{
  "best_metric": 5.1911420822143555,
  "best_model_checkpoint": "../checkpoints/nf-distilbart_6_6-rqnsf-lagtrain-augmented/checkpoint-212000",
  "epoch": 2.953510079549729,
  "global_step": 212000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03,
      "gate_score": 0.5014,
      "learning_rate": 0.0,
      "loss": 611.7,
      "nf_loss": 606.6464,
      "ppl": 183.8572,
      "step": 2000
    },
    {
      "epoch": 0.06,
      "gate_score": 0.4744,
      "learning_rate": 0.0,
      "loss": 379.5551,
      "nf_loss": 377.2968,
      "ppl": 13.9977,
      "step": 4000
    },
    {
      "epoch": 0.06,
      "eval_loss": 303.72564697265625,
      "eval_nf_loss": 301.4253845214844,
      "eval_perplexity": 10.682913780212402,
      "eval_runtime": 551.5576,
      "eval_samples_per_second": 24.237,
      "eval_steps_per_second": 3.03,
      "step": 4000
    },
    {
      "epoch": 0.08,
      "gate_score": 0.4514,
      "learning_rate": 0.0,
      "loss": 249.1765,
      "nf_loss": 247.7483,
      "ppl": 4.6289,
      "step": 6000
    },
    {
      "epoch": 0.11,
      "gate_score": 0.434,
      "learning_rate": 0.0,
      "loss": 169.2218,
      "nf_loss": 167.9433,
      "ppl": 3.9621,
      "step": 8000
    },
    {
      "epoch": 0.11,
      "eval_loss": 138.2334747314453,
      "eval_nf_loss": 136.36793518066406,
      "eval_perplexity": 6.839929103851318,
      "eval_runtime": 557.5174,
      "eval_samples_per_second": 23.978,
      "eval_steps_per_second": 2.997,
      "step": 8000
    },
    {
      "epoch": 0.14,
      "gate_score": 0.4184,
      "learning_rate": 0.0,
      "loss": 120.9341,
      "nf_loss": 119.7347,
      "ppl": 3.6249,
      "step": 10000
    },
    {
      "epoch": 0.17,
      "gate_score": 0.4011,
      "learning_rate": 0.0,
      "loss": 83.5963,
      "nf_loss": 82.4239,
      "ppl": 3.5277,
      "step": 12000
    },
    {
      "epoch": 0.17,
      "eval_loss": 65.74690246582031,
      "eval_nf_loss": 63.977783203125,
      "eval_perplexity": 6.194242477416992,
      "eval_runtime": 562.6949,
      "eval_samples_per_second": 23.757,
      "eval_steps_per_second": 2.97,
      "step": 12000
    },
    {
      "epoch": 0.2,
      "gate_score": 0.3846,
      "learning_rate": 0.0,
      "loss": 61.2072,
      "nf_loss": 60.0426,
      "ppl": 3.5018,
      "step": 14000
    },
    {
      "epoch": 0.22,
      "gate_score": 0.3634,
      "learning_rate": 0.0,
      "loss": 47.2789,
      "nf_loss": 46.1231,
      "ppl": 3.4605,
      "step": 16000
    },
    {
      "epoch": 0.22,
      "eval_loss": 40.041507720947266,
      "eval_nf_loss": 38.2694206237793,
      "eval_perplexity": 6.216037750244141,
      "eval_runtime": 547.0037,
      "eval_samples_per_second": 24.439,
      "eval_steps_per_second": 3.055,
      "step": 16000
    },
    {
      "epoch": 0.25,
      "gate_score": 0.3434,
      "learning_rate": 0.0,
      "loss": 37.6218,
      "nf_loss": 36.4738,
      "ppl": 3.4471,
      "step": 18000
    },
    {
      "epoch": 0.28,
      "gate_score": 0.3263,
      "learning_rate": 0.0,
      "loss": 29.1509,
      "nf_loss": 27.9897,
      "ppl": 3.4793,
      "step": 20000
    },
    {
      "epoch": 0.28,
      "eval_loss": 25.114910125732422,
      "eval_nf_loss": 23.323299407958984,
      "eval_perplexity": 6.345096111297607,
      "eval_runtime": 555.7818,
      "eval_samples_per_second": 24.053,
      "eval_steps_per_second": 3.007,
      "step": 20000
    },
    {
      "epoch": 0.31,
      "gate_score": 0.3067,
      "learning_rate": 0.0,
      "loss": 24.9587,
      "nf_loss": 23.7908,
      "ppl": 3.5116,
      "step": 22000
    },
    {
      "epoch": 0.33,
      "gate_score": 0.2877,
      "learning_rate": 0.0,
      "loss": 21.3109,
      "nf_loss": 20.1419,
      "ppl": 3.4998,
      "step": 24000
    },
    {
      "epoch": 0.33,
      "eval_loss": 19.053028106689453,
      "eval_nf_loss": 17.251123428344727,
      "eval_perplexity": 6.4101433753967285,
      "eval_runtime": 555.0721,
      "eval_samples_per_second": 24.083,
      "eval_steps_per_second": 3.01,
      "step": 24000
    },
    {
      "epoch": 0.36,
      "gate_score": 0.2675,
      "learning_rate": 0.0,
      "loss": 19.1247,
      "nf_loss": 17.9351,
      "ppl": 3.5921,
      "step": 26000
    },
    {
      "epoch": 0.39,
      "gate_score": 0.2498,
      "learning_rate": 0.0,
      "loss": 17.8051,
      "nf_loss": 16.637,
      "ppl": 3.5149,
      "step": 28000
    },
    {
