{
  "best_metric": 3.8748562335968018,
  "best_model_checkpoint": "/kaggle/working/checkpoint-669",
  "epoch": 50.0,
  "eval_steps": 500,
  "global_step": 11150,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_loss": 3.875725507736206,
      "eval_mean_perplexity": 366.3703887939453,
      "eval_perplexities": [
        335.80645751953125,
        377.47796630859375,
        475.9695129394531,
        276.7535705566406,
        294.2838134765625,
        264.73394775390625,
        394.9538269042969,
        325.2862854003906,
        525.1426391601562,
        393.2958679199219
      ],
      "eval_runtime": 2.3638,
      "eval_samples_per_second": 4.231,
      "eval_steps_per_second": 0.846,
      "step": 223
    },
    {
      "epoch": 2.0,
      "eval_loss": 3.877554416656494,
      "eval_mean_perplexity": 373.7104461669922,
      "eval_perplexities": [
        356.83929443359375,
        425.78094482421875,
        495.7661437988281,
        287.1143798828125,
        270.86737060546875,
        256.76934814453125,
        391.27044677734375,
        345.6097412109375,
        498.4945373535156,
        408.5922546386719
      ],
      "eval_runtime": 2.2583,
      "eval_samples_per_second": 4.428,
      "eval_steps_per_second": 0.886,
      "step": 446
    },
    {
      "epoch": 3.0,
      "eval_loss": 3.8748562335968018,
      "eval_mean_perplexity": 378.68709869384764,
      "eval_perplexities": [
        404.7926940917969,
        400.53558349609375,
        487.69427490234375,
        229.20298767089844,
        278.294189453125,
        259.6055908203125,
        387.93963623046875,
        365.1226501464844,
        508.79669189453125,
        464.8866882324219
      ],
      "eval_runtime": 2.4538,
      "eval_samples_per_second": 4.075,
      "eval_steps_per_second": 0.815,
      "step": 669
    },
    {
      "epoch": 4.0,
      "eval_loss": 3.889125108718872,
      "eval_mean_perplexity": 395.6389587402344,
      "eval_perplexities": [
        386.92559814453125,
        444.32073974609375,
        480.409912109375,
        278.6746520996094,
        279.8287353515625,
        286.8155212402344,
        426.2447204589844,
        352.7798767089844,
        545.115234375,
        475.27459716796875
      ],
      "eval_runtime": 2.2494,
      "eval_samples_per_second": 4.446,
      "eval_steps_per_second": 0.889,
      "step": 892
    },
    {
      "epoch": 5.0,
      "eval_loss": 3.9018893241882324,
      "eval_mean_perplexity": 418.4004302978516,
      "eval_perplexities": [
        426.7615051269531,
        474.9723815917969,
        516.2012329101562,
        285.6946105957031,
        314.9617919921875,
        305.4250793457031,
        411.5309143066406,
        412.8928527832031,
        537.16943359375,
        498.3945007324219
      ],
      "eval_runtime": 2.2536,
      "eval_samples_per_second": 4.437,
      "eval_steps_per_second": 0.887,
      "step": 1115
    },
    {
      "epoch": 6.0,
      "eval_loss": 3.9105305671691895,
      "eval_mean_perplexity": 423.3804229736328,
      "eval_perplexities": [
        421.5846862792969,
        457.750244140625,
        521.7881469726562,
        293.4595642089844,
        283.9613037109375,
        287.46807861328125,
        451.5904846191406,
        458.28509521484375,
        608.7252197265625,
        449.19140625
      ],
      "eval_runtime": 2.261,
      "eval_samples_per_second": 4.423,
      "eval_steps_per_second": 0.885,
      "step": 1338
    },
    {
      "epoch": 7.0,
      "eval_loss": 3.910905122756958,
      "eval_mean_perplexity": 437.6560882568359,
      "eval_perplexities": [
        459.6704406738281,
        471.1756591796875,
        508.8789367675781,
        278.22161865234375,
        305.4996337890625,
        335.0756530761719,
        449.5406494140625,
        402.190673828125,
        618.786376953125,
        547.521240234375
      ],
      "eval_runtime": 2.4541,
      "eval_samples_per_second": 4.075,
      "eval_steps_per_second": 0.815,
      "step": 1561
