{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.10001577535888942,
  "eval_steps": 20000,
  "global_step": 3170,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0003155071777882947,
      "grad_norm": 1.1314178705215454,
      "learning_rate": 3.1545741324921135e-06,
      "loss": 1.3533,
      "step": 10
    },
    {
      "epoch": 0.0006310143555765894,
      "grad_norm": 0.8602690696716309,
      "learning_rate": 6.309148264984227e-06,
      "loss": 1.3656,
      "step": 20
    },
    {
      "epoch": 0.000946521533364884,
      "grad_norm": 0.46021902561187744,
      "learning_rate": 9.46372239747634e-06,
      "loss": 1.3685,
      "step": 30
    },
    {
      "epoch": 0.0012620287111531788,
      "grad_norm": 0.6980212926864624,
      "learning_rate": 1.2618296529968454e-05,
      "loss": 1.2711,
      "step": 40
    },
    {
      "epoch": 0.0015775358889414735,
      "grad_norm": 0.674638569355011,
      "learning_rate": 1.577287066246057e-05,
      "loss": 1.1823,
      "step": 50
    },
    {
      "epoch": 0.001893043066729768,
      "grad_norm": 0.9752547740936279,
      "learning_rate": 1.892744479495268e-05,
      "loss": 1.3238,
      "step": 60
    },
    {
      "epoch": 0.0022085502445180626,
      "grad_norm": 0.3866540789604187,
      "learning_rate": 2.2082018927444796e-05,
      "loss": 1.2858,
      "step": 70
    },
    {
      "epoch": 0.0025240574223063575,
      "grad_norm": 0.7574986815452576,
      "learning_rate": 2.5236593059936908e-05,
      "loss": 1.2527,
      "step": 80
    },
    {
      "epoch": 0.002839564600094652,
      "grad_norm": 0.380484402179718,
      "learning_rate": 2.8391167192429026e-05,
      "loss": 1.2617,
      "step": 90
    },
    {
      "epoch": 0.003155071777882947,
      "grad_norm": 0.4931378960609436,
      "learning_rate": 3.154574132492114e-05,
      "loss": 1.145,
      "step": 100
    },
    {
      "epoch": 0.0034705789556712416,
      "grad_norm": 0.6048789620399475,
      "learning_rate": 3.470031545741325e-05,
      "loss": 1.1987,
      "step": 110
    },
    {
      "epoch": 0.003786086133459536,
      "grad_norm": 0.29029208421707153,
      "learning_rate": 3.785488958990536e-05,
      "loss": 1.2743,
      "step": 120
    },
    {
      "epoch": 0.004101593311247831,
      "grad_norm": 0.40250363945961,
      "learning_rate": 4.1009463722397477e-05,
      "loss": 1.1172,
      "step": 130
    },
    {
      "epoch": 0.004417100489036125,
      "grad_norm": 0.4228983223438263,
      "learning_rate": 4.416403785488959e-05,
      "loss": 1.2585,
      "step": 140
    },
    {
      "epoch": 0.00473260766682442,
      "grad_norm": 0.7012104392051697,
      "learning_rate": 4.731861198738171e-05,
      "loss": 1.2598,
      "step": 150
    },
    {
      "epoch": 0.005048114844612715,
      "grad_norm": 0.4556664526462555,
      "learning_rate": 5.0473186119873815e-05,
      "loss": 1.2826,
      "step": 160
    },
    {
      "epoch": 0.00536362202240101,
      "grad_norm": 0.4234769642353058,
      "learning_rate": 5.362776025236593e-05,
      "loss": 1.2544,
      "step": 170
    },
    {
      "epoch": 0.005679129200189304,
      "grad_norm": 0.3880709409713745,
      "learning_rate": 5.678233438485805e-05,
      "loss": 1.1758,
      "step": 180
    },
    {
      "epoch": 0.005994636377977599,
      "grad_norm": 0.5538777709007263,
      "learning_rate": 5.993690851735017e-05,
      "loss": 1.3851,
      "step": 190
    },
    {
      "epoch": 0.006310143555765894,
      "grad_norm": 0.3604615032672882,
      "learning_rate": 6.309148264984228e-05,
      "loss": 1.0848,
      "step": 200
    },
    {
      "epoch": 0.006625650733554188,
      "grad_norm": 0.49042531847953796,