      "epoch": 0.39,
      "eval_loss": 16.122459411621094,
      "eval_nf_loss": 14.319244384765625,
      "eval_perplexity": 6.42343282699585,
      "eval_runtime": 535.5188,
      "eval_samples_per_second": 24.963,
      "eval_steps_per_second": 3.12,
      "step": 28000
    },
    {
      "epoch": 0.42,
      "gate_score": 0.2419,
      "learning_rate": 0.0,
      "loss": 16.0983,
      "nf_loss": 14.8927,
      "ppl": 3.652,
      "step": 30000
    },
    {
      "epoch": 0.45,
      "gate_score": 0.2315,
      "learning_rate": 0.0,
      "loss": 15.1481,
      "nf_loss": 13.9579,
      "ppl": 3.6092,
      "step": 32000
    },
    {
      "epoch": 0.45,
      "eval_loss": 14.115809440612793,
      "eval_nf_loss": 12.312360763549805,
      "eval_perplexity": 6.42423677444458,
      "eval_runtime": 546.0274,
      "eval_samples_per_second": 24.482,
      "eval_steps_per_second": 3.06,
      "step": 32000
    },
    {
      "epoch": 0.47,
      "gate_score": 0.218,
      "learning_rate": 0.0,
      "loss": 14.0114,
      "nf_loss": 12.8229,
      "ppl": 3.6075,
      "step": 34000
    },
    {
      "epoch": 0.5,
      "gate_score": 0.2064,
      "learning_rate": 0.0,
      "loss": 13.2317,
      "nf_loss": 12.0527,
      "ppl": 3.5676,
      "step": 36000
    },
    {
      "epoch": 0.5,
      "eval_loss": 12.555018424987793,
      "eval_nf_loss": 10.7388916015625,
      "eval_perplexity": 6.519784927368164,
      "eval_runtime": 551.5879,
      "eval_samples_per_second": 24.235,
      "eval_steps_per_second": 3.029,
      "step": 36000
    },
    {
      "epoch": 0.53,
      "gate_score": 0.198,
      "learning_rate": 0.0,
      "loss": 12.6385,
      "nf_loss": 11.4663,
      "ppl": 3.5302,
      "step": 38000
    },
    {
      "epoch": 0.56,
      "gate_score": 0.1943,
      "learning_rate": 0.0,
      "loss": 12.0824,
      "nf_loss": 10.929,
      "ppl": 3.4804,
      "step": 40000
    },
    {
      "epoch": 0.56,
      "eval_loss": 11.536130905151367,
      "eval_nf_loss": 9.7628755569458,
      "eval_perplexity": 6.218123435974121,
      "eval_runtime": 552.328,
      "eval_samples_per_second": 24.203,
      "eval_steps_per_second": 3.025,
      "step": 40000
    },
    {
      "epoch": 0.59,
      "gate_score": 0.1868,
      "learning_rate": 0.0,
      "loss": 11.6608,
      "nf_loss": 10.5157,
      "ppl": 3.433,
      "step": 42000
    },
    {
      "epoch": 0.61,
      "gate_score": 0.18,
      "learning_rate": 0.0,
      "loss": 11.0564,
      "nf_loss": 9.9195,
      "ppl": 3.3953,
      "step": 44000
    },
    {
      "epoch": 0.61,
      "eval_loss": 10.314085006713867,
      "eval_nf_loss": 8.550851821899414,
      "eval_perplexity": 6.158840179443359,
      "eval_runtime": 545.2003,
      "eval_samples_per_second": 24.519,
      "eval_steps_per_second": 3.065,
      "step": 44000
    },
    {
      "epoch": 0.64,
      "gate_score": 0.1751,
      "learning_rate": 0.0,
      "loss": 10.8376,
      "nf_loss": 9.7071,
      "ppl": 3.3905,
      "step": 46000
    },
    {
      "epoch": 0.67,
      "gate_score": 0.1707,
      "learning_rate": 0.0,
      "loss": 10.544,
      "nf_loss": 9.4109,
      "ppl": 3.3956,
      "step": 48000
    },
    {
      "epoch": 0.67,
      "eval_loss": 9.926258087158203,
      "eval_nf_loss": 8.179387092590332,
      "eval_perplexity": 6.0638957023620605,
      "eval_runtime": 539.7929,
      "eval_samples_per_second": 24.765,
      "eval_steps_per_second": 3.096,
      "step": 48000
    },
    {
      "epoch": 0.7,
      "gate_score": 0.1675,
      "learning_rate": 0.0,
      "loss": 10.2486,
      "nf_loss": 9.1318,
      "ppl": 3.3367,
      "step": 50000
    },
    {
      "epoch": 0.72,
      "gate_score": 0.1654,
      "learning_rate": 0.0,
      "loss": 9.997,
      "nf_loss": 8.8935,
      "ppl": 3.2788,
      "step": 52000
    },
    {
      "epoch": 0.72,
      "eval_loss": 9.405996322631836,
      "eval_nf_loss": 7.684138774871826,
      "eval_perplexity": 5.903836727142334,
      "eval_runtime": 540.5809,
      "eval_samples_per_second": 24.729,
      "eval_steps_per_second": 3.091,
      "step": 52000
    },
    {
      "epoch": 0.75,
      "gate_score": 0.1641,
      "learning_rate": 0.0,
      "loss": 9.6968,
      "nf_loss": 8.5947,
      "ppl": 3.2795,
      "step": 54000
    },
    {
      "epoch": 0.78,
      "gate_score": 0.1588,
      "learning_rate": 0.0,
      "loss": 9.6519,
      "nf_loss": 8.5586,
      "ppl": 3.2488,