    },
    {
      "epoch": 8.0,
      "eval_loss": 3.9314892292022705,
      "eval_mean_perplexity": 451.49346618652345,
      "eval_perplexities": [
        432.6178283691406,
        494.5645446777344,
        518.4781494140625,
        289.39727783203125,
        317.63031005859375,
        327.0579528808594,
        491.4974670410156,
        445.9523620605469,
        667.2817993164062,
        530.4569702148438
      ],
      "eval_runtime": 3.1442,
      "eval_samples_per_second": 3.18,
      "eval_steps_per_second": 0.636,
      "step": 1784
    },
    {
      "epoch": 9.0,
      "eval_loss": 3.9446353912353516,
      "eval_mean_perplexity": 451.08067626953124,
      "eval_perplexities": [
        452.6179504394531,
        445.2970275878906,
        576.3162841796875,
        262.3009033203125,
        326.6243591308594,
        311.2583923339844,
        440.389404296875,
        448.5811767578125,
        677.4688110351562,
        569.9524536132812
      ],
      "eval_runtime": 2.2555,
      "eval_samples_per_second": 4.434,
      "eval_steps_per_second": 0.887,
      "step": 2007
    },
    {
      "epoch": 10.0,
      "eval_loss": 3.979356288909912,
      "eval_mean_perplexity": 466.0081390380859,
      "eval_perplexities": [
        479.5843505859375,
        542.103759765625,
        561.8331298828125,
        283.5645751953125,
        342.3445739746094,
        315.0624694824219,
        494.7655334472656,
        397.63812255859375,
        683.822265625,
        559.3626098632812
      ],
      "eval_runtime": 2.255,
      "eval_samples_per_second": 4.435,
      "eval_steps_per_second": 0.887,
      "step": 2230
    },
    {
      "epoch": 11.0,
      "eval_loss": 3.985069751739502,
      "eval_mean_perplexity": 471.23480529785155,
      "eval_perplexities": [
        442.65118408203125,
        578.3666381835938,
        530.2559204101562,
        292.97174072265625,
        325.9568786621094,
        331.6105651855469,
        466.1777648925781,
        453.5052490234375,
        693.37646484375,
        597.4756469726562
      ],
      "eval_runtime": 2.2568,
      "eval_samples_per_second": 4.431,
      "eval_steps_per_second": 0.886,
      "step": 2453
    },
    {
      "epoch": 12.0,
      "eval_loss": 4.011897087097168,
      "eval_mean_perplexity": 481.7092010498047,
      "eval_perplexities": [
        490.259765625,
        590.09716796875,
        524.1170043945312,
        292.7325134277344,
        362.09210205078125,
        320.8348388671875,
        483.432861328125,
        492.00469970703125,
        676.3094482421875,
        585.2116088867188
      ],
      "eval_runtime": 2.2714,
      "eval_samples_per_second": 4.402,
      "eval_steps_per_second": 0.88,
      "step": 2676
    },
    {
      "epoch": 13.0,
      "eval_loss": 4.028537750244141,
      "eval_mean_perplexity": 491.83470153808594,
      "eval_perplexities": [
        434.7430725097656,
        579.150390625,
        531.6643676757812,
        323.02447509765625,
        381.5349426269531,
        329.8095397949219,
        569.4808349609375,
        454.2198486328125,
        723.43798828125,
        591.2815551757812
      ],
      "eval_runtime": 2.268,
      "eval_samples_per_second": 4.409,
      "eval_steps_per_second": 0.882,
      "step": 2899
    },
    {
      "epoch": 14.0,
      "eval_loss": 4.046214580535889,
      "eval_mean_perplexity": 488.7431976318359,
      "eval_perplexities": [
        465.3315734863281,
        577.7365112304688,
        532.9083251953125,
        298.2411193847656,
        371.57415771484375,
        335.2172546386719,
        547.2592163085938,
        480.532470703125,
        732.7112426757812,
        545.9201049804688
      ],
      "eval_runtime": 2.2446,
      "eval_samples_per_second": 4.455,
      "eval_steps_per_second": 0.891,
      "step": 3122
    },
    {
      "epoch": 15.0,
      "eval_loss": 4.064830780029297,
      "eval_mean_perplexity": 493.6821044921875,
      "eval_perplexities": [
        508.1759948730469,
