 322.02,
      "learning_rate": 1.3559556786703603e-05,
      "loss": 3.7875,
      "step": 465000
    },
    {
      "epoch": 322.37,
      "learning_rate": 1.3552631578947371e-05,
      "loss": 3.789,
      "step": 465500
    },
    {
      "epoch": 322.71,
      "learning_rate": 1.3545706371191136e-05,
      "loss": 3.7871,
      "step": 466000
    },
    {
      "epoch": 323.0,
      "eval_loss": 3.707482099533081,
      "eval_runtime": 73.4812,
      "eval_samples_per_second": 6701.361,
      "eval_steps_per_second": 2.191,
      "step": 466412
    },
    {
      "epoch": 323.06,
      "learning_rate": 1.3538781163434903e-05,
      "loss": 3.7855,
      "step": 466500
    },
    {
      "epoch": 323.41,
      "learning_rate": 1.3531855955678672e-05,
      "loss": 3.7851,
      "step": 467000
    },
    {
      "epoch": 323.75,
      "learning_rate": 1.3524930747922437e-05,
      "loss": 3.7861,
      "step": 467500
    },
    {
      "epoch": 324.0,
      "eval_loss": 3.7149996757507324,
      "eval_runtime": 77.2505,
      "eval_samples_per_second": 6374.382,
      "eval_steps_per_second": 2.084,
      "step": 467856
    },
    {
      "epoch": 324.1,
      "learning_rate": 1.3518005540166206e-05,
      "loss": 3.7868,
      "step": 468000
    },
    {
      "epoch": 324.45,
      "learning_rate": 1.3511080332409974e-05,
      "loss": 3.7869,
      "step": 468500
    },
    {
      "epoch": 324.79,
      "learning_rate": 1.3504155124653741e-05,
      "loss": 3.7838,
      "step": 469000
    },
    {
      "epoch": 325.0,
      "eval_loss": 3.7106683254241943,
      "eval_runtime": 74.8849,
      "eval_samples_per_second": 6575.744,
      "eval_steps_per_second": 2.15,
      "step": 469300
    },
    {
      "epoch": 325.14,
      "learning_rate": 1.3497229916897507e-05,
      "loss": 3.7865,
      "step": 469500
    },
    {
      "epoch": 325.48,
      "learning_rate": 1.3490304709141275e-05,
      "loss": 3.7858,
      "step": 470000
    },
    {
      "epoch": 325.83,
      "learning_rate": 1.3483379501385044e-05,
      "loss": 3.784,
      "step": 470500
    },
    {
      "epoch": 326.0,
      "eval_loss": 3.7102861404418945,
      "eval_runtime": 74.7474,
      "eval_samples_per_second": 6587.842,
      "eval_steps_per_second": 2.154,
      "step": 470744
    },
    {
      "epoch": 326.18,
      "learning_rate": 1.3476454293628809e-05,
      "loss": 3.786,
      "step": 471000
    },
    {
      "epoch": 326.52,
      "learning_rate": 1.3469529085872578e-05,
      "loss": 3.7855,
      "step": 471500
    },
    {
      "epoch": 326.87,
      "learning_rate": 1.3462603878116345e-05,
      "loss": 3.7861,
      "step": 472000
    },
    {
      "epoch": 327.0,
      "eval_loss": 3.7090983390808105,
      "eval_runtime": 73.9182,
      "eval_samples_per_second": 6661.745,
      "eval_steps_per_second": 2.178,
      "step": 472188
    },
    {
      "epoch": 327.22,
      "learning_rate": 1.3455678670360113e-05,
      "loss": 3.7856,
      "step": 472500
    },
    {
      "epoch": 327.56,
      "learning_rate": 1.3448753462603878e-05,
      "loss": 3.7827,
      "step": 473000
    },
    {
      "epoch": 327.91,
      "learning_rate": 1.3441828254847647e-05,
      "loss": 3.7854,
      "step": 473500
    },
    {
      "epoch": 328.0,
      "eval_loss": 3.708383083343506,
      "eval_runtime": 74.4676,
      "eval_samples_per_second": 6612.59,
      "eval_steps_per_second": 2.162,
      "step": 473632
    },
    {
      "epoch": 328.25,
      "learning_rate": 1.3434903047091414e-05,
      "loss": 3.785,
      "step": 474000
    },
    {
      "epoch": 328.6,
      "learning_rate": 1.3427977839335181e-05,
      "loss": 3.7828,
      "step": 474500
    },
    {
      "epoch": 328.95,
      "learning_rate": 1.3421052631578948e-05,
      "loss": 3.7859,
      "step": 475000
    },
    {
      "epoch": 329.0,
      "eval_loss": 3.7049036026000977,
      "eval_runtime": 76.0122,
      "eval_samples_per_second": 6478.227,
      "eval_steps_per_second": 2.118,
      "step": 475076
    },
    {
      "epoch": 329.29,
      "learning_rate": 1.3414127423822716e-05,
      "loss": 3.7826,
      "step": 475500
    },
    {
      "epoch": 329.64,
      "learning_rate": 1.3407202216066483e-05,
      "loss": 3.7815,
      "step": 476000
    },
    {
      "epoch": 329.99,
      "learning_rate": 1.340027700831025e-05,
      "loss": 3.7841,
      "step": 476500
    },
    {
      "epoch": 330.0,
      "eval_loss": 3.7112672328948975,
      "eval_runtime": 73.3442,
      "eval_samples_per_second": 6713.874,
      "eval_steps_per_second": 2.195,
      "step": 476520
    },
    {
      "epoch": 330.33,
      "learning_rate": 1.3393351800554017e-05,
      "loss": 3.7814,
      "step": 477000
    },
    {
      "epoch": 330.68,
      "learning_rate": 1.3386426592797786e-05,
      "loss": 3.7823,
      "step": 477500
    },
    {
      "epoch": 331.0,
      "eval_loss": 3.7072346210479736,
      "eval_runtime": 76.1202,
      "eval_samples_per_second": 6469.031,
      "eval_steps_per_second": 2.115,
      "step": 477964
    },
    {
      "epoch": 331.02,
      "learning_rate": 1.3379501385041553e-05,
      "loss": 3.7845,
      "step": 478000
    },
    {
      "epoch": 331.37,
      "learning_rate": 1.337257617728532e-05,
      "loss": 3.7822,
      "step": 478500
    },
    {
      "epoch": 331.72,
      "learning_rate": 1.3365650969529087e-05,
      "loss": 3.7848,
      "step": 479000
    },
    {
      "epoch": 332.0,
      "eval_loss": 3.708578109741211,
      "eval_runtime": 73.6567,
      "eval_samples_per_second": 6685.396,
      "eval_steps_per_second": 2.186,
      "step": 479408
    },
    {
      "epoch": 332.06,
      "learning_rate": 1.3358725761772855e-05,
      "loss": 3.7811,
      "step": 479500
    },
    {
      "epoch": 332.41,
      "learning_rate": 1.335180055401662e-05,
      "loss": 3.7825,
      "step": 480000
    },
    {
      "epoch": 332.76,
      "learning_rate": 1.3344875346260389e-05,
      "loss": 3.7856,
      "step": 480500
    },
    {
      "epoch": 333.0,
      "eval_loss": 3.7044219970703125,
      "eval_runtime": 73.1824,
      "eval_samples_per_second": 6728.725,
      "eval_steps_per_second": 2.2,
      "step": 480852
    },
    {
      "epoch": 333.1,
      "learning_rate": 1.3337950138504156e-05,
      "loss": 3.7804,
      "step": 481000
    },
    {
      "epoch": 333.45,
      "learning_rate": 1.3331024930747925e-05,
      "loss": 3.7816,
      "step": 481500
    },
    {
      "epoch": 333.8,
      "learning_rate": 1.332409972299169e-05,
      "loss": 3.7793,
      "step": 482000
    },
    {
      "epoch": 334.0,
      "eval_loss": 3.7033491134643555,
      "eval_runtime": 74.3462,
      "eval_samples_per_second": 6623.388,
      "eval_steps_per_second": 2.166,
      "step": 482296
    },
    {
      "epoch": 334.14,
      "learning_rate": 1.3317174515235458e-05,
      "loss": 3.7816,
      "step": 482500
    },
    {
      "epoch": 334.49,
      "learning_rate": 1.3310249307479225e-05,
      "loss": 3.7796,
      "step": 483000
    },
    {
      "epoch": 334.83,
      "learning_rate": 1.3303324099722992e-05,
      "loss": 3.7823,
      "step": 483500
    },
    {
      "epoch": 335.0,
      "eval_loss": 3.712414503097534,
      "eval_runtime": 73.7811,
      "eval_samples_per_second": 6674.118,
      "eval_steps_per_second": 2.182,
      "step": 483740
    },
    {
      "epoch": 335.18,
      "learning_rate": 1.3296398891966759e-05,
      "loss": 3.7811,
      "step": 484000
    },
    {
      "epoch": 335.53,
      "learning_rate": 1.3289473684210528e-05,
      "loss": 3.7801,
      "step": 484500
    },
    {
      "epoch": 335.87,
      "learning_rate": 1.3282548476454296e-05,
      "loss": 3.7779,
      "step": 485000
    },
    {
      "epoch": 336.0,
      "eval_loss": 3.713364601135254,
      "eval_runtime": 74.5722,
      "eval_samples_per_second": 6603.322,
      "eval_steps_per_second": 2.159,
      "step": 485184
    },
    {
      "epoch": 336.22,
      "learning_rate": 1.3275623268698062e-05,
      "loss": 3.7814,
      "step": 485500
    },