      "step": 56000
    },
    {
      "epoch": 0.78,
      "eval_loss": 9.122363090515137,
      "eval_nf_loss": 7.420844554901123,
      "eval_perplexity": 5.77822732925415,
      "eval_runtime": 541.8983,
      "eval_samples_per_second": 24.669,
      "eval_steps_per_second": 3.084,
      "step": 56000
    },
    {
      "epoch": 0.81,
      "gate_score": 0.1589,
      "learning_rate": 0.0,
      "loss": 9.4534,
      "nf_loss": 8.3634,
      "ppl": 3.2306,
      "step": 58000
    },
    {
      "epoch": 0.84,
      "gate_score": 0.1554,
      "learning_rate": 0.0,
      "loss": 9.1261,
      "nf_loss": 8.0534,
      "ppl": 3.1834,
      "step": 60000
    },
    {
      "epoch": 0.84,
      "eval_loss": 8.732184410095215,
      "eval_nf_loss": 7.0388031005859375,
      "eval_perplexity": 5.730079650878906,
      "eval_runtime": 538.8984,
      "eval_samples_per_second": 24.806,
      "eval_steps_per_second": 3.101,
      "step": 60000
    },
    {
      "epoch": 0.86,
      "gate_score": 0.1552,
      "learning_rate": 0.0,
      "loss": 9.0458,
      "nf_loss": 7.9732,
      "ppl": 3.1887,
      "step": 62000
    },
    {
      "epoch": 0.89,
      "gate_score": 0.1527,
      "learning_rate": 0.0,
      "loss": 8.9264,
      "nf_loss": 7.8604,
      "ppl": 3.1517,
      "step": 64000
    },
    {
      "epoch": 0.89,
      "eval_loss": 8.456263542175293,
      "eval_nf_loss": 6.77412223815918,
      "eval_perplexity": 5.667126178741455,
      "eval_runtime": 546.1435,
      "eval_samples_per_second": 24.477,
      "eval_steps_per_second": 3.06,
      "step": 64000
    },
    {
      "epoch": 0.92,
      "gate_score": 0.1532,
      "learning_rate": 0.0,
      "loss": 8.8657,
      "nf_loss": 7.7956,
      "ppl": 3.1667,
      "step": 66000
    },
    {
      "epoch": 0.95,
      "gate_score": 0.1512,
      "learning_rate": 0.0,
      "loss": 8.7919,
      "nf_loss": 7.7484,
      "ppl": 3.0788,
      "step": 68000
    },
    {
      "epoch": 0.95,
      "eval_loss": 8.375990867614746,
      "eval_nf_loss": 6.704384803771973,
      "eval_perplexity": 5.604784965515137,
      "eval_runtime": 523.4289,
      "eval_samples_per_second": 25.539,
      "eval_steps_per_second": 3.192,
      "step": 68000
    },
    {
      "epoch": 0.98,
      "gate_score": 0.1507,
      "learning_rate": 0.0,
      "loss": 8.7686,
      "nf_loss": 7.7189,
      "ppl": 3.0927,
      "step": 70000
    },
    {
      "epoch": 1.0,
      "gate_score": 0.1504,
      "learning_rate": 5.5250000000000005e-06,
      "loss": 8.6434,
      "nf_loss": 7.5995,
      "ppl": 3.0833,
      "step": 72000
    },
    {
      "epoch": 1.0,
      "eval_loss": 8.295180320739746,
      "eval_nf_loss": 6.628587245941162,
      "eval_perplexity": 5.5753278732299805,
      "eval_runtime": 520.7294,
      "eval_samples_per_second": 25.672,
      "eval_steps_per_second": 3.209,
      "step": 72000
    },
    {
      "epoch": 1.03,
      "gate_score": 0.1439,
      "learning_rate": 4.9922293335593894e-05,
      "loss": 9.2496,
      "nf_loss": 8.1825,
      "ppl": 3.162,
      "step": 74000
    },
    {
      "epoch": 1.06,
      "gate_score": 0.1256,
      "learning_rate": 4.921622232583111e-05,
      "loss": 9.3761,
      "nf_loss": 8.2391,
      "ppl": 3.4304,
      "step": 76000
    },
    {
      "epoch": 1.06,
      "eval_loss": 8.703157424926758,
      "eval_nf_loss": 6.935546875,
      "eval_perplexity": 6.19848108291626,
      "eval_runtime": 523.4928,
      "eval_samples_per_second": 25.536,
      "eval_steps_per_second": 3.192,
      "step": 76000
    },
    {
      "epoch": 1.09,
      "gate_score": 0.1083,
      "learning_rate": 4.8510151316068326e-05,
      "loss": 8.9973,
      "nf_loss": 7.8753,
      "ppl": 3.3654,
      "step": 78000
    },
    {
      "epoch": 1.11,
      "gate_score": 0.0951,
      "learning_rate": 4.780372709419461e-05,
      "loss": 8.6673,
      "nf_loss": 7.5525,
      "ppl": 3.354,
      "step": 80000
    },
    {
      "epoch": 1.11,
      "eval_loss": 8.25178337097168,
      "eval_nf_loss": 6.506425857543945,
      "eval_perplexity": 6.058511257171631,
      "eval_runtime": 521.4601,
      "eval_samples_per_second": 25.636,
      "eval_steps_per_second": 3.204,
      "step": 80000
    },
    {
      "epoch": 1.14,
      "gate_score": 0.0857,
      "learning_rate": 4.7097302872320885e-05,
      "loss": 8.3294,
      "nf_loss": 7.2064,
      "ppl": 3.3779,
      "step": 82000
    },
    {
      "epoch": 1.17,