        588.5740356445312,
        494.7876892089844,
        310.3844299316406,
        357.8840026855469,
        349.0634460449219,
        574.1718139648438,
        489.5400085449219,
        730.5029907226562,
        533.7366333007812
      ],
      "eval_runtime": 2.2496,
      "eval_samples_per_second": 4.445,
      "eval_steps_per_second": 0.889,
      "step": 3345
    },
    {
      "epoch": 16.0,
      "eval_loss": 4.0865068435668945,
      "eval_mean_perplexity": 527.1822723388672,
      "eval_perplexities": [
        516.6371459960938,
        603.1266479492188,
        558.2211303710938,
        320.877685546875,
        440.86993408203125,
        384.0972595214844,
        586.449951171875,
        441.53643798828125,
        805.31201171875,
        614.6945190429688
      ],
      "eval_runtime": 2.2935,
      "eval_samples_per_second": 4.36,
      "eval_steps_per_second": 0.872,
      "step": 3568
    },
    {
      "epoch": 17.0,
      "eval_loss": 4.112179756164551,
      "eval_mean_perplexity": 517.5738647460937,
      "eval_perplexities": [
        484.16485595703125,
        623.9596557617188,
        544.3837890625,
        327.7882995605469,
        432.2279052734375,
        325.2957458496094,
        587.2503051757812,
        500.35589599609375,
        768.452392578125,
        581.8598022460938
      ],
      "eval_runtime": 2.6951,
      "eval_samples_per_second": 3.71,
      "eval_steps_per_second": 0.742,
      "step": 3791
    },
    {
      "epoch": 18.0,
      "eval_loss": 4.133630275726318,
      "eval_mean_perplexity": 532.7662231445313,
      "eval_perplexities": [
        459.6676025390625,
        644.7689819335938,
        559.6870727539062,
        313.52886962890625,
        407.5067443847656,
        358.4698486328125,
        635.401123046875,
        511.8723449707031,
        728.2614135742188,
        708.4982299804688
      ],
      "eval_runtime": 2.2702,
      "eval_samples_per_second": 4.405,
      "eval_steps_per_second": 0.881,
      "step": 4014
    },
    {
      "epoch": 19.0,
      "eval_loss": 4.14639949798584,
      "eval_mean_perplexity": 546.8367309570312,
      "eval_perplexities": [
        513.874267578125,
        633.9168090820312,
        623.620849609375,
        317.8645324707031,
        417.3320007324219,
        351.769287109375,
        557.8528442382812,
        530.8079223632812,
        834.164306640625,
        687.1644897460938
      ],
      "eval_runtime": 2.28,
      "eval_samples_per_second": 4.386,
      "eval_steps_per_second": 0.877,
      "step": 4237
    },
    {
      "epoch": 20.0,
      "eval_loss": 4.1683173179626465,
      "eval_mean_perplexity": 530.1545104980469,
      "eval_perplexities": [
        518.4563598632812,
        616.3810424804688,
        584.39404296875,
        357.87530517578125,
        424.8235778808594,
        363.0649719238281,
        572.78369140625,
        489.98748779296875,
        762.8850708007812,
        610.8935546875
      ],
      "eval_runtime": 2.5117,
      "eval_samples_per_second": 3.981,
      "eval_steps_per_second": 0.796,
      "step": 4460
    },
    {
      "epoch": 21.0,
      "eval_loss": 4.197434425354004,
      "eval_mean_perplexity": 552.9715087890625,
      "eval_perplexities": [
        496.9158630371094,
        622.5189208984375,
        620.3365478515625,
        315.29119873046875,
        508.8257751464844,
        386.271728515625,
        597.5241088867188,
        540.0391845703125,
        771.0007934570312,
        670.990966796875
      ],
      "eval_runtime": 2.2574,
      "eval_samples_per_second": 4.43,
      "eval_steps_per_second": 0.886,
      "step": 4683
    },
    {
      "epoch": 22.0,
      "eval_loss": 4.2334303855896,
      "eval_mean_perplexity": 591.004769897461,
      "eval_perplexities": [
        542.1947631835938,
        699.9281616210938,
        655.9802856445312,
        351.9259948730469,
        517.2227783203125,
        364.0205078125,
        738.9130859375,
