{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.999290780141844,
  "eval_steps": 500,
  "global_step": 7049,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0070921985815602835,
      "grad_norm": 5.779086589813232,
      "learning_rate": 5.6657223796034e-06,
      "loss": 0.8629,
      "step": 10
    },
    {
      "epoch": 0.014184397163120567,
      "grad_norm": 4.5390448570251465,
      "learning_rate": 1.13314447592068e-05,
      "loss": 0.6688,
      "step": 20
    },
    {
      "epoch": 0.02127659574468085,
      "grad_norm": 3.5783498287200928,
      "learning_rate": 1.69971671388102e-05,
      "loss": 0.4349,
      "step": 30
    },
    {
      "epoch": 0.028368794326241134,
      "grad_norm": 2.7119619846343994,
      "learning_rate": 2.26628895184136e-05,
      "loss": 0.3312,
      "step": 40
    },
    {
      "epoch": 0.03546099290780142,
      "grad_norm": 1.4443682432174683,
      "learning_rate": 2.8328611898017e-05,
      "loss": 0.275,
      "step": 50
    },
    {
      "epoch": 0.0425531914893617,
      "grad_norm": 1.6015079021453857,
      "learning_rate": 3.39943342776204e-05,
      "loss": 0.2297,
      "step": 60
    },
    {
      "epoch": 0.04964539007092199,
      "grad_norm": 1.588321566581726,
      "learning_rate": 3.96600566572238e-05,
      "loss": 0.1544,
      "step": 70
    },
    {
      "epoch": 0.05673758865248227,
      "grad_norm": 2.757812738418579,
      "learning_rate": 4.53257790368272e-05,
      "loss": 0.1773,
      "step": 80
    },
    {
      "epoch": 0.06382978723404255,
      "grad_norm": 2.5903780460357666,
      "learning_rate": 5.09915014164306e-05,
      "loss": 0.1427,
      "step": 90
    },
    {
      "epoch": 0.07092198581560284,
      "grad_norm": 0.9456828236579895,
      "learning_rate": 5.6657223796034e-05,
      "loss": 0.1267,
      "step": 100
    },
    {
      "epoch": 0.07801418439716312,
      "grad_norm": 1.6622958183288574,
      "learning_rate": 6.23229461756374e-05,
      "loss": 0.13,
      "step": 110
    },
    {
      "epoch": 0.0851063829787234,
      "grad_norm": 1.7187730073928833,
      "learning_rate": 6.79886685552408e-05,
      "loss": 0.1246,
      "step": 120
    },
    {
      "epoch": 0.09219858156028368,
      "grad_norm": 1.5915329456329346,
      "learning_rate": 7.36543909348442e-05,
      "loss": 0.0941,
      "step": 130
    },
    {
      "epoch": 0.09929078014184398,
      "grad_norm": 1.4188594818115234,
      "learning_rate": 7.93201133144476e-05,
      "loss": 0.0948,
      "step": 140
    },
    {
      "epoch": 0.10638297872340426,
      "grad_norm": 1.5538837909698486,
      "learning_rate": 8.4985835694051e-05,
      "loss": 0.108,
      "step": 150
    },
    {
      "epoch": 0.11347517730496454,
      "grad_norm": 1.3440709114074707,
      "learning_rate": 9.06515580736544e-05,
      "loss": 0.0956,
      "step": 160
    },
    {
      "epoch": 0.12056737588652482,
      "grad_norm": 0.9447225332260132,
      "learning_rate": 9.631728045325779e-05,
      "loss": 0.0904,
      "step": 170
    },
    {
      "epoch": 0.1276595744680851,
      "grad_norm": 0.77945876121521,
      "learning_rate": 0.0001019830028328612,
      "loss": 0.0892,
      "step": 180
    },
    {
      "epoch": 0.1347517730496454,
      "grad_norm": 1.136725664138794,
      "learning_rate": 0.00010764872521246458,
      "loss": 0.0899,
      "step": 190
    },
    {
      "epoch": 0.14184397163120568,
      "grad_norm": 1.6664468050003052,
      "learning_rate": 0.000113314447592068,
      "loss": 0.0944,
      "step": 200
    },
    {
      "epoch": 0.14893617021276595,
      "grad_norm": 0.7696484923362732,
      "learning_rate": 0.00011898016997167138,
      "loss": 0.0924,
      "step": 210
    },
    {
      "epoch": 0.15602836879432624,
      "grad_norm": 0.918065071105957,
      "learning_rate": 0.0001246458923512748,
      "loss": 0.0853,
      "step": 220
    },
    {
      "epoch": 0.16312056737588654,
      "grad_norm": 1.1535277366638184,
      "learning_rate": 0.0001303116147308782,
      "loss": 0.0739,
      "step": 230
    },
    {
      "epoch": 0.1702127659574468,
      "grad_norm": 1.3063223361968994,
      "learning_rate": 0.0001359773371104816,
      "loss": 0.0791,
      "step": 240
    },
    {
      "epoch": 0.1773049645390071,
      "grad_norm": 0.9549968838691711,
      "learning_rate": 0.00014164305949008498,
      "loss": 0.075,
      "step": 250
    },
    {
      "epoch": 0.18439716312056736,
      "grad_norm": 0.7243878841400146,
      "learning_rate": 0.0001473087818696884,
      "loss": 0.0802,
      "step": 260
    },
    {
      "epoch": 0.19148936170212766,
      "grad_norm": 0.5810666084289551,
      "learning_rate": 0.0001529745042492918,
      "loss": 0.0748,
      "step": 270
    },
    {
      "epoch": 0.19858156028368795,
      "grad_norm": 0.7565228343009949,
      "learning_rate": 0.0001586402266288952,
      "loss": 0.0841,
      "step": 280
    },
    {
      "epoch": 0.20567375886524822,
      "grad_norm": 0.8032036423683167,
      "learning_rate": 0.00016430594900849858,
      "loss": 0.0621,
      "step": 290
    },
    {
      "epoch": 0.2127659574468085,
      "grad_norm": 0.8412207365036011,
      "learning_rate": 0.000169971671388102,
      "loss": 0.0815,
      "step": 300
    },
    {
      "epoch": 0.2198581560283688,
      "grad_norm": 0.7461270093917847,
      "learning_rate": 0.00017563739376770537,
      "loss": 0.0836,
      "step": 310
    },
    {
      "epoch": 0.22695035460992907,
      "grad_norm": 0.686876118183136,
      "learning_rate": 0.0001813031161473088,
      "loss": 0.0799,
      "step": 320
    },
    {
      "epoch": 0.23404255319148937,
      "grad_norm": 1.616554617881775,
      "learning_rate": 0.0001869688385269122,
      "loss": 0.0819,
      "step": 330
    },
    {
      "epoch": 0.24113475177304963,
      "grad_norm": 1.098061203956604,
      "learning_rate": 0.00019263456090651558,
      "loss": 0.0649,
      "step": 340
    },
    {
      "epoch": 0.24822695035460993,
      "grad_norm": 0.9771316647529602,
      "learning_rate": 0.00019830028328611898,
      "loss": 0.0675,
      "step": 350
    },
    {
      "epoch": 0.2553191489361702,
      "grad_norm": 0.97850102186203,
      "learning_rate": 0.0001999994606948725,
      "loss": 0.0645,
      "step": 360
    },
    {
      "epoch": 0.2624113475177305,
      "grad_norm": 0.7112658023834229,
      "learning_rate": 0.0001999968192143738,
      "loss": 0.0694,
      "step": 370
    },
    {
      "epoch": 0.2695035460992908,
      "grad_norm": 0.708350419998169,
      "learning_rate": 0.00019999197656053288,
      "loss": 0.0624,
      "step": 380
    },
    {
      "epoch": 0.2765957446808511,
      "grad_norm": 0.5273390412330627,
      "learning_rate": 0.00019998493283994825,
      "loss": 0.074,
      "step": 390
    },
    {
      "epoch": 0.28368794326241137,
      "grad_norm": 1.1027687788009644,
      "learning_rate": 0.0001999756882076694,
      "loss": 0.0572,
      "step": 400
    },
    {
      "epoch": 0.2907801418439716,
      "grad_norm": 1.0205180644989014,
      "learning_rate": 0.00019996424286719317,
      "loss": 0.0654,
      "step": 410
    },
    {
      "epoch": 0.2978723404255319,
      "grad_norm": 0.5349808931350708,
      "learning_rate": 0.00019995059707045935,
      "loss": 0.0682,
      "step": 420
    },
    {
      "epoch": 0.3049645390070922,
      "grad_norm": 0.6187090277671814,
      "learning_rate": 0.00019993475111784514,
      "loss": 0.0611,
      "step": 430
    },
    {
      "epoch": 0.3120567375886525,
      "grad_norm": 0.5722566843032837,
      "learning_rate": 0.00019991670535815853,
      "loss": 0.0635,
      "step": 440
    },
    {
      "epoch": 0.3191489361702128,
      "grad_norm": 1.1226184368133545,
      "learning_rate": 0.00019989646018863058,
      "loss": 0.0632,
      "step": 450
    },
    {
      "epoch": 0.3262411347517731,
      "grad_norm": 0.814766526222229,
      "learning_rate": 0.00019987401605490676,
      "loss": 0.0546,
      "step": 460
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 0.8665221929550171,
      "learning_rate": 0.00019984937345103707,
      "loss": 0.0591,
      "step": 470
    },
    {
      "epoch": 0.3404255319148936,
      "grad_norm": 0.5462458729743958,
      "learning_rate": 0.0001998225329194652,
      "loss": 0.0543,
      "step": 480
    },
    {
      "epoch": 0.3475177304964539,
      "grad_norm": 0.3314288556575775,
      "learning_rate": 0.00019979349505101654,
      "loss": 0.0538,
      "step": 490
    },
    {
      "epoch": 0.3546099290780142,
      "grad_norm": 0.6016983985900879,
      "learning_rate": 0.00019976226048488526,
      "loss": 0.0417,
      "step": 500
    },
    {
      "epoch": 0.3617021276595745,
      "grad_norm": 0.5387014150619507,
      "learning_rate": 0.0001997288299086202,
      "loss": 0.0507,
      "step": 510
    },
    {
      "epoch": 0.36879432624113473,
      "grad_norm": 0.6287176609039307,
      "learning_rate": 0.00019969320405810966,
      "loss": 0.0497,
      "step": 520
    },
    {
      "epoch": 0.375886524822695,
      "grad_norm": 0.7354361414909363,
      "learning_rate": 0.0001996553837175653,
      "loss": 0.0556,
      "step": 530
    },
    {
      "epoch": 0.3829787234042553,
      "grad_norm": 0.5547359585762024,
      "learning_rate": 0.0001996153697195049,
      "loss": 0.056,
      "step": 540
    },
    {
      "epoch": 0.3900709219858156,
      "grad_norm": 0.612255871295929,
      "learning_rate": 0.00019957316294473389,
      "loss": 0.0504,
      "step": 550
    },
    {
      "epoch": 0.3971631205673759,
      "grad_norm": 0.7296609878540039,
      "learning_rate": 0.0001995287643223261,
      "loss": 0.0603,
      "step": 560
    },
    {
      "epoch": 0.40425531914893614,
      "grad_norm": 0.53786700963974,
      "learning_rate": 0.0001994821748296033,
      "loss": 0.0609,
      "step": 570
    },
    {
      "epoch": 0.41134751773049644,
      "grad_norm": 0.49047139286994934,
      "learning_rate": 0.00019943339549211359,
      "loss": 0.0581,
      "step": 580
    },
    {
      "epoch": 0.41843971631205673,
      "grad_norm": 0.4338454604148865,
      "learning_rate": 0.00019938242738360887,
      "loss": 0.0516,
      "step": 590
    },
    {
      "epoch": 0.425531914893617,
      "grad_norm": 0.7479408979415894,
      "learning_rate": 0.00019932927162602125,
      "loss": 0.0519,
      "step": 600
    },
    {
      "epoch": 0.4326241134751773,
      "grad_norm": 0.6909820437431335,
      "learning_rate": 0.00019927392938943835,
      "loss": 0.0577,
      "step": 610
    },
    {
      "epoch": 0.4397163120567376,
      "grad_norm": 0.44178247451782227,
      "learning_rate": 0.00019921640189207745,
      "loss": 0.0492,
      "step": 620
    },
    {
      "epoch": 0.44680851063829785,
      "grad_norm": 0.42631322145462036,
      "learning_rate": 0.00019915669040025872,
      "loss": 0.0442,
      "step": 630
    },
    {
      "epoch": 0.45390070921985815,
      "grad_norm": 0.36178159713745117,
      "learning_rate": 0.00019909479622837753,
      "loss": 0.0418,
      "step": 640
    },
    {
      "epoch": 0.46099290780141844,
      "grad_norm": 0.31890127062797546,
      "learning_rate": 0.00019903072073887507,
      "loss": 0.042,
      "step": 650
    },
    {
      "epoch": 0.46808510638297873,
      "grad_norm": 0.6327568292617798,
      "learning_rate": 0.00019896446534220893,
      "loss": 0.0519,
      "step": 660
    },
    {
      "epoch": 0.475177304964539,
      "grad_norm": 0.35806578397750854,
      "learning_rate": 0.00019889603149682156,
      "loss": 0.0396,
      "step": 670
    },
    {
      "epoch": 0.48226950354609927,
      "grad_norm": 0.7298017144203186,
      "learning_rate": 0.00019882542070910847,
      "loss": 0.049,
      "step": 680
    },
    {
      "epoch": 0.48936170212765956,
      "grad_norm": 0.497172474861145,
      "learning_rate": 0.00019875263453338494,
      "loss": 0.0426,
      "step": 690
    },
    {
      "epoch": 0.49645390070921985,
      "grad_norm": 0.7310464382171631,
      "learning_rate": 0.00019867767457185177,
      "loss": 0.0493,
      "step": 700
    },
    {
      "epoch": 0.5035460992907801,
      "grad_norm": 0.6526608467102051,
      "learning_rate": 0.0001986005424745602,
      "loss": 0.0516,
      "step": 710
    },
    {
      "epoch": 0.5106382978723404,
      "grad_norm": 0.6688535213470459,
      "learning_rate": 0.00019852123993937537,
      "loss": 0.0486,
      "step": 720
    },
    {
      "epoch": 0.5177304964539007,
      "grad_norm": 0.4904952049255371,
      "learning_rate": 0.0001984397687119391,
      "loss": 0.0528,
      "step": 730
    },
    {
      "epoch": 0.524822695035461,
      "grad_norm": 0.5391408801078796,
      "learning_rate": 0.00019835613058563132,
      "loss": 0.0492,
      "step": 740
    },
    {
      "epoch": 0.5319148936170213,
      "grad_norm": 0.6295509338378906,
      "learning_rate": 0.00019827032740153073,
      "loss": 0.0489,
      "step": 750
    },
    {
      "epoch": 0.5390070921985816,
      "grad_norm": 0.5825513005256653,
      "learning_rate": 0.0001981823610483743,
      "loss": 0.0476,
      "step": 760
    },
    {
      "epoch": 0.5460992907801419,
      "grad_norm": 0.45764580368995667,
      "learning_rate": 0.0001980922334625155,
      "loss": 0.0427,
      "step": 770
    },
    {
      "epoch": 0.5531914893617021,
      "grad_norm": 0.6424888372421265,
      "learning_rate": 0.00019799994662788177,
      "loss": 0.0507,
      "step": 780
    },
    {
      "epoch": 0.5602836879432624,
      "grad_norm": 0.5487046837806702,
      "learning_rate": 0.000197905502575931,
      "loss": 0.0396,
      "step": 790
    },
    {
      "epoch": 0.5673758865248227,
      "grad_norm": 0.4865322709083557,
      "learning_rate": 0.00019780890338560658,
      "loss": 0.0423,
      "step": 800
    },
    {
      "epoch": 0.574468085106383,
      "grad_norm": 0.3333698809146881,
      "learning_rate": 0.00019771015118329177,
      "loss": 0.0386,
      "step": 810
    },
    {
      "epoch": 0.5815602836879432,
      "grad_norm": 0.48208341002464294,
      "learning_rate": 0.0001976092481427629,
      "loss": 0.0626,
      "step": 820
    },
    {
      "epoch": 0.5886524822695035,
      "grad_norm": 0.41743791103363037,
      "learning_rate": 0.0001975061964851414,
      "loss": 0.0452,
      "step": 830
    },
    {
      "epoch": 0.5957446808510638,
      "grad_norm": 0.6599062085151672,
      "learning_rate": 0.00019740099847884507,
      "loss": 0.0375,
      "step": 840
    },
    {
      "epoch": 0.6028368794326241,
      "grad_norm": 0.33362510800361633,
      "learning_rate": 0.00019729365643953806,
      "loss": 0.0382,
      "step": 850
    },
    {
      "epoch": 0.6099290780141844,
      "grad_norm": 0.6667296886444092,
      "learning_rate": 0.00019718417273007982,
      "loss": 0.0541,
      "step": 860
    },
    {
      "epoch": 0.6170212765957447,
      "grad_norm": 0.3192714750766754,
      "learning_rate": 0.00019707254976047327,
      "loss": 0.0391,
      "step": 870
    },
    {
      "epoch": 0.624113475177305,
      "grad_norm": 0.6036109328269958,
      "learning_rate": 0.0001969587899878116,
      "loss": 0.0419,
      "step": 880