      "learning_rate": 6.624605678233438e-05,
      "loss": 1.3279,
      "step": 210
    },
    {
      "epoch": 0.006941157911342483,
      "grad_norm": 0.4594549238681793,
      "learning_rate": 6.94006309148265e-05,
      "loss": 1.0995,
      "step": 220
    },
    {
      "epoch": 0.007256665089130778,
      "grad_norm": 0.5176606178283691,
      "learning_rate": 7.255520504731861e-05,
      "loss": 1.1639,
      "step": 230
    },
    {
      "epoch": 0.007572172266919072,
      "grad_norm": 0.499743789434433,
      "learning_rate": 7.570977917981072e-05,
      "loss": 1.255,
      "step": 240
    },
    {
      "epoch": 0.007887679444707366,
      "grad_norm": 0.5964751243591309,
      "learning_rate": 7.886435331230284e-05,
      "loss": 1.1427,
      "step": 250
    },
    {
      "epoch": 0.008203186622495662,
      "grad_norm": 0.4749509394168854,
      "learning_rate": 8.201892744479495e-05,
      "loss": 1.3347,
      "step": 260
    },
    {
      "epoch": 0.008518693800283956,
      "grad_norm": 0.4441193640232086,
      "learning_rate": 8.517350157728708e-05,
      "loss": 1.1372,
      "step": 270
    },
    {
      "epoch": 0.00883420097807225,
      "grad_norm": 0.576790452003479,
      "learning_rate": 8.832807570977918e-05,
      "loss": 1.2528,
      "step": 280
    },
    {
      "epoch": 0.009149708155860546,
      "grad_norm": 0.4120502471923828,
      "learning_rate": 9.148264984227129e-05,
      "loss": 1.2139,
      "step": 290
    },
    {
      "epoch": 0.00946521533364884,
      "grad_norm": 0.3923511803150177,
      "learning_rate": 9.463722397476341e-05,
      "loss": 1.3012,
      "step": 300
    },
    {
      "epoch": 0.009780722511437134,
      "grad_norm": 0.5546993613243103,
      "learning_rate": 9.779179810725552e-05,
      "loss": 1.1328,
      "step": 310
    },
    {
      "epoch": 0.01009622968922543,
      "grad_norm": 0.4276692271232605,
      "learning_rate": 9.999972717866783e-05,
      "loss": 1.1742,
      "step": 320
    },
    {
      "epoch": 0.010411736867013724,
      "grad_norm": 0.5325655937194824,
      "learning_rate": 9.99948771044751e-05,
      "loss": 1.1614,
      "step": 330
    },
    {
      "epoch": 0.01072724404480202,
      "grad_norm": 0.43508264422416687,
      "learning_rate": 9.998396501092419e-05,
      "loss": 1.1259,
      "step": 340
    },
    {
      "epoch": 0.011042751222590314,
      "grad_norm": 0.5584490299224854,
      "learning_rate": 9.996699222113709e-05,
      "loss": 1.1327,
      "step": 350
    },
    {
      "epoch": 0.011358258400378608,
      "grad_norm": 0.8362613916397095,
      "learning_rate": 9.994396079311228e-05,
      "loss": 1.2165,
      "step": 360
    },
    {
      "epoch": 0.011673765578166904,
      "grad_norm": 0.5259298086166382,
      "learning_rate": 9.991487351947513e-05,
      "loss": 1.1041,
      "step": 370
    },
    {
      "epoch": 0.011989272755955198,
      "grad_norm": 0.3852245807647705,
      "learning_rate": 9.987973392713932e-05,
      "loss": 1.3298,
      "step": 380
    },
    {
      "epoch": 0.012304779933743492,
      "grad_norm": 0.4914740025997162,
      "learning_rate": 9.983854627687918e-05,
      "loss": 1.236,
      "step": 390
    },
    {
      "epoch": 0.012620287111531788,
      "grad_norm": 0.49665552377700806,
      "learning_rate": 9.979131556281304e-05,
      "loss": 1.1544,
      "step": 400
    },
    {
      "epoch": 0.012935794289320082,
      "grad_norm": 0.4329008162021637,
      "learning_rate": 9.973804751179779e-05,
      "loss": 1.2401,
      "step": 410
    },
    {
      "epoch": 0.013251301467108376,
      "grad_norm": 0.30973491072654724,
      "learning_rate": 9.967874858273423e-05,
      "loss": 1.2296,
      "step": 420