        537.7424926757812,
        836.2189331054688,
        665.9006958007812
      ],
      "eval_runtime": 2.2478,
      "eval_samples_per_second": 4.449,
      "eval_steps_per_second": 0.89,
      "step": 4906
    },
    {
      "epoch": 22.42,
      "grad_norm": 1.9367769956588745,
      "learning_rate": 2.7578475336322873e-05,
      "loss": 3.112,
      "step": 5000
    },
    {
      "epoch": 23.0,
      "eval_loss": 4.230788707733154,
      "eval_mean_perplexity": 571.9811889648438,
      "eval_perplexities": [
        529.1312255859375,
        677.629638671875,
        639.4375,
        351.9476623535156,
        471.0734558105469,
        388.02935791015625,
        630.8019409179688,
        550.748046875,
        854.4290771484375,
        626.583984375
      ],
      "eval_runtime": 2.3322,
      "eval_samples_per_second": 4.288,
      "eval_steps_per_second": 0.858,
      "step": 5129
    },
    {
      "epoch": 24.0,
      "eval_loss": 4.264792442321777,
      "eval_mean_perplexity": 588.0643798828125,
      "eval_perplexities": [
        532.3706665039062,
        650.5995483398438,
        640.096435546875,
        367.4383850097656,
        510.7347106933594,
        402.79925537109375,
        740.1875610351562,
        580.5010375976562,
        837.0123901367188,
        618.90380859375
      ],
      "eval_runtime": 2.414,
      "eval_samples_per_second": 4.142,
      "eval_steps_per_second": 0.828,
      "step": 5352
    },
    {
      "epoch": 25.0,
      "eval_loss": 4.280625820159912,
      "eval_mean_perplexity": 568.3704132080078,
      "eval_perplexities": [
        493.6273193359375,
        687.0193481445312,
        617.8013305664062,
        356.9241943359375,
        462.1817321777344,
        377.38800048828125,
        662.7510375976562,
        553.1047973632812,
        769.9375,
        702.9688720703125
      ],
      "eval_runtime": 2.2617,
      "eval_samples_per_second": 4.421,
      "eval_steps_per_second": 0.884,
      "step": 5575
    },
    {
      "epoch": 26.0,
      "eval_loss": 4.299654006958008,
      "eval_mean_perplexity": 585.3176055908203,
      "eval_perplexities": [
        507.7140808105469,
        719.7584228515625,
        610.5079956054688,
        355.23577880859375,
        474.7140197753906,
        343.5024719238281,
        644.6328125,
        599.3197631835938,
        868.6974487304688,
        729.09326171875
      ],
      "eval_runtime": 2.2752,
      "eval_samples_per_second": 4.395,
      "eval_steps_per_second": 0.879,
      "step": 5798
    },
    {
      "epoch": 27.0,
      "eval_loss": 4.309383869171143,
      "eval_mean_perplexity": 572.4837829589844,
      "eval_perplexities": [
        532.3777465820312,
        701.5357055664062,
        574.2848510742188,
        366.85614013671875,
        481.5206298828125,
        389.685546875,
        579.2503662109375,
        525.9729614257812,
        906.0205078125,
        667.3333740234375
      ],
      "eval_runtime": 2.4436,
      "eval_samples_per_second": 4.092,
      "eval_steps_per_second": 0.818,
      "step": 6021
    },
    {
      "epoch": 28.0,
      "eval_loss": 4.338665962219238,
      "eval_mean_perplexity": 617.5483917236328,
      "eval_perplexities": [
        530.51416015625,
        770.9551391601562,
        616.0216674804688,
        386.1516418457031,
        501.05426025390625,
        419.87841796875,
        735.8825073242188,
        594.3335571289062,
        911.9882202148438,
        708.704345703125
      ],
      "eval_runtime": 2.2597,
      "eval_samples_per_second": 4.425,
      "eval_steps_per_second": 0.885,
      "step": 6244
    },
    {
      "epoch": 29.0,
      "eval_loss": 4.354172229766846,
      "eval_mean_perplexity": 609.77958984375,
      "eval_perplexities": [
        571.1878662109375,
        744.3867797851562,
        607.4262084960938,
        374.7521057128906,
        501.3298034667969,
        401.453369140625,
        706.0652465820312,
        600.0166625976562,
        898.824951171875,