    },
    {
      "epoch": 0.6312056737588653,
      "grad_norm": 0.46062973141670227,
      "learning_rate": 0.0001968428959162243,
      "loss": 0.0399,
      "step": 890
    },
    {
      "epoch": 0.6382978723404256,
      "grad_norm": 0.45465371012687683,
      "learning_rate": 0.00019672487009682186,
      "loss": 0.0405,
      "step": 900
    },
    {
      "epoch": 0.6453900709219859,
      "grad_norm": 0.7598403692245483,
      "learning_rate": 0.00019660471512763982,
      "loss": 0.0467,
      "step": 910
    },
    {
      "epoch": 0.6524822695035462,
      "grad_norm": 0.5159363746643066,
      "learning_rate": 0.00019648243365358146,
      "loss": 0.05,
      "step": 920
    },
    {
      "epoch": 0.6595744680851063,
      "grad_norm": 0.6169264316558838,
      "learning_rate": 0.0001963580283663596,
      "loss": 0.0545,
      "step": 930
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.657867431640625,
      "learning_rate": 0.0001962315020044374,
      "loss": 0.0354,
      "step": 940
    },
    {
      "epoch": 0.6737588652482269,
      "grad_norm": 0.39899423718452454,
      "learning_rate": 0.00019610285735296797,
      "loss": 0.0432,
      "step": 950
    },
    {
      "epoch": 0.6808510638297872,
      "grad_norm": 0.39202943444252014,
      "learning_rate": 0.0001959720972437331,
      "loss": 0.0437,
      "step": 960
    },
    {
      "epoch": 0.6879432624113475,
      "grad_norm": 0.537214457988739,
      "learning_rate": 0.00019583922455508113,
      "loss": 0.0421,
      "step": 970
    },
    {
      "epoch": 0.6950354609929078,
      "grad_norm": 0.5166184902191162,
      "learning_rate": 0.0001957042422118632,
      "loss": 0.0377,
      "step": 980
    },
    {
      "epoch": 0.7021276595744681,
      "grad_norm": 0.6365172863006592,
      "learning_rate": 0.0001955671531853692,
      "loss": 0.0412,
      "step": 990
    },
    {
      "epoch": 0.7092198581560284,
      "grad_norm": 0.7188717126846313,
      "learning_rate": 0.00019542796049326223,
      "loss": 0.0437,
      "step": 1000
    },
    {
      "epoch": 0.7163120567375887,
      "grad_norm": 0.48798811435699463,
      "learning_rate": 0.00019528666719951222,
      "loss": 0.0594,
      "step": 1010
    },
    {
      "epoch": 0.723404255319149,
      "grad_norm": 0.7655798196792603,
      "learning_rate": 0.00019514327641432833,
      "loss": 0.0414,
      "step": 1020
    },
    {
      "epoch": 0.7304964539007093,
      "grad_norm": 0.5477094650268555,
      "learning_rate": 0.00019499779129409075,
      "loss": 0.0411,
      "step": 1030
    },
    {
      "epoch": 0.7375886524822695,
      "grad_norm": 0.5877737402915955,
      "learning_rate": 0.00019485021504128103,
      "loss": 0.0392,
      "step": 1040
    },
    {
      "epoch": 0.7446808510638298,
      "grad_norm": 0.42971503734588623,
      "learning_rate": 0.00019470055090441162,
      "loss": 0.0386,
      "step": 1050
    },
    {
      "epoch": 0.75177304964539,
      "grad_norm": 0.5418499708175659,
      "learning_rate": 0.00019454880217795441,
      "loss": 0.0467,
      "step": 1060
    },
    {
      "epoch": 0.7588652482269503,
      "grad_norm": 0.4252694845199585,
      "learning_rate": 0.00019439497220226818,
      "loss": 0.035,
      "step": 1070
    },
    {
      "epoch": 0.7659574468085106,
      "grad_norm": 0.39190033078193665,
      "learning_rate": 0.00019423906436352498,
      "loss": 0.043,
      "step": 1080
    },
    {
      "epoch": 0.7730496453900709,
      "grad_norm": 0.44122350215911865,
      "learning_rate": 0.0001940810820936358,
      "loss": 0.0448,
      "step": 1090
    },
    {
      "epoch": 0.7801418439716312,
      "grad_norm": 0.44722622632980347,
      "learning_rate": 0.0001939210288701749,
      "loss": 0.0377,
      "step": 1100
    },
    {
      "epoch": 0.7872340425531915,
      "grad_norm": 0.31242600083351135,
      "learning_rate": 0.00019375890821630315,
      "loss": 0.0443,
      "step": 1110
    },
    {
      "epoch": 0.7943262411347518,
      "grad_norm": 0.4650860130786896,
      "learning_rate": 0.0001935947237006908,
      "loss": 0.0361,
      "step": 1120
    },
    {
      "epoch": 0.8014184397163121,
      "grad_norm": 0.30178096890449524,
      "learning_rate": 0.00019342847893743847,
      "loss": 0.0366,
      "step": 1130
    },
    {
      "epoch": 0.8085106382978723,
      "grad_norm": 0.393303245306015,
      "learning_rate": 0.00019326017758599805,
      "loss": 0.0331,
      "step": 1140
    },
    {
      "epoch": 0.8156028368794326,
      "grad_norm": 0.48034635186195374,
      "learning_rate": 0.00019308982335109183,
      "loss": 0.042,
      "step": 1150
    },
    {
      "epoch": 0.8226950354609929,
      "grad_norm": 0.30449676513671875,
      "learning_rate": 0.00019291741998263108,
      "loss": 0.0355,
      "step": 1160
    },
    {
      "epoch": 0.8297872340425532,
      "grad_norm": 0.5498911738395691,
      "learning_rate": 0.00019274297127563355,
      "loss": 0.0495,
      "step": 1170
    },
    {
      "epoch": 0.8368794326241135,
      "grad_norm": 0.5336613059043884,
      "learning_rate": 0.00019256648107013978,
      "loss": 0.0415,
      "step": 1180
    },
    {
      "epoch": 0.8439716312056738,
      "grad_norm": 0.39611244201660156,
      "learning_rate": 0.0001923879532511287,
      "loss": 0.0387,
      "step": 1190
    },
    {
      "epoch": 0.851063829787234,
      "grad_norm": 0.4194435179233551,
      "learning_rate": 0.00019220739174843205,
      "loss": 0.0438,
      "step": 1200
    },
    {
      "epoch": 0.8581560283687943,
      "grad_norm": 0.2843916714191437,
      "learning_rate": 0.000192024800536648,
      "loss": 0.0448,
      "step": 1210
    },
    {
      "epoch": 0.8652482269503546,
      "grad_norm": 0.5037551522254944,
      "learning_rate": 0.00019184018363505335,
      "loss": 0.0393,
      "step": 1220
    },
    {
      "epoch": 0.8723404255319149,
      "grad_norm": 0.7623959183692932,
      "learning_rate": 0.00019165354510751548,
      "loss": 0.0302,
      "step": 1230
    },
    {
      "epoch": 0.8794326241134752,
      "grad_norm": 0.478982150554657,
      "learning_rate": 0.00019146488906240255,
      "loss": 0.043,
      "step": 1240
    },
    {
      "epoch": 0.8865248226950354,
      "grad_norm": 0.3730108141899109,
      "learning_rate": 0.00019127421965249323,
      "loss": 0.036,
      "step": 1250
    },
    {
      "epoch": 0.8936170212765957,
      "grad_norm": 0.32982558012008667,
      "learning_rate": 0.00019108154107488526,
      "loss": 0.0369,
      "step": 1260
    },
    {
      "epoch": 0.900709219858156,
      "grad_norm": 0.5964722633361816,
      "learning_rate": 0.000190886857570903,
      "loss": 0.0315,
      "step": 1270
    },
    {
      "epoch": 0.9078014184397163,
      "grad_norm": 0.4713937044143677,
      "learning_rate": 0.00019069017342600417,
      "loss": 0.0471,
      "step": 1280
    },
    {
      "epoch": 0.9148936170212766,
      "grad_norm": 0.19976341724395752,
      "learning_rate": 0.00019049149296968543,
      "loss": 0.0369,
      "step": 1290
    },
    {
      "epoch": 0.9219858156028369,
      "grad_norm": 0.5373408198356628,
      "learning_rate": 0.00019029082057538718,
      "loss": 0.04,
      "step": 1300
    },
    {
      "epoch": 0.9290780141843972,
      "grad_norm": 0.4725939631462097,
      "learning_rate": 0.0001900881606603971,
      "loss": 0.0384,
      "step": 1310
    },
    {
      "epoch": 0.9361702127659575,
      "grad_norm": 0.4426575303077698,
      "learning_rate": 0.0001898835176857532,
      "loss": 0.0357,
      "step": 1320
    },
    {
      "epoch": 0.9432624113475178,
      "grad_norm": 0.4588729441165924,
      "learning_rate": 0.0001896768961561453,
      "loss": 0.0397,
      "step": 1330
    },
    {
      "epoch": 0.950354609929078,
      "grad_norm": 0.5770094394683838,
      "learning_rate": 0.0001894683006198161,
      "loss": 0.0523,
      "step": 1340
    },
    {
      "epoch": 0.9574468085106383,
      "grad_norm": 0.5690935254096985,
      "learning_rate": 0.00018925773566846104,
      "loss": 0.0442,
      "step": 1350
    },
    {
      "epoch": 0.9645390070921985,
      "grad_norm": 0.415588915348053,
      "learning_rate": 0.0001890452059371271,
      "loss": 0.0465,
      "step": 1360
    },
    {
      "epoch": 0.9716312056737588,
      "grad_norm": 0.3456709682941437,
      "learning_rate": 0.00018883071610411082,
      "loss": 0.0388,
      "step": 1370
    },
    {
      "epoch": 0.9787234042553191,
      "grad_norm": 0.5673688650131226,
      "learning_rate": 0.00018861427089085552,
      "loss": 0.0426,
      "step": 1380
    },
    {
      "epoch": 0.9858156028368794,
      "grad_norm": 0.39077624678611755,
      "learning_rate": 0.00018839587506184699,
      "loss": 0.0458,
      "step": 1390
    },
    {
      "epoch": 0.9929078014184397,
      "grad_norm": 0.4722648561000824,
      "learning_rate": 0.00018817553342450897,
      "loss": 0.0353,
      "step": 1400
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.42179620265960693,
      "learning_rate": 0.00018795325082909708,
      "loss": 0.0488,
      "step": 1410
    },
    {
      "epoch": 1.0070921985815602,
      "grad_norm": 0.4976104497909546,
      "learning_rate": 0.0001877290321685922,
      "loss": 0.0433,
      "step": 1420
    },
    {
      "epoch": 1.0141843971631206,
      "grad_norm": 0.3866007924079895,
      "learning_rate": 0.00018750288237859275,
      "loss": 0.0411,
      "step": 1430
    },
    {
      "epoch": 1.0212765957446808,
      "grad_norm": 0.4193371534347534,
      "learning_rate": 0.00018727480643720588,
      "loss": 0.0378,
      "step": 1440
    },
    {
      "epoch": 1.0283687943262412,
      "grad_norm": 0.4635421931743622,
      "learning_rate": 0.00018704480936493817,
      "loss": 0.0463,
      "step": 1450
    },
    {
      "epoch": 1.0354609929078014,
      "grad_norm": 0.47600266337394714,
      "learning_rate": 0.00018681289622458485,
      "loss": 0.0371,
      "step": 1460
    },
    {
      "epoch": 1.0425531914893618,
      "grad_norm": 0.4182673394680023,
      "learning_rate": 0.00018657907212111857,
      "loss": 0.0346,
      "step": 1470
    },
    {
      "epoch": 1.049645390070922,
      "grad_norm": 0.2784164547920227,
      "learning_rate": 0.00018634334220157684,
      "loss": 0.0401,
      "step": 1480
    },
    {
      "epoch": 1.0567375886524824,
      "grad_norm": 0.6221792697906494,
      "learning_rate": 0.00018610571165494889,
      "loss": 0.0355,
      "step": 1490
    },
    {
      "epoch": 1.0638297872340425,
      "grad_norm": 0.5726498961448669,
      "learning_rate": 0.00018586618571206134,
      "loss": 0.0396,
      "step": 1500
    },
    {
      "epoch": 1.070921985815603,
      "grad_norm": 0.30378204584121704,
      "learning_rate": 0.00018562476964546307,
      "loss": 0.03,
      "step": 1510
    },
    {
      "epoch": 1.0780141843971631,
      "grad_norm": 0.48525288701057434,
      "learning_rate": 0.00018538146876930924,
      "loss": 0.0342,
      "step": 1520
    },
    {
      "epoch": 1.0851063829787233,
      "grad_norm": 0.30012843012809753,
      "learning_rate": 0.00018513628843924425,
      "loss": 0.0348,
      "step": 1530
    },
    {
      "epoch": 1.0921985815602837,
      "grad_norm": 0.32308411598205566,
      "learning_rate": 0.00018488923405228378,
      "loss": 0.0298,
      "step": 1540
    },
    {
      "epoch": 1.099290780141844,
      "grad_norm": 0.3836841881275177,
      "learning_rate": 0.00018464031104669615,
      "loss": 0.0422,
      "step": 1550
    },
    {
      "epoch": 1.1063829787234043,
      "grad_norm": 0.5509107708930969,
      "learning_rate": 0.0001843895249018825,
      "loss": 0.0296,
      "step": 1560
    },
    {
      "epoch": 1.1134751773049645,
      "grad_norm": 0.4632577896118164,
      "learning_rate": 0.0001841368811382562,
      "loss": 0.0327,
      "step": 1570
    },
    {
      "epoch": 1.1205673758865249,
      "grad_norm": 0.39413148164749146,
      "learning_rate": 0.00018388238531712124,
      "loss": 0.0291,
      "step": 1580
    },
    {
      "epoch": 1.127659574468085,
      "grad_norm": 0.3808094561100006,
      "learning_rate": 0.00018362604304055007,
      "loss": 0.0336,
      "step": 1590
    },
    {
      "epoch": 1.1347517730496455,
      "grad_norm": 0.36897146701812744,
      "learning_rate": 0.00018336785995125998,
      "loss": 0.035,
      "step": 1600
    },
    {
      "epoch": 1.1418439716312057,
      "grad_norm": 0.3850401043891907,
      "learning_rate": 0.00018310784173248908,
      "loss": 0.0366,
      "step": 1610
    },
    {
      "epoch": 1.148936170212766,
      "grad_norm": 0.3824830949306488,
      "learning_rate": 0.0001828459941078712,
      "loss": 0.0372,
      "step": 1620
    },
    {
      "epoch": 1.1560283687943262,
      "grad_norm": 0.47498247027397156,
      "learning_rate": 0.0001825823228413097,
      "loss": 0.0385,
      "step": 1630
    },
    {
      "epoch": 1.1631205673758864,
      "grad_norm": 0.32731738686561584,
      "learning_rate": 0.0001823168337368509,
      "loss": 0.0289,
      "step": 1640
    },
    {
      "epoch": 1.1702127659574468,
      "grad_norm": 0.27296149730682373,
      "learning_rate": 0.00018204953263855603,
      "loss": 0.0268,
      "step": 1650
    },
    {
      "epoch": 1.177304964539007,
      "grad_norm": 0.3830318748950958,
      "learning_rate": 0.00018178042543037277,
      "loss": 0.0358,
      "step": 1660
    },
    {
      "epoch": 1.1843971631205674,
      "grad_norm": 0.4035814106464386,
      "learning_rate": 0.00018150951803600572,
      "loss": 0.0246,
      "step": 1670
    },
    {
      "epoch": 1.1914893617021276,
      "grad_norm": 0.39326614141464233,
      "learning_rate": 0.00018123681641878583,
      "loss": 0.0343,
      "step": 1680
    },
    {
      "epoch": 1.198581560283688,
      "grad_norm": 0.4689919352531433,
      "learning_rate": 0.0001809623265815394,
      "loss": 0.0488,
      "step": 1690
    },
    {
      "epoch": 1.2056737588652482,
      "grad_norm": 0.4357942044734955,
      "learning_rate": 0.00018068605456645574,
      "loss": 0.0336,
      "step": 1700
    },
    {
      "epoch": 1.2127659574468086,
      "grad_norm": 0.2753821611404419,
      "learning_rate": 0.0001804080064549542,
      "loss": 0.028,
      "step": 1710
    },
    {
      "epoch": 1.2198581560283688,
      "grad_norm": 0.33889490365982056,
      "learning_rate": 0.0001801281883675504,
      "loss": 0.0367,
      "step": 1720
    },
    {
      "epoch": 1.226950354609929,
      "grad_norm": 0.3855700194835663,
      "learning_rate": 0.0001798466064637214,
      "loss": 0.0275,
      "step": 1730
    },
    {
      "epoch": 1.2340425531914894,
      "grad_norm": 0.343795508146286,
      "learning_rate": 0.00017956326694177014,
      "loss": 0.0367,
      "step": 1740
    },
    {
      "epoch": 1.2411347517730495,
      "grad_norm": 0.30135342478752136,
      "learning_rate": 0.00017927817603868902,
      "loss": 0.0237,
      "step": 1750
    },
    {
      "epoch": 1.24822695035461,
      "grad_norm": 0.357537180185318,
      "learning_rate": 0.00017899134003002267,
      "loss": 0.0281,
      "step": 1760
    },
    {
      "epoch": 1.2553191489361701,