    {
      "epoch": 336.57,
      "learning_rate": 1.3268698060941828e-05,
      "loss": 3.7797,
      "step": 486000
    },
    {
      "epoch": 336.91,
      "learning_rate": 1.3261772853185597e-05,
      "loss": 3.7817,
      "step": 486500
    },
    {
      "epoch": 337.0,
      "eval_loss": 3.704362392425537,
      "eval_runtime": 73.7818,
      "eval_samples_per_second": 6674.058,
      "eval_steps_per_second": 2.182,
      "step": 486628
    },
    {
      "epoch": 337.26,
      "learning_rate": 1.3254847645429362e-05,
      "loss": 3.7809,
      "step": 487000
    },
    {
      "epoch": 337.6,
      "learning_rate": 1.3247922437673131e-05,
      "loss": 3.7793,
      "step": 487500
    },
    {
      "epoch": 337.95,
      "learning_rate": 1.32409972299169e-05,
      "loss": 3.7766,
      "step": 488000
    },
    {
      "epoch": 338.0,
      "eval_loss": 3.7027761936187744,
      "eval_runtime": 74.7391,
      "eval_samples_per_second": 6588.57,
      "eval_steps_per_second": 2.154,
      "step": 488072
    },
    {
      "epoch": 338.3,
      "learning_rate": 1.3234072022160666e-05,
      "loss": 3.7818,
      "step": 488500
    },
    {
      "epoch": 338.64,
      "learning_rate": 1.3227146814404432e-05,
      "loss": 3.778,
      "step": 489000
    },
    {
      "epoch": 338.99,
      "learning_rate": 1.32202216066482e-05,
      "loss": 3.7803,
      "step": 489500
    },
    {
      "epoch": 339.0,
      "eval_loss": 3.7038166522979736,
      "eval_runtime": 73.5019,
      "eval_samples_per_second": 6699.476,
      "eval_steps_per_second": 2.19,
      "step": 489516
    },
    {
      "epoch": 339.34,
      "learning_rate": 1.3213296398891969e-05,
      "loss": 3.7785,
      "step": 490000
    },
    {
      "epoch": 339.68,
      "learning_rate": 1.3206371191135736e-05,
      "loss": 3.7806,
      "step": 490500
    },
    {
      "epoch": 340.0,
      "eval_loss": 3.707486867904663,
      "eval_runtime": 76.5029,
      "eval_samples_per_second": 6436.667,
      "eval_steps_per_second": 2.104,
      "step": 490960
    },
    {
      "epoch": 340.03,
      "learning_rate": 1.3199445983379501e-05,
      "loss": 3.7771,
      "step": 491000
    },
    {
      "epoch": 340.37,
      "learning_rate": 1.319252077562327e-05,
      "loss": 3.7764,
      "step": 491500
    },
    {
      "epoch": 340.72,
      "learning_rate": 1.3185595567867038e-05,
      "loss": 3.78,
      "step": 492000
    },
    {
      "epoch": 341.0,
      "eval_loss": 3.7045812606811523,
      "eval_runtime": 74.5994,
      "eval_samples_per_second": 6600.913,
      "eval_steps_per_second": 2.158,
      "step": 492404
    },
    {
      "epoch": 341.07,
      "learning_rate": 1.3178670360110804e-05,
      "loss": 3.7806,
      "step": 492500
    },
    {
      "epoch": 341.41,
      "learning_rate": 1.3171745152354572e-05,
      "loss": 3.7798,
      "step": 493000
    },
    {
      "epoch": 341.76,
      "learning_rate": 1.3164819944598339e-05,
      "loss": 3.7794,
      "step": 493500
    },
    {
      "epoch": 342.0,
      "eval_loss": 3.7000420093536377,
      "eval_runtime": 75.6658,
      "eval_samples_per_second": 6507.879,
      "eval_steps_per_second": 2.128,
      "step": 493848
    },
    {
      "epoch": 342.11,
      "learning_rate": 1.3157894736842108e-05,
      "loss": 3.7764,
      "step": 494000
    },
    {
      "epoch": 342.45,
      "learning_rate": 1.3150969529085873e-05,
      "loss": 3.7793,
      "step": 494500
    },
    {
      "epoch": 342.8,
      "learning_rate": 1.3144044321329642e-05,
      "loss": 3.7768,
      "step": 495000
    },
    {
      "epoch": 343.0,
      "eval_loss": 3.704115390777588,
      "eval_runtime": 75.5071,
      "eval_samples_per_second": 6521.559,
      "eval_steps_per_second": 2.132,
      "step": 495292
    },
    {
      "epoch": 343.14,
      "learning_rate": 1.3137119113573408e-05,
      "loss": 3.7755,
      "step": 495500
    },
    {
      "epoch": 343.49,
      "learning_rate": 1.3130193905817175e-05,
      "loss": 3.7773,
      "step": 496000
    },
    {
      "epoch": 343.84,
      "learning_rate": 1.3123268698060942e-05,
      "loss": 3.776,
      "step": 496500
    },
    {
      "epoch": 344.0,
      "eval_loss": 3.7043023109436035,
      "eval_runtime": 73.6566,
      "eval_samples_per_second": 6685.399,
      "eval_steps_per_second": 2.186,
      "step": 496736
    },
    {
      "epoch": 344.18,
      "learning_rate": 1.3116343490304711e-05,
      "loss": 3.7785,
      "step": 497000
    },
    {
      "epoch": 344.53,
      "learning_rate": 1.3109418282548478e-05,
      "loss": 3.7817,
      "step": 497500
    },
    {
      "epoch": 344.88,
      "learning_rate": 1.3102493074792245e-05,
      "loss": 3.7799,
      "step": 498000
    },
    {
      "epoch": 345.0,
      "eval_loss": 3.7023870944976807,
      "eval_runtime": 76.6367,
      "eval_samples_per_second": 6425.431,
      "eval_steps_per_second": 2.101,
      "step": 498180
    },
    {
      "epoch": 345.22,
      "learning_rate": 1.3095567867036012e-05,
      "loss": 3.776,
      "step": 498500
    },
    {
      "epoch": 345.57,
      "learning_rate": 1.308864265927978e-05,
      "loss": 3.7761,
      "step": 499000
    },
    {
      "epoch": 345.91,
      "learning_rate": 1.3081717451523547e-05,
      "loss": 3.773,
      "step": 499500
    },
    {
      "epoch": 346.0,
      "eval_loss": 3.698683500289917,
      "eval_runtime": 75.0471,
      "eval_samples_per_second": 6561.533,
      "eval_steps_per_second": 2.145,
      "step": 499624
    },
    {
      "epoch": 346.26,
      "learning_rate": 1.3074792243767314e-05,
      "loss": 3.776,
      "step": 500000
    },
    {
      "epoch": 346.61,
      "learning_rate": 1.3067867036011081e-05,
      "loss": 3.7738,
      "step": 500500
    },
    {
      "epoch": 346.95,
      "learning_rate": 1.306094182825485e-05,
      "loss": 3.7755,
      "step": 501000
    },
    {
      "epoch": 347.0,
      "eval_loss": 3.6997292041778564,
      "eval_runtime": 75.304,
      "eval_samples_per_second": 6539.146,
      "eval_steps_per_second": 2.138,
      "step": 501068
    },
    {
      "epoch": 347.3,
      "learning_rate": 1.3054016620498615e-05,
      "loss": 3.7746,
      "step": 501500
    },
    {
      "epoch": 347.65,
      "learning_rate": 1.3047091412742383e-05,
      "loss": 3.7746,
      "step": 502000
    },
    {
      "epoch": 347.99,
      "learning_rate": 1.304016620498615e-05,
      "loss": 3.7738,
      "step": 502500
    },
    {
      "epoch": 348.0,
      "eval_loss": 3.7015268802642822,
      "eval_runtime": 73.8099,
      "eval_samples_per_second": 6671.521,
      "eval_steps_per_second": 2.181,
      "step": 502512
    },
    {
      "epoch": 348.34,
      "learning_rate": 1.3033240997229919e-05,
      "loss": 3.7753,
      "step": 503000
    },
    {
      "epoch": 348.68,
      "learning_rate": 1.3026315789473684e-05,
      "loss": 3.7744,
      "step": 503500
    },
    {
      "epoch": 349.0,
      "eval_loss": 3.7030045986175537,
      "eval_runtime": 75.2471,
      "eval_samples_per_second": 6544.096,
      "eval_steps_per_second": 2.14,
      "step": 503956
    },
    {
      "epoch": 349.03,
      "learning_rate": 1.3019390581717453e-05,
      "loss": 3.7736,
      "step": 504000
    },
    {
      "epoch": 349.38,
      "learning_rate": 1.301246537396122e-05,
      "loss": 3.7759,
      "step": 504500
    },
    {
      "epoch": 349.72,
      "learning_rate": 1.3005540166204987e-05,
      "loss": 3.7721,
      "step": 505000
    },
    {
      "epoch": 350.0,
      "eval_loss": 3.7013323307037354,
      "eval_runtime": 74.9488,
      "eval_samples_per_second": 6570.141,
      "eval_steps_per_second": 2.148,
      "step": 505400
    },
    {
      "epoch": 350.07,
      "learning_rate": 1.2998614958448754e-05,
      "loss": 3.7721,
      "step": 505500
    },
    {
      "epoch": 350.42,
      "learning_rate": 1.2991689750692522e-05,
      "loss": 3.7719,
      "step": 506000
    },
    {
      "epoch": 350.76,
      "learning_rate": 1.298476454293629e-05,
      "loss": 3.7756,
      "step": 506500