        692.3529052734375
      ],
      "eval_runtime": 2.4521,
      "eval_samples_per_second": 4.078,
      "eval_steps_per_second": 0.816,
      "step": 6467
    },
    {
      "epoch": 30.0,
      "eval_loss": 4.371840000152588,
      "eval_mean_perplexity": 639.5644592285156,
      "eval_perplexities": [
        518.4427490234375,
        768.5755615234375,
        643.536376953125,
        434.2029724121094,
        553.5997924804688,
        401.9341735839844,
        738.0169677734375,
        627.0206909179688,
        923.0927734375,
        787.2225341796875
      ],
      "eval_runtime": 2.2759,
      "eval_samples_per_second": 4.394,
      "eval_steps_per_second": 0.879,
      "step": 6690
    },
    {
      "epoch": 31.0,
      "eval_loss": 4.397173881530762,
      "eval_mean_perplexity": 666.7775909423829,
      "eval_perplexities": [
        549.1520385742188,
        851.6005249023438,
        712.2999267578125,
        395.8175048828125,
        540.697021484375,
        451.8854064941406,
        723.3524780273438,
        630.9887084960938,
        1021.4713745117188,
        790.5109252929688
      ],
      "eval_runtime": 2.2555,
      "eval_samples_per_second": 4.434,
      "eval_steps_per_second": 0.887,
      "step": 6913
    },
    {
      "epoch": 32.0,
      "eval_loss": 4.415882110595703,
      "eval_mean_perplexity": 660.8353820800781,
      "eval_perplexities": [
        512.7852783203125,
        785.50439453125,
        668.9815063476562,
        457.69830322265625,
        539.0671997070312,
        459.60736083984375,
        751.9481201171875,
        625.0054321289062,
        999.4456787109375,
        808.310546875
      ],
      "eval_runtime": 2.2544,
      "eval_samples_per_second": 4.436,
      "eval_steps_per_second": 0.887,
      "step": 7136
    },
    {
      "epoch": 33.0,
      "eval_loss": 4.42371940612793,
      "eval_mean_perplexity": 662.7573822021484,
      "eval_perplexities": [
        559.4247436523438,
        747.0921020507812,
        678.9531860351562,
        437.0193786621094,
        544.0770874023438,
        431.26568603515625,
        756.7093505859375,
        626.279052734375,
        1022.3748168945312,
        824.37841796875
      ],
      "eval_runtime": 2.2901,
      "eval_samples_per_second": 4.367,
      "eval_steps_per_second": 0.873,
      "step": 7359
    },
    {
      "epoch": 34.0,
      "eval_loss": 4.434357166290283,
      "eval_mean_perplexity": 650.7052612304688,
      "eval_perplexities": [
        549.6817626953125,
        811.8685302734375,
        639.683349609375,
        410.87249755859375,
        510.8334045410156,
        438.6333923339844,
        779.5153198242188,
        584.6727294921875,
        1027.2081298828125,
        754.08349609375
      ],
      "eval_runtime": 2.3566,
      "eval_samples_per_second": 4.243,
      "eval_steps_per_second": 0.849,
      "step": 7582
    },
    {
      "epoch": 35.0,
      "eval_loss": 4.456236839294434,
      "eval_mean_perplexity": 683.213168334961,
      "eval_perplexities": [
        584.2586059570312,
        837.6990966796875,
        639.265869140625,
        431.7611389160156,
        558.5822143554688,
        446.50067138671875,
        781.8605346679688,
        649.52392578125,
        1031.22900390625,
        871.4506225585938
      ],
      "eval_runtime": 2.2758,
      "eval_samples_per_second": 4.394,
      "eval_steps_per_second": 0.879,
      "step": 7805
    },
    {
      "epoch": 36.0,
      "eval_loss": 4.476213455200195,
      "eval_mean_perplexity": 673.9735595703125,
      "eval_perplexities": [
        570.093505859375,
        876.3085327148438,
        637.6167602539062,
        410.8495788574219,
        565.6197509765625,
        431.1340637207031,
        784.1497192382812,
        635.3842163085938,
        1020.8198852539062,
        807.7595825195312
      ],
      "eval_runtime": 2.2666,
      "eval_samples_per_second": 4.412,
      "eval_steps_per_second": 0.882,
      "step": 8028
    },
    {