      "gate_score": 0.0771,
      "learning_rate": 4.63912318625581e-05,
      "loss": 7.9669,
      "nf_loss": 6.8593,
      "ppl": 3.3159,
      "step": 84000
    },
    {
      "epoch": 1.17,
      "eval_loss": 7.606574058532715,
      "eval_nf_loss": 5.845552444458008,
      "eval_perplexity": 6.1504316329956055,
      "eval_runtime": 522.2833,
      "eval_samples_per_second": 25.595,
      "eval_steps_per_second": 3.199,
      "step": 84000
    },
    {
      "epoch": 1.2,
      "gate_score": 0.0728,
      "learning_rate": 4.5685160852795324e-05,
      "loss": 7.6899,
      "nf_loss": 6.5963,
      "ppl": 3.2681,
      "step": 86000
    },
    {
      "epoch": 1.23,
      "gate_score": 0.0699,
      "learning_rate": 4.49787366309216e-05,
      "loss": 7.4537,
      "nf_loss": 6.3595,
      "ppl": 3.2686,
      "step": 88000
    },
    {
      "epoch": 1.23,
      "eval_loss": 7.2829508781433105,
      "eval_nf_loss": 5.5018086433410645,
      "eval_perplexity": 6.287156105041504,
      "eval_runtime": 523.2748,
      "eval_samples_per_second": 25.547,
      "eval_steps_per_second": 3.193,
      "step": 88000
    },
    {
      "epoch": 1.25,
      "gate_score": 0.0677,
      "learning_rate": 4.427266562115882e-05,
      "loss": 7.2129,
      "nf_loss": 6.1186,
      "ppl": 3.2734,
      "step": 90000
    },
    {
      "epoch": 1.28,
      "gate_score": 0.0629,
      "learning_rate": 4.356659461139604e-05,
      "loss": 6.9976,
      "nf_loss": 5.913,
      "ppl": 3.2358,
      "step": 92000
    },
    {
      "epoch": 1.28,
      "eval_loss": 6.838191986083984,
      "eval_nf_loss": 5.100852966308594,
      "eval_perplexity": 6.003511905670166,
      "eval_runtime": 521.2941,
      "eval_samples_per_second": 25.644,
      "eval_steps_per_second": 3.205,
      "step": 92000
    },
    {
      "epoch": 1.31,
      "gate_score": 0.0606,
      "learning_rate": 4.286017038952232e-05,
      "loss": 6.7417,
      "nf_loss": 5.6817,
      "ppl": 3.1591,
      "step": 94000
    },
    {
      "epoch": 1.34,
      "gate_score": 0.0583,
      "learning_rate": 4.215409937975954e-05,
      "loss": 6.626,
      "nf_loss": 5.5517,
      "ppl": 3.2084,
      "step": 96000
    },
    {
      "epoch": 1.34,
      "eval_loss": 6.460667610168457,
      "eval_nf_loss": 4.721277236938477,
      "eval_perplexity": 6.021188259124756,
      "eval_runtime": 522.1793,
      "eval_samples_per_second": 25.6,
      "eval_steps_per_second": 3.2,
      "step": 96000
    },
    {
      "epoch": 1.37,
      "gate_score": 0.054,
      "learning_rate": 4.144767515788582e-05,
      "loss": 6.4202,
      "nf_loss": 5.3655,
      "ppl": 3.1392,
      "step": 98000
    },
    {
      "epoch": 1.39,
      "gate_score": 0.0531,
      "learning_rate": 4.074160414812303e-05,
      "loss": 6.2782,
      "nf_loss": 5.2125,
      "ppl": 3.1798,
      "step": 100000
    },
    {
      "epoch": 1.39,
      "eval_loss": 6.316382884979248,
      "eval_nf_loss": 4.596680164337158,
      "eval_perplexity": 5.897133827209473,
      "eval_runtime": 522.484,
      "eval_samples_per_second": 25.585,
      "eval_steps_per_second": 3.198,
      "step": 100000
    },
    {
      "epoch": 1.42,
      "gate_score": 0.0518,
      "learning_rate": 4.003517992624931e-05,
      "loss": 6.1583,
      "nf_loss": 5.1099,
      "ppl": 3.1166,
      "step": 102000
    },
    {
      "epoch": 1.45,
      "gate_score": 0.0485,
      "learning_rate": 3.932946212859746e-05,
      "loss": 6.0752,
      "nf_loss": 5.0167,
      "ppl": 3.1635,
      "step": 104000
    },
    {
      "epoch": 1.45,
      "eval_loss": 6.336714267730713,
      "eval_nf_loss": 4.630500316619873,
      "eval_perplexity": 5.809221267700195,
      "eval_runtime": 528.0106,
      "eval_samples_per_second": 25.318,
      "eval_steps_per_second": 3.165,
      "step": 104000
    },
    {
      "epoch": 1.48,
      "gate_score": 0.0486,
      "learning_rate": 3.8623037906723746e-05,
      "loss": 5.8858,
      "nf_loss": 4.8328,
      "ppl": 3.1254,
      "step": 106000
    },
    {
      "epoch": 1.5,
      "gate_score": 0.0479,
      "learning_rate": 3.791696689696096e-05,
      "loss": 5.7518,
      "nf_loss": 4.7045,
      "ppl": 3.1126,
      "step": 108000
    },
    {
      "epoch": 1.5,
      "eval_loss": 5.817115783691406,
      "eval_nf_loss": 4.078435897827148,
      "eval_perplexity": 6.01722526550293,
      "eval_runtime": 523.2078,
      "eval_samples_per_second": 25.55,
      "eval_steps_per_second": 3.194,