      "grad_norm": 0.47195494174957275,
      "learning_rate": 0.00017870276522972964,
      "loss": 0.0342,
      "step": 1770
    },
    {
      "epoch": 1.2624113475177305,
      "grad_norm": 0.2944539487361908,
      "learning_rate": 0.00017841245799004357,
      "loss": 0.0282,
      "step": 1780
    },
    {
      "epoch": 1.2695035460992907,
      "grad_norm": 0.5694823861122131,
      "learning_rate": 0.00017812042470133327,
      "loss": 0.0403,
      "step": 1790
    },
    {
      "epoch": 1.2765957446808511,
      "grad_norm": 0.3958476781845093,
      "learning_rate": 0.00017782667179196216,
      "loss": 0.0385,
      "step": 1800
    },
    {
      "epoch": 1.2836879432624113,
      "grad_norm": 0.40212109684944153,
      "learning_rate": 0.0001775312057281466,
      "loss": 0.0303,
      "step": 1810
    },
    {
      "epoch": 1.2907801418439715,
      "grad_norm": 0.36823150515556335,
      "learning_rate": 0.0001772340330138137,
      "loss": 0.0295,
      "step": 1820
    },
    {
      "epoch": 1.297872340425532,
      "grad_norm": 0.25558122992515564,
      "learning_rate": 0.00017693516019045817,
      "loss": 0.0312,
      "step": 1830
    },
    {
      "epoch": 1.3049645390070923,
      "grad_norm": 0.5120919346809387,
      "learning_rate": 0.0001766345938369981,
      "loss": 0.0338,
      "step": 1840
    },
    {
      "epoch": 1.3120567375886525,
      "grad_norm": 0.32636040449142456,
      "learning_rate": 0.00017633234056963049,
      "loss": 0.0339,
      "step": 1850
    },
    {
      "epoch": 1.3191489361702127,
      "grad_norm": 0.3285328447818756,
      "learning_rate": 0.0001760284070416852,
      "loss": 0.0315,
      "step": 1860
    },
    {
      "epoch": 1.326241134751773,
      "grad_norm": 0.40622803568840027,
      "learning_rate": 0.00017572279994347887,
      "loss": 0.0282,
      "step": 1870
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.4662201702594757,
      "learning_rate": 0.00017541552600216747,
      "loss": 0.0413,
      "step": 1880
    },
    {
      "epoch": 1.3404255319148937,
      "grad_norm": 0.345679372549057,
      "learning_rate": 0.00017510659198159822,
      "loss": 0.0322,
      "step": 1890
    },
    {
      "epoch": 1.3475177304964538,
      "grad_norm": 0.34959903359413147,
      "learning_rate": 0.00017479600468216073,
      "loss": 0.0453,
      "step": 1900
    },
    {
      "epoch": 1.3546099290780143,
      "grad_norm": 0.4221673607826233,
      "learning_rate": 0.00017448377094063733,
      "loss": 0.0332,
      "step": 1910
    },
    {
      "epoch": 1.3617021276595744,
      "grad_norm": 0.4349720776081085,
      "learning_rate": 0.00017416989763005246,
      "loss": 0.03,
      "step": 1920
    },
    {
      "epoch": 1.3687943262411348,
      "grad_norm": 0.33618539571762085,
      "learning_rate": 0.00017385439165952158,
      "loss": 0.0333,
      "step": 1930
    },
    {
      "epoch": 1.375886524822695,
      "grad_norm": 0.23778650164604187,
      "learning_rate": 0.0001735372599740989,
      "loss": 0.029,
      "step": 1940
    },
    {
      "epoch": 1.3829787234042552,
      "grad_norm": 0.5169985890388489,
      "learning_rate": 0.00017321850955462454,
      "loss": 0.0406,
      "step": 1950
    },
    {
      "epoch": 1.3900709219858156,
      "grad_norm": 0.41695886850357056,
      "learning_rate": 0.00017289814741757099,
      "loss": 0.0289,
      "step": 1960
    },
    {
      "epoch": 1.397163120567376,
      "grad_norm": 0.33444464206695557,
      "learning_rate": 0.00017257618061488855,
      "loss": 0.0341,
      "step": 1970
    },
    {
      "epoch": 1.4042553191489362,
      "grad_norm": 0.37334710359573364,
      "learning_rate": 0.00017225261623385,
      "loss": 0.0362,
      "step": 1980
    },
    {
      "epoch": 1.4113475177304964,
      "grad_norm": 0.35155797004699707,
      "learning_rate": 0.0001719274613968948,
      "loss": 0.0344,
      "step": 1990
    },
    {
      "epoch": 1.4184397163120568,
      "grad_norm": 0.3294832408428192,
      "learning_rate": 0.0001716007232614723,
      "loss": 0.0347,
      "step": 2000
    },
    {
      "epoch": 1.425531914893617,
      "grad_norm": 0.3127981722354889,
      "learning_rate": 0.00017127240901988392,
      "loss": 0.0308,
      "step": 2010
    },
    {
      "epoch": 1.4326241134751774,
      "grad_norm": 0.30865317583084106,
      "learning_rate": 0.00017094252589912513,
      "loss": 0.0279,
      "step": 2020
    },
    {
      "epoch": 1.4397163120567376,
      "grad_norm": 0.30139267444610596,
      "learning_rate": 0.00017061108116072618,
      "loss": 0.0275,
      "step": 2030
    },
    {
      "epoch": 1.4468085106382977,
      "grad_norm": 0.3612299859523773,
      "learning_rate": 0.00017027808210059238,
      "loss": 0.0329,
      "step": 2040
    },
    {
      "epoch": 1.4539007092198581,
      "grad_norm": 0.2499615103006363,
      "learning_rate": 0.0001699435360488435,
      "loss": 0.0288,
      "step": 2050
    },
    {
      "epoch": 1.4609929078014185,
      "grad_norm": 0.4052152931690216,
      "learning_rate": 0.00016960745036965222,
      "loss": 0.0321,
      "step": 2060
    },
    {
      "epoch": 1.4680851063829787,
      "grad_norm": 0.40873101353645325,
      "learning_rate": 0.00016926983246108219,
      "loss": 0.0352,
      "step": 2070
    },
    {
      "epoch": 1.475177304964539,
      "grad_norm": 0.2873539328575134,
      "learning_rate": 0.0001689306897549253,
      "loss": 0.0313,
      "step": 2080
    },
    {
      "epoch": 1.4822695035460993,
      "grad_norm": 0.41596364974975586,
      "learning_rate": 0.00016859002971653778,
      "loss": 0.0261,
      "step": 2090
    },
    {
      "epoch": 1.4893617021276595,
      "grad_norm": 0.38014939427375793,
      "learning_rate": 0.00016824785984467612,
      "loss": 0.0277,
      "step": 2100
    },
    {
      "epoch": 1.49645390070922,
      "grad_norm": 0.396859347820282,
      "learning_rate": 0.00016790418767133192,
      "loss": 0.0327,
      "step": 2110
    },
    {
      "epoch": 1.50354609929078,
      "grad_norm": 0.3488304018974304,
      "learning_rate": 0.00016755902076156604,
      "loss": 0.0392,
      "step": 2120
    },
    {
      "epoch": 1.5106382978723403,
      "grad_norm": 0.25539007782936096,
      "learning_rate": 0.00016721236671334217,
      "loss": 0.0358,
      "step": 2130
    },
    {
      "epoch": 1.5177304964539007,
      "grad_norm": 0.29417455196380615,
      "learning_rate": 0.00016686423315735954,
      "loss": 0.0236,
      "step": 2140
    },
    {
      "epoch": 1.524822695035461,
      "grad_norm": 0.23996444046497345,
      "learning_rate": 0.00016651462775688493,
      "loss": 0.0283,
      "step": 2150
    },
    {
      "epoch": 1.5319148936170213,
      "grad_norm": 0.3587987422943115,
      "learning_rate": 0.00016616355820758398,
      "loss": 0.0455,
      "step": 2160
    },
    {
      "epoch": 1.5390070921985815,
      "grad_norm": 0.2270328402519226,
      "learning_rate": 0.00016581103223735187,
      "loss": 0.0228,
      "step": 2170
    },
    {
      "epoch": 1.5460992907801419,
      "grad_norm": 0.2835662364959717,
      "learning_rate": 0.00016545705760614307,
      "loss": 0.0222,
      "step": 2180
    },
    {
      "epoch": 1.5531914893617023,
      "grad_norm": 0.3576505780220032,
      "learning_rate": 0.00016510164210580065,
      "loss": 0.0277,
      "step": 2190
    },
    {
      "epoch": 1.5602836879432624,
      "grad_norm": 0.4545815885066986,
      "learning_rate": 0.0001647447935598847,
      "loss": 0.0285,
      "step": 2200
    },
    {
      "epoch": 1.5673758865248226,
      "grad_norm": 0.26356783509254456,
      "learning_rate": 0.00016438651982350017,
      "loss": 0.0317,
      "step": 2210
    },
    {
      "epoch": 1.574468085106383,
      "grad_norm": 0.3527631163597107,
      "learning_rate": 0.00016402682878312387,
      "loss": 0.0313,
      "step": 2220
    },
    {
      "epoch": 1.5815602836879432,
      "grad_norm": 0.16718344390392303,
      "learning_rate": 0.00016366572835643095,
      "loss": 0.0299,
      "step": 2230
    },
    {
      "epoch": 1.5886524822695036,
      "grad_norm": 0.24409593641757965,
      "learning_rate": 0.00016330322649212053,
      "loss": 0.0246,
      "step": 2240
    },
    {
      "epoch": 1.5957446808510638,
      "grad_norm": 0.2794688642024994,
      "learning_rate": 0.00016293933116974087,
      "loss": 0.027,
      "step": 2250
    },
    {
      "epoch": 1.602836879432624,
      "grad_norm": 0.29633232951164246,
      "learning_rate": 0.0001625740503995136,
      "loss": 0.0333,
      "step": 2260
    },
    {
      "epoch": 1.6099290780141844,
      "grad_norm": 0.3844369053840637,
      "learning_rate": 0.00016220739222215738,
      "loss": 0.0331,
      "step": 2270
    },
    {
      "epoch": 1.6170212765957448,
      "grad_norm": 0.7053892612457275,
      "learning_rate": 0.000161839364708711,
      "loss": 0.0357,
      "step": 2280
    },
    {
      "epoch": 1.624113475177305,
      "grad_norm": 0.5903012156486511,
      "learning_rate": 0.00016146997596035569,
      "loss": 0.0367,
      "step": 2290
    },
    {
      "epoch": 1.6312056737588652,
      "grad_norm": 0.41250288486480713,
      "learning_rate": 0.00016109923410823667,
      "loss": 0.0318,
      "step": 2300
    },
    {
      "epoch": 1.6382978723404256,
      "grad_norm": 0.4475861191749573,
      "learning_rate": 0.0001607271473132844,
      "loss": 0.0262,
      "step": 2310
    },
    {
      "epoch": 1.645390070921986,
      "grad_norm": 0.45323246717453003,
      "learning_rate": 0.00016035372376603475,
      "loss": 0.0324,
      "step": 2320
    },
    {
      "epoch": 1.6524822695035462,
      "grad_norm": 0.2859094440937042,
      "learning_rate": 0.0001599789716864487,
      "loss": 0.0253,
      "step": 2330
    },
    {
      "epoch": 1.6595744680851063,
      "grad_norm": 0.44032302498817444,
      "learning_rate": 0.00015960289932373157,
      "loss": 0.0329,
      "step": 2340
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.3153149485588074,
      "learning_rate": 0.00015922551495615124,
      "loss": 0.0352,
      "step": 2350
    },
    {
      "epoch": 1.673758865248227,
      "grad_norm": 0.34591788053512573,
      "learning_rate": 0.000158846826890856,
      "loss": 0.0265,
      "step": 2360
    },
    {
      "epoch": 1.6808510638297873,
      "grad_norm": 0.3234579563140869,
      "learning_rate": 0.00015846684346369186,
      "loss": 0.0271,
      "step": 2370
    },
    {
      "epoch": 1.6879432624113475,
      "grad_norm": 0.3714563548564911,
      "learning_rate": 0.00015808557303901866,
      "loss": 0.0301,
      "step": 2380
    },
    {
      "epoch": 1.6950354609929077,
      "grad_norm": 0.2916625142097473,
      "learning_rate": 0.00015770302400952639,
      "loss": 0.0373,
      "step": 2390
    },
    {
      "epoch": 1.702127659574468,
      "grad_norm": 0.2942117154598236,
      "learning_rate": 0.00015731920479605008,
      "loss": 0.0268,
      "step": 2400
    },
    {
      "epoch": 1.7092198581560285,
      "grad_norm": 0.2918272316455841,
      "learning_rate": 0.0001569341238473847,
      "loss": 0.029,
      "step": 2410
    },
    {
      "epoch": 1.7163120567375887,
      "grad_norm": 0.37234899401664734,
      "learning_rate": 0.0001565477896400991,
      "loss": 0.0263,
      "step": 2420
    },
    {
      "epoch": 1.7234042553191489,
      "grad_norm": 0.291821151971817,
      "learning_rate": 0.0001561602106783493,
      "loss": 0.0294,
      "step": 2430
    },
    {
      "epoch": 1.7304964539007093,
      "grad_norm": 0.6754502654075623,
      "learning_rate": 0.0001557713954936915,
      "loss": 0.0283,
      "step": 2440
    },
    {
      "epoch": 1.7375886524822695,
      "grad_norm": 0.25559020042419434,
      "learning_rate": 0.00015538135264489412,
      "loss": 0.0342,
      "step": 2450
    },
    {
      "epoch": 1.7446808510638299,
      "grad_norm": 0.4409390091896057,
      "learning_rate": 0.00015499009071774945,
      "loss": 0.0293,
      "step": 2460
    },
    {
      "epoch": 1.75177304964539,
      "grad_norm": 0.39615398645401,
      "learning_rate": 0.00015459761832488473,
      "loss": 0.0265,
      "step": 2470
    },
    {
      "epoch": 1.7588652482269502,
      "grad_norm": 0.26630377769470215,
      "learning_rate": 0.0001542039441055724,
      "loss": 0.0313,
      "step": 2480
    },
    {
      "epoch": 1.7659574468085106,
      "grad_norm": 0.31450021266937256,
      "learning_rate": 0.00015380907672554007,
      "loss": 0.0301,
      "step": 2490
    },
    {
      "epoch": 1.773049645390071,
      "grad_norm": 0.5041805505752563,
      "learning_rate": 0.00015341302487677973,
      "loss": 0.0304,
      "step": 2500
    },
    {
      "epoch": 1.7801418439716312,
      "grad_norm": 0.46769461035728455,
      "learning_rate": 0.0001530157972773564,
      "loss": 0.032,
      "step": 2510
    },
    {
      "epoch": 1.7872340425531914,
      "grad_norm": 0.3055451214313507,
      "learning_rate": 0.00015261740267121621,
      "loss": 0.0237,
      "step": 2520
    },
    {
      "epoch": 1.7943262411347518,
      "grad_norm": 0.2608392834663391,
      "learning_rate": 0.00015221784982799404,
      "loss": 0.0252,
      "step": 2530
    },
    {
      "epoch": 1.8014184397163122,
      "grad_norm": 0.4476111829280853,
      "learning_rate": 0.0001518171475428202,
      "loss": 0.0382,
      "step": 2540
    },
    {
      "epoch": 1.8085106382978724,
      "grad_norm": 0.3846855163574219,
      "learning_rate": 0.00015141530463612724,
      "loss": 0.0255,
      "step": 2550
    },
    {
      "epoch": 1.8156028368794326,
      "grad_norm": 0.3949633538722992,
      "learning_rate": 0.0001510123299534554,
      "loss": 0.0393,
      "step": 2560
    },
    {
      "epoch": 1.8226950354609928,
      "grad_norm": 0.2800884246826172,
      "learning_rate": 0.00015060823236525813,
      "loss": 0.0314,
      "step": 2570
    },
    {
      "epoch": 1.8297872340425532,
      "grad_norm": 0.26943808794021606,
      "learning_rate": 0.0001502030207667068,
      "loss": 0.0361,
      "step": 2580
    },
    {
      "epoch": 1.8368794326241136,
      "grad_norm": 0.29443779587745667,
      "learning_rate": 0.00014979670407749477,
      "loss": 0.0259,
      "step": 2590
    },
    {
      "epoch": 1.8439716312056738,
      "grad_norm": 0.27418988943099976,
      "learning_rate": 0.00014938929124164117,
      "loss": 0.0393,
      "step": 2600
    },
    {
      "epoch": 1.851063829787234,
      "grad_norm": 0.34443700313568115,
      "learning_rate": 0.00014898079122729402,
      "loss": 0.0283,
      "step": 2610
    },
    {
      "epoch": 1.8581560283687943,
      "grad_norm": 0.3798273205757141,
      "learning_rate": 0.00014857121302653273,
      "loss": 0.0287,
      "step": 2620
    },
    {
      "epoch": 1.8652482269503547,
      "grad_norm": 0.5379789471626282,
      "learning_rate": 0.0001481605656551703,
      "loss": 0.0307,
      "step": 2630
    },
    {
      "epoch": 1.872340425531915,
      "grad_norm": 0.31364014744758606,
      "learning_rate": 0.00014774885815255464,
      "loss": 0.0293,
      "step": 2640
    },
    {
      "epoch": 1.8794326241134751,
      "grad_norm": 0.3163762092590332,