    },
    {
      "epoch": 351.0,
      "eval_loss": 3.7012503147125244,
      "eval_runtime": 75.0555,
      "eval_samples_per_second": 6560.802,
      "eval_steps_per_second": 2.145,
      "step": 506844
    },
    {
      "epoch": 351.11,
      "learning_rate": 1.2977839335180056e-05,
      "loss": 3.7736,
      "step": 507000
    },
    {
      "epoch": 351.45,
      "learning_rate": 1.2970914127423823e-05,
      "loss": 3.7706,
      "step": 507500
    },
    {
      "epoch": 351.8,
      "learning_rate": 1.2963988919667592e-05,
      "loss": 3.7715,
      "step": 508000
    },
    {
      "epoch": 352.0,
      "eval_loss": 3.699119806289673,
      "eval_runtime": 74.874,
      "eval_samples_per_second": 6576.701,
      "eval_steps_per_second": 2.15,
      "step": 508288
    },
    {
      "epoch": 352.15,
      "learning_rate": 1.2957063711911357e-05,
      "loss": 3.7765,
      "step": 508500
    },
    {
      "epoch": 352.49,
      "learning_rate": 1.2950138504155125e-05,
      "loss": 3.7717,
      "step": 509000
    },
    {
      "epoch": 352.84,
      "learning_rate": 1.2943213296398894e-05,
      "loss": 3.7733,
      "step": 509500
    },
    {
      "epoch": 353.0,
      "eval_loss": 3.6972882747650146,
      "eval_runtime": 76.25,
      "eval_samples_per_second": 6458.018,
      "eval_steps_per_second": 2.111,
      "step": 509732
    },
    {
      "epoch": 353.19,
      "learning_rate": 1.2936288088642661e-05,
      "loss": 3.7728,
      "step": 510000
    },
    {
      "epoch": 353.53,
      "learning_rate": 1.2929362880886426e-05,
      "loss": 3.7725,
      "step": 510500
    },
    {
      "epoch": 353.88,
      "learning_rate": 1.2922437673130195e-05,
      "loss": 3.7701,
      "step": 511000
    },
    {
      "epoch": 354.0,
      "eval_loss": 3.700070858001709,
      "eval_runtime": 75.0621,
      "eval_samples_per_second": 6560.222,
      "eval_steps_per_second": 2.145,
      "step": 511176
    },
    {
      "epoch": 354.22,
      "learning_rate": 1.2915512465373963e-05,
      "loss": 3.7737,
      "step": 511500
    },
    {
      "epoch": 354.57,
      "learning_rate": 1.290858725761773e-05,
      "loss": 3.7703,
      "step": 512000
    },
    {
      "epoch": 354.92,
      "learning_rate": 1.2901662049861496e-05,
      "loss": 3.7733,
      "step": 512500
    },
    {
      "epoch": 355.0,
      "eval_loss": 3.697479248046875,
      "eval_runtime": 75.0016,
      "eval_samples_per_second": 6565.514,
      "eval_steps_per_second": 2.147,
      "step": 512620
    },
    {
      "epoch": 355.26,
      "learning_rate": 1.2894736842105264e-05,
      "loss": 3.7696,
      "step": 513000
    },
    {
      "epoch": 355.61,
      "learning_rate": 1.2887811634349033e-05,
      "loss": 3.771,
      "step": 513500
    },
    {
      "epoch": 355.96,
      "learning_rate": 1.2880886426592798e-05,
      "loss": 3.7723,
      "step": 514000
    },
    {
      "epoch": 356.0,
      "eval_loss": 3.695901870727539,
      "eval_runtime": 74.3951,
      "eval_samples_per_second": 6619.041,
      "eval_steps_per_second": 2.164,
      "step": 514064
    },
    {
      "epoch": 356.3,
      "learning_rate": 1.2873961218836567e-05,
      "loss": 3.771,
      "step": 514500
    },
    {
      "epoch": 356.65,
      "learning_rate": 1.2867036011080334e-05,
      "loss": 3.7708,
      "step": 515000
    },
    {
      "epoch": 356.99,
      "learning_rate": 1.2860110803324102e-05,
      "loss": 3.7671,
      "step": 515500
    },
    {
      "epoch": 357.0,
      "eval_loss": 3.696229934692383,
      "eval_runtime": 78.1947,
      "eval_samples_per_second": 6297.409,
      "eval_steps_per_second": 2.059,
      "step": 515508
    },
    {
      "epoch": 357.34,
      "learning_rate": 1.2853185595567867e-05,
      "loss": 3.7695,
      "step": 516000
    },
    {
      "epoch": 357.69,
      "learning_rate": 1.2846260387811636e-05,
      "loss": 3.7711,
      "step": 516500
    },
    {
      "epoch": 358.0,
      "eval_loss": 3.6941683292388916,
      "eval_runtime": 79.1242,
      "eval_samples_per_second": 6223.433,
      "eval_steps_per_second": 2.035,
      "step": 516952
    },
    {
      "epoch": 358.03,
      "learning_rate": 1.2839335180055403e-05,
      "loss": 3.7728,
      "step": 517000
    },
    {
      "epoch": 358.38,
      "learning_rate": 1.283240997229917e-05,
      "loss": 3.7696,
      "step": 517500
    },
    {
      "epoch": 358.73,
      "learning_rate": 1.2825484764542937e-05,
      "loss": 3.7721,
      "step": 518000
    },
    {
      "epoch": 359.0,
      "eval_loss": 3.6970505714416504,
      "eval_runtime": 78.5878,
      "eval_samples_per_second": 6265.908,
      "eval_steps_per_second": 2.049,
      "step": 518396
    },
    {
      "epoch": 359.07,
      "learning_rate": 1.2818559556786705e-05,
      "loss": 3.7704,
      "step": 518500
    },
    {
      "epoch": 359.42,
      "learning_rate": 1.2811634349030472e-05,
      "loss": 3.7671,
      "step": 519000
    },
    {
      "epoch": 359.76,
      "learning_rate": 1.280470914127424e-05,
      "loss": 3.7711,
      "step": 519500
    },
    {
      "epoch": 360.0,
      "eval_loss": 3.695997476577759,
      "eval_runtime": 74.3491,
      "eval_samples_per_second": 6623.133,
      "eval_steps_per_second": 2.165,
      "step": 519840
    },
    {
      "epoch": 360.11,
      "learning_rate": 1.2797783933518006e-05,
      "loss": 3.7685,
      "step": 520000
    },
    {
      "epoch": 360.46,
      "learning_rate": 1.2790858725761775e-05,
      "loss": 3.7706,
      "step": 520500
    },
    {
      "epoch": 360.8,
      "learning_rate": 1.278393351800554e-05,
      "loss": 3.7684,
      "step": 521000
    },
    {
      "epoch": 361.0,
      "eval_loss": 3.6960372924804688,
      "eval_runtime": 74.0632,
      "eval_samples_per_second": 6648.696,
      "eval_steps_per_second": 2.174,
      "step": 521284
    },
    {
      "epoch": 361.15,
      "learning_rate": 1.2777008310249309e-05,
      "loss": 3.7715,
      "step": 521500
    },
    {
      "epoch": 361.5,
      "learning_rate": 1.2770083102493075e-05,
      "loss": 3.7673,
      "step": 522000
    },
    {
      "epoch": 361.84,
      "learning_rate": 1.2763157894736844e-05,
      "loss": 3.7691,
      "step": 522500
    },
    {
      "epoch": 362.0,
      "eval_loss": 3.693530559539795,
      "eval_runtime": 74.6535,
      "eval_samples_per_second": 6596.124,
      "eval_steps_per_second": 2.157,
      "step": 522728
    },
    {
      "epoch": 362.19,
      "learning_rate": 1.275623268698061e-05,
      "loss": 3.7684,
      "step": 523000
    },
    {
      "epoch": 362.53,
      "learning_rate": 1.2749307479224378e-05,
      "loss": 3.7678,
      "step": 523500
    },
    {
      "epoch": 362.88,
      "learning_rate": 1.2742382271468145e-05,
      "loss": 3.7701,
      "step": 524000
    },
    {
      "epoch": 363.0,
      "eval_loss": 3.6938717365264893,
      "eval_runtime": 77.8737,
      "eval_samples_per_second": 6323.366,
      "eval_steps_per_second": 2.067,
      "step": 524172
    },
    {
      "epoch": 363.23,
      "learning_rate": 1.2735457063711913e-05,
      "loss": 3.7676,
      "step": 524500
    },
    {
      "epoch": 363.57,
      "learning_rate": 1.2728531855955679e-05,
      "loss": 3.7673,
      "step": 525000
    },
    {
      "epoch": 363.92,
      "learning_rate": 1.2721606648199447e-05,
      "loss": 3.7678,
      "step": 525500
    },
    {
      "epoch": 364.0,
      "eval_loss": 3.694033145904541,
      "eval_runtime": 75.4612,
      "eval_samples_per_second": 6525.529,
      "eval_steps_per_second": 2.134,
      "step": 525616
    },
    {
      "epoch": 364.27,
      "learning_rate": 1.2714681440443214e-05,
      "loss": 3.7668,
      "step": 526000
    },
    {
      "epoch": 364.61,
      "learning_rate": 1.2707756232686981e-05,
      "loss": 3.7671,
      "step": 526500
    },
    {
      "epoch": 364.96,
      "learning_rate": 1.2700831024930748e-05,
      "loss": 3.7685,
      "step": 527000
    },
    {
      "epoch": 365.0,
      "eval_loss": 3.6901051998138428,
      "eval_runtime": 74.9685,