      "step": 108000
    },
    {
      "epoch": 1.53,
      "gate_score": 0.0447,
      "learning_rate": 3.7210895887198185e-05,
      "loss": 5.6666,
      "nf_loss": 4.6289,
      "ppl": 3.0797,
      "step": 110000
    },
    {
      "epoch": 1.56,
      "gate_score": 0.0436,
      "learning_rate": 3.650447166532446e-05,
      "loss": 5.5464,
      "nf_loss": 4.5167,
      "ppl": 3.0513,
      "step": 112000
    },
    {
      "epoch": 1.56,
      "eval_loss": 5.787192344665527,
      "eval_nf_loss": 4.095870494842529,
      "eval_perplexity": 5.72128438949585,
      "eval_runtime": 542.9095,
      "eval_samples_per_second": 24.623,
      "eval_steps_per_second": 3.078,
      "step": 112000
    },
    {
      "epoch": 1.59,
      "gate_score": 0.0429,
      "learning_rate": 3.5798400655561684e-05,
      "loss": 5.4501,
      "nf_loss": 4.4131,
      "ppl": 3.0922,
      "step": 114000
    },
    {
      "epoch": 1.62,
      "gate_score": 0.0423,
      "learning_rate": 3.509197643368796e-05,
      "loss": 5.3478,
      "nf_loss": 4.3169,
      "ppl": 3.07,
      "step": 116000
    },
    {
      "epoch": 1.62,
      "eval_loss": 5.440142631530762,
      "eval_nf_loss": 3.742682695388794,
      "eval_perplexity": 5.755825519561768,
      "eval_runtime": 558.1397,
      "eval_samples_per_second": 23.951,
      "eval_steps_per_second": 2.994,
      "step": 116000
    },
    {
      "epoch": 1.64,
      "gate_score": 0.0385,
      "learning_rate": 3.438625863603611e-05,
      "loss": 5.2786,
      "nf_loss": 4.247,
      "ppl": 3.0576,
      "step": 118000
    },
    {
      "epoch": 1.67,
      "gate_score": 0.0397,
      "learning_rate": 3.368018762627333e-05,
      "loss": 5.2213,
      "nf_loss": 4.2011,
      "ppl": 3.0254,
      "step": 120000
    },
    {
      "epoch": 1.67,
      "eval_loss": 5.391260147094727,
      "eval_nf_loss": 3.704191207885742,
      "eval_perplexity": 5.694056034088135,
      "eval_runtime": 548.1395,
      "eval_samples_per_second": 24.388,
      "eval_steps_per_second": 3.048,
      "step": 120000
    },
    {
      "epoch": 1.7,
      "gate_score": 0.0399,
      "learning_rate": 3.297376340439961e-05,
      "loss": 5.0984,
      "nf_loss": 4.0796,
      "ppl": 3.0276,
      "step": 122000
    },
    {
      "epoch": 1.73,
      "gate_score": 0.0388,
      "learning_rate": 3.226733918252589e-05,
      "loss": 5.0362,
      "nf_loss": 4.0221,
      "ppl": 3.0027,
      "step": 124000
    },
    {
      "epoch": 1.73,
      "eval_loss": 5.266547203063965,
      "eval_nf_loss": 3.582312822341919,
      "eval_perplexity": 5.674898147583008,
      "eval_runtime": 547.5664,
      "eval_samples_per_second": 24.413,
      "eval_steps_per_second": 3.052,
      "step": 124000
    },
    {
      "epoch": 1.76,
      "gate_score": 0.0392,
      "learning_rate": 3.156126817276311e-05,
      "loss": 4.989,
      "nf_loss": 3.9664,
      "ppl": 3.025,
      "step": 126000
    },
    {
      "epoch": 1.78,
      "gate_score": 0.0387,
      "learning_rate": 3.085484395088939e-05,
      "loss": 4.8721,
      "nf_loss": 3.8658,
      "ppl": 2.9757,
      "step": 128000
    },
    {
      "epoch": 1.78,
      "eval_loss": 5.017204284667969,
      "eval_nf_loss": 3.343366861343384,
      "eval_perplexity": 5.61837100982666,
      "eval_runtime": 545.9238,
      "eval_samples_per_second": 24.487,
      "eval_steps_per_second": 3.061,
      "step": 128000
    },
    {
      "epoch": 1.81,
      "gate_score": 0.0387,
      "learning_rate": 3.0148772941126607e-05,
      "loss": 4.804,
      "nf_loss": 3.807,
      "ppl": 2.9565,
      "step": 130000
    },
    {
      "epoch": 1.84,
      "gate_score": 0.0387,
      "learning_rate": 2.9442701931363827e-05,
      "loss": 4.8048,
      "nf_loss": 3.7985,
      "ppl": 2.9827,
      "step": 132000
    },
    {
      "epoch": 1.84,
      "eval_loss": 4.92326545715332,
      "eval_nf_loss": 3.2530434131622314,
      "eval_perplexity": 5.5944623947143555,
      "eval_runtime": 555.9027,
      "eval_samples_per_second": 24.047,
      "eval_steps_per_second": 3.006,
      "step": 132000
    },
    {
      "epoch": 1.87,
      "gate_score": 0.0373,
      "learning_rate": 2.8736277709490106e-05,
      "loss": 4.721,
      "nf_loss": 3.7315,
      "ppl": 2.9249,
      "step": 134000
    },
    {
      "epoch": 1.89,
      "gate_score": 0.0358,
      "learning_rate": 2.8030559911838262e-05,
      "loss": 4.6625,
      "nf_loss": 3.6917,
