{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9203508421764635,
  "eval_steps": 500,
  "global_step": 83000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 3.2580924034118652,
      "learning_rate": 2.2172949002217296e-06,
      "loss": 10.2933,
      "step": 50
    },
    {
      "epoch": 0.0,
      "grad_norm": 2.4347386360168457,
      "learning_rate": 4.434589800443459e-06,
      "loss": 10.1894,
      "step": 100
    },
    {
      "epoch": 0.0,
      "grad_norm": 2.3895885944366455,
      "learning_rate": 6.651884700665188e-06,
      "loss": 10.1424,
      "step": 150
    },
    {
      "epoch": 0.0,
      "grad_norm": 2.129647731781006,
      "learning_rate": 8.869179600886918e-06,
      "loss": 10.0995,
      "step": 200
    },
    {
      "epoch": 0.0,
      "grad_norm": 2.3564186096191406,
      "learning_rate": 1.1086474501108649e-05,
      "loss": 10.0479,
      "step": 250
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.830551028251648,
      "learning_rate": 1.3303769401330377e-05,
      "loss": 9.9971,
      "step": 300
    },
    {
      "epoch": 0.0,
      "grad_norm": 2.1173911094665527,
      "learning_rate": 1.5521064301552106e-05,
      "loss": 9.9201,
      "step": 350
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.6636557579040527,
      "learning_rate": 1.7738359201773837e-05,
      "loss": 9.8562,
      "step": 400
    },
    {
      "epoch": 0.0,
      "grad_norm": 2.4503839015960693,
      "learning_rate": 1.9955654101995567e-05,
      "loss": 9.7599,
      "step": 450
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.822424054145813,
      "learning_rate": 2.2172949002217298e-05,
      "loss": 9.6608,
      "step": 500
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.6598998308181763,
      "learning_rate": 2.4390243902439026e-05,
      "loss": 9.55,
      "step": 550
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.8471707105636597,
      "learning_rate": 2.6607538802660753e-05,
      "loss": 9.4606,
      "step": 600
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.4833533763885498,
      "learning_rate": 2.8824833702882487e-05,
      "loss": 9.3283,
      "step": 650
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.688541054725647,
      "learning_rate": 3.104212860310421e-05,
      "loss": 9.2229,
      "step": 700
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.6466543674468994,
      "learning_rate": 3.325942350332594e-05,
      "loss": 9.1093,
      "step": 750
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.4169293642044067,
      "learning_rate": 3.547671840354767e-05,
      "loss": 8.9703,
      "step": 800
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.7079193592071533,
      "learning_rate": 3.7694013303769404e-05,
      "loss": 8.8351,
      "step": 850
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.5513204336166382,
      "learning_rate": 3.9911308203991135e-05,
      "loss": 8.7111,
      "step": 900
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.485573172569275,
      "learning_rate": 4.212860310421286e-05,
      "loss": 8.5627,
      "step": 950
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.511690616607666,
      "learning_rate": 4.4345898004434597e-05,
      "loss": 8.5042,
      "step": 1000
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.1478614807128906,
      "learning_rate": 4.656319290465632e-05,
      "loss": 8.3287,
      "step": 1050
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.4060652256011963,
      "learning_rate": 4.878048780487805e-05,
      "loss": 8.2341,
      "step": 1100
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.3950035572052002,
      "learning_rate": 5.099778270509978e-05,
      "loss": 8.1277,
      "step": 1150
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.5197688341140747,
      "learning_rate": 5.3215077605321506e-05,
      "loss": 8.0311,
      "step": 1200
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.3406693935394287,
      "learning_rate": 5.543237250554324e-05,
      "loss": 7.9824,
      "step": 1250
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.4520119428634644,
      "learning_rate": 5.7649667405764975e-05,
      "loss": 7.9948,
      "step": 1300
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.179124116897583,
      "learning_rate": 5.98669623059867e-05,
      "loss": 7.9144,
      "step": 1350
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.4039533138275146,
      "learning_rate": 6.208425720620842e-05,
      "loss": 7.8768,
      "step": 1400
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.5542700290679932,
      "learning_rate": 6.430155210643016e-05,
      "loss": 7.894,
      "step": 1450
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.4150550365447998,
      "learning_rate": 6.651884700665188e-05,
      "loss": 7.8409,
      "step": 1500
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.6647827625274658,
      "learning_rate": 6.873614190687362e-05,
      "loss": 7.91,
      "step": 1550
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.7795697450637817,
      "learning_rate": 7.095343680709535e-05,
      "loss": 7.8256,
      "step": 1600
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.933110237121582,
      "learning_rate": 7.317073170731707e-05,
      "loss": 7.8463,
      "step": 1650
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.1942570209503174,
      "learning_rate": 7.538802660753881e-05,
      "loss": 7.7827,
      "step": 1700
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.6759297847747803,
      "learning_rate": 7.760532150776053e-05,
      "loss": 7.8,
      "step": 1750
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.093256950378418,
      "learning_rate": 7.982261640798227e-05,
      "loss": 7.7461,
      "step": 1800
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.567872166633606,
      "learning_rate": 8.2039911308204e-05,
      "loss": 7.7338,
      "step": 1850
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.3017679452896118,
      "learning_rate": 8.425720620842572e-05,
      "loss": 7.804,
      "step": 1900
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.7510960102081299,
      "learning_rate": 8.647450110864746e-05,
      "loss": 7.7405,
      "step": 1950
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.7215120792388916,
      "learning_rate": 8.869179600886919e-05,
      "loss": 7.7429,
      "step": 2000
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.6202715635299683,
      "learning_rate": 9.090909090909092e-05,
      "loss": 7.6588,
      "step": 2050
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.5680756568908691,
      "learning_rate": 9.312638580931264e-05,
      "loss": 7.6224,
      "step": 2100
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.462240219116211,
      "learning_rate": 9.534368070953438e-05,
      "loss": 7.6851,
      "step": 2150
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.2018320560455322,
      "learning_rate": 9.75609756097561e-05,
      "loss": 7.6443,
      "step": 2200
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.9520208835601807,
      "learning_rate": 9.977827050997783e-05,
      "loss": 7.6456,
      "step": 2250
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.115421175956726,
      "learning_rate": 0.00010199556541019956,
      "loss": 7.5894,
      "step": 2300
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.6002250909805298,
      "learning_rate": 0.0001042128603104213,
      "loss": 7.6017,
      "step": 2350
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.6516796350479126,
      "learning_rate": 0.00010643015521064301,
      "loss": 7.4548,
      "step": 2400
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.2168257236480713,
      "learning_rate": 0.00010864745011086475,
      "loss": 7.5867,
      "step": 2450
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.5447593927383423,
      "learning_rate": 0.00011086474501108647,
      "loss": 7.5317,
      "step": 2500
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.6840906143188477,
      "learning_rate": 0.00011308203991130821,
      "loss": 7.5127,
      "step": 2550
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.2965503931045532,
      "learning_rate": 0.00011529933481152995,
      "loss": 7.4911,
      "step": 2600
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.643584966659546,
      "learning_rate": 0.00011751662971175166,
      "loss": 7.4416,
      "step": 2650
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.5419111251831055,
      "learning_rate": 0.0001197339246119734,
      "loss": 7.4944,
      "step": 2700
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.7774205207824707,
      "learning_rate": 0.00012195121951219512,
      "loss": 7.4244,
      "step": 2750
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.1709322929382324,
      "learning_rate": 0.00012416851441241685,
      "loss": 7.371,
      "step": 2800
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.5503411293029785,
      "learning_rate": 0.0001263858093126386,
      "loss": 7.3031,
      "step": 2850
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.7744035720825195,
      "learning_rate": 0.00012860310421286032,
      "loss": 7.3338,
      "step": 2900
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.2014000415802,
      "learning_rate": 0.00013082039911308205,
      "loss": 7.2962,
      "step": 2950
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.6716220378875732,
      "learning_rate": 0.00013303769401330377,
      "loss": 7.3348,
      "step": 3000
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.7045074701309204,
      "learning_rate": 0.0001352549889135255,
      "loss": 7.2864,
      "step": 3050
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.8933771848678589,
      "learning_rate": 0.00013747228381374724,
      "loss": 7.2744,
      "step": 3100
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.298779249191284,
      "learning_rate": 0.00013968957871396897,
      "loss": 7.2472,
      "step": 3150
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.3420922756195068,
      "learning_rate": 0.0001419068736141907,
      "loss": 7.3019,
      "step": 3200
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.9339039325714111,
      "learning_rate": 0.00014412416851441242,
      "loss": 7.2982,
      "step": 3250
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.69667387008667,
      "learning_rate": 0.00014634146341463414,
      "loss": 7.2851,
      "step": 3300
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.3124189376831055,
      "learning_rate": 0.0001485587583148559,
      "loss": 7.258,
      "step": 3350
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.975651741027832,
      "learning_rate": 0.00015077605321507762,
      "loss": 7.1275,
      "step": 3400
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.9704022407531738,
      "learning_rate": 0.00015299334811529934,
      "loss": 7.1473,
      "step": 3450
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.5047757625579834,
      "learning_rate": 0.00015521064301552106,
      "loss": 7.1096,
      "step": 3500
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.5465894937515259,
      "learning_rate": 0.0001574279379157428,
      "loss": 7.1501,
      "step": 3550
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.9557933807373047,
      "learning_rate": 0.00015964523281596454,
      "loss": 7.2033,
      "step": 3600
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.420116424560547,
      "learning_rate": 0.00016186252771618626,
      "loss": 7.1275,
      "step": 3650
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.114737033843994,
      "learning_rate": 0.000164079822616408,
      "loss": 7.0932,
      "step": 3700
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.3085389137268066,
      "learning_rate": 0.00016629711751662974,
      "loss": 7.0311,
      "step": 3750
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.5679140090942383,
      "learning_rate": 0.00016851441241685144,
      "loss": 6.9168,
      "step": 3800
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.8611838817596436,
      "learning_rate": 0.0001707317073170732,
      "loss": 7.0085,
      "step": 3850
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.8603994846343994,
      "learning_rate": 0.0001729490022172949,
      "loss": 6.9432,
      "step": 3900
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.4244627952575684,
      "learning_rate": 0.00017516629711751663,
      "loss": 6.9333,
      "step": 3950
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.177870750427246,
      "learning_rate": 0.00017738359201773839,
      "loss": 6.9499,
      "step": 4000
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.9320554733276367,
      "learning_rate": 0.00017960088691796008,
      "loss": 6.8204,
      "step": 4050
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.5062849521636963,
      "learning_rate": 0.00018181818181818183,
      "loss": 6.9505,
      "step": 4100
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.9272422790527344,
      "learning_rate": 0.00018403547671840356,
      "loss": 6.8701,
      "step": 4150
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.0309596061706543,
      "learning_rate": 0.00018625277161862528,
      "loss": 6.924,
      "step": 4200
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.0265886783599854,
      "learning_rate": 0.00018847006651884703,
      "loss": 6.9223,
      "step": 4250
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.5160486698150635,
      "learning_rate": 0.00019068736141906876,
      "loss": 6.8708,
      "step": 4300
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.613301992416382,
      "learning_rate": 0.00019290465631929045,
      "loss": 6.8937,
      "step": 4350
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.3031229972839355,
      "learning_rate": 0.0001951219512195122,
      "loss": 6.8337,
      "step": 4400
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.54779052734375,
      "learning_rate": 0.00019733924611973393,
      "loss": 6.8334,
      "step": 4450
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.8277971744537354,
      "learning_rate": 0.00019955654101995565,
      "loss": 6.7925,
      "step": 4500
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.0113885402679443,
      "learning_rate": 0.00019999989242739025,
      "loss": 6.8458,
      "step": 4550
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.2395377159118652,
      "learning_rate": 0.00019999945541405976,
      "loss": 6.6251,
      "step": 4600
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.445993423461914,
      "learning_rate": 0.0001999986822381884,
      "loss": 6.8099,
      "step": 4650
    },
    {
      "epoch": 0.05,
      "grad_norm": 4.077752590179443,
      "learning_rate": 0.0001999975729023753,
      "loss": 6.8053,
      "step": 4700
    },
    {
      "epoch": 0.05,
      "grad_norm": 3.167569875717163,
      "learning_rate": 0.00019999612741034963,
      "loss": 6.7706,
      "step": 4750
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.893659234046936,
      "learning_rate": 0.00019999434576697066,
      "loss": 6.8245,
      "step": 4800
    },
    {
      "epoch": 0.05,
      "grad_norm": 3.6101326942443848,
      "learning_rate": 0.00019999222797822762,
      "loss": 6.7407,
      "step": 4850
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.2858726978302,
      "learning_rate": 0.00019998977405123974,
      "loss": 6.74,
      "step": 4900
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.9325459003448486,
      "learning_rate": 0.0001999869839942563,
      "loss": 6.716,
      "step": 4950
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.0043437480926514,
      "learning_rate": 0.00019998385781665643,
      "loss": 6.6003,
      "step": 5000
    },
    {
      "epoch": 0.06,
      "grad_norm": 4.151523113250732,
      "learning_rate": 0.00019998039552894924,
      "loss": 6.6801,
      "step": 5050
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.8407771587371826,
      "learning_rate": 0.00019997659714277372,
      "loss": 6.608,
      "step": 5100
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.230713129043579,
      "learning_rate": 0.00019997246267089867,
      "loss": 6.6479,
      "step": 5150
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.2546942234039307,
      "learning_rate": 0.0001999679921272227,
      "loss": 6.6548,
      "step": 5200
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.180986166000366,
      "learning_rate": 0.00019996318552677425,
      "loss": 6.6851,
      "step": 5250
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.341231346130371,
      "learning_rate": 0.00019995804288571134,
      "loss": 6.547,
      "step": 5300
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.1117124557495117,
      "learning_rate": 0.00019995256422132172,
      "loss": 6.7072,
      "step": 5350
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.0082530975341797,
      "learning_rate": 0.0001999467495520227,
      "loss": 6.5422,
      "step": 5400
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.409489870071411,
      "learning_rate": 0.0001999405988973611,
      "loss": 6.3716,
      "step": 5450
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.649052381515503,
      "learning_rate": 0.00019993411227801328,
      "loss": 6.6434,
      "step": 5500
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.081116199493408,
      "learning_rate": 0.00019992728971578492,
      "loss": 6.4624,
      "step": 5550
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.1578280925750732,
      "learning_rate": 0.00019992013123361102,
      "loss": 6.5416,
      "step": 5600
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.7874557971954346,
      "learning_rate": 0.0001999126368555559,
      "loss": 6.4512,
      "step": 5650
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.7693099975585938,
      "learning_rate": 0.00019990480660681293,
      "loss": 6.5105,
      "step": 5700
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.4338185787200928,
      "learning_rate": 0.00019989680712666593,
      "loss": 6.5092,
      "step": 5750
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.656937837600708,
      "learning_rate": 0.00019988831193270577,
      "loss": 6.4269,
      "step": 5800
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.857292652130127,
      "learning_rate": 0.00019987948094982952,
      "loss": 6.4387,
      "step": 5850
    },
    {
      "epoch": 0.07,
      "grad_norm": 3.4963467121124268,
      "learning_rate": 0.00019987031420772385,
      "loss": 6.3851,
      "step": 5900
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.602522611618042,
      "learning_rate": 0.00019986081173720396,
      "loss": 6.3413,
      "step": 5950
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.6455273628234863,
      "learning_rate": 0.00019985097357021385,
      "loss": 6.2965,
      "step": 6000
    },
    {
      "epoch": 0.07,
      "grad_norm": 3.5592167377471924,
      "learning_rate": 0.0001998407997398259,
      "loss": 6.4293,
      "step": 6050
    },
    {
      "epoch": 0.07,
      "grad_norm": 3.6016533374786377,
      "learning_rate": 0.00019983029028024094,
      "loss": 6.2897,
      "step": 6100
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.5536839962005615,
      "learning_rate": 0.000199819445226788,
      "loss": 6.3157,
      "step": 6150
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.0514349937438965,
      "learning_rate": 0.00019980826461592427,
      "loss": 6.3847,
      "step": 6200
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.72495174407959,
      "learning_rate": 0.00019979674848523505,
      "loss": 6.3517,
      "step": 6250
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.4264872074127197,
      "learning_rate": 0.00019978489687343335,
      "loss": 6.2533,
      "step": 6300
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.8361423015594482,
      "learning_rate": 0.0001997727098203602,
      "loss": 6.3654,
      "step": 6350
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.9690892696380615,
      "learning_rate": 0.00019976018736698404,
      "loss": 6.3968,
      "step": 6400
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.6132867336273193,
      "learning_rate": 0.0001997473295554009,
      "loss": 6.3444,
      "step": 6450
    },
    {
      "epoch": 0.07,
      "grad_norm": 4.820697784423828,
      "learning_rate": 0.00019973413642883424,
      "loss": 6.2019,
      "step": 6500
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.2316782474517822,
      "learning_rate": 0.00019972060803163458,
      "loss": 6.2049,
      "step": 6550
    },
    {
      "epoch": 0.07,
      "grad_norm": 3.9528305530548096,
      "learning_rate": 0.00019970674440927957,
      "loss": 6.1718,
      "step": 6600
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.891073226928711,
      "learning_rate": 0.0001996925456083738,
      "loss": 6.2393,
      "step": 6650
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.813270092010498,
      "learning_rate": 0.00019967801167664853,
      "loss": 6.2116,
      "step": 6700
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.2726826667785645,
      "learning_rate": 0.00019966314266296173,
      "loss": 6.1521,
      "step": 6750
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.3895318508148193,
      "learning_rate": 0.00019964793861729772,
      "loss": 6.1072,
      "step": 6800
    },
    {
      "epoch": 0.08,
      "grad_norm": 3.190431833267212,
      "learning_rate": 0.000199632399590767,
      "loss": 6.2009,
      "step": 6850
    },
    {
      "epoch": 0.08,
      "grad_norm": 3.79266095161438,
      "learning_rate": 0.00019961652563560634,
      "loss": 6.028,
      "step": 6900
    },
    {
      "epoch": 0.08,
      "grad_norm": 3.260039806365967,
      "learning_rate": 0.00019960031680517826,
      "loss": 6.0733,
      "step": 6950
    },
    {
      "epoch": 0.08,
      "grad_norm": 3.0739686489105225,
      "learning_rate": 0.0001995837731539711,
      "loss": 6.0521,
      "step": 7000
    },
    {
      "epoch": 0.08,
      "grad_norm": 3.0517771244049072,
      "learning_rate": 0.00019956689473759872,
      "loss": 6.0544,
      "step": 7050
    },
    {
      "epoch": 0.08,
      "grad_norm": 3.9524648189544678,
      "learning_rate": 0.0001995496816128003,
      "loss": 6.1326,
      "step": 7100
    },
    {
      "epoch": 0.08,
      "grad_norm": 4.498497486114502,
      "learning_rate": 0.00019953213383744033,
      "loss": 6.236,
      "step": 7150
    },
    {
      "epoch": 0.08,
      "grad_norm": 4.157576084136963,
      "learning_rate": 0.00019951425147050807,
      "loss": 5.9898,
      "step": 7200
    },
    {
      "epoch": 0.08,
      "grad_norm": 3.9297516345977783,
      "learning_rate": 0.00019949603457211775,
      "loss": 6.086,
      "step": 7250
    },
    {
      "epoch": 0.08,
      "grad_norm": 3.3214786052703857,
      "learning_rate": 0.00019947748320350804,
      "loss": 5.9589,
      "step": 7300
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.8847291469573975,
      "learning_rate": 0.00019945859742704201,
      "loss": 6.1931,
      "step": 7350
    },
    {
      "epoch": 0.08,
      "grad_norm": 3.387896776199341,
      "learning_rate": 0.00019943937730620702,
      "loss": 6.0539,
      "step": 7400
    },
    {
      "epoch": 0.08,
      "grad_norm": 3.1214797496795654,
      "learning_rate": 0.00019941982290561417,
      "loss": 6.0288,
      "step": 7450
    },
    {
      "epoch": 0.08,
      "grad_norm": 3.7995123863220215,
      "learning_rate": 0.00019939993429099841,
      "loss": 6.0526,
      "step": 7500
    },
    {
      "epoch": 0.08,
      "grad_norm": 4.788393974304199,
      "learning_rate": 0.00019937971152921818,
      "loss": 5.9799,
      "step": 7550
    },
    {
      "epoch": 0.08,
      "grad_norm": 4.009220123291016,
      "learning_rate": 0.0001993591546882552,
      "loss": 6.1223,
      "step": 7600
    },
    {
      "epoch": 0.08,
      "grad_norm": 3.5576276779174805,
      "learning_rate": 0.00019933826383721428,
      "loss": 5.989,
      "step": 7650
    },
    {
      "epoch": 0.09,
      "grad_norm": 3.1287412643432617,
      "learning_rate": 0.00019931703904632294,
      "loss": 6.0542,
      "step": 7700
    },
    {
      "epoch": 0.09,
      "grad_norm": 3.6518595218658447,
      "learning_rate": 0.00019929548038693146,
      "loss": 6.041,
      "step": 7750
    },
    {
      "epoch": 0.09,
      "grad_norm": 3.268080472946167,
      "learning_rate": 0.0001992735879315123,
      "loss": 5.888,
      "step": 7800
    },
    {
      "epoch": 0.09,
      "grad_norm": 3.6055593490600586,
      "learning_rate": 0.00019925136175366007,
      "loss": 5.913,
      "step": 7850
    },
    {
      "epoch": 0.09,
      "grad_norm": 4.866463661193848,
      "learning_rate": 0.00019922880192809137,
      "loss": 5.9858,
      "step": 7900
    },
    {
      "epoch": 0.09,
      "grad_norm": 3.44808292388916,
      "learning_rate": 0.00019920590853064423,
      "loss": 5.7686,
      "step": 7950
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.9507765769958496,
      "learning_rate": 0.00019918268163827808,
      "loss": 5.8557,
      "step": 8000
    },
    {
      "epoch": 0.09,
      "grad_norm": 3.441870927810669,
      "learning_rate": 0.00019915912132907352,
      "loss": 5.8268,
      "step": 8050
    },
    {
      "epoch": 0.09,
      "grad_norm": 3.838809013366699,
      "learning_rate": 0.00019913522768223182,
      "loss": 5.9833,
      "step": 8100
    },
    {
      "epoch": 0.09,
      "grad_norm": 4.165487289428711,
      "learning_rate": 0.00019911100077807498,
      "loss": 5.7422,
      "step": 8150
    },
    {
      "epoch": 0.09,
      "grad_norm": 3.5947463512420654,
      "learning_rate": 0.0001990864406980452,
      "loss": 5.7479,
      "step": 8200
    },
    {
      "epoch": 0.09,
      "grad_norm": 4.130446434020996,
      "learning_rate": 0.00019906154752470472,
      "loss": 5.7767,
      "step": 8250
    },
    {
      "epoch": 0.09,
      "grad_norm": 4.866550922393799,
      "learning_rate": 0.00019903632134173554,
      "loss": 5.7681,
      "step": 8300
    },
    {
      "epoch": 0.09,
      "grad_norm": 3.2839725017547607,
      "learning_rate": 0.00019901076223393903,
      "loss": 5.6656,
      "step": 8350
    },
    {
      "epoch": 0.09,
      "grad_norm": 3.0762476921081543,
      "learning_rate": 0.0001989848702872359,
      "loss": 5.789,
      "step": 8400
    },
    {
      "epoch": 0.09,
      "grad_norm": 3.7109107971191406,
      "learning_rate": 0.00019895864558866556,
      "loss": 5.773,
      "step": 8450
    },
    {
      "epoch": 0.09,
      "grad_norm": 5.400998115539551,
      "learning_rate": 0.00019893208822638618,
      "loss": 5.7506,
      "step": 8500
    },
    {
      "epoch": 0.09,
      "grad_norm": 3.3062849044799805,
      "learning_rate": 0.00019890519828967413,
      "loss": 5.7515,
      "step": 8550
    },
    {
      "epoch": 0.1,
      "grad_norm": 4.109920501708984,
      "learning_rate": 0.00019887797586892373,
      "loss": 5.7972,
      "step": 8600
    },
    {
      "epoch": 0.1,
      "grad_norm": 3.4838390350341797,
      "learning_rate": 0.00019885042105564717,
      "loss": 5.6753,
      "step": 8650
    },
    {
      "epoch": 0.1,
      "grad_norm": 4.251760959625244,
      "learning_rate": 0.00019882253394247381,
      "loss": 5.6303,
      "step": 8700
    },
    {
      "epoch": 0.1,
      "grad_norm": 4.042376518249512,
      "learning_rate": 0.00019879431462315025,
      "loss": 5.5753,
      "step": 8750
    },
    {
      "epoch": 0.1,
      "grad_norm": 4.239652633666992,
      "learning_rate": 0.0001987657631925398,
      "loss": 5.5335,
      "step": 8800
    },
    {
      "epoch": 0.1,
      "grad_norm": 5.15481424331665,
      "learning_rate": 0.00019873687974662215,
      "loss": 5.5396,
      "step": 8850
    },
    {
      "epoch": 0.1,
      "grad_norm": 4.36835241317749,
      "learning_rate": 0.00019870766438249317,
      "loss": 5.6017,
      "step": 8900
    },
    {
      "epoch": 0.1,
      "grad_norm": 4.165258407592773,
      "learning_rate": 0.00019867811719836452,
      "loss": 5.7228,
      "step": 8950
    },
    {
      "epoch": 0.1,
      "grad_norm": 4.125988006591797,
      "learning_rate": 0.0001986482382935633,
      "loss": 5.5787,
      "step": 9000
    },
    {
      "epoch": 0.1,
      "grad_norm": 4.177731037139893,
      "learning_rate": 0.0001986180277685317,
      "loss": 5.5829,
      "step": 9050
    },
    {
      "epoch": 0.1,
      "grad_norm": 5.006561279296875,
      "learning_rate": 0.00019858748572482683,
      "loss": 5.5466,
      "step": 9100
    },
    {
      "epoch": 0.1,
      "grad_norm": 4.33070182800293,
      "learning_rate": 0.00019855661226512007,
      "loss": 5.5544,
      "step": 9150
    },
    {
      "epoch": 0.1,
      "grad_norm": 4.358560085296631,
      "learning_rate": 0.00019852540749319708,
      "loss": 5.4599,
      "step": 9200
    },
    {
      "epoch": 0.1,
      "grad_norm": 4.536096096038818,
      "learning_rate": 0.00019849387151395708,
      "loss": 5.4983,
      "step": 9250
    },
    {
      "epoch": 0.1,
      "grad_norm": 4.66163444519043,
      "learning_rate": 0.0001984620044334129,
      "loss": 5.4097,
      "step": 9300
    },
    {
      "epoch": 0.1,
      "grad_norm": 4.4319233894348145,
      "learning_rate": 0.00019842980635869024,
      "loss": 5.4093,
      "step": 9350
    },
    {
      "epoch": 0.1,
      "grad_norm": 4.98419713973999,
      "learning_rate": 0.0001983972773980276,
      "loss": 5.4056,
      "step": 9400
    },
    {
      "epoch": 0.1,
      "grad_norm": 3.6354339122772217,
      "learning_rate": 0.0001983644176607757,
      "loss": 5.3171,
      "step": 9450
    },
    {
      "epoch": 0.11,
      "grad_norm": 4.495342254638672,
      "learning_rate": 0.00019833122725739736,
      "loss": 5.4521,
      "step": 9500
    },
    {
      "epoch": 0.11,
      "grad_norm": 4.5558671951293945,
      "learning_rate": 0.00019829770629946678,
      "loss": 5.5158,
      "step": 9550
    },
    {
      "epoch": 0.11,
      "grad_norm": 3.7165732383728027,
      "learning_rate": 0.00019826385489966957,
      "loss": 5.301,
      "step": 9600
    },
    {
      "epoch": 0.11,
      "grad_norm": 6.030915260314941,
      "learning_rate": 0.00019822967317180204,
      "loss": 5.3316,
      "step": 9650
    },
    {
      "epoch": 0.11,
      "grad_norm": 5.385923385620117,
      "learning_rate": 0.00019819516123077094,
      "loss": 5.3844,
      "step": 9700
    },
    {
      "epoch": 0.11,
      "grad_norm": 4.383516788482666,
      "learning_rate": 0.00019816101926755305,
      "loss": 5.2995,
      "step": 9750
    },
    {
      "epoch": 0.11,
      "grad_norm": 4.446406364440918,
      "learning_rate": 0.00019812585384780055,
      "loss": 5.386,
      "step": 9800
    },
    {
      "epoch": 0.11,
      "grad_norm": 4.345483303070068,
      "learning_rate": 0.00019809035856388805,
      "loss": 5.2815,
      "step": 9850
    },
    {
      "epoch": 0.11,
      "grad_norm": 4.791261672973633,
      "learning_rate": 0.00019805453353513813,
      "loss": 5.3757,
      "step": 9900
    },
    {
      "epoch": 0.11,
      "grad_norm": 5.622151851654053,
      "learning_rate": 0.00019801837888198172,
      "loss": 5.4405,
      "step": 9950
    },
    {
      "epoch": 0.11,
      "grad_norm": 4.934606075286865,
      "learning_rate": 0.0001979818947259579,
      "loss": 5.139,
      "step": 10000
    },
    {
      "epoch": 0.11,
      "grad_norm": 3.9659693241119385,
      "learning_rate": 0.0001979450811897134,
      "loss": 5.1726,
      "step": 10050
    },
    {
      "epoch": 0.11,
      "grad_norm": 5.214992046356201,
      "learning_rate": 0.00019790793839700226,
      "loss": 5.2864,
      "step": 10100
    },
    {
      "epoch": 0.11,
      "grad_norm": 4.5359601974487305,
      "learning_rate": 0.00019787046647268524,
      "loss": 5.1443,
      "step": 10150
    },
    {
      "epoch": 0.11,
      "grad_norm": 4.26462984085083,
      "learning_rate": 0.00019783266554272962,
      "loss": 5.0597,
      "step": 10200
    },
    {
      "epoch": 0.11,
      "grad_norm": 5.053945064544678,
      "learning_rate": 0.00019779453573420873,
      "loss": 5.2946,
      "step": 10250
    },
    {
      "epoch": 0.11,
      "grad_norm": 6.082211494445801,
      "learning_rate": 0.00019775607717530127,
      "loss": 5.2075,
      "step": 10300
    },
    {
      "epoch": 0.11,
      "grad_norm": 4.107390403747559,
      "learning_rate": 0.00019771728999529132,
      "loss": 5.1394,
      "step": 10350
    },
    {
      "epoch": 0.12,
      "grad_norm": 4.58411169052124,
      "learning_rate": 0.00019767817432456752,
      "loss": 5.1064,
      "step": 10400
