{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500.0,
  "global_step": 36350,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.001375515818431912,
      "grad_norm": 0.06477122753858566,
      "learning_rate": 0.0001,
      "loss": 1.7497,
      "step": 50
    },
    {
      "epoch": 0.002751031636863824,
      "grad_norm": 0.08175177872180939,
      "learning_rate": 0.0001,
      "loss": 1.7218,
      "step": 100
    },
    {
      "epoch": 0.0041265474552957355,
      "grad_norm": 0.08908290416002274,
      "learning_rate": 0.0001,
      "loss": 1.7102,
      "step": 150
    },
    {
      "epoch": 0.005502063273727648,
      "grad_norm": 0.08645089715719223,
      "learning_rate": 0.0001,
      "loss": 1.6996,
      "step": 200
    },
    {
      "epoch": 0.0068775790921595595,
      "grad_norm": 0.07688608765602112,
      "learning_rate": 0.0001,
      "loss": 1.6968,
      "step": 250
    },
    {
      "epoch": 0.008253094910591471,
      "grad_norm": 0.08298292011022568,
      "learning_rate": 0.0001,
      "loss": 1.6922,
      "step": 300
    },
    {
      "epoch": 0.009628610729023384,
      "grad_norm": 0.07124519348144531,
      "learning_rate": 0.0001,
      "loss": 1.6874,
      "step": 350
    },
    {
      "epoch": 0.011004126547455296,
      "grad_norm": 0.0821714997291565,
      "learning_rate": 0.0001,
      "loss": 1.6838,
      "step": 400
    },
    {
      "epoch": 0.012379642365887207,
      "grad_norm": 0.11138464510440826,
      "learning_rate": 0.0001,
      "loss": 1.6807,
      "step": 450
    },
    {
      "epoch": 0.013755158184319119,
      "grad_norm": 0.09057251363992691,
      "learning_rate": 0.0001,
      "loss": 1.6772,
      "step": 500
    },
    {
      "epoch": 0.015130674002751032,
      "grad_norm": 0.10090968757867813,
      "learning_rate": 0.0001,
      "loss": 1.6718,
      "step": 550
    },
    {
      "epoch": 0.016506189821182942,
      "grad_norm": 0.08569780737161636,
      "learning_rate": 0.0001,
      "loss": 1.6702,
      "step": 600
    },
    {
      "epoch": 0.017881705639614855,
      "grad_norm": 0.07728252559900284,
      "learning_rate": 0.0001,
      "loss": 1.6671,
      "step": 650
    },
    {
      "epoch": 0.01925722145804677,
      "grad_norm": 0.08100250363349915,
      "learning_rate": 0.0001,
      "loss": 1.6641,
      "step": 700
    },
    {
      "epoch": 0.02063273727647868,
      "grad_norm": 0.09590116143226624,
      "learning_rate": 0.0001,
      "loss": 1.6616,
      "step": 750
    },
    {
      "epoch": 0.02200825309491059,
      "grad_norm": 0.10437134653329849,
      "learning_rate": 0.0001,
      "loss": 1.6607,
      "step": 800
    },
    {
      "epoch": 0.023383768913342505,
      "grad_norm": 0.08097755908966064,
      "learning_rate": 0.0001,
      "loss": 1.6578,
      "step": 850
    },
    {
      "epoch": 0.024759284731774415,
      "grad_norm": 0.08555827289819717,
      "learning_rate": 0.0001,
      "loss": 1.6548,
      "step": 900
    },
    {
      "epoch": 0.026134800550206328,
      "grad_norm": 0.10720808058977127,
      "learning_rate": 0.0001,
      "loss": 1.6528,
      "step": 950
    },
    {
      "epoch": 0.027510316368638238,
      "grad_norm": 0.11773797124624252,
      "learning_rate": 0.0001,
      "loss": 1.6511,
      "step": 1000
    },
    {
      "epoch": 0.02888583218707015,
      "grad_norm": 0.10159046947956085,
      "learning_rate": 0.0001,
      "loss": 1.6474,
      "step": 1050
    },
    {
      "epoch": 0.030261348005502064,
      "grad_norm": 0.08796145766973495,
      "learning_rate": 0.0001,
      "loss": 1.647,
      "step": 1100
    },
    {
      "epoch": 0.03163686382393398,
      "grad_norm": 0.08194500207901001,
      "learning_rate": 0.0001,
      "loss": 1.6459,
      "step": 1150
    },
    {
      "epoch": 0.033012379642365884,
      "grad_norm": 0.0940510481595993,
      "learning_rate": 0.0001,
      "loss": 1.6429,
      "step": 1200
    },
    {
      "epoch": 0.0343878954607978,
      "grad_norm": 0.08046701550483704,
      "learning_rate": 0.0001,
      "loss": 1.6407,
      "step": 1250
    },
    {
      "epoch": 0.03576341127922971,
      "grad_norm": 0.07953349500894547,
      "learning_rate": 0.0001,
      "loss": 1.6407,
      "step": 1300
    },
    {
      "epoch": 0.037138927097661624,
      "grad_norm": 0.0876886174082756,
      "learning_rate": 0.0001,
      "loss": 1.6378,
      "step": 1350
    },
    {
      "epoch": 0.03851444291609354,
      "grad_norm": 0.0870981439948082,
      "learning_rate": 0.0001,
      "loss": 1.6408,
      "step": 1400
    },
    {
      "epoch": 0.039889958734525444,
      "grad_norm": 0.09412265568971634,
      "learning_rate": 0.0001,
      "loss": 1.638,
      "step": 1450
    },
    {
      "epoch": 0.04126547455295736,
      "grad_norm": 0.08362641930580139,
      "learning_rate": 0.0001,
      "loss": 1.6344,
      "step": 1500
    },
    {
      "epoch": 0.04264099037138927,
      "grad_norm": 0.11198284476995468,
      "learning_rate": 0.0001,
      "loss": 1.6354,
      "step": 1550
    },
    {
      "epoch": 0.04401650618982118,
      "grad_norm": 0.09470899403095245,
      "learning_rate": 0.0001,
      "loss": 1.6337,
      "step": 1600
    },
    {
      "epoch": 0.0453920220082531,
      "grad_norm": 0.11157640069723129,
      "learning_rate": 0.0001,
      "loss": 1.6316,
      "step": 1650
    },
    {
      "epoch": 0.04676753782668501,
      "grad_norm": 0.08970475941896439,
      "learning_rate": 0.0001,
      "loss": 1.6324,
      "step": 1700
    },
    {
      "epoch": 0.048143053645116916,
      "grad_norm": 0.09438284486532211,
      "learning_rate": 0.0001,
      "loss": 1.6299,
      "step": 1750
    },
    {
      "epoch": 0.04951856946354883,
      "grad_norm": 0.09604686498641968,
      "learning_rate": 0.0001,
      "loss": 1.6278,
      "step": 1800
    },
    {
      "epoch": 0.05089408528198074,
      "grad_norm": 0.09955621510744095,
      "learning_rate": 0.0001,
      "loss": 1.6282,
      "step": 1850
    },
    {
      "epoch": 0.052269601100412656,
      "grad_norm": 0.10360520333051682,
      "learning_rate": 0.0001,
      "loss": 1.6265,
      "step": 1900
    },
    {
      "epoch": 0.05364511691884457,
      "grad_norm": 0.1229841411113739,
      "learning_rate": 0.0001,
      "loss": 1.6264,
      "step": 1950
    },
    {
      "epoch": 0.055020632737276476,
      "grad_norm": 0.09015832841396332,
      "learning_rate": 0.0001,
      "loss": 1.6248,
      "step": 2000
    },
    {
      "epoch": 0.05639614855570839,
      "grad_norm": 0.10285497456789017,
      "learning_rate": 0.0001,
      "loss": 1.6237,
      "step": 2050
    },
    {
      "epoch": 0.0577716643741403,
      "grad_norm": 0.07973627001047134,
      "learning_rate": 0.0001,
      "loss": 1.6262,
      "step": 2100
    },
    {
      "epoch": 0.059147180192572216,
      "grad_norm": 0.1072544977068901,
      "learning_rate": 0.0001,
      "loss": 1.6246,
      "step": 2150
    },
    {
      "epoch": 0.06052269601100413,
      "grad_norm": 0.11573298275470734,
      "learning_rate": 0.0001,
      "loss": 1.623,
      "step": 2200
    },
    {
      "epoch": 0.061898211829436035,
      "grad_norm": 0.1113864928483963,
      "learning_rate": 0.0001,
      "loss": 1.6189,
      "step": 2250
    },
    {
      "epoch": 0.06327372764786796,
      "grad_norm": 0.09252315014600754,
      "learning_rate": 0.0001,
      "loss": 1.6192,
      "step": 2300
    },
    {
      "epoch": 0.06464924346629987,
      "grad_norm": 0.09697891771793365,
      "learning_rate": 0.0001,
      "loss": 1.6191,
      "step": 2350
    },
    {
      "epoch": 0.06602475928473177,
      "grad_norm": 0.09384047985076904,
      "learning_rate": 0.0001,
      "loss": 1.6165,
      "step": 2400
    },
    {
      "epoch": 0.06740027510316368,
      "grad_norm": 0.10533461719751358,
      "learning_rate": 0.0001,
      "loss": 1.6202,
      "step": 2450
    },
    {
      "epoch": 0.0687757909215956,
      "grad_norm": 0.08703196048736572,
      "learning_rate": 0.0001,
      "loss": 1.618,
      "step": 2500
    },
    {
      "epoch": 0.07015130674002751,
      "grad_norm": 0.09502206742763519,
      "learning_rate": 0.0001,
      "loss": 1.6177,
      "step": 2550
    },
    {
      "epoch": 0.07152682255845942,
      "grad_norm": 0.09674184769392014,
      "learning_rate": 0.0001,
      "loss": 1.6143,
      "step": 2600
    },
    {
      "epoch": 0.07290233837689133,
      "grad_norm": 0.12614910304546356,
      "learning_rate": 0.0001,
      "loss": 1.6125,
      "step": 2650
    },
    {
      "epoch": 0.07427785419532325,
      "grad_norm": 0.10198106616735458,
      "learning_rate": 0.0001,
      "loss": 1.6158,
      "step": 2700
    },
    {
      "epoch": 0.07565337001375516,
      "grad_norm": 0.09061957895755768,
      "learning_rate": 0.0001,
      "loss": 1.6124,
      "step": 2750
    },
    {
      "epoch": 0.07702888583218707,
      "grad_norm": 0.08632820844650269,
      "learning_rate": 0.0001,
      "loss": 1.6113,
      "step": 2800
    },
    {
      "epoch": 0.07840440165061899,
      "grad_norm": 0.10429545491933823,
      "learning_rate": 0.0001,
      "loss": 1.6105,
      "step": 2850
    },
    {
      "epoch": 0.07977991746905089,
      "grad_norm": 0.104611337184906,
      "learning_rate": 0.0001,
      "loss": 1.6105,
      "step": 2900
    },
    {
      "epoch": 0.0811554332874828,
      "grad_norm": 0.11391541361808777,
      "learning_rate": 0.0001,
      "loss": 1.6078,
      "step": 2950
    },
    {
      "epoch": 0.08253094910591471,
      "grad_norm": 0.1170964241027832,
      "learning_rate": 0.0001,
      "loss": 1.6101,
      "step": 3000
    },
    {
      "epoch": 0.08390646492434663,
      "grad_norm": 0.10005070269107819,
      "learning_rate": 0.0001,
      "loss": 1.6096,
      "step": 3050
    },
    {
      "epoch": 0.08528198074277854,
      "grad_norm": 0.13063783943653107,
      "learning_rate": 0.0001,
      "loss": 1.6094,
      "step": 3100
    },
    {
      "epoch": 0.08665749656121045,
      "grad_norm": 0.10203906893730164,
      "learning_rate": 0.0001,
      "loss": 1.609,
      "step": 3150
    },
    {
      "epoch": 0.08803301237964237,
      "grad_norm": 0.11838550120592117,
      "learning_rate": 0.0001,
      "loss": 1.6068,
      "step": 3200
    },
    {
      "epoch": 0.08940852819807428,
      "grad_norm": 0.16624979674816132,
      "learning_rate": 0.0001,
      "loss": 1.6035,
      "step": 3250
    },
    {
      "epoch": 0.0907840440165062,
      "grad_norm": 0.11730783432722092,
      "learning_rate": 0.0001,
      "loss": 1.6074,
      "step": 3300
    },
    {
      "epoch": 0.0921595598349381,
      "grad_norm": 0.10523674637079239,
      "learning_rate": 0.0001,
      "loss": 1.6051,
      "step": 3350
    },
    {
      "epoch": 0.09353507565337002,
      "grad_norm": 0.10546988248825073,
      "learning_rate": 0.0001,
      "loss": 1.604,
      "step": 3400
    },
    {
      "epoch": 0.09491059147180192,
      "grad_norm": 0.13425269722938538,
      "learning_rate": 0.0001,
      "loss": 1.6044,
      "step": 3450
    },
    {
      "epoch": 0.09628610729023383,
      "grad_norm": 0.12492198497056961,
      "learning_rate": 0.0001,
      "loss": 1.6052,
      "step": 3500
    },
    {
      "epoch": 0.09766162310866575,
      "grad_norm": 0.09005106985569,
      "learning_rate": 0.0001,
      "loss": 1.603,
      "step": 3550
    },
    {
      "epoch": 0.09903713892709766,
      "grad_norm": 0.11914248019456863,
      "learning_rate": 0.0001,
      "loss": 1.6027,
      "step": 3600
    },
    {
      "epoch": 0.10041265474552957,
      "grad_norm": 0.12221172451972961,
      "learning_rate": 0.0001,
      "loss": 1.605,
      "step": 3650
    },
    {
      "epoch": 0.10178817056396149,
      "grad_norm": 0.13399210572242737,
      "learning_rate": 0.0001,
      "loss": 1.6039,
      "step": 3700
    },
    {
      "epoch": 0.1031636863823934,
      "grad_norm": 0.11565663665533066,
      "learning_rate": 0.0001,
      "loss": 1.6008,
      "step": 3750
    },
    {
      "epoch": 0.10453920220082531,
      "grad_norm": 0.12839622795581818,
      "learning_rate": 0.0001,
      "loss": 1.6004,
      "step": 3800
    },
    {
      "epoch": 0.10591471801925723,
      "grad_norm": 0.11184845864772797,
      "learning_rate": 0.0001,
      "loss": 1.5975,
      "step": 3850
    },
    {
      "epoch": 0.10729023383768914,
      "grad_norm": 0.11628763377666473,
      "learning_rate": 0.0001,
      "loss": 1.601,
      "step": 3900
    },
    {
      "epoch": 0.10866574965612105,
      "grad_norm": 0.11737735569477081,
      "learning_rate": 0.0001,
      "loss": 1.6011,
      "step": 3950
    },
    {
      "epoch": 0.11004126547455295,
      "grad_norm": 0.10090334713459015,
      "learning_rate": 0.0001,
      "loss": 1.5981,
      "step": 4000
    },
    {
      "epoch": 0.11141678129298486,
      "grad_norm": 0.11729908734560013,
      "learning_rate": 0.0001,
      "loss": 1.5972,
      "step": 4050
    },
    {
      "epoch": 0.11279229711141678,
      "grad_norm": 0.10134877264499664,
      "learning_rate": 0.0001,
      "loss": 1.5974,
      "step": 4100
    },
    {
      "epoch": 0.11416781292984869,
      "grad_norm": 0.150742307305336,
      "learning_rate": 0.0001,
      "loss": 1.5979,
      "step": 4150
    },
    {
      "epoch": 0.1155433287482806,
      "grad_norm": 0.1354828178882599,
      "learning_rate": 0.0001,
      "loss": 1.594,
      "step": 4200
    },
    {
      "epoch": 0.11691884456671252,
      "grad_norm": 0.10246012359857559,
      "learning_rate": 0.0001,
      "loss": 1.5944,
      "step": 4250
    },
    {
      "epoch": 0.11829436038514443,
      "grad_norm": 0.10707879811525345,
      "learning_rate": 0.0001,
      "loss": 1.5975,
      "step": 4300
    },
    {
      "epoch": 0.11966987620357634,
      "grad_norm": 0.09582670778036118,
      "learning_rate": 0.0001,
      "loss": 1.5931,
      "step": 4350
    },
    {
      "epoch": 0.12104539202200826,
      "grad_norm": 0.11471503973007202,
      "learning_rate": 0.0001,
      "loss": 1.5957,
      "step": 4400
    },
    {
      "epoch": 0.12242090784044017,
      "grad_norm": 0.14393934607505798,
      "learning_rate": 0.0001,
      "loss": 1.5947,
      "step": 4450
    },
    {
      "epoch": 0.12379642365887207,
      "grad_norm": 0.1267063319683075,
      "learning_rate": 0.0001,
      "loss": 1.5928,
      "step": 4500
    },
    {
      "epoch": 0.12517193947730398,