      "eval_samples_per_second": 6568.408,
      "eval_steps_per_second": 2.148,
      "step": 527060
    },
    {
      "epoch": 365.3,
      "learning_rate": 1.2693905817174517e-05,
      "loss": 3.7688,
      "step": 527500
    },
    {
      "epoch": 365.65,
      "learning_rate": 1.2686980609418285e-05,
      "loss": 3.7698,
      "step": 528000
    },
    {
      "epoch": 366.0,
      "learning_rate": 1.268005540166205e-05,
      "loss": 3.7679,
      "step": 528500
    },
    {
      "epoch": 366.0,
      "eval_loss": 3.6931703090667725,
      "eval_runtime": 74.1691,
      "eval_samples_per_second": 6639.205,
      "eval_steps_per_second": 2.171,
      "step": 528504
    },
    {
      "epoch": 366.34,
      "learning_rate": 1.2673130193905817e-05,
      "loss": 3.7677,
      "step": 529000
    },
    {
      "epoch": 366.69,
      "learning_rate": 1.2666204986149586e-05,
      "loss": 3.7684,
      "step": 529500
    },
    {
      "epoch": 367.0,
      "eval_loss": 3.687340497970581,
      "eval_runtime": 75.5717,
      "eval_samples_per_second": 6515.988,
      "eval_steps_per_second": 2.13,
      "step": 529948
    },
    {
      "epoch": 367.04,
      "learning_rate": 1.2659279778393351e-05,
      "loss": 3.7671,
      "step": 530000
    },
    {
      "epoch": 367.38,
      "learning_rate": 1.265235457063712e-05,
      "loss": 3.7657,
      "step": 530500
    },
    {
      "epoch": 367.73,
      "learning_rate": 1.2645429362880889e-05,
      "loss": 3.7647,
      "step": 531000
    },
    {
      "epoch": 368.0,
      "eval_loss": 3.690807342529297,
      "eval_runtime": 73.5177,
      "eval_samples_per_second": 6698.037,
      "eval_steps_per_second": 2.19,
      "step": 531392
    },
    {
      "epoch": 368.07,
      "learning_rate": 1.2638504155124655e-05,
      "loss": 3.7657,
      "step": 531500
    },
    {
      "epoch": 368.42,
      "learning_rate": 1.263157894736842e-05,
      "loss": 3.7676,
      "step": 532000
    },
    {
      "epoch": 368.77,
      "learning_rate": 1.262465373961219e-05,
      "loss": 3.7641,
      "step": 532500
    },
    {
      "epoch": 369.0,
      "eval_loss": 3.688164710998535,
      "eval_runtime": 75.7562,
      "eval_samples_per_second": 6500.118,
      "eval_steps_per_second": 2.125,
      "step": 532836
    },
    {
      "epoch": 369.11,
      "learning_rate": 1.2617728531855958e-05,
      "loss": 3.7655,
      "step": 533000
    },
    {
      "epoch": 369.46,
      "learning_rate": 1.2610803324099723e-05,
      "loss": 3.7655,
      "step": 533500
    },
    {
      "epoch": 369.81,
      "learning_rate": 1.2603878116343492e-05,
      "loss": 3.7648,
      "step": 534000
    },
    {
      "epoch": 370.0,
      "eval_loss": 3.6942825317382812,
      "eval_runtime": 75.9898,
      "eval_samples_per_second": 6480.131,
      "eval_steps_per_second": 2.119,
      "step": 534280
    },
    {
      "epoch": 370.15,
      "learning_rate": 1.2596952908587259e-05,
      "loss": 3.7682,
      "step": 534500
    },
    {
      "epoch": 370.5,
      "learning_rate": 1.2590027700831027e-05,
      "loss": 3.7648,
      "step": 535000
    },
    {
      "epoch": 370.84,
      "learning_rate": 1.2583102493074792e-05,
      "loss": 3.7688,
      "step": 535500
    },
    {
      "epoch": 371.0,
      "eval_loss": 3.6921067237854004,
      "eval_runtime": 76.7171,
      "eval_samples_per_second": 6418.698,
      "eval_steps_per_second": 2.099,
      "step": 535724
    },
    {
      "epoch": 371.19,
      "learning_rate": 1.2576177285318561e-05,
      "loss": 3.7628,
      "step": 536000
    },
    {
      "epoch": 371.54,
      "learning_rate": 1.2569252077562328e-05,
      "loss": 3.7644,
      "step": 536500
    },
    {
      "epoch": 371.88,
      "learning_rate": 1.2562326869806097e-05,
      "loss": 3.7672,
      "step": 537000
    },
    {
      "epoch": 372.0,
      "eval_loss": 3.6890618801116943,
      "eval_runtime": 74.1083,
      "eval_samples_per_second": 6644.652,
      "eval_steps_per_second": 2.172,
      "step": 537168
    },
    {
      "epoch": 372.23,
      "learning_rate": 1.2555401662049862e-05,
      "loss": 3.7629,
      "step": 537500
    },
    {
      "epoch": 372.58,
      "learning_rate": 1.254847645429363e-05,
      "loss": 3.7617,
      "step": 538000
    },
    {
      "epoch": 372.92,
      "learning_rate": 1.2541551246537397e-05,
      "loss": 3.7618,
      "step": 538500
    },
    {
      "epoch": 373.0,
      "eval_loss": 3.6865100860595703,
      "eval_runtime": 74.394,
      "eval_samples_per_second": 6619.133,
      "eval_steps_per_second": 2.164,
      "step": 538612
    },
    {
      "epoch": 373.27,
      "learning_rate": 1.2534626038781164e-05,
      "loss": 3.7647,
      "step": 539000
    },
    {
      "epoch": 373.61,
      "learning_rate": 1.2527700831024931e-05,
      "loss": 3.7625,
      "step": 539500
    },
    {
      "epoch": 373.96,
      "learning_rate": 1.25207756232687e-05,
      "loss": 3.7651,
      "step": 540000
    },
    {
      "epoch": 374.0,
      "eval_loss": 3.689401149749756,
      "eval_runtime": 73.8005,
      "eval_samples_per_second": 6672.368,
      "eval_steps_per_second": 2.182,
      "step": 540056
    },
    {
      "epoch": 374.31,
      "learning_rate": 1.2513850415512467e-05,
      "loss": 3.7643,
      "step": 540500
    },
    {
      "epoch": 374.65,
      "learning_rate": 1.2506925207756234e-05,
      "loss": 3.7638,
      "step": 541000
    },
    {
      "epoch": 375.0,
      "learning_rate": 1.25e-05,
      "loss": 3.7631,
      "step": 541500
    },
    {
      "epoch": 375.0,
      "eval_loss": 3.686310291290283,
      "eval_runtime": 73.7737,
      "eval_samples_per_second": 6674.791,
      "eval_steps_per_second": 2.182,
      "step": 541500
    },
    {
      "epoch": 375.35,
      "learning_rate": 1.249307479224377e-05,
      "loss": 3.7647,
      "step": 542000
    },
    {
      "epoch": 375.69,
      "learning_rate": 1.2486149584487534e-05,
      "loss": 3.7634,
      "step": 542500
    },
    {
      "epoch": 376.0,
      "eval_loss": 3.685969352722168,
      "eval_runtime": 75.4811,
      "eval_samples_per_second": 6523.809,
      "eval_steps_per_second": 2.133,
      "step": 542944
    },
    {
      "epoch": 376.04,
      "learning_rate": 1.2479224376731303e-05,
      "loss": 3.7621,
      "step": 543000
    },
    {
      "epoch": 376.39,
      "learning_rate": 1.247229916897507e-05,
      "loss": 3.7621,
      "step": 543500
    },
    {
      "epoch": 376.73,
      "learning_rate": 1.2465373961218839e-05,
      "loss": 3.7641,
      "step": 544000
    },
    {
      "epoch": 377.0,
      "eval_loss": 3.6860172748565674,
      "eval_runtime": 76.1045,
      "eval_samples_per_second": 6470.37,
      "eval_steps_per_second": 2.116,
      "step": 544388
    },
    {
      "epoch": 377.08,
      "learning_rate": 1.2458448753462604e-05,
      "loss": 3.7648,
      "step": 544500
    },
    {
      "epoch": 377.42,
      "learning_rate": 1.2451523545706372e-05,
      "loss": 3.7628,
      "step": 545000
    },
    {
      "epoch": 377.77,
      "learning_rate": 1.244459833795014e-05,
      "loss": 3.7637,
      "step": 545500
    },
    {
      "epoch": 378.0,
      "eval_loss": 3.6875898838043213,
      "eval_runtime": 74.9317,
      "eval_samples_per_second": 6571.636,
      "eval_steps_per_second": 2.149,
      "step": 545832
    },
    {
      "epoch": 378.12,
      "learning_rate": 1.2437673130193906e-05,
      "loss": 3.7598,
      "step": 546000
    },
    {
      "epoch": 378.46,
      "learning_rate": 1.2430747922437673e-05,
      "loss": 3.7616,
      "step": 546500
    },
    {
      "epoch": 378.81,
      "learning_rate": 1.2423822714681442e-05,
      "loss": 3.762,
      "step": 547000
    },
    {
      "epoch": 379.0,
      "eval_loss": 3.685751438140869,
      "eval_runtime": 73.6835,
      "eval_samples_per_second": 6682.965,
      "eval_steps_per_second": 2.185,
      "step": 547276
    },
    {
      "epoch": 379.16,
      "learning_rate": 1.241689750692521e-05,
      "loss": 3.7617,
      "step": 547500
    },
    {
      "epoch": 379.5,