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.38965892791748,
      "learning_rate": 0.00019763952239228627,
      "loss": 5.0808,
      "step": 10450
    },
    {
      "epoch": 0.12,
      "grad_norm": 3.885803699493408,
      "learning_rate": 0.00019759975669894338,
      "loss": 5.0664,
      "step": 10500
    },
    {
      "epoch": 0.12,
      "grad_norm": 4.1605916023254395,
      "learning_rate": 0.00019755966290999167,
      "loss": 5.2469,
      "step": 10550
    },
    {
      "epoch": 0.12,
      "grad_norm": 4.821887016296387,
      "learning_rate": 0.00019751924116021225,
      "loss": 5.2451,
      "step": 10600
    },
    {
      "epoch": 0.12,
      "grad_norm": 3.865694761276245,
      "learning_rate": 0.00019747849158548858,
      "loss": 5.2334,
      "step": 10650
    },
    {
      "epoch": 0.12,
      "grad_norm": 3.640681028366089,
      "learning_rate": 0.00019743741432280625,
      "loss": 5.1206,
      "step": 10700
    },
    {
      "epoch": 0.12,
      "grad_norm": 4.04166316986084,
      "learning_rate": 0.00019739600951025236,
      "loss": 5.0059,
      "step": 10750
    },
    {
      "epoch": 0.12,
      "grad_norm": 4.637605667114258,
      "learning_rate": 0.00019735427728701516,
      "loss": 5.0302,
      "step": 10800
    },
    {
      "epoch": 0.12,
      "grad_norm": 4.08723783493042,
      "learning_rate": 0.0001973122177933835,
      "loss": 5.1551,
      "step": 10850
    },
    {
      "epoch": 0.12,
      "grad_norm": 3.7944953441619873,
      "learning_rate": 0.00019726983117074643,
      "loss": 5.0665,
      "step": 10900
    },
    {
      "epoch": 0.12,
      "grad_norm": 5.2847371101379395,
      "learning_rate": 0.00019722711756159266,
      "loss": 5.2212,
      "step": 10950
    },
    {
      "epoch": 0.12,
      "grad_norm": 4.109150409698486,
      "learning_rate": 0.00019718407710951012,
      "loss": 5.2645,
      "step": 11000
    },
    {
      "epoch": 0.12,
      "grad_norm": 4.127768039703369,
      "learning_rate": 0.0001971407099591855,
      "loss": 5.0395,
      "step": 11050
    },
    {
      "epoch": 0.12,
      "grad_norm": 5.058667182922363,
      "learning_rate": 0.00019709701625640367,
      "loss": 5.0247,
      "step": 11100
    },
    {
      "epoch": 0.12,
      "grad_norm": 5.4407267570495605,
      "learning_rate": 0.00019705299614804732,
      "loss": 4.9935,
      "step": 11150
    },
    {
      "epoch": 0.12,
      "grad_norm": 3.7877707481384277,
      "learning_rate": 0.00019700864978209636,
      "loss": 5.074,
      "step": 11200
    },
    {
      "epoch": 0.12,
      "grad_norm": 3.777330160140991,
      "learning_rate": 0.00019696397730762746,
      "loss": 5.0458,
      "step": 11250
    },
    {
      "epoch": 0.13,
      "grad_norm": 4.143067836761475,
      "learning_rate": 0.0001969189788748136,
      "loss": 4.9375,
      "step": 11300
    },
    {
      "epoch": 0.13,
      "grad_norm": 5.560107231140137,
      "learning_rate": 0.00019687365463492344,
      "loss": 4.8285,
      "step": 11350
    },
    {
      "epoch": 0.13,
      "grad_norm": 4.057905197143555,
      "learning_rate": 0.00019682800474032095,
      "loss": 4.9753,
      "step": 11400
    },
    {
      "epoch": 0.13,
      "grad_norm": 3.835442066192627,
      "learning_rate": 0.00019678202934446482,
      "loss": 4.9368,
      "step": 11450
    },
    {
      "epoch": 0.13,
      "grad_norm": 5.135551929473877,
      "learning_rate": 0.0001967357286019079,
      "loss": 4.9994,
      "step": 11500
    },
    {
      "epoch": 0.13,
      "grad_norm": 4.615053653717041,
      "learning_rate": 0.00019668910266829685,
      "loss": 5.0182,
      "step": 11550
    },
    {
      "epoch": 0.13,
      "grad_norm": 4.474258899688721,
      "learning_rate": 0.0001966421517003714,
      "loss": 4.8704,
      "step": 11600
    },
    {
      "epoch": 0.13,
      "grad_norm": 4.264945030212402,
      "learning_rate": 0.00019659487585596406,
      "loss": 4.9076,
      "step": 11650
    },
    {
      "epoch": 0.13,
      "grad_norm": 4.091209411621094,
      "learning_rate": 0.00019654727529399925,
      "loss": 4.7135,
      "step": 11700
    },
    {
      "epoch": 0.13,
      "grad_norm": 4.154038429260254,
      "learning_rate": 0.00019649935017449318,
      "loss": 4.8239,
      "step": 11750
    },
    {
      "epoch": 0.13,
      "grad_norm": 3.697162628173828,
      "learning_rate": 0.00019645110065855305,
      "loss": 4.9972,
      "step": 11800
    },
    {
      "epoch": 0.13,
      "grad_norm": 4.0024847984313965,
      "learning_rate": 0.00019640252690837645,
      "loss": 4.8854,
      "step": 11850
    },
    {
      "epoch": 0.13,
      "grad_norm": 3.9416885375976562,
      "learning_rate": 0.0001963536290872511,
      "loss": 4.8547,
      "step": 11900
    },
    {
      "epoch": 0.13,
      "grad_norm": 3.978651285171509,
      "learning_rate": 0.000196304407359554,
      "loss": 4.7873,
      "step": 11950
    },
    {
      "epoch": 0.13,
      "grad_norm": 4.435175895690918,
      "learning_rate": 0.0001962548618907511,
      "loss": 4.8124,
      "step": 12000
    },
    {
      "epoch": 0.13,
      "grad_norm": 3.8776824474334717,
      "learning_rate": 0.00019620499284739662,
      "loss": 4.8896,
      "step": 12050
    },
    {
      "epoch": 0.13,
      "grad_norm": 5.041496276855469,
      "learning_rate": 0.00019615480039713248,
      "loss": 4.8343,
      "step": 12100
    },
    {
      "epoch": 0.13,
      "grad_norm": 4.18281888961792,
      "learning_rate": 0.00019610428470868784,
      "loss": 4.8559,
      "step": 12150
    },
    {
      "epoch": 0.14,
      "grad_norm": 4.223630905151367,
      "learning_rate": 0.00019605344595187844,
      "loss": 4.8153,
      "step": 12200
    },
    {
      "epoch": 0.14,
      "grad_norm": 4.63677453994751,
      "learning_rate": 0.0001960022842976061,
      "loss": 4.7951,
      "step": 12250
    },
    {
      "epoch": 0.14,
      "grad_norm": 4.188296794891357,
      "learning_rate": 0.00019595079991785802,
      "loss": 4.8904,
      "step": 12300
    },
    {
      "epoch": 0.14,
      "grad_norm": 4.402559280395508,
      "learning_rate": 0.00019589899298570634,
      "loss": 4.7851,
      "step": 12350
    },
    {
      "epoch": 0.14,
      "grad_norm": 5.976877212524414,
      "learning_rate": 0.00019584686367530755,
      "loss": 4.6431,
      "step": 12400
    },
    {
      "epoch": 0.14,
      "grad_norm": 4.849298477172852,
      "learning_rate": 0.0001957944121619018,
      "loss": 4.7544,
      "step": 12450
    },
    {
      "epoch": 0.14,
      "grad_norm": 4.932714462280273,
      "learning_rate": 0.0001957416386218124,
      "loss": 4.6811,
      "step": 12500
    },
    {
      "epoch": 0.14,
      "grad_norm": 4.682474136352539,
      "learning_rate": 0.00019568854323244515,
      "loss": 4.799,
      "step": 12550
    },
    {
      "epoch": 0.14,
      "grad_norm": 5.228520393371582,
      "learning_rate": 0.00019563619766470511,
      "loss": 4.7622,
      "step": 12600
    },
    {
      "epoch": 0.14,
      "grad_norm": 4.093870162963867,
      "learning_rate": 0.00019558246554138458,
      "loss": 4.7369,
      "step": 12650
    },
    {
      "epoch": 0.14,
      "grad_norm": 5.248356342315674,
      "learning_rate": 0.0001955284121038694,
      "loss": 4.7519,
      "step": 12700
    },
    {
      "epoch": 0.14,
      "grad_norm": 3.924299955368042,
      "learning_rate": 0.00019547403753386803,
      "loss": 4.6441,
      "step": 12750
    },
    {
      "epoch": 0.14,
      "grad_norm": 4.972569942474365,
      "learning_rate": 0.00019542043906868188,
      "loss": 4.7192,
      "step": 12800
    },
    {
      "epoch": 0.14,
      "grad_norm": 5.033604145050049,
      "learning_rate": 0.00019536542919665846,
      "loss": 4.6397,
      "step": 12850
    },
    {
      "epoch": 0.14,
      "grad_norm": 5.222695350646973,
      "learning_rate": 0.00019531009874003928,
      "loss": 4.6309,
      "step": 12900
    },
    {
      "epoch": 0.14,
      "grad_norm": 3.810999631881714,
      "learning_rate": 0.00019525444788482562,
      "loss": 4.6513,
      "step": 12950
    },
    {
      "epoch": 0.14,
      "grad_norm": 5.272600173950195,
      "learning_rate": 0.00019519847681809585,
      "loss": 4.8001,
      "step": 13000
    },
    {
      "epoch": 0.14,
      "grad_norm": 4.836308002471924,
      "learning_rate": 0.00019514218572800468,
      "loss": 4.7101,
      "step": 13050
    },
    {
      "epoch": 0.15,
      "grad_norm": 4.598148345947266,
      "learning_rate": 0.00019508557480378276,
      "loss": 4.5578,
      "step": 13100
    },
    {
      "epoch": 0.15,
      "grad_norm": 3.910820722579956,
      "learning_rate": 0.0001950286442357358,
      "loss": 4.7124,
      "step": 13150
    },
    {
      "epoch": 0.15,
      "grad_norm": 3.856081962585449,
      "learning_rate": 0.00019497139421524416,
      "loss": 4.7563,
      "step": 13200
    },
    {
      "epoch": 0.15,
      "grad_norm": 4.151907920837402,
      "learning_rate": 0.00019491382493476195,
      "loss": 4.6726,
      "step": 13250
    },
    {
      "epoch": 0.15,
      "grad_norm": 4.349935054779053,
      "learning_rate": 0.0001948559365878166,
      "loss": 4.6341,
      "step": 13300
    },
    {
      "epoch": 0.15,
      "grad_norm": 3.8229756355285645,
      "learning_rate": 0.00019479772936900811,
      "loss": 4.6183,
      "step": 13350
    },
    {
      "epoch": 0.15,
      "grad_norm": 5.495506286621094,
      "learning_rate": 0.0001947392034740084,
      "loss": 4.6608,
      "step": 13400
    },
    {
      "epoch": 0.15,
      "grad_norm": 4.307513236999512,
      "learning_rate": 0.00019468035909956072,
      "loss": 4.6805,
      "step": 13450
    },
    {
      "epoch": 0.15,
      "grad_norm": 3.939659595489502,
      "learning_rate": 0.0001946211964434788,
      "loss": 4.679,
      "step": 13500
    },
    {
      "epoch": 0.15,
      "grad_norm": 5.444967269897461,
      "learning_rate": 0.00019456171570464653,
      "loss": 4.7195,
      "step": 13550
    },
    {
      "epoch": 0.15,
      "grad_norm": 4.513270854949951,
      "learning_rate": 0.00019450191708301687,
      "loss": 4.5367,
      "step": 13600
    },
    {
      "epoch": 0.15,
      "grad_norm": 4.617405414581299,
      "learning_rate": 0.00019444180077961146,
      "loss": 4.5742,
      "step": 13650
    },
    {
      "epoch": 0.15,
      "grad_norm": 4.580646991729736,
      "learning_rate": 0.00019438136699652001,
      "loss": 4.4936,
      "step": 13700
    },
    {
      "epoch": 0.15,
      "grad_norm": 4.657532691955566,
      "learning_rate": 0.00019432061593689927,
      "loss": 4.6877,
      "step": 13750
    },
    {
      "epoch": 0.15,
      "grad_norm": 5.374803066253662,
      "learning_rate": 0.0001942595478049727,
      "loss": 4.6101,
      "step": 13800
    },
    {
      "epoch": 0.15,
      "grad_norm": 5.1111650466918945,
      "learning_rate": 0.00019419816280602962,
      "loss": 4.6185,
      "step": 13850
    },
    {
      "epoch": 0.15,
      "grad_norm": 5.18306303024292,
      "learning_rate": 0.00019413646114642446,
      "loss": 4.5524,
      "step": 13900
    },
    {
      "epoch": 0.15,
      "grad_norm": 4.411191463470459,
      "learning_rate": 0.00019407444303357624,
      "loss": 4.4346,
      "step": 13950
    },
    {
      "epoch": 0.16,
      "grad_norm": 4.161925792694092,
      "learning_rate": 0.0001940121086759678,
      "loss": 4.3702,
      "step": 14000
    },
    {
      "epoch": 0.16,
      "grad_norm": 5.059813022613525,
      "learning_rate": 0.000193949458283145,
      "loss": 4.5351,
      "step": 14050
    },
    {
      "epoch": 0.16,
      "grad_norm": 5.563150882720947,
      "learning_rate": 0.00019388649206571616,
      "loss": 4.477,
      "step": 14100
    },
    {
      "epoch": 0.16,
      "grad_norm": 5.1144609451293945,
      "learning_rate": 0.00019382321023535127,
      "loss": 4.6033,
      "step": 14150
    },
    {
      "epoch": 0.16,
      "grad_norm": 4.734794616699219,
      "learning_rate": 0.00019375961300478127,
      "loss": 4.5287,
      "step": 14200
    },
    {
      "epoch": 0.16,
      "grad_norm": 4.543684959411621,
      "learning_rate": 0.00019369570058779743,
      "loss": 4.4474,
      "step": 14250
    },
    {
      "epoch": 0.16,
      "grad_norm": 5.4647979736328125,
      "learning_rate": 0.00019363147319925047,
      "loss": 4.3806,
      "step": 14300
    },
    {
      "epoch": 0.16,
      "grad_norm": 5.058681964874268,
      "learning_rate": 0.00019356693105505006,
      "loss": 4.4998,
      "step": 14350
    },
    {
      "epoch": 0.16,
      "grad_norm": 5.494804859161377,
      "learning_rate": 0.00019350207437216386,
      "loss": 4.3911,
      "step": 14400
    },
    {
      "epoch": 0.16,
      "grad_norm": 5.227470397949219,
      "learning_rate": 0.00019343690336861687,
      "loss": 4.2557,
      "step": 14450
    },
    {
      "epoch": 0.16,
      "grad_norm": 3.7686829566955566,
      "learning_rate": 0.00019337141826349092,
      "loss": 4.313,
      "step": 14500
    },
    {
      "epoch": 0.16,
      "grad_norm": 4.975152492523193,
      "learning_rate": 0.00019330561927692345,
      "loss": 4.2914,
      "step": 14550
    },
    {
      "epoch": 0.16,
      "grad_norm": 5.811885356903076,
      "learning_rate": 0.00019323950663010733,
      "loss": 4.3566,
      "step": 14600
    },
    {
      "epoch": 0.16,
      "grad_norm": 5.566829204559326,
      "learning_rate": 0.00019317308054528966,
      "loss": 4.2847,
      "step": 14650
    },
    {
      "epoch": 0.16,
      "grad_norm": 5.977478504180908,
      "learning_rate": 0.0001931063412457713,
      "loss": 4.3034,
      "step": 14700
    },
    {
      "epoch": 0.16,
      "grad_norm": 4.601086616516113,
      "learning_rate": 0.00019303928895590596,
      "loss": 4.1929,
      "step": 14750
    },
    {
      "epoch": 0.16,
      "grad_norm": 5.051478385925293,
      "learning_rate": 0.0001929719239010996,
      "loss": 4.2749,
      "step": 14800
    },
    {
      "epoch": 0.16,
      "grad_norm": 6.248847961425781,
      "learning_rate": 0.00019290424630780947,
      "loss": 4.3419,
      "step": 14850
    },
    {
      "epoch": 0.17,
      "grad_norm": 5.392062664031982,
      "learning_rate": 0.0001928362564035436,
      "loss": 4.4038,
      "step": 14900
    },
    {
      "epoch": 0.17,
      "grad_norm": 5.6346211433410645,
      "learning_rate": 0.00019276795441685975,
      "loss": 4.3403,
      "step": 14950
    },
    {
      "epoch": 0.17,
      "grad_norm": 5.646982192993164,
      "learning_rate": 0.00019269934057736493,
      "loss": 4.252,
      "step": 15000
    },
    {
      "epoch": 0.17,
      "grad_norm": 5.455059051513672,
      "learning_rate": 0.00019263041511571438,
      "loss": 4.3809,
      "step": 15050
    },
    {
      "epoch": 0.17,
      "grad_norm": 5.478726387023926,
      "learning_rate": 0.00019256117826361096,
      "loss": 4.1885,
      "step": 15100
    },
    {
      "epoch": 0.17,
      "grad_norm": 5.029292106628418,
      "learning_rate": 0.0001924916302538043,
      "loss": 4.2615,
      "step": 15150
    },
    {
      "epoch": 0.17,
      "grad_norm": 5.6447978019714355,
      "learning_rate": 0.00019242177132009,
      "loss": 4.268,
      "step": 15200
    },
    {
      "epoch": 0.17,
      "grad_norm": 5.165138244628906,
      "learning_rate": 0.00019235160169730895,
      "loss": 4.3222,
      "step": 15250
    },
    {
      "epoch": 0.17,
      "grad_norm": 5.661884784698486,
      "learning_rate": 0.00019228112162134641,
      "loss": 4.3179,
      "step": 15300
    },
    {
      "epoch": 0.17,
      "grad_norm": 6.117990493774414,
      "learning_rate": 0.0001922103313291313,
      "loss": 4.2241,
      "step": 15350
    },
    {
      "epoch": 0.17,
      "grad_norm": 4.299765110015869,
      "learning_rate": 0.0001921392310586353,
      "loss": 4.2602,
      "step": 15400
    },
    {
      "epoch": 0.17,
      "grad_norm": 5.798460483551025,
      "learning_rate": 0.00019206782104887223,
      "loss": 4.3096,
      "step": 15450
    },
    {
      "epoch": 0.17,
      "grad_norm": 5.016506671905518,
      "learning_rate": 0.00019199610153989712,
      "loss": 4.2073,
      "step": 15500
    },
    {
      "epoch": 0.17,
      "grad_norm": 9.708767890930176,
      "learning_rate": 0.0001919240727728054,
      "loss": 4.2099,
      "step": 15550
    },
    {
      "epoch": 0.17,
      "grad_norm": 4.904361248016357,
      "learning_rate": 0.00019185173498973204,
      "loss": 4.2461,
      "step": 15600
    },
    {
      "epoch": 0.17,
      "grad_norm": 5.290199279785156,
      "learning_rate": 0.00019177908843385103,
      "loss": 4.115,
      "step": 15650
    },
    {
      "epoch": 0.17,
      "grad_norm": 6.290179252624512,
      "learning_rate": 0.00019170613334937406,
      "loss": 4.3295,
      "step": 15700
    },
    {
      "epoch": 0.17,
      "grad_norm": 5.071104526519775,
      "learning_rate": 0.00019163286998155027,
      "loss": 4.1532,
      "step": 15750
    },
    {
      "epoch": 0.18,
      "grad_norm": 4.5464067459106445,
      "learning_rate": 0.00019155929857666494,
      "loss": 4.0761,
      "step": 15800
    },
    {
      "epoch": 0.18,
      "grad_norm": 4.664229393005371,
      "learning_rate": 0.0001914854193820389,
      "loss": 4.1371,
      "step": 15850
    },
    {
      "epoch": 0.18,
      "grad_norm": 7.168484210968018,
      "learning_rate": 0.0001914112326460277,
      "loss": 4.178,
      "step": 15900
    },
    {
      "epoch": 0.18,
      "grad_norm": 6.570041179656982,
      "learning_rate": 0.0001913367386180207,
      "loss": 4.1536,
      "step": 15950
    },
    {
      "epoch": 0.18,
      "grad_norm": 5.298222064971924,
      "learning_rate": 0.00019126193754844036,
      "loss": 4.2089,
      "step": 16000
    },
    {
      "epoch": 0.18,
      "grad_norm": 7.139255523681641,
      "learning_rate": 0.0001911868296887411,
      "loss": 4.1362,
      "step": 16050
    },
    {
      "epoch": 0.18,
      "grad_norm": 5.763050556182861,
      "learning_rate": 0.00019111141529140887,
      "loss": 4.1106,
      "step": 16100
    },
    {
      "epoch": 0.18,
      "grad_norm": 6.586143493652344,
      "learning_rate": 0.00019103569460995998,
      "loss": 3.9519,
      "step": 16150
    },
    {
      "epoch": 0.18,
      "grad_norm": 5.827348232269287,
      "learning_rate": 0.00019095966789894038,
      "loss": 3.9598,
      "step": 16200
    },
    {
      "epoch": 0.18,
      "grad_norm": 5.121611595153809,
      "learning_rate": 0.00019088333541392478,
      "loss": 4.1347,
      "step": 16250
    },
    {
      "epoch": 0.18,
      "grad_norm": 5.110377788543701,
      "learning_rate": 0.00019080669741151581,
      "loss": 4.0088,
      "step": 16300
    },
    {
      "epoch": 0.18,
      "grad_norm": 6.672893047332764,
      "learning_rate": 0.00019072975414934318,
      "loss": 4.0916,
      "step": 16350
    },
    {
      "epoch": 0.18,
      "grad_norm": 5.667397499084473,
      "learning_rate": 0.00019065250588606262,
      "loss": 4.0695,
      "step": 16400
    },
    {
      "epoch": 0.18,
      "grad_norm": 6.404243469238281,
      "learning_rate": 0.0001905749528813553,
      "loss": 3.9728,
      "step": 16450
    },
    {
      "epoch": 0.18,
      "grad_norm": 6.912601470947266,
      "learning_rate": 0.00019049709539592686,
      "loss": 4.029,
      "step": 16500
    },
    {
      "epoch": 0.18,
      "grad_norm": 5.015479564666748,
      "learning_rate": 0.00019041893369150636,
      "loss": 4.0268,
      "step": 16550
    },
    {
      "epoch": 0.18,
      "grad_norm": 6.656422138214111,
      "learning_rate": 0.00019034046803084563,
      "loss": 4.0393,
      "step": 16600
    },
    {
      "epoch": 0.18,
      "grad_norm": 4.685242176055908,
      "learning_rate": 0.00019026169867771825,
      "loss": 4.1104,
      "step": 16650
    },
    {
      "epoch": 0.19,
      "grad_norm": 6.503780364990234,
      "learning_rate": 0.00019018262589691874,
      "loss": 4.0344,
      "step": 16700
    },
    {
      "epoch": 0.19,
      "grad_norm": 4.73757266998291,
      "learning_rate": 0.00019010324995426156,
      "loss": 4.1114,
      "step": 16750
    },
    {
      "epoch": 0.19,
      "grad_norm": 7.276214122772217,
      "learning_rate": 0.0001900235711165804,
      "loss": 3.8838,
      "step": 16800
    },
    {
      "epoch": 0.19,
      "grad_norm": 6.2224273681640625,
      "learning_rate": 0.00018994358965172717,
      "loss": 3.9479,
      "step": 16850
    },
    {
      "epoch": 0.19,
      "grad_norm": 6.4751996994018555,
      "learning_rate": 0.00018986330582857096,
      "loss": 4.0079,
      "step": 16900
    },
    {
      "epoch": 0.19,
      "grad_norm": 4.874088764190674,
      "learning_rate": 0.00018978271991699743,
      "loss": 4.1664,
      "step": 16950
    },
    {
      "epoch": 0.19,
      "grad_norm": 7.713326454162598,
      "learning_rate": 0.0001897018321879077,
      "loss": 3.9646,
      "step": 17000
    },
    {
      "epoch": 0.19,
      "grad_norm": 5.753252029418945,
      "learning_rate": 0.00018962064291321747,
      "loss": 3.8574,
      "step": 17050
    },
    {
      "epoch": 0.19,
      "grad_norm": 5.962434768676758,
      "learning_rate": 0.0001895391523658562,
      "loss": 3.9757,
      "step": 17100
    },
    {
      "epoch": 0.19,
      "grad_norm": 5.875513553619385,
      "learning_rate": 0.00018945736081976607,
      "loss": 4.0424,
      "step": 17150
    },
    {
      "epoch": 0.19,
      "grad_norm": 6.298293590545654,
      "learning_rate": 0.00018937526854990108,
      "loss": 3.958,
      "step": 17200
    },
    {
      "epoch": 0.19,
      "grad_norm": 4.98872184753418,
      "learning_rate": 0.00018929287583222625,
      "loss": 3.9225,
      "step": 17250
    },
    {
      "epoch": 0.19,
      "grad_norm": 6.467836380004883,
      "learning_rate": 0.00018921018294371645,
      "loss": 3.9369,
      "step": 17300
    },
    {
      "epoch": 0.19,
      "grad_norm": 5.920988082885742,
      "learning_rate": 0.0001891271901623558,
      "loss": 3.975,
      "step": 17350
    },
    {
      "epoch": 0.19,
      "grad_norm": 5.652931213378906,
      "learning_rate": 0.00018904389776713641,
      "loss": 3.9067,
      "step": 17400
    },
    {
      "epoch": 0.19,
      "grad_norm": 5.372093200683594,
      "learning_rate": 0.00018896030603805767,
      "loss": 3.9267,
      "step": 17450
    },
    {
      "epoch": 0.19,
      "grad_norm": 5.743618965148926,
      "learning_rate": 0.00018887641525612518,
      "loss": 3.8912,
      "step": 17500
    },
    {
      "epoch": 0.19,
      "grad_norm": 8.207468032836914,
      "learning_rate": 0.00018879222570334985,
      "loss": 3.9101,
      "step": 17550
    },
    {
      "epoch": 0.2,
      "grad_norm": 6.930370807647705,
      "learning_rate": 0.00018870773766274697,
      "loss": 3.8817,
      "step": 17600
    },
    {
      "epoch": 0.2,
      "grad_norm": 6.367077350616455,
      "learning_rate": 0.00018862295141833523,
      "loss": 3.8931,
      "step": 17650
    },
    {
      "epoch": 0.2,
      "grad_norm": 6.587210178375244,
      "learning_rate": 0.00018853786725513575,
      "loss": 3.9393,
      "step": 17700
    },
    {
      "epoch": 0.2,
      "grad_norm": 5.502545356750488,
      "learning_rate": 0.0001884524854591712,
      "loss": 3.8489,
      "step": 17750
    },
    {
      "epoch": 0.2,
      "grad_norm": 6.352043628692627,
      "learning_rate": 0.00018836680631746476,
      "loss": 3.8162,
      "step": 17800
    },
    {
      "epoch": 0.2,
      "grad_norm": 5.686196804046631,
      "learning_rate": 0.00018828083011803917,
      "loss": 3.9476,
      "step": 17850
    },
    {
      "epoch": 0.2,
      "grad_norm": 6.225170612335205,
      "learning_rate": 0.00018819455714991578,
      "loss": 3.9404,
      "step": 17900
    },
    {
      "epoch": 0.2,
      "grad_norm": 7.1347150802612305,
      "learning_rate": 0.0001881079877031136,
      "loss": 3.9798,
      "step": 17950
    },
    {
      "epoch": 0.2,
      "grad_norm": 5.343573093414307,
      "learning_rate": 0.0001880211220686482,
      "loss": 3.9038,
      "step": 18000
    },
    {
      "epoch": 0.2,
      "grad_norm": 6.858921051025391,
      "learning_rate": 0.00018793396053853098,
      "loss": 3.8792,
      "step": 18050
    },
    {
      "epoch": 0.2,
      "grad_norm": 6.721033573150635,
      "learning_rate": 0.0001878482554434291,
      "loss": 3.8421,
      "step": 18100
    },
    {
      "epoch": 0.2,
      "grad_norm": 6.173632621765137,
      "learning_rate": 0.00018776050890530516,
      "loss": 4.0233,
      "step": 18150
    },
    {
      "epoch": 0.2,
      "grad_norm": 5.996013164520264,
      "learning_rate": 0.00018767246734761796,
      "loss": 3.8057,
      "step": 18200
    },
    {
      "epoch": 0.2,
      "grad_norm": 5.707641124725342,
      "learning_rate": 0.00018758413106633186,
      "loss": 3.8299,
      "step": 18250
    },
    {
      "epoch": 0.2,
      "grad_norm": 7.221241474151611,
      "learning_rate": 0.00018749550035840193,
      "loss": 3.8828,
      "step": 18300
    },
    {
      "epoch": 0.2,
      "grad_norm": 5.554357528686523,
      "learning_rate": 0.00018740657552177305,
      "loss": 3.8553,
      "step": 18350
    },
    {
      "epoch": 0.2,
      "grad_norm": 4.664674282073975,
      "learning_rate": 0.00018731735685537885,
      "loss": 3.8838,
      "step": 18400
    },
    {
      "epoch": 0.2,
      "grad_norm": 5.485450267791748,
      "learning_rate": 0.00018722784465914071,
      "loss": 3.8165,
      "step": 18450
    },
    {
      "epoch": 0.21,
      "grad_norm": 5.825826644897461,
      "learning_rate": 0.00018713803923396668,
      "loss": 3.7588,
      "step": 18500
    },
    {
      "epoch": 0.21,
      "grad_norm": 5.392491817474365,
      "learning_rate": 0.0001870479408817507,
      "loss": 3.8001,
      "step": 18550
    },
    {
      "epoch": 0.21,
      "grad_norm": 5.493740081787109,
      "learning_rate": 0.00018695754990537123,
      "loss": 3.9735,
      "step": 18600
    },
    {
      "epoch": 0.21,
      "grad_norm": 5.905117511749268,
      "learning_rate": 0.00018686686660869062,
      "loss": 3.7334,
      "step": 18650
    },
    {
      "epoch": 0.21,
      "grad_norm": 5.598316192626953,
      "learning_rate": 0.0001867758912965537,
      "loss": 3.8269,
      "step": 18700
    },
    {
      "epoch": 0.21,
      "grad_norm": 5.979629039764404,
      "learning_rate": 0.00018668462427478714,
      "loss": 3.8713,
      "step": 18750
    },
    {
      "epoch": 0.21,
      "grad_norm": 6.480854511260986,
      "learning_rate": 0.00018659306585019813,
      "loss": 3.7792,
      "step": 18800
    },
    {
      "epoch": 0.21,
      "grad_norm": 5.820549488067627,
      "learning_rate": 0.00018650121633057346,
      "loss": 3.6656,
      "step": 18850
    },
    {
      "epoch": 0.21,
      "grad_norm": 6.579679012298584,
      "learning_rate": 0.0001864090760246785,
      "loss": 3.9109,
      "step": 18900
    },
    {
      "epoch": 0.21,
      "grad_norm": 5.669819355010986,
      "learning_rate": 0.00018631664524225615,
      "loss": 3.7815,
      "step": 18950
    },
    {
      "epoch": 0.21,
      "grad_norm": 6.644351005554199,
      "learning_rate": 0.0001862239242940257,
      "loss": 3.7529,
      "step": 19000
    },
    {
      "epoch": 0.21,
      "grad_norm": 5.022332191467285,
      "learning_rate": 0.00018613091349168205,
      "loss": 3.7001,
      "step": 19050
    },
    {
      "epoch": 0.21,
      "grad_norm": 6.376641273498535,
      "learning_rate": 0.00018603761314789425,
      "loss": 3.6871,
      "step": 19100
    },
    {
      "epoch": 0.21,
      "grad_norm": 6.298123359680176,
      "learning_rate": 0.00018594402357630495,
      "loss": 3.8095,
      "step": 19150
    },
    {
      "epoch": 0.21,
      "grad_norm": 4.590997695922852,
      "learning_rate": 0.00018585014509152882,
      "loss": 3.8069,
      "step": 19200
    },
    {
      "epoch": 0.21,
      "grad_norm": 6.710943222045898,
      "learning_rate": 0.00018575597800915198,
      "loss": 3.8547,
      "step": 19250
    },
    {
      "epoch": 0.21,
      "grad_norm": 5.5094499588012695,
      "learning_rate": 0.0001856615226457305,
      "loss": 3.7314,
      "step": 19300
    },
    {
      "epoch": 0.21,
      "grad_norm": 5.584799766540527,
      "learning_rate": 0.0001855667793187898,
      "loss": 3.7514,
      "step": 19350
    },
    {
      "epoch": 0.22,
      "grad_norm": 6.0391154289245605,
      "learning_rate": 0.00018547174834682308,
      "loss": 3.6231,
      "step": 19400
    },
    {
      "epoch": 0.22,
      "grad_norm": 5.92927885055542,
      "learning_rate": 0.00018537643004929067,
      "loss": 3.7008,
      "step": 19450
    },
    {
      "epoch": 0.22,
      "grad_norm": 6.359600067138672,
      "learning_rate": 0.00018528082474661867,
      "loss": 3.798,
      "step": 19500
    },
    {
      "epoch": 0.22,
      "grad_norm": 5.198579788208008,
      "learning_rate": 0.0001851849327601981,
      "loss": 3.7187,
      "step": 19550
    },
    {
      "epoch": 0.22,
      "grad_norm": 6.796758651733398,
      "learning_rate": 0.00018508875441238364,
      "loss": 3.7086,
      "step": 19600
    },
    {
      "epoch": 0.22,
      "grad_norm": 4.889728546142578,
      "learning_rate": 0.00018499229002649258,
      "loss": 3.7387,
      "step": 19650
    },
    {
      "epoch": 0.22,
      "grad_norm": 6.382203102111816,
      "learning_rate": 0.0001848955399268039,
      "loss": 3.5992,
      "step": 19700
    },
    {
      "epoch": 0.22,
      "grad_norm": 6.061376571655273,
      "learning_rate": 0.00018479850443855686,
      "loss": 3.6865,
      "step": 19750
    },
    {
      "epoch": 0.22,
      "grad_norm": 5.2180681228637695,
      "learning_rate": 0.0001847011838879503,
      "loss": 3.7467,
      "step": 19800
    },
    {
      "epoch": 0.22,
      "grad_norm": 5.063679218292236,
      "learning_rate": 0.0001846035786021412,
      "loss": 3.6894,
      "step": 19850
    },
    {
      "epoch": 0.22,
      "grad_norm": 5.036098480224609,
      "learning_rate": 0.00018450568890924373,
      "loss": 3.6412,
      "step": 19900
    },
    {
      "epoch": 0.22,
      "grad_norm": 4.86781644821167,
      "learning_rate": 0.00018440751513832822,
      "loss": 3.637,
      "step": 19950
    },
    {
      "epoch": 0.22,
      "grad_norm": 5.41668176651001,
      "learning_rate": 0.00018430905761941983,
      "loss": 3.6814,
      "step": 20000
    },
    {
      "epoch": 0.22,
      "grad_norm": 6.117024898529053,
      "learning_rate": 0.00018421031668349773,
      "loss": 3.6257,
      "step": 20050
    },
    {
      "epoch": 0.22,
      "grad_norm": 7.368699073791504,
      "learning_rate": 0.00018411129266249373,
      "loss": 3.7111,
      "step": 20100
    },
    {
      "epoch": 0.22,
      "grad_norm": 4.378394603729248,
      "learning_rate": 0.0001840119858892913,
      "loss": 3.7197,
      "step": 20150
    },
    {
      "epoch": 0.22,
      "grad_norm": 7.029990196228027,
      "learning_rate": 0.0001839123966977245,
      "loss": 3.7267,
      "step": 20200
    },
    {
      "epoch": 0.22,
      "grad_norm": 9.922813415527344,
      "learning_rate": 0.00018381252542257662,
      "loss": 3.7203,
      "step": 20250
    },
    {
      "epoch": 0.23,
      "grad_norm": 4.9374518394470215,
      "learning_rate": 0.00018371237239957932,
      "loss": 3.6876,
      "step": 20300
    },
    {
      "epoch": 0.23,
      "grad_norm": 5.682550430297852,
      "learning_rate": 0.00018361193796541142,
      "loss": 3.6862,
      "step": 20350
    },
    {
      "epoch": 0.23,
      "grad_norm": 5.477772235870361,
      "learning_rate": 0.00018351122245769771,
      "loss": 3.5982,
      "step": 20400
    },
    {
      "epoch": 0.23,
      "grad_norm": 7.745680332183838,
      "learning_rate": 0.00018341224888886997,
      "loss": 3.6978,
      "step": 20450
    },
    {
      "epoch": 0.23,
      "grad_norm": 6.691402912139893,
      "learning_rate": 0.0001833109778552932,
      "loss": 3.6693,
      "step": 20500
    },
    {
      "epoch": 0.23,
      "grad_norm": 6.229629993438721,
      "learning_rate": 0.00018320942675989125,
      "loss": 3.6327,
      "step": 20550
    },
    {
      "epoch": 0.23,
      "grad_norm": 5.655289649963379,
      "learning_rate": 0.0001831075959440427,
      "loss": 3.6032,
      "step": 20600
    },
    {
      "epoch": 0.23,
      "grad_norm": 5.4868927001953125,
      "learning_rate": 0.00018300548575006658,
      "loss": 3.7059,
      "step": 20650
    },
    {
      "epoch": 0.23,
      "grad_norm": 7.387706756591797,
      "learning_rate": 0.00018290309652122083,
      "loss": 3.6838,
      "step": 20700
    },
    {
      "epoch": 0.23,
      "grad_norm": 4.884798049926758,
      "learning_rate": 0.00018280042860170168,
      "loss": 3.665,
      "step": 20750
    },
    {
      "epoch": 0.23,
      "grad_norm": 7.185595512390137,