    },
    {
      "epoch": 0.013566808644896672,
      "grad_norm": 0.38413724303245544,
      "learning_rate": 9.961342596578422e-05,
      "loss": 1.0889,
      "step": 430
    },
    {
      "epoch": 0.013882315822684966,
      "grad_norm": 0.38513725996017456,
      "learning_rate": 9.954208758149867e-05,
      "loss": 1.2241,
      "step": 440
    },
    {
      "epoch": 0.01419782300047326,
      "grad_norm": 0.4003647565841675,
      "learning_rate": 9.946474207985713e-05,
      "loss": 1.2967,
      "step": 450
    },
    {
      "epoch": 0.014513330178261556,
      "grad_norm": 0.7440866827964783,
      "learning_rate": 9.938139883921906e-05,
      "loss": 1.2981,
      "step": 460
    },
    {
      "epoch": 0.01482883735604985,
      "grad_norm": 0.5401014685630798,
      "learning_rate": 9.929206796518662e-05,
      "loss": 1.2427,
      "step": 470
    },
    {
      "epoch": 0.015144344533838144,
      "grad_norm": 0.4463556408882141,
      "learning_rate": 9.919676028937936e-05,
      "loss": 1.1139,
      "step": 480
    },
    {
      "epoch": 0.01545985171162644,
      "grad_norm": 0.4856787621974945,
      "learning_rate": 9.909548736812085e-05,
      "loss": 1.2958,
      "step": 490
    },
    {
      "epoch": 0.015775358889414733,
      "grad_norm": 0.49902960658073425,
      "learning_rate": 9.898826148103737e-05,
      "loss": 1.1714,
      "step": 500
    },
    {
      "epoch": 0.01609086606720303,
      "grad_norm": 0.3542010486125946,
      "learning_rate": 9.887509562956911e-05,
      "loss": 1.1652,
      "step": 510
    },
    {
      "epoch": 0.016406373244991324,
      "grad_norm": 0.7918481826782227,
      "learning_rate": 9.875600353539363e-05,
      "loss": 1.0727,
      "step": 520
    },
    {
      "epoch": 0.01672188042277962,
      "grad_norm": 0.6717935800552368,
      "learning_rate": 9.863099963876201e-05,
      "loss": 1.2942,
      "step": 530
    },
    {
      "epoch": 0.017037387600567912,
      "grad_norm": 0.307864785194397,
      "learning_rate": 9.850009909674807e-05,
      "loss": 1.157,
      "step": 540
    },
    {
      "epoch": 0.017352894778356207,
      "grad_norm": 0.6204475164413452,
      "learning_rate": 9.83633177814104e-05,
      "loss": 1.2506,
      "step": 550
    },
    {
      "epoch": 0.0176684019561445,
      "grad_norm": 0.44475409388542175,
      "learning_rate": 9.822067227786793e-05,
      "loss": 1.1519,
      "step": 560
    },
    {
      "epoch": 0.017983909133932798,
      "grad_norm": 0.5524726510047913,
      "learning_rate": 9.807217988228884e-05,
      "loss": 1.0709,
      "step": 570
    },
    {
      "epoch": 0.018299416311721092,
      "grad_norm": 0.44522351026535034,
      "learning_rate": 9.791785859979344e-05,
      "loss": 1.1671,
      "step": 580
    },
    {
      "epoch": 0.018614923489509386,
      "grad_norm": 0.591107964515686,
      "learning_rate": 9.775772714227096e-05,
      "loss": 1.1114,
      "step": 590
    },
    {
      "epoch": 0.01893043066729768,
      "grad_norm": 0.5873256325721741,
      "learning_rate": 9.759180492611065e-05,
      "loss": 1.2766,
      "step": 600
    },
    {
      "epoch": 0.019245937845085975,
      "grad_norm": 0.559363842010498,
      "learning_rate": 9.742011206984756e-05,
      "loss": 1.1934,
      "step": 610
    },
    {
      "epoch": 0.01956144502287427,
      "grad_norm": 0.6068762540817261,
      "learning_rate": 9.724266939172302e-05,
      "loss": 1.0602,
      "step": 620
    },
    {
      "epoch": 0.019876952200662566,
      "grad_norm": 0.5704780220985413,
      "learning_rate": 9.705949840716047e-05,
      "loss": 1.3037,
      "step": 630
    },
    {
      "epoch": 0.02019245937845086,
      "grad_norm": 0.4779965579509735,