      "learning_rate": 1.2409972299168976e-05,
      "loss": 3.7617,
      "step": 548000
    },
    {
      "epoch": 379.85,
      "learning_rate": 1.2403047091412743e-05,
      "loss": 3.7584,
      "step": 548500
    },
    {
      "epoch": 380.0,
      "eval_loss": 3.6871654987335205,
      "eval_runtime": 79.6972,
      "eval_samples_per_second": 6178.684,
      "eval_steps_per_second": 2.02,
      "step": 548720
    },
    {
      "epoch": 380.19,
      "learning_rate": 1.2396121883656511e-05,
      "loss": 3.7617,
      "step": 549000
    },
    {
      "epoch": 380.54,
      "learning_rate": 1.238919667590028e-05,
      "loss": 3.7601,
      "step": 549500
    },
    {
      "epoch": 380.89,
      "learning_rate": 1.2382271468144045e-05,
      "loss": 3.7591,
      "step": 550000
    },
    {
      "epoch": 381.0,
      "eval_loss": 3.685418128967285,
      "eval_runtime": 76.3267,
      "eval_samples_per_second": 6451.53,
      "eval_steps_per_second": 2.109,
      "step": 550164
    },
    {
      "epoch": 381.23,
      "learning_rate": 1.2375346260387812e-05,
      "loss": 3.763,
      "step": 550500
    },
    {
      "epoch": 381.58,
      "learning_rate": 1.236842105263158e-05,
      "loss": 3.7639,
      "step": 551000
    },
    {
      "epoch": 381.93,
      "learning_rate": 1.2361495844875346e-05,
      "loss": 3.7582,
      "step": 551500
    },
    {
      "epoch": 382.0,
      "eval_loss": 3.681150197982788,
      "eval_runtime": 74.94,
      "eval_samples_per_second": 6570.908,
      "eval_steps_per_second": 2.148,
      "step": 551608
    },
    {
      "epoch": 382.27,
      "learning_rate": 1.2354570637119114e-05,
      "loss": 3.761,
      "step": 552000
    },
    {
      "epoch": 382.62,
      "learning_rate": 1.2347645429362883e-05,
      "loss": 3.7575,
      "step": 552500
    },
    {
      "epoch": 382.96,
      "learning_rate": 1.234072022160665e-05,
      "loss": 3.7603,
      "step": 553000
    },
    {
      "epoch": 383.0,
      "eval_loss": 3.6880075931549072,
      "eval_runtime": 77.7382,
      "eval_samples_per_second": 6334.39,
      "eval_steps_per_second": 2.071,
      "step": 553052
    },
    {
      "epoch": 383.31,
      "learning_rate": 1.2333795013850415e-05,
      "loss": 3.7594,
      "step": 553500
    },
    {
      "epoch": 383.66,
      "learning_rate": 1.2326869806094184e-05,
      "loss": 3.7627,
      "step": 554000
    },
    {
      "epoch": 384.0,
      "eval_loss": 3.684797763824463,
      "eval_runtime": 75.6595,
      "eval_samples_per_second": 6508.425,
      "eval_steps_per_second": 2.128,
      "step": 554496
    },
    {
      "epoch": 384.0,
      "learning_rate": 1.2319944598337952e-05,
      "loss": 3.761,
      "step": 554500
    },
    {
      "epoch": 384.35,
      "learning_rate": 1.2313019390581718e-05,
      "loss": 3.7601,
      "step": 555000
    },
    {
      "epoch": 384.7,
      "learning_rate": 1.2306094182825486e-05,
      "loss": 3.7575,
      "step": 555500
    },
    {
      "epoch": 385.0,
      "eval_loss": 3.683274745941162,
      "eval_runtime": 77.7247,
      "eval_samples_per_second": 6335.486,
      "eval_steps_per_second": 2.071,
      "step": 555940
    },
    {
      "epoch": 385.04,
      "learning_rate": 1.2299168975069253e-05,
      "loss": 3.7608,
      "step": 556000
    },
    {
      "epoch": 385.39,
      "learning_rate": 1.2292243767313022e-05,
      "loss": 3.7583,
      "step": 556500
    },
    {
      "epoch": 385.73,
      "learning_rate": 1.2285318559556787e-05,
      "loss": 3.7615,
      "step": 557000
    },
    {
      "epoch": 386.0,
      "eval_loss": 3.684981346130371,
      "eval_runtime": 73.8671,
      "eval_samples_per_second": 6666.348,
      "eval_steps_per_second": 2.18,
      "step": 557384
    },
    {
      "epoch": 386.08,
      "learning_rate": 1.2278393351800556e-05,
      "loss": 3.7573,
      "step": 557500
    },
    {
      "epoch": 386.43,
      "learning_rate": 1.2271468144044322e-05,
      "loss": 3.7603,
      "step": 558000
    },
    {
      "epoch": 386.77,
      "learning_rate": 1.2264542936288088e-05,
      "loss": 3.7589,
      "step": 558500
    },
    {
      "epoch": 387.0,
      "eval_loss": 3.682325601577759,
      "eval_runtime": 75.6044,
      "eval_samples_per_second": 6513.162,
      "eval_steps_per_second": 2.13,
      "step": 558828
    },
    {
      "epoch": 387.12,
      "learning_rate": 1.2257617728531856e-05,
      "loss": 3.7553,
      "step": 559000
    },
    {
      "epoch": 387.47,
      "learning_rate": 1.2250692520775625e-05,
      "loss": 3.7576,
      "step": 559500
    },
    {
      "epoch": 387.81,
      "learning_rate": 1.2243767313019392e-05,
      "loss": 3.7596,
      "step": 560000
    },
    {
      "epoch": 388.0,
      "eval_loss": 3.6864383220672607,
      "eval_runtime": 74.4247,
      "eval_samples_per_second": 6616.409,
      "eval_steps_per_second": 2.163,
      "step": 560272
    },
    {
      "epoch": 388.16,
      "learning_rate": 1.2236842105263159e-05,
      "loss": 3.7583,
      "step": 560500
    },
    {
      "epoch": 388.5,
      "learning_rate": 1.2229916897506926e-05,
      "loss": 3.7577,
      "step": 561000
    },
    {
      "epoch": 388.85,
      "learning_rate": 1.2222991689750694e-05,
      "loss": 3.7575,
      "step": 561500
    },
    {
      "epoch": 389.0,
      "eval_loss": 3.680553674697876,
      "eval_runtime": 74.7418,
      "eval_samples_per_second": 6588.339,
      "eval_steps_per_second": 2.154,
      "step": 561716
    },
    {
      "epoch": 389.2,
      "learning_rate": 1.2216066481994461e-05,
      "loss": 3.7591,
      "step": 562000
    },
    {
      "epoch": 389.54,
      "learning_rate": 1.2209141274238228e-05,
      "loss": 3.7586,
      "step": 562500
    },
    {
      "epoch": 389.89,
      "learning_rate": 1.2202216066481995e-05,
      "loss": 3.7578,
      "step": 563000
    },
    {
      "epoch": 390.0,
      "eval_loss": 3.685154438018799,
      "eval_runtime": 73.6614,
      "eval_samples_per_second": 6684.968,
      "eval_steps_per_second": 2.186,
      "step": 563160
    },
    {
      "epoch": 390.24,
      "learning_rate": 1.2195290858725764e-05,
      "loss": 3.7573,
      "step": 563500
    },
    {
      "epoch": 390.58,
      "learning_rate": 1.2188365650969529e-05,
      "loss": 3.7584,
      "step": 564000
    },
    {
      "epoch": 390.93,
      "learning_rate": 1.2181440443213298e-05,
      "loss": 3.756,
      "step": 564500
    },
    {
      "epoch": 391.0,
      "eval_loss": 3.683967113494873,
      "eval_runtime": 74.0988,
      "eval_samples_per_second": 6645.504,
      "eval_steps_per_second": 2.173,
      "step": 564604
    },
    {
      "epoch": 391.27,
      "learning_rate": 1.2174515235457064e-05,
      "loss": 3.7563,
      "step": 565000
    },
    {
      "epoch": 391.62,
      "learning_rate": 1.2167590027700833e-05,
      "loss": 3.7564,
      "step": 565500
    },
    {
      "epoch": 391.97,
      "learning_rate": 1.2160664819944598e-05,
      "loss": 3.7574,
      "step": 566000
    },
    {
      "epoch": 392.0,
      "eval_loss": 3.6856420040130615,
      "eval_runtime": 74.3539,
      "eval_samples_per_second": 6622.705,
      "eval_steps_per_second": 2.165,
      "step": 566048
    },
    {
      "epoch": 392.31,
      "learning_rate": 1.2153739612188367e-05,
      "loss": 3.7559,
      "step": 566500
    },
    {
      "epoch": 392.66,
      "learning_rate": 1.2146814404432134e-05,
      "loss": 3.7576,
      "step": 567000
    },
    {
      "epoch": 393.0,
      "eval_loss": 3.68398118019104,
      "eval_runtime": 75.3631,
      "eval_samples_per_second": 6534.022,
      "eval_steps_per_second": 2.136,
      "step": 567492
    },
    {
      "epoch": 393.01,
      "learning_rate": 1.21398891966759e-05,
      "loss": 3.7578,
      "step": 567500
    },
    {
      "epoch": 393.35,
      "learning_rate": 1.2132963988919668e-05,
      "loss": 3.7571,
      "step": 568000
    },
    {
      "epoch": 393.7,
      "learning_rate": 1.2126038781163436e-05,
      "loss": 3.7559,
      "step": 568500
    },
    {