      "grad_norm": 0.10451563447713852,
      "learning_rate": 0.0001,
      "loss": 1.5944,
      "step": 4550
    },
    {
      "epoch": 0.1265474552957359,
      "grad_norm": 0.13244299590587616,
      "learning_rate": 0.0001,
      "loss": 1.5935,
      "step": 4600
    },
    {
      "epoch": 0.1279229711141678,
      "grad_norm": 0.14042487740516663,
      "learning_rate": 0.0001,
      "loss": 1.5929,
      "step": 4650
    },
    {
      "epoch": 0.12929848693259974,
      "grad_norm": 0.12199941277503967,
      "learning_rate": 0.0001,
      "loss": 1.5933,
      "step": 4700
    },
    {
      "epoch": 0.13067400275103164,
      "grad_norm": 0.13133960962295532,
      "learning_rate": 0.0001,
      "loss": 1.5904,
      "step": 4750
    },
    {
      "epoch": 0.13204951856946354,
      "grad_norm": 0.12281449884176254,
      "learning_rate": 0.0001,
      "loss": 1.5909,
      "step": 4800
    },
    {
      "epoch": 0.13342503438789546,
      "grad_norm": 0.1380591243505478,
      "learning_rate": 0.0001,
      "loss": 1.5899,
      "step": 4850
    },
    {
      "epoch": 0.13480055020632736,
      "grad_norm": 0.13320781290531158,
      "learning_rate": 0.0001,
      "loss": 1.5924,
      "step": 4900
    },
    {
      "epoch": 0.1361760660247593,
      "grad_norm": 0.10719151794910431,
      "learning_rate": 0.0001,
      "loss": 1.5909,
      "step": 4950
    },
    {
      "epoch": 0.1375515818431912,
      "grad_norm": 0.17885592579841614,
      "learning_rate": 0.0001,
      "loss": 1.591,
      "step": 5000
    },
    {
      "epoch": 0.13892709766162312,
      "grad_norm": 0.15455111861228943,
      "learning_rate": 0.0001,
      "loss": 1.587,
      "step": 5050
    },
    {
      "epoch": 0.14030261348005502,
      "grad_norm": 0.12887494266033173,
      "learning_rate": 0.0001,
      "loss": 1.5886,
      "step": 5100
    },
    {
      "epoch": 0.14167812929848694,
      "grad_norm": 0.13535436987876892,
      "learning_rate": 0.0001,
      "loss": 1.5901,
      "step": 5150
    },
    {
      "epoch": 0.14305364511691884,
      "grad_norm": 0.12412004172801971,
      "learning_rate": 0.0001,
      "loss": 1.5884,
      "step": 5200
    },
    {
      "epoch": 0.14442916093535077,
      "grad_norm": 0.1510736644268036,
      "learning_rate": 0.0001,
      "loss": 1.5879,
      "step": 5250
    },
    {
      "epoch": 0.14580467675378267,
      "grad_norm": 0.128033846616745,
      "learning_rate": 0.0001,
      "loss": 1.5868,
      "step": 5300
    },
    {
      "epoch": 0.14718019257221457,
      "grad_norm": 0.11286512017250061,
      "learning_rate": 0.0001,
      "loss": 1.5859,
      "step": 5350
    },
    {
      "epoch": 0.1485557083906465,
      "grad_norm": 0.11637207865715027,
      "learning_rate": 0.0001,
      "loss": 1.5837,
      "step": 5400
    },
    {
      "epoch": 0.1499312242090784,
      "grad_norm": 0.13789626955986023,
      "learning_rate": 0.0001,
      "loss": 1.5894,
      "step": 5450
    },
    {
      "epoch": 0.15130674002751032,
      "grad_norm": 0.12487693876028061,
      "learning_rate": 0.0001,
      "loss": 1.5851,
      "step": 5500
    },
    {
      "epoch": 0.15268225584594222,
      "grad_norm": 0.14437325298786163,
      "learning_rate": 0.0001,
      "loss": 1.5879,
      "step": 5550
    },
    {
      "epoch": 0.15405777166437415,
      "grad_norm": 0.10904733836650848,
      "learning_rate": 0.0001,
      "loss": 1.5838,
      "step": 5600
    },
    {
      "epoch": 0.15543328748280605,
      "grad_norm": 0.10461211949586868,
      "learning_rate": 0.0001,
      "loss": 1.5833,
      "step": 5650
    },
    {
      "epoch": 0.15680880330123798,
      "grad_norm": 0.1489093005657196,
      "learning_rate": 0.0001,
      "loss": 1.5823,
      "step": 5700
    },
    {
      "epoch": 0.15818431911966988,
      "grad_norm": 0.15630511939525604,
      "learning_rate": 0.0001,
      "loss": 1.5844,
      "step": 5750
    },
    {
      "epoch": 0.15955983493810177,
      "grad_norm": 0.15836940705776215,
      "learning_rate": 0.0001,
      "loss": 1.584,
      "step": 5800
    },
    {
      "epoch": 0.1609353507565337,
      "grad_norm": 0.12032505124807358,
      "learning_rate": 0.0001,
      "loss": 1.5848,
      "step": 5850
    },
    {
      "epoch": 0.1623108665749656,
      "grad_norm": 0.15543417632579803,
      "learning_rate": 0.0001,
      "loss": 1.5843,
      "step": 5900
    },
    {
      "epoch": 0.16368638239339753,
      "grad_norm": 0.11939691752195358,
      "learning_rate": 0.0001,
      "loss": 1.5818,
      "step": 5950
    },
    {
      "epoch": 0.16506189821182943,
      "grad_norm": 0.13943925499916077,
      "learning_rate": 0.0001,
      "loss": 1.5821,
      "step": 6000
    },
    {
      "epoch": 0.16643741403026135,
      "grad_norm": 0.1273224651813507,
      "learning_rate": 0.0001,
      "loss": 1.5807,
      "step": 6050
    },
    {
      "epoch": 0.16781292984869325,
      "grad_norm": 0.1731129139661789,
      "learning_rate": 0.0001,
      "loss": 1.5828,
      "step": 6100
    },
    {
      "epoch": 0.16918844566712518,
      "grad_norm": 0.11023139208555222,
      "learning_rate": 0.0001,
      "loss": 1.5806,
      "step": 6150
    },
    {
      "epoch": 0.17056396148555708,
      "grad_norm": 0.15180650353431702,
      "learning_rate": 0.0001,
      "loss": 1.5805,
      "step": 6200
    },
    {
      "epoch": 0.171939477303989,
      "grad_norm": 0.1235494539141655,
      "learning_rate": 0.0001,
      "loss": 1.5811,
      "step": 6250
    },
    {
      "epoch": 0.1733149931224209,
      "grad_norm": 0.12696652114391327,
      "learning_rate": 0.0001,
      "loss": 1.58,
      "step": 6300
    },
    {
      "epoch": 0.1746905089408528,
      "grad_norm": 0.1397417187690735,
      "learning_rate": 0.0001,
      "loss": 1.5806,
      "step": 6350
    },
    {
      "epoch": 0.17606602475928473,
      "grad_norm": 0.15651826560497284,
      "learning_rate": 0.0001,
      "loss": 1.5774,
      "step": 6400
    },
    {
      "epoch": 0.17744154057771663,
      "grad_norm": 0.10367725789546967,
      "learning_rate": 0.0001,
      "loss": 1.5793,
      "step": 6450
    },
    {
      "epoch": 0.17881705639614856,
      "grad_norm": 0.15408000349998474,
      "learning_rate": 0.0001,
      "loss": 1.5791,
      "step": 6500
    },
    {
      "epoch": 0.18019257221458046,
      "grad_norm": 0.10724977403879166,
      "learning_rate": 0.0001,
      "loss": 1.5799,
      "step": 6550
    },
    {
      "epoch": 0.1815680880330124,
      "grad_norm": 0.14652323722839355,
      "learning_rate": 0.0001,
      "loss": 1.5784,
      "step": 6600
    },
    {
      "epoch": 0.1829436038514443,
      "grad_norm": 0.11810048669576645,
      "learning_rate": 0.0001,
      "loss": 1.5783,
      "step": 6650
    },
    {
      "epoch": 0.1843191196698762,
      "grad_norm": 0.1892373412847519,
      "learning_rate": 0.0001,
      "loss": 1.5811,
      "step": 6700
    },
    {
      "epoch": 0.1856946354883081,
      "grad_norm": 0.1516016721725464,
      "learning_rate": 0.0001,
      "loss": 1.5781,
      "step": 6750
    },
    {
      "epoch": 0.18707015130674004,
      "grad_norm": 0.14342574775218964,
      "learning_rate": 0.0001,
      "loss": 1.5759,
      "step": 6800
    },
    {
      "epoch": 0.18844566712517194,
      "grad_norm": 0.1327650249004364,
      "learning_rate": 0.0001,
      "loss": 1.5779,
      "step": 6850
    },
    {
      "epoch": 0.18982118294360384,
      "grad_norm": 0.137595072388649,
      "learning_rate": 0.0001,
      "loss": 1.5761,
      "step": 6900
    },
    {
      "epoch": 0.19119669876203577,
      "grad_norm": 0.1387586146593094,
      "learning_rate": 0.0001,
      "loss": 1.5768,
      "step": 6950
    },
    {
      "epoch": 0.19257221458046767,
      "grad_norm": 0.1557263284921646,
      "learning_rate": 0.0001,
      "loss": 1.5775,
      "step": 7000
    },
    {
      "epoch": 0.1939477303988996,
      "grad_norm": 0.14735980331897736,
      "learning_rate": 0.0001,
      "loss": 1.5771,
      "step": 7050
    },
    {
      "epoch": 0.1953232462173315,
      "grad_norm": 0.18839861452579498,
      "learning_rate": 0.0001,
      "loss": 1.5748,
      "step": 7100
    },
    {
      "epoch": 0.19669876203576342,
      "grad_norm": 0.17223089933395386,
      "learning_rate": 0.0001,
      "loss": 1.5795,
      "step": 7150
    },
    {
      "epoch": 0.19807427785419532,
      "grad_norm": 0.11284028738737106,
      "learning_rate": 0.0001,
      "loss": 1.5745,
      "step": 7200
    },
    {
      "epoch": 0.19944979367262725,
      "grad_norm": 0.16285105049610138,
      "learning_rate": 0.0001,
      "loss": 1.5763,
      "step": 7250
    },
    {
      "epoch": 0.20082530949105915,
      "grad_norm": 0.15286004543304443,
      "learning_rate": 0.0001,
      "loss": 1.5734,
      "step": 7300
    },
    {
      "epoch": 0.20220082530949107,
      "grad_norm": 0.15827025473117828,
      "learning_rate": 0.0001,
      "loss": 1.5736,
      "step": 7350
    },
    {
      "epoch": 0.20357634112792297,
      "grad_norm": 0.13479341566562653,
      "learning_rate": 0.0001,
      "loss": 1.5755,
      "step": 7400
    },
    {
      "epoch": 0.20495185694635487,
      "grad_norm": 0.11652766913175583,
      "learning_rate": 0.0001,
      "loss": 1.5745,
      "step": 7450
    },
    {
      "epoch": 0.2063273727647868,
      "grad_norm": 0.1466943770647049,
      "learning_rate": 0.0001,
      "loss": 1.5748,
      "step": 7500
    },
    {
      "epoch": 0.2077028885832187,
      "grad_norm": 0.16038121283054352,
      "learning_rate": 0.0001,
      "loss": 1.572,
      "step": 7550
    },
    {
      "epoch": 0.20907840440165062,
      "grad_norm": 0.1869979202747345,
      "learning_rate": 0.0001,
      "loss": 1.5762,
      "step": 7600
    },
    {
      "epoch": 0.21045392022008252,
      "grad_norm": 0.14036841690540314,
      "learning_rate": 0.0001,
      "loss": 1.5754,
      "step": 7650
    },
    {
      "epoch": 0.21182943603851445,
      "grad_norm": 0.18491779267787933,
      "learning_rate": 0.0001,
      "loss": 1.5757,
      "step": 7700
    },
    {
      "epoch": 0.21320495185694635,
      "grad_norm": 0.13815288245677948,
      "learning_rate": 0.0001,
      "loss": 1.5754,
      "step": 7750
    },
    {
      "epoch": 0.21458046767537828,
      "grad_norm": 0.13334764540195465,
      "learning_rate": 0.0001,
      "loss": 1.5706,
      "step": 7800
    },
    {
      "epoch": 0.21595598349381018,
      "grad_norm": 0.15366512537002563,
      "learning_rate": 0.0001,
      "loss": 1.5731,
      "step": 7850
    },
    {
      "epoch": 0.2173314993122421,
      "grad_norm": 0.16366422176361084,
      "learning_rate": 0.0001,
      "loss": 1.5715,
      "step": 7900
    },
    {
      "epoch": 0.218707015130674,
      "grad_norm": 0.14637479186058044,
      "learning_rate": 0.0001,
      "loss": 1.5715,
      "step": 7950
    },
    {
      "epoch": 0.2200825309491059,
      "grad_norm": 0.1257038414478302,
      "learning_rate": 0.0001,
      "loss": 1.5712,
      "step": 8000
    },
    {
      "epoch": 0.22145804676753783,
      "grad_norm": 0.13014163076877594,
      "learning_rate": 0.0001,
      "loss": 1.5711,
      "step": 8050
    },
    {
      "epoch": 0.22283356258596973,
      "grad_norm": 0.13101409375667572,
      "learning_rate": 0.0001,
      "loss": 1.5734,
      "step": 8100
    },
    {
      "epoch": 0.22420907840440166,
      "grad_norm": 0.1509891152381897,
      "learning_rate": 0.0001,
      "loss": 1.5698,
      "step": 8150
    },
    {
      "epoch": 0.22558459422283356,
      "grad_norm": 0.16276001930236816,
      "learning_rate": 0.0001,
      "loss": 1.5714,
      "step": 8200
    },
    {
      "epoch": 0.22696011004126548,
      "grad_norm": 0.16040217876434326,
      "learning_rate": 0.0001,
      "loss": 1.5701,
      "step": 8250
    },
    {
      "epoch": 0.22833562585969738,
      "grad_norm": 0.160230815410614,
      "learning_rate": 0.0001,
      "loss": 1.5705,
      "step": 8300
    },
    {
      "epoch": 0.2297111416781293,
      "grad_norm": 0.18454241752624512,
      "learning_rate": 0.0001,
      "loss": 1.571,
      "step": 8350
    },
    {
      "epoch": 0.2310866574965612,
      "grad_norm": 0.17411856353282928,
      "learning_rate": 0.0001,
      "loss": 1.5679,
      "step": 8400
    },
    {
      "epoch": 0.2324621733149931,
      "grad_norm": 0.16710075736045837,
      "learning_rate": 0.0001,
      "loss": 1.5674,
      "step": 8450
    },
    {
      "epoch": 0.23383768913342504,
      "grad_norm": 0.12378160655498505,
      "learning_rate": 0.0001,
      "loss": 1.5671,
      "step": 8500
    },
    {
      "epoch": 0.23521320495185694,
      "grad_norm": 0.11550536751747131,
      "learning_rate": 0.0001,
      "loss": 1.5698,
      "step": 8550
    },
    {
      "epoch": 0.23658872077028886,
      "grad_norm": 0.17768432199954987,
      "learning_rate": 0.0001,
      "loss": 1.5699,
      "step": 8600
    },
    {
      "epoch": 0.23796423658872076,
      "grad_norm": 0.15126097202301025,
      "learning_rate": 0.0001,
      "loss": 1.5694,
      "step": 8650
    },
    {
      "epoch": 0.2393397524071527,
      "grad_norm": 0.1827315390110016,
      "learning_rate": 0.0001,
      "loss": 1.5671,
      "step": 8700
    },
    {
      "epoch": 0.2407152682255846,
      "grad_norm": 0.11432069540023804,
      "learning_rate": 0.0001,
      "loss": 1.5685,
      "step": 8750
    },
    {
      "epoch": 0.24209078404401652,
      "grad_norm": 0.14279188215732574,
      "learning_rate": 0.0001,
      "loss": 1.5677,
      "step": 8800
    },
    {
      "epoch": 0.24346629986244842,
      "grad_norm": 0.13771188259124756,
      "learning_rate": 0.0001,
      "loss": 1.5667,
      "step": 8850
    },
    {
      "epoch": 0.24484181568088034,
      "grad_norm": 0.12438327074050903,
      "learning_rate": 0.0001,
      "loss": 1.5649,
      "step": 8900
    },
    {
      "epoch": 0.24621733149931224,
      "grad_norm": 0.146587535738945,
      "learning_rate": 0.0001,
      "loss": 1.5689,
      "step": 8950
    },
    {
      "epoch": 0.24759284731774414,
      "grad_norm": 0.13684628903865814,
      "learning_rate": 0.0001,
      "loss": 1.5662,
      "step": 9000
    },
    {
      "epoch": 0.24896836313617607,
      "grad_norm": 0.1465720385313034,
      "learning_rate": 0.0001,
      "loss": 1.5666,
      "step": 9050
    },
    {
      "epoch": 0.25034387895460797,
      "grad_norm": 0.1553189605474472,