      "learning_rate": 0.00014733609958136987,
      "loss": 0.0336,
      "step": 2650
    },
    {
      "epoch": 1.8865248226950353,
      "grad_norm": 0.24602492153644562,
      "learning_rate": 0.00014692229902743662,
      "loss": 0.0213,
      "step": 2660
    },
    {
      "epoch": 1.8936170212765957,
      "grad_norm": 0.32489195466041565,
      "learning_rate": 0.00014650746559951205,
      "loss": 0.0321,
      "step": 2670
    },
    {
      "epoch": 1.900709219858156,
      "grad_norm": 0.40947791934013367,
      "learning_rate": 0.00014609160842908952,
      "loss": 0.025,
      "step": 2680
    },
    {
      "epoch": 1.9078014184397163,
      "grad_norm": 0.17635834217071533,
      "learning_rate": 0.00014567473667019735,
      "loss": 0.0207,
      "step": 2690
    },
    {
      "epoch": 1.9148936170212765,
      "grad_norm": 0.30374202132225037,
      "learning_rate": 0.00014525685949919745,
      "loss": 0.0294,
      "step": 2700
    },
    {
      "epoch": 1.9219858156028369,
      "grad_norm": 0.3767811357975006,
      "learning_rate": 0.00014483798611458326,
      "loss": 0.0235,
      "step": 2710
    },
    {
      "epoch": 1.9290780141843973,
      "grad_norm": 0.425862580537796,
      "learning_rate": 0.00014441812573677738,
      "loss": 0.0298,
      "step": 2720
    },
    {
      "epoch": 1.9361702127659575,
      "grad_norm": 0.28912046551704407,
      "learning_rate": 0.00014399728760792848,
      "loss": 0.0215,
      "step": 2730
    },
    {
      "epoch": 1.9432624113475176,
      "grad_norm": 0.4757988452911377,
      "learning_rate": 0.00014357548099170795,
      "loss": 0.0283,
      "step": 2740
    },
    {
      "epoch": 1.950354609929078,
      "grad_norm": 0.38002464175224304,
      "learning_rate": 0.0001431527151731059,
      "loss": 0.0242,
      "step": 2750
    },
    {
      "epoch": 1.9574468085106385,
      "grad_norm": 0.2170061618089676,
      "learning_rate": 0.0001427289994582268,
      "loss": 0.0251,
      "step": 2760
    },
    {
      "epoch": 1.9645390070921986,
      "grad_norm": 0.2710418105125427,
      "learning_rate": 0.00014230434317408474,
      "loss": 0.0276,
      "step": 2770
    },
    {
      "epoch": 1.9716312056737588,
      "grad_norm": 0.3072097599506378,
      "learning_rate": 0.00014187875566839793,
      "loss": 0.0241,
      "step": 2780
    },
    {
      "epoch": 1.978723404255319,
      "grad_norm": 0.28988635540008545,
      "learning_rate": 0.0001414522463093831,
      "loss": 0.0308,
      "step": 2790
    },
    {
      "epoch": 1.9858156028368794,
      "grad_norm": 0.2880018353462219,
      "learning_rate": 0.0001410248244855492,
      "loss": 0.0321,
      "step": 2800
    },
    {
      "epoch": 1.9929078014184398,
      "grad_norm": 0.4506809115409851,
      "learning_rate": 0.0001405964996054907,
      "loss": 0.0247,
      "step": 2810
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.37970393896102905,
      "learning_rate": 0.00014016728109768066,
      "loss": 0.0239,
      "step": 2820
    },
    {
      "epoch": 2.00709219858156,
      "grad_norm": 0.34335339069366455,
      "learning_rate": 0.0001397371784102629,
      "loss": 0.0288,
      "step": 2830
    },
    {
      "epoch": 2.0141843971631204,
      "grad_norm": 0.2879555821418762,
      "learning_rate": 0.00013930620101084432,
      "loss": 0.0244,
      "step": 2840
    },
    {
      "epoch": 2.021276595744681,
      "grad_norm": 0.29170095920562744,
      "learning_rate": 0.0001388743583862862,
      "loss": 0.0298,
      "step": 2850
    },
    {
      "epoch": 2.028368794326241,
      "grad_norm": 0.4589764177799225,
      "learning_rate": 0.00013844166004249578,
      "loss": 0.0248,
      "step": 2860
    },
    {
      "epoch": 2.0354609929078014,
      "grad_norm": 0.32960548996925354,
      "learning_rate": 0.00013800811550421655,
      "loss": 0.0296,
      "step": 2870
    },
    {
      "epoch": 2.0425531914893615,
      "grad_norm": 0.3347654938697815,
      "learning_rate": 0.00013757373431481882,
      "loss": 0.0278,
      "step": 2880
    },
    {
      "epoch": 2.049645390070922,
      "grad_norm": 0.3965177536010742,
      "learning_rate": 0.00013713852603608977,
      "loss": 0.0296,
      "step": 2890
    },
    {
      "epoch": 2.0567375886524824,
      "grad_norm": 0.2053564339876175,
      "learning_rate": 0.00013670250024802266,
      "loss": 0.0279,
      "step": 2900
    },
    {
      "epoch": 2.0638297872340425,
      "grad_norm": 0.6006909012794495,
      "learning_rate": 0.0001362656665486062,
      "loss": 0.0342,
      "step": 2910
    },
    {
      "epoch": 2.0709219858156027,
      "grad_norm": 0.24472111463546753,
      "learning_rate": 0.00013582803455361323,
      "loss": 0.0244,
      "step": 2920
    },
    {
      "epoch": 2.078014184397163,
      "grad_norm": 0.2269747108221054,
      "learning_rate": 0.00013538961389638897,
      "loss": 0.0256,
      "step": 2930
    },
    {
      "epoch": 2.0851063829787235,
      "grad_norm": 0.31606826186180115,
      "learning_rate": 0.00013495041422763898,
      "loss": 0.0305,
      "step": 2940
    },
    {
      "epoch": 2.0921985815602837,
      "grad_norm": 0.2274671196937561,
      "learning_rate": 0.0001345104452152168,
      "loss": 0.0307,
      "step": 2950
    },
    {
      "epoch": 2.099290780141844,
      "grad_norm": 0.2778410017490387,
      "learning_rate": 0.00013406971654391115,
      "loss": 0.0274,
      "step": 2960
    },
    {
      "epoch": 2.106382978723404,
      "grad_norm": 0.545807421207428,
      "learning_rate": 0.00013362823791523266,
      "loss": 0.028,
      "step": 2970
    },
    {
      "epoch": 2.1134751773049647,
      "grad_norm": 0.41192397475242615,
      "learning_rate": 0.00013318601904720024,
      "loss": 0.0247,
      "step": 2980
    },
    {
      "epoch": 2.120567375886525,
      "grad_norm": 0.4247974157333374,
      "learning_rate": 0.00013274306967412736,
      "loss": 0.0308,
      "step": 2990
    },
    {
      "epoch": 2.127659574468085,
      "grad_norm": 0.49807772040367126,
      "learning_rate": 0.0001322993995464078,
      "loss": 0.0379,
      "step": 3000
    },
    {
      "epoch": 2.1347517730496453,
      "grad_norm": 0.2732253968715668,
      "learning_rate": 0.0001318550184303006,
      "loss": 0.0242,
      "step": 3010
    },
    {
      "epoch": 2.141843971631206,
      "grad_norm": 0.3375590145587921,
      "learning_rate": 0.00013140993610771573,
      "loss": 0.0281,
      "step": 3020
    },
    {
      "epoch": 2.148936170212766,
      "grad_norm": 0.24619796872138977,
      "learning_rate": 0.0001309641623759982,
      "loss": 0.0261,
      "step": 3030
    },
    {
      "epoch": 2.1560283687943262,
      "grad_norm": 0.39782339334487915,
      "learning_rate": 0.0001305177070477126,
      "loss": 0.0268,
      "step": 3040
    },
    {
      "epoch": 2.1631205673758864,
      "grad_norm": 0.5394360423088074,
      "learning_rate": 0.00013007057995042732,
      "loss": 0.0305,
      "step": 3050
    },
    {
      "epoch": 2.1702127659574466,
      "grad_norm": 0.3109723925590515,
      "learning_rate": 0.00012962279092649783,
      "loss": 0.0224,
      "step": 3060
    },
    {
      "epoch": 2.1773049645390072,
      "grad_norm": 0.24598295986652374,
      "learning_rate": 0.0001291743498328503,
      "loss": 0.0239,
      "step": 3070
    },
    {
      "epoch": 2.1843971631205674,
      "grad_norm": 0.37659743428230286,
      "learning_rate": 0.0001287252665407645,
      "loss": 0.0322,
      "step": 3080
    },
    {
      "epoch": 2.1914893617021276,
      "grad_norm": 0.4015694260597229,
      "learning_rate": 0.00012827555093565657,
      "loss": 0.0304,
      "step": 3090
    },
    {
      "epoch": 2.198581560283688,
      "grad_norm": 0.32140156626701355,
      "learning_rate": 0.00012782521291686147,
      "loss": 0.0302,
      "step": 3100
    },
    {
      "epoch": 2.2056737588652484,
      "grad_norm": 0.3436850905418396,
      "learning_rate": 0.00012737426239741484,
      "loss": 0.0265,
      "step": 3110
    },
    {
      "epoch": 2.2127659574468086,
      "grad_norm": 0.2246638983488083,
      "learning_rate": 0.00012692270930383508,
      "loss": 0.0253,
      "step": 3120
    },
    {
      "epoch": 2.219858156028369,
      "grad_norm": 0.25144413113594055,
      "learning_rate": 0.00012647056357590472,
      "loss": 0.0274,
      "step": 3130
    },
    {
      "epoch": 2.226950354609929,
      "grad_norm": 0.3174673020839691,
      "learning_rate": 0.0001260178351664515,
      "loss": 0.0288,
      "step": 3140
    },
    {
      "epoch": 2.2340425531914896,
      "grad_norm": 0.36320146918296814,
      "learning_rate": 0.0001255645340411295,
      "loss": 0.0294,
      "step": 3150
    },
    {
      "epoch": 2.2411347517730498,
      "grad_norm": 0.22592422366142273,
      "learning_rate": 0.00012511067017819964,
      "loss": 0.0245,
      "step": 3160
    },
    {
      "epoch": 2.24822695035461,
      "grad_norm": 0.27350878715515137,
      "learning_rate": 0.00012465625356831007,
      "loss": 0.0297,
      "step": 3170
    },
    {
      "epoch": 2.25531914893617,
      "grad_norm": 0.3462134003639221,
      "learning_rate": 0.00012420129421427623,
      "loss": 0.0295,
      "step": 3180
    },
    {
      "epoch": 2.2624113475177303,
      "grad_norm": 0.23078316450119019,
      "learning_rate": 0.00012374580213086065,
      "loss": 0.0236,
      "step": 3190
    },
    {
      "epoch": 2.269503546099291,
      "grad_norm": 0.41071340441703796,
      "learning_rate": 0.00012328978734455265,
      "loss": 0.0303,
      "step": 3200
    },
    {
      "epoch": 2.276595744680851,
      "grad_norm": 0.3645431697368622,
      "learning_rate": 0.00012283325989334736,
      "loss": 0.0215,
      "step": 3210
    },
    {
      "epoch": 2.2836879432624113,
      "grad_norm": 0.3068390488624573,
      "learning_rate": 0.00012237622982652497,
      "loss": 0.0216,
      "step": 3220
    },
    {
      "epoch": 2.2907801418439715,
      "grad_norm": 0.2963109314441681,
      "learning_rate": 0.00012191870720442953,
      "loss": 0.024,
      "step": 3230
    },
    {
      "epoch": 2.297872340425532,
      "grad_norm": 0.34965890645980835,
      "learning_rate": 0.00012146070209824736,
      "loss": 0.0208,
      "step": 3240
    },
    {
      "epoch": 2.3049645390070923,
      "grad_norm": 0.3760829269886017,
      "learning_rate": 0.00012100222458978541,
      "loss": 0.0231,
      "step": 3250
    },
    {
      "epoch": 2.3120567375886525,
      "grad_norm": 0.39412423968315125,
      "learning_rate": 0.00012054328477124943,
      "loss": 0.0263,
      "step": 3260
    },
    {
      "epoch": 2.3191489361702127,
      "grad_norm": 0.47879675030708313,
      "learning_rate": 0.0001200838927450217,
      "loss": 0.0253,
      "step": 3270
    },
    {
      "epoch": 2.326241134751773,
      "grad_norm": 0.32520684599876404,
      "learning_rate": 0.00011962405862343866,
      "loss": 0.0221,
      "step": 3280
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 0.2687060236930847,
      "learning_rate": 0.00011916379252856838,
      "loss": 0.0251,
      "step": 3290
    },
    {
      "epoch": 2.3404255319148937,
      "grad_norm": 0.246451273560524,
      "learning_rate": 0.00011870310459198772,
      "loss": 0.0301,
      "step": 3300
    },
    {
      "epoch": 2.347517730496454,
      "grad_norm": 0.2831878960132599,
      "learning_rate": 0.00011824200495455929,
      "loss": 0.0207,
      "step": 3310
    },
    {
      "epoch": 2.354609929078014,
      "grad_norm": 0.2363816499710083,
      "learning_rate": 0.00011778050376620822,
      "loss": 0.0199,
      "step": 3320
    },
    {
      "epoch": 2.3617021276595747,
      "grad_norm": 0.30800870060920715,
      "learning_rate": 0.0001173186111856988,
      "loss": 0.0234,
      "step": 3330
    },
    {
      "epoch": 2.368794326241135,
      "grad_norm": 0.3593873679637909,
      "learning_rate": 0.00011685633738041075,
      "loss": 0.0218,
      "step": 3340
    },
    {
      "epoch": 2.375886524822695,
      "grad_norm": 0.1976577192544937,
      "learning_rate": 0.00011639369252611552,
      "loss": 0.0248,
      "step": 3350
    },
    {
      "epoch": 2.382978723404255,
      "grad_norm": 0.21267811954021454,
      "learning_rate": 0.00011593068680675228,
      "loss": 0.0245,
      "step": 3360
    },
    {
      "epoch": 2.3900709219858154,
      "grad_norm": 0.3738497197628021,
      "learning_rate": 0.00011546733041420362,
      "loss": 0.0282,
      "step": 3370
    },
    {
      "epoch": 2.397163120567376,
      "grad_norm": 0.37859100103378296,
      "learning_rate": 0.0001150036335480714,
      "loss": 0.0189,
      "step": 3380
    },
    {
      "epoch": 2.404255319148936,
      "grad_norm": 0.32795801758766174,
      "learning_rate": 0.00011453960641545206,
      "loss": 0.0236,
      "step": 3390
    },
    {
      "epoch": 2.4113475177304964,
      "grad_norm": 0.2747984528541565,
      "learning_rate": 0.00011407525923071207,
      "loss": 0.0223,
      "step": 3400
    },
    {
      "epoch": 2.4184397163120566,
      "grad_norm": 0.22304099798202515,
      "learning_rate": 0.00011361060221526301,
      "loss": 0.0256,
      "step": 3410
    },
    {
      "epoch": 2.425531914893617,
      "grad_norm": 0.15323618054389954,
      "learning_rate": 0.00011314564559733654,
      "loss": 0.0235,
      "step": 3420
    },
    {
      "epoch": 2.4326241134751774,
      "grad_norm": 0.4732156991958618,
      "learning_rate": 0.00011268039961175936,
      "loss": 0.0278,
      "step": 3430
    },
    {
      "epoch": 2.4397163120567376,
      "grad_norm": 0.3939105272293091,
      "learning_rate": 0.0001122148744997278,
      "loss": 0.0365,
      "step": 3440
    },
    {
      "epoch": 2.4468085106382977,
      "grad_norm": 0.3410981595516205,
      "learning_rate": 0.00011174908050858251,
      "loss": 0.0229,
      "step": 3450
    },
    {
      "epoch": 2.453900709219858,
      "grad_norm": 0.34085023403167725,
      "learning_rate": 0.00011128302789158273,
      "loss": 0.0246,
      "step": 3460
    },
    {
      "epoch": 2.4609929078014185,
      "grad_norm": 0.3545391261577606,
      "learning_rate": 0.00011081672690768084,
      "loss": 0.0241,
      "step": 3470
    },
    {
      "epoch": 2.4680851063829787,
      "grad_norm": 0.3405749797821045,
      "learning_rate": 0.00011035018782129618,
      "loss": 0.0238,
      "step": 3480
    },
    {
      "epoch": 2.475177304964539,
      "grad_norm": 0.7175484895706177,
      "learning_rate": 0.00010988342090208945,
      "loss": 0.0303,
      "step": 3490
    },
    {
      "epoch": 2.482269503546099,
      "grad_norm": 0.4182647168636322,
      "learning_rate": 0.00010941643642473646,
      "loss": 0.0219,
      "step": 3500
    },
    {
      "epoch": 2.4893617021276597,
      "grad_norm": 0.2711605131626129,
      "learning_rate": 0.00010894924466870205,
      "loss": 0.0282,
      "step": 3510
    },
    {
      "epoch": 2.49645390070922,
      "grad_norm": 0.3410244584083557,
      "learning_rate": 0.00010848185591801365,
      "loss": 0.0263,
      "step": 3520
    },
    {
      "epoch": 2.50354609929078,
      "grad_norm": 0.2878762483596802,
      "learning_rate": 0.0001080142804610351,