      "epoch": 394.0,
      "eval_loss": 3.6840593814849854,
      "eval_runtime": 75.017,
      "eval_samples_per_second": 6564.166,
      "eval_steps_per_second": 2.146,
      "step": 568936
    },
    {
      "epoch": 394.04,
      "learning_rate": 1.2119113573407205e-05,
      "loss": 3.758,
      "step": 569000
    },
    {
      "epoch": 394.39,
      "learning_rate": 1.211218836565097e-05,
      "loss": 3.7553,
      "step": 569500
    },
    {
      "epoch": 394.74,
      "learning_rate": 1.2105263157894737e-05,
      "loss": 3.7552,
      "step": 570000
    },
    {
      "epoch": 395.0,
      "eval_loss": 3.6834726333618164,
      "eval_runtime": 78.0649,
      "eval_samples_per_second": 6307.88,
      "eval_steps_per_second": 2.062,
      "step": 570380
    },
    {
      "epoch": 395.08,
      "learning_rate": 1.2098337950138506e-05,
      "loss": 3.7579,
      "step": 570500
    },
    {
      "epoch": 395.43,
      "learning_rate": 1.2091412742382271e-05,
      "loss": 3.7579,
      "step": 571000
    },
    {
      "epoch": 395.78,
      "learning_rate": 1.208448753462604e-05,
      "loss": 3.7547,
      "step": 571500
    },
    {
      "epoch": 396.0,
      "eval_loss": 3.6800179481506348,
      "eval_runtime": 73.5113,
      "eval_samples_per_second": 6698.614,
      "eval_steps_per_second": 2.19,
      "step": 571824
    },
    {
      "epoch": 396.12,
      "learning_rate": 1.2077562326869806e-05,
      "loss": 3.7539,
      "step": 572000
    },
    {
      "epoch": 396.47,
      "learning_rate": 1.2070637119113575e-05,
      "loss": 3.7574,
      "step": 572500
    },
    {
      "epoch": 396.81,
      "learning_rate": 1.206371191135734e-05,
      "loss": 3.7572,
      "step": 573000
    },
    {
      "epoch": 397.0,
      "eval_loss": 3.678455352783203,
      "eval_runtime": 74.8356,
      "eval_samples_per_second": 6580.081,
      "eval_steps_per_second": 2.151,
      "step": 573268
    },
    {
      "epoch": 397.16,
      "learning_rate": 1.2056786703601109e-05,
      "loss": 3.7534,
      "step": 573500
    },
    {
      "epoch": 397.51,
      "learning_rate": 1.2049861495844877e-05,
      "loss": 3.7549,
      "step": 574000
    },
    {
      "epoch": 397.85,
      "learning_rate": 1.2042936288088644e-05,
      "loss": 3.7546,
      "step": 574500
    },
    {
      "epoch": 398.0,
      "eval_loss": 3.6750450134277344,
      "eval_runtime": 73.6423,
      "eval_samples_per_second": 6686.702,
      "eval_steps_per_second": 2.186,
      "step": 574712
    },
    {
      "epoch": 398.2,
      "learning_rate": 1.203601108033241e-05,
      "loss": 3.7518,
      "step": 575000
    },
    {
      "epoch": 398.55,
      "learning_rate": 1.2029085872576178e-05,
      "loss": 3.7571,
      "step": 575500
    },
    {
      "epoch": 398.89,
      "learning_rate": 1.2022160664819947e-05,
      "loss": 3.7555,
      "step": 576000
    },
    {
      "epoch": 399.0,
      "eval_loss": 3.678826093673706,
      "eval_runtime": 76.3543,
      "eval_samples_per_second": 6449.196,
      "eval_steps_per_second": 2.109,
      "step": 576156
    },
    {
      "epoch": 399.24,
      "learning_rate": 1.2015235457063712e-05,
      "loss": 3.7544,
      "step": 576500
    },
    {
      "epoch": 399.58,
      "learning_rate": 1.200831024930748e-05,
      "loss": 3.7527,
      "step": 577000
    },
    {
      "epoch": 399.93,
      "learning_rate": 1.2001385041551248e-05,
      "loss": 3.7565,
      "step": 577500
    },
    {
      "epoch": 400.0,
      "eval_loss": 3.6734468936920166,
      "eval_runtime": 74.6685,
      "eval_samples_per_second": 6594.802,
      "eval_steps_per_second": 2.156,
      "step": 577600
    },
    {
      "epoch": 400.28,
      "learning_rate": 1.1994459833795016e-05,
      "loss": 3.7545,
      "step": 578000
    },
    {
      "epoch": 400.62,
      "learning_rate": 1.1987534626038781e-05,
      "loss": 3.754,
      "step": 578500
    },
    {
      "epoch": 400.97,
      "learning_rate": 1.198060941828255e-05,
      "loss": 3.754,
      "step": 579000
    },
    {
      "epoch": 401.0,
      "eval_loss": 3.6801302433013916,
      "eval_runtime": 75.4548,
      "eval_samples_per_second": 6526.079,
      "eval_steps_per_second": 2.134,
      "step": 579044
    },
    {
      "epoch": 401.32,
      "learning_rate": 1.1973684210526317e-05,
      "loss": 3.7545,
      "step": 579500
    },
    {
      "epoch": 401.66,
      "learning_rate": 1.1966759002770084e-05,
      "loss": 3.7534,
      "step": 580000
    },
    {
      "epoch": 402.0,
      "eval_loss": 3.6784117221832275,
      "eval_runtime": 74.9008,
      "eval_samples_per_second": 6574.347,
      "eval_steps_per_second": 2.15,
      "step": 580488
    },
    {
      "epoch": 402.01,
      "learning_rate": 1.195983379501385e-05,
      "loss": 3.7546,
      "step": 580500
    },
    {
      "epoch": 402.35,
      "learning_rate": 1.195290858725762e-05,
      "loss": 3.7524,
      "step": 581000
    },
    {
      "epoch": 402.7,
      "learning_rate": 1.1945983379501386e-05,
      "loss": 3.7499,
      "step": 581500
    },
    {
      "epoch": 403.0,
      "eval_loss": 3.6806490421295166,
      "eval_runtime": 77.3538,
      "eval_samples_per_second": 6365.865,
      "eval_steps_per_second": 2.081,
      "step": 581932
    },
    {
      "epoch": 403.05,
      "learning_rate": 1.1939058171745153e-05,
      "loss": 3.7531,
      "step": 582000
    },
    {
      "epoch": 403.39,
      "learning_rate": 1.193213296398892e-05,
      "loss": 3.7501,
      "step": 582500
    },
    {
      "epoch": 403.74,
      "learning_rate": 1.1925207756232689e-05,
      "loss": 3.7507,
      "step": 583000
    },
    {
      "epoch": 404.0,
      "eval_loss": 3.6763315200805664,
      "eval_runtime": 74.1317,
      "eval_samples_per_second": 6642.552,
      "eval_steps_per_second": 2.172,
      "step": 583376
    },
    {
      "epoch": 404.09,
      "learning_rate": 1.1918282548476454e-05,
      "loss": 3.7527,
      "step": 583500
    },
    {
      "epoch": 404.43,
      "learning_rate": 1.1911357340720223e-05,
      "loss": 3.7528,
      "step": 584000
    },
    {
      "epoch": 404.78,
      "learning_rate": 1.190443213296399e-05,
      "loss": 3.7485,
      "step": 584500
    },
    {
      "epoch": 405.0,
      "eval_loss": 3.6807456016540527,
      "eval_runtime": 76.3544,
      "eval_samples_per_second": 6449.191,
      "eval_steps_per_second": 2.109,
      "step": 584820
    },
    {
      "epoch": 405.12,
      "learning_rate": 1.1897506925207758e-05,
      "loss": 3.7503,
      "step": 585000
    },
    {
      "epoch": 405.47,
      "learning_rate": 1.1890581717451523e-05,
      "loss": 3.7524,
      "step": 585500
    },
    {
      "epoch": 405.82,
      "learning_rate": 1.1883656509695292e-05,
      "loss": 3.7516,
      "step": 586000
    },
    {
      "epoch": 406.0,
      "eval_loss": 3.6764273643493652,
      "eval_runtime": 75.8758,
      "eval_samples_per_second": 6489.866,
      "eval_steps_per_second": 2.122,
      "step": 586264
    },
    {
      "epoch": 406.16,
      "learning_rate": 1.1876731301939059e-05,
      "loss": 3.7525,
      "step": 586500
    },
    {
      "epoch": 406.51,
      "learning_rate": 1.1869806094182828e-05,
      "loss": 3.7519,
      "step": 587000
    },
    {
      "epoch": 406.86,
      "learning_rate": 1.1862880886426593e-05,
      "loss": 3.7544,
      "step": 587500
    },
    {
      "epoch": 407.0,
      "eval_loss": 3.679368495941162,
      "eval_runtime": 73.5888,
      "eval_samples_per_second": 6691.56,
      "eval_steps_per_second": 2.188,
      "step": 587708
    },
    {
      "epoch": 407.2,
      "learning_rate": 1.1855955678670361e-05,
      "loss": 3.7496,
      "step": 588000
    },
    {
      "epoch": 407.55,
      "learning_rate": 1.1849030470914128e-05,
      "loss": 3.7509,
      "step": 588500
    },
    {
      "epoch": 407.89,
      "learning_rate": 1.1842105263157895e-05,
      "loss": 3.7532,
      "step": 589000
    },
    {
      "epoch": 408.0,
      "eval_loss": 3.678314208984375,
      "eval_runtime": 73.6357,
      "eval_samples_per_second": 6687.301,