      "learning_rate": 0.00018269748233664204,
      "loss": 3.6057,
      "step": 20800
    },
    {
      "epoch": 0.23,
      "grad_norm": 6.449123382568359,
      "learning_rate": 0.0001825942580721106,
      "loss": 3.6262,
      "step": 20850
    },
    {
      "epoch": 0.23,
      "grad_norm": 5.469310283660889,
      "learning_rate": 0.00018249075615511053,
      "loss": 3.522,
      "step": 20900
    },
    {
      "epoch": 0.23,
      "grad_norm": 5.678877353668213,
      "learning_rate": 0.0001823869769335784,
      "loss": 3.6757,
      "step": 20950
    },
    {
      "epoch": 0.23,
      "grad_norm": 5.033955097198486,
      "learning_rate": 0.000182282920756383,
      "loss": 3.7316,
      "step": 21000
    },
    {
      "epoch": 0.23,
      "grad_norm": 6.790628433227539,
      "learning_rate": 0.00018217858797332413,
      "loss": 3.545,
      "step": 21050
    },
    {
      "epoch": 0.23,
      "grad_norm": 6.205599308013916,
      "learning_rate": 0.00018207397893513143,
      "loss": 3.6035,
      "step": 21100
    },
    {
      "epoch": 0.23,
      "grad_norm": 5.7604804039001465,
      "learning_rate": 0.00018196909399346316,
      "loss": 3.6869,
      "step": 21150
    },
    {
      "epoch": 0.24,
      "grad_norm": 5.528883457183838,
      "learning_rate": 0.0001818639335009052,
      "loss": 3.6493,
      "step": 21200
    },
    {
      "epoch": 0.24,
      "grad_norm": 6.46929407119751,
      "learning_rate": 0.00018175849781096966,
      "loss": 3.639,
      "step": 21250
    },
    {
      "epoch": 0.24,
      "grad_norm": 5.487035274505615,
      "learning_rate": 0.00018165278727809368,
      "loss": 3.5755,
      "step": 21300
    },
    {
      "epoch": 0.24,
      "grad_norm": 6.251669406890869,
      "learning_rate": 0.00018154680225763848,
      "loss": 3.704,
      "step": 21350
    },
    {
      "epoch": 0.24,
      "grad_norm": 6.204404830932617,
      "learning_rate": 0.00018144054310588792,
      "loss": 3.6071,
      "step": 21400
    },
    {
      "epoch": 0.24,
      "grad_norm": 5.7311482429504395,
      "learning_rate": 0.00018133401018004743,
      "loss": 3.5395,
      "step": 21450
    },
    {
      "epoch": 0.24,
      "grad_norm": 7.110382556915283,
      "learning_rate": 0.00018122720383824273,
      "loss": 3.6643,
      "step": 21500
    },
    {
      "epoch": 0.24,
      "grad_norm": 5.991401672363281,
      "learning_rate": 0.0001811201244395187,
      "loss": 3.6752,
      "step": 21550
    },
    {
      "epoch": 0.24,
      "grad_norm": 5.788415431976318,
      "learning_rate": 0.0001810127723438381,
      "loss": 3.6362,
      "step": 21600
    },
    {
      "epoch": 0.24,
      "grad_norm": 4.827778339385986,
      "learning_rate": 0.00018090514791208043,
      "loss": 3.7298,
      "step": 21650
    },
    {
      "epoch": 0.24,
      "grad_norm": 5.7845916748046875,
      "learning_rate": 0.0001807972515060407,
      "loss": 3.543,
      "step": 21700
    },
    {
      "epoch": 0.24,
      "grad_norm": 5.071081638336182,
      "learning_rate": 0.00018068908348842818,
      "loss": 3.5706,
      "step": 21750
    },
    {
      "epoch": 0.24,
      "grad_norm": 5.189342021942139,
      "learning_rate": 0.00018058064422286525,
      "loss": 3.667,
      "step": 21800
    },
    {
      "epoch": 0.24,
      "grad_norm": 7.787344455718994,
      "learning_rate": 0.00018047193407388603,
      "loss": 3.4985,
      "step": 21850
    },
    {
      "epoch": 0.24,
      "grad_norm": 4.411252975463867,
      "learning_rate": 0.00018036295340693531,
      "loss": 3.6719,
      "step": 21900
    },
    {
      "epoch": 0.24,
      "grad_norm": 5.700460433959961,
      "learning_rate": 0.00018025370258836732,
      "loss": 3.5075,
      "step": 21950
    },
    {
      "epoch": 0.24,
      "grad_norm": 5.121459007263184,
      "learning_rate": 0.00018014418198544432,
      "loss": 3.5511,
      "step": 22000
    },
    {
      "epoch": 0.24,
      "grad_norm": 5.29133415222168,
      "learning_rate": 0.0001800343919663356,
      "loss": 3.7063,
      "step": 22050
    },
    {
      "epoch": 0.25,
      "grad_norm": 6.53157377243042,
      "learning_rate": 0.00017992433290011604,
      "loss": 3.5146,
      "step": 22100
    },
    {
      "epoch": 0.25,
      "grad_norm": 6.442373275756836,
      "learning_rate": 0.00017981400515676508,
      "loss": 3.5431,
      "step": 22150
    },
    {
      "epoch": 0.25,
      "grad_norm": 5.247061729431152,
      "learning_rate": 0.00017970340910716522,
      "loss": 3.604,
      "step": 22200
    },
    {
      "epoch": 0.25,
      "grad_norm": 6.570899486541748,
      "learning_rate": 0.000179592545123101,
      "loss": 3.6034,
      "step": 22250
    },
    {
      "epoch": 0.25,
      "grad_norm": 6.012238502502441,
      "learning_rate": 0.00017948141357725764,
      "loss": 3.4793,
      "step": 22300
    },
    {
      "epoch": 0.25,
      "grad_norm": 5.5325422286987305,
      "learning_rate": 0.0001793700148432198,
      "loss": 3.563,
      "step": 22350
    },
    {
      "epoch": 0.25,
      "grad_norm": 4.889975070953369,
      "learning_rate": 0.00017925834929547035,
      "loss": 3.5512,
      "step": 22400
    },
    {
      "epoch": 0.25,
      "grad_norm": 5.225555896759033,
      "learning_rate": 0.00017914641730938907,
      "loss": 3.5521,
      "step": 22450
    },
    {
      "epoch": 0.25,
      "grad_norm": 5.430109024047852,
      "learning_rate": 0.0001790342192612514,
      "loss": 3.4549,
      "step": 22500
    },
    {
      "epoch": 0.25,
      "grad_norm": 5.8808274269104,
      "learning_rate": 0.00017892175552822716,
      "loss": 3.5518,
      "step": 22550
    },
    {
      "epoch": 0.25,
      "grad_norm": 6.657894134521484,
      "learning_rate": 0.00017880902648837946,
      "loss": 3.4643,
      "step": 22600
    },
    {
      "epoch": 0.25,
      "grad_norm": 4.968985557556152,
      "learning_rate": 0.00017869603252066308,
      "loss": 3.5022,
      "step": 22650
    },
    {
      "epoch": 0.25,
      "grad_norm": 5.369678497314453,
      "learning_rate": 0.00017858277400492357,
      "loss": 3.6906,
      "step": 22700
    },
    {
      "epoch": 0.25,
      "grad_norm": 6.433826446533203,
      "learning_rate": 0.0001784692513218956,
      "loss": 3.4281,
      "step": 22750
    },
    {
      "epoch": 0.25,
      "grad_norm": 5.364591121673584,
      "learning_rate": 0.00017835546485320202,
      "loss": 3.6194,
      "step": 22800
    },
    {
      "epoch": 0.25,
      "grad_norm": 5.889247894287109,
      "learning_rate": 0.00017824141498135244,
      "loss": 3.7013,
      "step": 22850
    },
    {
      "epoch": 0.25,
      "grad_norm": 5.254469394683838,
      "learning_rate": 0.0001781271020897419,
      "loss": 3.4107,
      "step": 22900
    },
    {
      "epoch": 0.25,
      "grad_norm": 5.486823558807373,
      "learning_rate": 0.0001780125265626495,
      "loss": 3.5453,
      "step": 22950
    },
    {
      "epoch": 0.26,
      "grad_norm": 5.4713311195373535,
      "learning_rate": 0.0001778976887852375,
      "loss": 3.5482,
      "step": 23000
    },
    {
      "epoch": 0.26,
      "grad_norm": 6.2519731521606445,
      "learning_rate": 0.00017778258914354946,
      "loss": 3.6251,
      "step": 23050
    },
    {
      "epoch": 0.26,
      "grad_norm": 5.657818794250488,
      "learning_rate": 0.00017766722802450944,
      "loss": 3.5081,
      "step": 23100
    },
    {
      "epoch": 0.26,
      "grad_norm": 6.176442623138428,
      "learning_rate": 0.0001775516058159204,
      "loss": 3.45,
      "step": 23150
    },
    {
      "epoch": 0.26,
      "grad_norm": 5.838647365570068,
      "learning_rate": 0.00017743572290646303,
      "loss": 3.419,
      "step": 23200
    },
    {
      "epoch": 0.26,
      "grad_norm": 6.912227630615234,
      "learning_rate": 0.00017731957968569436,
      "loss": 3.4892,
      "step": 23250
    },
    {
      "epoch": 0.26,
      "grad_norm": 7.276485919952393,
      "learning_rate": 0.0001772031765440465,
      "loss": 3.5143,
      "step": 23300
    },
    {
      "epoch": 0.26,
      "grad_norm": 5.351586818695068,
      "learning_rate": 0.0001770865138728254,
      "loss": 3.5467,
      "step": 23350
    },
    {
      "epoch": 0.26,
      "grad_norm": 5.220416069030762,
      "learning_rate": 0.00017696959206420937,
      "loss": 3.4736,
      "step": 23400
    },
    {
      "epoch": 0.26,
      "grad_norm": 5.282609462738037,
      "learning_rate": 0.00017685241151124781,
      "loss": 3.4181,
      "step": 23450
    },
    {
      "epoch": 0.26,
      "grad_norm": 4.958062171936035,
      "learning_rate": 0.00017673497260786006,
      "loss": 3.4309,
      "step": 23500
    },
    {
      "epoch": 0.26,
      "grad_norm": 6.3785929679870605,
      "learning_rate": 0.00017661727574883388,
      "loss": 3.3805,
      "step": 23550
    },
    {
      "epoch": 0.26,
      "grad_norm": 5.702798366546631,
      "learning_rate": 0.00017649932132982415,
      "loss": 3.5371,
      "step": 23600
    },
    {
      "epoch": 0.26,
      "grad_norm": 6.496365547180176,
      "learning_rate": 0.0001763811097473516,
      "loss": 3.4107,
      "step": 23650
    },
    {
      "epoch": 0.26,
      "grad_norm": 5.093421936035156,
      "learning_rate": 0.00017626264139880148,
      "loss": 3.5514,
      "step": 23700
    },
    {
      "epoch": 0.26,
      "grad_norm": 5.6509480476379395,
      "learning_rate": 0.0001761439166824221,
      "loss": 3.5612,
      "step": 23750
    },
    {
      "epoch": 0.26,
      "grad_norm": 5.662957191467285,
      "learning_rate": 0.00017602493599732372,
      "loss": 3.5515,
      "step": 23800
    },
    {
      "epoch": 0.26,
      "grad_norm": 7.548245429992676,
      "learning_rate": 0.000175905699743477,
      "loss": 3.5552,
      "step": 23850
    },
    {
      "epoch": 0.27,
      "grad_norm": 5.2797112464904785,
      "learning_rate": 0.00017578620832171173,
      "loss": 3.4159,
      "step": 23900
    },
    {
      "epoch": 0.27,
      "grad_norm": 6.431013584136963,
      "learning_rate": 0.0001756664621337155,
      "loss": 3.4257,
      "step": 23950
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.88436222076416,
      "learning_rate": 0.00017554646158203236,
      "loss": 3.5517,
      "step": 24000
    },
    {
      "epoch": 0.27,
      "grad_norm": 5.538012981414795,
      "learning_rate": 0.00017542620707006136,
      "loss": 3.4451,
      "step": 24050
    },
    {
      "epoch": 0.27,
      "grad_norm": 6.702478408813477,
      "learning_rate": 0.00017530569900205538,
      "loss": 3.5453,
      "step": 24100
    },
    {
      "epoch": 0.27,
      "grad_norm": 5.236027240753174,
      "learning_rate": 0.00017518493778311957,
      "loss": 3.4483,
      "step": 24150
    },
    {
      "epoch": 0.27,
      "grad_norm": 5.824537754058838,
      "learning_rate": 0.00017506392381921014,
      "loss": 3.507,
      "step": 24200
    },
    {
      "epoch": 0.27,
      "grad_norm": 6.682642459869385,
      "learning_rate": 0.0001749426575171329,
      "loss": 3.4624,
      "step": 24250
    },
    {
      "epoch": 0.27,
      "grad_norm": 6.02097225189209,
      "learning_rate": 0.00017482113928454196,
      "loss": 3.4782,
      "step": 24300
    },
    {
      "epoch": 0.27,
      "grad_norm": 5.951188087463379,
      "learning_rate": 0.00017469936952993834,
      "loss": 3.5305,
      "step": 24350
    },
    {
      "epoch": 0.27,
      "grad_norm": 5.47694730758667,
      "learning_rate": 0.00017457734866266854,
      "loss": 3.4653,
      "step": 24400
    },
    {
      "epoch": 0.27,
      "grad_norm": 5.073057651519775,
      "learning_rate": 0.0001744575249785453,
      "loss": 3.4969,
      "step": 24450
    },
    {
      "epoch": 0.27,
      "grad_norm": 6.532285690307617,
      "learning_rate": 0.00017433500811915326,
      "loss": 3.3932,
      "step": 24500
    },
    {
      "epoch": 0.27,
      "grad_norm": 6.016458988189697,
      "learning_rate": 0.00017421224137194837,
      "loss": 3.4828,
      "step": 24550
    },
    {
      "epoch": 0.27,
      "grad_norm": 7.032898902893066,
      "learning_rate": 0.0001740892251496286,
      "loss": 3.4347,
      "step": 24600
    },
    {
      "epoch": 0.27,
      "grad_norm": 6.3446431159973145,
      "learning_rate": 0.00017396595986573065,
      "loss": 3.4101,
      "step": 24650
    },
    {
      "epoch": 0.27,
      "grad_norm": 5.952356815338135,
      "learning_rate": 0.00017384244593462859,
      "loss": 3.4296,
      "step": 24700
    },
    {
      "epoch": 0.27,
      "grad_norm": 5.403810501098633,
      "learning_rate": 0.00017371868377153216,
      "loss": 3.4264,
      "step": 24750
    },
    {
      "epoch": 0.27,
      "grad_norm": 6.429996490478516,
      "learning_rate": 0.00017359467379248568,
      "loss": 3.4341,
      "step": 24800
    },
    {
      "epoch": 0.28,
      "grad_norm": 5.618744850158691,
      "learning_rate": 0.00017347041641436653,
      "loss": 3.3357,
      "step": 24850
    },
    {
      "epoch": 0.28,
      "grad_norm": 6.539459705352783,
      "learning_rate": 0.00017334840455978504,
      "loss": 3.5718,
      "step": 24900
    },
    {
      "epoch": 0.28,
      "grad_norm": 4.993662357330322,
      "learning_rate": 0.00017322365856462736,
      "loss": 3.4774,
      "step": 24950
    },
    {
      "epoch": 0.28,
      "grad_norm": 5.6996002197265625,
      "learning_rate": 0.00017309866641761798,
      "loss": 3.311,
      "step": 25000
    },
    {
      "epoch": 0.28,
      "grad_norm": 4.32814884185791,
      "learning_rate": 0.00017297342853893604,
      "loss": 3.5558,
      "step": 25050
    },
    {
      "epoch": 0.28,
      "grad_norm": 5.538712501525879,
      "learning_rate": 0.0001728479453495866,
      "loss": 3.3261,
      "step": 25100
    },
    {
      "epoch": 0.28,
      "grad_norm": 4.975490093231201,
      "learning_rate": 0.00017272221727139946,
      "loss": 3.5,
      "step": 25150
    },
    {
      "epoch": 0.28,
      "grad_norm": 4.377697467803955,
      "learning_rate": 0.00017259624472702764,
      "loss": 3.4562,
      "step": 25200
    },
    {
      "epoch": 0.28,
      "grad_norm": 6.686251163482666,
      "learning_rate": 0.00017247002813994592,
      "loss": 3.3968,
      "step": 25250
    },
    {
      "epoch": 0.28,
      "grad_norm": 7.116434097290039,
      "learning_rate": 0.00017234356793444954,
      "loss": 3.3161,
      "step": 25300
    },
    {
      "epoch": 0.28,
      "grad_norm": 5.755252838134766,
      "learning_rate": 0.0001722168645356526,
      "loss": 3.4195,
      "step": 25350
    },
    {
      "epoch": 0.28,
      "grad_norm": 6.647252559661865,
      "learning_rate": 0.00017208991836948685,
      "loss": 3.1887,
      "step": 25400
    },
    {
      "epoch": 0.28,
      "grad_norm": 4.997719764709473,
      "learning_rate": 0.0001719627298627,
      "loss": 3.4098,
      "step": 25450
    },
    {
      "epoch": 0.28,
      "grad_norm": 6.054971218109131,
      "learning_rate": 0.00017183529944285456,
      "loss": 3.4159,
      "step": 25500
    },
    {
      "epoch": 0.28,
      "grad_norm": 4.706241130828857,
      "learning_rate": 0.00017170762753832615,
      "loss": 3.4024,
      "step": 25550
    },
    {
      "epoch": 0.28,
      "grad_norm": 7.701054096221924,
      "learning_rate": 0.00017157971457830226,
      "loss": 3.3564,
      "step": 25600
    },
    {
      "epoch": 0.28,
      "grad_norm": 5.441225528717041,
      "learning_rate": 0.00017145156099278067,
      "loss": 3.5887,
      "step": 25650
    },
    {
      "epoch": 0.28,
      "grad_norm": 5.631026268005371,
      "learning_rate": 0.0001713231672125681,
      "loss": 3.352,
      "step": 25700
    },
    {
      "epoch": 0.29,
      "grad_norm": 4.974308967590332,
      "learning_rate": 0.0001711945336692786,
      "loss": 3.3959,
      "step": 25750
    },
    {
      "epoch": 0.29,
      "grad_norm": 7.063317775726318,
      "learning_rate": 0.00017106566079533246,
      "loss": 3.3942,
      "step": 25800
    },
    {
      "epoch": 0.29,
      "grad_norm": 6.312389850616455,
      "learning_rate": 0.0001709365490239543,
      "loss": 3.3928,
      "step": 25850
    },
    {
      "epoch": 0.29,
      "grad_norm": 5.022332668304443,
      "learning_rate": 0.00017080719878917182,
      "loss": 3.4401,
      "step": 25900
    },
    {
      "epoch": 0.29,
      "grad_norm": 4.356366157531738,
      "learning_rate": 0.00017067761052581455,
      "loss": 3.4353,
      "step": 25950
    },
    {
      "epoch": 0.29,
      "grad_norm": 5.611413478851318,
      "learning_rate": 0.00017054778466951196,
      "loss": 3.2737,
      "step": 26000
    },
    {
      "epoch": 0.29,
      "grad_norm": 7.244396686553955,
      "learning_rate": 0.0001704177216566924,
      "loss": 3.2309,
      "step": 26050
    },
    {
      "epoch": 0.29,
      "grad_norm": 6.024662017822266,
      "learning_rate": 0.00017028742192458132,
      "loss": 3.3593,
      "step": 26100
    },
    {
      "epoch": 0.29,
      "grad_norm": 7.399158954620361,
      "learning_rate": 0.00017015688591120006,
      "loss": 3.2026,
      "step": 26150
    },
    {
      "epoch": 0.29,
      "grad_norm": 7.96980619430542,
      "learning_rate": 0.00017002611405536413,
      "loss": 3.4413,
      "step": 26200
    },
    {
      "epoch": 0.29,
      "grad_norm": 5.538659572601318,
      "learning_rate": 0.00016989510679668194,
      "loss": 3.3497,
      "step": 26250
    },
    {
      "epoch": 0.29,
      "grad_norm": 5.8960394859313965,
      "learning_rate": 0.00016976386457555323,
      "loss": 3.3708,
      "step": 26300
    },
    {
      "epoch": 0.29,
      "grad_norm": 5.486491680145264,
      "learning_rate": 0.00016963238783316754,
      "loss": 3.4697,
      "step": 26350
    },
    {
      "epoch": 0.29,
      "grad_norm": 5.217641353607178,
      "learning_rate": 0.0001695006770115029,
      "loss": 3.4249,
      "step": 26400
    },
    {
      "epoch": 0.29,
      "grad_norm": 5.6906938552856445,
      "learning_rate": 0.00016936873255332413,
      "loss": 3.5343,
      "step": 26450
    },
    {
      "epoch": 0.29,
      "grad_norm": 6.275619983673096,
      "learning_rate": 0.00016923655490218149,
      "loss": 3.3991,
      "step": 26500
    },
    {
      "epoch": 0.29,
      "grad_norm": 5.75913667678833,
      "learning_rate": 0.00016910414450240917,
      "loss": 3.3861,
      "step": 26550
    },
    {
      "epoch": 0.29,
      "grad_norm": 5.206583499908447,
      "learning_rate": 0.0001689715017991237,
      "loss": 3.369,
      "step": 26600
    },
    {
      "epoch": 0.3,
      "grad_norm": 5.569302082061768,
      "learning_rate": 0.0001688386272382227,
      "loss": 3.4837,
      "step": 26650
    },
    {
      "epoch": 0.3,
      "grad_norm": 5.360637187957764,
      "learning_rate": 0.00016870552126638298,
      "loss": 3.3299,
      "step": 26700
    },
    {
      "epoch": 0.3,
      "grad_norm": 5.168808937072754,
      "learning_rate": 0.00016857218433105945,
      "loss": 3.3613,
      "step": 26750
    },
    {
      "epoch": 0.3,
      "grad_norm": 5.515918731689453,
      "learning_rate": 0.0001684386168804834,
      "loss": 3.159,
      "step": 26800
    },
    {
      "epoch": 0.3,
      "grad_norm": 5.885009765625,
      "learning_rate": 0.000168304819363661,
      "loss": 3.3029,
      "step": 26850
    },
    {
      "epoch": 0.3,
      "grad_norm": 5.174962997436523,
      "learning_rate": 0.0001681707922303718,
      "loss": 3.3289,
      "step": 26900
    },
    {
      "epoch": 0.3,
      "grad_norm": 4.86044454574585,
      "learning_rate": 0.0001680365359311673,
      "loss": 3.3132,
      "step": 26950
    },
    {
      "epoch": 0.3,
      "grad_norm": 6.267008304595947,
      "learning_rate": 0.00016790205091736935,
      "loss": 3.3649,
      "step": 27000
    },
    {
      "epoch": 0.3,
      "grad_norm": 6.221423625946045,
      "learning_rate": 0.00016776733764106862,
      "loss": 3.3311,
      "step": 27050
    },
    {
      "epoch": 0.3,
      "grad_norm": 5.069894790649414,
      "learning_rate": 0.00016763239655512318,
      "loss": 3.3157,
      "step": 27100
    },
    {
      "epoch": 0.3,
      "grad_norm": 8.425812721252441,
      "learning_rate": 0.00016749722811315688,
      "loss": 3.2714,
      "step": 27150
    },
    {
      "epoch": 0.3,
      "grad_norm": 5.80504846572876,
      "learning_rate": 0.00016736183276955783,
      "loss": 3.3274,
      "step": 27200
    },
    {
      "epoch": 0.3,
      "grad_norm": 7.617208003997803,
      "learning_rate": 0.00016722621097947697,
      "loss": 3.1857,
      "step": 27250
    },
    {
      "epoch": 0.3,
      "grad_norm": 5.363246440887451,
      "learning_rate": 0.00016709036319882646,
      "loss": 3.4673,
      "step": 27300
    },
    {
      "epoch": 0.3,
      "grad_norm": 6.790156364440918,
      "learning_rate": 0.00016695428988427807,
      "loss": 3.3016,
      "step": 27350
    },
    {
      "epoch": 0.3,
      "grad_norm": 4.9824981689453125,
      "learning_rate": 0.00016681799149326185,
      "loss": 3.4103,
      "step": 27400
    },
    {
      "epoch": 0.3,
      "grad_norm": 4.915642738342285,
      "learning_rate": 0.00016668146848396442,
      "loss": 3.4356,
      "step": 27450
    },
    {
      "epoch": 0.3,
      "grad_norm": 4.938210487365723,
      "learning_rate": 0.0001665447213153275,
      "loss": 3.299,
      "step": 27500
    },
    {
      "epoch": 0.31,
      "grad_norm": 5.116371154785156,
      "learning_rate": 0.00016640775044704634,
      "loss": 3.3231,
      "step": 27550
    },
    {
      "epoch": 0.31,
      "grad_norm": 6.796716213226318,
      "learning_rate": 0.0001662705563395682,
      "loss": 3.3685,
      "step": 27600
    },
    {
      "epoch": 0.31,
      "grad_norm": 9.196764945983887,
      "learning_rate": 0.0001661331394540908,
      "loss": 3.2807,
      "step": 27650
    },
    {
      "epoch": 0.31,
      "grad_norm": 4.64096736907959,
      "learning_rate": 0.00016599550025256076,
      "loss": 3.2909,
      "step": 27700
    },
    {
      "epoch": 0.31,
      "grad_norm": 5.615271091461182,
      "learning_rate": 0.000165857639197672,
      "loss": 3.2044,
      "step": 27750
    },
    {
      "epoch": 0.31,
      "grad_norm": 6.112679481506348,
      "learning_rate": 0.0001657195567528643,
      "loss": 3.2377,
      "step": 27800
    },
    {
      "epoch": 0.31,
      "grad_norm": 4.882411003112793,
      "learning_rate": 0.0001655812533823216,
      "loss": 3.4462,
      "step": 27850
    },
    {
      "epoch": 0.31,
      "grad_norm": 4.716900825500488,
      "learning_rate": 0.00016544272955097063,
      "loss": 3.3563,
      "step": 27900
    },
    {
      "epoch": 0.31,
      "grad_norm": 5.415688991546631,
      "learning_rate": 0.0001653039857244791,
      "loss": 3.2475,
      "step": 27950
    },
    {
      "epoch": 0.31,
      "grad_norm": 5.752101421356201,
      "learning_rate": 0.00016516502236925434,
      "loss": 3.3646,
      "step": 28000
    },
    {
      "epoch": 0.31,
      "grad_norm": 7.952321529388428,
      "learning_rate": 0.00016502583995244163,
      "loss": 3.2835,
      "step": 28050
    },
    {
      "epoch": 0.31,
      "grad_norm": 5.233190536499023,
      "learning_rate": 0.00016488643894192268,
      "loss": 3.3653,
      "step": 28100
    },
    {
      "epoch": 0.31,
      "grad_norm": 5.358859062194824,
      "learning_rate": 0.00016474681980631402,
      "loss": 3.2425,
      "step": 28150
    },
    {
      "epoch": 0.31,
      "grad_norm": 5.584342002868652,
      "learning_rate": 0.0001646069830149654,
      "loss": 3.3139,
      "step": 28200
    },
    {
      "epoch": 0.31,
      "grad_norm": 6.155908584594727,
      "learning_rate": 0.00016446692903795837,
      "loss": 3.2732,
      "step": 28250
    },
    {
      "epoch": 0.31,
      "grad_norm": 6.454020023345947,
      "learning_rate": 0.00016432665834610445,
      "loss": 3.2121,
      "step": 28300
    },
    {
      "epoch": 0.31,
      "grad_norm": 5.337937831878662,
      "learning_rate": 0.00016418617141094374,
      "loss": 3.3123,
      "step": 28350
    },
    {
      "epoch": 0.31,
      "grad_norm": 4.727194786071777,
      "learning_rate": 0.00016404546870474324,
      "loss": 3.2558,
      "step": 28400
    },
    {
      "epoch": 0.32,
      "grad_norm": 5.887204647064209,
      "learning_rate": 0.00016390455070049536,
      "loss": 3.377,
      "step": 28450
    },
    {
      "epoch": 0.32,
      "grad_norm": 5.2247724533081055,
      "learning_rate": 0.0001637634178719162,
      "loss": 3.1277,
      "step": 28500
    },
    {
      "epoch": 0.32,
      "grad_norm": 5.014094829559326,
      "learning_rate": 0.00016362207069344403,
      "loss": 3.142,
      "step": 28550
    },
    {
      "epoch": 0.32,
      "grad_norm": 7.58336067199707,
      "learning_rate": 0.00016348050964023773,
      "loss": 3.3156,
      "step": 28600
    },
    {
      "epoch": 0.32,
      "grad_norm": 4.806646823883057,
      "learning_rate": 0.00016333873518817514,
      "loss": 3.2814,
      "step": 28650
    },
    {
      "epoch": 0.32,
      "grad_norm": 5.387387275695801,
      "learning_rate": 0.00016319674781385143,
      "loss": 3.3977,
      "step": 28700
    },
    {
      "epoch": 0.32,
      "grad_norm": 5.4912638664245605,
      "learning_rate": 0.00016305454799457755,
      "loss": 3.2323,
      "step": 28750
    },
    {
      "epoch": 0.32,
      "grad_norm": 4.692640781402588,
      "learning_rate": 0.00016291213620837867,
      "loss": 3.3033,
      "step": 28800
    },
    {
      "epoch": 0.32,
      "grad_norm": 5.484092712402344,
      "learning_rate": 0.0001627695129339924,
      "loss": 3.1466,
      "step": 28850
    },
    {
      "epoch": 0.32,
      "grad_norm": 5.305532932281494,
      "learning_rate": 0.00016262667865086746,
      "loss": 3.4111,
      "step": 28900
    },
    {
      "epoch": 0.32,
      "grad_norm": 5.511746883392334,
      "learning_rate": 0.00016248363383916182,
      "loss": 3.2535,
      "step": 28950
    },
    {
      "epoch": 0.32,
      "grad_norm": 7.3723649978637695,
      "learning_rate": 0.00016234037897974108,
      "loss": 3.3265,
      "step": 29000
    },
    {
      "epoch": 0.32,
      "grad_norm": 5.085361003875732,
      "learning_rate": 0.0001621997858933184,
      "loss": 3.3003,
      "step": 29050
    },
    {
      "epoch": 0.32,
      "grad_norm": 4.863938331604004,
      "learning_rate": 0.000162056116560834,
      "loss": 3.2411,
      "step": 29100
    },
    {
      "epoch": 0.32,
      "grad_norm": 5.351004123687744,
      "learning_rate": 0.00016191223861779529,
      "loss": 3.2409,
      "step": 29150
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.316920280456543,
      "learning_rate": 0.0001617681525478687,
      "loss": 3.0881,
      "step": 29200
    },
    {
      "epoch": 0.32,
      "grad_norm": 7.093586444854736,
      "learning_rate": 0.0001616238588354203,
      "loss": 3.2573,
      "step": 29250
    },
    {
      "epoch": 0.32,
      "grad_norm": 6.5853705406188965,
      "learning_rate": 0.00016147935796551405,
      "loss": 3.3215,
      "step": 29300
    },
    {
      "epoch": 0.33,
      "grad_norm": 4.667483329772949,
      "learning_rate": 0.00016133465042391046,
      "loss": 3.3032,
      "step": 29350
    },
    {
      "epoch": 0.33,
      "grad_norm": 5.287956237792969,
      "learning_rate": 0.00016118973669706468,
      "loss": 3.2255,
      "step": 29400
    },
    {
      "epoch": 0.33,
      "grad_norm": 5.676483631134033,
      "learning_rate": 0.0001610446172721251,
      "loss": 3.3663,
      "step": 29450
    },
    {
      "epoch": 0.33,
      "grad_norm": 6.739033222198486,
      "learning_rate": 0.00016089929263693144,
      "loss": 3.2537,
      "step": 29500
    },
    {
      "epoch": 0.33,
      "grad_norm": 5.158905982971191,
      "learning_rate": 0.00016075376328001344,
      "loss": 3.2336,
      "step": 29550
    },
    {
      "epoch": 0.33,
      "grad_norm": 5.2512526512146,
      "learning_rate": 0.00016060802969058885,
      "loss": 3.2982,
      "step": 29600
    },
    {
      "epoch": 0.33,
      "grad_norm": 8.520125389099121,
      "learning_rate": 0.00016046209235856212,
      "loss": 3.3153,
      "step": 29650
    },
    {
      "epoch": 0.33,
      "grad_norm": 6.461225509643555,
      "learning_rate": 0.00016031595177452257,
      "loss": 3.2629,
      "step": 29700
    },
    {
      "epoch": 0.33,
      "grad_norm": 6.117869853973389,
      "learning_rate": 0.00016016960842974278,
      "loss": 3.2225,
      "step": 29750
    },
    {
      "epoch": 0.33,
      "grad_norm": 4.035212993621826,
      "learning_rate": 0.00016002306281617692,
      "loss": 3.3866,
      "step": 29800
    },
    {
      "epoch": 0.33,
      "grad_norm": 5.271117687225342,
      "learning_rate": 0.00015987631542645913,
      "loss": 3.2602,
      "step": 29850
    },
    {
      "epoch": 0.33,
      "grad_norm": 4.860154151916504,
      "learning_rate": 0.00015972936675390185,
      "loss": 3.288,
      "step": 29900
    },
    {
      "epoch": 0.33,
      "grad_norm": 5.154600143432617,
      "learning_rate": 0.0001595822172924942,
      "loss": 3.1941,
      "step": 29950
    },
    {
      "epoch": 0.33,
      "grad_norm": 6.897374153137207,
      "learning_rate": 0.00015943486753690017,
      "loss": 3.2323,
      "step": 30000
    },
    {
      "epoch": 0.33,
      "grad_norm": 6.063130855560303,
      "learning_rate": 0.00015928731798245721,
      "loss": 3.1718,
      "step": 30050
    },
    {
      "epoch": 0.33,
      "grad_norm": 5.736262321472168,
      "learning_rate": 0.00015913956912517432,
      "loss": 3.3035,
      "step": 30100
    },
    {
      "epoch": 0.33,
      "grad_norm": 5.386317253112793,
      "learning_rate": 0.00015899162146173053,
      "loss": 3.2879,
      "step": 30150
    },
    {
      "epoch": 0.33,
      "grad_norm": 5.535543441772461,
      "learning_rate": 0.00015884347548947314,
      "loss": 3.2266,
      "step": 30200
    },
    {
      "epoch": 0.34,
      "grad_norm": 6.057496070861816,
      "learning_rate": 0.00015869513170641616,
      "loss": 3.1668,
      "step": 30250
    },
    {
      "epoch": 0.34,
      "grad_norm": 5.6912055015563965,
      "learning_rate": 0.00015854659061123854,
      "loss": 3.1562,
      "step": 30300
    },
    {
      "epoch": 0.34,
      "grad_norm": 5.563050270080566,
      "learning_rate": 0.0001583978527032825,
      "loss": 3.1819,
      "step": 30350
    },
    {
      "epoch": 0.34,
      "grad_norm": 6.75504732131958,
      "learning_rate": 0.0001582489184825519,
      "loss": 3.1891,
      "step": 30400
    },
    {
      "epoch": 0.34,
      "grad_norm": 5.390425205230713,
      "learning_rate": 0.00015809978844971053,
      "loss": 3.1856,
      "step": 30450
    },
    {
      "epoch": 0.34,
      "grad_norm": 6.183398246765137,
      "learning_rate": 0.0001579504631060804,
      "loss": 3.3115,
      "step": 30500
    },
    {
      "epoch": 0.34,
      "grad_norm": 5.40380859375,
      "learning_rate": 0.00015780094295364015,
      "loss": 3.162,
      "step": 30550
    },
    {
      "epoch": 0.34,
      "grad_norm": 5.078804016113281,
      "learning_rate": 0.00015765122849502325,
      "loss": 3.2046,
      "step": 30600
    },
    {
      "epoch": 0.34,
      "grad_norm": 6.183681964874268,
      "learning_rate": 0.00015750132023351638,
      "loss": 3.0689,
      "step": 30650
    },
    {
      "epoch": 0.34,
      "grad_norm": 12.730826377868652,
      "learning_rate": 0.00015735121867305768,
      "loss": 3.2468,
      "step": 30700
    },
    {
      "epoch": 0.34,
      "grad_norm": 6.127053260803223,
      "learning_rate": 0.00015720092431823515,
      "loss": 3.1628,
      "step": 30750
    },
    {
      "epoch": 0.34,
      "grad_norm": 5.8310089111328125,
      "learning_rate": 0.00015705043767428483,
      "loss": 3.2047,
      "step": 30800
    },
    {
      "epoch": 0.34,
      "grad_norm": 7.505776882171631,
      "learning_rate": 0.0001568997592470892,
      "loss": 3.2827,
      "step": 30850
    },
    {
      "epoch": 0.34,
      "grad_norm": 5.399072170257568,
      "learning_rate": 0.00015674888954317549,
      "loss": 3.1483,
      "step": 30900
    },
    {
      "epoch": 0.34,
      "grad_norm": 5.224669456481934,
      "learning_rate": 0.00015659782906971383,
      "loss": 3.2698,
      "step": 30950
    },
    {
      "epoch": 0.34,
      "grad_norm": 7.958742618560791,
      "learning_rate": 0.00015644657833451577,
      "loss": 3.0145,
      "step": 31000
    },
    {
      "epoch": 0.34,
      "grad_norm": 6.20373010635376,
      "learning_rate": 0.0001562981685120925,
      "loss": 3.1598,
      "step": 31050
    },
    {
      "epoch": 0.34,
      "grad_norm": 5.2080159187316895,
      "learning_rate": 0.00015614654255930347,
      "loss": 3.2801,
      "step": 31100
    },
    {
      "epoch": 0.35,
      "grad_norm": 4.195250988006592,
      "learning_rate": 0.00015599472786184245,