      "loss": 0.0213,
      "step": 3530
    },
    {
      "epoch": 2.5106382978723403,
      "grad_norm": 0.33405497670173645,
      "learning_rate": 0.00010754652859024017,
      "loss": 0.0316,
      "step": 3540
    },
    {
      "epoch": 2.5177304964539005,
      "grad_norm": 0.2429051399230957,
      "learning_rate": 0.00010707861060198581,
      "loss": 0.0223,
      "step": 3550
    },
    {
      "epoch": 2.524822695035461,
      "grad_norm": 0.16325709223747253,
      "learning_rate": 0.00010661053679628561,
      "loss": 0.0254,
      "step": 3560
    },
    {
      "epoch": 2.5319148936170213,
      "grad_norm": 0.301996648311615,
      "learning_rate": 0.0001061423174765832,
      "loss": 0.0309,
      "step": 3570
    },
    {
      "epoch": 2.5390070921985815,
      "grad_norm": 0.16105884313583374,
      "learning_rate": 0.00010567396294952527,
      "loss": 0.025,
      "step": 3580
    },
    {
      "epoch": 2.546099290780142,
      "grad_norm": 0.2534509003162384,
      "learning_rate": 0.00010520548352473468,
      "loss": 0.026,
      "step": 3590
    },
    {
      "epoch": 2.5531914893617023,
      "grad_norm": 0.17970708012580872,
      "learning_rate": 0.00010473688951458369,
      "loss": 0.022,
      "step": 3600
    },
    {
      "epoch": 2.5602836879432624,
      "grad_norm": 0.2073034942150116,
      "learning_rate": 0.00010426819123396687,
      "loss": 0.0253,
      "step": 3610
    },
    {
      "epoch": 2.5673758865248226,
      "grad_norm": 0.2955438494682312,
      "learning_rate": 0.00010379939900007393,
      "loss": 0.0235,
      "step": 3620
    },
    {
      "epoch": 2.574468085106383,
      "grad_norm": 0.19330568611621857,
      "learning_rate": 0.00010333052313216281,
      "loss": 0.0211,
      "step": 3630
    },
    {
      "epoch": 2.581560283687943,
      "grad_norm": 0.20802323520183563,
      "learning_rate": 0.00010286157395133244,
      "loss": 0.0186,
      "step": 3640
    },
    {
      "epoch": 2.5886524822695036,
      "grad_norm": 0.27056363224983215,
      "learning_rate": 0.00010239256178029548,
      "loss": 0.0233,
      "step": 3650
    },
    {
      "epoch": 2.595744680851064,
      "grad_norm": 0.20938947796821594,
      "learning_rate": 0.00010192349694315124,
      "loss": 0.0202,
      "step": 3660
    },
    {
      "epoch": 2.602836879432624,
      "grad_norm": 0.22657611966133118,
      "learning_rate": 0.00010145438976515828,
      "loss": 0.0223,
      "step": 3670
    },
    {
      "epoch": 2.6099290780141846,
      "grad_norm": 0.32048580050468445,
      "learning_rate": 0.00010098525057250718,
      "loss": 0.0208,
      "step": 3680
    },
    {
      "epoch": 2.617021276595745,
      "grad_norm": 0.30489683151245117,
      "learning_rate": 0.0001005160896920933,
      "loss": 0.0197,
      "step": 3690
    },
    {
      "epoch": 2.624113475177305,
      "grad_norm": 0.339277058839798,
      "learning_rate": 0.00010004691745128933,
      "loss": 0.0252,
      "step": 3700
    },
    {
      "epoch": 2.631205673758865,
      "grad_norm": 0.3218172490596771,
      "learning_rate": 9.957774417771809e-05,
      "loss": 0.0267,
      "step": 3710
    },
    {
      "epoch": 2.6382978723404253,
      "grad_norm": 0.33065009117126465,
      "learning_rate": 9.91085801990251e-05,
      "loss": 0.0258,
      "step": 3720
    },
    {
      "epoch": 2.645390070921986,
      "grad_norm": 0.32526910305023193,
      "learning_rate": 9.863943584265125e-05,
      "loss": 0.025,
      "step": 3730
    },
    {
      "epoch": 2.652482269503546,
      "grad_norm": 0.3356036841869354,
      "learning_rate": 9.817032143560561e-05,
      "loss": 0.0249,
      "step": 3740
    },
    {
      "epoch": 2.6595744680851063,
      "grad_norm": 0.29925739765167236,
      "learning_rate": 9.77012473042378e-05,
      "loss": 0.0284,
      "step": 3750
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.19530752301216125,
      "learning_rate": 9.72322237740111e-05,
      "loss": 0.0241,
      "step": 3760
    },
    {
      "epoch": 2.673758865248227,
      "grad_norm": 0.30117693543434143,
      "learning_rate": 9.676326116927472e-05,
      "loss": 0.0216,
      "step": 3770
    },
    {
      "epoch": 2.6808510638297873,
      "grad_norm": 0.32530850172042847,
      "learning_rate": 9.629436981303696e-05,
      "loss": 0.0213,
      "step": 3780
    },
    {
      "epoch": 2.6879432624113475,
      "grad_norm": 0.26192381978034973,
      "learning_rate": 9.582556002673756e-05,
      "loss": 0.0229,
      "step": 3790
    },
    {
      "epoch": 2.6950354609929077,
      "grad_norm": 0.2634004056453705,
      "learning_rate": 9.535684213002087e-05,
      "loss": 0.0317,
      "step": 3800
    },
    {
      "epoch": 2.702127659574468,
      "grad_norm": 0.31830480694770813,
      "learning_rate": 9.488822644050844e-05,
      "loss": 0.0229,
      "step": 3810
    },
    {
      "epoch": 2.7092198581560285,
      "grad_norm": 0.26657938957214355,
      "learning_rate": 9.441972327357203e-05,
      "loss": 0.025,
      "step": 3820
    },
    {
      "epoch": 2.7163120567375887,
      "grad_norm": 0.24714654684066772,
      "learning_rate": 9.395134294210648e-05,
      "loss": 0.0197,
      "step": 3830
    },
    {
      "epoch": 2.723404255319149,
      "grad_norm": 0.2549240291118622,
      "learning_rate": 9.348309575630271e-05,
      "loss": 0.0255,
      "step": 3840
    },
    {
      "epoch": 2.7304964539007095,
      "grad_norm": 0.19039757549762726,
      "learning_rate": 9.301499202342086e-05,
      "loss": 0.0224,
      "step": 3850
    },
    {
      "epoch": 2.7375886524822697,
      "grad_norm": 0.24891236424446106,
      "learning_rate": 9.254704204756319e-05,
      "loss": 0.0252,
      "step": 3860
    },
    {
      "epoch": 2.74468085106383,
      "grad_norm": 0.30138468742370605,
      "learning_rate": 9.207925612944752e-05,
      "loss": 0.0233,
      "step": 3870
    },
    {
      "epoch": 2.75177304964539,
      "grad_norm": 0.24954111874103546,
      "learning_rate": 9.161164456618027e-05,
      "loss": 0.0262,
      "step": 3880
    },
    {
      "epoch": 2.7588652482269502,
      "grad_norm": 0.24573440849781036,
      "learning_rate": 9.114421765102999e-05,
      "loss": 0.0276,
      "step": 3890
    },
    {
      "epoch": 2.7659574468085104,
      "grad_norm": 0.2604183256626129,
      "learning_rate": 9.06769856732005e-05,
      "loss": 0.0235,
      "step": 3900
    },
    {
      "epoch": 2.773049645390071,
      "grad_norm": 0.30546581745147705,
      "learning_rate": 9.02099589176048e-05,
      "loss": 0.0178,
      "step": 3910
    },
    {
      "epoch": 2.780141843971631,
      "grad_norm": 0.4359723627567291,
      "learning_rate": 8.974314766463825e-05,
      "loss": 0.0178,
      "step": 3920
    },
    {
      "epoch": 2.7872340425531914,
      "grad_norm": 0.27847006916999817,
      "learning_rate": 8.927656218995261e-05,
      "loss": 0.0223,
      "step": 3930
    },
    {
      "epoch": 2.794326241134752,
      "grad_norm": 0.21782971918582916,
      "learning_rate": 8.88102127642296e-05,
      "loss": 0.0271,
      "step": 3940
    },
    {
      "epoch": 2.801418439716312,
      "grad_norm": 0.3402385711669922,
      "learning_rate": 8.834410965295506e-05,
      "loss": 0.0276,
      "step": 3950
    },
    {
      "epoch": 2.8085106382978724,
      "grad_norm": 0.2527099847793579,
      "learning_rate": 8.78782631161927e-05,
      "loss": 0.0352,
      "step": 3960
    },
    {
      "epoch": 2.8156028368794326,
      "grad_norm": 0.4070510268211365,
      "learning_rate": 8.741268340835848e-05,
      "loss": 0.0237,
      "step": 3970
    },
    {
      "epoch": 2.8226950354609928,
      "grad_norm": 0.4537751078605652,
      "learning_rate": 8.694738077799488e-05,
      "loss": 0.0248,
      "step": 3980
    },
    {
      "epoch": 2.829787234042553,
      "grad_norm": 0.25030770897865295,
      "learning_rate": 8.648236546754502e-05,
      "loss": 0.0216,
      "step": 3990
    },
    {
      "epoch": 2.8368794326241136,
      "grad_norm": 0.2886297404766083,
      "learning_rate": 8.601764771312765e-05,
      "loss": 0.019,
      "step": 4000
    },
    {
      "epoch": 2.8439716312056738,
      "grad_norm": 0.2288312315940857,
      "learning_rate": 8.555323774431143e-05,
      "loss": 0.0184,
      "step": 4010
    },
    {
      "epoch": 2.851063829787234,
      "grad_norm": 0.36999377608299255,
      "learning_rate": 8.508914578389e-05,
      "loss": 0.019,
      "step": 4020
    },
    {
      "epoch": 2.8581560283687946,
      "grad_norm": 0.27074334025382996,
      "learning_rate": 8.462538204765675e-05,
      "loss": 0.0201,
      "step": 4030
    },
    {
      "epoch": 2.8652482269503547,
      "grad_norm": 0.32663315534591675,
      "learning_rate": 8.416195674418017e-05,
      "loss": 0.0217,
      "step": 4040
    },
    {
      "epoch": 2.872340425531915,
      "grad_norm": 0.2872985601425171,
      "learning_rate": 8.369888007457895e-05,
      "loss": 0.0252,
      "step": 4050
    },
    {
      "epoch": 2.879432624113475,
      "grad_norm": 0.2996869683265686,
      "learning_rate": 8.323616223229761e-05,
      "loss": 0.0247,
      "step": 4060
    },
    {
      "epoch": 2.8865248226950353,
      "grad_norm": 0.2089969664812088,
      "learning_rate": 8.277381340288182e-05,
      "loss": 0.0277,
      "step": 4070
    },
    {
      "epoch": 2.8936170212765955,
      "grad_norm": 0.4416872262954712,
      "learning_rate": 8.231184376375451e-05,
      "loss": 0.0186,
      "step": 4080
    },
    {
      "epoch": 2.900709219858156,
      "grad_norm": 0.27074792981147766,
      "learning_rate": 8.185026348399174e-05,
      "loss": 0.0191,
      "step": 4090
    },
    {
      "epoch": 2.9078014184397163,
      "grad_norm": 0.17527180910110474,
      "learning_rate": 8.138908272409869e-05,
      "loss": 0.0206,
      "step": 4100
    },
    {
      "epoch": 2.9148936170212765,
      "grad_norm": 0.2025071531534195,
      "learning_rate": 8.092831163578625e-05,
      "loss": 0.0178,
      "step": 4110
    },
    {
      "epoch": 2.921985815602837,
      "grad_norm": 0.2685222625732422,
      "learning_rate": 8.046796036174738e-05,
      "loss": 0.0174,
      "step": 4120
    },
    {
      "epoch": 2.9290780141843973,
      "grad_norm": 0.34887775778770447,
      "learning_rate": 8.000803903543398e-05,
      "loss": 0.0158,
      "step": 4130
    },
    {
      "epoch": 2.9361702127659575,
      "grad_norm": 0.2305716872215271,
      "learning_rate": 7.95485577808336e-05,
      "loss": 0.0167,
      "step": 4140
    },
    {
      "epoch": 2.9432624113475176,
      "grad_norm": 0.4017210006713867,
      "learning_rate": 7.908952671224698e-05,
      "loss": 0.0246,
      "step": 4150
    },
    {
      "epoch": 2.950354609929078,
      "grad_norm": 0.16771471500396729,
      "learning_rate": 7.863095593406491e-05,
      "loss": 0.0169,
      "step": 4160
    },
    {
      "epoch": 2.9574468085106385,
      "grad_norm": 0.3576452136039734,
      "learning_rate": 7.817285554054628e-05,
      "loss": 0.0176,
      "step": 4170
    },
    {
      "epoch": 2.9645390070921986,
      "grad_norm": 0.2420264184474945,
      "learning_rate": 7.771523561559555e-05,
      "loss": 0.019,
      "step": 4180
    },
    {
      "epoch": 2.971631205673759,
      "grad_norm": 0.15221752226352692,
      "learning_rate": 7.725810623254088e-05,
      "loss": 0.0168,
      "step": 4190
    },
    {
      "epoch": 2.978723404255319,
      "grad_norm": 0.18496164679527283,
      "learning_rate": 7.680147745391255e-05,
      "loss": 0.0128,
      "step": 4200
    },
    {
      "epoch": 2.9858156028368796,
      "grad_norm": 0.25904592871665955,
      "learning_rate": 7.634535933122111e-05,
      "loss": 0.0194,
      "step": 4210
    },
    {
      "epoch": 2.99290780141844,
      "grad_norm": 0.19931963086128235,
      "learning_rate": 7.588976190473657e-05,
      "loss": 0.0133,
      "step": 4220
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.5127366781234741,
      "learning_rate": 7.543469520326694e-05,
      "loss": 0.0178,
      "step": 4230
    },
    {
      "epoch": 3.00709219858156,
      "grad_norm": 0.19438482820987701,
      "learning_rate": 7.498016924393778e-05,
      "loss": 0.0133,
      "step": 4240
    },
    {
      "epoch": 3.0141843971631204,
      "grad_norm": 0.43314528465270996,
      "learning_rate": 7.45261940319716e-05,
      "loss": 0.0203,
      "step": 4250
    },
    {
      "epoch": 3.021276595744681,
      "grad_norm": 0.33415085077285767,
      "learning_rate": 7.407277956046763e-05,
      "loss": 0.0153,
      "step": 4260
    },
    {
      "epoch": 3.028368794326241,
      "grad_norm": 0.2561874985694885,
      "learning_rate": 7.361993581018173e-05,
      "loss": 0.0189,
      "step": 4270
    },
    {
      "epoch": 3.0354609929078014,
      "grad_norm": 0.2553478181362152,
      "learning_rate": 7.316767274930699e-05,
      "loss": 0.0178,
      "step": 4280
    },
    {
      "epoch": 3.0425531914893615,
      "grad_norm": 0.38511422276496887,
      "learning_rate": 7.271600033325393e-05,
      "loss": 0.033,
      "step": 4290
    },
    {
      "epoch": 3.049645390070922,
      "grad_norm": 0.27205201983451843,
      "learning_rate": 7.226492850443161e-05,
      "loss": 0.0156,
      "step": 4300
    },
    {
      "epoch": 3.0567375886524824,
      "grad_norm": 0.22120597958564758,
      "learning_rate": 7.181446719202873e-05,
      "loss": 0.0189,
      "step": 4310
    },
    {
      "epoch": 3.0638297872340425,
      "grad_norm": 0.28937768936157227,
      "learning_rate": 7.136462631179502e-05,
      "loss": 0.019,
      "step": 4320
    },
    {
      "epoch": 3.0709219858156027,
      "grad_norm": 0.24847429990768433,
      "learning_rate": 7.0915415765823e-05,
      "loss": 0.0205,
      "step": 4330
    },
    {
      "epoch": 3.078014184397163,
      "grad_norm": 0.16116857528686523,
      "learning_rate": 7.04668454423299e-05,
      "loss": 0.0166,
      "step": 4340
    },
    {
      "epoch": 3.0851063829787235,
      "grad_norm": 0.4110172688961029,
      "learning_rate": 7.001892521544027e-05,
      "loss": 0.0213,
      "step": 4350
    },
    {
      "epoch": 3.0921985815602837,
      "grad_norm": 0.3403964936733246,
      "learning_rate": 6.957166494496828e-05,
      "loss": 0.0179,
      "step": 4360
    },
    {
      "epoch": 3.099290780141844,
      "grad_norm": 0.3022840917110443,
      "learning_rate": 6.912507447620102e-05,
      "loss": 0.018,
      "step": 4370
    },
    {
      "epoch": 3.106382978723404,
      "grad_norm": 0.14554759860038757,
      "learning_rate": 6.867916363968143e-05,
      "loss": 0.0158,
      "step": 4380
    },
    {
      "epoch": 3.1134751773049647,
      "grad_norm": 0.2720215916633606,
      "learning_rate": 6.82339422509923e-05,
      "loss": 0.0221,
      "step": 4390
    },
    {
      "epoch": 3.120567375886525,
      "grad_norm": 0.23295775055885315,
      "learning_rate": 6.778942011053976e-05,
      "loss": 0.0149,
      "step": 4400
    },
    {
      "epoch": 3.127659574468085,
      "grad_norm": 0.24804311990737915,
      "learning_rate": 6.7345607003338e-05,
      "loss": 0.0161,
      "step": 4410
    },
    {
      "epoch": 3.1347517730496453,