      "eval_steps_per_second": 2.186,
      "step": 589152
    },
    {
      "epoch": 408.24,
      "learning_rate": 1.1835180055401662e-05,
      "loss": 3.7535,
      "step": 589500
    },
    {
      "epoch": 408.59,
      "learning_rate": 1.182825484764543e-05,
      "loss": 3.7506,
      "step": 590000
    },
    {
      "epoch": 408.93,
      "learning_rate": 1.18213296398892e-05,
      "loss": 3.7532,
      "step": 590500
    },
    {
      "epoch": 409.0,
      "eval_loss": 3.676326036453247,
      "eval_runtime": 73.8096,
      "eval_samples_per_second": 6671.541,
      "eval_steps_per_second": 2.181,
      "step": 590596
    },
    {
      "epoch": 409.28,
      "learning_rate": 1.1814404432132965e-05,
      "loss": 3.7545,
      "step": 591000
    },
    {
      "epoch": 409.63,
      "learning_rate": 1.1807479224376732e-05,
      "loss": 3.7496,
      "step": 591500
    },
    {
      "epoch": 409.97,
      "learning_rate": 1.18005540166205e-05,
      "loss": 3.7494,
      "step": 592000
    },
    {
      "epoch": 410.0,
      "eval_loss": 3.6734423637390137,
      "eval_runtime": 73.4761,
      "eval_samples_per_second": 6701.822,
      "eval_steps_per_second": 2.191,
      "step": 592040
    },
    {
      "epoch": 410.32,
      "learning_rate": 1.1793628808864265e-05,
      "loss": 3.7547,
      "step": 592500
    },
    {
      "epoch": 410.66,
      "learning_rate": 1.1786703601108034e-05,
      "loss": 3.7501,
      "step": 593000
    },
    {
      "epoch": 411.0,
      "eval_loss": 3.6773502826690674,
      "eval_runtime": 75.1449,
      "eval_samples_per_second": 6552.994,
      "eval_steps_per_second": 2.143,
      "step": 593484
    },
    {
      "epoch": 411.01,
      "learning_rate": 1.1779778393351803e-05,
      "loss": 3.7512,
      "step": 593500
    },
    {
      "epoch": 411.36,
      "learning_rate": 1.177285318559557e-05,
      "loss": 3.7481,
      "step": 594000
    },
    {
      "epoch": 411.7,
      "learning_rate": 1.1765927977839335e-05,
      "loss": 3.7523,
      "step": 594500
    },
    {
      "epoch": 412.0,
      "eval_loss": 3.6761634349823,
      "eval_runtime": 74.2595,
      "eval_samples_per_second": 6631.127,
      "eval_steps_per_second": 2.168,
      "step": 594928
    },
    {
      "epoch": 412.05,
      "learning_rate": 1.1759002770083103e-05,
      "loss": 3.7488,
      "step": 595000
    },
    {
      "epoch": 412.4,
      "learning_rate": 1.1752077562326872e-05,
      "loss": 3.7468,
      "step": 595500
    },
    {
      "epoch": 412.74,
      "learning_rate": 1.1745152354570639e-05,
      "loss": 3.7527,
      "step": 596000
    },
    {
      "epoch": 413.0,
      "eval_loss": 3.677605390548706,
      "eval_runtime": 75.7462,
      "eval_samples_per_second": 6500.971,
      "eval_steps_per_second": 2.126,
      "step": 596372
    },
    {
      "epoch": 413.09,
      "learning_rate": 1.1738227146814404e-05,
      "loss": 3.7511,
      "step": 596500
    },
    {
      "epoch": 413.43,
      "learning_rate": 1.1731301939058173e-05,
      "loss": 3.7485,
      "step": 597000
    },
    {
      "epoch": 413.78,
      "learning_rate": 1.1724376731301941e-05,
      "loss": 3.751,
      "step": 597500
    },
    {
      "epoch": 414.0,
      "eval_loss": 3.6766717433929443,
      "eval_runtime": 73.9242,
      "eval_samples_per_second": 6661.202,
      "eval_steps_per_second": 2.178,
      "step": 597816
    },
    {
      "epoch": 414.13,
      "learning_rate": 1.1717451523545707e-05,
      "loss": 3.7492,
      "step": 598000
    },
    {
      "epoch": 414.47,
      "learning_rate": 1.1710526315789475e-05,
      "loss": 3.749,
      "step": 598500
    },
    {
      "epoch": 414.82,
      "learning_rate": 1.1703601108033242e-05,
      "loss": 3.7475,
      "step": 599000
    },
    {
      "epoch": 415.0,
      "eval_loss": 3.672994375228882,
      "eval_runtime": 75.3605,
      "eval_samples_per_second": 6534.25,
      "eval_steps_per_second": 2.136,
      "step": 599260
    },
    {
      "epoch": 415.17,
      "learning_rate": 1.169667590027701e-05,
      "loss": 3.7479,
      "step": 599500
    },
    {
      "epoch": 415.51,
      "learning_rate": 1.1689750692520776e-05,
      "loss": 3.7481,
      "step": 600000
    },
    {
      "epoch": 415.86,
      "learning_rate": 1.1682825484764545e-05,
      "loss": 3.7493,
      "step": 600500
    },
    {
      "epoch": 416.0,
      "eval_loss": 3.6774232387542725,
      "eval_runtime": 74.4096,
      "eval_samples_per_second": 6617.752,
      "eval_steps_per_second": 2.164,
      "step": 600704
    },
    {
      "epoch": 416.2,
      "learning_rate": 1.1675900277008311e-05,
      "loss": 3.7479,
      "step": 601000
    },
    {
      "epoch": 416.55,
      "learning_rate": 1.1668975069252078e-05,
      "loss": 3.7478,
      "step": 601500
    },
    {
      "epoch": 416.9,
      "learning_rate": 1.1662049861495845e-05,
      "loss": 3.7468,
      "step": 602000
    },
    {
      "epoch": 417.0,
      "eval_loss": 3.6730477809906006,
      "eval_runtime": 77.6666,
      "eval_samples_per_second": 6340.227,
      "eval_steps_per_second": 2.073,
      "step": 602148
    },
    {
      "epoch": 417.24,
      "learning_rate": 1.1655124653739614e-05,
      "loss": 3.7497,
      "step": 602500
    },
    {
      "epoch": 417.59,
      "learning_rate": 1.164819944598338e-05,
      "loss": 3.7491,
      "step": 603000
    },
    {
      "epoch": 417.94,
      "learning_rate": 1.1641274238227148e-05,
      "loss": 3.7457,
      "step": 603500
    },
    {
      "epoch": 418.0,
      "eval_loss": 3.6768581867218018,
      "eval_runtime": 74.7091,
      "eval_samples_per_second": 6591.216,
      "eval_steps_per_second": 2.155,
      "step": 603592
    },
    {
      "epoch": 418.28,
      "learning_rate": 1.1634349030470915e-05,
      "loss": 3.746,
      "step": 604000
    },
    {
      "epoch": 418.63,
      "learning_rate": 1.1627423822714683e-05,
      "loss": 3.7471,
      "step": 604500
    },
    {
      "epoch": 418.98,
      "learning_rate": 1.1620498614958448e-05,
      "loss": 3.749,
      "step": 605000
    },
    {
      "epoch": 419.0,
      "eval_loss": 3.674077272415161,
      "eval_runtime": 74.7647,
      "eval_samples_per_second": 6586.318,
      "eval_steps_per_second": 2.153,
      "step": 605036
    },
    {
      "epoch": 419.32,
      "learning_rate": 1.1613573407202217e-05,
      "loss": 3.7479,
      "step": 605500
    },
    {
      "epoch": 419.67,
      "learning_rate": 1.1606648199445984e-05,
      "loss": 3.7512,
      "step": 606000
    },
    {
      "epoch": 420.0,
      "eval_loss": 3.6696927547454834,
      "eval_runtime": 74.4025,
      "eval_samples_per_second": 6618.38,
      "eval_steps_per_second": 2.164,
      "step": 606480
    },
    {
      "epoch": 420.01,
      "learning_rate": 1.1599722991689753e-05,
      "loss": 3.7462,
      "step": 606500
    },
    {
      "epoch": 420.36,
      "learning_rate": 1.1592797783933518e-05,
      "loss": 3.7487,
      "step": 607000
    },
    {
      "epoch": 420.71,
      "learning_rate": 1.1585872576177286e-05,
      "loss": 3.7491,
      "step": 607500
    },
    {
      "epoch": 421.0,
      "eval_loss": 3.670689105987549,
      "eval_runtime": 73.7952,
      "eval_samples_per_second": 6672.842,
      "eval_steps_per_second": 2.182,
      "step": 607924
    },
    {
      "epoch": 421.05,
      "learning_rate": 1.1578947368421053e-05,
      "loss": 3.7472,
      "step": 608000
    },
    {
      "epoch": 421.4,
      "learning_rate": 1.1572022160664822e-05,
      "loss": 3.7462,
      "step": 608500
    },
    {
      "epoch": 421.75,
      "learning_rate": 1.1565096952908587e-05,
      "loss": 3.7464,
      "step": 609000
    },
    {
      "epoch": 422.0,
      "eval_loss": 3.6667892932891846,
      "eval_runtime": 75.7682,
      "eval_samples_per_second": 6499.083,
      "eval_steps_per_second": 2.125,
      "step": 609368
    },
    {
      "epoch": 422.09,
      "learning_rate": 1.1558171745152356e-05,
      "loss": 3.7461,
      "step": 609500
    },
    {
      "epoch": 422.44,