      "ppl": 2.8665,
      "step": 136000
    },
    {
      "epoch": 1.89,
      "eval_loss": 4.809460639953613,
      "eval_nf_loss": 3.1493453979492188,
      "eval_perplexity": 5.54146671295166,
      "eval_runtime": 554.8074,
      "eval_samples_per_second": 24.095,
      "eval_steps_per_second": 3.012,
      "step": 136000
    },
    {
      "epoch": 1.92,
      "gate_score": 0.0359,
      "learning_rate": 2.7324135689964535e-05,
      "loss": 4.6193,
      "nf_loss": 3.6298,
      "ppl": 2.9235,
      "step": 138000
    },
    {
      "epoch": 1.95,
      "gate_score": 0.0359,
      "learning_rate": 2.6617711468090818e-05,
      "loss": 4.5676,
      "nf_loss": 3.5739,
      "ppl": 2.9433,
      "step": 140000
    },
    {
      "epoch": 1.95,
      "eval_loss": 4.744356632232666,
      "eval_nf_loss": 3.076949119567871,
      "eval_perplexity": 5.5782151222229,
      "eval_runtime": 553.0927,
      "eval_samples_per_second": 24.17,
      "eval_steps_per_second": 3.021,
      "step": 140000
    },
    {
      "epoch": 1.98,
      "gate_score": 0.0362,
      "learning_rate": 2.591199367043897e-05,
      "loss": 4.5088,
      "nf_loss": 3.5319,
      "ppl": 2.8899,
      "step": 142000
    },
    {
      "epoch": 2.01,
      "gate_score": 0.0341,
      "learning_rate": 2.5205569448565254e-05,
      "loss": 4.4715,
      "nf_loss": 3.5168,
      "ppl": 2.8169,
      "step": 144000
    },
    {
      "epoch": 2.01,
      "eval_loss": 4.6942243576049805,
      "eval_nf_loss": 3.0113067626953125,
      "eval_perplexity": 5.677947521209717,
      "eval_runtime": 545.2448,
      "eval_samples_per_second": 24.517,
      "eval_steps_per_second": 3.065,
      "step": 144000
    },
    {
      "epoch": 2.03,
      "gate_score": 0.0338,
      "learning_rate": 2.449949843880247e-05,
      "loss": 4.3094,
      "nf_loss": 3.4433,
      "ppl": 2.5464,
      "step": 146000
    },
    {
      "epoch": 2.06,
      "gate_score": 0.0356,
      "learning_rate": 2.379342742903969e-05,
      "loss": 4.2547,
      "nf_loss": 3.3975,
      "ppl": 2.515,
      "step": 148000
    },
    {
      "epoch": 2.06,
      "eval_loss": 4.599600791931152,
      "eval_nf_loss": 2.9459831714630127,
      "eval_perplexity": 5.511317729949951,
      "eval_runtime": 545.8913,
      "eval_samples_per_second": 24.488,
      "eval_steps_per_second": 3.061,
      "step": 148000
    },
    {
      "epoch": 2.09,
      "gate_score": 0.037,
      "learning_rate": 2.308700320716597e-05,
      "loss": 4.2072,
      "nf_loss": 3.3509,
      "ppl": 2.523,
      "step": 150000
    },
    {
      "epoch": 2.12,
      "gate_score": 0.0352,
      "learning_rate": 2.2380932197403185e-05,
      "loss": 4.2326,
      "nf_loss": 3.3816,
      "ppl": 2.5079,
      "step": 152000
    },
    {
      "epoch": 2.12,
      "eval_loss": 4.591421127319336,
      "eval_nf_loss": 2.9333114624023438,
      "eval_perplexity": 5.539112567901611,
      "eval_runtime": 534.7032,
      "eval_samples_per_second": 25.001,
      "eval_steps_per_second": 3.125,
      "step": 152000
    },
    {
      "epoch": 2.15,
      "gate_score": 0.0347,
      "learning_rate": 2.16748611876404e-05,
      "loss": 4.1787,
      "nf_loss": 3.3198,
      "ppl": 2.5258,
      "step": 154000
    },
    {
      "epoch": 2.17,
      "gate_score": 0.0354,
      "learning_rate": 2.096879017787762e-05,
      "loss": 4.1297,
      "nf_loss": 3.2735,
      "ppl": 2.5189,
      "step": 156000
    },
    {
      "epoch": 2.17,
      "eval_loss": 4.490691661834717,
      "eval_nf_loss": 2.837989091873169,
      "eval_perplexity": 5.503147602081299,
      "eval_runtime": 566.0681,
      "eval_samples_per_second": 23.616,
      "eval_steps_per_second": 2.952,
      "step": 156000
    },
    {
      "epoch": 2.2,
      "gate_score": 0.0354,
      "learning_rate": 2.02623659560039e-05,
      "loss": 4.0949,
      "nf_loss": 3.2404,
      "ppl": 2.5092,
      "step": 158000
    },
    {
      "epoch": 2.23,
      "gate_score": 0.0364,
      "learning_rate": 1.955629494624112e-05,
      "loss": 4.0613,
      "nf_loss": 3.2097,
      "ppl": 2.5052,
      "step": 160000
    },
    {
      "epoch": 2.23,
      "eval_loss": 4.413490295410156,
      "eval_nf_loss": 2.7680394649505615,
      "eval_perplexity": 5.459636688232422,
      "eval_runtime": 574.0313,
      "eval_samples_per_second": 23.288,
      "eval_steps_per_second": 2.911,
      "step": 160000
    },
    {
      "epoch": 2.26,
      "gate_score": 0.0349,