      "learning_rate": 0.0001,
      "loss": 1.5647,
      "step": 9100
    },
    {
      "epoch": 0.2517193947730399,
      "grad_norm": 0.12973164021968842,
      "learning_rate": 0.0001,
      "loss": 1.5647,
      "step": 9150
    },
    {
      "epoch": 0.2530949105914718,
      "grad_norm": 0.17071610689163208,
      "learning_rate": 0.0001,
      "loss": 1.5691,
      "step": 9200
    },
    {
      "epoch": 0.2544704264099037,
      "grad_norm": 0.1424863487482071,
      "learning_rate": 0.0001,
      "loss": 1.5654,
      "step": 9250
    },
    {
      "epoch": 0.2558459422283356,
      "grad_norm": 0.13117440044879913,
      "learning_rate": 0.0001,
      "loss": 1.5668,
      "step": 9300
    },
    {
      "epoch": 0.25722145804676755,
      "grad_norm": 0.14353643357753754,
      "learning_rate": 0.0001,
      "loss": 1.567,
      "step": 9350
    },
    {
      "epoch": 0.2585969738651995,
      "grad_norm": 0.18137438595294952,
      "learning_rate": 0.0001,
      "loss": 1.5648,
      "step": 9400
    },
    {
      "epoch": 0.25997248968363135,
      "grad_norm": 0.1453561782836914,
      "learning_rate": 0.0001,
      "loss": 1.5631,
      "step": 9450
    },
    {
      "epoch": 0.2613480055020633,
      "grad_norm": 0.13514567911624908,
      "learning_rate": 0.0001,
      "loss": 1.5633,
      "step": 9500
    },
    {
      "epoch": 0.2627235213204952,
      "grad_norm": 0.20019495487213135,
      "learning_rate": 0.0001,
      "loss": 1.5655,
      "step": 9550
    },
    {
      "epoch": 0.2640990371389271,
      "grad_norm": 0.18167296051979065,
      "learning_rate": 0.0001,
      "loss": 1.5634,
      "step": 9600
    },
    {
      "epoch": 0.265474552957359,
      "grad_norm": 0.1335984319448471,
      "learning_rate": 0.0001,
      "loss": 1.5609,
      "step": 9650
    },
    {
      "epoch": 0.2668500687757909,
      "grad_norm": 0.12064065039157867,
      "learning_rate": 0.0001,
      "loss": 1.5619,
      "step": 9700
    },
    {
      "epoch": 0.26822558459422285,
      "grad_norm": 0.16066288948059082,
      "learning_rate": 0.0001,
      "loss": 1.5639,
      "step": 9750
    },
    {
      "epoch": 0.2696011004126547,
      "grad_norm": 0.18084204196929932,
      "learning_rate": 0.0001,
      "loss": 1.5597,
      "step": 9800
    },
    {
      "epoch": 0.27097661623108665,
      "grad_norm": 0.14845338463783264,
      "learning_rate": 0.0001,
      "loss": 1.5626,
      "step": 9850
    },
    {
      "epoch": 0.2723521320495186,
      "grad_norm": 0.13293515145778656,
      "learning_rate": 0.0001,
      "loss": 1.5648,
      "step": 9900
    },
    {
      "epoch": 0.2737276478679505,
      "grad_norm": 0.14939668774604797,
      "learning_rate": 0.0001,
      "loss": 1.5612,
      "step": 9950
    },
    {
      "epoch": 0.2751031636863824,
      "grad_norm": 0.1553388386964798,
      "learning_rate": 0.0001,
      "loss": 1.5629,
      "step": 10000
    },
    {
      "epoch": 0.2764786795048143,
      "grad_norm": 0.22416375577449799,
      "learning_rate": 0.0001,
      "loss": 1.5621,
      "step": 10050
    },
    {
      "epoch": 0.27785419532324623,
      "grad_norm": 0.2197302132844925,
      "learning_rate": 0.0001,
      "loss": 1.5635,
      "step": 10100
    },
    {
      "epoch": 0.2792297111416781,
      "grad_norm": 0.17688524723052979,
      "learning_rate": 0.0001,
      "loss": 1.5616,
      "step": 10150
    },
    {
      "epoch": 0.28060522696011003,
      "grad_norm": 0.1495491862297058,
      "learning_rate": 0.0001,
      "loss": 1.5614,
      "step": 10200
    },
    {
      "epoch": 0.28198074277854196,
      "grad_norm": 0.15716291964054108,
      "learning_rate": 0.0001,
      "loss": 1.5592,
      "step": 10250
    },
    {
      "epoch": 0.2833562585969739,
      "grad_norm": 0.14116239547729492,
      "learning_rate": 0.0001,
      "loss": 1.5586,
      "step": 10300
    },
    {
      "epoch": 0.28473177441540576,
      "grad_norm": 0.11010037362575531,
      "learning_rate": 0.0001,
      "loss": 1.5603,
      "step": 10350
    },
    {
      "epoch": 0.2861072902338377,
      "grad_norm": 0.1838681697845459,
      "learning_rate": 0.0001,
      "loss": 1.561,
      "step": 10400
    },
    {
      "epoch": 0.2874828060522696,
      "grad_norm": 0.19001850485801697,
      "learning_rate": 0.0001,
      "loss": 1.5588,
      "step": 10450
    },
    {
      "epoch": 0.28885832187070154,
      "grad_norm": 0.20800583064556122,
      "learning_rate": 0.0001,
      "loss": 1.5607,
      "step": 10500
    },
    {
      "epoch": 0.2902338376891334,
      "grad_norm": 0.17948520183563232,
      "learning_rate": 0.0001,
      "loss": 1.56,
      "step": 10550
    },
    {
      "epoch": 0.29160935350756534,
      "grad_norm": 0.16178689897060394,
      "learning_rate": 0.0001,
      "loss": 1.5603,
      "step": 10600
    },
    {
      "epoch": 0.29298486932599727,
      "grad_norm": 0.1580880880355835,
      "learning_rate": 0.0001,
      "loss": 1.5606,
      "step": 10650
    },
    {
      "epoch": 0.29436038514442914,
      "grad_norm": 0.14434567093849182,
      "learning_rate": 0.0001,
      "loss": 1.5618,
      "step": 10700
    },
    {
      "epoch": 0.29573590096286106,
      "grad_norm": 0.17610964179039001,
      "learning_rate": 0.0001,
      "loss": 1.5613,
      "step": 10750
    },
    {
      "epoch": 0.297111416781293,
      "grad_norm": 0.15156705677509308,
      "learning_rate": 0.0001,
      "loss": 1.5563,
      "step": 10800
    },
    {
      "epoch": 0.2984869325997249,
      "grad_norm": 0.1466618925333023,
      "learning_rate": 0.0001,
      "loss": 1.5616,
      "step": 10850
    },
    {
      "epoch": 0.2998624484181568,
      "grad_norm": 0.1162666529417038,
      "learning_rate": 0.0001,
      "loss": 1.559,
      "step": 10900
    },
    {
      "epoch": 0.3012379642365887,
      "grad_norm": 0.15534426271915436,
      "learning_rate": 0.0001,
      "loss": 1.5594,
      "step": 10950
    },
    {
      "epoch": 0.30261348005502064,
      "grad_norm": 0.15940657258033752,
      "learning_rate": 0.0001,
      "loss": 1.5613,
      "step": 11000
    },
    {
      "epoch": 0.30398899587345257,
      "grad_norm": 0.1757323294878006,
      "learning_rate": 0.0001,
      "loss": 1.5588,
      "step": 11050
    },
    {
      "epoch": 0.30536451169188444,
      "grad_norm": 0.11815246194601059,
      "learning_rate": 0.0001,
      "loss": 1.5589,
      "step": 11100
    },
    {
      "epoch": 0.30674002751031637,
      "grad_norm": 0.2773960828781128,
      "learning_rate": 0.0001,
      "loss": 1.5584,
      "step": 11150
    },
    {
      "epoch": 0.3081155433287483,
      "grad_norm": 0.12601600587368011,
      "learning_rate": 0.0001,
      "loss": 1.5572,
      "step": 11200
    },
    {
      "epoch": 0.30949105914718017,
      "grad_norm": 0.1593768298625946,
      "learning_rate": 0.0001,
      "loss": 1.5575,
      "step": 11250
    },
    {
      "epoch": 0.3108665749656121,
      "grad_norm": 0.149438738822937,
      "learning_rate": 0.0001,
      "loss": 1.557,
      "step": 11300
    },
    {
      "epoch": 0.312242090784044,
      "grad_norm": 0.11111125349998474,
      "learning_rate": 0.0001,
      "loss": 1.5587,
      "step": 11350
    },
    {
      "epoch": 0.31361760660247595,
      "grad_norm": 0.1610383540391922,
      "learning_rate": 0.0001,
      "loss": 1.5572,
      "step": 11400
    },
    {
      "epoch": 0.3149931224209078,
      "grad_norm": 0.17420324683189392,
      "learning_rate": 0.0001,
      "loss": 1.5581,
      "step": 11450
    },
    {
      "epoch": 0.31636863823933975,
      "grad_norm": 0.16623131930828094,
      "learning_rate": 0.0001,
      "loss": 1.5561,
      "step": 11500
    },
    {
      "epoch": 0.3177441540577717,
      "grad_norm": 0.15828974545001984,
      "learning_rate": 0.0001,
      "loss": 1.5544,
      "step": 11550
    },
    {
      "epoch": 0.31911966987620355,
      "grad_norm": 0.15183350443840027,
      "learning_rate": 0.0001,
      "loss": 1.5555,
      "step": 11600
    },
    {
      "epoch": 0.3204951856946355,
      "grad_norm": 0.16378933191299438,
      "learning_rate": 0.0001,
      "loss": 1.5532,
      "step": 11650
    },
    {
      "epoch": 0.3218707015130674,
      "grad_norm": 0.15861773490905762,
      "learning_rate": 0.0001,
      "loss": 1.5568,
      "step": 11700
    },
    {
      "epoch": 0.32324621733149933,
      "grad_norm": 0.13385528326034546,
      "learning_rate": 0.0001,
      "loss": 1.5568,
      "step": 11750
    },
    {
      "epoch": 0.3246217331499312,
      "grad_norm": 0.16392391920089722,
      "learning_rate": 0.0001,
      "loss": 1.5548,
      "step": 11800
    },
    {
      "epoch": 0.32599724896836313,
      "grad_norm": 0.14662721753120422,
      "learning_rate": 0.0001,
      "loss": 1.5539,
      "step": 11850
    },
    {
      "epoch": 0.32737276478679506,
      "grad_norm": 0.13727930188179016,
      "learning_rate": 0.0001,
      "loss": 1.5552,
      "step": 11900
    },
    {
      "epoch": 0.328748280605227,
      "grad_norm": 0.15576840937137604,
      "learning_rate": 0.0001,
      "loss": 1.5552,
      "step": 11950
    },
    {
      "epoch": 0.33012379642365886,
      "grad_norm": 0.1717185378074646,
      "learning_rate": 0.0001,
      "loss": 1.5538,
      "step": 12000
    },
    {
      "epoch": 0.3314993122420908,
      "grad_norm": 0.16970685124397278,
      "learning_rate": 0.0001,
      "loss": 1.5556,
      "step": 12050
    },
    {
      "epoch": 0.3328748280605227,
      "grad_norm": 0.1489485800266266,
      "learning_rate": 0.0001,
      "loss": 1.5527,
      "step": 12100
    },
    {
      "epoch": 0.3342503438789546,
      "grad_norm": 0.1374077945947647,
      "learning_rate": 0.0001,
      "loss": 1.5528,
      "step": 12150
    },
    {
      "epoch": 0.3356258596973865,
      "grad_norm": 0.19402620196342468,
      "learning_rate": 0.0001,
      "loss": 1.5547,
      "step": 12200
    },
    {
      "epoch": 0.33700137551581844,
      "grad_norm": 0.1642199009656906,
      "learning_rate": 0.0001,
      "loss": 1.5538,
      "step": 12250
    },
    {
      "epoch": 0.33837689133425036,
      "grad_norm": 0.13107603788375854,
      "learning_rate": 0.0001,
      "loss": 1.5547,
      "step": 12300
    },
    {
      "epoch": 0.33975240715268223,
      "grad_norm": 0.1858353465795517,
      "learning_rate": 0.0001,
      "loss": 1.5526,
      "step": 12350
    },
    {
      "epoch": 0.34112792297111416,
      "grad_norm": 0.1422649323940277,
      "learning_rate": 0.0001,
      "loss": 1.5523,
      "step": 12400
    },
    {
      "epoch": 0.3425034387895461,
      "grad_norm": 0.16968269646167755,
      "learning_rate": 0.0001,
      "loss": 1.554,
      "step": 12450
    },
    {
      "epoch": 0.343878954607978,
      "grad_norm": 0.1434723138809204,
      "learning_rate": 0.0001,
      "loss": 1.5544,
      "step": 12500
    },
    {
      "epoch": 0.3452544704264099,
      "grad_norm": 0.18616297841072083,
      "learning_rate": 0.0001,
      "loss": 1.5506,
      "step": 12550
    },
    {
      "epoch": 0.3466299862448418,
      "grad_norm": 0.16946491599082947,
      "learning_rate": 0.0001,
      "loss": 1.5524,
      "step": 12600
    },
    {
      "epoch": 0.34800550206327374,
      "grad_norm": 0.17658023536205292,
      "learning_rate": 0.0001,
      "loss": 1.5536,
      "step": 12650
    },
    {
      "epoch": 0.3493810178817056,
      "grad_norm": 0.15203554928302765,
      "learning_rate": 0.0001,
      "loss": 1.5507,
      "step": 12700
    },
    {
      "epoch": 0.35075653370013754,
      "grad_norm": 0.13097505271434784,
      "learning_rate": 0.0001,
      "loss": 1.5542,
      "step": 12750
    },
    {
      "epoch": 0.35213204951856947,
      "grad_norm": 0.14317452907562256,
      "learning_rate": 0.0001,
      "loss": 1.5534,
      "step": 12800
    },
    {
      "epoch": 0.3535075653370014,
      "grad_norm": 0.12445474416017532,
      "learning_rate": 0.0001,
      "loss": 1.5535,
      "step": 12850
    },
    {
      "epoch": 0.35488308115543327,
      "grad_norm": 0.1327485293149948,
      "learning_rate": 0.0001,
      "loss": 1.5521,
      "step": 12900
    },
    {
      "epoch": 0.3562585969738652,
      "grad_norm": 0.15487389266490936,
      "learning_rate": 0.0001,
      "loss": 1.553,
      "step": 12950
    },
    {
      "epoch": 0.3576341127922971,
      "grad_norm": 0.23483023047447205,
      "learning_rate": 0.0001,
      "loss": 1.5502,
      "step": 13000
    },
    {
      "epoch": 0.35900962861072905,
      "grad_norm": 0.14994105696678162,
      "learning_rate": 0.0001,
      "loss": 1.5518,
      "step": 13050
    },
    {
      "epoch": 0.3603851444291609,
      "grad_norm": 0.12222074717283249,
      "learning_rate": 0.0001,
      "loss": 1.5508,
      "step": 13100
    },
    {
      "epoch": 0.36176066024759285,
      "grad_norm": 0.1246858537197113,
      "learning_rate": 0.0001,
      "loss": 1.552,
      "step": 13150
    },
    {
      "epoch": 0.3631361760660248,
      "grad_norm": 0.15825419127941132,
      "learning_rate": 0.0001,
      "loss": 1.5499,
      "step": 13200
    },
    {
      "epoch": 0.36451169188445665,
      "grad_norm": 0.17960667610168457,
      "learning_rate": 0.0001,
      "loss": 1.5551,
      "step": 13250
    },
    {
      "epoch": 0.3658872077028886,
      "grad_norm": 0.1628105491399765,
      "learning_rate": 0.0001,
      "loss": 1.5544,
      "step": 13300
    },
    {
      "epoch": 0.3672627235213205,
      "grad_norm": 0.15981099009513855,
      "learning_rate": 0.0001,
      "loss": 1.5527,
      "step": 13350
    },
    {
      "epoch": 0.3686382393397524,
      "grad_norm": 0.11882206797599792,
      "learning_rate": 0.0001,
      "loss": 1.5505,
      "step": 13400
    },
    {
      "epoch": 0.3700137551581843,
      "grad_norm": 0.1369376927614212,
      "learning_rate": 0.0001,
      "loss": 1.5487,
      "step": 13450
    },
    {
      "epoch": 0.3713892709766162,
      "grad_norm": 0.1341916173696518,
      "learning_rate": 0.0001,
      "loss": 1.5489,
      "step": 13500
    },
    {
      "epoch": 0.37276478679504815,
      "grad_norm": 0.1692420095205307,
      "learning_rate": 0.0001,
      "loss": 1.5486,
      "step": 13550
    },
    {
      "epoch": 0.3741403026134801,
      "grad_norm": 0.12764231860637665,
      "learning_rate": 0.0001,
      "loss": 1.5479,
      "step": 13600
    },
    {
      "epoch": 0.37551581843191195,
      "grad_norm": 0.1610202044248581,