      "loss": 3.156,
      "step": 31150
    },
    {
      "epoch": 0.35,
      "grad_norm": 5.0389204025268555,
      "learning_rate": 0.00015584272493005642,
      "loss": 3.1345,
      "step": 31200
    },
    {
      "epoch": 0.35,
      "grad_norm": 7.003210544586182,
      "learning_rate": 0.00015569053427492505,
      "loss": 3.2186,
      "step": 31250
    },
    {
      "epoch": 0.35,
      "grad_norm": 5.47674036026001,
      "learning_rate": 0.00015553815640805907,
      "loss": 3.3211,
      "step": 31300
    },
    {
      "epoch": 0.35,
      "grad_norm": 5.2981648445129395,
      "learning_rate": 0.00015538559184169863,
      "loss": 3.2454,
      "step": 31350
    },
    {
      "epoch": 0.35,
      "grad_norm": 5.528575420379639,
      "learning_rate": 0.00015523284108871142,
      "loss": 3.1963,
      "step": 31400
    },
    {
      "epoch": 0.35,
      "grad_norm": 5.553009033203125,
      "learning_rate": 0.0001550799046625911,
      "loss": 3.1682,
      "step": 31450
    },
    {
      "epoch": 0.35,
      "grad_norm": 5.608404636383057,
      "learning_rate": 0.0001549267830774553,
      "loss": 3.1461,
      "step": 31500
    },
    {
      "epoch": 0.35,
      "grad_norm": 6.462625503540039,
      "learning_rate": 0.00015477347684804445,
      "loss": 3.2772,
      "step": 31550
    },
    {
      "epoch": 0.35,
      "grad_norm": 5.327962875366211,
      "learning_rate": 0.00015461998648971928,
      "loss": 3.2144,
      "step": 31600
    },
    {
      "epoch": 0.35,
      "grad_norm": 6.94124174118042,
      "learning_rate": 0.00015446631251845978,
      "loss": 3.2227,
      "step": 31650
    },
    {
      "epoch": 0.35,
      "grad_norm": 5.351782321929932,
      "learning_rate": 0.00015431245545086307,
      "loss": 3.2687,
      "step": 31700
    },
    {
      "epoch": 0.35,
      "grad_norm": 4.562844276428223,
      "learning_rate": 0.00015415841580414185,
      "loss": 3.1332,
      "step": 31750
    },
    {
      "epoch": 0.35,
      "grad_norm": 5.023700714111328,
      "learning_rate": 0.00015400419409612243,
      "loss": 3.2272,
      "step": 31800
    },
    {
      "epoch": 0.35,
      "grad_norm": 5.127398490905762,
      "learning_rate": 0.0001538497908452433,
      "loss": 3.2843,
      "step": 31850
    },
    {
      "epoch": 0.35,
      "grad_norm": 5.577905178070068,
      "learning_rate": 0.0001536952065705532,
      "loss": 3.2635,
      "step": 31900
    },
    {
      "epoch": 0.35,
      "grad_norm": 6.119299411773682,
      "learning_rate": 0.00015354044179170933,
      "loss": 3.126,
      "step": 31950
    },
    {
      "epoch": 0.35,
      "grad_norm": 4.827983856201172,
      "learning_rate": 0.0001533854970289758,
      "loss": 3.2345,
      "step": 32000
    },
    {
      "epoch": 0.36,
      "grad_norm": 4.499656677246094,
      "learning_rate": 0.00015323037280322166,
      "loss": 3.0808,
      "step": 32050
    },
    {
      "epoch": 0.36,
      "grad_norm": 5.260239601135254,
      "learning_rate": 0.00015307506963591923,
      "loss": 3.1234,
      "step": 32100
    },
    {
      "epoch": 0.36,
      "grad_norm": 5.486075401306152,
      "learning_rate": 0.00015291958804914256,
      "loss": 3.1769,
      "step": 32150
    },
    {
      "epoch": 0.36,
      "grad_norm": 4.572110176086426,
      "learning_rate": 0.00015276392856556527,
      "loss": 3.2166,
      "step": 32200
    },
    {
      "epoch": 0.36,
      "grad_norm": 5.288125991821289,
      "learning_rate": 0.0001526080917084591,
      "loss": 3.0781,
      "step": 32250
    },
    {
      "epoch": 0.36,
      "grad_norm": 6.380829334259033,
      "learning_rate": 0.000152452078001692,
      "loss": 3.1178,
      "step": 32300
    },
    {
      "epoch": 0.36,
      "grad_norm": 5.039462566375732,
      "learning_rate": 0.00015229588796972652,
      "loss": 3.2808,
      "step": 32350
    },
    {
      "epoch": 0.36,
      "grad_norm": 7.323626518249512,
      "learning_rate": 0.00015213952213761787,
      "loss": 3.1391,
      "step": 32400
    },
    {
      "epoch": 0.36,
      "grad_norm": 5.912395000457764,
      "learning_rate": 0.00015198298103101228,
      "loss": 3.1744,
      "step": 32450
    },
    {
      "epoch": 0.36,
      "grad_norm": 5.694441795349121,
      "learning_rate": 0.00015182626517614518,
      "loss": 3.0576,
      "step": 32500
    },
    {
      "epoch": 0.36,
      "grad_norm": 6.467188358306885,
      "learning_rate": 0.00015166937509983943,
      "loss": 3.2361,
      "step": 32550
    },
    {
      "epoch": 0.36,
      "grad_norm": 5.292226791381836,
      "learning_rate": 0.00015151231132950357,
      "loss": 3.1376,
      "step": 32600
    },
    {
      "epoch": 0.36,
      "grad_norm": 6.369929313659668,
      "learning_rate": 0.00015135507439313005,
      "loss": 3.1406,
      "step": 32650
    },
    {
      "epoch": 0.36,
      "grad_norm": 5.252573013305664,
      "learning_rate": 0.00015119766481929342,
      "loss": 3.123,
      "step": 32700
    },
    {
      "epoch": 0.36,
      "grad_norm": 6.9053730964660645,
      "learning_rate": 0.00015104008313714858,
      "loss": 3.0018,
      "step": 32750
    },
    {
      "epoch": 0.36,
      "grad_norm": 6.080839157104492,
      "learning_rate": 0.00015088232987642898,
      "loss": 3.2106,
      "step": 32800
    },
    {
      "epoch": 0.36,
      "grad_norm": 5.92653226852417,
      "learning_rate": 0.00015072440556744492,
      "loss": 3.2095,
      "step": 32850
    },
    {
      "epoch": 0.36,
      "grad_norm": 5.627429008483887,
      "learning_rate": 0.00015056631074108166,
      "loss": 3.089,
      "step": 32900
    },
    {
      "epoch": 0.37,
      "grad_norm": 5.694194793701172,
      "learning_rate": 0.00015040804592879762,
      "loss": 3.0885,
      "step": 32950
    },
    {
      "epoch": 0.37,
      "grad_norm": 5.04107666015625,
      "learning_rate": 0.00015024961166262276,
      "loss": 3.0906,
      "step": 33000
    },
    {
      "epoch": 0.37,
      "grad_norm": 5.280002117156982,
      "learning_rate": 0.0001500910084751567,
      "loss": 3.2142,
      "step": 33050
    },
    {
      "epoch": 0.37,
      "grad_norm": 4.803068161010742,
      "learning_rate": 0.00014993223689956672,
      "loss": 3.2014,
      "step": 33100
    },
    {
      "epoch": 0.37,
      "grad_norm": 5.611780643463135,
      "learning_rate": 0.00014977329746958636,
      "loss": 3.2491,
      "step": 33150
    },
    {
      "epoch": 0.37,
      "grad_norm": 5.6669020652771,
      "learning_rate": 0.00014961737449079314,
      "loss": 3.2,
      "step": 33200
    },
    {
      "epoch": 0.37,
      "grad_norm": 5.894138336181641,
      "learning_rate": 0.00014945810428594703,
      "loss": 3.0321,
      "step": 33250
    },
    {
      "epoch": 0.37,
      "grad_norm": 6.354518413543701,
      "learning_rate": 0.0001492986678205755,
      "loss": 3.1314,
      "step": 33300
    },
    {
      "epoch": 0.37,
      "grad_norm": 4.786489009857178,
      "learning_rate": 0.00014913906563064706,
      "loss": 3.1937,
      "step": 33350
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.759417533874512,
      "learning_rate": 0.00014897929825268745,
      "loss": 3.1069,
      "step": 33400
    },
    {
      "epoch": 0.37,
      "grad_norm": 5.354910850524902,
      "learning_rate": 0.00014881936622377766,
      "loss": 3.1519,
      "step": 33450
    },
    {
      "epoch": 0.37,
      "grad_norm": 5.169478416442871,
      "learning_rate": 0.0001486592700815522,
      "loss": 3.1414,
      "step": 33500
    },
    {
      "epoch": 0.37,
      "grad_norm": 6.09418249130249,
      "learning_rate": 0.00014849901036419723,
      "loss": 3.0954,
      "step": 33550
    },
    {
      "epoch": 0.37,
      "grad_norm": 4.986037731170654,
      "learning_rate": 0.00014833858761044883,
      "loss": 3.2445,
      "step": 33600
    },
    {
      "epoch": 0.37,
      "grad_norm": 12.803654670715332,
      "learning_rate": 0.00014817800235959118,
      "loss": 3.0699,
      "step": 33650
    },
    {
      "epoch": 0.37,
      "grad_norm": 6.191990852355957,
      "learning_rate": 0.00014801725515145467,
      "loss": 3.2574,
      "step": 33700
    },
    {
      "epoch": 0.37,
      "grad_norm": 5.233778476715088,
      "learning_rate": 0.00014785634652641412,
      "loss": 3.1152,
      "step": 33750
    },
    {
      "epoch": 0.37,
      "grad_norm": 6.065474987030029,
      "learning_rate": 0.000147695277025387,
      "loss": 3.1178,
      "step": 33800
    },
    {
      "epoch": 0.38,
      "grad_norm": 5.4664435386657715,
      "learning_rate": 0.00014753404718983158,
      "loss": 3.0627,
      "step": 33850
    },
    {
      "epoch": 0.38,
      "grad_norm": 4.926270484924316,
      "learning_rate": 0.00014737265756174515,
      "loss": 3.0182,
      "step": 33900
    },
    {
      "epoch": 0.38,
      "grad_norm": 6.013931751251221,
      "learning_rate": 0.0001472111086836621,
      "loss": 3.0801,
      "step": 33950
    },
    {
      "epoch": 0.38,
      "grad_norm": 6.25607442855835,
      "learning_rate": 0.00014704940109865224,
      "loss": 3.1227,
      "step": 34000
    },
    {
      "epoch": 0.38,
      "grad_norm": 5.084470272064209,
      "learning_rate": 0.00014688753535031882,
      "loss": 3.1786,
      "step": 34050
    },
    {
      "epoch": 0.38,
      "grad_norm": 5.16444730758667,
      "learning_rate": 0.00014672551198279687,
      "loss": 3.0651,
      "step": 34100
    },
    {
      "epoch": 0.38,
      "grad_norm": 5.316379070281982,
      "learning_rate": 0.00014656333154075118,
      "loss": 2.9925,
      "step": 34150
    },
    {
      "epoch": 0.38,
      "grad_norm": 5.331335067749023,
      "learning_rate": 0.00014640099456937462,
      "loss": 3.097,
      "step": 34200
    },
    {
      "epoch": 0.38,
      "grad_norm": 4.452579498291016,
      "learning_rate": 0.00014623850161438626,
      "loss": 3.1222,
      "step": 34250
    },
    {
      "epoch": 0.38,
      "grad_norm": 5.550851345062256,
      "learning_rate": 0.00014607585322202953,
      "loss": 3.2343,
      "step": 34300
    },
    {
      "epoch": 0.38,
      "grad_norm": 6.5701375007629395,
      "learning_rate": 0.00014591304993907033,
      "loss": 3.1558,
      "step": 34350
    },
    {
      "epoch": 0.38,
      "grad_norm": 5.140341758728027,
      "learning_rate": 0.00014575009231279534,
      "loss": 3.0036,
      "step": 34400
    },
    {
      "epoch": 0.38,
      "grad_norm": 4.767757415771484,
      "learning_rate": 0.00014558698089101003,
      "loss": 3.1355,
      "step": 34450
    },
    {
      "epoch": 0.38,
      "grad_norm": 6.189707279205322,
      "learning_rate": 0.00014542371622203689,
      "loss": 3.0721,
      "step": 34500
    },
    {
      "epoch": 0.38,
      "grad_norm": 4.492117881774902,
      "learning_rate": 0.00014526029885471355,
      "loss": 3.1083,
      "step": 34550
    },
    {
      "epoch": 0.38,
      "grad_norm": 5.021632671356201,
      "learning_rate": 0.000145096729338391,
      "loss": 3.1291,
      "step": 34600
    },
    {
      "epoch": 0.38,
      "grad_norm": 5.501243591308594,
      "learning_rate": 0.00014493300822293164,
      "loss": 3.0654,
      "step": 34650
    },
    {
      "epoch": 0.38,
      "grad_norm": 5.056375026702881,
      "learning_rate": 0.0001447691360587076,
      "loss": 3.054,
      "step": 34700
    },
    {
      "epoch": 0.39,
      "grad_norm": 5.866014003753662,
      "learning_rate": 0.0001446051133965986,
      "loss": 3.1111,
      "step": 34750
    },
    {
      "epoch": 0.39,
      "grad_norm": 5.377662658691406,
      "learning_rate": 0.0001444409407879905,
      "loss": 3.2083,
      "step": 34800
    },
    {
      "epoch": 0.39,
      "grad_norm": 5.692511558532715,
      "learning_rate": 0.00014427661878477305,
      "loss": 3.1613,
      "step": 34850
    },
    {
      "epoch": 0.39,
      "grad_norm": 5.763862609863281,
      "learning_rate": 0.0001441121479393383,
      "loss": 3.1422,
      "step": 34900
    },
    {
      "epoch": 0.39,
      "grad_norm": 6.247575283050537,
      "learning_rate": 0.00014394752880457867,
      "loss": 3.1826,
      "step": 34950
    },
    {
      "epoch": 0.39,
      "grad_norm": 6.18945837020874,
      "learning_rate": 0.00014378276193388498,
      "loss": 3.0887,
      "step": 35000
    },
    {
      "epoch": 0.39,
      "grad_norm": 6.006703853607178,
      "learning_rate": 0.00014361784788114476,
      "loss": 3.0888,
      "step": 35050
    },
    {
      "epoch": 0.39,
      "grad_norm": 5.308047771453857,
      "learning_rate": 0.00014345278720074032,
      "loss": 3.1604,
      "step": 35100
    },
    {
      "epoch": 0.39,
      "grad_norm": 4.780633926391602,
      "learning_rate": 0.00014329088601053276,
      "loss": 3.0979,
      "step": 35150
    },
    {
      "epoch": 0.39,
      "grad_norm": 5.292586803436279,
      "learning_rate": 0.0001431255366448192,
      "loss": 3.1392,
      "step": 35200
    },
    {
      "epoch": 0.39,
      "grad_norm": 15.803680419921875,
      "learning_rate": 0.0001429600423064164,
      "loss": 3.1114,
      "step": 35250
    },
    {
      "epoch": 0.39,
      "grad_norm": 6.428526401519775,
      "learning_rate": 0.00014279440355165733,
      "loss": 2.9971,
      "step": 35300
    },
    {
      "epoch": 0.39,
      "grad_norm": 6.179375171661377,
      "learning_rate": 0.0001426286209373605,
      "loss": 3.0888,
      "step": 35350
    },
    {
      "epoch": 0.39,
      "grad_norm": 5.588586330413818,
      "learning_rate": 0.000142462695020828,
      "loss": 3.0704,
      "step": 35400
    },
    {
      "epoch": 0.39,
      "grad_norm": 4.596306324005127,
      "learning_rate": 0.00014229662635984358,
      "loss": 3.1505,
      "step": 35450
    },
    {
      "epoch": 0.39,
      "grad_norm": 5.788643836975098,
      "learning_rate": 0.00014213041551267098,
      "loss": 2.9664,
      "step": 35500
    },
    {
      "epoch": 0.39,
      "grad_norm": 6.529501438140869,
      "learning_rate": 0.00014196406303805181,
      "loss": 3.0943,
      "step": 35550
    },
    {
      "epoch": 0.39,
      "grad_norm": 6.052688121795654,
      "learning_rate": 0.00014179756949520385,
      "loss": 3.1545,
      "step": 35600
    },
    {
      "epoch": 0.4,
      "grad_norm": 4.716464042663574,
      "learning_rate": 0.00014163093544381904,
      "loss": 3.1625,
      "step": 35650
    },
    {
      "epoch": 0.4,
      "grad_norm": 5.132946014404297,
      "learning_rate": 0.0001414641614440617,
      "loss": 3.0922,
      "step": 35700
    },
    {
      "epoch": 0.4,
      "grad_norm": 5.718029975891113,
      "learning_rate": 0.00014129724805656666,
      "loss": 3.0912,
      "step": 35750
    },
    {
      "epoch": 0.4,
      "grad_norm": 5.161171913146973,
      "learning_rate": 0.00014113019584243716,
      "loss": 3.1888,
      "step": 35800
    },
    {
      "epoch": 0.4,
      "grad_norm": 5.421058177947998,
      "learning_rate": 0.0001409630053632433,
      "loss": 2.9738,
      "step": 35850
    },
    {
      "epoch": 0.4,
      "grad_norm": 5.962222576141357,
      "learning_rate": 0.00014079567718101987,
      "loss": 3.0353,
      "step": 35900
    },
    {
      "epoch": 0.4,
      "grad_norm": 5.4867753982543945,
      "learning_rate": 0.0001406282118582646,
      "loss": 3.0904,
      "step": 35950
    },
    {
      "epoch": 0.4,
      "grad_norm": 7.088072299957275,
      "learning_rate": 0.00014046060995793626,
      "loss": 3.0246,
      "step": 36000
    },
    {
      "epoch": 0.4,
      "grad_norm": 5.093292236328125,
      "learning_rate": 0.0001402928720434527,
      "loss": 2.9939,
      "step": 36050
    },
    {
      "epoch": 0.4,
      "grad_norm": 5.068340301513672,
      "learning_rate": 0.00014012499867868905,
      "loss": 3.118,
      "step": 36100
    },
    {
      "epoch": 0.4,
      "grad_norm": 5.125149726867676,
      "learning_rate": 0.00013995699042797574,
      "loss": 3.0984,
      "step": 36150
    },
    {
      "epoch": 0.4,
      "grad_norm": 5.709127902984619,
      "learning_rate": 0.00013978884785609668,
      "loss": 3.0354,
      "step": 36200
    },
    {
      "epoch": 0.4,
      "grad_norm": 4.788537979125977,
      "learning_rate": 0.00013962057152828726,
      "loss": 3.0648,
      "step": 36250
    },
    {
      "epoch": 0.4,
      "grad_norm": 6.115299701690674,
      "learning_rate": 0.00013945216201023252,
      "loss": 3.191,
      "step": 36300
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.774588584899902,
      "learning_rate": 0.0001392836198680653,
      "loss": 3.2434,
      "step": 36350
    },
    {
      "epoch": 0.4,
      "grad_norm": 5.179059982299805,
      "learning_rate": 0.00013911494566836417,
      "loss": 3.0184,
      "step": 36400
    },
    {
      "epoch": 0.4,
      "grad_norm": 5.080101490020752,
      "learning_rate": 0.00013894613997815174,
      "loss": 3.0847,
      "step": 36450
    },
    {
      "epoch": 0.4,
      "grad_norm": 5.972216606140137,
      "learning_rate": 0.00013877720336489264,
      "loss": 3.0745,
      "step": 36500
    },
    {
      "epoch": 0.41,
      "grad_norm": 4.853601455688477,
      "learning_rate": 0.0001386081363964915,
      "loss": 3.0191,
      "step": 36550
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.077406883239746,
      "learning_rate": 0.00013843893964129127,
      "loss": 3.0575,
      "step": 36600
    },
    {
      "epoch": 0.41,
      "grad_norm": 5.7632904052734375,
      "learning_rate": 0.00013826961366807114,
      "loss": 3.1308,
      "step": 36650
    },
    {
      "epoch": 0.41,
      "grad_norm": 7.178746700286865,
      "learning_rate": 0.00013810015904604472,
      "loss": 3.0838,
      "step": 36700
    },
    {
      "epoch": 0.41,
      "grad_norm": 5.9257402420043945,
      "learning_rate": 0.0001379305763448581,
      "loss": 3.1233,
      "step": 36750
    },
    {
      "epoch": 0.41,
      "grad_norm": 5.543473720550537,
      "learning_rate": 0.00013776086613458783,
      "loss": 3.0783,
      "step": 36800
    },
    {
      "epoch": 0.41,
      "grad_norm": 6.271590709686279,
      "learning_rate": 0.00013759102898573924,
      "loss": 2.9985,
      "step": 36850
    },
    {
      "epoch": 0.41,
      "grad_norm": 6.419349193572998,
      "learning_rate": 0.00013742106546924427,
      "loss": 3.093,
      "step": 36900
    },
    {
      "epoch": 0.41,
      "grad_norm": 6.282319068908691,
      "learning_rate": 0.00013725097615645973,
      "loss": 3.0827,
      "step": 36950
    },
    {
      "epoch": 0.41,
      "grad_norm": 5.733994960784912,
      "learning_rate": 0.00013708076161916529,
      "loss": 2.9967,
      "step": 37000
    },
    {
      "epoch": 0.41,
      "grad_norm": 5.090719699859619,
      "learning_rate": 0.00013691042242956156,
      "loss": 3.1321,
      "step": 37050
    },
    {
      "epoch": 0.41,
      "grad_norm": 4.850597381591797,
      "learning_rate": 0.00013673995916026823,
      "loss": 3.0506,
      "step": 37100
    },
    {
      "epoch": 0.41,
      "grad_norm": 5.414620876312256,
      "learning_rate": 0.00013656937238432207,
      "loss": 2.9907,
      "step": 37150
    },
    {
      "epoch": 0.41,
      "grad_norm": 5.465592384338379,
      "learning_rate": 0.00013639866267517502,
      "loss": 3.0623,
      "step": 37200
    },
    {
      "epoch": 0.41,
      "grad_norm": 5.292060375213623,
      "learning_rate": 0.00013622783060669236,
      "loss": 3.1343,
      "step": 37250
    },
    {
      "epoch": 0.41,
      "grad_norm": 5.4502644538879395,
      "learning_rate": 0.00013605687675315056,
      "loss": 3.0725,
      "step": 37300
    },
    {
      "epoch": 0.41,
      "grad_norm": 4.803048133850098,
      "learning_rate": 0.00013588580168923564,
      "loss": 3.0884,
      "step": 37350
    },
    {
      "epoch": 0.41,
      "grad_norm": 5.469101428985596,
      "learning_rate": 0.00013571460599004097,
      "loss": 3.0724,
      "step": 37400
    },
    {
      "epoch": 0.42,
      "grad_norm": 4.901301383972168,
      "learning_rate": 0.00013554329023106549,
      "loss": 3.1341,
      "step": 37450
    },
    {
      "epoch": 0.42,
      "grad_norm": 6.607944011688232,
      "learning_rate": 0.00013537185498821178,
      "loss": 3.003,
      "step": 37500
    },
    {
      "epoch": 0.42,
      "grad_norm": 5.942598342895508,
      "learning_rate": 0.00013520030083778403,
      "loss": 3.1121,
      "step": 37550
    },
    {
      "epoch": 0.42,
      "grad_norm": 4.80432653427124,
      "learning_rate": 0.00013502862835648618,
      "loss": 3.0313,
      "step": 37600
    },
    {
      "epoch": 0.42,
      "grad_norm": 5.739936828613281,
      "learning_rate": 0.00013485683812141993,
      "loss": 3.1094,
      "step": 37650
    },
    {
      "epoch": 0.42,
      "grad_norm": 5.018519401550293,
      "learning_rate": 0.00013468493071008286,
      "loss": 3.0696,
      "step": 37700
    },
    {
      "epoch": 0.42,
      "grad_norm": 6.4165825843811035,
      "learning_rate": 0.00013451290670036641,
      "loss": 3.1152,
      "step": 37750
    },
    {
      "epoch": 0.42,
      "grad_norm": 4.517113208770752,
      "learning_rate": 0.00013434076667055407,
      "loss": 2.886,
      "step": 37800
    },
    {
      "epoch": 0.42,
      "grad_norm": 7.886168956756592,
      "learning_rate": 0.00013416851119931922,
      "loss": 3.1064,
      "step": 37850
    },
    {
      "epoch": 0.42,
      "grad_norm": 6.000852108001709,
      "learning_rate": 0.0001339961408657235,
      "loss": 2.9484,
      "step": 37900
    },
    {
      "epoch": 0.42,
      "grad_norm": 5.562610626220703,
      "learning_rate": 0.00013382365624921438,
      "loss": 3.0847,
      "step": 37950
    },
    {
      "epoch": 0.42,
      "grad_norm": 5.29053258895874,
      "learning_rate": 0.00013365105792962384,
      "loss": 3.0938,
      "step": 38000
    },
    {
      "epoch": 0.42,
      "grad_norm": 6.465425491333008,
      "learning_rate": 0.0001334783464871658,
      "loss": 3.1012,
      "step": 38050
    },
    {
      "epoch": 0.42,
      "grad_norm": 5.68001651763916,
      "learning_rate": 0.00013330552250243475,
      "loss": 3.0371,
      "step": 38100
    },
    {
      "epoch": 0.42,
      "grad_norm": 6.374165058135986,
      "learning_rate": 0.00013313258655640325,
      "loss": 3.141,
      "step": 38150
    },
    {
      "epoch": 0.42,
      "grad_norm": 6.1900410652160645,
      "learning_rate": 0.00013295953923042034,
      "loss": 2.9999,
      "step": 38200
    },
    {
      "epoch": 0.42,
      "grad_norm": 5.627082824707031,
      "learning_rate": 0.00013278638110620952,
      "loss": 2.9751,
      "step": 38250
    },
    {
      "epoch": 0.42,
      "grad_norm": 6.941247463226318,
      "learning_rate": 0.0001326131127658667,
      "loss": 3.0579,
      "step": 38300
    },
    {
      "epoch": 0.43,
      "grad_norm": 5.677605152130127,
      "learning_rate": 0.00013243973479185828,
      "loss": 3.0327,
      "step": 38350
    },
    {
      "epoch": 0.43,
      "grad_norm": 5.137176513671875,
      "learning_rate": 0.00013226624776701925,
      "loss": 3.0782,
      "step": 38400
    },
    {
      "epoch": 0.43,
      "grad_norm": 7.087386131286621,
      "learning_rate": 0.00013209265227455118,
      "loss": 3.0633,
      "step": 38450
    },
    {
      "epoch": 0.43,
      "grad_norm": 4.575850486755371,
      "learning_rate": 0.0001319189488980203,
      "loss": 3.1038,
      "step": 38500
    },
    {
      "epoch": 0.43,
      "grad_norm": 4.31092643737793,
      "learning_rate": 0.0001317451382213554,
      "loss": 3.0153,
      "step": 38550
    },
    {
      "epoch": 0.43,
      "grad_norm": 5.346449375152588,
      "learning_rate": 0.0001315712208288461,
      "loss": 2.998,
      "step": 38600
    },
    {
      "epoch": 0.43,
      "grad_norm": 5.6449198722839355,
      "learning_rate": 0.0001313971973051407,
      "loss": 3.0801,
      "step": 38650
    },
    {
      "epoch": 0.43,
      "grad_norm": 5.223085403442383,
      "learning_rate": 0.00013122306823524425,
      "loss": 2.9869,
      "step": 38700
    },
    {
      "epoch": 0.43,
      "grad_norm": 5.5236053466796875,
      "learning_rate": 0.00013104883420451673,
      "loss": 2.865,
      "step": 38750
    },
    {
      "epoch": 0.43,
      "grad_norm": 6.014589786529541,
      "learning_rate": 0.00013087449579867082,
      "loss": 2.9073,
      "step": 38800
    },
    {
      "epoch": 0.43,
      "grad_norm": 4.373612403869629,
      "learning_rate": 0.0001307000536037701,
      "loss": 2.982,
      "step": 38850
    },
    {
      "epoch": 0.43,
      "grad_norm": 4.678324222564697,
      "learning_rate": 0.00013052550820622712,
      "loss": 3.0649,
      "step": 38900
    },
    {
      "epoch": 0.43,
      "grad_norm": 5.631078720092773,
      "learning_rate": 0.0001303508601928013,
      "loss": 3.0235,
      "step": 38950
    },
    {
      "epoch": 0.43,
      "grad_norm": 5.745089054107666,
      "learning_rate": 0.00013017611015059703,
      "loss": 3.017,
      "step": 39000
    },
    {
      "epoch": 0.43,
      "grad_norm": 4.9316086769104,
      "learning_rate": 0.00013000125866706173,
      "loss": 2.8922,
      "step": 39050
    },
    {
      "epoch": 0.43,
      "grad_norm": 4.83328104019165,
      "learning_rate": 0.00012982630632998375,
      "loss": 2.9944,
      "step": 39100
    },
    {
      "epoch": 0.43,
      "grad_norm": 4.858354568481445,
      "learning_rate": 0.00012965125372749048,
      "loss": 3.2066,
      "step": 39150
    },
    {
      "epoch": 0.43,
      "grad_norm": 4.85454797744751,
      "learning_rate": 0.00012947610144804653,
      "loss": 3.0025,
      "step": 39200
    },
    {
      "epoch": 0.44,
      "grad_norm": 4.420687675476074,
      "learning_rate": 0.00012930085008045134,
      "loss": 3.0974,
      "step": 39250
    },
    {
      "epoch": 0.44,
      "grad_norm": 5.3282012939453125,
      "learning_rate": 0.0001291255002138376,
      "loss": 2.9854,
      "step": 39300
    },
    {
      "epoch": 0.44,
      "grad_norm": 5.671788692474365,
      "learning_rate": 0.0001289500524376691,
      "loss": 2.9885,
      "step": 39350
    },
    {
      "epoch": 0.44,
      "grad_norm": 4.758500099182129,
      "learning_rate": 0.00012877450734173875,
      "loss": 3.0249,
      "step": 39400
    },
    {
      "epoch": 0.44,
      "grad_norm": 5.179042339324951,
      "learning_rate": 0.00012859886551616664,
      "loss": 2.9665,
      "step": 39450
    },
    {
      "epoch": 0.44,
      "grad_norm": 4.027754306793213,
      "learning_rate": 0.00012842312755139795,
      "loss": 2.9796,
      "step": 39500
    },
    {
      "epoch": 0.44,
      "grad_norm": 4.407317638397217,
      "learning_rate": 0.00012824729403820118,
      "loss": 3.0149,
      "step": 39550
    },
    {
      "epoch": 0.44,
      "grad_norm": 4.677199840545654,
      "learning_rate": 0.00012807136556766587,
      "loss": 3.0432,
      "step": 39600
    },
    {
      "epoch": 0.44,
      "grad_norm": 4.903664588928223,
      "learning_rate": 0.00012789534273120094,
      "loss": 2.9611,
      "step": 39650
    },
    {
      "epoch": 0.44,
      "grad_norm": 4.799653053283691,
      "learning_rate": 0.0001277192261205324,
      "loss": 3.0132,
      "step": 39700
    },
    {
      "epoch": 0.44,
      "grad_norm": 5.403965473175049,
      "learning_rate": 0.00012754301632770157,
      "loss": 2.9755,
      "step": 39750
    },
    {
      "epoch": 0.44,
      "grad_norm": 15.81489086151123,
      "learning_rate": 0.000127366713945063,
      "loss": 2.9749,
      "step": 39800
    },
    {
      "epoch": 0.44,
      "grad_norm": 5.791144847869873,
      "learning_rate": 0.00012719031956528247,
      "loss": 2.9958,
      "step": 39850
    },
    {
      "epoch": 0.44,
      "grad_norm": 6.054769992828369,
      "learning_rate": 0.000127013833781335,
      "loss": 3.0009,
      "step": 39900
    },
    {
      "epoch": 0.44,
      "grad_norm": 4.867542743682861,
      "learning_rate": 0.00012683725718650298,
      "loss": 3.0065,
      "step": 39950
    },
    {
      "epoch": 0.44,
      "grad_norm": 6.006402015686035,
      "learning_rate": 0.00012666059037437402,
      "loss": 2.9425,
      "step": 40000
    },
    {
      "epoch": 0.44,
      "grad_norm": 4.8633551597595215,
      "learning_rate": 0.00012648383393883893,
      "loss": 2.9857,
      "step": 40050
    },
    {
      "epoch": 0.44,
      "grad_norm": 4.701941013336182,
      "learning_rate": 0.0001263069884740899,
      "loss": 2.9212,
      "step": 40100
    },
    {
      "epoch": 0.45,
      "grad_norm": 4.5726237297058105,
      "learning_rate": 0.0001261300545746184,
      "loss": 2.9625,
      "step": 40150
    },
    {
      "epoch": 0.45,
      "grad_norm": 7.131462574005127,
      "learning_rate": 0.00012595303283521316,
      "loss": 3.0548,
      "step": 40200
    },
    {
      "epoch": 0.45,
      "grad_norm": 5.9713826179504395,
      "learning_rate": 0.00012577592385095814,
      "loss": 2.9929,
      "step": 40250
    },
    {
      "epoch": 0.45,
      "grad_norm": 4.615350723266602,
      "learning_rate": 0.0001255987282172307,
      "loss": 3.0768,
      "step": 40300
    },
    {
      "epoch": 0.45,
      "grad_norm": 5.178917407989502,
      "learning_rate": 0.00012542144652969944,
      "loss": 3.0278,
      "step": 40350
    },
    {
      "epoch": 0.45,
      "grad_norm": 6.033947944641113,
      "learning_rate": 0.0001252440793843222,
      "loss": 2.984,
      "step": 40400
    },
    {
      "epoch": 0.45,
      "grad_norm": 4.493157863616943,
      "learning_rate": 0.00012506662737734413,
      "loss": 3.0221,
      "step": 40450
    },
    {
      "epoch": 0.45,
      "grad_norm": 5.638075351715088,
      "learning_rate": 0.00012488909110529563,
      "loss": 3.0525,
      "step": 40500
    },
    {
      "epoch": 0.45,
      "grad_norm": 6.92527437210083,
      "learning_rate": 0.00012471147116499044,
      "loss": 2.9417,
      "step": 40550
    },
    {
      "epoch": 0.45,
      "grad_norm": 5.838274955749512,
      "learning_rate": 0.0001245337681535235,
      "loss": 2.9705,
      "step": 40600
    },
    {
      "epoch": 0.45,
      "grad_norm": 5.925703525543213,
      "learning_rate": 0.00012435598266826906,
      "loss": 2.9178,
      "step": 40650
    },
    {
      "epoch": 0.45,
      "grad_norm": 5.92296028137207,
      "learning_rate": 0.0001241781153068785,
      "loss": 3.0117,
      "step": 40700
    },
    {
      "epoch": 0.45,
      "grad_norm": 5.870546817779541,
      "learning_rate": 0.00012400016666727856,
      "loss": 2.9036,
      "step": 40750
    },
    {
      "epoch": 0.45,
      "grad_norm": 4.947800159454346,
      "learning_rate": 0.00012382213734766918,
      "loss": 2.9763,
      "step": 40800
    },
    {
      "epoch": 0.45,
      "grad_norm": 5.460611820220947,
      "learning_rate": 0.00012364402794652146,
      "loss": 2.9834,
      "step": 40850
    },
    {
      "epoch": 0.45,
      "grad_norm": 4.415246486663818,
      "learning_rate": 0.00012346583906257578,
      "loss": 2.9728,
      "step": 40900
    },
    {
      "epoch": 0.45,
      "grad_norm": 4.871602535247803,
      "learning_rate": 0.00012328757129483967,
      "loss": 3.1404,
      "step": 40950
    },
    {
      "epoch": 0.45,
      "grad_norm": 5.851921558380127,
      "learning_rate": 0.00012310922524258588,
      "loss": 2.9854,
      "step": 41000
    },
    {
      "epoch": 0.46,
      "grad_norm": 5.699054718017578,
      "learning_rate": 0.00012293080150535025,
      "loss": 3.0183,
      "step": 41050
    },
    {
      "epoch": 0.46,
      "grad_norm": 5.255423069000244,
      "learning_rate": 0.00012275230068292987,
      "loss": 3.0225,
      "step": 41100
    },
    {
      "epoch": 0.46,
      "grad_norm": 6.410534858703613,
      "learning_rate": 0.00012257372337538087,
      "loss": 2.9212,
      "step": 41150
    },
    {
      "epoch": 0.46,
      "grad_norm": 5.757706642150879,
      "learning_rate": 0.0001223950701830165,
      "loss": 2.9113,
      "step": 41200
    },
    {
      "epoch": 0.46,
      "grad_norm": 5.773000717163086,
      "learning_rate": 0.00012221634170640525,
      "loss": 3.0744,
      "step": 41250
    },
    {
      "epoch": 0.46,
      "grad_norm": 5.893006801605225,
      "learning_rate": 0.00012203753854636848,
      "loss": 2.9178,
      "step": 41300
    },
    {
      "epoch": 0.46,
      "grad_norm": 4.773557186126709,
      "learning_rate": 0.00012185866130397874,
      "loss": 3.0172,
      "step": 41350
    },
    {
      "epoch": 0.46,
      "grad_norm": 5.521202087402344,
      "learning_rate": 0.00012167971058055758,
      "loss": 3.0125,
      "step": 41400
    },
    {
      "epoch": 0.46,
      "grad_norm": 4.666463375091553,
      "learning_rate": 0.00012150068697767361,
      "loss": 3.01,
      "step": 41450
    },
    {
      "epoch": 0.46,
      "grad_norm": 6.500729084014893,
      "learning_rate": 0.00012132159109714036,
      "loss": 2.9112,
      "step": 41500