      "learning_rate": 1.8850223936478333e-05,
      "loss": 4.0192,
      "nf_loss": 3.1647,
      "ppl": 2.5177,
      "step": 162000
    },
    {
      "epoch": 2.28,
      "gate_score": 0.0356,
      "learning_rate": 1.8144152926715553e-05,
      "loss": 3.9985,
      "nf_loss": 3.1476,
      "ppl": 2.5005,
      "step": 164000
    },
    {
      "epoch": 2.28,
      "eval_loss": 4.352046966552734,
      "eval_nf_loss": 2.7076711654663086,
      "eval_perplexity": 5.455787658691406,
      "eval_runtime": 559.4532,
      "eval_samples_per_second": 23.895,
      "eval_steps_per_second": 2.987,
      "step": 164000
    },
    {
      "epoch": 2.31,
      "gate_score": 0.0364,
      "learning_rate": 1.7437728704841832e-05,
      "loss": 4.0077,
      "nf_loss": 3.1494,
      "ppl": 2.5183,
      "step": 166000
    },
    {
      "epoch": 2.34,
      "gate_score": 0.0361,
      "learning_rate": 1.673201090718999e-05,
      "loss": 3.943,
      "nf_loss": 3.0934,
      "ppl": 2.4942,
      "step": 168000
    },
    {
      "epoch": 2.34,
      "eval_loss": 4.351120948791504,
      "eval_nf_loss": 2.7190158367156982,
      "eval_perplexity": 5.387732028961182,
      "eval_runtime": 555.7968,
      "eval_samples_per_second": 24.052,
      "eval_steps_per_second": 3.006,
      "step": 168000
    },
    {
      "epoch": 2.37,
      "gate_score": 0.0351,
      "learning_rate": 1.6025586685316265e-05,
      "loss": 3.9552,
      "nf_loss": 3.1077,
      "ppl": 2.4925,
      "step": 170000
    },
    {
      "epoch": 2.4,
      "gate_score": 0.0348,
      "learning_rate": 1.5319162463442547e-05,
      "loss": 3.9297,
      "nf_loss": 3.0888,
      "ppl": 2.4752,
      "step": 172000
    },
    {
      "epoch": 2.4,
      "eval_loss": 4.315845966339111,
      "eval_nf_loss": 2.6691393852233887,
      "eval_perplexity": 5.4654669761657715,
      "eval_runtime": 551.5389,
      "eval_samples_per_second": 24.238,
      "eval_steps_per_second": 3.03,
      "step": 172000
    },
    {
      "epoch": 2.42,
      "gate_score": 0.0366,
      "learning_rate": 1.4613091453679764e-05,
      "loss": 3.8766,
      "nf_loss": 3.0339,
      "ppl": 2.4785,
      "step": 174000
    },
    {
      "epoch": 2.45,
      "gate_score": 0.0367,
      "learning_rate": 1.3907020443916982e-05,
      "loss": 3.836,
      "nf_loss": 2.9959,
      "ppl": 2.4729,
      "step": 176000
    },
    {
      "epoch": 2.45,
      "eval_loss": 4.235720157623291,
      "eval_nf_loss": 2.6040456295013428,
      "eval_perplexity": 5.383092880249023,
      "eval_runtime": 542.8059,
      "eval_samples_per_second": 24.628,
      "eval_steps_per_second": 3.078,
      "step": 176000
    },
    {
      "epoch": 2.48,
      "gate_score": 0.0365,
      "learning_rate": 1.32009494341542e-05,
      "loss": 3.8171,
      "nf_loss": 2.9705,
      "ppl": 2.4948,
      "step": 178000
    },
    {
      "epoch": 2.51,
      "gate_score": 0.0365,
      "learning_rate": 1.2494878424391417e-05,
      "loss": 3.7842,
      "nf_loss": 2.9479,
      "ppl": 2.4626,
      "step": 180000
    },
    {
      "epoch": 2.51,
      "eval_loss": 4.155655860900879,
      "eval_nf_loss": 2.538012981414795,
      "eval_perplexity": 5.301826000213623,
      "eval_runtime": 533.4404,
      "eval_samples_per_second": 25.06,
      "eval_steps_per_second": 3.132,
      "step": 180000
    },
    {
      "epoch": 2.54,
      "gate_score": 0.0361,
      "learning_rate": 1.1788454202517695e-05,
      "loss": 3.7683,
      "nf_loss": 2.932,
      "ppl": 2.4647,
      "step": 182000
    },
    {
      "epoch": 2.56,
      "gate_score": 0.0367,
      "learning_rate": 1.1082029980643976e-05,
      "loss": 3.7123,
      "nf_loss": 2.878,
      "ppl": 2.4572,
      "step": 184000
    },
    {
      "epoch": 2.56,
      "eval_loss": 4.167212963104248,
      "eval_nf_loss": 2.5371882915496826,
      "eval_perplexity": 5.372892379760742,
      "eval_runtime": 539.6678,
      "eval_samples_per_second": 24.771,
      "eval_steps_per_second": 3.096,
      "step": 184000
    },
    {
      "epoch": 2.59,
      "gate_score": 0.0363,
      "learning_rate": 1.0375958970881194e-05,
      "loss": 3.6904,
      "nf_loss": 2.8501,
      "ppl": 2.4696,
      "step": 186000
    },
    {
      "epoch": 2.62,
      "gate_score": 0.0366,
      "learning_rate": 9.66988796111841e-06,
      "loss": 3.6854,
      "nf_loss": 2.8601,
      "ppl": 2.437,
      "step": 188000
    },
    {
      "epoch": 2.62,