      "learning_rate": 0.0001,
      "loss": 1.5493,
      "step": 13650
    },
    {
      "epoch": 0.3768913342503439,
      "grad_norm": 0.20008735358715057,
      "learning_rate": 0.0001,
      "loss": 1.5504,
      "step": 13700
    },
    {
      "epoch": 0.3782668500687758,
      "grad_norm": 0.14668354392051697,
      "learning_rate": 0.0001,
      "loss": 1.5459,
      "step": 13750
    },
    {
      "epoch": 0.3796423658872077,
      "grad_norm": 0.16147159039974213,
      "learning_rate": 0.0001,
      "loss": 1.5497,
      "step": 13800
    },
    {
      "epoch": 0.3810178817056396,
      "grad_norm": 0.2127738893032074,
      "learning_rate": 0.0001,
      "loss": 1.5496,
      "step": 13850
    },
    {
      "epoch": 0.38239339752407153,
      "grad_norm": 0.14936117827892303,
      "learning_rate": 0.0001,
      "loss": 1.5487,
      "step": 13900
    },
    {
      "epoch": 0.38376891334250346,
      "grad_norm": 0.1460547298192978,
      "learning_rate": 0.0001,
      "loss": 1.5513,
      "step": 13950
    },
    {
      "epoch": 0.38514442916093533,
      "grad_norm": 0.1418396234512329,
      "learning_rate": 0.0001,
      "loss": 1.5489,
      "step": 14000
    },
    {
      "epoch": 0.38651994497936726,
      "grad_norm": 0.12608648836612701,
      "learning_rate": 0.0001,
      "loss": 1.5478,
      "step": 14050
    },
    {
      "epoch": 0.3878954607977992,
      "grad_norm": 0.12352428585290909,
      "learning_rate": 0.0001,
      "loss": 1.5472,
      "step": 14100
    },
    {
      "epoch": 0.3892709766162311,
      "grad_norm": 0.140400692820549,
      "learning_rate": 0.0001,
      "loss": 1.5471,
      "step": 14150
    },
    {
      "epoch": 0.390646492434663,
      "grad_norm": 0.14015322923660278,
      "learning_rate": 0.0001,
      "loss": 1.5495,
      "step": 14200
    },
    {
      "epoch": 0.3920220082530949,
      "grad_norm": 0.13664819300174713,
      "learning_rate": 0.0001,
      "loss": 1.5515,
      "step": 14250
    },
    {
      "epoch": 0.39339752407152684,
      "grad_norm": 0.19558057188987732,
      "learning_rate": 0.0001,
      "loss": 1.5493,
      "step": 14300
    },
    {
      "epoch": 0.3947730398899587,
      "grad_norm": 0.14744845032691956,
      "learning_rate": 0.0001,
      "loss": 1.547,
      "step": 14350
    },
    {
      "epoch": 0.39614855570839064,
      "grad_norm": 0.13610410690307617,
      "learning_rate": 0.0001,
      "loss": 1.5499,
      "step": 14400
    },
    {
      "epoch": 0.39752407152682256,
      "grad_norm": 0.16850556433200836,
      "learning_rate": 0.0001,
      "loss": 1.5475,
      "step": 14450
    },
    {
      "epoch": 0.3988995873452545,
      "grad_norm": 0.11494544893503189,
      "learning_rate": 0.0001,
      "loss": 1.5441,
      "step": 14500
    },
    {
      "epoch": 0.40027510316368636,
      "grad_norm": 0.1311003863811493,
      "learning_rate": 0.0001,
      "loss": 1.5451,
      "step": 14550
    },
    {
      "epoch": 0.4016506189821183,
      "grad_norm": 0.16432379186153412,
      "learning_rate": 0.0001,
      "loss": 1.5483,
      "step": 14600
    },
    {
      "epoch": 0.4030261348005502,
      "grad_norm": 0.16200096905231476,
      "learning_rate": 0.0001,
      "loss": 1.5458,
      "step": 14650
    },
    {
      "epoch": 0.40440165061898214,
      "grad_norm": 0.15324008464813232,
      "learning_rate": 0.0001,
      "loss": 1.5486,
      "step": 14700
    },
    {
      "epoch": 0.405777166437414,
      "grad_norm": 0.2114071398973465,
      "learning_rate": 0.0001,
      "loss": 1.5463,
      "step": 14750
    },
    {
      "epoch": 0.40715268225584594,
      "grad_norm": 0.1691250056028366,
      "learning_rate": 0.0001,
      "loss": 1.5449,
      "step": 14800
    },
    {
      "epoch": 0.40852819807427787,
      "grad_norm": 0.15044333040714264,
      "learning_rate": 0.0001,
      "loss": 1.5454,
      "step": 14850
    },
    {
      "epoch": 0.40990371389270974,
      "grad_norm": 0.14457371830940247,
      "learning_rate": 0.0001,
      "loss": 1.5475,
      "step": 14900
    },
    {
      "epoch": 0.41127922971114167,
      "grad_norm": 0.15145525336265564,
      "learning_rate": 0.0001,
      "loss": 1.5474,
      "step": 14950
    },
    {
      "epoch": 0.4126547455295736,
      "grad_norm": 0.1273120492696762,
      "learning_rate": 0.0001,
      "loss": 1.5446,
      "step": 15000
    },
    {
      "epoch": 0.4140302613480055,
      "grad_norm": 0.1621488630771637,
      "learning_rate": 0.0001,
      "loss": 1.5464,
      "step": 15050
    },
    {
      "epoch": 0.4154057771664374,
      "grad_norm": 0.1621532440185547,
      "learning_rate": 0.0001,
      "loss": 1.5472,
      "step": 15100
    },
    {
      "epoch": 0.4167812929848693,
      "grad_norm": 0.13030585646629333,
      "learning_rate": 0.0001,
      "loss": 1.5416,
      "step": 15150
    },
    {
      "epoch": 0.41815680880330125,
      "grad_norm": 0.18759876489639282,
      "learning_rate": 0.0001,
      "loss": 1.5448,
      "step": 15200
    },
    {
      "epoch": 0.4195323246217332,
      "grad_norm": 0.12614044547080994,
      "learning_rate": 0.0001,
      "loss": 1.5459,
      "step": 15250
    },
    {
      "epoch": 0.42090784044016505,
      "grad_norm": 0.11533529311418533,
      "learning_rate": 0.0001,
      "loss": 1.5446,
      "step": 15300
    },
    {
      "epoch": 0.422283356258597,
      "grad_norm": 0.1886916160583496,
      "learning_rate": 0.0001,
      "loss": 1.5466,
      "step": 15350
    },
    {
      "epoch": 0.4236588720770289,
      "grad_norm": 0.2204965353012085,
      "learning_rate": 0.0001,
      "loss": 1.5436,
      "step": 15400
    },
    {
      "epoch": 0.4250343878954608,
      "grad_norm": 0.12042222172021866,
      "learning_rate": 0.0001,
      "loss": 1.5425,
      "step": 15450
    },
    {
      "epoch": 0.4264099037138927,
      "grad_norm": 0.135628342628479,
      "learning_rate": 0.0001,
      "loss": 1.5464,
      "step": 15500
    },
    {
      "epoch": 0.42778541953232463,
      "grad_norm": 0.15042053163051605,
      "learning_rate": 0.0001,
      "loss": 1.5441,
      "step": 15550
    },
    {
      "epoch": 0.42916093535075656,
      "grad_norm": 0.1294483244419098,
      "learning_rate": 0.0001,
      "loss": 1.5468,
      "step": 15600
    },
    {
      "epoch": 0.4305364511691884,
      "grad_norm": 0.153069868683815,
      "learning_rate": 0.0001,
      "loss": 1.5416,
      "step": 15650
    },
    {
      "epoch": 0.43191196698762035,
      "grad_norm": 0.129000723361969,
      "learning_rate": 0.0001,
      "loss": 1.5434,
      "step": 15700
    },
    {
      "epoch": 0.4332874828060523,
      "grad_norm": 0.1890910267829895,
      "learning_rate": 0.0001,
      "loss": 1.5426,
      "step": 15750
    },
    {
      "epoch": 0.4346629986244842,
      "grad_norm": 0.14907212555408478,
      "learning_rate": 0.0001,
      "loss": 1.5447,
      "step": 15800
    },
    {
      "epoch": 0.4360385144429161,
      "grad_norm": 0.1549520045518875,
      "learning_rate": 0.0001,
      "loss": 1.5438,
      "step": 15850
    },
    {
      "epoch": 0.437414030261348,
      "grad_norm": 0.1726304590702057,
      "learning_rate": 0.0001,
      "loss": 1.5431,
      "step": 15900
    },
    {
      "epoch": 0.43878954607977994,
      "grad_norm": 0.14929509162902832,
      "learning_rate": 0.0001,
      "loss": 1.5408,
      "step": 15950
    },
    {
      "epoch": 0.4401650618982118,
      "grad_norm": 0.1404862105846405,
      "learning_rate": 0.0001,
      "loss": 1.5431,
      "step": 16000
    },
    {
      "epoch": 0.44154057771664373,
      "grad_norm": 0.1365077942609787,
      "learning_rate": 0.0001,
      "loss": 1.5434,
      "step": 16050
    },
    {
      "epoch": 0.44291609353507566,
      "grad_norm": 0.16866528987884521,
      "learning_rate": 0.0001,
      "loss": 1.5425,
      "step": 16100
    },
    {
      "epoch": 0.4442916093535076,
      "grad_norm": 0.13150258362293243,
      "learning_rate": 0.0001,
      "loss": 1.5418,
      "step": 16150
    },
    {
      "epoch": 0.44566712517193946,
      "grad_norm": 0.17333872616291046,
      "learning_rate": 0.0001,
      "loss": 1.5415,
      "step": 16200
    },
    {
      "epoch": 0.4470426409903714,
      "grad_norm": 0.2110324501991272,
      "learning_rate": 0.0001,
      "loss": 1.5434,
      "step": 16250
    },
    {
      "epoch": 0.4484181568088033,
      "grad_norm": 0.19441699981689453,
      "learning_rate": 0.0001,
      "loss": 1.5408,
      "step": 16300
    },
    {
      "epoch": 0.4497936726272352,
      "grad_norm": 0.1581384241580963,
      "learning_rate": 0.0001,
      "loss": 1.5428,
      "step": 16350
    },
    {
      "epoch": 0.4511691884456671,
      "grad_norm": 0.14479832351207733,
      "learning_rate": 0.0001,
      "loss": 1.5444,
      "step": 16400
    },
    {
      "epoch": 0.45254470426409904,
      "grad_norm": 0.16739803552627563,
      "learning_rate": 0.0001,
      "loss": 1.541,
      "step": 16450
    },
    {
      "epoch": 0.45392022008253097,
      "grad_norm": 0.14801441133022308,
      "learning_rate": 0.0001,
      "loss": 1.54,
      "step": 16500
    },
    {
      "epoch": 0.45529573590096284,
      "grad_norm": 0.13265211880207062,
      "learning_rate": 0.0001,
      "loss": 1.5417,
      "step": 16550
    },
    {
      "epoch": 0.45667125171939477,
      "grad_norm": 0.1164972111582756,
      "learning_rate": 0.0001,
      "loss": 1.5411,
      "step": 16600
    },
    {
      "epoch": 0.4580467675378267,
      "grad_norm": 0.1256764531135559,
      "learning_rate": 0.0001,
      "loss": 1.538,
      "step": 16650
    },
    {
      "epoch": 0.4594222833562586,
      "grad_norm": 0.13301979005336761,
      "learning_rate": 0.0001,
      "loss": 1.5409,
      "step": 16700
    },
    {
      "epoch": 0.4607977991746905,
      "grad_norm": 0.1520063877105713,
      "learning_rate": 0.0001,
      "loss": 1.5406,
      "step": 16750
    },
    {
      "epoch": 0.4621733149931224,
      "grad_norm": 0.12742547690868378,
      "learning_rate": 0.0001,
      "loss": 1.5405,
      "step": 16800
    },
    {
      "epoch": 0.46354883081155435,
      "grad_norm": 0.17311689257621765,
      "learning_rate": 0.0001,
      "loss": 1.5416,
      "step": 16850
    },
    {
      "epoch": 0.4649243466299862,
      "grad_norm": 0.14269371330738068,
      "learning_rate": 0.0001,
      "loss": 1.5413,
      "step": 16900
    },
    {
      "epoch": 0.46629986244841815,
      "grad_norm": 0.14457383751869202,
      "learning_rate": 0.0001,
      "loss": 1.5415,
      "step": 16950
    },
    {
      "epoch": 0.4676753782668501,
      "grad_norm": 0.13189777731895447,
      "learning_rate": 0.0001,
      "loss": 1.5388,
      "step": 17000
    },
    {
      "epoch": 0.469050894085282,
      "grad_norm": 0.16488979756832123,
      "learning_rate": 0.0001,
      "loss": 1.5398,
      "step": 17050
    },
    {
      "epoch": 0.47042640990371387,
      "grad_norm": 0.15953794121742249,
      "learning_rate": 0.0001,
      "loss": 1.5387,
      "step": 17100
    },
    {
      "epoch": 0.4718019257221458,
      "grad_norm": 0.11922045797109604,
      "learning_rate": 0.0001,
      "loss": 1.5389,
      "step": 17150
    },
    {
      "epoch": 0.4731774415405777,
      "grad_norm": 0.13724352419376373,
      "learning_rate": 0.0001,
      "loss": 1.5399,
      "step": 17200
    },
    {
      "epoch": 0.47455295735900965,
      "grad_norm": 0.14968377351760864,
      "learning_rate": 0.0001,
      "loss": 1.5419,
      "step": 17250
    },
    {
      "epoch": 0.4759284731774415,
      "grad_norm": 0.17267867922782898,
      "learning_rate": 0.0001,
      "loss": 1.5395,
      "step": 17300
    },
    {
      "epoch": 0.47730398899587345,
      "grad_norm": 0.14226895570755005,
      "learning_rate": 0.0001,
      "loss": 1.5386,
      "step": 17350
    },
    {
      "epoch": 0.4786795048143054,
      "grad_norm": 0.15129058063030243,
      "learning_rate": 0.0001,
      "loss": 1.5424,
      "step": 17400
    },
    {
      "epoch": 0.48005502063273725,
      "grad_norm": 0.2448931634426117,
      "learning_rate": 0.0001,
      "loss": 1.5396,
      "step": 17450
    },
    {
      "epoch": 0.4814305364511692,
      "grad_norm": 0.2225511074066162,
      "learning_rate": 0.0001,
      "loss": 1.5404,
      "step": 17500
    },
    {
      "epoch": 0.4828060522696011,
      "grad_norm": 0.1891157031059265,
      "learning_rate": 0.0001,
      "loss": 1.5394,
      "step": 17550
    },
    {
      "epoch": 0.48418156808803303,
      "grad_norm": 0.1472170352935791,
      "learning_rate": 0.0001,
      "loss": 1.5417,
      "step": 17600
    },
    {
      "epoch": 0.4855570839064649,
      "grad_norm": 0.1682361215353012,
      "learning_rate": 0.0001,
      "loss": 1.5377,
      "step": 17650
    },
    {
      "epoch": 0.48693259972489683,
      "grad_norm": 0.18433457612991333,
      "learning_rate": 0.0001,
      "loss": 1.5396,
      "step": 17700
    },
    {
      "epoch": 0.48830811554332876,
      "grad_norm": 0.15077999234199524,
      "learning_rate": 0.0001,
      "loss": 1.5392,
      "step": 17750
    },
    {
      "epoch": 0.4896836313617607,
      "grad_norm": 0.16640494763851166,
      "learning_rate": 0.0001,
      "loss": 1.5381,
      "step": 17800
    },
    {
      "epoch": 0.49105914718019256,
      "grad_norm": 0.1587841510772705,
      "learning_rate": 0.0001,
      "loss": 1.5386,
      "step": 17850
    },
    {
      "epoch": 0.4924346629986245,
      "grad_norm": 0.15444575250148773,
      "learning_rate": 0.0001,
      "loss": 1.5389,
      "step": 17900
    },
    {
      "epoch": 0.4938101788170564,
      "grad_norm": 0.18525558710098267,
      "learning_rate": 0.0001,
      "loss": 1.5404,
      "step": 17950
    },
    {
      "epoch": 0.4951856946354883,
      "grad_norm": 0.12790025770664215,
      "learning_rate": 0.0001,
      "loss": 1.5394,
      "step": 18000
    },
    {
      "epoch": 0.4965612104539202,
      "grad_norm": 0.12284336239099503,
      "learning_rate": 0.0001,
      "loss": 1.5389,
      "step": 18050
    },
    {
      "epoch": 0.49793672627235214,
      "grad_norm": 0.12023458629846573,
      "learning_rate": 0.0001,
      "loss": 1.5345,
      "step": 18100
    },
    {
      "epoch": 0.49931224209078406,
      "grad_norm": 0.220647931098938,
      "learning_rate": 0.0001,
      "loss": 1.5398,
      "step": 18150
    },
    {
      "epoch": 0.5006877579092159,
      "grad_norm": 0.1563023179769516,