    },
    {
      "epoch": 0.46,
      "grad_norm": 4.394867897033691,
      "learning_rate": 0.0001211424235410144,
      "loss": 2.9984,
      "step": 41550
    },
    {
      "epoch": 0.46,
      "grad_norm": 5.369923114776611,
      "learning_rate": 0.00012096318491159317,
      "loss": 3.0215,
      "step": 41600
    },
    {
      "epoch": 0.46,
      "grad_norm": 4.682898998260498,
      "learning_rate": 0.00012078387581141315,
      "loss": 3.0607,
      "step": 41650
    },
    {
      "epoch": 0.46,
      "grad_norm": 6.522764205932617,
      "learning_rate": 0.00012060449684324761,
      "loss": 2.8554,
      "step": 41700
    },
    {
      "epoch": 0.46,
      "grad_norm": 4.820666790008545,
      "learning_rate": 0.00012042504861010472,
      "loss": 2.9452,
      "step": 41750
    },
    {
      "epoch": 0.46,
      "grad_norm": 4.502612113952637,
      "learning_rate": 0.00012024553171522557,
      "loss": 2.8537,
      "step": 41800
    },
    {
      "epoch": 0.46,
      "grad_norm": 6.1411919593811035,
      "learning_rate": 0.00012006594676208191,
      "loss": 2.8307,
      "step": 41850
    },
    {
      "epoch": 0.46,
      "grad_norm": 6.050970554351807,
      "learning_rate": 0.00011988629435437444,
      "loss": 3.0305,
      "step": 41900
    },
    {
      "epoch": 0.47,
      "grad_norm": 5.647627830505371,
      "learning_rate": 0.0001197065750960305,
      "loss": 2.8381,
      "step": 41950
    },
    {
      "epoch": 0.47,
      "grad_norm": 6.286309719085693,
      "learning_rate": 0.00011952678959120224,
      "loss": 2.9592,
      "step": 42000
    },
    {
      "epoch": 0.47,
      "grad_norm": 5.323746204376221,
      "learning_rate": 0.00011935053610658573,
      "loss": 2.9537,
      "step": 42050
    },
    {
      "epoch": 0.47,
      "grad_norm": 5.923861503601074,
      "learning_rate": 0.00011917062121695774,
      "loss": 2.9324,
      "step": 42100
    },
    {
      "epoch": 0.47,
      "grad_norm": 5.180017948150635,
      "learning_rate": 0.00011899064188253139,
      "loss": 3.0805,
      "step": 42150
    },
    {
      "epoch": 0.47,
      "grad_norm": 5.270453929901123,
      "learning_rate": 0.00011881059870833311,
      "loss": 2.9936,
      "step": 42200
    },
    {
      "epoch": 0.47,
      "grad_norm": 4.614379405975342,
      "learning_rate": 0.0001186304922996039,
      "loss": 3.0418,
      "step": 42250
    },
    {
      "epoch": 0.47,
      "grad_norm": 5.335697174072266,
      "learning_rate": 0.00011845032326179733,
      "loss": 2.8929,
      "step": 42300
    },
    {
      "epoch": 0.47,
      "grad_norm": 4.8122968673706055,
      "learning_rate": 0.00011827009220057747,
      "loss": 2.9899,
      "step": 42350
    },
    {
      "epoch": 0.47,
      "grad_norm": 5.42422342300415,
      "learning_rate": 0.00011808979972181702,
      "loss": 2.9105,
      "step": 42400
    },
    {
      "epoch": 0.47,
      "grad_norm": 5.390098571777344,
      "learning_rate": 0.00011790944643159498,
      "loss": 2.9243,
      "step": 42450
    },
    {
      "epoch": 0.47,
      "grad_norm": 4.957045078277588,
      "learning_rate": 0.0001177290329361949,
      "loss": 2.8823,
      "step": 42500
    },
    {
      "epoch": 0.47,
      "grad_norm": 4.895153522491455,
      "learning_rate": 0.00011754855984210267,
      "loss": 2.9154,
      "step": 42550
    },
    {
      "epoch": 0.47,
      "grad_norm": 5.1375651359558105,
      "learning_rate": 0.00011736802775600453,
      "loss": 2.8391,
      "step": 42600
    },
    {
      "epoch": 0.47,
      "grad_norm": 5.678954601287842,
      "learning_rate": 0.000117187437284785,
      "loss": 3.0804,
      "step": 42650
    },
    {
      "epoch": 0.47,
      "grad_norm": 5.407025337219238,
      "learning_rate": 0.00011700678903552496,
      "loss": 2.8453,
      "step": 42700
    },
    {
      "epoch": 0.47,
      "grad_norm": 4.979137420654297,
      "learning_rate": 0.00011682608361549941,
      "loss": 2.9804,
      "step": 42750
    },
    {
      "epoch": 0.47,
      "grad_norm": 6.545903205871582,
      "learning_rate": 0.00011664532163217561,
      "loss": 2.9162,
      "step": 42800
    },
    {
      "epoch": 0.48,
      "grad_norm": 4.6958513259887695,
      "learning_rate": 0.00011646450369321096,
      "loss": 2.9211,
      "step": 42850
    },
    {
      "epoch": 0.48,
      "grad_norm": 5.4765777587890625,
      "learning_rate": 0.00011628363040645092,
      "loss": 3.0789,
      "step": 42900
    },
    {
      "epoch": 0.48,
      "grad_norm": 4.468578338623047,
      "learning_rate": 0.00011610270237992707,
      "loss": 2.8729,
      "step": 42950
    },
    {
      "epoch": 0.48,
      "grad_norm": 5.818888187408447,
      "learning_rate": 0.00011592172022185495,
      "loss": 3.0,
      "step": 43000
    },
    {
      "epoch": 0.48,
      "grad_norm": 4.613924503326416,
      "learning_rate": 0.00011574068454063209,
      "loss": 2.9263,
      "step": 43050
    },
    {
      "epoch": 0.48,
      "grad_norm": 5.543806076049805,
      "learning_rate": 0.00011555959594483597,
      "loss": 3.0417,
      "step": 43100
    },
    {
      "epoch": 0.48,
      "grad_norm": 6.07559871673584,
      "learning_rate": 0.00011537845504322191,
      "loss": 2.8919,
      "step": 43150
    },
    {
      "epoch": 0.48,
      "grad_norm": 5.199263572692871,
      "learning_rate": 0.00011519726244472111,
      "loss": 2.9026,
      "step": 43200
    },
    {
      "epoch": 0.48,
      "grad_norm": 5.044497966766357,
      "learning_rate": 0.0001150160187584385,
      "loss": 2.9563,
      "step": 43250
    },
    {
      "epoch": 0.48,
      "grad_norm": 4.860003471374512,
      "learning_rate": 0.00011483472459365079,
      "loss": 3.041,
      "step": 43300
    },
    {
      "epoch": 0.48,
      "grad_norm": 5.622586727142334,
      "learning_rate": 0.00011465338055980439,
      "loss": 2.7987,
      "step": 43350
    },
    {
      "epoch": 0.48,
      "grad_norm": 6.760509490966797,
      "learning_rate": 0.00011447198726651332,
      "loss": 2.9494,
      "step": 43400
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.510196685791016,
      "learning_rate": 0.00011429054532355719,
      "loss": 2.9445,
      "step": 43450
    },
    {
      "epoch": 0.48,
      "grad_norm": 5.452268123626709,
      "learning_rate": 0.00011410905534087917,
      "loss": 2.8942,
      "step": 43500
    },
    {
      "epoch": 0.48,
      "grad_norm": 6.183492183685303,
      "learning_rate": 0.00011392751792858391,
      "loss": 3.044,
      "step": 43550
    },
    {
      "epoch": 0.48,
      "grad_norm": 6.01775598526001,
      "learning_rate": 0.00011374593369693555,
      "loss": 2.9628,
      "step": 43600
    },
    {
      "epoch": 0.48,
      "grad_norm": 4.965917110443115,
      "learning_rate": 0.00011356430325635556,
      "loss": 2.8986,
      "step": 43650
    },
    {
      "epoch": 0.48,
      "grad_norm": 4.815042018890381,
      "learning_rate": 0.00011338262721742076,
      "loss": 2.8902,
      "step": 43700
    },
    {
      "epoch": 0.49,
      "grad_norm": 6.838842391967773,
      "learning_rate": 0.0001132009061908613,
      "loss": 2.8702,
      "step": 43750
    },
    {
      "epoch": 0.49,
      "grad_norm": 4.264350891113281,
      "learning_rate": 0.00011301914078755846,
      "loss": 2.9841,
      "step": 43800
    },
    {
      "epoch": 0.49,
      "grad_norm": 5.292375564575195,
      "learning_rate": 0.00011283733161854284,
      "loss": 2.9599,
      "step": 43850
    },
    {
      "epoch": 0.49,
      "grad_norm": 6.021354675292969,
      "learning_rate": 0.00011265547929499205,
      "loss": 2.9701,
      "step": 43900
    },
    {
      "epoch": 0.49,
      "grad_norm": 5.289261817932129,
      "learning_rate": 0.0001124735844282288,
      "loss": 2.9529,
      "step": 43950
    },
    {
      "epoch": 0.49,
      "grad_norm": 5.365589141845703,
      "learning_rate": 0.00011229164762971884,
      "loss": 2.8719,
      "step": 44000
    },
    {
      "epoch": 0.49,
      "grad_norm": 5.0670599937438965,
      "learning_rate": 0.00011210966951106889,
      "loss": 2.901,
      "step": 44050
    },
    {
      "epoch": 0.49,
      "grad_norm": 5.8635573387146,
      "learning_rate": 0.00011193129145555024,
      "loss": 2.8916,
      "step": 44100
    },
    {
      "epoch": 0.49,
      "grad_norm": 5.172703266143799,
      "learning_rate": 0.00011174923332792682,
      "loss": 2.9745,
      "step": 44150
    },
    {
      "epoch": 0.49,
      "grad_norm": 5.451959609985352,
      "learning_rate": 0.000111567135703567,
      "loss": 2.9827,
      "step": 44200
    },
    {
      "epoch": 0.49,
      "grad_norm": 4.874610900878906,
      "learning_rate": 0.00011138499919461816,
      "loss": 2.9749,
      "step": 44250
    },
    {
      "epoch": 0.49,
      "grad_norm": 4.621652603149414,
      "learning_rate": 0.0001112028244133583,
      "loss": 2.8464,
      "step": 44300
    },
    {
      "epoch": 0.49,
      "grad_norm": 5.568783760070801,
      "learning_rate": 0.00011102061197219413,
      "loss": 2.923,
      "step": 44350
    },
    {
      "epoch": 0.49,
      "grad_norm": 6.537819862365723,
      "learning_rate": 0.00011083836248365901,
      "loss": 2.9042,
      "step": 44400
    },
    {
      "epoch": 0.49,
      "grad_norm": 4.626901626586914,
      "learning_rate": 0.00011065607656041075,
      "loss": 2.9843,
      "step": 44450
    },
    {
      "epoch": 0.49,
      "grad_norm": 5.579078197479248,
      "learning_rate": 0.00011047375481522969,
      "loss": 2.9371,
      "step": 44500
    },
    {
      "epoch": 0.49,
      "grad_norm": 5.379307746887207,
      "learning_rate": 0.00011029139786101657,
      "loss": 2.8835,
      "step": 44550
    },
    {
      "epoch": 0.49,
      "grad_norm": 5.236015796661377,
      "learning_rate": 0.00011010900631079049,
      "loss": 2.8995,
      "step": 44600
    },
    {
      "epoch": 0.5,
      "grad_norm": 4.900421619415283,
      "learning_rate": 0.00010992658077768689,
      "loss": 2.9544,
      "step": 44650
    },
    {
      "epoch": 0.5,
      "grad_norm": 4.1276445388793945,
      "learning_rate": 0.0001097441218749554,
      "loss": 2.9357,
      "step": 44700
    },
    {
      "epoch": 0.5,
      "grad_norm": 5.315830707550049,
      "learning_rate": 0.00010956163021595782,
      "loss": 2.9812,
      "step": 44750
    },
    {
      "epoch": 0.5,
      "grad_norm": 4.422897815704346,
      "learning_rate": 0.00010937910641416613,
      "loss": 2.7807,
      "step": 44800
    },
    {
      "epoch": 0.5,
      "grad_norm": 6.341892242431641,
      "learning_rate": 0.0001091965510831603,
      "loss": 3.0267,
      "step": 44850
    },
    {
      "epoch": 0.5,
      "grad_norm": 4.295689105987549,
      "learning_rate": 0.00010901396483662629,
      "loss": 2.9204,
      "step": 44900
    },
    {
      "epoch": 0.5,
      "grad_norm": 5.320792198181152,
      "learning_rate": 0.00010883134828835406,
      "loss": 2.8956,
      "step": 44950
    },
    {
      "epoch": 0.5,
      "grad_norm": 5.325866222381592,
      "learning_rate": 0.00010864870205223534,
      "loss": 2.8959,
      "step": 45000
    },
    {
      "epoch": 0.5,
      "grad_norm": 7.466726303100586,
      "learning_rate": 0.00010846602674226174,
      "loss": 3.0028,
      "step": 45050
    },
    {
      "epoch": 0.5,
      "grad_norm": 5.151381015777588,
      "learning_rate": 0.0001082833229725225,
      "loss": 2.9644,
      "step": 45100
    },
    {
      "epoch": 0.5,
      "grad_norm": 4.178279399871826,
      "learning_rate": 0.00010810059135720268,
      "loss": 3.0683,
      "step": 45150
    },
    {
      "epoch": 0.5,
      "grad_norm": 6.167341709136963,
      "learning_rate": 0.00010791783251058083,
      "loss": 2.8418,
      "step": 45200
    },
    {
      "epoch": 0.5,
      "grad_norm": 4.541141033172607,
      "learning_rate": 0.00010773504704702707,
      "loss": 2.9832,
      "step": 45250
    },
    {
      "epoch": 0.5,
      "grad_norm": 5.082434177398682,
      "learning_rate": 0.00010755223558100099,
      "loss": 2.9448,
      "step": 45300
    },
    {
      "epoch": 0.5,
      "grad_norm": 4.632344722747803,
      "learning_rate": 0.00010736939872704965,
      "loss": 2.8956,
      "step": 45350
    },
    {
      "epoch": 0.5,
      "grad_norm": 4.984410762786865,
      "learning_rate": 0.00010718653709980537,
      "loss": 2.7704,
      "step": 45400
    },
    {
      "epoch": 0.5,
      "grad_norm": 4.388175964355469,
      "learning_rate": 0.00010700365131398384,
      "loss": 2.9454,
      "step": 45450
    },
    {
      "epoch": 0.5,
      "grad_norm": 5.920036792755127,
      "learning_rate": 0.00010682074198438188,
      "loss": 2.7829,
      "step": 45500
    },
    {
      "epoch": 0.51,
      "grad_norm": 7.379148006439209,
      "learning_rate": 0.00010663780972587546,
      "loss": 2.8758,
      "step": 45550
    },
    {
      "epoch": 0.51,
      "grad_norm": 4.970609664916992,
      "learning_rate": 0.00010645485515341773,
      "loss": 2.8542,
      "step": 45600
    },
    {
      "epoch": 0.51,
      "grad_norm": 5.384859561920166,
      "learning_rate": 0.00010627187888203671,
      "loss": 3.0344,
      "step": 45650
    },
    {
      "epoch": 0.51,
      "grad_norm": 6.491629123687744,
      "learning_rate": 0.00010608888152683345,
      "loss": 2.929,
      "step": 45700
    },
    {
      "epoch": 0.51,
      "grad_norm": 4.58522367477417,
      "learning_rate": 0.00010590586370297987,
      "loss": 2.7649,
      "step": 45750
    },
    {
      "epoch": 0.51,
      "grad_norm": 5.452960014343262,
      "learning_rate": 0.0001057228260257167,
      "loss": 2.8287,
      "step": 45800
    },
    {
      "epoch": 0.51,
      "grad_norm": 5.445278167724609,
      "learning_rate": 0.0001055397691103514,
      "loss": 2.9186,
      "step": 45850
    },
    {
      "epoch": 0.51,
      "grad_norm": 4.806256294250488,
      "learning_rate": 0.00010535669357225606,
      "loss": 2.8372,
      "step": 45900
    },
    {
      "epoch": 0.51,
      "grad_norm": 8.205987930297852,
      "learning_rate": 0.00010517360002686542,
      "loss": 2.8849,
      "step": 45950
    },
    {
      "epoch": 0.51,
      "grad_norm": 5.005978107452393,
      "learning_rate": 0.00010499048908967479,
      "loss": 2.9383,
      "step": 46000
    },
    {
      "epoch": 0.51,
      "grad_norm": 5.5991530418396,
      "learning_rate": 0.00010480736137623785,
      "loss": 2.8669,
      "step": 46050
    },
    {
      "epoch": 0.51,
      "grad_norm": 5.811635494232178,
      "learning_rate": 0.00010462788053403845,
      "loss": 2.8846,
      "step": 46100
    },
    {
      "epoch": 0.51,
      "grad_norm": 5.274057388305664,
      "learning_rate": 0.00010444472141985921,
      "loss": 2.8283,
      "step": 46150
    },
    {
      "epoch": 0.51,
      "grad_norm": 5.665740013122559,
      "learning_rate": 0.0001042615473641102,
      "loss": 2.9235,
      "step": 46200
    },
    {
      "epoch": 0.51,
      "grad_norm": 5.276669502258301,
      "learning_rate": 0.00010407835898255722,
      "loss": 2.8764,
      "step": 46250
    },
    {
      "epoch": 0.51,
      "grad_norm": 4.8003716468811035,
      "learning_rate": 0.00010389515689101443,
      "loss": 2.9182,
      "step": 46300
    },
    {
      "epoch": 0.51,
      "grad_norm": 6.221768379211426,
      "learning_rate": 0.00010371194170534193,
      "loss": 2.8737,
      "step": 46350
    },
    {
      "epoch": 0.51,
      "grad_norm": 6.831139087677002,
      "learning_rate": 0.00010352871404144393,
      "loss": 2.9114,
      "step": 46400
    },
    {
      "epoch": 0.52,
      "grad_norm": 4.927224636077881,
      "learning_rate": 0.00010334547451526655,
      "loss": 2.8759,
      "step": 46450
    },
    {
      "epoch": 0.52,
      "grad_norm": 5.206568717956543,
      "learning_rate": 0.00010316222374279579,
      "loss": 2.8074,
      "step": 46500
    },
    {
      "epoch": 0.52,
      "grad_norm": 5.590527057647705,
      "learning_rate": 0.00010297896234005547,
      "loss": 2.9832,
      "step": 46550
    },
    {
      "epoch": 0.52,
      "grad_norm": 5.250126838684082,
      "learning_rate": 0.00010279569092310509,
      "loss": 2.8367,
      "step": 46600
    },
    {
      "epoch": 0.52,
      "grad_norm": 5.71900749206543,
      "learning_rate": 0.0001026124101080379,
      "loss": 2.8001,
      "step": 46650
    },
    {
      "epoch": 0.52,
      "grad_norm": 4.7858428955078125,
      "learning_rate": 0.00010242912051097868,
      "loss": 2.8666,
      "step": 46700
    },
    {
      "epoch": 0.52,
      "grad_norm": 4.844628810882568,
      "learning_rate": 0.00010224582274808179,
      "loss": 2.8775,
      "step": 46750
    },
    {
      "epoch": 0.52,
      "grad_norm": 5.725827217102051,
      "learning_rate": 0.00010206251743552893,
      "loss": 2.9606,
      "step": 46800
    },
    {
      "epoch": 0.52,
      "grad_norm": 4.768713474273682,
      "learning_rate": 0.00010187920518952732,
      "loss": 2.958,
      "step": 46850
    },
    {
      "epoch": 0.52,
      "grad_norm": 3.9493703842163086,
      "learning_rate": 0.00010169588662630737,
      "loss": 2.8679,
      "step": 46900
    },
    {
      "epoch": 0.52,
      "grad_norm": 5.172091484069824,
      "learning_rate": 0.00010151256236212083,
      "loss": 2.782,
      "step": 46950
    },
    {
      "epoch": 0.52,
      "grad_norm": 5.498044490814209,
      "learning_rate": 0.00010132923301323851,
      "loss": 2.9971,
      "step": 47000
    },
    {
      "epoch": 0.52,
      "grad_norm": 5.295326232910156,
      "learning_rate": 0.00010114589919594841,
      "loss": 2.7663,
      "step": 47050
    },
    {
      "epoch": 0.52,
      "grad_norm": 5.579819202423096,
      "learning_rate": 0.00010096256152655347,
      "loss": 2.8452,
      "step": 47100
    },
    {
      "epoch": 0.52,
      "grad_norm": 6.318177223205566,
      "learning_rate": 0.00010077922062136962,
      "loss": 2.8623,
      "step": 47150
    },
    {
      "epoch": 0.52,
      "grad_norm": 5.403071403503418,
      "learning_rate": 0.00010059587709672367,
      "loss": 2.908,
      "step": 47200
    },
    {
      "epoch": 0.52,
      "grad_norm": 4.94517707824707,
      "learning_rate": 0.00010041253156895121,
      "loss": 2.7523,
      "step": 47250
    },
    {
      "epoch": 0.52,
      "grad_norm": 5.644167423248291,
      "learning_rate": 0.00010022918465439459,
      "loss": 2.7601,
      "step": 47300
    },
    {
      "epoch": 0.53,
      "grad_norm": 5.638683795928955,
      "learning_rate": 0.00010004583696940082,
      "loss": 2.8464,
      "step": 47350
    },
    {
      "epoch": 0.53,
      "grad_norm": 6.358886241912842,
      "learning_rate": 9.986248913031951e-05,
      "loss": 2.8521,
      "step": 47400
    },
    {
      "epoch": 0.53,
      "grad_norm": 5.6009674072265625,
      "learning_rate": 9.967914175350072e-05,
      "loss": 2.8016,
      "step": 47450
    },
    {
      "epoch": 0.53,
      "grad_norm": 5.444103717803955,
      "learning_rate": 9.949579545529306e-05,
      "loss": 2.8864,
      "step": 47500
    },
    {
      "epoch": 0.53,
      "grad_norm": 4.8751020431518555,
      "learning_rate": 9.931245085204141e-05,
      "loss": 2.898,
      "step": 47550
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.828577995300293,
      "learning_rate": 9.912910856008507e-05,
      "loss": 2.8747,
      "step": 47600
    },
    {
      "epoch": 0.53,
      "grad_norm": 4.8726420402526855,
      "learning_rate": 9.894576919575544e-05,
      "loss": 2.9534,
      "step": 47650
    },
    {
      "epoch": 0.53,
      "grad_norm": 5.384267807006836,
      "learning_rate": 9.876243337537419e-05,
      "loss": 2.8939,
      "step": 47700
    },
    {
      "epoch": 0.53,
      "grad_norm": 7.15346097946167,
      "learning_rate": 9.857910171525097e-05,
      "loss": 2.837,
      "step": 47750
    },
    {
      "epoch": 0.53,
      "grad_norm": 5.236895561218262,
      "learning_rate": 9.839577483168159e-05,
      "loss": 2.9125,
      "step": 47800
    },
    {
      "epoch": 0.53,
      "grad_norm": 5.851171493530273,
      "learning_rate": 9.821245334094561e-05,
      "loss": 2.8393,
      "step": 47850
    },
    {
      "epoch": 0.53,
      "grad_norm": 7.324557304382324,
      "learning_rate": 9.802913785930468e-05,
      "loss": 2.8172,
      "step": 47900
    },
    {
      "epoch": 0.53,
      "grad_norm": 5.504180908203125,
      "learning_rate": 9.784582900300003e-05,
      "loss": 2.9226,
      "step": 47950
    },
    {
      "epoch": 0.53,
      "grad_norm": 6.177040100097656,
      "learning_rate": 9.766252738825082e-05,
      "loss": 2.9002,
      "step": 48000
    },
    {
      "epoch": 0.53,
      "grad_norm": 5.5708184242248535,
      "learning_rate": 9.74792336312517e-05,
      "loss": 2.944,
      "step": 48050
    },
    {
      "epoch": 0.53,
      "grad_norm": 5.0200958251953125,
      "learning_rate": 9.729594834817101e-05,
      "loss": 2.7737,
      "step": 48100
    },
    {
      "epoch": 0.53,
      "grad_norm": 4.931415557861328,
      "learning_rate": 9.71163375859414e-05,
      "loss": 2.8364,
      "step": 48150
    },
    {
      "epoch": 0.53,
      "grad_norm": 5.3155837059021,
      "learning_rate": 9.693307089892542e-05,
      "loss": 2.847,
      "step": 48200
    },
    {
      "epoch": 0.54,
      "grad_norm": 4.763603687286377,
      "learning_rate": 9.674981452183242e-05,
      "loss": 2.7749,
      "step": 48250
    },
    {
      "epoch": 0.54,
      "grad_norm": 5.666341781616211,
      "learning_rate": 9.656656907070497e-05,
      "loss": 2.8714,
      "step": 48300
    },
    {
      "epoch": 0.54,
      "grad_norm": 5.25208044052124,
      "learning_rate": 9.638333516154907e-05,
      "loss": 2.8369,
      "step": 48350
    },
    {
      "epoch": 0.54,
      "grad_norm": 5.267052173614502,
      "learning_rate": 9.620011341033175e-05,
      "loss": 2.9775,
      "step": 48400
    },
    {
      "epoch": 0.54,
      "grad_norm": 5.161477088928223,
      "learning_rate": 9.601690443297935e-05,
      "loss": 2.7449,
      "step": 48450
    },
    {
      "epoch": 0.54,
      "grad_norm": 4.944921016693115,
      "learning_rate": 9.583370884537509e-05,
      "loss": 2.8609,
      "step": 48500
    },
    {
      "epoch": 0.54,
      "grad_norm": 5.221665859222412,
      "learning_rate": 9.565052726335734e-05,
      "loss": 2.8469,
      "step": 48550
    },
    {
      "epoch": 0.54,
      "grad_norm": 4.500278949737549,
      "learning_rate": 9.546736030271725e-05,
      "loss": 2.8544,
      "step": 48600
    },
    {
      "epoch": 0.54,
      "grad_norm": 4.985889911651611,
      "learning_rate": 9.528420857919697e-05,
      "loss": 2.8036,
      "step": 48650
    },
    {
      "epoch": 0.54,
      "grad_norm": 4.907659530639648,
      "learning_rate": 9.510107270848724e-05,
      "loss": 2.8904,
      "step": 48700
    },
    {
      "epoch": 0.54,
      "grad_norm": 5.073204040527344,
      "learning_rate": 9.491795330622567e-05,
      "loss": 2.6987,
      "step": 48750
    },
    {
      "epoch": 0.54,
      "grad_norm": 4.93405294418335,
      "learning_rate": 9.473485098799443e-05,
      "loss": 2.8249,
      "step": 48800
    },
    {
      "epoch": 0.54,
      "grad_norm": 4.795543670654297,
      "learning_rate": 9.455176636931828e-05,
      "loss": 2.8581,
      "step": 48850
    },
    {
      "epoch": 0.54,
      "grad_norm": 5.811830520629883,
      "learning_rate": 9.436870006566246e-05,
      "loss": 2.906,
      "step": 48900
    },
    {
      "epoch": 0.54,
      "grad_norm": 5.670967102050781,
      "learning_rate": 9.418565269243071e-05,
      "loss": 2.9754,
      "step": 48950
    },
    {
      "epoch": 0.54,
      "grad_norm": 5.656855583190918,
      "learning_rate": 9.400262486496302e-05,
      "loss": 2.7556,
      "step": 49000
    },
    {
      "epoch": 0.54,
      "grad_norm": 7.969732284545898,
      "learning_rate": 9.38196171985338e-05,
      "loss": 2.7671,
      "step": 49050
    },
    {
      "epoch": 0.54,
      "grad_norm": 6.888683795928955,
      "learning_rate": 9.363663030834952e-05,
      "loss": 2.8166,
      "step": 49100
    },
    {
      "epoch": 0.55,
      "grad_norm": 5.012565612792969,
      "learning_rate": 9.345366480954705e-05,
      "loss": 2.8049,
      "step": 49150
    },
    {
      "epoch": 0.55,
      "grad_norm": 3.9672935009002686,
      "learning_rate": 9.327072131719107e-05,
      "loss": 2.7632,
      "step": 49200
    },
    {
      "epoch": 0.55,
      "grad_norm": 4.249286651611328,
      "learning_rate": 9.308780044627255e-05,
      "loss": 2.7411,
      "step": 49250
    },
    {
      "epoch": 0.55,
      "grad_norm": 4.749345302581787,
      "learning_rate": 9.290490281170615e-05,
      "loss": 2.7056,
      "step": 49300
    },
    {
      "epoch": 0.55,
      "grad_norm": 4.181809902191162,
      "learning_rate": 9.272202902832869e-05,
      "loss": 2.924,
      "step": 49350
    },
    {
      "epoch": 0.55,
      "grad_norm": 6.044013977050781,
      "learning_rate": 9.253917971089656e-05,
      "loss": 2.9458,
      "step": 49400
    },
    {
      "epoch": 0.55,
      "grad_norm": 5.671850204467773,
      "learning_rate": 9.235635547408412e-05,
      "loss": 2.8085,
      "step": 49450
    },
    {
      "epoch": 0.55,
      "grad_norm": 6.044562339782715,
      "learning_rate": 9.217355693248123e-05,
      "loss": 2.8216,
      "step": 49500
    },
    {
      "epoch": 0.55,
      "grad_norm": 6.342569828033447,
      "learning_rate": 9.199078470059152e-05,
      "loss": 2.8485,
      "step": 49550
    },
    {
      "epoch": 0.55,
      "grad_norm": 4.90534782409668,
      "learning_rate": 9.180803939283008e-05,
      "loss": 2.8404,
      "step": 49600
    },
    {
      "epoch": 0.55,
      "grad_norm": 4.511479377746582,
      "learning_rate": 9.162532162352155e-05,
      "loss": 2.8728,
      "step": 49650
    },
    {
      "epoch": 0.55,
      "grad_norm": 5.42932653427124,
      "learning_rate": 9.144263200689793e-05,
      "loss": 2.8323,
      "step": 49700
    },
    {
      "epoch": 0.55,
      "grad_norm": 5.7290568351745605,
      "learning_rate": 9.125997115709666e-05,
      "loss": 2.8467,
      "step": 49750
    },
    {
      "epoch": 0.55,
      "grad_norm": 5.0781168937683105,
      "learning_rate": 9.107733968815843e-05,
      "loss": 2.8715,
      "step": 49800
    },
    {
      "epoch": 0.55,
      "grad_norm": 6.262780666351318,
      "learning_rate": 9.089473821402519e-05,
      "loss": 2.8191,
      "step": 49850
    },
    {
      "epoch": 0.55,
      "grad_norm": 4.868679523468018,
      "learning_rate": 9.071216734853796e-05,
      "loss": 2.766,
      "step": 49900
    },
    {
      "epoch": 0.55,
      "grad_norm": 4.828547477722168,
      "learning_rate": 9.052962770543505e-05,
      "loss": 2.7919,
      "step": 49950
    },
    {
      "epoch": 0.55,
      "grad_norm": 5.066901683807373,
      "learning_rate": 9.03471198983496e-05,
      "loss": 2.9057,
      "step": 50000
    },
    {
      "epoch": 0.55,
      "grad_norm": 5.19648551940918,
      "learning_rate": 9.016464454080793e-05,
      "loss": 2.8718,
      "step": 50050
    },
    {
      "epoch": 0.56,
      "grad_norm": 5.266683101654053,
      "learning_rate": 8.998220224622708e-05,
      "loss": 2.8969,
      "step": 50100
    },
    {
      "epoch": 0.56,
      "grad_norm": 4.854067325592041,
      "learning_rate": 8.979979362791311e-05,
      "loss": 2.8398,
      "step": 50150
    },
    {
      "epoch": 0.56,
      "grad_norm": 4.818492889404297,
      "learning_rate": 8.961741929905883e-05,
      "loss": 2.8047,
      "step": 50200
    },
    {
      "epoch": 0.56,
      "grad_norm": 5.926136016845703,
      "learning_rate": 8.943507987274168e-05,
      "loss": 2.8119,
      "step": 50250
    },
    {
      "epoch": 0.56,
      "grad_norm": 5.289876937866211,
      "learning_rate": 8.925277596192192e-05,
      "loss": 2.8051,
      "step": 50300
    },
    {
      "epoch": 0.56,
      "grad_norm": 5.575194835662842,
      "learning_rate": 8.907050817944029e-05,
      "loss": 2.8371,
      "step": 50350
    },
    {
      "epoch": 0.56,
      "grad_norm": 5.982607841491699,
      "learning_rate": 8.888827713801616e-05,
      "loss": 2.7407,
      "step": 50400
    },
    {
      "epoch": 0.56,
      "grad_norm": 4.438473224639893,
      "learning_rate": 8.870608345024536e-05,
      "loss": 2.7715,
      "step": 50450
    },
    {
      "epoch": 0.56,
      "grad_norm": 4.853867530822754,
      "learning_rate": 8.852392772859816e-05,
      "loss": 2.7517,
      "step": 50500
    },
    {
      "epoch": 0.56,
      "grad_norm": 4.544076919555664,
      "learning_rate": 8.834181058541719e-05,
      "loss": 2.8274,
      "step": 50550
    },
    {
      "epoch": 0.56,
      "grad_norm": 5.546023368835449,
      "learning_rate": 8.815973263291542e-05,
      "loss": 2.8477,
      "step": 50600
    },
    {
      "epoch": 0.56,
      "grad_norm": 6.0551934242248535,
      "learning_rate": 8.7977694483174e-05,
      "loss": 2.8963,
      "step": 50650
    },
    {
      "epoch": 0.56,
      "grad_norm": 4.458075046539307,
      "learning_rate": 8.77956967481404e-05,
      "loss": 2.8206,
      "step": 50700
    },
    {
      "epoch": 0.56,
      "grad_norm": 4.1557111740112305,
      "learning_rate": 8.761374003962613e-05,
      "loss": 2.8098,
      "step": 50750
    },
    {
      "epoch": 0.56,
      "grad_norm": 5.024372100830078,
      "learning_rate": 8.743182496930482e-05,
      "loss": 2.8212,
      "step": 50800
    },
    {
      "epoch": 0.56,
      "grad_norm": 4.921896934509277,
      "learning_rate": 8.72499521487101e-05,
      "loss": 2.8694,
      "step": 50850
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.456253051757812,
      "learning_rate": 8.706812218923365e-05,
      "loss": 2.8235,
      "step": 50900
    },
    {
      "epoch": 0.56,
      "grad_norm": 4.885873794555664,
      "learning_rate": 8.688633570212294e-05,
      "loss": 2.7026,
      "step": 50950
    },
    {
      "epoch": 0.57,
      "grad_norm": 5.612597942352295,
      "learning_rate": 8.670459329847947e-05,
      "loss": 2.8504,
      "step": 51000
    },
    {
      "epoch": 0.57,
      "grad_norm": 5.247095584869385,
      "learning_rate": 8.652289558925632e-05,
      "loss": 2.7814,
      "step": 51050
    },
    {
      "epoch": 0.57,
      "grad_norm": 4.934078693389893,
      "learning_rate": 8.634124318525656e-05,
      "loss": 2.9126,
      "step": 51100
    },
    {
      "epoch": 0.57,
      "grad_norm": 5.127668857574463,
      "learning_rate": 8.615963669713078e-05,
      "loss": 2.7345,
      "step": 51150
    },
    {
      "epoch": 0.57,
      "grad_norm": 4.752030372619629,
      "learning_rate": 8.597807673537537e-05,
      "loss": 2.8596,
      "step": 51200
    },
    {
      "epoch": 0.57,
      "grad_norm": 5.056651592254639,
      "learning_rate": 8.579656391033014e-05,
      "loss": 2.8417,
      "step": 51250
    },
    {
      "epoch": 0.57,
      "grad_norm": 5.137719631195068,
      "learning_rate": 8.561509883217663e-05,
      "loss": 2.7885,
      "step": 51300
    },
    {
      "epoch": 0.57,
      "grad_norm": 4.433664321899414,
      "learning_rate": 8.543368211093571e-05,
      "loss": 2.9204,
      "step": 51350
    },
    {
      "epoch": 0.57,
      "grad_norm": 4.627000331878662,
      "learning_rate": 8.525231435646577e-05,
      "loss": 2.8252,
      "step": 51400
    },
    {
      "epoch": 0.57,
      "grad_norm": 5.376331329345703,
      "learning_rate": 8.507099617846061e-05,
      "loss": 2.7879,
      "step": 51450
    },
    {
      "epoch": 0.57,
      "grad_norm": 4.6834259033203125,
      "learning_rate": 8.488972818644737e-05,
      "loss": 2.9127,
      "step": 51500
    },
    {
      "epoch": 0.57,
      "grad_norm": 4.654898166656494,
      "learning_rate": 8.470851098978436e-05,
      "loss": 2.9049,
      "step": 51550
    },
    {
      "epoch": 0.57,
      "grad_norm": 5.21427583694458,
      "learning_rate": 8.452734519765933e-05,
      "loss": 2.7647,
      "step": 51600
    },
    {
      "epoch": 0.57,
      "grad_norm": 5.709262371063232,
      "learning_rate": 8.434623141908703e-05,
      "loss": 2.847,
      "step": 51650
    },
    {
      "epoch": 0.57,
      "grad_norm": 4.660146713256836,
      "learning_rate": 8.416517026290756e-05,
      "loss": 2.761,
      "step": 51700
    },
    {
      "epoch": 0.57,
      "grad_norm": 5.906465530395508,
      "learning_rate": 8.398416233778394e-05,
      "loss": 2.7357,
      "step": 51750
    },
    {
      "epoch": 0.57,
      "grad_norm": 5.650059223175049,
      "learning_rate": 8.380320825220037e-05,
      "loss": 2.7828,
      "step": 51800
    },
    {
      "epoch": 0.57,
      "grad_norm": 6.86538028717041,
      "learning_rate": 8.362230861445996e-05,
      "loss": 2.7406,
      "step": 51850
    },
    {
      "epoch": 0.58,
      "grad_norm": 5.082160949707031,
      "learning_rate": 8.344146403268294e-05,