      "eval_loss": 4.0906524658203125,
      "eval_nf_loss": 2.4722979068756104,
      "eval_perplexity": 5.306558609008789,
      "eval_runtime": 540.495,
      "eval_samples_per_second": 24.733,
      "eval_steps_per_second": 3.092,
      "step": 188000
    },
    {
      "epoch": 2.65,
      "gate_score": 0.0373,
      "learning_rate": 8.963816951355628e-06,
      "loss": 3.6947,
      "nf_loss": 2.8627,
      "ppl": 2.4485,
      "step": 190000
    },
    {
      "epoch": 2.67,
      "gate_score": 0.0361,
      "learning_rate": 8.25739272948191e-06,
      "loss": 3.6339,
      "nf_loss": 2.8122,
      "ppl": 2.4183,
      "step": 192000
    },
    {
      "epoch": 2.67,
      "eval_loss": 4.026703357696533,
      "eval_nf_loss": 2.4067482948303223,
      "eval_perplexity": 5.3168044090271,
      "eval_runtime": 520.3796,
      "eval_samples_per_second": 25.689,
      "eval_steps_per_second": 3.211,
      "step": 192000
    },
    {
      "epoch": 2.7,
      "gate_score": 0.0346,
      "learning_rate": 7.5513217197191265e-06,
      "loss": 3.6295,
      "nf_loss": 2.8102,
      "ppl": 2.4127,
      "step": 194000
    },
    {
      "epoch": 2.73,
      "gate_score": 0.0357,
      "learning_rate": 6.845250709956344e-06,
      "loss": 3.6169,
      "nf_loss": 2.7965,
      "ppl": 2.4179,
      "step": 196000
    },
    {
      "epoch": 2.73,
      "eval_loss": 4.014979839324951,
      "eval_nf_loss": 2.3965229988098145,
      "eval_perplexity": 5.308826446533203,
      "eval_runtime": 519.5159,
      "eval_samples_per_second": 25.732,
      "eval_steps_per_second": 3.216,
      "step": 196000
    },
    {
      "epoch": 2.76,
      "gate_score": 0.0354,
      "learning_rate": 6.138826488082624e-06,
      "loss": 3.6326,
      "nf_loss": 2.8126,
      "ppl": 2.4235,
      "step": 198000
    },
    {
      "epoch": 2.79,
      "gate_score": 0.0345,
      "learning_rate": 5.432755478319841e-06,
      "loss": 3.5858,
      "nf_loss": 2.7795,
      "ppl": 2.3812,
      "step": 200000
    },
    {
      "epoch": 2.79,
      "eval_loss": 3.991497039794922,
      "eval_nf_loss": 2.37373948097229,
      "eval_perplexity": 5.302868366241455,
      "eval_runtime": 521.0118,
      "eval_samples_per_second": 25.658,
      "eval_steps_per_second": 3.207,
      "step": 200000
    },
    {
      "epoch": 2.81,
      "gate_score": 0.0345,
      "learning_rate": 4.726684468557059e-06,
      "loss": 3.5398,
      "nf_loss": 2.7304,
      "ppl": 2.3993,
      "step": 202000
    },
    {
      "epoch": 2.84,
      "gate_score": 0.0346,
      "learning_rate": 4.020613458794276e-06,
      "loss": 3.5681,
      "nf_loss": 2.7478,
      "ppl": 2.4182,
      "step": 204000
    },
    {
      "epoch": 2.84,
      "eval_loss": 3.961500644683838,
      "eval_nf_loss": 2.357841730117798,
      "eval_perplexity": 5.2263312339782715,
      "eval_runtime": 522.7896,
      "eval_samples_per_second": 25.571,
      "eval_steps_per_second": 3.196,
      "step": 204000
    },
    {
      "epoch": 2.87,
      "gate_score": 0.0344,
      "learning_rate": 3.3141892369205557e-06,
      "loss": 3.5525,
      "nf_loss": 2.7408,
      "ppl": 2.3967,
      "step": 206000
    },
    {
      "epoch": 2.9,
      "gate_score": 0.0342,
      "learning_rate": 2.6077650150468363e-06,
      "loss": 3.5232,
      "nf_loss": 2.7209,
      "ppl": 2.3736,
      "step": 208000
    },
    {
      "epoch": 2.9,
      "eval_loss": 3.9473981857299805,
      "eval_nf_loss": 2.347726583480835,
      "eval_perplexity": 5.205344200134277,
      "eval_runtime": 522.9583,
      "eval_samples_per_second": 25.562,
      "eval_steps_per_second": 3.195,
      "step": 208000
    },
    {
      "epoch": 2.93,
      "gate_score": 0.0341,
      "learning_rate": 1.9013407931731166e-06,
      "loss": 3.494,
      "nf_loss": 2.6823,
      "ppl": 2.3989,
      "step": 210000
    },
    {
      "epoch": 2.95,
      "gate_score": 0.0342,
      "learning_rate": 1.1956229955212706e-06,
      "loss": 3.4907,
      "nf_loss": 2.6875,
      "ppl": 2.3756,
      "step": 212000
    },
    {
      "epoch": 2.95,
      "eval_loss": 3.90411639213562,
      "eval_nf_loss": 2.3072071075439453,
      "eval_perplexity": 5.1911420822143555,
      "eval_runtime": 522.2632,
      "eval_samples_per_second": 25.596,
      "eval_steps_per_second": 3.2,
      "step": 212000
    }
  ],
  "max_steps": 215337,
  "num_train_epochs": 3,
  "total_flos": 1.9082412644524078e+18,
  "trial_name": null,
  "trial_params": null
}