      "learning_rate": 9.687062132615658e-05,
      "loss": 1.1023,
      "step": 640
    },
    {
      "epoch": 0.020507966556239154,
      "grad_norm": 0.46647870540618896,
      "learning_rate": 9.667606105058828e-05,
      "loss": 1.1474,
      "step": 650
    },
    {
      "epoch": 0.02082347373402745,
      "grad_norm": 0.4536419212818146,
      "learning_rate": 9.647584117143576e-05,
      "loss": 1.2257,
      "step": 660
    },
    {
      "epoch": 0.021138980911815743,
      "grad_norm": 0.4657559096813202,
      "learning_rate": 9.626998596592209e-05,
      "loss": 1.0884,
      "step": 670
    },
    {
      "epoch": 0.02145448808960404,
      "grad_norm": 0.489685595035553,
      "learning_rate": 9.605852039456947e-05,
      "loss": 1.1375,
      "step": 680
    },
    {
      "epoch": 0.021769995267392334,
      "grad_norm": 0.9330468773841858,
      "learning_rate": 9.584147009817277e-05,
      "loss": 1.1861,
      "step": 690
    },
    {
      "epoch": 0.02208550244518063,
      "grad_norm": 0.610003650188446,
      "learning_rate": 9.561886139469045e-05,
      "loss": 1.2251,
      "step": 700
    },
    {
      "epoch": 0.022401009622968922,
      "grad_norm": 0.687454104423523,
      "learning_rate": 9.539072127605342e-05,
      "loss": 1.2597,
      "step": 710
    },
    {
      "epoch": 0.022716516800757217,
      "grad_norm": 0.39726585149765015,
      "learning_rate": 9.515707740489224e-05,
      "loss": 1.2133,
      "step": 720
    },
    {
      "epoch": 0.02303202397854551,
      "grad_norm": 0.4841748774051666,
      "learning_rate": 9.491795811118296e-05,
      "loss": 1.175,
      "step": 730
    },
    {
      "epoch": 0.023347531156333808,
      "grad_norm": 0.6921291947364807,
      "learning_rate": 9.467339238881199e-05,
      "loss": 1.0637,
      "step": 740
    },
    {
      "epoch": 0.023663038334122102,
      "grad_norm": 0.45699384808540344,
      "learning_rate": 9.442340989206047e-05,
      "loss": 1.0974,
      "step": 750
    },
    {
      "epoch": 0.023978545511910396,
      "grad_norm": 0.5954831838607788,
      "learning_rate": 9.416804093200874e-05,
      "loss": 1.2166,
      "step": 760
    },
    {
      "epoch": 0.02429405268969869,
      "grad_norm": 0.6534472703933716,
      "learning_rate": 9.390731647286089e-05,
      "loss": 1.25,
      "step": 770
    },
    {
      "epoch": 0.024609559867486985,
      "grad_norm": 0.4319377839565277,
      "learning_rate": 9.364126812819038e-05,
      "loss": 1.0963,
      "step": 780
    },
    {
      "epoch": 0.02492506704527528,
      "grad_norm": 0.3469117283821106,
      "learning_rate": 9.336992815710668e-05,
      "loss": 1.1501,
      "step": 790
    },
    {
      "epoch": 0.025240574223063576,
      "grad_norm": 0.5852107405662537,
      "learning_rate": 9.309332946034392e-05,
      "loss": 1.1365,
      "step": 800
    },
    {
      "epoch": 0.02555608140085187,
      "grad_norm": 0.364041268825531,
      "learning_rate": 9.281150557627143e-05,
      "loss": 1.1154,
      "step": 810
    },
    {
      "epoch": 0.025871588578640164,
      "grad_norm": 0.3952679932117462,
      "learning_rate": 9.252449067682721e-05,
      "loss": 1.1984,
      "step": 820
    },
    {
      "epoch": 0.02618709575642846,
      "grad_norm": 0.3946315050125122,
      "learning_rate": 9.22323195633745e-05,
      "loss": 1.1346,
      "step": 830
    },
    {
      "epoch": 0.026502602934216753,
      "grad_norm": 0.38489851355552673,
      "learning_rate": 9.19350276624819e-05,
      "loss": 1.1597,
      "step": 840
    },
    {
      "epoch": 0.026818110112005047,
      "grad_norm": 0.37699270248413086,
      "learning_rate": 9.163265102162794e-05,
      "loss": 1.1,