      "loss": 2.8243,
      "step": 51900
    },
    {
      "epoch": 0.58,
      "grad_norm": 4.921591758728027,
      "learning_rate": 8.326067511480423e-05,
      "loss": 2.7735,
      "step": 51950
    },
    {
      "epoch": 0.58,
      "grad_norm": 7.034480094909668,
      "learning_rate": 8.307994246857189e-05,
      "loss": 2.737,
      "step": 52000
    },
    {
      "epoch": 0.58,
      "grad_norm": 4.324589729309082,
      "learning_rate": 8.289926670154456e-05,
      "loss": 2.8362,
      "step": 52050
    },
    {
      "epoch": 0.58,
      "grad_norm": 5.1923370361328125,
      "learning_rate": 8.271864842108992e-05,
      "loss": 2.7732,
      "step": 52100
    },
    {
      "epoch": 0.58,
      "grad_norm": 4.978728771209717,
      "learning_rate": 8.253808823438218e-05,
      "loss": 2.8185,
      "step": 52150
    },
    {
      "epoch": 0.58,
      "grad_norm": 5.956799507141113,
      "learning_rate": 8.235758674840042e-05,
      "loss": 2.8004,
      "step": 52200
    },
    {
      "epoch": 0.58,
      "grad_norm": 5.126569747924805,
      "learning_rate": 8.217714456992628e-05,
      "loss": 2.7992,
      "step": 52250
    },
    {
      "epoch": 0.58,
      "grad_norm": 4.2657599449157715,
      "learning_rate": 8.19967623055421e-05,
      "loss": 2.7826,
      "step": 52300
    },
    {
      "epoch": 0.58,
      "grad_norm": 5.124692440032959,
      "learning_rate": 8.181644056162877e-05,
      "loss": 2.8173,
      "step": 52350
    },
    {
      "epoch": 0.58,
      "grad_norm": 4.610875129699707,
      "learning_rate": 8.163617994436379e-05,
      "loss": 2.7106,
      "step": 52400
    },
    {
      "epoch": 0.58,
      "grad_norm": 5.901200294494629,
      "learning_rate": 8.145598105971906e-05,
      "loss": 2.8315,
      "step": 52450
    },
    {
      "epoch": 0.58,
      "grad_norm": 4.964183330535889,
      "learning_rate": 8.12758445134591e-05,
      "loss": 2.7531,
      "step": 52500
    },
    {
      "epoch": 0.58,
      "grad_norm": 4.925256729125977,
      "learning_rate": 8.109577091113874e-05,
      "loss": 2.7022,
      "step": 52550
    },
    {
      "epoch": 0.58,
      "grad_norm": 4.919853687286377,
      "learning_rate": 8.091576085810133e-05,
      "loss": 2.8289,
      "step": 52600
    },
    {
      "epoch": 0.58,
      "grad_norm": 4.8153300285339355,
      "learning_rate": 8.073581495947647e-05,
      "loss": 2.8629,
      "step": 52650
    },
    {
      "epoch": 0.58,
      "grad_norm": 4.402426719665527,
      "learning_rate": 8.055593382017822e-05,
      "loss": 2.7992,
      "step": 52700
    },
    {
      "epoch": 0.58,
      "grad_norm": 5.2513933181762695,
      "learning_rate": 8.037611804490281e-05,
      "loss": 2.7252,
      "step": 52750
    },
    {
      "epoch": 0.59,
      "grad_norm": 5.595922470092773,
      "learning_rate": 8.019636823812692e-05,
      "loss": 2.8512,
      "step": 52800
    },
    {
      "epoch": 0.59,
      "grad_norm": 5.7623982429504395,
      "learning_rate": 8.001668500410524e-05,
      "loss": 2.8289,
      "step": 52850
    },
    {
      "epoch": 0.59,
      "grad_norm": 4.7036004066467285,
      "learning_rate": 7.983706894686887e-05,
      "loss": 2.7803,
      "step": 52900
    },
    {
      "epoch": 0.59,
      "grad_norm": 4.882056713104248,
      "learning_rate": 7.965752067022295e-05,
      "loss": 2.8064,
      "step": 52950
    },
    {
      "epoch": 0.59,
      "grad_norm": 7.131617546081543,
      "learning_rate": 7.947804077774486e-05,
      "loss": 2.6916,
      "step": 53000
    },
    {
      "epoch": 0.59,
      "grad_norm": 5.684287071228027,
      "learning_rate": 7.929862987278198e-05,
      "loss": 2.7387,
      "step": 53050
    },
    {
      "epoch": 0.59,
      "grad_norm": 5.956545829772949,
      "learning_rate": 7.911928855844995e-05,
      "loss": 2.8002,
      "step": 53100
    },
    {
      "epoch": 0.59,
      "grad_norm": 4.755430698394775,
      "learning_rate": 7.894001743763029e-05,
      "loss": 2.6729,
      "step": 53150
    },
    {
      "epoch": 0.59,
      "grad_norm": 6.940397262573242,
      "learning_rate": 7.876081711296868e-05,
      "loss": 2.8161,
      "step": 53200
    },
    {
      "epoch": 0.59,
      "grad_norm": 5.435543537139893,
      "learning_rate": 7.858168818687275e-05,
      "loss": 2.7731,
      "step": 53250
    },
    {
      "epoch": 0.59,
      "grad_norm": 5.857146263122559,
      "learning_rate": 7.840263126151012e-05,
      "loss": 2.7879,
      "step": 53300
    },
    {
      "epoch": 0.59,
      "grad_norm": 6.037492752075195,
      "learning_rate": 7.82236469388064e-05,
      "loss": 2.6745,
      "step": 53350
    },
    {
      "epoch": 0.59,
      "grad_norm": 6.296214580535889,
      "learning_rate": 7.804473582044311e-05,
      "loss": 2.8119,
      "step": 53400
    },
    {
      "epoch": 0.59,
      "grad_norm": 6.708181858062744,
      "learning_rate": 7.786589850785566e-05,
      "loss": 2.7203,
      "step": 53450
    },
    {
      "epoch": 0.59,
      "grad_norm": 4.357819080352783,
      "learning_rate": 7.768713560223142e-05,
      "loss": 2.7626,
      "step": 53500
    },
    {
      "epoch": 0.59,
      "grad_norm": 6.6953020095825195,
      "learning_rate": 7.750844770450753e-05,
      "loss": 2.6817,
      "step": 53550
    },
    {
      "epoch": 0.59,
      "grad_norm": 5.347495079040527,
      "learning_rate": 7.732983541536912e-05,
      "loss": 2.8281,
      "step": 53600
    },
    {
      "epoch": 0.59,
      "grad_norm": 6.330661773681641,
      "learning_rate": 7.715129933524697e-05,
      "loss": 2.8623,
      "step": 53650
    },
    {
      "epoch": 0.6,
      "grad_norm": 5.908714294433594,
      "learning_rate": 7.697284006431589e-05,
      "loss": 2.7086,
      "step": 53700
    },
    {
      "epoch": 0.6,
      "grad_norm": 5.155627727508545,
      "learning_rate": 7.679445820249223e-05,
      "loss": 2.7711,
      "step": 53750
    },
    {
      "epoch": 0.6,
      "grad_norm": 5.00850248336792,
      "learning_rate": 7.661615434943239e-05,
      "loss": 2.7969,
      "step": 53800
    },
    {
      "epoch": 0.6,
      "grad_norm": 5.916557312011719,
      "learning_rate": 7.643792910453029e-05,
      "loss": 2.7986,
      "step": 53850
    },
    {
      "epoch": 0.6,
      "grad_norm": 4.54706335067749,
      "learning_rate": 7.62597830669158e-05,
      "loss": 2.6953,
      "step": 53900
    },
    {
      "epoch": 0.6,
      "grad_norm": 5.8201904296875,
      "learning_rate": 7.60817168354523e-05,
      "loss": 2.7541,
      "step": 53950
    },
    {
      "epoch": 0.6,
      "grad_norm": 5.3225417137146,
      "learning_rate": 7.590373100873517e-05,
      "loss": 2.7687,
      "step": 54000
    },
    {
      "epoch": 0.6,
      "grad_norm": 5.088871002197266,
      "learning_rate": 7.572582618508924e-05,
      "loss": 2.7016,
      "step": 54050
    },
    {
      "epoch": 0.6,
      "grad_norm": 5.487249851226807,
      "learning_rate": 7.554800296256717e-05,
      "loss": 2.7251,
      "step": 54100
    },
    {
      "epoch": 0.6,
      "grad_norm": 6.182318210601807,
      "learning_rate": 7.537026193894729e-05,
      "loss": 2.8398,
      "step": 54150
    },
    {
      "epoch": 0.6,
      "grad_norm": 4.5475850105285645,
      "learning_rate": 7.519260371173157e-05,
      "loss": 2.8044,
      "step": 54200
    },
    {
      "epoch": 0.6,
      "grad_norm": 5.055111885070801,
      "learning_rate": 7.50150288781437e-05,
      "loss": 2.8092,
      "step": 54250
    },
    {
      "epoch": 0.6,
      "grad_norm": 6.259912014007568,
      "learning_rate": 7.4837538035127e-05,
      "loss": 2.7299,
      "step": 54300
    },
    {
      "epoch": 0.6,
      "grad_norm": 5.038516998291016,
      "learning_rate": 7.466013177934242e-05,
      "loss": 2.7027,
      "step": 54350
    },
    {
      "epoch": 0.6,
      "grad_norm": 4.430378437042236,
      "learning_rate": 7.448281070716664e-05,
      "loss": 2.7285,
      "step": 54400
    },
    {
      "epoch": 0.6,
      "grad_norm": 5.217715263366699,
      "learning_rate": 7.430557541468986e-05,
      "loss": 2.6564,
      "step": 54450
    },
    {
      "epoch": 0.6,
      "grad_norm": 4.507577419281006,
      "learning_rate": 7.412842649771404e-05,
      "loss": 2.6815,
      "step": 54500
    },
    {
      "epoch": 0.6,
      "grad_norm": 4.967469692230225,
      "learning_rate": 7.395136455175067e-05,
      "loss": 2.8376,
      "step": 54550
    },
    {
      "epoch": 0.61,
      "grad_norm": 6.287683486938477,
      "learning_rate": 7.377439017201902e-05,
      "loss": 2.7887,
      "step": 54600
    },
    {
      "epoch": 0.61,
      "grad_norm": 5.037525653839111,
      "learning_rate": 7.359750395344377e-05,
      "loss": 2.7634,
      "step": 54650
    },
    {
      "epoch": 0.61,
      "grad_norm": 6.68649959564209,
      "learning_rate": 7.342070649065351e-05,
      "loss": 2.7991,
      "step": 54700
    },
    {
      "epoch": 0.61,
      "grad_norm": 4.452758312225342,
      "learning_rate": 7.324399837797818e-05,
      "loss": 2.7953,
      "step": 54750
    },
    {
      "epoch": 0.61,
      "grad_norm": 6.111091613769531,
      "learning_rate": 7.306738020944762e-05,
      "loss": 2.7969,
      "step": 54800
    },
    {
      "epoch": 0.61,
      "grad_norm": 4.303292274475098,
      "learning_rate": 7.289085257878909e-05,
      "loss": 2.7704,
      "step": 54850
    },
    {
      "epoch": 0.61,
      "grad_norm": 5.180151462554932,
      "learning_rate": 7.271441607942567e-05,
      "loss": 2.6844,
      "step": 54900
    },
    {
      "epoch": 0.61,
      "grad_norm": 5.898430824279785,
      "learning_rate": 7.253807130447394e-05,
      "loss": 2.7816,
      "step": 54950
    },
    {
      "epoch": 0.61,
      "grad_norm": 5.069900989532471,
      "learning_rate": 7.236181884674226e-05,
      "loss": 2.7799,
      "step": 55000
    },
    {
      "epoch": 0.61,
      "grad_norm": 5.311367034912109,
      "learning_rate": 7.218565929872855e-05,
      "loss": 2.7918,
      "step": 55050
    },
    {
      "epoch": 0.61,
      "grad_norm": 5.394289016723633,
      "learning_rate": 7.200959325261849e-05,
      "loss": 2.719,
      "step": 55100
    },
    {
      "epoch": 0.61,
      "grad_norm": 4.599655628204346,
      "learning_rate": 7.183362130028334e-05,
      "loss": 2.6619,
      "step": 55150
    },
    {
      "epoch": 0.61,
      "grad_norm": 4.696221351623535,
      "learning_rate": 7.165774403327817e-05,
      "loss": 2.7063,
      "step": 55200
    },
    {
      "epoch": 0.61,
      "grad_norm": 5.125298023223877,
      "learning_rate": 7.148196204283968e-05,
      "loss": 2.7743,
      "step": 55250
    },
    {
      "epoch": 0.61,
      "grad_norm": 5.986608028411865,
      "learning_rate": 7.130627591988424e-05,
      "loss": 2.7537,
      "step": 55300
    },
    {
      "epoch": 0.61,
      "grad_norm": 6.218910217285156,
      "learning_rate": 7.113068625500607e-05,
      "loss": 2.6841,
      "step": 55350
    },
    {
      "epoch": 0.61,
      "grad_norm": 5.147339820861816,
      "learning_rate": 7.0955193638475e-05,
      "loss": 2.7363,
      "step": 55400
    },
    {
      "epoch": 0.61,
      "grad_norm": 4.607673645019531,
      "learning_rate": 7.077979866023477e-05,
      "loss": 2.8394,
      "step": 55450
    },
    {
      "epoch": 0.62,
      "grad_norm": 3.8307926654815674,
      "learning_rate": 7.060450190990069e-05,
      "loss": 2.8604,
      "step": 55500
    },
    {
      "epoch": 0.62,
      "grad_norm": 4.884772777557373,
      "learning_rate": 7.042930397675808e-05,
      "loss": 2.7944,
      "step": 55550
    },
    {
      "epoch": 0.62,
      "grad_norm": 6.111611366271973,
      "learning_rate": 7.02542054497599e-05,
      "loss": 2.7035,
      "step": 55600
    },
    {
      "epoch": 0.62,
      "grad_norm": 5.194789409637451,
      "learning_rate": 7.007920691752507e-05,
      "loss": 2.7799,
      "step": 55650
    },
    {
      "epoch": 0.62,
      "grad_norm": 5.869514465332031,
      "learning_rate": 6.990430896833628e-05,
      "loss": 2.8049,
      "step": 55700
    },
    {
      "epoch": 0.62,
      "grad_norm": 6.182733535766602,
      "learning_rate": 6.972951219013812e-05,
      "loss": 2.6971,
      "step": 55750
    },
    {
      "epoch": 0.62,
      "grad_norm": 4.943215847015381,
      "learning_rate": 6.955481717053507e-05,
      "loss": 2.8311,
      "step": 55800
    },
    {
      "epoch": 0.62,
      "grad_norm": 5.8995585441589355,
      "learning_rate": 6.938022449678956e-05,
      "loss": 2.8054,
      "step": 55850
    },
    {
      "epoch": 0.62,
      "grad_norm": 4.764339923858643,
      "learning_rate": 6.920573475581995e-05,
      "loss": 2.8206,
      "step": 55900
    },
    {
      "epoch": 0.62,
      "grad_norm": 5.303322792053223,
      "learning_rate": 6.90313485341986e-05,
      "loss": 2.7161,
      "step": 55950
    },
    {
      "epoch": 0.62,
      "grad_norm": 6.062620162963867,
      "learning_rate": 6.88570664181498e-05,
      "loss": 2.7988,
      "step": 56000
    },
    {
      "epoch": 0.62,
      "grad_norm": 4.624374866485596,
      "learning_rate": 6.868288899354802e-05,
      "loss": 2.777,
      "step": 56050
    },
    {
      "epoch": 0.62,
      "grad_norm": 5.287914752960205,
      "learning_rate": 6.85088168459156e-05,
      "loss": 2.7806,
      "step": 56100
    },
    {
      "epoch": 0.62,
      "grad_norm": 4.46606969833374,
      "learning_rate": 6.833485056042118e-05,
      "loss": 2.7335,
      "step": 56150
    },
    {
      "epoch": 0.62,
      "grad_norm": 4.997598648071289,
      "learning_rate": 6.816099072187735e-05,
      "loss": 2.7451,
      "step": 56200
    },
    {
      "epoch": 0.62,
      "grad_norm": 5.153663158416748,
      "learning_rate": 6.798723791473903e-05,
      "loss": 2.7516,
      "step": 56250
    },
    {
      "epoch": 0.62,
      "grad_norm": 5.8853278160095215,
      "learning_rate": 6.781359272310113e-05,
      "loss": 2.7484,
      "step": 56300
    },
    {
      "epoch": 0.62,
      "grad_norm": 5.8589372634887695,
      "learning_rate": 6.764005573069704e-05,
      "loss": 2.7018,
      "step": 56350
    },
    {
      "epoch": 0.63,
      "grad_norm": 4.759388446807861,
      "learning_rate": 6.746662752089623e-05,
      "loss": 2.7278,
      "step": 56400
    },
    {
      "epoch": 0.63,
      "grad_norm": 6.161319255828857,
      "learning_rate": 6.729330867670259e-05,
      "loss": 2.7544,
      "step": 56450
    },
    {
      "epoch": 0.63,
      "grad_norm": 4.967267036437988,
      "learning_rate": 6.712009978075227e-05,
      "loss": 2.8378,
      "step": 56500
    },
    {
      "epoch": 0.63,
      "grad_norm": 5.112069129943848,
      "learning_rate": 6.694700141531194e-05,
      "loss": 2.7669,
      "step": 56550
    },
    {
      "epoch": 0.63,
      "grad_norm": 5.643813610076904,
      "learning_rate": 6.677401416227656e-05,
      "loss": 2.6907,
      "step": 56600
    },
    {
      "epoch": 0.63,
      "grad_norm": 4.927740097045898,
      "learning_rate": 6.66011386031677e-05,
      "loss": 2.7377,
      "step": 56650
    },
    {
      "epoch": 0.63,
      "grad_norm": 4.542614459991455,
      "learning_rate": 6.642837531913136e-05,
      "loss": 2.6761,
      "step": 56700
    },
    {
      "epoch": 0.63,
      "grad_norm": 5.79432487487793,
      "learning_rate": 6.625572489093617e-05,
      "loss": 2.7382,
      "step": 56750
    },
    {
      "epoch": 0.63,
      "grad_norm": 5.034465312957764,
      "learning_rate": 6.608318789897134e-05,
      "loss": 2.6606,
      "step": 56800
    },
    {
      "epoch": 0.63,
      "grad_norm": 4.9932732582092285,
      "learning_rate": 6.591076492324478e-05,
      "loss": 2.6977,
      "step": 56850
    },
    {
      "epoch": 0.63,
      "grad_norm": 4.95469856262207,
      "learning_rate": 6.573845654338106e-05,
      "loss": 2.71,
      "step": 56900
    },
    {
      "epoch": 0.63,
      "grad_norm": 5.425679683685303,
      "learning_rate": 6.556626333861964e-05,
      "loss": 2.7526,
      "step": 56950
    },
    {
      "epoch": 0.63,
      "grad_norm": 5.906972885131836,
      "learning_rate": 6.539418588781265e-05,
      "loss": 2.7878,
      "step": 57000
    },
    {
      "epoch": 0.63,
      "grad_norm": 5.378124237060547,
      "learning_rate": 6.522222476942324e-05,
      "loss": 2.6774,
      "step": 57050
    },
    {
      "epoch": 0.63,
      "grad_norm": 5.601273059844971,
      "learning_rate": 6.505038056152333e-05,
      "loss": 2.779,
      "step": 57100
    },
    {
      "epoch": 0.63,
      "grad_norm": 4.701582431793213,
      "learning_rate": 6.487865384179204e-05,
      "loss": 2.778,
      "step": 57150
    },
    {
      "epoch": 0.63,
      "grad_norm": 4.828241348266602,
      "learning_rate": 6.470704518751329e-05,
      "loss": 2.7221,
      "step": 57200
    },
    {
      "epoch": 0.63,
      "grad_norm": 6.58074951171875,
      "learning_rate": 6.453555517557437e-05,
      "loss": 2.6962,
      "step": 57250
    },
    {
      "epoch": 0.64,
      "grad_norm": 5.19921350479126,
      "learning_rate": 6.436418438246346e-05,
      "loss": 2.6619,
      "step": 57300
    },
    {
      "epoch": 0.64,
      "grad_norm": 7.129755020141602,
      "learning_rate": 6.419293338426824e-05,
      "loss": 2.7651,
      "step": 57350
    },
    {
      "epoch": 0.64,
      "grad_norm": 5.713471412658691,
      "learning_rate": 6.402180275667345e-05,
      "loss": 2.7451,
      "step": 57400
    },
    {
      "epoch": 0.64,
      "grad_norm": 5.965299606323242,
      "learning_rate": 6.385079307495938e-05,
      "loss": 2.6373,
      "step": 57450
    },
    {
      "epoch": 0.64,
      "grad_norm": 5.371218204498291,
      "learning_rate": 6.367990491399954e-05,
      "loss": 2.8015,
      "step": 57500
    },
    {
      "epoch": 0.64,
      "grad_norm": 6.4079365730285645,
      "learning_rate": 6.350913884825915e-05,
      "loss": 2.8006,
      "step": 57550
    },
    {
      "epoch": 0.64,
      "grad_norm": 5.489929676055908,
      "learning_rate": 6.333849545179283e-05,
      "loss": 2.6912,
      "step": 57600
    },
    {
      "epoch": 0.64,
      "grad_norm": 5.011102199554443,
      "learning_rate": 6.316797529824288e-05,
      "loss": 2.744,
      "step": 57650
    },
    {
      "epoch": 0.64,
      "grad_norm": 5.367973327636719,
      "learning_rate": 6.299757896083733e-05,
      "loss": 2.7401,
      "step": 57700
    },
    {
      "epoch": 0.64,
      "grad_norm": 5.436025619506836,
      "learning_rate": 6.282730701238794e-05,
      "loss": 2.7494,
      "step": 57750
    },
    {
      "epoch": 0.64,
      "grad_norm": 5.955270767211914,
      "learning_rate": 6.26571600252883e-05,
      "loss": 2.7513,
      "step": 57800
    },
    {
      "epoch": 0.64,
      "grad_norm": 5.521211624145508,
      "learning_rate": 6.248713857151203e-05,
      "loss": 2.6837,
      "step": 57850
    },
    {
      "epoch": 0.64,
      "grad_norm": 6.505960941314697,
      "learning_rate": 6.23172432226106e-05,
      "loss": 2.6164,
      "step": 57900
    },
    {
      "epoch": 0.64,
      "grad_norm": 4.359035968780518,
      "learning_rate": 6.214747454971174e-05,
      "loss": 2.7221,
      "step": 57950
    },
    {
      "epoch": 0.64,
      "grad_norm": 5.607531547546387,
      "learning_rate": 6.197783312351713e-05,
      "loss": 2.6201,
      "step": 58000
    },
    {
      "epoch": 0.64,
      "grad_norm": 3.8494324684143066,
      "learning_rate": 6.181170853019237e-05,
      "loss": 2.7321,
      "step": 58050
    },
    {
      "epoch": 0.64,
      "grad_norm": 5.898524761199951,
      "learning_rate": 6.164232073448073e-05,
      "loss": 2.598,
      "step": 58100
    },
    {
      "epoch": 0.64,
      "grad_norm": 5.539312362670898,
      "learning_rate": 6.14730618836205e-05,
      "loss": 2.7442,
      "step": 58150
    },
    {
      "epoch": 0.65,
      "grad_norm": 4.616910457611084,
      "learning_rate": 6.130393254659972e-05,
      "loss": 2.8174,
      "step": 58200
    },
    {
      "epoch": 0.65,
      "grad_norm": 5.510334491729736,
      "learning_rate": 6.113493329197089e-05,
      "loss": 2.7456,
      "step": 58250
    },
    {
      "epoch": 0.65,
      "grad_norm": 4.171205520629883,
      "learning_rate": 6.096944077588434e-05,
      "loss": 2.7062,
      "step": 58300
    },
    {
      "epoch": 0.65,
      "grad_norm": 3.8376502990722656,
      "learning_rate": 6.080070076002211e-05,
      "loss": 2.7123,
      "step": 58350
    },
    {
      "epoch": 0.65,
      "grad_norm": 5.743969917297363,
      "learning_rate": 6.0632092518237806e-05,
      "loss": 2.621,
      "step": 58400
    },
    {
      "epoch": 0.65,
      "grad_norm": 6.348492622375488,
      "learning_rate": 6.046361661733233e-05,
      "loss": 2.6932,
      "step": 58450
    },
    {
      "epoch": 0.65,
      "grad_norm": 5.0930495262146,
      "learning_rate": 6.029527362366156e-05,
      "loss": 2.673,
      "step": 58500
    },
    {
      "epoch": 0.65,
      "grad_norm": 4.009668827056885,
      "learning_rate": 6.0127064103134736e-05,
      "loss": 2.6655,
      "step": 58550
    },
    {
      "epoch": 0.65,
      "grad_norm": 5.569638729095459,
      "learning_rate": 5.9958988621212276e-05,
      "loss": 2.721,
      "step": 58600
    },
    {
      "epoch": 0.65,
      "grad_norm": 7.03887939453125,
      "learning_rate": 5.979104774290413e-05,
      "loss": 2.7568,
      "step": 58650
    },
    {
      "epoch": 0.65,
      "grad_norm": 4.858248233795166,
      "learning_rate": 5.9623242032767615e-05,
      "loss": 2.5367,
      "step": 58700
    },
    {
      "epoch": 0.65,
      "grad_norm": 6.241574287414551,
      "learning_rate": 5.945557205490584e-05,
      "loss": 2.6444,
      "step": 58750
    },
    {
      "epoch": 0.65,
      "grad_norm": 6.078144550323486,
      "learning_rate": 5.9288038372965435e-05,
      "loss": 2.7933,
      "step": 58800
    },
    {
      "epoch": 0.65,
      "grad_norm": 5.697675704956055,
      "learning_rate": 5.912064155013503e-05,
      "loss": 2.7129,
      "step": 58850
    },
    {
      "epoch": 0.65,
      "grad_norm": 5.177536964416504,
      "learning_rate": 5.8953382149143055e-05,
      "loss": 2.6881,
      "step": 58900
    },
    {
      "epoch": 0.65,
      "grad_norm": 5.119016647338867,
      "learning_rate": 5.878626073225601e-05,
      "loss": 2.7928,
      "step": 58950
    },
    {
      "epoch": 0.65,
      "grad_norm": 5.4692888259887695,
      "learning_rate": 5.861927786127661e-05,
      "loss": 2.6978,
      "step": 59000
    },
    {
      "epoch": 0.65,
      "grad_norm": 6.175304412841797,
      "learning_rate": 5.845243409754174e-05,
      "loss": 2.6446,
      "step": 59050
    },
    {
      "epoch": 0.66,
      "grad_norm": 5.197447299957275,
      "learning_rate": 5.8285730001920704e-05,
      "loss": 2.6267,
      "step": 59100
    },
    {
      "epoch": 0.66,
      "grad_norm": 4.362741947174072,
      "learning_rate": 5.8119166134813253e-05,
      "loss": 2.7157,
      "step": 59150
    },
    {
      "epoch": 0.66,
      "grad_norm": 5.375911235809326,
      "learning_rate": 5.7952743056147784e-05,
      "loss": 2.6706,
      "step": 59200
    },
    {
      "epoch": 0.66,
      "grad_norm": 4.943705081939697,
      "learning_rate": 5.778646132537939e-05,
      "loss": 2.65,
      "step": 59250
    },
    {
      "epoch": 0.66,
      "grad_norm": 5.501734256744385,
      "learning_rate": 5.762032150148802e-05,
      "loss": 2.7627,
      "step": 59300
    },
    {
      "epoch": 0.66,
      "grad_norm": 5.921018600463867,
      "learning_rate": 5.7454324142976534e-05,
      "loss": 2.7103,
      "step": 59350
    },
    {
      "epoch": 0.66,
      "grad_norm": 4.590132236480713,
      "learning_rate": 5.7288469807868936e-05,
      "loss": 2.6617,
      "step": 59400
    },
    {
      "epoch": 0.66,
      "grad_norm": 4.6414055824279785,
      "learning_rate": 5.712275905370841e-05,
      "loss": 2.6071,
      "step": 59450
    },
    {
      "epoch": 0.66,
      "grad_norm": 7.030457973480225,
      "learning_rate": 5.6957192437555464e-05,
      "loss": 2.7209,
      "step": 59500
    },
    {
      "epoch": 0.66,
      "grad_norm": 5.596695899963379,
      "learning_rate": 5.679177051598611e-05,
      "loss": 2.774,
      "step": 59550
    },
    {
      "epoch": 0.66,
      "grad_norm": 5.1710638999938965,
      "learning_rate": 5.66264938450898e-05,
      "loss": 2.686,
      "step": 59600
    },
    {
      "epoch": 0.66,
      "grad_norm": 5.006558418273926,
      "learning_rate": 5.6461362980467956e-05,
      "loss": 2.7063,
      "step": 59650
    },
    {
      "epoch": 0.66,
      "grad_norm": 4.40914249420166,
      "learning_rate": 5.629637847723158e-05,
      "loss": 2.6988,
      "step": 59700
    },
    {
      "epoch": 0.66,
      "grad_norm": 4.632761001586914,
      "learning_rate": 5.6131540889999946e-05,
      "loss": 2.6615,
      "step": 59750
    },
    {
      "epoch": 0.66,
      "grad_norm": 5.840973854064941,
      "learning_rate": 5.596685077289811e-05,
      "loss": 2.7795,
      "step": 59800
    },
    {
      "epoch": 0.66,
      "grad_norm": 4.624524116516113,
      "learning_rate": 5.580230867955574e-05,
      "loss": 2.6971,
      "step": 59850
    },
    {
      "epoch": 0.66,
      "grad_norm": 4.625621795654297,
      "learning_rate": 5.563791516310458e-05,
      "loss": 2.7315,
      "step": 59900
    },
    {
      "epoch": 0.66,
      "grad_norm": 7.469701766967773,
      "learning_rate": 5.547367077617721e-05,
      "loss": 2.6429,
      "step": 59950
    },
    {
      "epoch": 0.67,
      "grad_norm": 5.144496917724609,
      "learning_rate": 5.530957607090459e-05,
      "loss": 2.7822,
      "step": 60000
    },
    {
      "epoch": 0.67,
      "grad_norm": 4.293697357177734,
      "learning_rate": 5.514563159891481e-05,
      "loss": 2.756,
      "step": 60050
    },
    {
      "epoch": 0.67,
      "grad_norm": 4.150298118591309,
      "learning_rate": 5.4981837911330626e-05,
      "loss": 2.7322,
      "step": 60100
    },
    {
      "epoch": 0.67,
      "grad_norm": 5.827974796295166,
      "learning_rate": 5.481819555876819e-05,
      "loss": 2.6922,
      "step": 60150
    },
    {
      "epoch": 0.67,
      "grad_norm": 5.333983898162842,
      "learning_rate": 5.465470509133466e-05,
      "loss": 2.7122,
      "step": 60200
    },
    {
      "epoch": 0.67,
      "grad_norm": 7.231263637542725,
      "learning_rate": 5.449136705862687e-05,
      "loss": 2.7259,
      "step": 60250
    },
    {
      "epoch": 0.67,
      "grad_norm": 6.030016899108887,
      "learning_rate": 5.432818200972899e-05,
      "loss": 2.6265,
      "step": 60300
    },
    {
      "epoch": 0.67,
      "grad_norm": 5.128213405609131,
      "learning_rate": 5.416515049321105e-05,
      "loss": 2.7022,
      "step": 60350
    },
    {
      "epoch": 0.67,
      "grad_norm": 6.4167585372924805,
      "learning_rate": 5.400227305712693e-05,
      "loss": 2.6602,
      "step": 60400
    },
    {
      "epoch": 0.67,
      "grad_norm": 5.0454535484313965,
      "learning_rate": 5.383955024901253e-05,
      "loss": 2.8175,
      "step": 60450
    },
    {
      "epoch": 0.67,
      "grad_norm": 6.559722423553467,
      "learning_rate": 5.367698261588394e-05,
      "loss": 2.5561,
      "step": 60500
    },
    {
      "epoch": 0.67,
      "grad_norm": 4.881749629974365,
      "learning_rate": 5.351457070423564e-05,
      "loss": 2.5868,
      "step": 60550
    },
    {
      "epoch": 0.67,
      "grad_norm": 4.018120288848877,
      "learning_rate": 5.335231506003864e-05,
      "loss": 2.761,
      "step": 60600
    },
    {
      "epoch": 0.67,
      "grad_norm": 4.554300308227539,
      "learning_rate": 5.3190216228738566e-05,
      "loss": 2.6393,
      "step": 60650
    },
    {
      "epoch": 0.67,
      "grad_norm": 5.151977062225342,
      "learning_rate": 5.3028274755253936e-05,
      "loss": 2.7029,
      "step": 60700
    },
    {
      "epoch": 0.67,
      "grad_norm": 5.380387783050537,
      "learning_rate": 5.286649118397431e-05,
      "loss": 2.681,
      "step": 60750
    },
    {
      "epoch": 0.67,
      "grad_norm": 4.952496528625488,
      "learning_rate": 5.2704866058758396e-05,
      "loss": 2.6099,
      "step": 60800
    },
    {
      "epoch": 0.67,
      "grad_norm": 4.520925998687744,
      "learning_rate": 5.254339992293229e-05,
      "loss": 2.6787,
      "step": 60850
    },
    {
      "epoch": 0.68,
      "grad_norm": 5.067198276519775,
      "learning_rate": 5.238209331928761e-05,
      "loss": 2.739,
      "step": 60900
    },
    {
      "epoch": 0.68,
      "grad_norm": 5.477843761444092,
      "learning_rate": 5.222094679007967e-05,
      "loss": 2.6496,
      "step": 60950
    },
    {
      "epoch": 0.68,
      "grad_norm": 5.660317897796631,
      "learning_rate": 5.2059960877025723e-05,
      "loss": 2.7858,
      "step": 61000
    },
    {
      "epoch": 0.68,
      "grad_norm": 4.7009992599487305,
      "learning_rate": 5.189913612130304e-05,
      "loss": 2.7136,
      "step": 61050
    },
    {
      "epoch": 0.68,
      "grad_norm": 4.286222457885742,
      "learning_rate": 5.173847306354712e-05,
      "loss": 2.5391,
      "step": 61100
    },
    {
      "epoch": 0.68,
      "grad_norm": 5.181181907653809,
      "learning_rate": 5.157797224384996e-05,
      "loss": 2.6648,
      "step": 61150
    },
    {
      "epoch": 0.68,
      "grad_norm": 6.731600284576416,
      "learning_rate": 5.141763420175811e-05,
      "loss": 2.6739,
      "step": 61200
    },
    {
      "epoch": 0.68,
      "grad_norm": 5.451928615570068,
      "learning_rate": 5.125745947627094e-05,
      "loss": 2.6651,
      "step": 61250
    },
    {
      "epoch": 0.68,
      "grad_norm": 4.888402938842773,
      "learning_rate": 5.1097448605838826e-05,
      "loss": 2.5952,
      "step": 61300
    },
    {
      "epoch": 0.68,
      "grad_norm": 5.415097236633301,
      "learning_rate": 5.0937602128361294e-05,
      "loss": 2.6332,
      "step": 61350
    },
    {
      "epoch": 0.68,
      "grad_norm": 5.228392601013184,
      "learning_rate": 5.077792058118527e-05,
      "loss": 2.6299,
      "step": 61400
    },
    {
      "epoch": 0.68,
      "grad_norm": 4.857380390167236,
      "learning_rate": 5.061840450110322e-05,
      "loss": 2.6149,
      "step": 61450
    },
    {
      "epoch": 0.68,
      "grad_norm": 4.560726165771484,
      "learning_rate": 5.0459054424351296e-05,
      "loss": 2.7187,
      "step": 61500
    },
    {
      "epoch": 0.68,
      "grad_norm": 5.016136646270752,
      "learning_rate": 5.029987088660781e-05,
      "loss": 2.7901,
      "step": 61550
    },
    {
      "epoch": 0.68,
      "grad_norm": 4.029599666595459,
      "learning_rate": 5.014085442299097e-05,
      "loss": 2.6107,
      "step": 61600
    },
    {
      "epoch": 0.68,
      "grad_norm": 4.8952741622924805,
      "learning_rate": 4.998200556805761e-05,
      "loss": 2.6566,
      "step": 61650
    },
    {
      "epoch": 0.68,
      "grad_norm": 5.930116653442383,
      "learning_rate": 4.9823324855800836e-05,
      "loss": 2.6534,
      "step": 61700
    },
    {
      "epoch": 0.68,
      "grad_norm": 4.591596603393555,
      "learning_rate": 4.966481281964882e-05,
      "loss": 2.6903,
      "step": 61750
    },
    {
      "epoch": 0.69,
      "grad_norm": 4.6077375411987305,
      "learning_rate": 4.950646999246239e-05,
      "loss": 2.707,
      "step": 61800
    },
    {
      "epoch": 0.69,
      "grad_norm": 4.62371826171875,
      "learning_rate": 4.93482969065339e-05,
      "loss": 2.7319,
      "step": 61850
    },
    {
      "epoch": 0.69,
      "grad_norm": 5.955516815185547,
      "learning_rate": 4.919029409358472e-05,
      "loss": 2.6313,
      "step": 61900
    },
    {
      "epoch": 0.69,
      "grad_norm": 6.556937217712402,
      "learning_rate": 4.9032462084764184e-05,
      "loss": 2.7019,
      "step": 61950
    },
    {
      "epoch": 0.69,
      "grad_norm": 6.456974506378174,
      "learning_rate": 4.887480141064712e-05,
      "loss": 2.7097,
      "step": 62000
    },
    {
      "epoch": 0.69,
      "grad_norm": 4.723240852355957,
      "learning_rate": 4.871731260123269e-05,
      "loss": 2.7978,
      "step": 62050
    },
    {
      "epoch": 0.69,
      "grad_norm": 5.934678554534912,
      "learning_rate": 4.855999618594201e-05,
      "loss": 2.7178,
      "step": 62100
    },
    {
      "epoch": 0.69,
      "grad_norm": 5.596593379974365,
      "learning_rate": 4.840285269361698e-05,
      "loss": 2.626,
      "step": 62150
    },
    {
      "epoch": 0.69,
      "grad_norm": 5.818239688873291,
      "learning_rate": 4.8245882652517905e-05,
      "loss": 2.5804,
      "step": 62200
    },
    {
      "epoch": 0.69,
      "grad_norm": 4.86491584777832,
      "learning_rate": 4.8089086590322186e-05,
      "loss": 2.7239,
      "step": 62250
    },
    {
      "epoch": 0.69,
      "grad_norm": 4.0567450523376465,
      "learning_rate": 4.793559575168114e-05,