      "grad_norm": 0.1949310600757599,
      "learning_rate": 6.690251269879355e-05,
      "loss": 0.0188,
      "step": 4420
    },
    {
      "epoch": 3.141843971631206,
      "grad_norm": 0.2794622778892517,
      "learning_rate": 6.64601469504903e-05,
      "loss": 0.0141,
      "step": 4430
    },
    {
      "epoch": 3.148936170212766,
      "grad_norm": 0.17811664938926697,
      "learning_rate": 6.601851949597497e-05,
      "loss": 0.0189,
      "step": 4440
    },
    {
      "epoch": 3.1560283687943262,
      "grad_norm": 0.248806893825531,
      "learning_rate": 6.557764005654254e-05,
      "loss": 0.0172,
      "step": 4450
    },
    {
      "epoch": 3.1631205673758864,
      "grad_norm": 0.2596416175365448,
      "learning_rate": 6.513751833702244e-05,
      "loss": 0.0198,
      "step": 4460
    },
    {
      "epoch": 3.1702127659574466,
      "grad_norm": 0.3255382776260376,
      "learning_rate": 6.469816402556475e-05,
      "loss": 0.0221,
      "step": 4470
    },
    {
      "epoch": 3.1773049645390072,
      "grad_norm": 0.3366486430168152,
      "learning_rate": 6.425958679342708e-05,
      "loss": 0.0145,
      "step": 4480
    },
    {
      "epoch": 3.1843971631205674,
      "grad_norm": 0.24371476471424103,
      "learning_rate": 6.382179629476161e-05,
      "loss": 0.02,
      "step": 4490
    },
    {
      "epoch": 3.1914893617021276,
      "grad_norm": 0.16515839099884033,
      "learning_rate": 6.338480216640271e-05,
      "loss": 0.025,
      "step": 4500
    },
    {
      "epoch": 3.198581560283688,
      "grad_norm": 0.2598879635334015,
      "learning_rate": 6.294861402765448e-05,
      "loss": 0.0139,
      "step": 4510
    },
    {
      "epoch": 3.2056737588652484,
      "grad_norm": 0.25636476278305054,
      "learning_rate": 6.251324148007944e-05,
      "loss": 0.0266,
      "step": 4520
    },
    {
      "epoch": 3.2127659574468086,
      "grad_norm": 0.23822833597660065,
      "learning_rate": 6.207869410728689e-05,
      "loss": 0.0158,
      "step": 4530
    },
    {
      "epoch": 3.219858156028369,
      "grad_norm": 0.24598778784275055,
      "learning_rate": 6.164498147472194e-05,
      "loss": 0.0134,
      "step": 4540
    },
    {
      "epoch": 3.226950354609929,
      "grad_norm": 0.21434803307056427,
      "learning_rate": 6.121211312945518e-05,
      "loss": 0.0223,
      "step": 4550
    },
    {
      "epoch": 3.2340425531914896,
      "grad_norm": 0.2982153296470642,
      "learning_rate": 6.078009859997227e-05,
      "loss": 0.0194,
      "step": 4560
    },
    {
      "epoch": 3.2411347517730498,
      "grad_norm": 0.29648178815841675,
      "learning_rate": 6.034894739596442e-05,
      "loss": 0.0188,
      "step": 4570
    },
    {
      "epoch": 3.24822695035461,
      "grad_norm": 0.27711746096611023,
      "learning_rate": 5.991866900811876e-05,
      "loss": 0.0182,
      "step": 4580
    },
    {
      "epoch": 3.25531914893617,
      "grad_norm": 0.23644205927848816,
      "learning_rate": 5.9489272907909845e-05,
      "loss": 0.0153,
      "step": 4590
    },
    {
      "epoch": 3.2624113475177303,
      "grad_norm": 0.2650356590747833,
      "learning_rate": 5.906076854739074e-05,
      "loss": 0.0154,
      "step": 4600
    },
    {
      "epoch": 3.269503546099291,
      "grad_norm": 0.3089609146118164,
      "learning_rate": 5.863316535898527e-05,
      "loss": 0.0225,
      "step": 4610
    },
    {
      "epoch": 3.276595744680851,
      "grad_norm": 0.29182925820350647,
      "learning_rate": 5.8206472755280206e-05,
      "loss": 0.025,
      "step": 4620
    },
    {
      "epoch": 3.2836879432624113,
      "grad_norm": 0.14777465164661407,
      "learning_rate": 5.778070012881813e-05,
      "loss": 0.0177,
      "step": 4630
    },
    {
      "epoch": 3.2907801418439715,
      "grad_norm": 0.29156795144081116,
      "learning_rate": 5.735585685189075e-05,
      "loss": 0.0157,
      "step": 4640
    },
    {
      "epoch": 3.297872340425532,
      "grad_norm": 0.3428608179092407,
      "learning_rate": 5.6931952276332436e-05,
      "loss": 0.0204,
      "step": 4650
    },
    {
      "epoch": 3.3049645390070923,
      "grad_norm": 0.14621183276176453,
      "learning_rate": 5.6508995733314576e-05,
      "loss": 0.0222,
      "step": 4660
    },
    {
      "epoch": 3.3120567375886525,
      "grad_norm": 0.2761157155036926,
      "learning_rate": 5.6086996533139894e-05,
      "loss": 0.027,
      "step": 4670
    },
    {
      "epoch": 3.3191489361702127,
      "grad_norm": 0.25750598311424255,
      "learning_rate": 5.566596396503785e-05,
      "loss": 0.0128,
      "step": 4680
    },
    {
      "epoch": 3.326241134751773,
      "grad_norm": 0.14341020584106445,
      "learning_rate": 5.5245907296959795e-05,
      "loss": 0.0227,
      "step": 4690
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.3929702043533325,
      "learning_rate": 5.4826835775375285e-05,
      "loss": 0.0211,
      "step": 4700
    },
    {
      "epoch": 3.3404255319148937,
      "grad_norm": 0.16511176526546478,
      "learning_rate": 5.440875862506838e-05,
      "loss": 0.0174,
      "step": 4710
    },
    {
      "epoch": 3.347517730496454,
      "grad_norm": 0.3370586335659027,
      "learning_rate": 5.3991685048934604e-05,
      "loss": 0.0154,
      "step": 4720
    },
    {
      "epoch": 3.354609929078014,
      "grad_norm": 0.2349589467048645,
      "learning_rate": 5.357562422777829e-05,
      "loss": 0.0158,
      "step": 4730
    },
    {
      "epoch": 3.3617021276595747,
      "grad_norm": 0.14333926141262054,
      "learning_rate": 5.316058532011068e-05,
      "loss": 0.0217,
      "step": 4740
    },
    {
      "epoch": 3.368794326241135,
      "grad_norm": 0.277540385723114,
      "learning_rate": 5.2746577461948166e-05,
      "loss": 0.0155,
      "step": 4750
    },
    {
      "epoch": 3.375886524822695,
      "grad_norm": 0.20732851326465607,
      "learning_rate": 5.233360976661117e-05,
      "loss": 0.0145,
      "step": 4760
    },
    {
      "epoch": 3.382978723404255,
      "grad_norm": 0.21173764765262604,
      "learning_rate": 5.1921691324523716e-05,
      "loss": 0.0166,
      "step": 4770
    },
    {
      "epoch": 3.3900709219858154,
      "grad_norm": 0.1785140335559845,
      "learning_rate": 5.151083120301314e-05,
      "loss": 0.0243,
      "step": 4780
    },
    {
      "epoch": 3.397163120567376,
      "grad_norm": 0.22786866128444672,
      "learning_rate": 5.110103844611064e-05,
      "loss": 0.0211,
      "step": 4790
    },
    {
      "epoch": 3.404255319148936,
      "grad_norm": 0.3462839424610138,
      "learning_rate": 5.069232207435196e-05,
      "loss": 0.0172,
      "step": 4800
    },
    {
      "epoch": 3.4113475177304964,
      "grad_norm": 0.35770383477211,
      "learning_rate": 5.0284691084579186e-05,
      "loss": 0.0178,
      "step": 4810
    },
    {
      "epoch": 3.4184397163120566,
      "grad_norm": 0.18918779492378235,
      "learning_rate": 4.987815444974233e-05,
      "loss": 0.0121,
      "step": 4820
    },
    {
      "epoch": 3.425531914893617,
      "grad_norm": 0.16905491054058075,
      "learning_rate": 4.94727211187021e-05,
      "loss": 0.0163,
      "step": 4830
    },
    {
      "epoch": 3.4326241134751774,
      "grad_norm": 0.15049125254154205,
      "learning_rate": 4.906840001603279e-05,
      "loss": 0.0183,
      "step": 4840
    },
    {
      "epoch": 3.4397163120567376,
      "grad_norm": 0.24548375606536865,
      "learning_rate": 4.866520004182583e-05,
      "loss": 0.0149,
      "step": 4850
    },
    {
      "epoch": 3.4468085106382977,
      "grad_norm": 0.2544180452823639,
      "learning_rate": 4.826313007149387e-05,
      "loss": 0.0167,
      "step": 4860
    },
    {
      "epoch": 3.453900709219858,
      "grad_norm": 0.29529479146003723,
      "learning_rate": 4.7862198955575366e-05,
      "loss": 0.0221,
      "step": 4870
    },
    {
      "epoch": 3.4609929078014185,
      "grad_norm": 0.37530678510665894,
      "learning_rate": 4.746241551953997e-05,
      "loss": 0.0188,
      "step": 4880
    },
    {
      "epoch": 3.4680851063829787,
      "grad_norm": 0.17742417752742767,
      "learning_rate": 4.7063788563593946e-05,
      "loss": 0.0173,
      "step": 4890
    },
    {
      "epoch": 3.475177304964539,
      "grad_norm": 0.1323331743478775,
      "learning_rate": 4.666632686248673e-05,
      "loss": 0.0139,
      "step": 4900
    },
    {
      "epoch": 3.482269503546099,
      "grad_norm": 0.2831038236618042,
      "learning_rate": 4.6270039165317605e-05,
      "loss": 0.0223,
      "step": 4910
    },
    {
      "epoch": 3.4893617021276597,
      "grad_norm": 0.31867706775665283,
      "learning_rate": 4.587493419534324e-05,
      "loss": 0.022,
      "step": 4920
    },
    {
      "epoch": 3.49645390070922,
      "grad_norm": 0.2805313766002655,
      "learning_rate": 4.5481020649785464e-05,
      "loss": 0.0188,
      "step": 4930
    },
    {
      "epoch": 3.50354609929078,
      "grad_norm": 0.23637081682682037,
      "learning_rate": 4.508830719964012e-05,
      "loss": 0.0153,
      "step": 4940
    },
    {
      "epoch": 3.5106382978723403,
      "grad_norm": 0.2689560055732727,
      "learning_rate": 4.469680248948589e-05,
      "loss": 0.0186,
      "step": 4950
    },
    {
      "epoch": 3.5177304964539005,
      "grad_norm": 0.26819270849227905,
      "learning_rate": 4.4306515137294243e-05,
      "loss": 0.0177,
      "step": 4960
    },
    {
      "epoch": 3.524822695035461,
      "grad_norm": 0.397055059671402,
      "learning_rate": 4.3917453734239566e-05,
      "loss": 0.0149,
      "step": 4970
    },
    {
      "epoch": 3.5319148936170213,
      "grad_norm": 0.2141522765159607,
      "learning_rate": 4.352962684451016e-05,
      "loss": 0.0142,
      "step": 4980
    },
    {
      "epoch": 3.5390070921985815,
      "grad_norm": 0.19512633979320526,
      "learning_rate": 4.314304300511975e-05,
      "loss": 0.0191,
      "step": 4990
    },
    {
      "epoch": 3.546099290780142,
      "grad_norm": 0.32705944776535034,
      "learning_rate": 4.275771072571932e-05,
      "loss": 0.018,
      "step": 5000
    },
    {
      "epoch": 3.5531914893617023,
      "grad_norm": 0.21208684146404266,
      "learning_rate": 4.237363848841018e-05,
      "loss": 0.02,
      "step": 5010
    },
    {
      "epoch": 3.5602836879432624,
      "grad_norm": 0.20367835462093353,
      "learning_rate": 4.1990834747556874e-05,
      "loss": 0.0173,
      "step": 5020
    },
    {
      "epoch": 3.5673758865248226,
      "grad_norm": 0.175355926156044,
      "learning_rate": 4.160930792960138e-05,
      "loss": 0.0158,
      "step": 5030
    },
    {
      "epoch": 3.574468085106383,
      "grad_norm": 0.2119171917438507,
      "learning_rate": 4.122906643287735e-05,
      "loss": 0.0142,
      "step": 5040
    },
    {
      "epoch": 3.581560283687943,
      "grad_norm": 0.20758385956287384,
      "learning_rate": 4.0850118627425615e-05,
      "loss": 0.0137,
      "step": 5050
    },
    {
      "epoch": 3.5886524822695036,
      "grad_norm": 0.292469322681427,
      "learning_rate": 4.0472472854809475e-05,
      "loss": 0.0125,
      "step": 5060
    },
    {
      "epoch": 3.595744680851064,
      "grad_norm": 0.26145458221435547,
      "learning_rate": 4.009613742793149e-05,
      "loss": 0.0193,
      "step": 5070
    },
    {
      "epoch": 3.602836879432624,
      "grad_norm": 0.21884608268737793,
      "learning_rate": 3.972112063085017e-05,
      "loss": 0.0195,
      "step": 5080
    },
    {
      "epoch": 3.6099290780141846,
      "grad_norm": 0.2939169406890869,
      "learning_rate": 3.934743071859794e-05,
      "loss": 0.0207,
      "step": 5090
    },
    {
      "epoch": 3.617021276595745,
      "grad_norm": 0.2042369693517685,
      "learning_rate": 3.8975075916999106e-05,
      "loss": 0.0229,
      "step": 5100
    },
    {
      "epoch": 3.624113475177305,
      "grad_norm": 0.2654969394207001,
      "learning_rate": 3.8604064422489036e-05,
      "loss": 0.0173,
      "step": 5110
    },
    {
      "epoch": 3.631205673758865,
      "grad_norm": 0.3279620110988617,
      "learning_rate": 3.8234404401933666e-05,
      "loss": 0.021,
      "step": 5120
    },
    {
      "epoch": 3.6382978723404253,
      "grad_norm": 0.2576730251312256,
      "learning_rate": 3.7866103992449566e-05,
      "loss": 0.0138,
      "step": 5130
    },
    {
      "epoch": 3.645390070921986,
      "grad_norm": 0.3340737223625183,
      "learning_rate": 3.749917130122511e-05,
      "loss": 0.0146,
      "step": 5140
    },
    {
      "epoch": 3.652482269503546,
      "grad_norm": 0.18983782827854156,
      "learning_rate": 3.713361440534173e-05,
      "loss": 0.0156,
      "step": 5150
    },
    {
      "epoch": 3.6595744680851063,
      "grad_norm": 0.21235822141170502,
      "learning_rate": 3.67694413515964e-05,
      "loss": 0.0159,
      "step": 5160
    },
    {
      "epoch": 3.6666666666666665,
      "grad_norm": 0.322186678647995,
      "learning_rate": 3.64066601563242e-05,
      "loss": 0.0184,
      "step": 5170
    },
    {
      "epoch": 3.673758865248227,
      "grad_norm": 0.30648860335350037,
      "learning_rate": 3.604527880522214e-05,
      "loss": 0.0175,
      "step": 5180
    },
    {
      "epoch": 3.6808510638297873,
      "grad_norm": 0.31347954273223877,
      "learning_rate": 3.56853052531732e-05,
      "loss": 0.0113,
      "step": 5190
    },
    {
      "epoch": 3.6879432624113475,
      "grad_norm": 0.2934507727622986,
      "learning_rate": 3.532674742407134e-05,
      "loss": 0.0184,
      "step": 5200
    },
    {
      "epoch": 3.6950354609929077,
      "grad_norm": 0.141712948679924,
      "learning_rate": 3.496961321064691e-05,
      "loss": 0.018,
      "step": 5210
    },
    {
      "epoch": 3.702127659574468,
      "grad_norm": 0.1768782138824463,
      "learning_rate": 3.461391047429304e-05,
      "loss": 0.0113,
      "step": 5220
    },
    {
      "epoch": 3.7092198581560285,
      "grad_norm": 0.1911313533782959,
      "learning_rate": 3.425964704489267e-05,
      "loss": 0.0157,
      "step": 5230
    },
    {
      "epoch": 3.7163120567375887,
      "grad_norm": 0.1564929336309433,
      "learning_rate": 3.390683072064594e-05,
      "loss": 0.013,
      "step": 5240
    },
    {
      "epoch": 3.723404255319149,
      "grad_norm": 0.3047170639038086,
      "learning_rate": 3.3555469267898784e-05,
      "loss": 0.0138,
      "step": 5250
    },
    {
      "epoch": 3.7304964539007095,
      "grad_norm": 0.10271797329187393,
      "learning_rate": 3.3205570420971855e-05,
      "loss": 0.0149,
      "step": 5260
    },
    {
      "epoch": 3.7375886524822697,
      "grad_norm": 0.2823294997215271,
      "learning_rate": 3.285714188199031e-05,
      "loss": 0.018,
      "step": 5270
    },
    {
      "epoch": 3.74468085106383,
      "grad_norm": 0.19177910685539246,
      "learning_rate": 3.251019132071415e-05,
      "loss": 0.0124,
      "step": 5280
    },
    {
      "epoch": 3.75177304964539,
      "grad_norm": 0.16738402843475342,
      "learning_rate": 3.2164726374369655e-05,
      "loss": 0.0106,
      "step": 5290
    },
    {
      "epoch": 3.7588652482269502,
      "grad_norm": 0.21252532303333282,
      "learning_rate": 3.182075464748094e-05,