      "step": 850
    },
    {
      "epoch": 0.027133617289793344,
      "grad_norm": 0.6026850342750549,
      "learning_rate": 9.132522630483017e-05,
      "loss": 1.0557,
      "step": 860
    },
    {
      "epoch": 0.02744912446758164,
      "grad_norm": 0.7021874785423279,
      "learning_rate": 9.101279078819949e-05,
      "loss": 1.146,
      "step": 870
    },
    {
      "epoch": 0.027764631645369933,
      "grad_norm": 0.4263465404510498,
      "learning_rate": 9.069538235542037e-05,
      "loss": 1.1513,
      "step": 880
    },
    {
      "epoch": 0.028080138823158227,
      "grad_norm": 0.41765096783638,
      "learning_rate": 9.037303949315736e-05,
      "loss": 1.1684,
      "step": 890
    },
    {
      "epoch": 0.02839564600094652,
      "grad_norm": 0.44986680150032043,
      "learning_rate": 9.004580128638846e-05,
      "loss": 1.1054,
      "step": 900
    },
    {
      "epoch": 0.028711153178734815,
      "grad_norm": 0.39543870091438293,
      "learning_rate": 8.971370741366592e-05,
      "loss": 1.2186,
      "step": 910
    },
    {
      "epoch": 0.029026660356523112,
      "grad_norm": 0.37330368161201477,
      "learning_rate": 8.937679814230517e-05,
      "loss": 1.2054,
      "step": 920
    },
    {
      "epoch": 0.029342167534311406,
      "grad_norm": 0.8773049712181091,
      "learning_rate": 8.903511432350221e-05,
      "loss": 1.1217,
      "step": 930
    },
    {
      "epoch": 0.0296576747120997,
      "grad_norm": 1.5796482563018799,
      "learning_rate": 8.868869738738038e-05,
      "loss": 1.0987,
      "step": 940
    },
    {
      "epoch": 0.029973181889887995,
      "grad_norm": 0.36434420943260193,
      "learning_rate": 8.833758933796678e-05,
      "loss": 1.1457,
      "step": 950
    },
    {
      "epoch": 0.03028868906767629,
      "grad_norm": 0.5418674945831299,
      "learning_rate": 8.798183274809917e-05,
      "loss": 1.2404,
      "step": 960
    },
    {
      "epoch": 0.030604196245464583,
      "grad_norm": 0.4581955075263977,
      "learning_rate": 8.762147075426392e-05,
      "loss": 1.0919,
      "step": 970
    },
    {
      "epoch": 0.03091970342325288,
      "grad_norm": 0.5733634829521179,
      "learning_rate": 8.725654705136558e-05,
      "loss": 1.1757,
      "step": 980
    },
    {
      "epoch": 0.031235210601041175,
      "grad_norm": 0.4178631603717804,
      "learning_rate": 8.688710588742872e-05,
      "loss": 1.25,
      "step": 990
    },
    {
      "epoch": 0.031550717778829465,
      "grad_norm": 0.4076519012451172,
      "learning_rate": 8.651319205823278e-05,
      "loss": 1.1356,
      "step": 1000
    },
    {
      "epoch": 0.03186622495661776,
      "grad_norm": 0.7019752264022827,
      "learning_rate": 8.613485090188042e-05,
      "loss": 1.2086,
      "step": 1010
    },
    {
      "epoch": 0.03218173213440606,
      "grad_norm": 0.3069887161254883,
      "learning_rate": 8.575212829330019e-05,
      "loss": 1.1486,
      "step": 1020
    },
    {
      "epoch": 0.03249723931219435,
      "grad_norm": 0.5908852815628052,
      "learning_rate": 8.536507063868397e-05,
      "loss": 1.1095,
      "step": 1030
    },
    {
      "epoch": 0.03281274648998265,
      "grad_norm": 0.6867959499359131,
      "learning_rate": 8.497372486986024e-05,
      "loss": 1.1468,
      "step": 1040
    },
    {
      "epoch": 0.03312825366777094,
      "grad_norm": 0.5001075863838196,
      "learning_rate": 8.45781384386033e-05,
      "loss": 1.0803,
      "step": 1050
    },
    {
      "epoch": 0.03344376084555924,
      "grad_norm": 0.46148380637168884,
      "learning_rate": 8.417835931087975e-05,
      "loss": 1.1178,
      "step": 1060
    },
    {
      "epoch": 0.033759268023347534,