      "loss": 2.6851,
      "step": 62300
    },
    {
      "epoch": 0.69,
      "grad_norm": 5.087411403656006,
      "learning_rate": 4.7779145722176965e-05,
      "loss": 2.6683,
      "step": 62350
    },
    {
      "epoch": 0.69,
      "grad_norm": 6.674807071685791,
      "learning_rate": 4.762287124057948e-05,
      "loss": 2.597,
      "step": 62400
    },
    {
      "epoch": 0.69,
      "grad_norm": 4.746437072753906,
      "learning_rate": 4.7466772832227666e-05,
      "loss": 2.6623,
      "step": 62450
    },
    {
      "epoch": 0.69,
      "grad_norm": 5.101381778717041,
      "learning_rate": 4.731085102186891e-05,
      "loss": 2.5818,
      "step": 62500
    },
    {
      "epoch": 0.69,
      "grad_norm": 4.619199275970459,
      "learning_rate": 4.715510633365664e-05,
      "loss": 2.7697,
      "step": 62550
    },
    {
      "epoch": 0.69,
      "grad_norm": 5.254241943359375,
      "learning_rate": 4.6999539291149154e-05,
      "loss": 2.7071,
      "step": 62600
    },
    {
      "epoch": 0.69,
      "grad_norm": 4.435392379760742,
      "learning_rate": 4.684415041730727e-05,
      "loss": 2.6527,
      "step": 62650
    },
    {
      "epoch": 0.7,
      "grad_norm": 4.774613857269287,
      "learning_rate": 4.668894023449317e-05,
      "loss": 2.6708,
      "step": 62700
    },
    {
      "epoch": 0.7,
      "grad_norm": 5.6753010749816895,
      "learning_rate": 4.653390926446806e-05,
      "loss": 2.5888,
      "step": 62750
    },
    {
      "epoch": 0.7,
      "grad_norm": 4.8252949714660645,
      "learning_rate": 4.6379058028390935e-05,
      "loss": 2.6198,
      "step": 62800
    },
    {
      "epoch": 0.7,
      "grad_norm": 5.033131122589111,
      "learning_rate": 4.62243870468164e-05,
      "loss": 2.5929,
      "step": 62850
    },
    {
      "epoch": 0.7,
      "grad_norm": 5.117373943328857,
      "learning_rate": 4.606989683969322e-05,
      "loss": 2.627,
      "step": 62900
    },
    {
      "epoch": 0.7,
      "grad_norm": 5.87759256362915,
      "learning_rate": 4.591558792636242e-05,
      "loss": 2.734,
      "step": 62950
    },
    {
      "epoch": 0.7,
      "grad_norm": 4.762241840362549,
      "learning_rate": 4.576146082555559e-05,
      "loss": 2.6727,
      "step": 63000
    },
    {
      "epoch": 0.7,
      "grad_norm": 4.74021053314209,
      "learning_rate": 4.560751605539314e-05,
      "loss": 2.7416,
      "step": 63050
    },
    {
      "epoch": 0.7,
      "grad_norm": 4.793142318725586,
      "learning_rate": 4.545375413338253e-05,
      "loss": 2.6885,
      "step": 63100
    },
    {
      "epoch": 0.7,
      "grad_norm": 5.351389408111572,
      "learning_rate": 4.530017557641656e-05,
      "loss": 2.7941,
      "step": 63150
    },
    {
      "epoch": 0.7,
      "grad_norm": 5.015024662017822,
      "learning_rate": 4.5146780900771636e-05,
      "loss": 2.5848,
      "step": 63200
    },
    {
      "epoch": 0.7,
      "grad_norm": 4.458579063415527,
      "learning_rate": 4.499357062210597e-05,
      "loss": 2.6863,
      "step": 63250
    },
    {
      "epoch": 0.7,
      "grad_norm": 5.938669681549072,
      "learning_rate": 4.4840545255457976e-05,
      "loss": 2.6753,
      "step": 63300
    },
    {
      "epoch": 0.7,
      "grad_norm": 4.693249225616455,
      "learning_rate": 4.468770531524439e-05,
      "loss": 2.6635,
      "step": 63350
    },
    {
      "epoch": 0.7,
      "grad_norm": 4.526652812957764,
      "learning_rate": 4.453505131525862e-05,
      "loss": 2.6117,
      "step": 63400
    },
    {
      "epoch": 0.7,
      "grad_norm": 5.427757740020752,
      "learning_rate": 4.4382583768669075e-05,
      "loss": 2.7516,
      "step": 63450
    },
    {
      "epoch": 0.7,
      "grad_norm": 5.505650997161865,
      "learning_rate": 4.423030318801728e-05,
      "loss": 2.6203,
      "step": 63500
    },
    {
      "epoch": 0.7,
      "grad_norm": 5.214341640472412,
      "learning_rate": 4.407821008521632e-05,
      "loss": 2.6517,
      "step": 63550
    },
    {
      "epoch": 0.71,
      "grad_norm": 5.016122817993164,
      "learning_rate": 4.3926304971549004e-05,
      "loss": 2.6629,
      "step": 63600
    },
    {
      "epoch": 0.71,
      "grad_norm": 4.789595603942871,
      "learning_rate": 4.37745883576662e-05,
      "loss": 2.5895,
      "step": 63650
    },
    {
      "epoch": 0.71,
      "grad_norm": 4.732479572296143,
      "learning_rate": 4.362306075358512e-05,
      "loss": 2.5952,
      "step": 63700
    },
    {
      "epoch": 0.71,
      "grad_norm": 7.306616306304932,
      "learning_rate": 4.347172266868758e-05,
      "loss": 2.5092,
      "step": 63750
    },
    {
      "epoch": 0.71,
      "grad_norm": 4.95161771774292,
      "learning_rate": 4.33205746117183e-05,
      "loss": 2.7244,
      "step": 63800
    },
    {
      "epoch": 0.71,
      "grad_norm": 4.876521110534668,
      "learning_rate": 4.316961709078319e-05,
      "loss": 2.7332,
      "step": 63850
    },
    {
      "epoch": 0.71,
      "grad_norm": 5.317580699920654,
      "learning_rate": 4.301885061334766e-05,
      "loss": 2.6448,
      "step": 63900
    },
    {
      "epoch": 0.71,
      "grad_norm": 5.089837551116943,
      "learning_rate": 4.286827568623489e-05,
      "loss": 2.7677,
      "step": 63950
    },
    {
      "epoch": 0.71,
      "grad_norm": 4.458879470825195,
      "learning_rate": 4.271789281562417e-05,
      "loss": 2.6661,
      "step": 64000
    },
    {
      "epoch": 0.71,
      "grad_norm": 4.508949279785156,
      "learning_rate": 4.256770250704902e-05,
      "loss": 2.5478,
      "step": 64050
    },
    {
      "epoch": 0.71,
      "grad_norm": 5.3414306640625,
      "learning_rate": 4.241770526539591e-05,
      "loss": 2.6117,
      "step": 64100
    },
    {
      "epoch": 0.71,
      "grad_norm": 5.041897296905518,
      "learning_rate": 4.226790159490196e-05,
      "loss": 2.7514,
      "step": 64150
    },
    {
      "epoch": 0.71,
      "grad_norm": 6.20094633102417,
      "learning_rate": 4.211829199915389e-05,
      "loss": 2.6139,
      "step": 64200
    },
    {
      "epoch": 0.71,
      "grad_norm": 5.264660358428955,
      "learning_rate": 4.196887698108573e-05,
      "loss": 2.665,
      "step": 64250
    },
    {
      "epoch": 0.71,
      "grad_norm": 5.019193649291992,
      "learning_rate": 4.181965704297764e-05,
      "loss": 2.6762,
      "step": 64300
    },
    {
      "epoch": 0.71,
      "grad_norm": 4.88646125793457,
      "learning_rate": 4.167063268645377e-05,
      "loss": 2.6067,
      "step": 64350
    },
    {
      "epoch": 0.71,
      "grad_norm": 4.774142265319824,
      "learning_rate": 4.152180441248106e-05,
      "loss": 2.7637,
      "step": 64400
    },
    {
      "epoch": 0.71,
      "grad_norm": 4.547387599945068,
      "learning_rate": 4.137317272136699e-05,
      "loss": 2.7124,
      "step": 64450
    },
    {
      "epoch": 0.72,
      "grad_norm": 4.326138973236084,
      "learning_rate": 4.1224738112758485e-05,
      "loss": 2.7023,
      "step": 64500
    },
    {
      "epoch": 0.72,
      "grad_norm": 5.459550380706787,
      "learning_rate": 4.1076501085639695e-05,
      "loss": 2.7996,
      "step": 64550
    },
    {
      "epoch": 0.72,
      "grad_norm": 5.483733654022217,
      "learning_rate": 4.09284621383308e-05,
      "loss": 2.6413,
      "step": 64600
    },
    {
      "epoch": 0.72,
      "grad_norm": 5.255352973937988,
      "learning_rate": 4.0780621768485927e-05,
      "loss": 2.5948,
      "step": 64650
    },
    {
      "epoch": 0.72,
      "grad_norm": 4.731751918792725,
      "learning_rate": 4.06329804730917e-05,
      "loss": 2.5661,
      "step": 64700
    },
    {
      "epoch": 0.72,
      "grad_norm": 4.457857131958008,
      "learning_rate": 4.048553874846567e-05,
      "loss": 2.5307,
      "step": 64750
    },
    {
      "epoch": 0.72,
      "grad_norm": 4.675079345703125,
      "learning_rate": 4.033829709025423e-05,
      "loss": 2.6014,
      "step": 64800
    },
    {
      "epoch": 0.72,
      "grad_norm": 5.134871482849121,
      "learning_rate": 4.019125599343153e-05,
      "loss": 2.5904,
      "step": 64850
    },
    {
      "epoch": 0.72,
      "grad_norm": 4.3691253662109375,
      "learning_rate": 4.004441595229721e-05,
      "loss": 2.698,
      "step": 64900
    },
    {
      "epoch": 0.72,
      "grad_norm": 4.799289703369141,
      "learning_rate": 3.9897777460475306e-05,
      "loss": 2.7371,
      "step": 64950
    },
    {
      "epoch": 0.72,
      "grad_norm": 5.261094093322754,
      "learning_rate": 3.975134101091203e-05,
      "loss": 2.6594,
      "step": 65000
    },
    {
      "epoch": 0.72,
      "grad_norm": 4.8293585777282715,
      "learning_rate": 3.960802978615641e-05,
      "loss": 2.6577,
      "step": 65050
    },
    {
      "epoch": 0.72,
      "grad_norm": 4.705104827880859,
      "learning_rate": 3.9461994831895845e-05,
      "loss": 2.643,
      "step": 65100
    },
    {
      "epoch": 0.72,
      "grad_norm": 5.281649112701416,
      "learning_rate": 3.931616338483999e-05,
      "loss": 2.7382,
      "step": 65150
    },
    {
      "epoch": 0.72,
      "grad_norm": 4.238641262054443,
      "learning_rate": 3.9170535935222084e-05,
      "loss": 2.6859,
      "step": 65200
    },
    {
      "epoch": 0.72,
      "grad_norm": 3.864513397216797,
      "learning_rate": 3.902511297258985e-05,
      "loss": 2.6514,
      "step": 65250
    },
    {
      "epoch": 0.72,
      "grad_norm": 5.5272932052612305,
      "learning_rate": 3.8879894985803355e-05,
      "loss": 2.5481,
      "step": 65300
    },
    {
      "epoch": 0.72,
      "grad_norm": 5.717905521392822,
      "learning_rate": 3.873488246303387e-05,
      "loss": 2.7319,
      "step": 65350
    },
    {
      "epoch": 0.73,
      "grad_norm": 4.693354606628418,
      "learning_rate": 3.8590075891761715e-05,
      "loss": 2.5399,
      "step": 65400
    },
    {
      "epoch": 0.73,
      "grad_norm": 5.992891788482666,
      "learning_rate": 3.844547575877503e-05,
      "loss": 2.6001,
      "step": 65450
    },
    {
      "epoch": 0.73,
      "grad_norm": 5.528879165649414,
      "learning_rate": 3.8301082550167946e-05,
      "loss": 2.7701,
      "step": 65500
    },
    {
      "epoch": 0.73,
      "grad_norm": 7.080493927001953,
      "learning_rate": 3.8156896751338976e-05,
      "loss": 2.6873,
      "step": 65550
    },
    {
      "epoch": 0.73,
      "grad_norm": 5.8174333572387695,
      "learning_rate": 3.801291884698942e-05,
      "loss": 2.6757,
      "step": 65600
    },
    {
      "epoch": 0.73,
      "grad_norm": 6.21406888961792,
      "learning_rate": 3.786914932112169e-05,
      "loss": 2.6236,
      "step": 65650
    },
    {
      "epoch": 0.73,
      "grad_norm": 5.273697376251221,
      "learning_rate": 3.7725588657037695e-05,
      "loss": 2.5544,
      "step": 65700
    },
    {
      "epoch": 0.73,
      "grad_norm": 6.8518595695495605,
      "learning_rate": 3.7582237337337256e-05,
      "loss": 2.6315,
      "step": 65750
    },
    {
      "epoch": 0.73,
      "grad_norm": 5.1203293800354,
      "learning_rate": 3.743909584391643e-05,
      "loss": 2.4959,
      "step": 65800
    },
    {
      "epoch": 0.73,
      "grad_norm": 5.127179145812988,
      "learning_rate": 3.729616465796593e-05,
      "loss": 2.7377,
      "step": 65850
    },
    {
      "epoch": 0.73,
      "grad_norm": 4.408633708953857,
      "learning_rate": 3.715344425996945e-05,
      "loss": 2.6407,
      "step": 65900
    },
    {
      "epoch": 0.73,
      "grad_norm": 4.980579376220703,
      "learning_rate": 3.701093512970214e-05,
      "loss": 2.7636,
      "step": 65950
    },
    {
      "epoch": 0.73,
      "grad_norm": 7.047328472137451,
      "learning_rate": 3.686863774622893e-05,
      "loss": 2.5239,
      "step": 66000
    },
    {
      "epoch": 0.73,
      "grad_norm": 4.069336891174316,
      "learning_rate": 3.6726552587902906e-05,
      "loss": 2.6355,
      "step": 66050
    },
    {
      "epoch": 0.73,
      "grad_norm": 4.68779182434082,
      "learning_rate": 3.6584680132363766e-05,
      "loss": 2.7245,
      "step": 66100
    },
    {
      "epoch": 0.73,
      "grad_norm": 6.041943550109863,
      "learning_rate": 3.6443020856536147e-05,
      "loss": 2.6538,
      "step": 66150
    },
    {
      "epoch": 0.73,
      "grad_norm": 6.336287975311279,
      "learning_rate": 3.630157523662806e-05,
      "loss": 2.761,
      "step": 66200
    },
    {
      "epoch": 0.73,
      "grad_norm": 4.71364688873291,
      "learning_rate": 3.6160343748129285e-05,
      "loss": 2.5481,
      "step": 66250
    },
    {
      "epoch": 0.74,
      "grad_norm": 5.510044574737549,
      "learning_rate": 3.601932686580979e-05,
      "loss": 2.6258,
      "step": 66300
    },
    {
      "epoch": 0.74,
      "grad_norm": 5.9190497398376465,
      "learning_rate": 3.587852506371806e-05,
      "loss": 2.6299,
      "step": 66350
    },
    {
      "epoch": 0.74,
      "grad_norm": 5.256284713745117,
      "learning_rate": 3.573793881517961e-05,
      "loss": 2.6011,
      "step": 66400
    },
    {
      "epoch": 0.74,
      "grad_norm": 5.074404716491699,
      "learning_rate": 3.5597568592795295e-05,
      "loss": 2.5963,
      "step": 66450
    },
    {
      "epoch": 0.74,
      "grad_norm": 6.2140398025512695,
      "learning_rate": 3.545741486843982e-05,
      "loss": 2.6901,
      "step": 66500
    },
    {
      "epoch": 0.74,
      "grad_norm": 5.4605302810668945,
      "learning_rate": 3.5317478113260064e-05,
      "loss": 2.651,
      "step": 66550
    },
    {
      "epoch": 0.74,
      "grad_norm": 5.9920196533203125,
      "learning_rate": 3.5177758797673524e-05,
      "loss": 2.6192,
      "step": 66600
    },
    {
      "epoch": 0.74,
      "grad_norm": 5.6968817710876465,
      "learning_rate": 3.503825739136677e-05,
      "loss": 2.6061,
      "step": 66650
    },
    {
      "epoch": 0.74,
      "grad_norm": 6.247105121612549,
      "learning_rate": 3.4898974363293835e-05,
      "loss": 2.571,
      "step": 66700
    },
    {
      "epoch": 0.74,
      "grad_norm": 5.774839401245117,
      "learning_rate": 3.475991018167463e-05,
      "loss": 2.7125,
      "step": 66750
    },
    {
      "epoch": 0.74,
      "grad_norm": 6.337151527404785,
      "learning_rate": 3.4621065313993376e-05,
      "loss": 2.5568,
      "step": 66800
    },
    {
      "epoch": 0.74,
      "grad_norm": 4.821563720703125,
      "learning_rate": 3.448244022699706e-05,
      "loss": 2.7574,
      "step": 66850
    },
    {
      "epoch": 0.74,
      "grad_norm": 5.146304607391357,
      "learning_rate": 3.434403538669384e-05,
      "loss": 2.5438,
      "step": 66900
    },
    {
      "epoch": 0.74,
      "grad_norm": 5.046156883239746,
      "learning_rate": 3.420585125835146e-05,
      "loss": 2.5751,
      "step": 66950
    },
    {
      "epoch": 0.74,
      "grad_norm": 5.2903547286987305,
      "learning_rate": 3.4067888306495755e-05,
      "loss": 2.5946,
      "step": 67000
    },
    {
      "epoch": 0.74,
      "grad_norm": 6.126682758331299,
      "learning_rate": 3.3930146994909e-05,
      "loss": 2.6725,
      "step": 67050
    },
    {
      "epoch": 0.74,
      "grad_norm": 4.8487958908081055,
      "learning_rate": 3.379262778662839e-05,
      "loss": 2.5634,
      "step": 67100
    },
    {
      "epoch": 0.74,
      "grad_norm": 4.605789661407471,
      "learning_rate": 3.3655331143944536e-05,
      "loss": 2.5278,
      "step": 67150
    },
    {
      "epoch": 0.75,
      "grad_norm": 5.337599754333496,
      "learning_rate": 3.351825752839986e-05,
      "loss": 2.5888,
      "step": 67200
    },
    {
      "epoch": 0.75,
      "grad_norm": 5.273902416229248,
      "learning_rate": 3.338140740078691e-05,
      "loss": 2.5882,
      "step": 67250
    },
    {
      "epoch": 0.75,
      "grad_norm": 4.65778923034668,
      "learning_rate": 3.32447812211472e-05,
      "loss": 2.6465,
      "step": 67300
    },
    {
      "epoch": 0.75,
      "grad_norm": 5.418448448181152,
      "learning_rate": 3.3108379448769134e-05,
      "loss": 2.5445,
      "step": 67350
    },
    {
      "epoch": 0.75,
      "grad_norm": 4.730058193206787,
      "learning_rate": 3.2972202542187004e-05,
      "loss": 2.6927,
      "step": 67400
    },
    {
      "epoch": 0.75,
      "grad_norm": 4.912891387939453,
      "learning_rate": 3.2836250959178916e-05,
      "loss": 2.6355,
      "step": 67450
    },
    {
      "epoch": 0.75,
      "grad_norm": 4.967782497406006,
      "learning_rate": 3.270052515676581e-05,
      "loss": 2.6581,
      "step": 67500
    },
    {
      "epoch": 0.75,
      "grad_norm": 6.024198055267334,
      "learning_rate": 3.256502559120931e-05,
      "loss": 2.6022,
      "step": 67550
    },
    {
      "epoch": 0.75,
      "grad_norm": 5.38806676864624,
      "learning_rate": 3.2429752718010816e-05,
      "loss": 2.6499,
      "step": 67600
    },
    {
      "epoch": 0.75,
      "grad_norm": 6.4937214851379395,
      "learning_rate": 3.2294706991909426e-05,
      "loss": 2.6135,
      "step": 67650
    },
    {
      "epoch": 0.75,
      "grad_norm": 5.324350833892822,
      "learning_rate": 3.2159888866880885e-05,
      "loss": 2.4966,
      "step": 67700
    },
    {
      "epoch": 0.75,
      "grad_norm": 5.561586856842041,
      "learning_rate": 3.202529879613554e-05,
      "loss": 2.4383,
      "step": 67750
    },
    {
      "epoch": 0.75,
      "grad_norm": 5.771432399749756,
      "learning_rate": 3.189093723211739e-05,
      "loss": 2.5363,
      "step": 67800
    },
    {
      "epoch": 0.75,
      "grad_norm": 4.824429035186768,
      "learning_rate": 3.1756804626502036e-05,
      "loss": 2.5594,
      "step": 67850
    },
    {
      "epoch": 0.75,
      "grad_norm": 5.729842185974121,
      "learning_rate": 3.162290143019563e-05,
      "loss": 2.6484,
      "step": 67900
    },
    {
      "epoch": 0.75,
      "grad_norm": 5.200084686279297,
      "learning_rate": 3.148922809333289e-05,
      "loss": 2.509,
      "step": 67950
    },
    {
      "epoch": 0.75,
      "grad_norm": 4.5528388023376465,
      "learning_rate": 3.135578506527601e-05,
      "loss": 2.5466,
      "step": 68000
    },
    {
      "epoch": 0.75,
      "grad_norm": 5.316043853759766,
      "learning_rate": 3.12225727946129e-05,
      "loss": 2.6108,
      "step": 68050
    },
    {
      "epoch": 0.76,
      "grad_norm": 4.202846527099609,
      "learning_rate": 3.108959172915574e-05,
      "loss": 2.6763,
      "step": 68100
    },
    {
      "epoch": 0.76,
      "grad_norm": 5.51525354385376,
      "learning_rate": 3.095684231593952e-05,
      "loss": 2.6869,
      "step": 68150
    },
    {
      "epoch": 0.76,
      "grad_norm": 5.375635623931885,
      "learning_rate": 3.0824325001220436e-05,
      "loss": 2.5884,
      "step": 68200
    },
    {
      "epoch": 0.76,
      "grad_norm": 5.150887489318848,
      "learning_rate": 3.069204023047451e-05,
      "loss": 2.5735,
      "step": 68250
    },
    {
      "epoch": 0.76,
      "grad_norm": 5.633454322814941,
      "learning_rate": 3.055998844839599e-05,
      "loss": 2.6382,
      "step": 68300
    },
    {
      "epoch": 0.76,
      "grad_norm": 5.733946800231934,
      "learning_rate": 3.0428170098895937e-05,
      "loss": 2.5457,
      "step": 68350
    },
    {
      "epoch": 0.76,
      "grad_norm": 5.607996940612793,
      "learning_rate": 3.0296585625100682e-05,
      "loss": 2.641,
      "step": 68400
    },
    {
      "epoch": 0.76,
      "grad_norm": 5.576999664306641,
      "learning_rate": 3.0165235469350327e-05,
      "loss": 2.524,
      "step": 68450
    },
    {
      "epoch": 0.76,
      "grad_norm": 5.019045352935791,
      "learning_rate": 3.0034120073197313e-05,
      "loss": 2.5996,
      "step": 68500
    },
    {
      "epoch": 0.76,
      "grad_norm": 5.5556321144104,
      "learning_rate": 2.9903239877404877e-05,
      "loss": 2.6136,
      "step": 68550
    },
    {
      "epoch": 0.76,
      "grad_norm": 4.786374092102051,
      "learning_rate": 2.9772595321945605e-05,
      "loss": 2.6759,
      "step": 68600
    },
    {
      "epoch": 0.76,
      "grad_norm": 5.464079856872559,
      "learning_rate": 2.9642186845999954e-05,
      "loss": 2.7705,
      "step": 68650
    },
    {
      "epoch": 0.76,
      "grad_norm": 5.239927291870117,
      "learning_rate": 2.9512014887954752e-05,
      "loss": 2.7256,
      "step": 68700
    },
    {
      "epoch": 0.76,
      "grad_norm": 5.427574157714844,
      "learning_rate": 2.9382079885401748e-05,
      "loss": 2.6021,
      "step": 68750
    },
    {
      "epoch": 0.76,
      "grad_norm": 4.713808059692383,
      "learning_rate": 2.9252382275136124e-05,
      "loss": 2.5717,
      "step": 68800
    },
    {
      "epoch": 0.76,
      "grad_norm": 4.801243305206299,
      "learning_rate": 2.9122922493155046e-05,
      "loss": 2.5959,
      "step": 68850
    },
    {
      "epoch": 0.76,
      "grad_norm": 5.959753513336182,
      "learning_rate": 2.8993700974656157e-05,
      "loss": 2.4835,
      "step": 68900
    },
    {
      "epoch": 0.76,
      "grad_norm": 6.386477947235107,
      "learning_rate": 2.8864718154036185e-05,
      "loss": 2.5146,
      "step": 68950
    },
    {
      "epoch": 0.77,
      "grad_norm": 5.259350776672363,
      "learning_rate": 2.8735974464889404e-05,
      "loss": 2.7085,
      "step": 69000
    },
    {
      "epoch": 0.77,
      "grad_norm": 5.164392948150635,
      "learning_rate": 2.8610038071978694e-05,
      "loss": 2.6228,
      "step": 69050
    },
    {
      "epoch": 0.77,
      "grad_norm": 4.889065742492676,
      "learning_rate": 2.8481769139191007e-05,
      "loss": 2.6869,
      "step": 69100
    },
    {
      "epoch": 0.77,
      "grad_norm": 5.981446266174316,
      "learning_rate": 2.8353740625214655e-05,
      "loss": 2.6408,
      "step": 69150
    },
    {
      "epoch": 0.77,
      "grad_norm": 5.691173553466797,
      "learning_rate": 2.8225952960435886e-05,
      "loss": 2.6958,
      "step": 69200
    },
    {
      "epoch": 0.77,
      "grad_norm": 4.624173641204834,
      "learning_rate": 2.8098406574431322e-05,
      "loss": 2.5914,
      "step": 69250
    },
    {
      "epoch": 0.77,
      "grad_norm": 5.377009391784668,
      "learning_rate": 2.7971101895966457e-05,
      "loss": 2.5921,
      "step": 69300
    },
    {
      "epoch": 0.77,
      "grad_norm": 4.812516689300537,
      "learning_rate": 2.7844039352994255e-05,
      "loss": 2.5843,
      "step": 69350
    },
    {
      "epoch": 0.77,
      "grad_norm": 5.23838996887207,
      "learning_rate": 2.7717219372653736e-05,
      "loss": 2.6546,
      "step": 69400
    },
    {
      "epoch": 0.77,
      "grad_norm": 5.0327677726745605,
      "learning_rate": 2.759064238126847e-05,
      "loss": 2.569,
      "step": 69450
    },
    {
      "epoch": 0.77,
      "grad_norm": 5.910417079925537,
      "learning_rate": 2.7464308804345217e-05,
      "loss": 2.5743,
      "step": 69500
    },
    {
      "epoch": 0.77,
      "grad_norm": 5.872021675109863,
      "learning_rate": 2.7338219066572457e-05,
      "loss": 2.6469,
      "step": 69550
    },
    {
      "epoch": 0.77,
      "grad_norm": 5.492809772491455,
      "learning_rate": 2.7212373591818942e-05,
      "loss": 2.6901,
      "step": 69600
    },
    {
      "epoch": 0.77,
      "grad_norm": 5.253983974456787,
      "learning_rate": 2.7086772803132344e-05,
      "loss": 2.6407,
      "step": 69650
    },
    {
      "epoch": 0.77,
      "grad_norm": 4.084805488586426,
      "learning_rate": 2.6961417122737753e-05,
      "loss": 2.5014,
      "step": 69700
    },
    {
      "epoch": 0.77,
      "grad_norm": 5.427392959594727,
      "learning_rate": 2.683630697203634e-05,
      "loss": 2.6135,
      "step": 69750
    },
    {
      "epoch": 0.77,
      "grad_norm": 6.176027774810791,
      "learning_rate": 2.6711442771603757e-05,
      "loss": 2.6338,
      "step": 69800
    },
    {
      "epoch": 0.77,
      "grad_norm": 5.913874626159668,
      "learning_rate": 2.6586824941189092e-05,
      "loss": 2.6866,
      "step": 69850
    },
    {
      "epoch": 0.78,
      "grad_norm": 6.058617115020752,
      "learning_rate": 2.6462453899712956e-05,
      "loss": 2.6602,
      "step": 69900
    },
    {
      "epoch": 0.78,
      "grad_norm": 4.49558687210083,
      "learning_rate": 2.6338330065266626e-05,
      "loss": 2.5515,
      "step": 69950
    },
    {
      "epoch": 0.78,
      "grad_norm": 5.014800548553467,
      "learning_rate": 2.6214453855110056e-05,
      "loss": 2.6239,
      "step": 70000
    },
    {
      "epoch": 0.78,
      "grad_norm": 4.493063926696777,
      "learning_rate": 2.6090825685671073e-05,
      "loss": 2.5475,
      "step": 70050
    },
    {
      "epoch": 0.78,
      "grad_norm": 5.5358171463012695,
      "learning_rate": 2.5967445972543414e-05,
      "loss": 2.6701,
      "step": 70100
    },
    {
      "epoch": 0.78,
      "grad_norm": 5.248385906219482,
      "learning_rate": 2.584431513048583e-05,
      "loss": 2.5959,
      "step": 70150
    },
    {
      "epoch": 0.78,
      "grad_norm": 6.446710109710693,
      "learning_rate": 2.572143357342022e-05,
      "loss": 2.6035,
      "step": 70200
    },
    {
      "epoch": 0.78,
      "grad_norm": 4.519626617431641,
      "learning_rate": 2.559880171443072e-05,
      "loss": 2.5648,
      "step": 70250
    },
    {
      "epoch": 0.78,
      "grad_norm": 5.842326641082764,
      "learning_rate": 2.5476419965761834e-05,
      "loss": 2.6762,
      "step": 70300
    },
    {
      "epoch": 0.78,
      "grad_norm": 5.328994274139404,
      "learning_rate": 2.53542887388175e-05,
      "loss": 2.6056,
      "step": 70350
    },
    {
      "epoch": 0.78,
      "grad_norm": 4.1011152267456055,
      "learning_rate": 2.5234843588264645e-05,
      "loss": 2.6015,
      "step": 70400
    },
    {
      "epoch": 0.78,
      "grad_norm": 6.344366550445557,
      "learning_rate": 2.511320960476089e-05,
      "loss": 2.6078,
      "step": 70450
    },
    {
      "epoch": 0.78,
      "grad_norm": 6.164771556854248,
      "learning_rate": 2.4991827363965515e-05,
      "loss": 2.5715,
      "step": 70500
    },
    {
      "epoch": 0.78,
      "grad_norm": 4.802293300628662,
      "learning_rate": 2.4870697273922274e-05,
      "loss": 2.5835,
      "step": 70550
    },
    {
      "epoch": 0.78,
      "grad_norm": 4.860954284667969,
      "learning_rate": 2.4749819741827474e-05,
      "loss": 2.5601,
      "step": 70600
    },
    {
      "epoch": 0.78,
      "grad_norm": 5.093806266784668,
      "learning_rate": 2.4629195174028195e-05,
      "loss": 2.587,
      "step": 70650
    },
    {
      "epoch": 0.78,
      "grad_norm": 5.063363075256348,
      "learning_rate": 2.450882397602138e-05,
      "loss": 2.5513,
      "step": 70700
    },
    {
      "epoch": 0.78,
      "grad_norm": 4.237406253814697,
      "learning_rate": 2.438870655245199e-05,
      "loss": 2.5971,
      "step": 70750
    },
    {
      "epoch": 0.79,
      "grad_norm": 4.2623610496521,
      "learning_rate": 2.4268843307112132e-05,
      "loss": 2.5996,
      "step": 70800
    },
    {
      "epoch": 0.79,
      "grad_norm": 4.136486530303955,
      "learning_rate": 2.41492346429392e-05,
      "loss": 2.5447,
      "step": 70850
    },
    {
      "epoch": 0.79,
      "grad_norm": 4.980117321014404,
      "learning_rate": 2.4029880962015038e-05,
      "loss": 2.6504,
      "step": 70900
    },
    {
      "epoch": 0.79,
      "grad_norm": 5.2773518562316895,
      "learning_rate": 2.391078266556407e-05,
      "loss": 2.6319,
      "step": 70950
    },
    {
      "epoch": 0.79,
      "grad_norm": 4.6080732345581055,
      "learning_rate": 2.3791940153952464e-05,
      "loss": 2.6571,
      "step": 71000
    },
    {
      "epoch": 0.79,
      "grad_norm": 6.255810260772705,
      "learning_rate": 2.3673353826686275e-05,
      "loss": 2.5552,
      "step": 71050
    },
    {
      "epoch": 0.79,
      "grad_norm": 4.48773193359375,
      "learning_rate": 2.3555024082410603e-05,
      "loss": 2.6354,
      "step": 71100
    },
    {
      "epoch": 0.79,
      "grad_norm": 4.960320949554443,
      "learning_rate": 2.343695131890782e-05,
      "loss": 2.6678,
      "step": 71150
    },
    {
      "epoch": 0.79,
      "grad_norm": 6.516583442687988,
      "learning_rate": 2.3319135933096493e-05,
      "loss": 2.5373,
      "step": 71200
    },
    {
      "epoch": 0.79,
      "grad_norm": 5.552701950073242,
      "learning_rate": 2.320157832102998e-05,
      "loss": 2.6271,
      "step": 71250
    },
    {
      "epoch": 0.79,
      "grad_norm": 4.759840488433838,
      "learning_rate": 2.3084278877895106e-05,
      "loss": 2.5934,
      "step": 71300
    },
    {
      "epoch": 0.79,
      "grad_norm": 7.348878860473633,
      "learning_rate": 2.2967237998010784e-05,
      "loss": 2.5784,
      "step": 71350
    },
    {
      "epoch": 0.79,
      "grad_norm": 4.939338207244873,
      "learning_rate": 2.2850456074826766e-05,
      "loss": 2.6144,
      "step": 71400
    },
    {
      "epoch": 0.79,
      "grad_norm": 5.545421600341797,
      "learning_rate": 2.2733933500922254e-05,
      "loss": 2.735,
      "step": 71450
    },
    {
      "epoch": 0.79,
      "grad_norm": 4.873772621154785,
      "learning_rate": 2.2617670668004653e-05,
      "loss": 2.6888,
      "step": 71500
    },
    {
      "epoch": 0.79,
      "grad_norm": 5.401042938232422,
      "learning_rate": 2.250166796690817e-05,
      "loss": 2.6143,
      "step": 71550
    },
    {
      "epoch": 0.79,
      "grad_norm": 5.7700653076171875,
      "learning_rate": 2.2385925787592543e-05,
      "loss": 2.477,
      "step": 71600
    },
    {
      "epoch": 0.79,
      "grad_norm": 5.366491794586182,
      "learning_rate": 2.2270444519141763e-05,
      "loss": 2.5567,
      "step": 71650
    },
    {
      "epoch": 0.8,
      "grad_norm": 4.462210178375244,
      "learning_rate": 2.215522454976271e-05,
      "loss": 2.5026,
      "step": 71700
    },
    {
      "epoch": 0.8,
      "grad_norm": 4.724314212799072,
      "learning_rate": 2.204026626678386e-05,
      "loss": 2.5007,
      "step": 71750
    },
    {
      "epoch": 0.8,
      "grad_norm": 5.366270542144775,
      "learning_rate": 2.1925570056654e-05,
      "loss": 2.5524,
      "step": 71800
    },
    {
      "epoch": 0.8,
      "grad_norm": 5.3839921951293945,
      "learning_rate": 2.1811136304940928e-05,
      "loss": 2.6468,
      "step": 71850
    },
    {
      "epoch": 0.8,
      "grad_norm": 5.34608268737793,
      "learning_rate": 2.1696965396330148e-05,
      "loss": 2.5045,
      "step": 71900
    },
    {
      "epoch": 0.8,
      "grad_norm": 4.68846321105957,
      "learning_rate": 2.158305771462359e-05,
      "loss": 2.5885,
      "step": 71950
    },
    {
      "epoch": 0.8,
      "grad_norm": 5.150451183319092,
      "learning_rate": 2.1469413642738267e-05,
      "loss": 2.6345,
      "step": 72000
    },
    {
      "epoch": 0.8,
      "grad_norm": 5.348923206329346,
      "learning_rate": 2.135603356270508e-05,
      "loss": 2.5886,
      "step": 72050
    },
    {
      "epoch": 0.8,
      "grad_norm": 4.919592380523682,
      "learning_rate": 2.1242917855667465e-05,
      "loss": 2.5777,
      "step": 72100
    },
    {
      "epoch": 0.8,
      "grad_norm": 5.09805154800415,
      "learning_rate": 2.113006690188011e-05,
      "loss": 2.637,
      "step": 72150
    },
    {
      "epoch": 0.8,
      "grad_norm": 6.317235469818115,
      "learning_rate": 2.1017481080707757e-05,
      "loss": 2.5614,
      "step": 72200
    },
    {
      "epoch": 0.8,
      "grad_norm": 6.027612686157227,
      "learning_rate": 2.0905160770623743e-05,
      "loss": 2.5736,
      "step": 72250
    },
    {
      "epoch": 0.8,
      "grad_norm": 5.0004777908325195,
      "learning_rate": 2.0793106349209033e-05,
      "loss": 2.4603,
      "step": 72300
    },
    {
      "epoch": 0.8,
      "grad_norm": 5.560655117034912,
      "learning_rate": 2.0681318193150557e-05,
      "loss": 2.521,
      "step": 72350
    },
    {
      "epoch": 0.8,
      "grad_norm": 4.0777268409729,
      "learning_rate": 2.0569796678240396e-05,
      "loss": 2.7248,
      "step": 72400
    },
    {
      "epoch": 0.8,
      "grad_norm": 4.989569187164307,
      "learning_rate": 2.0458542179374018e-05,
      "loss": 2.5748,
      "step": 72450
    },
    {
      "epoch": 0.8,
      "grad_norm": 6.184603214263916,
      "learning_rate": 2.034755507054953e-05,
      "loss": 2.5971,
      "step": 72500
    },
    {
      "epoch": 0.8,
      "grad_norm": 5.676340579986572,
      "learning_rate": 2.0236835724865932e-05,
      "loss": 2.5565,
      "step": 72550
    },
    {
      "epoch": 0.81,
      "grad_norm": 5.117308616638184,
      "learning_rate": 2.0126384514522345e-05,
      "loss": 2.6551,
      "step": 72600
    },
    {
      "epoch": 0.81,
      "grad_norm": 4.88153600692749,
      "learning_rate": 2.001620181081626e-05,
      "loss": 2.6353,
      "step": 72650
    },
    {
      "epoch": 0.81,
      "grad_norm": 4.366287708282471,