      "epoch": 37.0,
      "eval_loss": 4.485024452209473,
      "eval_mean_perplexity": 678.1103912353516,
      "eval_perplexities": [
        524.7376708984375,
        897.5795288085938,
        604.8717651367188,
        421.68621826171875,
        562.4842529296875,
        447.8833312988281,
        743.5353393554688,
        641.0125122070312,
        1038.24658203125,
        899.0667114257812
      ],
      "eval_runtime": 2.2714,
      "eval_samples_per_second": 4.403,
      "eval_steps_per_second": 0.881,
      "step": 8251
    },
    {
      "epoch": 38.0,
      "eval_loss": 4.500662326812744,
      "eval_mean_perplexity": 667.2204986572266,
      "eval_perplexities": [
        517.3828125,
        849.9619140625,
        632.2133178710938,
        421.2346496582031,
        562.2227783203125,
        412.367919921875,
        801.9597778320312,
        634.914794921875,
        955.3936767578125,
        884.5533447265625
      ],
      "eval_runtime": 2.2586,
      "eval_samples_per_second": 4.428,
      "eval_steps_per_second": 0.886,
      "step": 8474
    },
    {
      "epoch": 39.0,
      "eval_loss": 4.507510662078857,
      "eval_mean_perplexity": 674.6702362060547,
      "eval_perplexities": [
        523.6810913085938,
        871.532958984375,
        702.7888793945312,
        412.113525390625,
        559.32421875,
        436.3467712402344,
        741.5978393554688,
        612.4104614257812,
        969.3262329101562,
        917.5803833007812
      ],
      "eval_runtime": 2.2926,
      "eval_samples_per_second": 4.362,
      "eval_steps_per_second": 0.872,
      "step": 8697
    },
    {
      "epoch": 40.0,
      "eval_loss": 4.5184006690979,
      "eval_mean_perplexity": 690.1411804199219,
      "eval_perplexities": [
        533.463623046875,
        880.5369262695312,
        677.1242065429688,
        468.8255920410156,
        524.9594116210938,
        481.3450012207031,
        724.6446533203125,
        644.5559692382812,
        1083.451416015625,
        882.5050048828125
      ],
      "eval_runtime": 2.2507,
      "eval_samples_per_second": 4.443,
      "eval_steps_per_second": 0.889,
      "step": 8920
    },
    {
      "epoch": 41.0,
      "eval_loss": 4.528339385986328,
      "eval_mean_perplexity": 686.3480163574219,
      "eval_perplexities": [
        540.8236083984375,
        941.376708984375,
        641.9276733398438,
        446.772216796875,
        562.5877685546875,
        455.73175048828125,
        764.4960327148438,
        647.960205078125,
        983.826416015625,
        877.977783203125
      ],
      "eval_runtime": 2.2901,
      "eval_samples_per_second": 4.367,
      "eval_steps_per_second": 0.873,
      "step": 9143
    },
    {
      "epoch": 42.0,
      "eval_loss": 4.525776386260986,
      "eval_mean_perplexity": 692.94345703125,
      "eval_perplexities": [
        512.2537841796875,
        996.9277954101562,
        626.796142578125,
        407.5407409667969,
        572.6943969726562,
        439.5016174316406,
        776.8684692382812,
        660.894775390625,
        1015.8809204101562,
        920.075927734375
      ],
      "eval_runtime": 2.2958,
      "eval_samples_per_second": 4.356,
      "eval_steps_per_second": 0.871,
      "step": 9366
    },
    {
      "epoch": 43.0,
      "eval_loss": 4.5432844161987305,
      "eval_mean_perplexity": 698.305697631836,
      "eval_perplexities": [
        520.623779296875,
        898.4200439453125,
        635.4287719726562,
        451.1359558105469,
        590.0744018554688,
        448.103759765625,
        752.5787353515625,
        682.3783569335938,
        1070.27294921875,
        934.0402221679688
      ],
      "eval_runtime": 2.8416,
      "eval_samples_per_second": 3.519,
      "eval_steps_per_second": 0.704,
      "step": 9589
    },
    {
      "epoch": 44.0,
      "eval_loss": 4.546627998352051,
      "eval_mean_perplexity": 698.2709533691407,
      "eval_perplexities": [
        496.5508728027344,
        900.9185180664062,