      "grad_norm": 0.5462920665740967,
      "learning_rate": 8.377443596103239e-05,
      "loss": 1.2744,
      "step": 1070
    },
    {
      "epoch": 0.034074775201135825,
      "grad_norm": 0.4107511043548584,
      "learning_rate": 8.336641736590258e-05,
      "loss": 1.0937,
      "step": 1080
    },
    {
      "epoch": 0.03439028237892412,
      "grad_norm": 0.3727000057697296,
      "learning_rate": 8.295435299889173e-05,
      "loss": 1.2042,
      "step": 1090
    },
    {
      "epoch": 0.03470578955671241,
      "grad_norm": 0.48867249488830566,
      "learning_rate": 8.253829282396245e-05,
      "loss": 1.161,
      "step": 1100
    },
    {
      "epoch": 0.03502129673450071,
      "grad_norm": 0.5484758019447327,
      "learning_rate": 8.211828728958027e-05,
      "loss": 1.1609,
      "step": 1110
    },
    {
      "epoch": 0.035336803912289,
      "grad_norm": 0.6911703944206238,
      "learning_rate": 8.16943873225966e-05,
      "loss": 1.0674,
      "step": 1120
    },
    {
      "epoch": 0.0356523110900773,
      "grad_norm": 0.563105583190918,
      "learning_rate": 8.126664432207385e-05,
      "loss": 1.2002,
      "step": 1130
    },
    {
      "epoch": 0.035967818267865596,
      "grad_norm": 0.6931987404823303,
      "learning_rate": 8.083511015305299e-05,
      "loss": 1.1209,
      "step": 1140
    },
    {
      "epoch": 0.03628332544565389,
      "grad_norm": 0.7918395400047302,
      "learning_rate": 8.039983714026478e-05,
      "loss": 1.1997,
      "step": 1150
    },
    {
      "epoch": 0.036598832623442185,
      "grad_norm": 0.4152407944202423,
      "learning_rate": 7.99608780617854e-05,
      "loss": 1.0481,
      "step": 1160
    },
    {
      "epoch": 0.036914339801230475,
      "grad_norm": 0.3731307089328766,
      "learning_rate": 7.951828614263679e-05,
      "loss": 1.1326,
      "step": 1170
    },
    {
      "epoch": 0.03722984697901877,
      "grad_norm": 0.5490486025810242,
      "learning_rate": 7.907211504833301e-05,
      "loss": 1.2251,
      "step": 1180
    },
    {
      "epoch": 0.03754535415680707,
      "grad_norm": 0.699494481086731,
      "learning_rate": 7.862241887837322e-05,
      "loss": 1.1822,
      "step": 1190
    },
    {
      "epoch": 0.03786086133459536,
      "grad_norm": 0.48040771484375,
      "learning_rate": 7.816925215968189e-05,
      "loss": 1.084,
      "step": 1200
    },
    {
      "epoch": 0.03817636851238366,
      "grad_norm": 0.5420158505439758,
      "learning_rate": 7.771266983999725e-05,
      "loss": 1.2134,
      "step": 1210
    },
    {
      "epoch": 0.03849187569017195,
      "grad_norm": 0.6118718385696411,
      "learning_rate": 7.72527272812088e-05,
      "loss": 1.184,
      "step": 1220
    },
    {
      "epoch": 0.03880738286796025,
      "grad_norm": 0.5210500955581665,
      "learning_rate": 7.678948025264443e-05,
      "loss": 1.1628,
      "step": 1230
    },
    {
      "epoch": 0.03912289004574854,
      "grad_norm": 0.4329388439655304,
      "learning_rate": 7.632298492430831e-05,
      "loss": 1.0391,
      "step": 1240
    },
    {
      "epoch": 0.039438397223536835,
      "grad_norm": 0.5016342997550964,
      "learning_rate": 7.585329786007006e-05,
      "loss": 1.2102,
      "step": 1250
    },
    {
      "epoch": 0.03975390440132513,
      "grad_norm": 0.4969483017921448,
      "learning_rate": 7.538047601080629e-05,
      "loss": 1.0538,
      "step": 1260
    },
    {
      "epoch": 0.04006941157911342,
      "grad_norm": 0.6484401226043701,
      "learning_rate": 7.490457670749503e-05,
      "loss": 1.1832,
      "step": 1270
    },
    {
      "epoch": 0.04038491875690172,
      "grad_norm": 0.6064450740814209,
      "learning_rate": 7.442565765426436e-05,