      "loss": 0.0209,
      "step": 5300
    },
    {
      "epoch": 3.7659574468085104,
      "grad_norm": 0.21184132993221283,
      "learning_rate": 3.1478283711702825e-05,
      "loss": 0.0228,
      "step": 5310
    },
    {
      "epoch": 3.773049645390071,
      "grad_norm": 0.17201204597949982,
      "learning_rate": 3.1137321105654074e-05,
      "loss": 0.0175,
      "step": 5320
    },
    {
      "epoch": 3.780141843971631,
      "grad_norm": 0.1786358505487442,
      "learning_rate": 3.079787433475135e-05,
      "loss": 0.0191,
      "step": 5330
    },
    {
      "epoch": 3.7872340425531914,
      "grad_norm": 0.18210577964782715,
      "learning_rate": 3.0459950871044217e-05,
      "loss": 0.0186,
      "step": 5340
    },
    {
      "epoch": 3.794326241134752,
      "grad_norm": 0.2807255685329437,
      "learning_rate": 3.0123558153050423e-05,
      "loss": 0.0194,
      "step": 5350
    },
    {
      "epoch": 3.801418439716312,
      "grad_norm": 0.18478602170944214,
      "learning_rate": 2.9788703585592382e-05,
      "loss": 0.0159,
      "step": 5360
    },
    {
      "epoch": 3.8085106382978724,
      "grad_norm": 0.14877133071422577,
      "learning_rate": 2.9455394539633974e-05,
      "loss": 0.0123,
      "step": 5370
    },
    {
      "epoch": 3.8156028368794326,
      "grad_norm": 0.2453097105026245,
      "learning_rate": 2.9123638352118455e-05,
      "loss": 0.0163,
      "step": 5380
    },
    {
      "epoch": 3.8226950354609928,
      "grad_norm": 0.3394005596637726,
      "learning_rate": 2.8793442325806886e-05,
      "loss": 0.0157,
      "step": 5390
    },
    {
      "epoch": 3.829787234042553,
      "grad_norm": 0.21939441561698914,
      "learning_rate": 2.846481372911739e-05,
      "loss": 0.0191,
      "step": 5400
    },
    {
      "epoch": 3.8368794326241136,
      "grad_norm": 0.22955432534217834,
      "learning_rate": 2.8137759795965068e-05,
      "loss": 0.0145,
      "step": 5410
    },
    {
      "epoch": 3.8439716312056738,
      "grad_norm": 0.1423809826374054,
      "learning_rate": 2.781228772560297e-05,
      "loss": 0.016,
      "step": 5420
    },
    {
      "epoch": 3.851063829787234,
      "grad_norm": 0.2662895619869232,
      "learning_rate": 2.7488404682463353e-05,
      "loss": 0.0144,
      "step": 5430
    },
    {
      "epoch": 3.8581560283687946,
      "grad_norm": 0.26522767543792725,
      "learning_rate": 2.716611779600028e-05,
      "loss": 0.0196,
      "step": 5440
    },
    {
      "epoch": 3.8652482269503547,
      "grad_norm": 0.16856004297733307,
      "learning_rate": 2.6845434160532356e-05,
      "loss": 0.0161,
      "step": 5450
    },
    {
      "epoch": 3.872340425531915,
      "grad_norm": 0.24976351857185364,
      "learning_rate": 2.652636083508686e-05,
      "loss": 0.0128,
      "step": 5460
    },
    {
      "epoch": 3.879432624113475,
      "grad_norm": 0.18612882494926453,
      "learning_rate": 2.6208904843244186e-05,
      "loss": 0.0158,
      "step": 5470
    },
    {
      "epoch": 3.8865248226950353,
      "grad_norm": 0.2666490077972412,
      "learning_rate": 2.589307317298324e-05,
      "loss": 0.0174,
      "step": 5480
    },
    {
      "epoch": 3.8936170212765955,
      "grad_norm": 0.17859995365142822,
      "learning_rate": 2.5578872776527717e-05,
      "loss": 0.011,
      "step": 5490
    },
    {
      "epoch": 3.900709219858156,
      "grad_norm": 0.18200767040252686,
      "learning_rate": 2.526631057019294e-05,
      "loss": 0.0119,
      "step": 5500
    },
    {
      "epoch": 3.9078014184397163,
      "grad_norm": 0.15403611958026886,
      "learning_rate": 2.4955393434233754e-05,
      "loss": 0.0131,
      "step": 5510
    },
    {
      "epoch": 3.9148936170212765,
      "grad_norm": 0.13963736593723297,
      "learning_rate": 2.4646128212692952e-05,
      "loss": 0.0132,
      "step": 5520
    },
    {
      "epoch": 3.921985815602837,
      "grad_norm": 0.19664759933948517,
      "learning_rate": 2.433852171325072e-05,
      "loss": 0.016,
      "step": 5530
    },
    {
      "epoch": 3.9290780141843973,
      "grad_norm": 0.25922274589538574,
      "learning_rate": 2.4032580707074637e-05,
      "loss": 0.014,
      "step": 5540
    },
    {
      "epoch": 3.9361702127659575,
      "grad_norm": 0.09897946566343307,
      "learning_rate": 2.3728311928670845e-05,
      "loss": 0.0108,
      "step": 5550
    },
    {
      "epoch": 3.9432624113475176,
      "grad_norm": 0.25340279936790466,
      "learning_rate": 2.342572207573559e-05,
      "loss": 0.02,
      "step": 5560
    },
    {
      "epoch": 3.950354609929078,
      "grad_norm": 0.2628706395626068,
      "learning_rate": 2.3124817809007903e-05,
      "loss": 0.0161,
      "step": 5570
    },
    {
      "epoch": 3.9574468085106385,
      "grad_norm": 0.27684804797172546,
      "learning_rate": 2.282560575212298e-05,
      "loss": 0.0123,
      "step": 5580
    },
    {
      "epoch": 3.9645390070921986,
      "grad_norm": 0.14096932113170624,
      "learning_rate": 2.2528092491466358e-05,
      "loss": 0.0143,
      "step": 5590
    },
    {
      "epoch": 3.971631205673759,
      "grad_norm": 0.22769907116889954,
      "learning_rate": 2.223228457602897e-05,
      "loss": 0.0192,
      "step": 5600
    },
    {
      "epoch": 3.978723404255319,
      "grad_norm": 0.2954142391681671,
      "learning_rate": 2.1938188517262838e-05,
      "loss": 0.0118,
      "step": 5610
    },
    {
      "epoch": 3.9858156028368796,
      "grad_norm": 0.23976144194602966,
      "learning_rate": 2.1645810788937993e-05,
      "loss": 0.0161,
      "step": 5620
    },
    {
      "epoch": 3.99290780141844,
      "grad_norm": 0.4473976790904999,
      "learning_rate": 2.1355157826999704e-05,
      "loss": 0.0115,
      "step": 5630
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.18849952518939972,
      "learning_rate": 2.1066236029427043e-05,
      "loss": 0.0134,
      "step": 5640
    },
    {
      "epoch": 4.00709219858156,
      "grad_norm": 0.3046376705169678,
      "learning_rate": 2.077905175609184e-05,
      "loss": 0.0145,
      "step": 5650
    },
    {
      "epoch": 4.01418439716312,
      "grad_norm": 0.2120821177959442,
      "learning_rate": 2.0493611328618876e-05,
      "loss": 0.0129,
      "step": 5660
    },
    {
      "epoch": 4.0212765957446805,
      "grad_norm": 0.14745740592479706,
      "learning_rate": 2.0209921030246613e-05,
      "loss": 0.0099,
      "step": 5670
    },
    {
      "epoch": 4.028368794326241,
      "grad_norm": 0.14253804087638855,
      "learning_rate": 1.9927987105688872e-05,
      "loss": 0.0212,
      "step": 5680
    },
    {
      "epoch": 4.035460992907802,
      "grad_norm": 0.19978418946266174,
      "learning_rate": 1.964781576099749e-05,
      "loss": 0.0093,
      "step": 5690
    },
    {
      "epoch": 4.042553191489362,
      "grad_norm": 0.16095592081546783,
      "learning_rate": 1.936941316342553e-05,
      "loss": 0.0133,
      "step": 5700
    },
    {
      "epoch": 4.049645390070922,
      "grad_norm": 0.2865225672721863,
      "learning_rate": 1.9092785441291738e-05,
      "loss": 0.0182,
      "step": 5710
    },
    {
      "epoch": 4.056737588652482,
      "grad_norm": 0.10646691173315048,
      "learning_rate": 1.8817938683845394e-05,
      "loss": 0.0103,
      "step": 5720
    },
    {
      "epoch": 4.0638297872340425,
      "grad_norm": 0.16811849176883698,
      "learning_rate": 1.8544878941132605e-05,
      "loss": 0.0151,
      "step": 5730
    },
    {
      "epoch": 4.070921985815603,
      "grad_norm": 0.23703442513942719,
      "learning_rate": 1.8273612223862734e-05,
      "loss": 0.0102,
      "step": 5740
    },
    {
      "epoch": 4.078014184397163,
      "grad_norm": 0.19330421090126038,
      "learning_rate": 1.8004144503276443e-05,
      "loss": 0.0152,
      "step": 5750
    },
    {
      "epoch": 4.085106382978723,
      "grad_norm": 0.20537878572940826,
      "learning_rate": 1.7736481711013984e-05,
      "loss": 0.0118,
      "step": 5760
    },
    {
      "epoch": 4.092198581560283,
      "grad_norm": 0.13630573451519012,
      "learning_rate": 1.7470629738984856e-05,
      "loss": 0.0102,
      "step": 5770
    },
    {
      "epoch": 4.099290780141844,
      "grad_norm": 0.38417133688926697,
      "learning_rate": 1.7206594439237865e-05,
      "loss": 0.0163,
      "step": 5780
    },
    {
      "epoch": 4.1063829787234045,
      "grad_norm": 0.3182227313518524,
      "learning_rate": 1.694438162383255e-05,
      "loss": 0.0149,
      "step": 5790
    },
    {
      "epoch": 4.113475177304965,
      "grad_norm": 0.16077764332294464,
      "learning_rate": 1.6683997064711132e-05,
      "loss": 0.0126,
      "step": 5800
    },
    {
      "epoch": 4.120567375886525,
      "grad_norm": 0.4243755340576172,
      "learning_rate": 1.6425446493571363e-05,
      "loss": 0.0187,
      "step": 5810
    },
    {
      "epoch": 4.127659574468085,
      "grad_norm": 0.22879810631275177,
      "learning_rate": 1.616873560174057e-05,
      "loss": 0.0213,
      "step": 5820
    },
    {
      "epoch": 4.134751773049645,
      "grad_norm": 0.1545029729604721,
      "learning_rate": 1.5913870040050162e-05,
      "loss": 0.0137,
      "step": 5830
    },
    {
      "epoch": 4.141843971631205,
      "grad_norm": 0.2279650866985321,
      "learning_rate": 1.566085541871145e-05,
      "loss": 0.0115,
      "step": 5840
    },
    {
      "epoch": 4.148936170212766,
      "grad_norm": 0.26870498061180115,
      "learning_rate": 1.540969730719193e-05,
      "loss": 0.011,
      "step": 5850
    },
    {
      "epoch": 4.156028368794326,
      "grad_norm": 0.17166127264499664,
      "learning_rate": 1.5160401234092868e-05,
      "loss": 0.0144,
      "step": 5860
    },
    {
      "epoch": 4.163120567375887,
      "grad_norm": 0.20089754462242126,
      "learning_rate": 1.4912972687027527e-05,
      "loss": 0.0122,
      "step": 5870
    },
    {
      "epoch": 4.170212765957447,
      "grad_norm": 0.14342001080513,
      "learning_rate": 1.4667417112500393e-05,
      "loss": 0.0176,
      "step": 5880
    },
    {
      "epoch": 4.177304964539007,
      "grad_norm": 0.2057809829711914,
      "learning_rate": 1.4423739915787183e-05,
      "loss": 0.0114,
      "step": 5890
    },
    {
      "epoch": 4.184397163120567,
      "grad_norm": 0.19354157149791718,
      "learning_rate": 1.4181946460816064e-05,
      "loss": 0.0114,
      "step": 5900
    },
    {
      "epoch": 4.191489361702128,
      "grad_norm": 0.23628641664981842,
      "learning_rate": 1.3942042070049389e-05,
      "loss": 0.0197,
      "step": 5910
    },
    {
      "epoch": 4.198581560283688,
      "grad_norm": 0.10524198412895203,
      "learning_rate": 1.3704032024366597e-05,
      "loss": 0.0135,
      "step": 5920
    },
    {
      "epoch": 4.205673758865248,
      "grad_norm": 0.19477611780166626,
      "learning_rate": 1.3467921562948039e-05,
      "loss": 0.0134,
      "step": 5930
    },
    {
      "epoch": 4.212765957446808,
      "grad_norm": 0.15423716604709625,
      "learning_rate": 1.3233715883159559e-05,
      "loss": 0.0169,
      "step": 5940
    },
    {
      "epoch": 4.219858156028369,
      "grad_norm": 0.18021738529205322,
      "learning_rate": 1.3001420140438158e-05,
      "loss": 0.0101,
      "step": 5950
    },
    {
      "epoch": 4.226950354609929,
      "grad_norm": 0.20674963295459747,
      "learning_rate": 1.27710394481784e-05,
      "loss": 0.012,
      "step": 5960
    },
    {
      "epoch": 4.23404255319149,
      "grad_norm": 0.2461792379617691,
      "learning_rate": 1.2542578877620026e-05,
      "loss": 0.0116,
      "step": 5970
    },
    {
      "epoch": 4.24113475177305,
      "grad_norm": 0.20844022929668427,
      "learning_rate": 1.2316043457736127e-05,
      "loss": 0.0091,
      "step": 5980
    },
    {
      "epoch": 4.24822695035461,
      "grad_norm": 0.16048000752925873,
      "learning_rate": 1.2091438175122593e-05,
      "loss": 0.0114,
      "step": 5990
    },
    {
      "epoch": 4.25531914893617,
      "grad_norm": 0.22476951777935028,
      "learning_rate": 1.1868767973888317e-05,
      "loss": 0.0193,
      "step": 6000
    },
    {
      "epoch": 4.26241134751773,
      "grad_norm": 0.1669793426990509,
      "learning_rate": 1.1648037755546325e-05,
      "loss": 0.0115,
      "step": 6010
    },
    {
      "epoch": 4.2695035460992905,
      "grad_norm": 0.07553170621395111,
      "learning_rate": 1.1429252378905875e-05,
      "loss": 0.0108,
      "step": 6020
    },
    {
      "epoch": 4.276595744680851,
      "grad_norm": 0.24128834903240204,
      "learning_rate": 1.1212416659965507e-05,
      "loss": 0.0181,
      "step": 6030
    },
    {
      "epoch": 4.283687943262412,
      "grad_norm": 0.24184225499629974,
      "learning_rate": 1.0997535371807143e-05,
      "loss": 0.0151,
      "step": 6040
    },
    {
      "epoch": 4.290780141843972,
      "grad_norm": 0.22201502323150635,
      "learning_rate": 1.0784613244490816e-05,
      "loss": 0.011,
      "step": 6050
    },
    {
      "epoch": 4.297872340425532,
      "grad_norm": 0.12740255892276764,
      "learning_rate": 1.057365496495073e-05,
      "loss": 0.013,
      "step": 6060
    },
    {
      "epoch": 4.304964539007092,
      "grad_norm": 0.2595222592353821,
      "learning_rate": 1.036466517689202e-05,
      "loss": 0.0116,
      "step": 6070
    },
    {
      "epoch": 4.3120567375886525,
      "grad_norm": 0.12973222136497498,
      "learning_rate": 1.015764848068852e-05,
      "loss": 0.0099,
      "step": 6080
    },
    {
      "epoch": 4.319148936170213,
      "grad_norm": 0.2121812403202057,
      "learning_rate": 9.95260943328148e-06,
      "loss": 0.0096,
      "step": 6090
    },
    {
      "epoch": 4.326241134751773,
      "grad_norm": 0.24669013917446136,
      "learning_rate": 9.749552548079344e-06,
      "loss": 0.0186,
      "step": 6100
    },
    {
      "epoch": 4.333333333333333,
      "grad_norm": 0.24903731048107147,
      "learning_rate": 9.548482294858285e-06,
      "loss": 0.0154,
      "step": 6110
    },
    {
      "epoch": 4.340425531914893,
      "grad_norm": 0.15923161804676056,
      "learning_rate": 9.349403099663923e-06,
      "loss": 0.0125,
      "step": 6120
    },
    {
      "epoch": 4.347517730496454,
      "grad_norm": 0.2829233407974243,
      "learning_rate": 9.1523193447138e-06,
      "loss": 0.0178,
      "step": 6130
    },
    {
      "epoch": 4.3546099290780145,
      "grad_norm": 0.27781230211257935,
      "learning_rate": 8.957235368300986e-06,
      "loss": 0.0133,
      "step": 6140
    },
    {
      "epoch": 4.361702127659575,
      "grad_norm": 0.14426802098751068,
      "learning_rate": 8.764155464698597e-06,
      "loss": 0.0106,
      "step": 6150
    },
    {
      "epoch": 4.368794326241135,
      "grad_norm": 0.17137634754180908,
      "learning_rate": 8.573083884065126e-06,
      "loss": 0.015,
      "step": 6160
    },
    {
      "epoch": 4.375886524822695,
      "grad_norm": 0.1472645103931427,
      "learning_rate": 8.384024832351112e-06,
      "loss": 0.0134,
      "step": 6170
    },
    {
      "epoch": 4.382978723404255,
      "grad_norm": 0.23173321783542633,
      "learning_rate": 8.196982471206326e-06,
      "loss": 0.0175,
      "step": 6180