      "learning_rate": 1.1551246537396123e-05,
      "loss": 3.7467,
      "step": 610000
    },
    {
      "epoch": 422.78,
      "learning_rate": 1.154432132963989e-05,
      "loss": 3.7476,
      "step": 610500
    },
    {
      "epoch": 423.0,
      "eval_loss": 3.6749465465545654,
      "eval_runtime": 74.4127,
      "eval_samples_per_second": 6617.471,
      "eval_steps_per_second": 2.164,
      "step": 610812
    },
    {
      "epoch": 423.13,
      "learning_rate": 1.1537396121883657e-05,
      "loss": 3.7449,
      "step": 611000
    },
    {
      "epoch": 423.48,
      "learning_rate": 1.1530470914127425e-05,
      "loss": 3.7459,
      "step": 611500
    },
    {
      "epoch": 423.82,
      "learning_rate": 1.1523545706371194e-05,
      "loss": 3.7478,
      "step": 612000
    },
    {
      "epoch": 424.0,
      "eval_loss": 3.673068046569824,
      "eval_runtime": 76.5927,
      "eval_samples_per_second": 6429.124,
      "eval_steps_per_second": 2.102,
      "step": 612256
    },
    {
      "epoch": 424.17,
      "learning_rate": 1.1516620498614959e-05,
      "loss": 3.7464,
      "step": 612500
    },
    {
      "epoch": 424.52,
      "learning_rate": 1.1509695290858726e-05,
      "loss": 3.7449,
      "step": 613000
    },
    {
      "epoch": 424.86,
      "learning_rate": 1.1502770083102495e-05,
      "loss": 3.7474,
      "step": 613500
    },
    {
      "epoch": 425.0,
      "eval_loss": 3.6703741550445557,
      "eval_runtime": 75.4887,
      "eval_samples_per_second": 6523.149,
      "eval_steps_per_second": 2.133,
      "step": 613700
    },
    {
      "epoch": 425.21,
      "learning_rate": 1.149584487534626e-05,
      "loss": 3.7467,
      "step": 614000
    },
    {
      "epoch": 425.55,
      "learning_rate": 1.1488919667590028e-05,
      "loss": 3.7443,
      "step": 614500
    },
    {
      "epoch": 425.9,
      "learning_rate": 1.1481994459833797e-05,
      "loss": 3.7466,
      "step": 615000
    },
    {
      "epoch": 426.0,
      "eval_loss": 3.6724741458892822,
      "eval_runtime": 77.1998,
      "eval_samples_per_second": 6378.57,
      "eval_steps_per_second": 2.085,
      "step": 615144
    },
    {
      "epoch": 426.25,
      "learning_rate": 1.1475069252077564e-05,
      "loss": 3.7465,
      "step": 615500
    },
    {
      "epoch": 426.59,
      "learning_rate": 1.146814404432133e-05,
      "loss": 3.7436,
      "step": 616000
    },
    {
      "epoch": 426.94,
      "learning_rate": 1.1461218836565098e-05,
      "loss": 3.7468,
      "step": 616500
    },
    {
      "epoch": 427.0,
      "eval_loss": 3.67818021774292,
      "eval_runtime": 76.3088,
      "eval_samples_per_second": 6453.046,
      "eval_steps_per_second": 2.11,
      "step": 616588
    },
    {
      "epoch": 427.29,
      "learning_rate": 1.1454293628808866e-05,
      "loss": 3.7458,
      "step": 617000
    },
    {
      "epoch": 427.63,
      "learning_rate": 1.1447368421052632e-05,
      "loss": 3.7447,
      "step": 617500
    },
    {
      "epoch": 427.98,
      "learning_rate": 1.1440443213296399e-05,
      "loss": 3.7472,
      "step": 618000
    },
    {
      "epoch": 428.0,
      "eval_loss": 3.6692676544189453,
      "eval_runtime": 77.2341,
      "eval_samples_per_second": 6375.732,
      "eval_steps_per_second": 2.085,
      "step": 618032
    },
    {
      "epoch": 428.32,
      "learning_rate": 1.1433518005540167e-05,
      "loss": 3.7432,
      "step": 618500
    },
    {
      "epoch": 428.67,
      "learning_rate": 1.1426592797783936e-05,
      "loss": 3.7454,
      "step": 619000
    },
    {
      "epoch": 429.0,
      "eval_loss": 3.6682395935058594,
      "eval_runtime": 81.127,
      "eval_samples_per_second": 6069.794,
      "eval_steps_per_second": 1.985,
      "step": 619476
    },
    {
      "epoch": 429.02,
      "learning_rate": 1.1419667590027701e-05,
      "loss": 3.7452,
      "step": 619500
    },
    {
      "epoch": 429.36,
      "learning_rate": 1.141274238227147e-05,
      "loss": 3.7429,
      "step": 620000
    },
    {
      "epoch": 429.71,
      "learning_rate": 1.1405817174515237e-05,
      "loss": 3.7424,
      "step": 620500
    },
    {
      "epoch": 430.0,
      "eval_loss": 3.669628858566284,
      "eval_runtime": 73.5755,
      "eval_samples_per_second": 6692.774,
      "eval_steps_per_second": 2.188,
      "step": 620920
    },
    {
      "epoch": 430.06,
      "learning_rate": 1.1398891966759005e-05,
      "loss": 3.7433,
      "step": 621000
    },
    {
      "epoch": 430.4,
      "learning_rate": 1.139196675900277e-05,
      "loss": 3.7426,
      "step": 621500
    },
    {
      "epoch": 430.75,
      "learning_rate": 1.1385041551246539e-05,
      "loss": 3.7435,
      "step": 622000
    },
    {
      "epoch": 431.0,
      "eval_loss": 3.6648757457733154,
      "eval_runtime": 77.6023,
      "eval_samples_per_second": 6345.481,
      "eval_steps_per_second": 2.075,
      "step": 622364
    },
    {
      "epoch": 431.09,
      "learning_rate": 1.1378116343490306e-05,
      "loss": 3.7437,
      "step": 622500
    },
    {
      "epoch": 431.44,
      "learning_rate": 1.1371191135734073e-05,
      "loss": 3.7417,
      "step": 623000
    },
    {
      "epoch": 431.79,
      "learning_rate": 1.136426592797784e-05,
      "loss": 3.7441,
      "step": 623500
    },
    {
      "epoch": 432.0,
      "eval_loss": 3.6664254665374756,
      "eval_runtime": 74.4125,
      "eval_samples_per_second": 6617.491,
      "eval_steps_per_second": 2.164,
      "step": 623808
    },
    {
      "epoch": 432.13,
      "learning_rate": 1.1357340720221608e-05,
      "loss": 3.7417,
      "step": 624000
    },
    {
      "epoch": 432.48,
      "learning_rate": 1.1350415512465375e-05,
      "loss": 3.7433,
      "step": 624500
    },
    {
      "epoch": 432.83,
      "learning_rate": 1.1343490304709142e-05,
      "loss": 3.7415,
      "step": 625000
    },
    {
      "epoch": 433.0,
      "eval_loss": 3.6667163372039795,
      "eval_runtime": 73.9564,
      "eval_samples_per_second": 6658.305,
      "eval_steps_per_second": 2.177,
      "step": 625252
    },
    {
      "epoch": 433.17,
      "learning_rate": 1.1336565096952909e-05,
      "loss": 3.7433,
      "step": 625500
    },
    {
      "epoch": 433.52,
      "learning_rate": 1.1329639889196678e-05,
      "loss": 3.7397,
      "step": 626000
    },
    {
      "epoch": 433.86,
      "learning_rate": 1.1322714681440443e-05,
      "loss": 3.7418,
      "step": 626500
    },
    {
      "epoch": 434.0,
      "eval_loss": 3.6659815311431885,
      "eval_runtime": 79.156,
      "eval_samples_per_second": 6220.929,
      "eval_steps_per_second": 2.034,
      "step": 626696
    },
    {
      "epoch": 434.21,
      "learning_rate": 1.1315789473684212e-05,
      "loss": 3.7422,
      "step": 627000
    },
    {
      "epoch": 434.56,
      "learning_rate": 1.1308864265927979e-05,
      "loss": 3.7429,
      "step": 627500
    },
    {
      "epoch": 434.9,
      "learning_rate": 1.1301939058171747e-05,
      "loss": 3.7442,
      "step": 628000
    },
    {
      "epoch": 435.0,
      "eval_loss": 3.669684886932373,
      "eval_runtime": 75.7717,
      "eval_samples_per_second": 6498.787,
      "eval_steps_per_second": 2.125,
      "step": 628140
    },
    {
      "epoch": 435.25,
      "learning_rate": 1.1295013850415512e-05,
      "loss": 3.7411,
      "step": 628500
    },
    {
      "epoch": 435.6,
      "learning_rate": 1.1288088642659281e-05,
      "loss": 3.7416,
      "step": 629000
    },
    {
      "epoch": 435.94,
      "learning_rate": 1.1281163434903048e-05,
      "loss": 3.7423,
      "step": 629500
    },
    {
      "epoch": 436.0,
      "eval_loss": 3.667738914489746,
      "eval_runtime": 75.7672,
      "eval_samples_per_second": 6499.168,
      "eval_steps_per_second": 2.125,
      "step": 629584
    },
    {
      "epoch": 436.29,
      "learning_rate": 1.1274238227146815e-05,
      "loss": 3.7399,
      "step": 630000
    }
  ],
  "max_steps": 1444000,
  "num_train_epochs": 1000,
  "total_flos": 2.9616940967712937e+18,
  "trial_name": null,
  "trial_params": null
}