      "loss": 1.1387,
      "step": 1280
    },
    {
      "epoch": 0.04070042593469001,
      "grad_norm": 0.4169582724571228,
      "learning_rate": 7.394377692139543e-05,
      "loss": 1.058,
      "step": 1290
    },
    {
      "epoch": 0.04101593311247831,
      "grad_norm": 0.609321117401123,
      "learning_rate": 7.345899293828144e-05,
      "loss": 1.2175,
      "step": 1300
    },
    {
      "epoch": 0.041331440290266606,
      "grad_norm": 0.4401857554912567,
      "learning_rate": 7.297136448634282e-05,
      "loss": 1.1076,
      "step": 1310
    },
    {
      "epoch": 0.0416469474680549,
      "grad_norm": 0.6151861548423767,
      "learning_rate": 7.248095069189981e-05,
      "loss": 1.1631,
      "step": 1320
    },
    {
      "epoch": 0.041962454645843195,
      "grad_norm": 0.6283815503120422,
      "learning_rate": 7.19878110190033e-05,
      "loss": 1.1071,
      "step": 1330
    },
    {
      "epoch": 0.042277961823631485,
      "grad_norm": 0.38593342900276184,
      "learning_rate": 7.149200526222459e-05,
      "loss": 1.1593,
      "step": 1340
    },
    {
      "epoch": 0.04259346900141978,
      "grad_norm": 0.3702864348888397,
      "learning_rate": 7.099359353940509e-05,
      "loss": 1.1549,
      "step": 1350
    },
    {
      "epoch": 0.04290897617920808,
      "grad_norm": 0.7573534250259399,
      "learning_rate": 7.049263628436702e-05,
      "loss": 1.139,
      "step": 1360
    },
    {
      "epoch": 0.04322448335699637,
      "grad_norm": 0.4059164524078369,
      "learning_rate": 6.998919423958547e-05,
      "loss": 1.2294,
      "step": 1370
    },
    {
      "epoch": 0.04353999053478467,
      "grad_norm": 0.5392842888832092,
      "learning_rate": 6.948332844882326e-05,
      "loss": 1.1949,
      "step": 1380
    },
    {
      "epoch": 0.04385549771257296,
      "grad_norm": 0.40931907296180725,
      "learning_rate": 6.897510024972925e-05,
      "loss": 1.1734,
      "step": 1390
    },
    {
      "epoch": 0.04417100489036126,
      "grad_norm": 0.4209199845790863,
      "learning_rate": 6.846457126640096e-05,
      "loss": 1.0708,
      "step": 1400
    },
    {
      "epoch": 0.04448651206814955,
      "grad_norm": 0.600908637046814,
      "learning_rate": 6.795180340191243e-05,
      "loss": 1.0657,
      "step": 1410
    },
    {
      "epoch": 0.044802019245937845,
      "grad_norm": 0.31521522998809814,
      "learning_rate": 6.743685883080835e-05,
      "loss": 1.0177,
      "step": 1420
    },
    {
      "epoch": 0.04511752642372614,
      "grad_norm": 0.4605425000190735,
      "learning_rate": 6.691979999156521e-05,
      "loss": 1.1436,
      "step": 1430
    },
    {
      "epoch": 0.04543303360151443,
      "grad_norm": 0.43967700004577637,
      "learning_rate": 6.640068957902043e-05,
      "loss": 1.2336,
      "step": 1440
    },
    {
      "epoch": 0.04574854077930273,
      "grad_norm": 0.4047873616218567,
      "learning_rate": 6.587959053677051e-05,
      "loss": 1.2193,
      "step": 1450
    },
    {
      "epoch": 0.04606404795709102,
      "grad_norm": 0.6791156530380249,
      "learning_rate": 6.535656604953884e-05,
      "loss": 1.1857,
      "step": 1460
    },
    {
      "epoch": 0.04637955513487932,
      "grad_norm": 0.43025296926498413,
      "learning_rate": 6.483167953551442e-05,
      "loss": 1.2128,
      "step": 1470
    },
    {
      "epoch": 0.046695062312667616,
      "grad_norm": 0.9802103638648987,
      "learning_rate": 6.430499463866231e-05,
      "loss": 1.1581,
      "step": 1480
    },
    {
      "epoch": 0.04701056949045591,
      "grad_norm": 0.5605840086936951,
      "learning_rate": 6.377657522100644e-05,
      "loss": 1.1955,
      "step": 1490