    },
    {
      "epoch": 4.390070921985815,
      "grad_norm": 0.2875099182128906,
      "learning_rate": 8.011960917888339e-06,
      "loss": 0.0132,
      "step": 6190
    },
    {
      "epoch": 4.397163120567376,
      "grad_norm": 0.17926155030727386,
      "learning_rate": 7.82896424517181e-06,
      "loss": 0.0141,
      "step": 6200
    },
    {
      "epoch": 4.404255319148936,
      "grad_norm": 0.2918379306793213,
      "learning_rate": 7.647996481258868e-06,
      "loss": 0.0166,
      "step": 6210
    },
    {
      "epoch": 4.411347517730497,
      "grad_norm": 0.1396457999944687,
      "learning_rate": 7.469061609690353e-06,
      "loss": 0.0181,
      "step": 6220
    },
    {
      "epoch": 4.418439716312057,
      "grad_norm": 0.2047400027513504,
      "learning_rate": 7.292163569258281e-06,
      "loss": 0.0122,
      "step": 6230
    },
    {
      "epoch": 4.425531914893617,
      "grad_norm": 0.18662038445472717,
      "learning_rate": 7.117306253918999e-06,
      "loss": 0.0142,
      "step": 6240
    },
    {
      "epoch": 4.432624113475177,
      "grad_norm": 0.1682143360376358,
      "learning_rate": 6.94449351270754e-06,
      "loss": 0.0106,
      "step": 6250
    },
    {
      "epoch": 4.439716312056738,
      "grad_norm": 0.08281824737787247,
      "learning_rate": 6.773729149652907e-06,
      "loss": 0.0175,
      "step": 6260
    },
    {
      "epoch": 4.446808510638298,
      "grad_norm": 0.25101786851882935,
      "learning_rate": 6.605016923694307e-06,
      "loss": 0.0148,
      "step": 6270
    },
    {
      "epoch": 4.453900709219858,
      "grad_norm": 0.4814547598361969,
      "learning_rate": 6.43836054859841e-06,
      "loss": 0.0137,
      "step": 6280
    },
    {
      "epoch": 4.460992907801418,
      "grad_norm": 0.18742209672927856,
      "learning_rate": 6.273763692877588e-06,
      "loss": 0.0106,
      "step": 6290
    },
    {
      "epoch": 4.468085106382979,
      "grad_norm": 0.15110346674919128,
      "learning_rate": 6.11122997970921e-06,
      "loss": 0.0156,
      "step": 6300
    },
    {
      "epoch": 4.475177304964539,
      "grad_norm": 0.21841980516910553,
      "learning_rate": 5.9507629868558404e-06,
      "loss": 0.0087,
      "step": 6310
    },
    {
      "epoch": 4.4822695035460995,
      "grad_norm": 0.2773948013782501,
      "learning_rate": 5.792366246586511e-06,
      "loss": 0.0111,
      "step": 6320
    },
    {
      "epoch": 4.48936170212766,
      "grad_norm": 0.168336421251297,
      "learning_rate": 5.636043245598932e-06,
      "loss": 0.0122,
      "step": 6330
    },
    {
      "epoch": 4.49645390070922,
      "grad_norm": 0.22542957961559296,
      "learning_rate": 5.4817974249427914e-06,
      "loss": 0.0175,
      "step": 6340
    },
    {
      "epoch": 4.50354609929078,
      "grad_norm": 0.22203271090984344,
      "learning_rate": 5.329632179943977e-06,
      "loss": 0.0089,
      "step": 6350
    },
    {
      "epoch": 4.51063829787234,
      "grad_norm": 0.07676347345113754,
      "learning_rate": 5.179550860129823e-06,
      "loss": 0.0123,
      "step": 6360
    },
    {
      "epoch": 4.5177304964539005,
      "grad_norm": 0.1949254721403122,
      "learning_rate": 5.031556769155444e-06,
      "loss": 0.0132,
      "step": 6370
    },
    {
      "epoch": 4.524822695035461,
      "grad_norm": 0.2822635769844055,
      "learning_rate": 4.885653164730908e-06,
      "loss": 0.0128,
      "step": 6380
    },
    {
      "epoch": 4.531914893617021,
      "grad_norm": 0.22830314934253693,
      "learning_rate": 4.741843258549639e-06,
      "loss": 0.0169,
      "step": 6390
    },
    {
      "epoch": 4.539007092198582,
      "grad_norm": 0.3081783354282379,
      "learning_rate": 4.600130216217613e-06,
      "loss": 0.0134,
      "step": 6400
    },
    {
      "epoch": 4.546099290780142,
      "grad_norm": 0.2014082372188568,
      "learning_rate": 4.460517157183819e-06,
      "loss": 0.0139,
      "step": 6410
    },
    {
      "epoch": 4.553191489361702,
      "grad_norm": 0.2044125497341156,
      "learning_rate": 4.323007154671399e-06,
      "loss": 0.0134,
      "step": 6420
    },
    {
      "epoch": 4.560283687943262,
      "grad_norm": 0.17926239967346191,
      "learning_rate": 4.1876032356101825e-06,
      "loss": 0.0131,
      "step": 6430
    },
    {
      "epoch": 4.567375886524823,
      "grad_norm": 0.1999116986989975,
      "learning_rate": 4.054308380569927e-06,
      "loss": 0.0089,
      "step": 6440
    },
    {
      "epoch": 4.574468085106383,
      "grad_norm": 0.3048093020915985,
      "learning_rate": 3.9231255236947925e-06,
      "loss": 0.0126,
      "step": 6450
    },
    {
      "epoch": 4.581560283687943,
      "grad_norm": 0.11860720068216324,
      "learning_rate": 3.7940575526386857e-06,
      "loss": 0.0111,
      "step": 6460
    },
    {
      "epoch": 4.588652482269503,
      "grad_norm": 0.25165706872940063,
      "learning_rate": 3.667107308501749e-06,
      "loss": 0.0099,
      "step": 6470
    },
    {
      "epoch": 4.595744680851064,
      "grad_norm": 0.18574100732803345,
      "learning_rate": 3.542277585767828e-06,
      "loss": 0.0093,
      "step": 6480
    },
    {
      "epoch": 4.602836879432624,
      "grad_norm": 0.23088154196739197,
      "learning_rate": 3.4195711322428716e-06,
      "loss": 0.008,
      "step": 6490
    },
    {
      "epoch": 4.609929078014185,
      "grad_norm": 0.13895224034786224,
      "learning_rate": 3.298990648994571e-06,
      "loss": 0.0098,
      "step": 6500
    },
    {
      "epoch": 4.617021276595745,
      "grad_norm": 0.2849128246307373,
      "learning_rate": 3.1805387902927754e-06,
      "loss": 0.0225,
      "step": 6510
    },
    {
      "epoch": 4.624113475177305,
      "grad_norm": 0.2238398641347885,
      "learning_rate": 3.06421816355118e-06,
      "loss": 0.0158,
      "step": 6520
    },
    {
      "epoch": 4.631205673758865,
      "grad_norm": 0.2368791699409485,
      "learning_rate": 2.950031329269831e-06,
      "loss": 0.0104,
      "step": 6530
    },
    {
      "epoch": 4.638297872340425,
      "grad_norm": 0.2200658917427063,
      "learning_rate": 2.8379808009788677e-06,
      "loss": 0.0158,
      "step": 6540
    },
    {
      "epoch": 4.6453900709219855,
      "grad_norm": 0.17251890897750854,
      "learning_rate": 2.728069045183068e-06,
      "loss": 0.0139,
      "step": 6550
    },
    {
      "epoch": 4.652482269503546,
      "grad_norm": 0.13242077827453613,
      "learning_rate": 2.620298481307704e-06,
      "loss": 0.0083,
      "step": 6560
    },
    {
      "epoch": 4.659574468085106,
      "grad_norm": 0.2232428342103958,
      "learning_rate": 2.5146714816451033e-06,
      "loss": 0.0165,
      "step": 6570
    },
    {
      "epoch": 4.666666666666667,
      "grad_norm": 0.164382204413414,
      "learning_rate": 2.4111903713026294e-06,
      "loss": 0.0111,
      "step": 6580
    },
    {
      "epoch": 4.673758865248227,
      "grad_norm": 0.14152784645557404,
      "learning_rate": 2.3098574281513185e-06,
      "loss": 0.0114,
      "step": 6590
    },
    {
      "epoch": 4.680851063829787,
      "grad_norm": 0.2582566440105438,
      "learning_rate": 2.210674882775854e-06,
      "loss": 0.0103,
      "step": 6600
    },
    {
      "epoch": 4.6879432624113475,
      "grad_norm": 0.12841904163360596,
      "learning_rate": 2.1136449184254393e-06,
      "loss": 0.0075,
      "step": 6610
    },
    {
      "epoch": 4.695035460992908,
      "grad_norm": 0.14777736365795135,
      "learning_rate": 2.0187696709657145e-06,
      "loss": 0.0133,
      "step": 6620
    },
    {
      "epoch": 4.702127659574468,
      "grad_norm": 0.14513316750526428,
      "learning_rate": 1.9260512288317823e-06,
      "loss": 0.0108,
      "step": 6630
    },
    {
      "epoch": 4.709219858156028,
      "grad_norm": 0.21794749796390533,
      "learning_rate": 1.835491632982178e-06,
      "loss": 0.0105,
      "step": 6640
    },
    {
      "epoch": 4.716312056737589,
      "grad_norm": 0.31420379877090454,
      "learning_rate": 1.747092876853984e-06,
      "loss": 0.0108,
      "step": 6650
    },
    {
      "epoch": 4.723404255319149,
      "grad_norm": 0.1831444799900055,
      "learning_rate": 1.6608569063189638e-06,
      "loss": 0.012,
      "step": 6660
    },
    {
      "epoch": 4.7304964539007095,
      "grad_norm": 0.17518573999404907,
      "learning_rate": 1.5767856196406749e-06,
      "loss": 0.016,
      "step": 6670
    },
    {
      "epoch": 4.73758865248227,
      "grad_norm": 0.33270004391670227,
      "learning_rate": 1.4948808674327464e-06,
      "loss": 0.0137,
      "step": 6680
    },
    {
      "epoch": 4.74468085106383,
      "grad_norm": 0.14303705096244812,
      "learning_rate": 1.4151444526181112e-06,
      "loss": 0.0108,
      "step": 6690
    },
    {
      "epoch": 4.75177304964539,
      "grad_norm": 0.37515661120414734,
      "learning_rate": 1.337578130389272e-06,
      "loss": 0.0162,
      "step": 6700
    },
    {
      "epoch": 4.75886524822695,
      "grad_norm": 0.12578299641609192,
      "learning_rate": 1.2621836081697647e-06,
      "loss": 0.0093,
      "step": 6710
    },
    {
      "epoch": 4.76595744680851,
      "grad_norm": 0.23699671030044556,
      "learning_rate": 1.1889625455764997e-06,
      "loss": 0.0102,
      "step": 6720
    },
    {
      "epoch": 4.773049645390071,
      "grad_norm": 0.1644619107246399,
      "learning_rate": 1.117916554383258e-06,
      "loss": 0.0179,
      "step": 6730
    },
    {
      "epoch": 4.780141843971631,
      "grad_norm": 0.139451801776886,
      "learning_rate": 1.0490471984851868e-06,
      "loss": 0.0167,
      "step": 6740
    },
    {
      "epoch": 4.787234042553192,
      "grad_norm": 0.2277289628982544,
      "learning_rate": 9.823559938644255e-07,
      "loss": 0.0139,
      "step": 6750
    },
    {
      "epoch": 4.794326241134752,
      "grad_norm": 0.16943824291229248,
      "learning_rate": 9.178444085566895e-07,
      "loss": 0.0148,
      "step": 6760
    },
    {
      "epoch": 4.801418439716312,
      "grad_norm": 0.1669815182685852,
      "learning_rate": 8.555138626189618e-07,
      "loss": 0.0173,
      "step": 6770
    },
    {
      "epoch": 4.808510638297872,
      "grad_norm": 0.380924791097641,
      "learning_rate": 7.953657280982407e-07,
      "loss": 0.0131,
      "step": 6780
    },
    {
      "epoch": 4.815602836879433,
      "grad_norm": 0.14774571359157562,
      "learning_rate": 7.374013290013415e-07,
      "loss": 0.0108,
      "step": 6790
    },
    {
      "epoch": 4.822695035460993,
      "grad_norm": 0.26690518856048584,
      "learning_rate": 6.816219412657531e-07,
      "loss": 0.0191,
      "step": 6800
    },
    {
      "epoch": 4.829787234042553,
      "grad_norm": 0.34515464305877686,
      "learning_rate": 6.280287927315498e-07,
      "loss": 0.0138,
      "step": 6810
    },
    {
      "epoch": 4.836879432624113,
      "grad_norm": 0.1968574970960617,
      "learning_rate": 5.766230631143455e-07,
      "loss": 0.0125,
      "step": 6820
    },
    {
      "epoch": 4.843971631205674,
      "grad_norm": 0.24029354751110077,
      "learning_rate": 5.274058839793594e-07,
      "loss": 0.0134,
      "step": 6830
    },
    {
      "epoch": 4.851063829787234,
      "grad_norm": 0.19054192304611206,
      "learning_rate": 4.803783387164696e-07,
      "loss": 0.0156,
      "step": 6840
    },
    {
      "epoch": 4.858156028368795,
      "grad_norm": 0.08260553330183029,
      "learning_rate": 4.355414625163978e-07,
      "loss": 0.0097,
      "step": 6850
    },
    {
      "epoch": 4.865248226950355,
      "grad_norm": 0.16334979236125946,
      "learning_rate": 3.9289624234790656e-07,
      "loss": 0.0109,
      "step": 6860
    },
    {
      "epoch": 4.872340425531915,
      "grad_norm": 0.2816116213798523,
      "learning_rate": 3.524436169360601e-07,
      "loss": 0.0104,
      "step": 6870
    },
    {
      "epoch": 4.879432624113475,
      "grad_norm": 0.28883612155914307,
      "learning_rate": 3.141844767415969e-07,
      "loss": 0.0176,
      "step": 6880
    },
    {
      "epoch": 4.886524822695035,
      "grad_norm": 0.154693603515625,
      "learning_rate": 2.7811966394131195e-07,
      "loss": 0.0119,
      "step": 6890
    },
    {
      "epoch": 4.8936170212765955,
      "grad_norm": 0.1789465993642807,
      "learning_rate": 2.4424997240948266e-07,
      "loss": 0.0103,
      "step": 6900
    },
    {
      "epoch": 4.900709219858156,
      "grad_norm": 0.2560679018497467,
      "learning_rate": 2.1257614770046063e-07,
      "loss": 0.0171,
      "step": 6910
    },
    {
      "epoch": 4.907801418439716,
      "grad_norm": 0.14788945019245148,
      "learning_rate": 1.830988870321848e-07,
      "loss": 0.0129,
      "step": 6920
    },
    {
      "epoch": 4.914893617021277,
      "grad_norm": 0.2480207085609436,
      "learning_rate": 1.55818839270927e-07,
      "loss": 0.013,
      "step": 6930
    },
    {
      "epoch": 4.921985815602837,
      "grad_norm": 0.29826200008392334,
      "learning_rate": 1.3073660491691452e-07,
      "loss": 0.0105,
      "step": 6940
    },
    {
      "epoch": 4.929078014184397,
      "grad_norm": 0.1601097583770752,
      "learning_rate": 1.0785273609115188e-07,
      "loss": 0.0097,
      "step": 6950
    },
    {
      "epoch": 4.9361702127659575,
      "grad_norm": 0.15834511816501617,
      "learning_rate": 8.716773652330812e-08,
      "loss": 0.0147,
      "step": 6960
    },
    {
      "epoch": 4.943262411347518,
      "grad_norm": 0.06883599609136581,
      "learning_rate": 6.868206154054812e-08,
      "loss": 0.0109,
      "step": 6970
    },
    {
      "epoch": 4.950354609929078,
      "grad_norm": 0.13073433935642242,
      "learning_rate": 5.2396118057562724e-08,
      "loss": 0.0113,
      "step": 6980
    },
    {
      "epoch": 4.957446808510638,
      "grad_norm": 0.19870930910110474,
      "learning_rate": 3.831026456760922e-08,
      "loss": 0.0217,
      "step": 6990
    },
    {
      "epoch": 4.964539007092198,
      "grad_norm": 0.09622751176357269,
      "learning_rate": 2.6424811134584393e-08,
      "loss": 0.0155,
      "step": 7000
    },
    {
      "epoch": 4.971631205673759,
      "grad_norm": 0.16659928858280182,
      "learning_rate": 1.674001938626324e-08,
      "loss": 0.0162,
      "step": 7010
    },
    {
      "epoch": 4.9787234042553195,
      "grad_norm": 0.2666969299316406,
      "learning_rate": 9.256102508481412e-09,
      "loss": 0.0121,
      "step": 7020
    },
    {
      "epoch": 4.98581560283688,
      "grad_norm": 0.1449095904827118,
      "learning_rate": 3.973225240450073e-09,
      "loss": 0.0118,
      "step": 7030
    },
    {
      "epoch": 4.99290780141844,
      "grad_norm": 0.13001736998558044,
      "learning_rate": 8.915038711476697e-10,
      "loss": 0.0109,
      "step": 7040
    },
    {
      "epoch": 4.999290780141844,
      "step": 7049,
      "total_flos": 2.471599871280288e+17,
      "train_loss": 0.03227324262610191,
      "train_runtime": 3151.6449,
      "train_samples_per_second": 35.786,
      "train_steps_per_second": 2.237
    }
  ],
  "logging_steps": 10,
  "max_steps": 7049,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.471599871280288e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}