      "learning_rate": 0.0001,
      "loss": 1.5361,
      "step": 18200
    },
    {
      "epoch": 0.5020632737276479,
      "grad_norm": 0.15485098958015442,
      "learning_rate": 0.0001,
      "loss": 1.539,
      "step": 18250
    },
    {
      "epoch": 0.5034387895460798,
      "grad_norm": 0.21312743425369263,
      "learning_rate": 0.0001,
      "loss": 1.5378,
      "step": 18300
    },
    {
      "epoch": 0.5048143053645117,
      "grad_norm": 0.1381313055753708,
      "learning_rate": 0.0001,
      "loss": 1.5396,
      "step": 18350
    },
    {
      "epoch": 0.5061898211829436,
      "grad_norm": 0.1357322335243225,
      "learning_rate": 0.0001,
      "loss": 1.5398,
      "step": 18400
    },
    {
      "epoch": 0.5075653370013755,
      "grad_norm": 0.16733530163764954,
      "learning_rate": 0.0001,
      "loss": 1.5381,
      "step": 18450
    },
    {
      "epoch": 0.5089408528198074,
      "grad_norm": 0.12985962629318237,
      "learning_rate": 0.0001,
      "loss": 1.5391,
      "step": 18500
    },
    {
      "epoch": 0.5103163686382394,
      "grad_norm": 0.17726540565490723,
      "learning_rate": 0.0001,
      "loss": 1.5406,
      "step": 18550
    },
    {
      "epoch": 0.5116918844566712,
      "grad_norm": 0.1869622766971588,
      "learning_rate": 0.0001,
      "loss": 1.5379,
      "step": 18600
    },
    {
      "epoch": 0.5130674002751031,
      "grad_norm": 0.19111870229244232,
      "learning_rate": 0.0001,
      "loss": 1.5373,
      "step": 18650
    },
    {
      "epoch": 0.5144429160935351,
      "grad_norm": 0.16479162871837616,
      "learning_rate": 0.0001,
      "loss": 1.5346,
      "step": 18700
    },
    {
      "epoch": 0.515818431911967,
      "grad_norm": 0.17092610895633698,
      "learning_rate": 0.0001,
      "loss": 1.5387,
      "step": 18750
    },
    {
      "epoch": 0.517193947730399,
      "grad_norm": 0.1678820550441742,
      "learning_rate": 0.0001,
      "loss": 1.5376,
      "step": 18800
    },
    {
      "epoch": 0.5185694635488308,
      "grad_norm": 0.14618681371212006,
      "learning_rate": 0.0001,
      "loss": 1.5353,
      "step": 18850
    },
    {
      "epoch": 0.5199449793672627,
      "grad_norm": 0.192416712641716,
      "learning_rate": 0.0001,
      "loss": 1.54,
      "step": 18900
    },
    {
      "epoch": 0.5213204951856947,
      "grad_norm": 0.17582687735557556,
      "learning_rate": 0.0001,
      "loss": 1.5346,
      "step": 18950
    },
    {
      "epoch": 0.5226960110041265,
      "grad_norm": 0.19511322677135468,
      "learning_rate": 0.0001,
      "loss": 1.5371,
      "step": 19000
    },
    {
      "epoch": 0.5240715268225584,
      "grad_norm": 0.15874715149402618,
      "learning_rate": 0.0001,
      "loss": 1.5362,
      "step": 19050
    },
    {
      "epoch": 0.5254470426409904,
      "grad_norm": 0.17555968463420868,
      "learning_rate": 0.0001,
      "loss": 1.5342,
      "step": 19100
    },
    {
      "epoch": 0.5268225584594223,
      "grad_norm": 0.17204701900482178,
      "learning_rate": 0.0001,
      "loss": 1.5356,
      "step": 19150
    },
    {
      "epoch": 0.5281980742778541,
      "grad_norm": 0.1334696263074875,
      "learning_rate": 0.0001,
      "loss": 1.5378,
      "step": 19200
    },
    {
      "epoch": 0.5295735900962861,
      "grad_norm": 0.12202008068561554,
      "learning_rate": 0.0001,
      "loss": 1.536,
      "step": 19250
    },
    {
      "epoch": 0.530949105914718,
      "grad_norm": 0.1914770007133484,
      "learning_rate": 0.0001,
      "loss": 1.5361,
      "step": 19300
    },
    {
      "epoch": 0.53232462173315,
      "grad_norm": 0.18114732205867767,
      "learning_rate": 0.0001,
      "loss": 1.5391,
      "step": 19350
    },
    {
      "epoch": 0.5337001375515819,
      "grad_norm": 0.13230808079242706,
      "learning_rate": 0.0001,
      "loss": 1.5398,
      "step": 19400
    },
    {
      "epoch": 0.5350756533700137,
      "grad_norm": 0.24269579350948334,
      "learning_rate": 0.0001,
      "loss": 1.535,
      "step": 19450
    },
    {
      "epoch": 0.5364511691884457,
      "grad_norm": 0.14454102516174316,
      "learning_rate": 0.0001,
      "loss": 1.5339,
      "step": 19500
    },
    {
      "epoch": 0.5378266850068776,
      "grad_norm": 0.17638514935970306,
      "learning_rate": 0.0001,
      "loss": 1.5385,
      "step": 19550
    },
    {
      "epoch": 0.5392022008253095,
      "grad_norm": 0.1496788114309311,
      "learning_rate": 0.0001,
      "loss": 1.5389,
      "step": 19600
    },
    {
      "epoch": 0.5405777166437414,
      "grad_norm": 0.1927812695503235,
      "learning_rate": 0.0001,
      "loss": 1.5357,
      "step": 19650
    },
    {
      "epoch": 0.5419532324621733,
      "grad_norm": 0.1372377574443817,
      "learning_rate": 0.0001,
      "loss": 1.5363,
      "step": 19700
    },
    {
      "epoch": 0.5433287482806052,
      "grad_norm": 0.15738138556480408,
      "learning_rate": 0.0001,
      "loss": 1.5358,
      "step": 19750
    },
    {
      "epoch": 0.5447042640990372,
      "grad_norm": 0.13599953055381775,
      "learning_rate": 0.0001,
      "loss": 1.5357,
      "step": 19800
    },
    {
      "epoch": 0.546079779917469,
      "grad_norm": 0.16571839153766632,
      "learning_rate": 0.0001,
      "loss": 1.5343,
      "step": 19850
    },
    {
      "epoch": 0.547455295735901,
      "grad_norm": 0.14264202117919922,
      "learning_rate": 0.0001,
      "loss": 1.5315,
      "step": 19900
    },
    {
      "epoch": 0.5488308115543329,
      "grad_norm": 0.15331332385540009,
      "learning_rate": 0.0001,
      "loss": 1.5344,
      "step": 19950
    },
    {
      "epoch": 0.5502063273727648,
      "grad_norm": 0.1380966752767563,
      "learning_rate": 0.0001,
      "loss": 1.5357,
      "step": 20000
    },
    {
      "epoch": 0.5515818431911967,
      "grad_norm": 0.198713481426239,
      "learning_rate": 0.0001,
      "loss": 1.5323,
      "step": 20050
    },
    {
      "epoch": 0.5529573590096286,
      "grad_norm": 0.12092329561710358,
      "learning_rate": 0.0001,
      "loss": 1.5328,
      "step": 20100
    },
    {
      "epoch": 0.5543328748280605,
      "grad_norm": 0.13770416378974915,
      "learning_rate": 0.0001,
      "loss": 1.5346,
      "step": 20150
    },
    {
      "epoch": 0.5557083906464925,
      "grad_norm": 0.12443804740905762,
      "learning_rate": 0.0001,
      "loss": 1.5312,
      "step": 20200
    },
    {
      "epoch": 0.5570839064649243,
      "grad_norm": 0.15430398285388947,
      "learning_rate": 0.0001,
      "loss": 1.5322,
      "step": 20250
    },
    {
      "epoch": 0.5584594222833562,
      "grad_norm": 0.1415732502937317,
      "learning_rate": 0.0001,
      "loss": 1.5338,
      "step": 20300
    },
    {
      "epoch": 0.5598349381017882,
      "grad_norm": 0.2753756642341614,
      "learning_rate": 0.0001,
      "loss": 1.5329,
      "step": 20350
    },
    {
      "epoch": 0.5612104539202201,
      "grad_norm": 0.1666756421327591,
      "learning_rate": 0.0001,
      "loss": 1.5337,
      "step": 20400
    },
    {
      "epoch": 0.562585969738652,
      "grad_norm": 0.17720907926559448,
      "learning_rate": 0.0001,
      "loss": 1.5312,
      "step": 20450
    },
    {
      "epoch": 0.5639614855570839,
      "grad_norm": 0.18275785446166992,
      "learning_rate": 0.0001,
      "loss": 1.5333,
      "step": 20500
    },
    {
      "epoch": 0.5653370013755158,
      "grad_norm": 0.20009452104568481,
      "learning_rate": 0.0001,
      "loss": 1.5301,
      "step": 20550
    },
    {
      "epoch": 0.5667125171939478,
      "grad_norm": 0.18812476098537445,
      "learning_rate": 0.0001,
      "loss": 1.5332,
      "step": 20600
    },
    {
      "epoch": 0.5680880330123796,
      "grad_norm": 0.15448282659053802,
      "learning_rate": 0.0001,
      "loss": 1.5323,
      "step": 20650
    },
    {
      "epoch": 0.5694635488308115,
      "grad_norm": 0.1646738499403,
      "learning_rate": 0.0001,
      "loss": 1.5335,
      "step": 20700
    },
    {
      "epoch": 0.5708390646492435,
      "grad_norm": 0.15908415615558624,
      "learning_rate": 0.0001,
      "loss": 1.5319,
      "step": 20750
    },
    {
      "epoch": 0.5722145804676754,
      "grad_norm": 0.15112848579883575,
      "learning_rate": 0.0001,
      "loss": 1.5342,
      "step": 20800
    },
    {
      "epoch": 0.5735900962861072,
      "grad_norm": 0.3316288888454437,
      "learning_rate": 0.0001,
      "loss": 1.5344,
      "step": 20850
    },
    {
      "epoch": 0.5749656121045392,
      "grad_norm": 0.13579101860523224,
      "learning_rate": 0.0001,
      "loss": 1.5321,
      "step": 20900
    },
    {
      "epoch": 0.5763411279229711,
      "grad_norm": 0.2203134000301361,
      "learning_rate": 0.0001,
      "loss": 1.5324,
      "step": 20950
    },
    {
      "epoch": 0.5777166437414031,
      "grad_norm": 0.1271039992570877,
      "learning_rate": 0.0001,
      "loss": 1.5328,
      "step": 21000
    },
    {
      "epoch": 0.579092159559835,
      "grad_norm": 0.3165966272354126,
      "learning_rate": 0.0001,
      "loss": 1.5349,
      "step": 21050
    },
    {
      "epoch": 0.5804676753782668,
      "grad_norm": 0.1456591635942459,
      "learning_rate": 0.0001,
      "loss": 1.5343,
      "step": 21100
    },
    {
      "epoch": 0.5818431911966988,
      "grad_norm": 0.16555163264274597,
      "learning_rate": 0.0001,
      "loss": 1.5349,
      "step": 21150
    },
    {
      "epoch": 0.5832187070151307,
      "grad_norm": 0.22577494382858276,
      "learning_rate": 0.0001,
      "loss": 1.5342,
      "step": 21200
    },
    {
      "epoch": 0.5845942228335625,
      "grad_norm": 0.23455490171909332,
      "learning_rate": 0.0001,
      "loss": 1.5346,
      "step": 21250
    },
    {
      "epoch": 0.5859697386519945,
      "grad_norm": 0.2247081696987152,
      "learning_rate": 0.0001,
      "loss": 1.5316,
      "step": 21300
    },
    {
      "epoch": 0.5873452544704264,
      "grad_norm": 0.15159213542938232,
      "learning_rate": 0.0001,
      "loss": 1.534,
      "step": 21350
    },
    {
      "epoch": 0.5887207702888583,
      "grad_norm": 0.20483700931072235,
      "learning_rate": 0.0001,
      "loss": 1.5295,
      "step": 21400
    },
    {
      "epoch": 0.5900962861072903,
      "grad_norm": 0.16780568659305573,
      "learning_rate": 0.0001,
      "loss": 1.5341,
      "step": 21450
    },
    {
      "epoch": 0.5914718019257221,
      "grad_norm": 0.15840616822242737,
      "learning_rate": 0.0001,
      "loss": 1.5339,
      "step": 21500
    },
    {
      "epoch": 0.5928473177441541,
      "grad_norm": 0.1488318294286728,
      "learning_rate": 0.0001,
      "loss": 1.5341,
      "step": 21550
    },
    {
      "epoch": 0.594222833562586,
      "grad_norm": 0.13899248838424683,
      "learning_rate": 0.0001,
      "loss": 1.5323,
      "step": 21600
    },
    {
      "epoch": 0.5955983493810179,
      "grad_norm": 0.15024836361408234,
      "learning_rate": 0.0001,
      "loss": 1.5318,
      "step": 21650
    },
    {
      "epoch": 0.5969738651994498,
      "grad_norm": 0.19209244847297668,
      "learning_rate": 0.0001,
      "loss": 1.5325,
      "step": 21700
    },
    {
      "epoch": 0.5983493810178817,
      "grad_norm": 0.20580926537513733,
      "learning_rate": 0.0001,
      "loss": 1.5324,
      "step": 21750
    },
    {
      "epoch": 0.5997248968363136,
      "grad_norm": 0.2091200202703476,
      "learning_rate": 0.0001,
      "loss": 1.5282,
      "step": 21800
    },
    {
      "epoch": 0.6011004126547456,
      "grad_norm": 0.1571815311908722,
      "learning_rate": 0.0001,
      "loss": 1.532,
      "step": 21850
    },
    {
      "epoch": 0.6024759284731774,
      "grad_norm": 0.17794279754161835,
      "learning_rate": 0.0001,
      "loss": 1.5326,
      "step": 21900
    },
    {
      "epoch": 0.6038514442916093,
      "grad_norm": 0.1439165472984314,
      "learning_rate": 0.0001,
      "loss": 1.5325,
      "step": 21950
    },
    {
      "epoch": 0.6052269601100413,
      "grad_norm": 0.15884612500667572,
      "learning_rate": 0.0001,
      "loss": 1.5329,
      "step": 22000
    },
    {
      "epoch": 0.6066024759284732,
      "grad_norm": 0.26263782382011414,
      "learning_rate": 0.0001,
      "loss": 1.5315,
      "step": 22050
    },
    {
      "epoch": 0.6079779917469051,
      "grad_norm": 0.19535377621650696,
      "learning_rate": 0.0001,
      "loss": 1.5308,
      "step": 22100
    },
    {
      "epoch": 0.609353507565337,
      "grad_norm": 0.14018963277339935,
      "learning_rate": 0.0001,
      "loss": 1.5332,
      "step": 22150
    },
    {
      "epoch": 0.6107290233837689,
      "grad_norm": 0.15927653014659882,
      "learning_rate": 0.0001,
      "loss": 1.5299,
      "step": 22200
    },
    {
      "epoch": 0.6121045392022009,
      "grad_norm": 0.143597811460495,
      "learning_rate": 0.0001,
      "loss": 1.532,
      "step": 22250
    },
    {
      "epoch": 0.6134800550206327,
      "grad_norm": 0.15887697041034698,
      "learning_rate": 0.0001,
      "loss": 1.5313,
      "step": 22300
    },
    {
      "epoch": 0.6148555708390646,
      "grad_norm": 0.1907578855752945,
      "learning_rate": 0.0001,
      "loss": 1.5323,
      "step": 22350
    },
    {
      "epoch": 0.6162310866574966,
      "grad_norm": 0.189689502120018,
      "learning_rate": 0.0001,
      "loss": 1.5319,
      "step": 22400
    },
    {
      "epoch": 0.6176066024759285,
      "grad_norm": 0.15399134159088135,
      "learning_rate": 0.0001,
      "loss": 1.5291,
      "step": 22450
    },
    {
      "epoch": 0.6189821182943603,
      "grad_norm": 0.16801948845386505,
      "learning_rate": 0.0001,
      "loss": 1.5319,
      "step": 22500
    },
    {
      "epoch": 0.6203576341127923,
      "grad_norm": 0.21341322362422943,
      "learning_rate": 0.0001,
      "loss": 1.5311,
      "step": 22550
    },
    {
      "epoch": 0.6217331499312242,
      "grad_norm": 0.19961433112621307,
      "learning_rate": 0.0001,
      "loss": 1.529,
      "step": 22600
    },
    {
      "epoch": 0.6231086657496562,
      "grad_norm": 0.1254952847957611,
      "learning_rate": 0.0001,
      "loss": 1.528,
      "step": 22650
    },
    {
      "epoch": 0.624484181568088,
      "grad_norm": 0.21346162259578705,
      "learning_rate": 0.0001,
      "loss": 1.5323,
      "step": 22700
    },
    {
      "epoch": 0.6258596973865199,
      "grad_norm": 0.1551300436258316,
      "learning_rate": 0.0001,