      "learning_rate": 1.990628798414278e-05,
      "loss": 2.4753,
      "step": 72700
    },
    {
      "epoch": 0.81,
      "grad_norm": 5.386646747589111,
      "learning_rate": 1.979664340399292e-05,
      "loss": 2.6531,
      "step": 72750
    },
    {
      "epoch": 0.81,
      "grad_norm": 5.51713228225708,
      "learning_rate": 1.9687268438952812e-05,
      "loss": 2.536,
      "step": 72800
    },
    {
      "epoch": 0.81,
      "grad_norm": 6.721363067626953,
      "learning_rate": 1.9578163456702002e-05,
      "loss": 2.4521,
      "step": 72850
    },
    {
      "epoch": 0.81,
      "grad_norm": 5.048563003540039,
      "learning_rate": 1.946932882401269e-05,
      "loss": 2.5679,
      "step": 72900
    },
    {
      "epoch": 0.81,
      "grad_norm": 4.4516282081604,
      "learning_rate": 1.9360764906748042e-05,
      "loss": 2.6384,
      "step": 72950
    },
    {
      "epoch": 0.81,
      "grad_norm": 5.141951084136963,
      "learning_rate": 1.9252472069861295e-05,
      "loss": 2.4803,
      "step": 73000
    },
    {
      "epoch": 0.81,
      "grad_norm": 5.912957668304443,
      "learning_rate": 1.914445067739439e-05,
      "loss": 2.7067,
      "step": 73050
    },
    {
      "epoch": 0.81,
      "grad_norm": 4.5161309242248535,
      "learning_rate": 1.903670109247675e-05,
      "loss": 2.5582,
      "step": 73100
    },
    {
      "epoch": 0.81,
      "grad_norm": 5.558721542358398,
      "learning_rate": 1.8929223677324104e-05,
      "loss": 2.6032,
      "step": 73150
    },
    {
      "epoch": 0.81,
      "grad_norm": 4.807974338531494,
      "learning_rate": 1.8822018793237227e-05,
      "loss": 2.6008,
      "step": 73200
    },
    {
      "epoch": 0.81,
      "grad_norm": 5.234206199645996,
      "learning_rate": 1.8715086800600734e-05,
      "loss": 2.595,
      "step": 73250
    },
    {
      "epoch": 0.81,
      "grad_norm": 5.292330265045166,
      "learning_rate": 1.8608428058881898e-05,
      "loss": 2.419,
      "step": 73300
    },
    {
      "epoch": 0.81,
      "grad_norm": 6.529526233673096,
      "learning_rate": 1.8502042926629404e-05,
      "loss": 2.6273,
      "step": 73350
    },
    {
      "epoch": 0.81,
      "grad_norm": 6.344511985778809,
      "learning_rate": 1.8395931761472153e-05,
      "loss": 2.6277,
      "step": 73400
    },
    {
      "epoch": 0.81,
      "grad_norm": 4.8768744468688965,
      "learning_rate": 1.8290094920118094e-05,
      "loss": 2.5871,
      "step": 73450
    },
    {
      "epoch": 0.82,
      "grad_norm": 5.642852783203125,
      "learning_rate": 1.8184532758352968e-05,
      "loss": 2.5379,
      "step": 73500
    },
    {
      "epoch": 0.82,
      "grad_norm": 5.099608421325684,
      "learning_rate": 1.8079245631039154e-05,
      "loss": 2.5403,
      "step": 73550
    },
    {
      "epoch": 0.82,
      "grad_norm": 5.815806865692139,
      "learning_rate": 1.797423389211447e-05,
      "loss": 2.5712,
      "step": 73600
    },
    {
      "epoch": 0.82,
      "grad_norm": 4.7188029289245605,
      "learning_rate": 1.7869497894590958e-05,
      "loss": 2.609,
      "step": 73650
    },
    {
      "epoch": 0.82,
      "grad_norm": 5.4554362297058105,
      "learning_rate": 1.7765037990553723e-05,
      "loss": 2.5684,
      "step": 73700
    },
    {
      "epoch": 0.82,
      "grad_norm": 5.408052444458008,
      "learning_rate": 1.7660854531159765e-05,
      "loss": 2.6032,
      "step": 73750
    },
    {
      "epoch": 0.82,
      "grad_norm": 4.356540679931641,
      "learning_rate": 1.7556947866636732e-05,
      "loss": 2.6344,
      "step": 73800
    },
    {
      "epoch": 0.82,
      "grad_norm": 5.066736221313477,
      "learning_rate": 1.745331834628183e-05,
      "loss": 2.647,
      "step": 73850
    },
    {
      "epoch": 0.82,
      "grad_norm": 5.2489728927612305,
      "learning_rate": 1.7349966318460565e-05,
      "loss": 2.5774,
      "step": 73900
    },
    {
      "epoch": 0.82,
      "grad_norm": 5.3656005859375,
      "learning_rate": 1.724689213060565e-05,
      "loss": 2.4268,
      "step": 73950
    },
    {
      "epoch": 0.82,
      "grad_norm": 5.382472038269043,
      "learning_rate": 1.714409612921577e-05,
      "loss": 2.6417,
      "step": 74000
    },
    {
      "epoch": 0.82,
      "grad_norm": 5.097740650177002,
      "learning_rate": 1.704157865985446e-05,
      "loss": 2.5104,
      "step": 74050
    },
    {
      "epoch": 0.82,
      "grad_norm": 4.66810941696167,
      "learning_rate": 1.6939340067148945e-05,
      "loss": 2.6314,
      "step": 74100
    },
    {
      "epoch": 0.82,
      "grad_norm": 6.123903751373291,
      "learning_rate": 1.6837380694788886e-05,
      "loss": 2.5331,
      "step": 74150
    },
    {
      "epoch": 0.82,
      "grad_norm": 5.872149467468262,
      "learning_rate": 1.673570088552544e-05,
      "loss": 2.5468,
      "step": 74200
    },
    {
      "epoch": 0.82,
      "grad_norm": 4.637653350830078,
      "learning_rate": 1.6634300981169858e-05,
      "loss": 2.482,
      "step": 74250
    },
    {
      "epoch": 0.82,
      "grad_norm": 5.8766913414001465,
      "learning_rate": 1.6533181322592506e-05,
      "loss": 2.608,
      "step": 74300
    },
    {
      "epoch": 0.82,
      "grad_norm": 4.490416526794434,
      "learning_rate": 1.6432342249721644e-05,
      "loss": 2.5773,
      "step": 74350
    },
    {
      "epoch": 0.82,
      "grad_norm": 5.313449859619141,
      "learning_rate": 1.6331784101542314e-05,
      "loss": 2.5967,
      "step": 74400
    },
    {
      "epoch": 0.83,
      "grad_norm": 5.736847877502441,
      "learning_rate": 1.6231507216095165e-05,
      "loss": 2.6087,
      "step": 74450
    },
    {
      "epoch": 0.83,
      "grad_norm": 5.328598499298096,
      "learning_rate": 1.6131511930475386e-05,
      "loss": 2.5679,
      "step": 74500
    },
    {
      "epoch": 0.83,
      "grad_norm": 4.367438793182373,
      "learning_rate": 1.603179858083147e-05,
      "loss": 2.495,
      "step": 74550
    },
    {
      "epoch": 0.83,
      "grad_norm": 4.41094446182251,
      "learning_rate": 1.59323675023642e-05,
      "loss": 2.5749,
      "step": 74600
    },
    {
      "epoch": 0.83,
      "grad_norm": 4.8286566734313965,
      "learning_rate": 1.583321902932542e-05,
      "loss": 2.5806,
      "step": 74650
    },
    {
      "epoch": 0.83,
      "grad_norm": 4.310670375823975,
      "learning_rate": 1.5734353495016972e-05,
      "loss": 2.6343,
      "step": 74700
    },
    {
      "epoch": 0.83,
      "grad_norm": 6.212695598602295,
      "learning_rate": 1.5635771231789564e-05,
      "loss": 2.5887,
      "step": 74750
    },
    {
      "epoch": 0.83,
      "grad_norm": 4.295767784118652,
      "learning_rate": 1.5537472571041632e-05,
      "loss": 2.5137,
      "step": 74800
    },
    {
      "epoch": 0.83,
      "grad_norm": 5.152267932891846,
      "learning_rate": 1.5439457843218296e-05,
      "loss": 2.683,
      "step": 74850
    },
    {
      "epoch": 0.83,
      "grad_norm": 5.38249397277832,
      "learning_rate": 1.5341727377810057e-05,
      "loss": 2.5565,
      "step": 74900
    },
    {
      "epoch": 0.83,
      "grad_norm": 4.609240531921387,
      "learning_rate": 1.524428150335202e-05,
      "loss": 2.5312,
      "step": 74950
    },
    {
      "epoch": 0.83,
      "grad_norm": 8.245000839233398,
      "learning_rate": 1.51471205474224e-05,
      "loss": 2.6335,
      "step": 75000
    },
    {
      "epoch": 0.83,
      "grad_norm": 6.447003364562988,
      "learning_rate": 1.5050244836641824e-05,
      "loss": 2.5568,
      "step": 75050
    },
    {
      "epoch": 0.83,
      "grad_norm": 5.601855278015137,
      "learning_rate": 1.4953654696671803e-05,
      "loss": 2.5073,
      "step": 75100
    },
    {
      "epoch": 0.83,
      "grad_norm": 6.112193584442139,
      "learning_rate": 1.485735045221407e-05,
      "loss": 2.4316,
      "step": 75150
    },
    {
      "epoch": 0.83,
      "grad_norm": 4.9077606201171875,
      "learning_rate": 1.47613324270091e-05,
      "loss": 2.5999,
      "step": 75200
    },
    {
      "epoch": 0.83,
      "grad_norm": 4.854934215545654,
      "learning_rate": 1.4665600943835378e-05,
      "loss": 2.6205,
      "step": 75250
    },
    {
      "epoch": 0.83,
      "grad_norm": 6.032494068145752,
      "learning_rate": 1.4570156324507944e-05,
      "loss": 2.6094,
      "step": 75300
    },
    {
      "epoch": 0.84,
      "grad_norm": 5.131882667541504,
      "learning_rate": 1.4474998889877689e-05,
      "loss": 2.5278,
      "step": 75350
    },
    {
      "epoch": 0.84,
      "grad_norm": 5.608442783355713,
      "learning_rate": 1.4380128959829919e-05,
      "loss": 2.5813,
      "step": 75400
    },
    {
      "epoch": 0.84,
      "grad_norm": 5.487002849578857,
      "learning_rate": 1.4285546853283639e-05,
      "loss": 2.4209,
      "step": 75450
    },
    {
      "epoch": 0.84,
      "grad_norm": 4.697822570800781,
      "learning_rate": 1.4191252888190088e-05,
      "loss": 2.5251,
      "step": 75500
    },
    {
      "epoch": 0.84,
      "grad_norm": 5.489309310913086,
      "learning_rate": 1.4097247381532086e-05,
      "loss": 2.5596,
      "step": 75550
    },
    {
      "epoch": 0.84,
      "grad_norm": 6.367151737213135,
      "learning_rate": 1.400353064932256e-05,
      "loss": 2.5409,
      "step": 75600
    },
    {
      "epoch": 0.84,
      "grad_norm": 4.946759223937988,
      "learning_rate": 1.3910103006603825e-05,
      "loss": 2.477,
      "step": 75650
    },
    {
      "epoch": 0.84,
      "grad_norm": 5.678277492523193,
      "learning_rate": 1.3816964767446316e-05,
      "loss": 2.4527,
      "step": 75700
    },
    {
      "epoch": 0.84,
      "grad_norm": 6.542437553405762,
      "learning_rate": 1.3724116244947615e-05,
      "loss": 2.6315,
      "step": 75750
    },
    {
      "epoch": 0.84,
      "grad_norm": 5.304154872894287,
      "learning_rate": 1.3633406076809507e-05,
      "loss": 2.5684,
      "step": 75800
    },
    {
      "epoch": 0.84,
      "grad_norm": 5.378647804260254,
      "learning_rate": 1.3541132113182808e-05,
      "loss": 2.587,
      "step": 75850
    },
    {
      "epoch": 0.84,
      "grad_norm": 4.8275146484375,
      "learning_rate": 1.344914879346606e-05,
      "loss": 2.5758,
      "step": 75900
    },
    {
      "epoch": 0.84,
      "grad_norm": 5.953795909881592,
      "learning_rate": 1.3357456426874326e-05,
      "loss": 2.534,
      "step": 75950
    },
    {
      "epoch": 0.84,
      "grad_norm": 5.681417465209961,
      "learning_rate": 1.3266055321644767e-05,
      "loss": 2.6101,
      "step": 76000
    },
    {
      "epoch": 0.84,
      "grad_norm": 5.7463154792785645,
      "learning_rate": 1.3174945785035253e-05,
      "loss": 2.5352,
      "step": 76050
    },
    {
      "epoch": 0.84,
      "grad_norm": 4.927527904510498,
      "learning_rate": 1.3084128123323657e-05,
      "loss": 2.4799,
      "step": 76100
    },
    {
      "epoch": 0.84,
      "grad_norm": 4.909834861755371,
      "learning_rate": 1.2993602641806525e-05,
      "loss": 2.5069,
      "step": 76150
    },
    {
      "epoch": 0.84,
      "grad_norm": 5.693231582641602,
      "learning_rate": 1.2903369644798357e-05,
      "loss": 2.5283,
      "step": 76200
    },
    {
      "epoch": 0.85,
      "grad_norm": 5.905484199523926,
      "learning_rate": 1.2813429435630264e-05,
      "loss": 2.5575,
      "step": 76250
    },
    {
      "epoch": 0.85,
      "grad_norm": 5.348565101623535,
      "learning_rate": 1.2723782316649224e-05,
      "loss": 2.5726,
      "step": 76300
    },
    {
      "epoch": 0.85,
      "grad_norm": 5.240434646606445,
      "learning_rate": 1.2634428589216906e-05,
      "loss": 2.4842,
      "step": 76350
    },
    {
      "epoch": 0.85,
      "grad_norm": 5.337642669677734,
      "learning_rate": 1.2545368553708703e-05,
      "loss": 2.5963,
      "step": 76400
    },
    {
      "epoch": 0.85,
      "grad_norm": 5.730823040008545,
      "learning_rate": 1.2456602509512728e-05,
      "loss": 2.6206,
      "step": 76450
    },
    {
      "epoch": 0.85,
      "grad_norm": 5.830763339996338,
      "learning_rate": 1.2368130755028796e-05,
      "loss": 2.5722,
      "step": 76500
    },
    {
      "epoch": 0.85,
      "grad_norm": 5.059443473815918,
      "learning_rate": 1.2279953587667414e-05,
      "loss": 2.462,
      "step": 76550
    },
    {
      "epoch": 0.85,
      "grad_norm": 9.116705894470215,
      "learning_rate": 1.2192071303848806e-05,
      "loss": 2.5282,
      "step": 76600
    },
    {
      "epoch": 0.85,
      "grad_norm": 5.363910675048828,
      "learning_rate": 1.2104484199001909e-05,
      "loss": 2.5181,
      "step": 76650
    },
    {
      "epoch": 0.85,
      "grad_norm": 4.729151725769043,
      "learning_rate": 1.2017192567563351e-05,
      "loss": 2.5399,
      "step": 76700
    },
    {
      "epoch": 0.85,
      "grad_norm": 5.173400402069092,
      "learning_rate": 1.1930196702976514e-05,
      "loss": 2.5551,
      "step": 76750
    },
    {
      "epoch": 0.85,
      "grad_norm": 5.20831298828125,
      "learning_rate": 1.1843496897690486e-05,
      "loss": 2.5992,
      "step": 76800
    },
    {
      "epoch": 0.85,
      "grad_norm": 7.35823392868042,
      "learning_rate": 1.1757093443159117e-05,
      "loss": 2.528,
      "step": 76850
    },
    {
      "epoch": 0.85,
      "grad_norm": 4.642600059509277,
      "learning_rate": 1.1670986629840053e-05,
      "loss": 2.5585,
      "step": 76900
    },
    {
      "epoch": 0.85,
      "grad_norm": 4.8710713386535645,
      "learning_rate": 1.1585176747193704e-05,
      "loss": 2.6434,
      "step": 76950
    },
    {
      "epoch": 0.85,
      "grad_norm": 5.579038143157959,
      "learning_rate": 1.1499664083682348e-05,
      "loss": 2.6397,
      "step": 77000
    },
    {
      "epoch": 0.85,
      "grad_norm": 5.042369365692139,
      "learning_rate": 1.1414448926769061e-05,
      "loss": 2.4907,
      "step": 77050
    },
    {
      "epoch": 0.85,
      "grad_norm": 4.497950077056885,
      "learning_rate": 1.1329531562916862e-05,
      "loss": 2.5528,
      "step": 77100
    },
    {
      "epoch": 0.86,
      "grad_norm": 5.472546577453613,
      "learning_rate": 1.1244912277587672e-05,
      "loss": 2.4866,
      "step": 77150
    },
    {
      "epoch": 0.86,
      "grad_norm": 6.027166366577148,
      "learning_rate": 1.116059135524138e-05,
      "loss": 2.7039,
      "step": 77200
    },
    {
      "epoch": 0.86,
      "grad_norm": 5.426793098449707,
      "learning_rate": 1.1076569079334864e-05,
      "loss": 2.5509,
      "step": 77250
    },
    {
      "epoch": 0.86,
      "grad_norm": 5.257969379425049,
      "learning_rate": 1.099284573232111e-05,
      "loss": 2.4837,
      "step": 77300
    },
    {
      "epoch": 0.86,
      "grad_norm": 5.376690864562988,
      "learning_rate": 1.0909421595648161e-05,
      "loss": 2.5693,
      "step": 77350
    },
    {
      "epoch": 0.86,
      "grad_norm": 5.130237102508545,
      "learning_rate": 1.0826296949758275e-05,
      "loss": 2.4905,
      "step": 77400
    },
    {
      "epoch": 0.86,
      "grad_norm": 7.183712005615234,
      "learning_rate": 1.0743472074086824e-05,
      "loss": 2.5907,
      "step": 77450
    },
    {
      "epoch": 0.86,
      "grad_norm": 5.350920677185059,
      "learning_rate": 1.0660947247061615e-05,
      "loss": 2.6006,
      "step": 77500
    },
    {
      "epoch": 0.86,
      "grad_norm": 4.992347717285156,
      "learning_rate": 1.0578722746101632e-05,
      "loss": 2.5076,
      "step": 77550
    },
    {
      "epoch": 0.86,
      "grad_norm": 5.391441345214844,
      "learning_rate": 1.049679884761644e-05,
      "loss": 2.5131,
      "step": 77600
    },
    {
      "epoch": 0.86,
      "grad_norm": 5.465065002441406,
      "learning_rate": 1.0415175827004909e-05,
      "loss": 2.5767,
      "step": 77650
    },
    {
      "epoch": 0.86,
      "grad_norm": 6.124163627624512,
      "learning_rate": 1.0333853958654638e-05,
      "loss": 2.58,
      "step": 77700
    },
    {
      "epoch": 0.86,
      "grad_norm": 5.160919666290283,
      "learning_rate": 1.0252833515940685e-05,
      "loss": 2.4377,
      "step": 77750
    },
    {
      "epoch": 0.86,
      "grad_norm": 6.924820423126221,
      "learning_rate": 1.0172114771225017e-05,
      "loss": 2.5339,
      "step": 77800
    },
    {
      "epoch": 0.86,
      "grad_norm": 5.10518217086792,
      "learning_rate": 1.0091697995855186e-05,
      "loss": 2.5654,
      "step": 77850
    },
    {
      "epoch": 0.86,
      "grad_norm": 6.116388320922852,
      "learning_rate": 1.0011583460163831e-05,
      "loss": 2.497,
      "step": 77900
    },
    {
      "epoch": 0.86,
      "grad_norm": 5.105760097503662,
      "learning_rate": 9.931771433467408e-06,
      "loss": 2.5719,
      "step": 77950
    },
    {
      "epoch": 0.86,
      "grad_norm": 4.459804534912109,
      "learning_rate": 9.852262184065575e-06,
      "loss": 2.5743,
      "step": 78000
    },
    {
      "epoch": 0.87,
      "grad_norm": 5.1481828689575195,
      "learning_rate": 9.773055979240054e-06,
      "loss": 2.5321,
      "step": 78050
    },
    {
      "epoch": 0.87,
      "grad_norm": 4.6464152336120605,
      "learning_rate": 9.6941530852539e-06,
      "loss": 2.5022,
      "step": 78100
    },
    {
      "epoch": 0.87,
      "grad_norm": 5.2632222175598145,
      "learning_rate": 9.615553767350538e-06,
      "loss": 2.5179,
      "step": 78150
    },
    {
      "epoch": 0.87,
      "grad_norm": 5.13355827331543,
      "learning_rate": 9.53725828975287e-06,
      "loss": 2.4673,
      "step": 78200
    },
    {
      "epoch": 0.87,
      "grad_norm": 4.451653480529785,
      "learning_rate": 9.459266915662379e-06,
      "loss": 2.6681,
      "step": 78250
    },
    {
      "epoch": 0.87,
      "grad_norm": 5.235838413238525,
      "learning_rate": 9.381579907258309e-06,
      "loss": 2.6428,
      "step": 78300
    },
    {
      "epoch": 0.87,
      "grad_norm": 4.433870792388916,
      "learning_rate": 9.304197525696679e-06,
      "loss": 2.4684,
      "step": 78350
    },
    {
      "epoch": 0.87,
      "grad_norm": 6.210980415344238,
      "learning_rate": 9.227120031109482e-06,
      "loss": 2.5359,
      "step": 78400
    },
    {
      "epoch": 0.87,
      "grad_norm": 6.603292942047119,
      "learning_rate": 9.150347682603811e-06,
      "loss": 2.5951,
      "step": 78450
    },
    {
      "epoch": 0.87,
      "grad_norm": 4.311337947845459,
      "learning_rate": 9.073880738260942e-06,
      "loss": 2.5774,
      "step": 78500
    },
    {
      "epoch": 0.87,
      "grad_norm": 5.452199935913086,
      "learning_rate": 8.997719455135501e-06,
      "loss": 2.4391,
      "step": 78550
    },
    {
      "epoch": 0.87,
      "grad_norm": 6.8268656730651855,
      "learning_rate": 8.921864089254584e-06,
      "loss": 2.5437,
      "step": 78600
    },
    {
      "epoch": 0.87,
      "grad_norm": 5.306812286376953,
      "learning_rate": 8.846314895616902e-06,
      "loss": 2.585,
      "step": 78650
    },
    {
      "epoch": 0.87,
      "grad_norm": 5.723698139190674,
      "learning_rate": 8.771072128191926e-06,
      "loss": 2.56,
      "step": 78700
    },
    {
      "epoch": 0.87,
      "grad_norm": 4.836296558380127,
      "learning_rate": 8.696136039919035e-06,
      "loss": 2.4498,
      "step": 78750
    },
    {
      "epoch": 0.87,
      "grad_norm": 4.454384803771973,
      "learning_rate": 8.621506882706665e-06,
      "loss": 2.5153,
      "step": 78800
    },
    {
      "epoch": 0.87,
      "grad_norm": 6.231645584106445,
      "learning_rate": 8.54718490743145e-06,
      "loss": 2.4977,
      "step": 78850
    },
    {
      "epoch": 0.87,
      "grad_norm": 7.251499652862549,
      "learning_rate": 8.473170363937389e-06,
      "loss": 2.5264,
      "step": 78900
    },
    {
      "epoch": 0.88,
      "grad_norm": 5.47197961807251,
      "learning_rate": 8.399463501034998e-06,
      "loss": 2.5653,
      "step": 78950
    },
    {
      "epoch": 0.88,
      "grad_norm": 5.263489246368408,
      "learning_rate": 8.326064566500514e-06,
      "loss": 2.5783,
      "step": 79000
    },
    {
      "epoch": 0.88,
      "grad_norm": 5.1917219161987305,
      "learning_rate": 8.252973807074981e-06,
      "loss": 2.5063,
      "step": 79050
    },
    {
      "epoch": 0.88,
      "grad_norm": 5.902822494506836,
      "learning_rate": 8.180191468463495e-06,
      "loss": 2.5589,
      "step": 79100
    },
    {
      "epoch": 0.88,
      "grad_norm": 5.971536636352539,
      "learning_rate": 8.107717795334358e-06,
      "loss": 2.5798,
      "step": 79150
    },
    {
      "epoch": 0.88,
      "grad_norm": 5.564624786376953,
      "learning_rate": 8.035553031318243e-06,
      "loss": 2.5032,
      "step": 79200
    },
    {
      "epoch": 0.88,
      "grad_norm": 7.015138149261475,
      "learning_rate": 7.963697419007332e-06,
      "loss": 2.5977,
      "step": 79250
    },
    {
      "epoch": 0.88,
      "grad_norm": 5.202609539031982,
      "learning_rate": 7.892151199954666e-06,
      "loss": 2.4845,
      "step": 79300
    },
    {
      "epoch": 0.88,
      "grad_norm": 4.632216453552246,
      "learning_rate": 7.820914614673058e-06,
      "loss": 2.4535,
      "step": 79350
    },
    {
      "epoch": 0.88,
      "grad_norm": 4.653558254241943,
      "learning_rate": 7.749987902634604e-06,
      "loss": 2.5467,
      "step": 79400
    },
    {
      "epoch": 0.88,
      "grad_norm": 5.157925128936768,
      "learning_rate": 7.679371302269578e-06,
      "loss": 2.4726,
      "step": 79450
    },
    {
      "epoch": 0.88,
      "grad_norm": 6.130363941192627,
      "learning_rate": 7.609065050965891e-06,
      "loss": 2.5956,
      "step": 79500
    },
    {
      "epoch": 0.88,
      "grad_norm": 5.374567985534668,
      "learning_rate": 7.539069385068043e-06,
      "loss": 2.5357,
      "step": 79550
    },
    {
      "epoch": 0.88,
      "grad_norm": 5.50112247467041,
      "learning_rate": 7.469384539876589e-06,
      "loss": 2.5185,
      "step": 79600
    },
    {
      "epoch": 0.88,
      "grad_norm": 4.573812484741211,
      "learning_rate": 7.400010749647101e-06,
      "loss": 2.5178,
      "step": 79650
    },
    {
      "epoch": 0.88,
      "grad_norm": 5.003163814544678,
      "learning_rate": 7.330948247589586e-06,
      "loss": 2.6177,
      "step": 79700
    },
    {
      "epoch": 0.88,
      "grad_norm": 4.302044868469238,
      "learning_rate": 7.262197265867532e-06,
      "loss": 2.3993,
      "step": 79750
    },
    {
      "epoch": 0.88,
      "grad_norm": 6.77066707611084,
      "learning_rate": 7.195123763548639e-06,
      "loss": 2.563,
      "step": 79800
    },
    {
      "epoch": 0.89,
      "grad_norm": 6.391753196716309,
      "learning_rate": 7.126990272920198e-06,
      "loss": 2.5765,
      "step": 79850
    },
    {
      "epoch": 0.89,
      "grad_norm": 5.601807594299316,
      "learning_rate": 7.059168988261244e-06,
      "loss": 2.5157,
      "step": 79900
    },
    {
      "epoch": 0.89,
      "grad_norm": 4.471221923828125,
      "learning_rate": 6.991660137562839e-06,
      "loss": 2.6151,
      "step": 79950
    },
    {
      "epoch": 0.89,
      "grad_norm": 5.325886249542236,
      "learning_rate": 6.924463947765636e-06,
      "loss": 2.5217,
      "step": 80000
    },
    {
      "epoch": 0.89,
      "grad_norm": 5.799530982971191,
      "learning_rate": 6.857580644759332e-06,
      "loss": 2.5052,
      "step": 80050
    },
    {
      "epoch": 0.89,
      "grad_norm": 4.483170509338379,
      "learning_rate": 6.791010453381696e-06,
      "loss": 2.4436,
      "step": 80100
    },
    {
      "epoch": 0.89,
      "grad_norm": 5.7187981605529785,
      "learning_rate": 6.7247535974180655e-06,
      "loss": 2.53,
      "step": 80150
    },
    {
      "epoch": 0.89,
      "grad_norm": 4.269189357757568,
      "learning_rate": 6.658810299600327e-06,
      "loss": 2.5459,
      "step": 80200
    },
    {
      "epoch": 0.89,
      "grad_norm": 5.858389377593994,
      "learning_rate": 6.5931807816064005e-06,
      "loss": 2.4309,
      "step": 80250
    },
    {
      "epoch": 0.89,
      "grad_norm": 5.224058628082275,
      "learning_rate": 6.5278652640592855e-06,
      "loss": 2.6175,
      "step": 80300
    },
    {
      "epoch": 0.89,
      "grad_norm": 5.791477680206299,
      "learning_rate": 6.462863966526545e-06,
      "loss": 2.5597,
      "step": 80350
    },
    {
      "epoch": 0.89,
      "grad_norm": 4.741706848144531,
      "learning_rate": 6.398177107519332e-06,
      "loss": 2.4946,
      "step": 80400
    },
    {
      "epoch": 0.89,
      "grad_norm": 5.0054450035095215,
      "learning_rate": 6.333804904491858e-06,
      "loss": 2.5819,
      "step": 80450
    },
    {
      "epoch": 0.89,
      "grad_norm": 5.472426891326904,
      "learning_rate": 6.26974757384049e-06,
      "loss": 2.4771,
      "step": 80500
    },
    {
      "epoch": 0.89,
      "grad_norm": 6.89616060256958,
      "learning_rate": 6.206005330903208e-06,
      "loss": 2.4928,
      "step": 80550
    },
    {
      "epoch": 0.89,
      "grad_norm": 4.742382526397705,
      "learning_rate": 6.142578389958686e-06,
      "loss": 2.6554,
      "step": 80600
    },
    {
      "epoch": 0.89,
      "grad_norm": 6.414988040924072,
      "learning_rate": 6.0794669642256885e-06,
      "loss": 2.4839,
      "step": 80650
    },
    {
      "epoch": 0.89,
      "grad_norm": 5.960868835449219,
      "learning_rate": 6.0166712658623725e-06,
      "loss": 2.4859,
      "step": 80700
    },
    {
      "epoch": 0.9,
      "grad_norm": 5.40272331237793,
      "learning_rate": 5.9541915059655e-06,
      "loss": 2.5507,
      "step": 80750
    },
    {
      "epoch": 0.9,
      "grad_norm": 5.389658451080322,
      "learning_rate": 5.892027894569741e-06,
      "loss": 2.587,
      "step": 80800
    },
    {
      "epoch": 0.9,
      "grad_norm": 5.898835182189941,
      "learning_rate": 5.830180640647031e-06,
      "loss": 2.5249,
      "step": 80850
    },
    {
      "epoch": 0.9,
      "grad_norm": 4.355838298797607,
      "learning_rate": 5.768649952105798e-06,
      "loss": 2.587,
      "step": 80900
    },
    {
      "epoch": 0.9,
      "grad_norm": 4.786905288696289,
      "learning_rate": 5.707436035790281e-06,
      "loss": 2.5946,
      "step": 80950
    },
    {
      "epoch": 0.9,
      "grad_norm": 5.000050067901611,
      "learning_rate": 5.6465390974798746e-06,
      "loss": 2.5569,
      "step": 81000
    },
    {
      "epoch": 0.9,
      "grad_norm": 4.742411136627197,
      "learning_rate": 5.58595934188838e-06,
      "loss": 2.6298,
      "step": 81050
    },
    {
      "epoch": 0.9,
      "grad_norm": 4.782742023468018,
      "learning_rate": 5.525696972663352e-06,
      "loss": 2.4946,
      "step": 81100
    },
    {
      "epoch": 0.9,
      "grad_norm": 5.483613014221191,
      "learning_rate": 5.46575219238541e-06,
      "loss": 2.5737,
      "step": 81150
    },
    {
      "epoch": 0.9,
      "grad_norm": 5.547824859619141,
      "learning_rate": 5.406125202567547e-06,
      "loss": 2.4482,
      "step": 81200
    },
    {
      "epoch": 0.9,
      "grad_norm": 5.587399959564209,
      "learning_rate": 5.346816203654448e-06,
      "loss": 2.4292,
      "step": 81250
    },
    {
      "epoch": 0.9,
      "grad_norm": 5.2117438316345215,
      "learning_rate": 5.28782539502185e-06,
      "loss": 2.5498,
      "step": 81300
    },
    {
      "epoch": 0.9,
      "grad_norm": 4.911371231079102,
      "learning_rate": 5.229152974975838e-06,
      "loss": 2.5179,
      "step": 81350
    },
    {
      "epoch": 0.9,
      "grad_norm": 4.5971999168396,
      "learning_rate": 5.170799140752158e-06,
      "loss": 2.5086,
      "step": 81400
    },
    {
      "epoch": 0.9,
      "grad_norm": 4.789346694946289,
      "learning_rate": 5.112764088515643e-06,
      "loss": 2.5681,
      "step": 81450
    },
    {
      "epoch": 0.9,
      "grad_norm": 5.9746503829956055,
      "learning_rate": 5.055048013359442e-06,
      "loss": 2.5565,
      "step": 81500
    },
    {
      "epoch": 0.9,
      "grad_norm": 5.581201076507568,
      "learning_rate": 4.9976511093044355e-06,
      "loss": 2.6655,
      "step": 81550
    },
    {
      "epoch": 0.9,
      "grad_norm": 5.030186176300049,
      "learning_rate": 4.940573569298568e-06,
      "loss": 2.6387,
      "step": 81600
    },
    {
      "epoch": 0.91,
      "grad_norm": 5.522227764129639,
      "learning_rate": 4.883815585216201e-06,
      "loss": 2.4928,
      "step": 81650
    },
    {
      "epoch": 0.91,
      "grad_norm": 5.1960344314575195,
      "learning_rate": 4.827377347857454e-06,
      "loss": 2.5224,
      "step": 81700
    },
    {
      "epoch": 0.91,
      "grad_norm": 5.583174705505371,
      "learning_rate": 4.7712590469475735e-06,
      "loss": 2.4716,
      "step": 81750
    },
    {
      "epoch": 0.91,
      "grad_norm": 4.959489345550537,
      "learning_rate": 4.715460871136268e-06,
      "loss": 2.5839,
      "step": 81800
    },
    {
      "epoch": 0.91,
      "grad_norm": 5.3739542961120605,
      "learning_rate": 4.661089424987708e-06,
      "loss": 2.4676,
      "step": 81850
    },
    {
      "epoch": 0.91,
      "grad_norm": 5.230461597442627,
      "learning_rate": 4.605925649213516e-06,
      "loss": 2.5171,
      "step": 81900
    },
    {
      "epoch": 0.91,
      "grad_norm": 6.518977642059326,
      "learning_rate": 4.551082554329877e-06,
      "loss": 2.4782,
      "step": 81950
    },
    {
      "epoch": 0.91,
      "grad_norm": 6.236433029174805,
      "learning_rate": 4.496560324699772e-06,
      "loss": 2.5228,
      "step": 82000
    },
    {
      "epoch": 0.91,
      "grad_norm": 4.416147232055664,
      "learning_rate": 4.442359143607488e-06,
      "loss": 2.6116,
      "step": 82050
    },
    {
      "epoch": 0.91,
      "grad_norm": 5.622953414916992,
      "learning_rate": 4.388479193258088e-06,
      "loss": 2.4052,
      "step": 82100
    },
    {
      "epoch": 0.91,
      "grad_norm": 4.993776798248291,
      "learning_rate": 4.334920654776775e-06,
      "loss": 2.5948,
      "step": 82150
    },
    {
      "epoch": 0.91,
      "grad_norm": 4.830969333648682,
      "learning_rate": 4.2816837082082835e-06,
      "loss": 2.5543,
      "step": 82200
    },
    {
      "epoch": 0.91,
      "grad_norm": 4.715949058532715,
      "learning_rate": 4.228768532516236e-06,
      "loss": 2.5163,
      "step": 82250
    },
    {
      "epoch": 0.91,
      "grad_norm": 5.530038833618164,
      "learning_rate": 4.176175305582608e-06,
      "loss": 2.4745,
      "step": 82300
    },
    {
      "epoch": 0.91,
      "grad_norm": 5.377774715423584,
      "learning_rate": 4.123904204207107e-06,
      "loss": 2.5355,
      "step": 82350
    },
    {
      "epoch": 0.91,
      "grad_norm": 6.267026901245117,
      "learning_rate": 4.071955404106542e-06,
      "loss": 2.5235,
      "step": 82400
    },
    {
      "epoch": 0.91,
      "grad_norm": 5.261455535888672,
      "learning_rate": 4.020329079914253e-06,
      "loss": 2.6404,
      "step": 82450
    },
    {
      "epoch": 0.91,
      "grad_norm": 5.198992729187012,
      "learning_rate": 3.969025405179583e-06,
      "loss": 2.5232,
      "step": 82500
    },
    {
      "epoch": 0.92,
      "grad_norm": 5.214979648590088,
      "learning_rate": 3.918044552367162e-06,
      "loss": 2.4884,
      "step": 82550
    },
    {
      "epoch": 0.92,
      "grad_norm": 4.916982650756836,
      "learning_rate": 3.867386692856511e-06,
      "loss": 2.4918,
      "step": 82600
    },
    {
      "epoch": 0.92,
      "grad_norm": 5.684042930603027,
      "learning_rate": 3.817051996941245e-06,
      "loss": 2.5103,
      "step": 82650
    },
    {
      "epoch": 0.92,
      "grad_norm": 4.156066417694092,
      "learning_rate": 3.7670406338287332e-06,
      "loss": 2.4773,
      "step": 82700
    },
    {
      "epoch": 0.92,
      "grad_norm": 4.400490760803223,
      "learning_rate": 3.717352771639304e-06,
      "loss": 2.5679,
      "step": 82750
    },
    {
      "epoch": 0.92,
      "grad_norm": 4.964176654815674,
      "learning_rate": 3.667988577405912e-06,
      "loss": 2.4718,
      "step": 82800
    },
    {
      "epoch": 0.92,
      "grad_norm": 4.562291145324707,
      "learning_rate": 3.6189482170733256e-06,
      "loss": 2.5083,
      "step": 82850
    },
    {
      "epoch": 0.92,
      "grad_norm": 5.713414192199707,
      "learning_rate": 3.570231855497819e-06,
      "loss": 2.5654,
      "step": 82900
    },
    {
      "epoch": 0.92,
      "grad_norm": 4.2873311042785645,
      "learning_rate": 3.5218396564463908e-06,
      "loss": 2.5527,
      "step": 82950
    },
    {
      "epoch": 0.92,
      "grad_norm": 4.6480841636657715,
      "learning_rate": 3.4737717825964243e-06,
      "loss": 2.5647,
      "step": 83000
    }
  ],
  "logging_steps": 50,
  "max_steps": 90183,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "total_flos": 2.9846013739008e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}