        651.6212768554688,
        429.4538879394531,
        572.0066528320312,
        429.689208984375,
        786.3671264648438,
        686.788818359375,
        1078.3935546875,
        950.9196166992188
      ],
      "eval_runtime": 2.2997,
      "eval_samples_per_second": 4.348,
      "eval_steps_per_second": 0.87,
      "step": 9812
    },
    {
      "epoch": 44.84,
      "grad_norm": 2.2081024646759033,
      "learning_rate": 5.15695067264574e-06,
      "loss": 2.4732,
      "step": 10000
    },
    {
      "epoch": 45.0,
      "eval_loss": 4.5482940673828125,
      "eval_mean_perplexity": 683.0953948974609,
      "eval_perplexities": [
        513.7635498046875,
        881.8916625976562,
        650.1325073242188,
        426.90740966796875,
        583.7874145507812,
        441.8877258300781,
        793.0048828125,
        660.0620727539062,
        992.5838012695312,
        886.9329223632812
      ],
      "eval_runtime": 2.6736,
      "eval_samples_per_second": 3.74,
      "eval_steps_per_second": 0.748,
      "step": 10035
    },
    {
      "epoch": 46.0,
      "eval_loss": 4.556254863739014,
      "eval_mean_perplexity": 694.8246520996094,
      "eval_perplexities": [
        484.3752136230469,
        864.9951171875,
        721.4644775390625,
        440.7280578613281,
        583.1522216796875,
        458.59246826171875,
        795.0460815429688,
        658.3065795898438,
        1043.74267578125,
        897.8436279296875
      ],
      "eval_runtime": 2.6592,
      "eval_samples_per_second": 3.761,
      "eval_steps_per_second": 0.752,
      "step": 10258
    },
    {
      "epoch": 47.0,
      "eval_loss": 4.565934181213379,
      "eval_mean_perplexity": 697.1294586181641,
      "eval_perplexities": [
        521.9780883789062,
        891.4306030273438,
        694.7803344726562,
        465.2270812988281,
        585.0350341796875,
        454.22308349609375,
        784.2135620117188,
        672.1557006835938,
        1016.0956420898438,
        886.1554565429688
      ],
      "eval_runtime": 2.3829,
      "eval_samples_per_second": 4.197,
      "eval_steps_per_second": 0.839,
      "step": 10481
    },
    {
      "epoch": 48.0,
      "eval_loss": 4.569299221038818,
      "eval_mean_perplexity": 698.8598907470703,
      "eval_perplexities": [
        523.372802734375,
        930.4688720703125,
        683.875732421875,
        451.5999755859375,
        597.7974243164062,
        447.0304870605469,
        766.9913940429688,
        671.3782348632812,
        1039.4984130859375,
        876.5855712890625
      ],
      "eval_runtime": 2.2797,
      "eval_samples_per_second": 4.387,
      "eval_steps_per_second": 0.877,
      "step": 10704
    },
    {
      "epoch": 49.0,
      "eval_loss": 4.569689750671387,
      "eval_mean_perplexity": 694.5519165039062,
      "eval_perplexities": [
        508.9573059082031,
        927.446533203125,
        672.3489379882812,
        473.5368347167969,
        568.0377807617188,
        454.00933837890625,
        733.6597900390625,
        695.4989624023438,
        1047.673828125,
        864.349853515625
      ],
      "eval_runtime": 2.3973,
      "eval_samples_per_second": 4.171,
      "eval_steps_per_second": 0.834,
      "step": 10927
    },
    {
      "epoch": 50.0,
      "eval_loss": 4.569555759429932,
      "eval_mean_perplexity": 698.6035186767579,
      "eval_perplexities": [
        517.4149169921875,
        924.535888671875,
        704.73291015625,
        465.9677429199219,
        577.629150390625,
        443.994140625,
        770.1861572265625,
        683.028076171875,
        1017.7510375976562,
        880.795166015625
      ],
      "eval_runtime": 2.2771,
      "eval_samples_per_second": 4.391,
      "eval_steps_per_second": 0.878,
      "step": 11150
    }
  ],
  "logging_steps": 5000,
  "max_steps": 11150,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 500,
  "total_flos": 1.1355181056e+16,
  "train_batch_size": 10,
  "trial_name": null,
  "trial_params": null
}