      "loss": 1.5302,
      "step": 22750
    },
    {
      "epoch": 0.6272352132049519,
      "grad_norm": 0.1974526047706604,
      "learning_rate": 0.0001,
      "loss": 1.5294,
      "step": 22800
    },
    {
      "epoch": 0.6286107290233838,
      "grad_norm": 0.130974680185318,
      "learning_rate": 0.0001,
      "loss": 1.5303,
      "step": 22850
    },
    {
      "epoch": 0.6299862448418156,
      "grad_norm": 0.17787273228168488,
      "learning_rate": 0.0001,
      "loss": 1.5299,
      "step": 22900
    },
    {
      "epoch": 0.6313617606602476,
      "grad_norm": 0.19317127764225006,
      "learning_rate": 0.0001,
      "loss": 1.5295,
      "step": 22950
    },
    {
      "epoch": 0.6327372764786795,
      "grad_norm": 0.2229757010936737,
      "learning_rate": 0.0001,
      "loss": 1.5307,
      "step": 23000
    },
    {
      "epoch": 0.6341127922971114,
      "grad_norm": 0.17582648992538452,
      "learning_rate": 0.0001,
      "loss": 1.5294,
      "step": 23050
    },
    {
      "epoch": 0.6354883081155434,
      "grad_norm": 0.17122450470924377,
      "learning_rate": 0.0001,
      "loss": 1.5291,
      "step": 23100
    },
    {
      "epoch": 0.6368638239339752,
      "grad_norm": 0.16124916076660156,
      "learning_rate": 0.0001,
      "loss": 1.5268,
      "step": 23150
    },
    {
      "epoch": 0.6382393397524071,
      "grad_norm": 0.18122687935829163,
      "learning_rate": 0.0001,
      "loss": 1.5274,
      "step": 23200
    },
    {
      "epoch": 0.6396148555708391,
      "grad_norm": 0.17480894923210144,
      "learning_rate": 0.0001,
      "loss": 1.5276,
      "step": 23250
    },
    {
      "epoch": 0.640990371389271,
      "grad_norm": 0.1798102855682373,
      "learning_rate": 0.0001,
      "loss": 1.5267,
      "step": 23300
    },
    {
      "epoch": 0.6423658872077029,
      "grad_norm": 0.19186878204345703,
      "learning_rate": 0.0001,
      "loss": 1.5294,
      "step": 23350
    },
    {
      "epoch": 0.6437414030261348,
      "grad_norm": 0.1212744414806366,
      "learning_rate": 0.0001,
      "loss": 1.527,
      "step": 23400
    },
    {
      "epoch": 0.6451169188445667,
      "grad_norm": 0.16844585537910461,
      "learning_rate": 0.0001,
      "loss": 1.5265,
      "step": 23450
    },
    {
      "epoch": 0.6464924346629987,
      "grad_norm": 0.16216999292373657,
      "learning_rate": 0.0001,
      "loss": 1.5288,
      "step": 23500
    },
    {
      "epoch": 0.6478679504814305,
      "grad_norm": 0.157547265291214,
      "learning_rate": 0.0001,
      "loss": 1.5298,
      "step": 23550
    },
    {
      "epoch": 0.6492434662998624,
      "grad_norm": 0.20760610699653625,
      "learning_rate": 0.0001,
      "loss": 1.5264,
      "step": 23600
    },
    {
      "epoch": 0.6506189821182944,
      "grad_norm": 0.19178840517997742,
      "learning_rate": 0.0001,
      "loss": 1.5251,
      "step": 23650
    },
    {
      "epoch": 0.6519944979367263,
      "grad_norm": 0.17904846370220184,
      "learning_rate": 0.0001,
      "loss": 1.5293,
      "step": 23700
    },
    {
      "epoch": 0.6533700137551581,
      "grad_norm": 0.14902061223983765,
      "learning_rate": 0.0001,
      "loss": 1.5278,
      "step": 23750
    },
    {
      "epoch": 0.6547455295735901,
      "grad_norm": 0.1306075155735016,
      "learning_rate": 0.0001,
      "loss": 1.5274,
      "step": 23800
    },
    {
      "epoch": 0.656121045392022,
      "grad_norm": 0.14361289143562317,
      "learning_rate": 0.0001,
      "loss": 1.5259,
      "step": 23850
    },
    {
      "epoch": 0.657496561210454,
      "grad_norm": 0.23775485157966614,
      "learning_rate": 0.0001,
      "loss": 1.528,
      "step": 23900
    },
    {
      "epoch": 0.6588720770288858,
      "grad_norm": 0.12788158655166626,
      "learning_rate": 0.0001,
      "loss": 1.5285,
      "step": 23950
    },
    {
      "epoch": 0.6602475928473177,
      "grad_norm": 0.11719505488872528,
      "learning_rate": 0.0001,
      "loss": 1.5275,
      "step": 24000
    },
    {
      "epoch": 0.6616231086657497,
      "grad_norm": 0.2011108100414276,
      "learning_rate": 0.0001,
      "loss": 1.5276,
      "step": 24050
    },
    {
      "epoch": 0.6629986244841816,
      "grad_norm": 0.16335125267505646,
      "learning_rate": 0.0001,
      "loss": 1.5305,
      "step": 24100
    },
    {
      "epoch": 0.6643741403026134,
      "grad_norm": 0.15488557517528534,
      "learning_rate": 0.0001,
      "loss": 1.5259,
      "step": 24150
    },
    {
      "epoch": 0.6657496561210454,
      "grad_norm": 0.2333500236272812,
      "learning_rate": 0.0001,
      "loss": 1.5269,
      "step": 24200
    },
    {
      "epoch": 0.6671251719394773,
      "grad_norm": 0.14059284329414368,
      "learning_rate": 0.0001,
      "loss": 1.5298,
      "step": 24250
    },
    {
      "epoch": 0.6685006877579092,
      "grad_norm": 0.24036471545696259,
      "learning_rate": 0.0001,
      "loss": 1.5274,
      "step": 24300
    },
    {
      "epoch": 0.6698762035763411,
      "grad_norm": 0.13437625765800476,
      "learning_rate": 0.0001,
      "loss": 1.529,
      "step": 24350
    },
    {
      "epoch": 0.671251719394773,
      "grad_norm": 0.25569766759872437,
      "learning_rate": 0.0001,
      "loss": 1.5259,
      "step": 24400
    },
    {
      "epoch": 0.672627235213205,
      "grad_norm": 0.14324542880058289,
      "learning_rate": 0.0001,
      "loss": 1.5286,
      "step": 24450
    },
    {
      "epoch": 0.6740027510316369,
      "grad_norm": 0.2062855213880539,
      "learning_rate": 0.0001,
      "loss": 1.5259,
      "step": 24500
    },
    {
      "epoch": 0.6753782668500687,
      "grad_norm": 0.18274646997451782,
      "learning_rate": 0.0001,
      "loss": 1.5293,
      "step": 24550
    },
    {
      "epoch": 0.6767537826685007,
      "grad_norm": 0.16611768305301666,
      "learning_rate": 0.0001,
      "loss": 1.5283,
      "step": 24600
    },
    {
      "epoch": 0.6781292984869326,
      "grad_norm": 0.2058711051940918,
      "learning_rate": 0.0001,
      "loss": 1.5253,
      "step": 24650
    },
    {
      "epoch": 0.6795048143053645,
      "grad_norm": 0.16299676895141602,
      "learning_rate": 0.0001,
      "loss": 1.5281,
      "step": 24700
    },
    {
      "epoch": 0.6808803301237965,
      "grad_norm": 0.17875225841999054,
      "learning_rate": 0.0001,
      "loss": 1.5266,
      "step": 24750
    },
    {
      "epoch": 0.6822558459422283,
      "grad_norm": 0.18055297434329987,
      "learning_rate": 0.0001,
      "loss": 1.5269,
      "step": 24800
    },
    {
      "epoch": 0.6836313617606602,
      "grad_norm": 0.22491872310638428,
      "learning_rate": 0.0001,
      "loss": 1.5236,
      "step": 24850
    },
    {
      "epoch": 0.6850068775790922,
      "grad_norm": 0.17760007083415985,
      "learning_rate": 0.0001,
      "loss": 1.5249,
      "step": 24900
    },
    {
      "epoch": 0.686382393397524,
      "grad_norm": 0.19768892228603363,
      "learning_rate": 0.0001,
      "loss": 1.5254,
      "step": 24950
    },
    {
      "epoch": 0.687757909215956,
      "grad_norm": 0.16851931810379028,
      "learning_rate": 0.0001,
      "loss": 1.5284,
      "step": 25000
    },
    {
      "epoch": 0.6891334250343879,
      "grad_norm": 0.16162404417991638,
      "learning_rate": 0.0001,
      "loss": 1.5278,
      "step": 25050
    },
    {
      "epoch": 0.6905089408528198,
      "grad_norm": 0.1808663010597229,
      "learning_rate": 0.0001,
      "loss": 1.5239,
      "step": 25100
    },
    {
      "epoch": 0.6918844566712518,
      "grad_norm": 0.15550534427165985,
      "learning_rate": 0.0001,
      "loss": 1.5266,
      "step": 25150
    },
    {
      "epoch": 0.6932599724896836,
      "grad_norm": 0.22426332533359528,
      "learning_rate": 0.0001,
      "loss": 1.5226,
      "step": 25200
    },
    {
      "epoch": 0.6946354883081155,
      "grad_norm": 0.11868047714233398,
      "learning_rate": 0.0001,
      "loss": 1.5256,
      "step": 25250
    },
    {
      "epoch": 0.6960110041265475,
      "grad_norm": 0.21659235656261444,
      "learning_rate": 0.0001,
      "loss": 1.5284,
      "step": 25300
    },
    {
      "epoch": 0.6973865199449794,
      "grad_norm": 0.1800456941127777,
      "learning_rate": 0.0001,
      "loss": 1.5235,
      "step": 25350
    },
    {
      "epoch": 0.6987620357634112,
      "grad_norm": 0.21043701469898224,
      "learning_rate": 0.0001,
      "loss": 1.5275,
      "step": 25400
    },
    {
      "epoch": 0.7001375515818432,
      "grad_norm": 0.18925617635250092,
      "learning_rate": 0.0001,
      "loss": 1.5279,
      "step": 25450
    },
    {
      "epoch": 0.7015130674002751,
      "grad_norm": 0.1537819653749466,
      "learning_rate": 0.0001,
      "loss": 1.5243,
      "step": 25500
    },
    {
      "epoch": 0.7028885832187071,
      "grad_norm": 0.1832038164138794,
      "learning_rate": 0.0001,
      "loss": 1.5255,
      "step": 25550
    },
    {
      "epoch": 0.7042640990371389,
      "grad_norm": 0.186794713139534,
      "learning_rate": 0.0001,
      "loss": 1.5261,
      "step": 25600
    },
    {
      "epoch": 0.7056396148555708,
      "grad_norm": 0.12374402582645416,
      "learning_rate": 0.0001,
      "loss": 1.526,
      "step": 25650
    },
    {
      "epoch": 0.7070151306740028,
      "grad_norm": 0.16702401638031006,
      "learning_rate": 0.0001,
      "loss": 1.5245,
      "step": 25700
    },
    {
      "epoch": 0.7083906464924347,
      "grad_norm": 0.1393430233001709,
      "learning_rate": 0.0001,
      "loss": 1.5254,
      "step": 25750
    },
    {
      "epoch": 0.7097661623108665,
      "grad_norm": 0.1630173921585083,
      "learning_rate": 0.0001,
      "loss": 1.5251,
      "step": 25800
    },
    {
      "epoch": 0.7111416781292985,
      "grad_norm": 0.1440727412700653,
      "learning_rate": 0.0001,
      "loss": 1.5282,
      "step": 25850
    },
    {
      "epoch": 0.7125171939477304,
      "grad_norm": 0.17978446185588837,
      "learning_rate": 0.0001,
      "loss": 1.5262,
      "step": 25900
    },
    {
      "epoch": 0.7138927097661623,
      "grad_norm": 0.151292085647583,
      "learning_rate": 0.0001,
      "loss": 1.527,
      "step": 25950
    },
    {
      "epoch": 0.7152682255845942,
      "grad_norm": 0.24109718203544617,
      "learning_rate": 0.0001,
      "loss": 1.5235,
      "step": 26000
    },
    {
      "epoch": 0.7166437414030261,
      "grad_norm": 0.15700335800647736,
      "learning_rate": 0.0001,
      "loss": 1.5245,
      "step": 26050
    },
    {
      "epoch": 0.7180192572214581,
      "grad_norm": 0.14807374775409698,
      "learning_rate": 0.0001,
      "loss": 1.5224,
      "step": 26100
    },
    {
      "epoch": 0.71939477303989,
      "grad_norm": 0.13032929599285126,
      "learning_rate": 0.0001,
      "loss": 1.5221,
      "step": 26150
    },
    {
      "epoch": 0.7207702888583218,
      "grad_norm": 0.1900160163640976,
      "learning_rate": 0.0001,
      "loss": 1.5259,
      "step": 26200
    },
    {
      "epoch": 0.7221458046767538,
      "grad_norm": 0.20619365572929382,
      "learning_rate": 0.0001,
      "loss": 1.5261,
      "step": 26250
    },
    {
      "epoch": 0.7235213204951857,
      "grad_norm": 0.17259658873081207,
      "learning_rate": 0.0001,
      "loss": 1.5272,
      "step": 26300
    },
    {
      "epoch": 0.7248968363136176,
      "grad_norm": 0.1594364494085312,
      "learning_rate": 0.0001,
      "loss": 1.5242,
      "step": 26350
    },
    {
      "epoch": 0.7262723521320495,
      "grad_norm": 0.16156145930290222,
      "learning_rate": 0.0001,
      "loss": 1.5263,
      "step": 26400
    },
    {
      "epoch": 0.7276478679504814,
      "grad_norm": 0.15612217783927917,
      "learning_rate": 0.0001,
      "loss": 1.5232,
      "step": 26450
    },
    {
      "epoch": 0.7290233837689133,
      "grad_norm": 0.2097177803516388,
      "learning_rate": 0.0001,
      "loss": 1.5265,
      "step": 26500
    },
    {
      "epoch": 0.7303988995873453,
      "grad_norm": 0.18174001574516296,
      "learning_rate": 0.0001,
      "loss": 1.5235,
      "step": 26550
    },
    {
      "epoch": 0.7317744154057771,
      "grad_norm": 0.15661188960075378,
      "learning_rate": 0.0001,
      "loss": 1.5239,
      "step": 26600
    },
    {
      "epoch": 0.7331499312242091,
      "grad_norm": 0.17666810750961304,
      "learning_rate": 0.0001,
      "loss": 1.5244,
      "step": 26650
    },
    {
      "epoch": 0.734525447042641,
      "grad_norm": 0.135247141122818,
      "learning_rate": 0.0001,
      "loss": 1.5228,
      "step": 26700
    },
    {
      "epoch": 0.7359009628610729,
      "grad_norm": 0.17839883267879486,
      "learning_rate": 0.0001,
      "loss": 1.522,
      "step": 26750
    },
    {
      "epoch": 0.7372764786795049,
      "grad_norm": 0.1601705551147461,
      "learning_rate": 0.0001,
      "loss": 1.5258,
      "step": 26800
    },
    {
      "epoch": 0.7386519944979367,
      "grad_norm": 0.21927671134471893,
      "learning_rate": 0.0001,
      "loss": 1.5234,
      "step": 26850
    },
    {
      "epoch": 0.7400275103163686,
      "grad_norm": 0.18870490789413452,
      "learning_rate": 0.0001,
      "loss": 1.5222,
      "step": 26900
    },
    {
      "epoch": 0.7414030261348006,
      "grad_norm": 0.17285650968551636,
      "learning_rate": 0.0001,
      "loss": 1.5243,
      "step": 26950
    },
    {
      "epoch": 0.7427785419532325,
      "grad_norm": 0.14226007461547852,
      "learning_rate": 0.0001,
      "loss": 1.5265,
      "step": 27000
    },
    {
      "epoch": 0.7441540577716643,
      "grad_norm": 0.17631758749485016,
      "learning_rate": 0.0001,
      "loss": 1.5209,
      "step": 27050
    },
    {
      "epoch": 0.7455295735900963,
      "grad_norm": 0.22787536680698395,
      "learning_rate": 0.0001,
      "loss": 1.5233,
      "step": 27100
    },
    {
      "epoch": 0.7469050894085282,
      "grad_norm": 0.14378662407398224,
      "learning_rate": 0.0001,
      "loss": 1.5214,
      "step": 27150
    },
    {
      "epoch": 0.7482806052269602,
      "grad_norm": 0.21862713992595673,
      "learning_rate": 0.0001,
      "loss": 1.5211,
      "step": 27200
    },
    {
      "epoch": 0.749656121045392,
      "grad_norm": 0.15041618049144745,
      "learning_rate": 0.0001,
      "loss": 1.5233,
      "step": 27250
    },
    {
      "epoch": 0.7510316368638239,
      "grad_norm": 0.15543252229690552,
      "learning_rate": 0.0001,
      "loss": 1.5216,
      "step": 27300
    },
    {
      "epoch": 0.7524071526822559,
      "grad_norm": 0.1488107591867447,
      "learning_rate": 0.0001,
      "loss": 1.5237,
      "step": 27350
    },
    {
      "epoch": 0.7537826685006878,
      "grad_norm": 0.2412855178117752,
      "learning_rate": 0.0001,
      "loss": 1.5236,
      "step": 27400
    },
    {
      "epoch": 0.7551581843191196,
      "grad_norm": 0.21001331508159637,
      "learning_rate": 0.0001,
      "loss": 1.5227,
      "step": 27450
    },
    {
      "epoch": 0.7565337001375516,
      "grad_norm": 0.16884082555770874,
      "learning_rate": 0.0001,
      "loss": 1.523,
      "step": 27500
    },
    {
      "epoch": 0.7579092159559835,
      "grad_norm": 0.1195225790143013,
      "learning_rate": 0.0001,
      "loss": 1.5223,
      "step": 27550
    },
    {
      "epoch": 0.7592847317744154,
      "grad_norm": 0.2539023160934448,
      "learning_rate": 0.0001,
      "loss": 1.5223,
      "step": 27600
    },
    {
      "epoch": 0.7606602475928473,
      "grad_norm": 0.17333871126174927,
      "learning_rate": 0.0001,
      "loss": 1.5207,
      "step": 27650
    },
    {
      "epoch": 0.7620357634112792,
      "grad_norm": 0.14636480808258057,
      "learning_rate": 0.0001,
      "loss": 1.5241,
      "step": 27700
    },
    {
      "epoch": 0.7634112792297112,
      "grad_norm": 0.13305403292179108,
      "learning_rate": 0.0001,
      "loss": 1.5224,
      "step": 27750
    },
    {
      "epoch": 0.7647867950481431,
      "grad_norm": 0.18532030284404755,
      "learning_rate": 0.0001,
      "loss": 1.5234,
      "step": 27800
    },
    {
      "epoch": 0.7661623108665749,
      "grad_norm": 0.1548730880022049,
      "learning_rate": 0.0001,
      "loss": 1.5224,
      "step": 27850
    },
    {
      "epoch": 0.7675378266850069,
      "grad_norm": 0.20586071908473969,
      "learning_rate": 0.0001,
      "loss": 1.5219,
      "step": 27900
    },
    {
      "epoch": 0.7689133425034388,
      "grad_norm": 0.13693679869174957,
      "learning_rate": 0.0001,
      "loss": 1.5226,
      "step": 27950
    },
    {
      "epoch": 0.7702888583218707,
      "grad_norm": 0.17651352286338806,
      "learning_rate": 0.0001,
      "loss": 1.5198,
      "step": 28000
    },
    {
      "epoch": 0.7716643741403026,
      "grad_norm": 0.19794145226478577,
      "learning_rate": 0.0001,
      "loss": 1.5243,
      "step": 28050
    },
    {
      "epoch": 0.7730398899587345,
      "grad_norm": 0.14593897759914398,
      "learning_rate": 0.0001,
      "loss": 1.5203,
      "step": 28100
    },
    {
      "epoch": 0.7744154057771664,
      "grad_norm": 0.18138128519058228,
      "learning_rate": 0.0001,
      "loss": 1.5189,
      "step": 28150
    },
    {
      "epoch": 0.7757909215955984,
      "grad_norm": 0.15987426042556763,
      "learning_rate": 0.0001,
      "loss": 1.5209,
      "step": 28200
    },
    {
      "epoch": 0.7771664374140302,
      "grad_norm": 0.15444040298461914,
      "learning_rate": 0.0001,
      "loss": 1.5187,
      "step": 28250
    },
    {
      "epoch": 0.7785419532324622,
      "grad_norm": 0.22651028633117676,
      "learning_rate": 0.0001,
      "loss": 1.5201,
      "step": 28300
    },
    {
      "epoch": 0.7799174690508941,
      "grad_norm": 0.1889326423406601,
      "learning_rate": 0.0001,
      "loss": 1.522,
      "step": 28350
    },
    {
      "epoch": 0.781292984869326,
      "grad_norm": 0.1659088283777237,
      "learning_rate": 0.0001,
      "loss": 1.5211,
      "step": 28400
    },
    {
      "epoch": 0.782668500687758,
      "grad_norm": 0.20580235123634338,
      "learning_rate": 0.0001,
      "loss": 1.5215,
      "step": 28450
    },
    {
      "epoch": 0.7840440165061898,
      "grad_norm": 0.1748579442501068,
      "learning_rate": 0.0001,
      "loss": 1.5199,
      "step": 28500
    },
    {
      "epoch": 0.7854195323246217,
      "grad_norm": 0.20172914862632751,
      "learning_rate": 0.0001,
      "loss": 1.5228,
      "step": 28550
    },
    {
      "epoch": 0.7867950481430537,
      "grad_norm": 0.1552000194787979,
      "learning_rate": 0.0001,
      "loss": 1.5205,
      "step": 28600
    },
    {
      "epoch": 0.7881705639614855,
      "grad_norm": 0.18557365238666534,
      "learning_rate": 0.0001,
      "loss": 1.5234,
      "step": 28650
    },
    {
      "epoch": 0.7895460797799174,
      "grad_norm": 0.17085815966129303,
      "learning_rate": 0.0001,
      "loss": 1.522,
      "step": 28700
    },
    {
      "epoch": 0.7909215955983494,
      "grad_norm": 0.19171683490276337,
      "learning_rate": 0.0001,
      "loss": 1.5187,
      "step": 28750
    },
    {
      "epoch": 0.7922971114167813,
      "grad_norm": 0.3197721838951111,
      "learning_rate": 0.0001,
      "loss": 1.5228,
      "step": 28800
    },
    {
      "epoch": 0.7936726272352133,
      "grad_norm": 0.21279697120189667,
      "learning_rate": 0.0001,
      "loss": 1.5181,
      "step": 28850
    },
    {
      "epoch": 0.7950481430536451,
      "grad_norm": 0.2184215933084488,
      "learning_rate": 0.0001,
      "loss": 1.5214,
      "step": 28900
    },
    {
      "epoch": 0.796423658872077,
      "grad_norm": 0.21635691821575165,
      "learning_rate": 0.0001,
      "loss": 1.523,
      "step": 28950
    },
    {
      "epoch": 0.797799174690509,
      "grad_norm": 0.15319493412971497,
      "learning_rate": 0.0001,
      "loss": 1.5197,
      "step": 29000
    },
    {
      "epoch": 0.7991746905089409,
      "grad_norm": 0.22083012759685516,
      "learning_rate": 0.0001,
      "loss": 1.5219,
      "step": 29050
    },
    {
      "epoch": 0.8005502063273727,
      "grad_norm": 0.15193097293376923,
      "learning_rate": 0.0001,
      "loss": 1.5195,
      "step": 29100
    },
    {
      "epoch": 0.8019257221458047,
      "grad_norm": 0.19553427398204803,
      "learning_rate": 0.0001,
      "loss": 1.5205,
      "step": 29150
    },
    {
      "epoch": 0.8033012379642366,
      "grad_norm": 0.2117278128862381,
      "learning_rate": 0.0001,
      "loss": 1.5203,
      "step": 29200
    },
    {
      "epoch": 0.8046767537826685,
      "grad_norm": 0.15601006150245667,
      "learning_rate": 0.0001,
      "loss": 1.5199,
      "step": 29250
    },
    {
      "epoch": 0.8060522696011004,
      "grad_norm": 0.15379014611244202,
      "learning_rate": 0.0001,
      "loss": 1.5222,
      "step": 29300
    },
    {
      "epoch": 0.8074277854195323,
      "grad_norm": 0.1712176352739334,
      "learning_rate": 0.0001,
      "loss": 1.5204,
      "step": 29350
    },
    {
      "epoch": 0.8088033012379643,
      "grad_norm": 0.19847099483013153,
      "learning_rate": 0.0001,
      "loss": 1.5203,
      "step": 29400
    },
    {
      "epoch": 0.8101788170563962,
      "grad_norm": 0.15735092759132385,
      "learning_rate": 0.0001,
      "loss": 1.5181,
      "step": 29450
    },
    {
      "epoch": 0.811554332874828,
      "grad_norm": 0.2128709852695465,
      "learning_rate": 0.0001,
      "loss": 1.52,
      "step": 29500
    },
    {
      "epoch": 0.81292984869326,
      "grad_norm": 0.23607073724269867,
      "learning_rate": 0.0001,
      "loss": 1.5222,
      "step": 29550
    },
    {
      "epoch": 0.8143053645116919,
      "grad_norm": 0.15351270139217377,
      "learning_rate": 0.0001,
      "loss": 1.5186,
      "step": 29600
    },
    {
      "epoch": 0.8156808803301238,
      "grad_norm": 0.18421980738639832,
      "learning_rate": 0.0001,
      "loss": 1.5189,
      "step": 29650
    },
    {
      "epoch": 0.8170563961485557,
      "grad_norm": 0.15863709151744843,
      "learning_rate": 0.0001,
      "loss": 1.5191,
      "step": 29700
    },
    {
      "epoch": 0.8184319119669876,
      "grad_norm": 0.1642359048128128,
      "learning_rate": 0.0001,
      "loss": 1.5188,
      "step": 29750
    },
    {
      "epoch": 0.8198074277854195,
      "grad_norm": 0.2115437388420105,
      "learning_rate": 0.0001,
      "loss": 1.5193,
      "step": 29800
    },
    {
      "epoch": 0.8211829436038515,
      "grad_norm": 0.1653752475976944,
      "learning_rate": 0.0001,
      "loss": 1.5196,
      "step": 29850
    },
    {
      "epoch": 0.8225584594222833,
      "grad_norm": 0.25687387585639954,
      "learning_rate": 0.0001,
      "loss": 1.5193,
      "step": 29900
    },
    {
      "epoch": 0.8239339752407153,
      "grad_norm": 0.22497384250164032,
      "learning_rate": 0.0001,
      "loss": 1.519,
      "step": 29950
    },
    {
      "epoch": 0.8253094910591472,
      "grad_norm": 0.16616137325763702,
      "learning_rate": 0.0001,
      "loss": 1.5204,
      "step": 30000
    },
    {
      "epoch": 0.8266850068775791,
      "grad_norm": 0.14630819857120514,
      "learning_rate": 0.0001,
      "loss": 1.5208,
      "step": 30050
    },
    {
      "epoch": 0.828060522696011,
      "grad_norm": 0.19977807998657227,
      "learning_rate": 0.0001,
      "loss": 1.5187,
      "step": 30100
    },
    {
      "epoch": 0.8294360385144429,
      "grad_norm": 0.21963287889957428,
      "learning_rate": 0.0001,
      "loss": 1.5181,
      "step": 30150
    },
    {
      "epoch": 0.8308115543328748,
      "grad_norm": 0.2047349214553833,
      "learning_rate": 0.0001,
      "loss": 1.5184,
      "step": 30200
    },
    {
      "epoch": 0.8321870701513068,
      "grad_norm": 0.1430223435163498,
      "learning_rate": 0.0001,
      "loss": 1.5187,
      "step": 30250
    },
    {
      "epoch": 0.8335625859697386,
      "grad_norm": 0.2075473666191101,
      "learning_rate": 0.0001,
      "loss": 1.5185,
      "step": 30300
    },
    {
      "epoch": 0.8349381017881705,
      "grad_norm": 0.22520440816879272,
      "learning_rate": 0.0001,
      "loss": 1.5207,
      "step": 30350
    },
    {
      "epoch": 0.8363136176066025,
      "grad_norm": 0.2137775719165802,
      "learning_rate": 0.0001,
      "loss": 1.5174,
      "step": 30400
    },
    {
      "epoch": 0.8376891334250344,
      "grad_norm": 0.1777603179216385,
      "learning_rate": 0.0001,
      "loss": 1.5189,
      "step": 30450
    },
    {
      "epoch": 0.8390646492434664,
      "grad_norm": 0.13343022763729095,
      "learning_rate": 0.0001,
      "loss": 1.5196,
      "step": 30500
    },
    {
      "epoch": 0.8404401650618982,
      "grad_norm": 0.223526269197464,
      "learning_rate": 0.0001,
      "loss": 1.5201,
      "step": 30550
    },
    {
      "epoch": 0.8418156808803301,
      "grad_norm": 0.2005707323551178,
      "learning_rate": 0.0001,
      "loss": 1.5182,
      "step": 30600
    },
    {
      "epoch": 0.8431911966987621,
      "grad_norm": 0.1620023101568222,
      "learning_rate": 0.0001,
      "loss": 1.5194,
      "step": 30650
    },
    {
      "epoch": 0.844566712517194,
      "grad_norm": 0.1359826922416687,
      "learning_rate": 0.0001,
      "loss": 1.5186,
      "step": 30700
    },
    {
      "epoch": 0.8459422283356258,
      "grad_norm": 0.23660969734191895,
      "learning_rate": 0.0001,
      "loss": 1.5208,
      "step": 30750
    },
    {
      "epoch": 0.8473177441540578,
      "grad_norm": 0.22223958373069763,
      "learning_rate": 0.0001,
      "loss": 1.5167,
      "step": 30800
    },
    {
      "epoch": 0.8486932599724897,
      "grad_norm": 0.22506959736347198,
      "learning_rate": 0.0001,
      "loss": 1.5166,
      "step": 30850
    },
    {
      "epoch": 0.8500687757909215,
      "grad_norm": 0.20386451482772827,
      "learning_rate": 0.0001,
      "loss": 1.5181,
      "step": 30900
    },
    {
      "epoch": 0.8514442916093535,
      "grad_norm": 0.21547478437423706,
      "learning_rate": 0.0001,
      "loss": 1.5184,
      "step": 30950
    },
    {
      "epoch": 0.8528198074277854,
      "grad_norm": 0.2500711977481842,
      "learning_rate": 0.0001,
      "loss": 1.5188,
      "step": 31000
    },
    {
      "epoch": 0.8541953232462174,
      "grad_norm": 0.17289701104164124,
      "learning_rate": 0.0001,
      "loss": 1.5182,
      "step": 31050
    },
    {
      "epoch": 0.8555708390646493,
      "grad_norm": 0.24792905151844025,
      "learning_rate": 0.0001,
      "loss": 1.5201,
      "step": 31100
    },
    {
      "epoch": 0.8569463548830811,
      "grad_norm": 0.16410884261131287,
      "learning_rate": 0.0001,
      "loss": 1.5191,
      "step": 31150
    },
    {
      "epoch": 0.8583218707015131,
      "grad_norm": 0.20413684844970703,
      "learning_rate": 0.0001,
      "loss": 1.5207,
      "step": 31200
    },
    {
      "epoch": 0.859697386519945,
      "grad_norm": 0.1622382253408432,
      "learning_rate": 0.0001,
      "loss": 1.5191,
      "step": 31250
    },
    {
      "epoch": 0.8610729023383769,
      "grad_norm": 0.19682924449443817,
      "learning_rate": 0.0001,
      "loss": 1.5195,
      "step": 31300
    },
    {
      "epoch": 0.8624484181568088,
      "grad_norm": 0.17585939168930054,
      "learning_rate": 0.0001,
      "loss": 1.5182,
      "step": 31350
    },
    {
      "epoch": 0.8638239339752407,
      "grad_norm": 0.3021407127380371,
      "learning_rate": 0.0001,
      "loss": 1.5177,
      "step": 31400
    },
    {
      "epoch": 0.8651994497936726,
      "grad_norm": 0.25355300307273865,
      "learning_rate": 0.0001,
      "loss": 1.5179,
      "step": 31450
    },
    {
      "epoch": 0.8665749656121046,
      "grad_norm": 0.19390764832496643,
      "learning_rate": 0.0001,
      "loss": 1.5146,
      "step": 31500
    },
    {
      "epoch": 0.8679504814305364,
      "grad_norm": 0.14198362827301025,
      "learning_rate": 0.0001,
      "loss": 1.5194,
      "step": 31550
    },
    {
      "epoch": 0.8693259972489684,
      "grad_norm": 0.21591129899024963,
      "learning_rate": 0.0001,
      "loss": 1.516,
      "step": 31600
    },
    {
      "epoch": 0.8707015130674003,
      "grad_norm": 0.142410010099411,
      "learning_rate": 0.0001,
      "loss": 1.5164,
      "step": 31650
    },
    {
      "epoch": 0.8720770288858322,
      "grad_norm": 0.14241962134838104,
      "learning_rate": 0.0001,
      "loss": 1.5144,
      "step": 31700
    },
    {
      "epoch": 0.8734525447042641,
      "grad_norm": 0.1909308135509491,
      "learning_rate": 0.0001,
      "loss": 1.5182,
      "step": 31750
    },
    {
      "epoch": 0.874828060522696,
      "grad_norm": 0.1649756282567978,
      "learning_rate": 0.0001,
      "loss": 1.5145,
      "step": 31800
    },
    {
      "epoch": 0.8762035763411279,
      "grad_norm": 0.26334628462791443,
      "learning_rate": 0.0001,
      "loss": 1.5157,
      "step": 31850
    },
    {
      "epoch": 0.8775790921595599,
      "grad_norm": 0.1725001484155655,
      "learning_rate": 0.0001,
      "loss": 1.5191,
      "step": 31900
    },
    {
      "epoch": 0.8789546079779917,
      "grad_norm": 0.18799418210983276,
      "learning_rate": 0.0001,
      "loss": 1.5171,
      "step": 31950
    },
    {
      "epoch": 0.8803301237964236,
      "grad_norm": 0.15485192835330963,
      "learning_rate": 0.0001,
      "loss": 1.5147,
      "step": 32000
    },
    {
      "epoch": 0.8817056396148556,
      "grad_norm": 0.13494554162025452,
      "learning_rate": 0.0001,
      "loss": 1.5147,
      "step": 32050
    },
    {
      "epoch": 0.8830811554332875,
      "grad_norm": 0.22909484803676605,
      "learning_rate": 0.0001,
      "loss": 1.5154,
      "step": 32100
    },
    {
      "epoch": 0.8844566712517193,
      "grad_norm": 0.2062431126832962,
      "learning_rate": 0.0001,
      "loss": 1.5135,
      "step": 32150
    },
    {
      "epoch": 0.8858321870701513,
      "grad_norm": 0.17063121497631073,
      "learning_rate": 0.0001,
      "loss": 1.517,
      "step": 32200
    },
    {
      "epoch": 0.8872077028885832,
      "grad_norm": 0.1380726397037506,
      "learning_rate": 0.0001,
      "loss": 1.5134,
      "step": 32250
    },
    {
      "epoch": 0.8885832187070152,
      "grad_norm": 0.18543638288974762,
      "learning_rate": 0.0001,
      "loss": 1.5186,
      "step": 32300
    },
    {
      "epoch": 0.889958734525447,
      "grad_norm": 0.28441041707992554,
      "learning_rate": 0.0001,
      "loss": 1.5179,
      "step": 32350
    },
    {
      "epoch": 0.8913342503438789,
      "grad_norm": 0.2097078114748001,
      "learning_rate": 0.0001,
      "loss": 1.518,
      "step": 32400
    },
    {
      "epoch": 0.8927097661623109,
      "grad_norm": 0.16976235806941986,
      "learning_rate": 0.0001,
      "loss": 1.5147,
      "step": 32450
    },
    {
      "epoch": 0.8940852819807428,
      "grad_norm": 0.20023608207702637,
      "learning_rate": 0.0001,
      "loss": 1.5209,
      "step": 32500
    },
    {
      "epoch": 0.8954607977991746,
      "grad_norm": 0.1981000006198883,
      "learning_rate": 0.0001,
      "loss": 1.5161,
      "step": 32550
    },
    {
      "epoch": 0.8968363136176066,
      "grad_norm": 0.24770237505435944,
      "learning_rate": 0.0001,
      "loss": 1.5145,
      "step": 32600
    },
    {
      "epoch": 0.8982118294360385,
      "grad_norm": 0.27108198404312134,
      "learning_rate": 0.0001,
      "loss": 1.5157,
      "step": 32650
    },
    {
      "epoch": 0.8995873452544704,
      "grad_norm": 0.21742689609527588,
      "learning_rate": 0.0001,
      "loss": 1.5176,
      "step": 32700
    },
    {
      "epoch": 0.9009628610729024,
      "grad_norm": 0.18256455659866333,
      "learning_rate": 0.0001,
      "loss": 1.5153,
      "step": 32750
    },
    {
      "epoch": 0.9023383768913342,
      "grad_norm": 0.1812065690755844,
      "learning_rate": 0.0001,
      "loss": 1.517,
      "step": 32800
    },
    {
      "epoch": 0.9037138927097662,
      "grad_norm": 0.1624094694852829,
      "learning_rate": 0.0001,
      "loss": 1.5184,
      "step": 32850
    },
    {
      "epoch": 0.9050894085281981,
      "grad_norm": 0.12931875884532928,
      "learning_rate": 0.0001,
      "loss": 1.5187,
      "step": 32900
    },
    {
      "epoch": 0.90646492434663,
      "grad_norm": 0.15731951594352722,
      "learning_rate": 0.0001,
      "loss": 1.515,
      "step": 32950
    },
    {
      "epoch": 0.9078404401650619,
      "grad_norm": 0.2222890406847,
      "learning_rate": 0.0001,
      "loss": 1.5167,
      "step": 33000
    },
    {
      "epoch": 0.9092159559834938,
      "grad_norm": 0.33150213956832886,
      "learning_rate": 0.0001,
      "loss": 1.5166,
      "step": 33050
    },
    {
      "epoch": 0.9105914718019257,
      "grad_norm": 0.27547687292099,
      "learning_rate": 0.0001,
      "loss": 1.5151,
      "step": 33100
    },
    {
      "epoch": 0.9119669876203577,
      "grad_norm": 0.1873897761106491,
      "learning_rate": 0.0001,
      "loss": 1.5132,
      "step": 33150
    },
    {
      "epoch": 0.9133425034387895,
      "grad_norm": 0.1707950383424759,
      "learning_rate": 0.0001,
      "loss": 1.5149,
      "step": 33200
    },
    {
      "epoch": 0.9147180192572214,
      "grad_norm": 0.1721598356962204,
      "learning_rate": 0.0001,
      "loss": 1.5135,
      "step": 33250
    },
    {
      "epoch": 0.9160935350756534,
      "grad_norm": 0.31545665860176086,
      "learning_rate": 0.0001,
      "loss": 1.5142,
      "step": 33300
    },
    {
      "epoch": 0.9174690508940853,
      "grad_norm": 0.19677673280239105,
      "learning_rate": 0.0001,
      "loss": 1.5114,
      "step": 33350
    },
    {
      "epoch": 0.9188445667125172,
      "grad_norm": 0.19303210079669952,
      "learning_rate": 0.0001,
      "loss": 1.5126,
      "step": 33400
    },
    {
      "epoch": 0.9202200825309491,
      "grad_norm": 0.14599211513996124,
      "learning_rate": 0.0001,
      "loss": 1.5149,
      "step": 33450
    },
    {
      "epoch": 0.921595598349381,
      "grad_norm": 0.2020881623029709,
      "learning_rate": 0.0001,
      "loss": 1.5169,
      "step": 33500
    },
    {
      "epoch": 0.922971114167813,
      "grad_norm": 0.1755484640598297,
      "learning_rate": 0.0001,
      "loss": 1.5146,
      "step": 33550
    },
    {
      "epoch": 0.9243466299862448,
      "grad_norm": 0.15174026787281036,
      "learning_rate": 0.0001,
      "loss": 1.5164,
      "step": 33600
    },
    {
      "epoch": 0.9257221458046767,
      "grad_norm": 0.21369625627994537,
      "learning_rate": 0.0001,
      "loss": 1.5161,
      "step": 33650
    },
    {
      "epoch": 0.9270976616231087,
      "grad_norm": 0.23643817007541656,
      "learning_rate": 0.0001,
      "loss": 1.5129,
      "step": 33700
    },
    {
      "epoch": 0.9284731774415406,
      "grad_norm": 0.22748377919197083,
      "learning_rate": 0.0001,
      "loss": 1.5169,
      "step": 33750
    },
    {
      "epoch": 0.9298486932599724,
      "grad_norm": 0.24398982524871826,
      "learning_rate": 0.0001,
      "loss": 1.5137,
      "step": 33800
    },
    {
      "epoch": 0.9312242090784044,
      "grad_norm": 0.16090893745422363,
      "learning_rate": 0.0001,
      "loss": 1.5126,
      "step": 33850
    },
    {
      "epoch": 0.9325997248968363,
      "grad_norm": 0.1766052097082138,
      "learning_rate": 0.0001,
      "loss": 1.5149,
      "step": 33900
    },
    {
      "epoch": 0.9339752407152683,
      "grad_norm": 0.15594764053821564,
      "learning_rate": 0.0001,
      "loss": 1.5139,
      "step": 33950
    },
    {
      "epoch": 0.9353507565337001,
      "grad_norm": 0.22842876613140106,
      "learning_rate": 0.0001,
      "loss": 1.5152,
      "step": 34000
    },
    {
      "epoch": 0.936726272352132,
      "grad_norm": 0.17382940649986267,
      "learning_rate": 0.0001,
      "loss": 1.5138,
      "step": 34050
    },
    {
      "epoch": 0.938101788170564,
      "grad_norm": 0.19100262224674225,
      "learning_rate": 0.0001,
      "loss": 1.5136,
      "step": 34100
    },
    {
      "epoch": 0.9394773039889959,
      "grad_norm": 0.13861484825611115,
      "learning_rate": 0.0001,
      "loss": 1.5118,
      "step": 34150
    },
    {
      "epoch": 0.9408528198074277,
      "grad_norm": 0.22483597695827484,
      "learning_rate": 0.0001,
      "loss": 1.5119,
      "step": 34200
    },
    {
      "epoch": 0.9422283356258597,
      "grad_norm": 0.20615430176258087,
      "learning_rate": 0.0001,
      "loss": 1.512,
      "step": 34250
    },
    {
      "epoch": 0.9436038514442916,
      "grad_norm": 0.18101869523525238,
      "learning_rate": 0.0001,
      "loss": 1.5142,
      "step": 34300
    },
    {
      "epoch": 0.9449793672627235,
      "grad_norm": 0.19411496818065643,
      "learning_rate": 0.0001,
      "loss": 1.512,
      "step": 34350
    },
    {
      "epoch": 0.9463548830811555,
      "grad_norm": 0.2966468334197998,
      "learning_rate": 0.0001,
      "loss": 1.5121,
      "step": 34400
    },
    {
      "epoch": 0.9477303988995873,
      "grad_norm": 0.2614442706108093,
      "learning_rate": 0.0001,
      "loss": 1.5127,
      "step": 34450
    },
    {
      "epoch": 0.9491059147180193,
      "grad_norm": 0.3327767252922058,
      "learning_rate": 0.0001,
      "loss": 1.5136,
      "step": 34500
    },
    {
      "epoch": 0.9504814305364512,
      "grad_norm": 0.1958717554807663,
      "learning_rate": 0.0001,
      "loss": 1.5133,
      "step": 34550
    },
    {
      "epoch": 0.951856946354883,
      "grad_norm": 0.15711049735546112,
      "learning_rate": 0.0001,
      "loss": 1.5121,
      "step": 34600
    },
    {
      "epoch": 0.953232462173315,
      "grad_norm": 0.2362435758113861,
      "learning_rate": 0.0001,
      "loss": 1.514,
      "step": 34650
    },
    {
      "epoch": 0.9546079779917469,
      "grad_norm": 0.17552147805690765,
      "learning_rate": 0.0001,
      "loss": 1.5115,
      "step": 34700
    },
    {
      "epoch": 0.9559834938101788,
      "grad_norm": 0.16898372769355774,
      "learning_rate": 0.0001,
      "loss": 1.5131,
      "step": 34750
    },
    {
      "epoch": 0.9573590096286108,
      "grad_norm": 0.18677185475826263,
      "learning_rate": 0.0001,
      "loss": 1.5146,
      "step": 34800
    },
    {
      "epoch": 0.9587345254470426,
      "grad_norm": 0.1758512556552887,
      "learning_rate": 0.0001,
      "loss": 1.5141,
      "step": 34850
    },
    {
      "epoch": 0.9601100412654745,
      "grad_norm": 0.18687918782234192,
      "learning_rate": 0.0001,
      "loss": 1.5134,
      "step": 34900
    },
    {
      "epoch": 0.9614855570839065,
      "grad_norm": 0.2375195175409317,
      "learning_rate": 0.0001,
      "loss": 1.5129,
      "step": 34950
    },
    {
      "epoch": 0.9628610729023384,
      "grad_norm": 0.24082688987255096,
      "learning_rate": 0.0001,
      "loss": 1.514,
      "step": 35000
    },
    {
      "epoch": 0.9642365887207703,
      "grad_norm": 0.2279283106327057,
      "learning_rate": 0.0001,
      "loss": 1.5129,
      "step": 35050
    },
    {
      "epoch": 0.9656121045392022,
      "grad_norm": 0.267251193523407,
      "learning_rate": 0.0001,
      "loss": 1.5139,
      "step": 35100
    },
    {
      "epoch": 0.9669876203576341,
      "grad_norm": 0.1902667135000229,
      "learning_rate": 0.0001,
      "loss": 1.5127,
      "step": 35150
    },
    {
      "epoch": 0.9683631361760661,
      "grad_norm": 0.20134538412094116,
      "learning_rate": 0.0001,
      "loss": 1.5137,
      "step": 35200
    },
    {
      "epoch": 0.9697386519944979,
      "grad_norm": 0.21791616082191467,
      "learning_rate": 0.0001,
      "loss": 1.5148,
      "step": 35250
    },
    {
      "epoch": 0.9711141678129298,
      "grad_norm": 0.2014089673757553,
      "learning_rate": 0.0001,
      "loss": 1.5135,
      "step": 35300
    },
    {
      "epoch": 0.9724896836313618,
      "grad_norm": 0.1704970896244049,
      "learning_rate": 0.0001,
      "loss": 1.5148,
      "step": 35350
    },
    {
      "epoch": 0.9738651994497937,
      "grad_norm": 0.15112122893333435,
      "learning_rate": 0.0001,
      "loss": 1.512,
      "step": 35400
    },
    {
      "epoch": 0.9752407152682255,
      "grad_norm": 0.1649782657623291,
      "learning_rate": 0.0001,
      "loss": 1.5107,
      "step": 35450
    },
    {
      "epoch": 0.9766162310866575,
      "grad_norm": 0.2087404876947403,
      "learning_rate": 0.0001,
      "loss": 1.5149,
      "step": 35500
    },
    {
      "epoch": 0.9779917469050894,
      "grad_norm": 0.2056160867214203,
      "learning_rate": 0.0001,
      "loss": 1.511,
      "step": 35550
    },
    {
      "epoch": 0.9793672627235214,
      "grad_norm": 0.2275388538837433,
      "learning_rate": 0.0001,
      "loss": 1.5147,
      "step": 35600
    },
    {
      "epoch": 0.9807427785419532,
      "grad_norm": 0.24389615654945374,
      "learning_rate": 0.0001,
      "loss": 1.5122,
      "step": 35650
    },
    {
      "epoch": 0.9821182943603851,
      "grad_norm": 0.21413607895374298,
      "learning_rate": 0.0001,
      "loss": 1.5119,
      "step": 35700
    },
    {
      "epoch": 0.9834938101788171,
      "grad_norm": 0.19716958701610565,
      "learning_rate": 0.0001,
      "loss": 1.5127,
      "step": 35750
    },
    {
      "epoch": 0.984869325997249,
      "grad_norm": 0.22444148361682892,
      "learning_rate": 0.0001,
      "loss": 1.5128,
      "step": 35800
    },
    {
      "epoch": 0.9862448418156808,
      "grad_norm": 0.15065211057662964,
      "learning_rate": 0.0001,
      "loss": 1.512,
      "step": 35850
    },
    {
      "epoch": 0.9876203576341128,
      "grad_norm": 0.3378779888153076,
      "learning_rate": 0.0001,
      "loss": 1.5108,
      "step": 35900
    },
    {
      "epoch": 0.9889958734525447,
      "grad_norm": 0.17586860060691833,
      "learning_rate": 0.0001,
      "loss": 1.5144,
      "step": 35950
    },
    {
      "epoch": 0.9903713892709766,
      "grad_norm": 0.270921915769577,
      "learning_rate": 0.0001,
      "loss": 1.5142,
      "step": 36000
    },
    {
      "epoch": 0.9917469050894085,
      "grad_norm": 0.18357771635055542,
      "learning_rate": 0.0001,
      "loss": 1.513,
      "step": 36050
    },
    {
      "epoch": 0.9931224209078404,
      "grad_norm": 0.33356377482414246,
      "learning_rate": 0.0001,
      "loss": 1.5129,
      "step": 36100
    },
    {
      "epoch": 0.9944979367262724,
      "grad_norm": 0.19254672527313232,
      "learning_rate": 0.0001,
      "loss": 1.511,
      "step": 36150
    },
    {
      "epoch": 0.9958734525447043,
      "grad_norm": 0.2596052289009094,
      "learning_rate": 0.0001,
      "loss": 1.5113,
      "step": 36200
    },
    {
      "epoch": 0.9972489683631361,
      "grad_norm": 0.3195280432701111,
      "learning_rate": 0.0001,
      "loss": 1.514,
      "step": 36250
    },
    {
      "epoch": 0.9986244841815681,
      "grad_norm": 0.2321728765964508,
      "learning_rate": 0.0001,
      "loss": 1.5121,
      "step": 36300
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.2551921010017395,
      "learning_rate": 0.0001,
      "loss": 1.5127,
      "step": 36350
    }
  ],
  "logging_steps": 50,
  "max_steps": 36350,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "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": 1.3730675995865063e+22,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}