    },
    {
      "epoch": 0.047326076668244205,
      "grad_norm": 0.7271468043327332,
      "learning_rate": 6.324648535488631e-05,
      "loss": 1.1685,
      "step": 1500
    },
    {
      "epoch": 0.047641583846032495,
      "grad_norm": 0.46086645126342773,
      "learning_rate": 6.271478931518802e-05,
      "loss": 1.1762,
      "step": 1510
    },
    {
      "epoch": 0.04795709102382079,
      "grad_norm": 0.41231945157051086,
      "learning_rate": 6.21815515715507e-05,
      "loss": 1.1528,
      "step": 1520
    },
    {
      "epoch": 0.048272598201609083,
      "grad_norm": 0.4973437190055847,
      "learning_rate": 6.164683678054938e-05,
      "loss": 1.141,
      "step": 1530
    },
    {
      "epoch": 0.04858810537939738,
      "grad_norm": 0.5368614792823792,
      "learning_rate": 6.11107097778553e-05,
      "loss": 1.112,
      "step": 1540
    },
    {
      "epoch": 0.04890361255718568,
      "grad_norm": 0.4287970960140228,
      "learning_rate": 6.057323557037431e-05,
      "loss": 1.1357,
      "step": 1550
    },
    {
      "epoch": 0.04921911973497397,
      "grad_norm": 0.8291701078414917,
      "learning_rate": 6.00344793283646e-05,
      "loss": 1.0729,
      "step": 1560
    },
    {
      "epoch": 0.04953462691276227,
      "grad_norm": 0.4257969856262207,
      "learning_rate": 5.9494506377534695e-05,
      "loss": 1.1132,
      "step": 1570
    },
    {
      "epoch": 0.04985013409055056,
      "grad_norm": 0.46945375204086304,
      "learning_rate": 5.895338219112246e-05,
      "loss": 1.155,
      "step": 1580
    },
    {
      "epoch": 0.050165641268338855,
      "grad_norm": 0.450214684009552,
      "learning_rate": 5.841117238195631e-05,
      "loss": 1.2993,
      "step": 1590
    },
    {
      "epoch": 0.05048114844612715,
      "grad_norm": 0.41964858770370483,
      "learning_rate": 5.786794269449947e-05,
      "loss": 1.1301,
      "step": 1600
    },
    {
      "epoch": 0.05079665562391544,
      "grad_norm": 0.4688185453414917,
      "learning_rate": 5.732375899687827e-05,
      "loss": 1.1262,
      "step": 1610
    },
    {
      "epoch": 0.05111216280170374,
      "grad_norm": 0.40953758358955383,
      "learning_rate": 5.677868727289548e-05,
      "loss": 1.1443,
      "step": 1620
    },
    {
      "epoch": 0.05142766997949203,
      "grad_norm": 0.7492470741271973,
      "learning_rate": 5.623279361402952e-05,
      "loss": 1.0674,
      "step": 1630
    },
    {
      "epoch": 0.05174317715728033,
      "grad_norm": 0.5046231746673584,
      "learning_rate": 5.5686144211420775e-05,
      "loss": 1.1484,
      "step": 1640
    },
    {
      "epoch": 0.05205868433506862,
      "grad_norm": 0.5525899529457092,
      "learning_rate": 5.513880534784562e-05,
      "loss": 1.1305,
      "step": 1650
    },
    {
      "epoch": 0.05237419151285692,
      "grad_norm": 0.44825589656829834,
      "learning_rate": 5.459084338967958e-05,
      "loss": 1.174,
      "step": 1660
    },
    {
      "epoch": 0.052689698690645215,
      "grad_norm": 0.8712518215179443,
      "learning_rate": 5.40423247788501e-05,
      "loss": 1.0862,
      "step": 1670
    },
    {
      "epoch": 0.053005205868433505,
      "grad_norm": 0.46037644147872925,
      "learning_rate": 5.349331602478032e-05,
      "loss": 1.2612,
      "step": 1680
    },
    {
      "epoch": 0.0533207130462218,
      "grad_norm": 0.3993786871433258,
      "learning_rate": 5.294388369632466e-05,
      "loss": 1.0599,
      "step": 1690
    },
    {
      "epoch": 0.053636220224010094,
      "grad_norm": 0.4509952664375305,
      "learning_rate": 5.2394094413697136e-05,
      "loss": 1.1323,
      "step": 1700
    },
    {
      "epoch": 0.05395172740179839,
      "grad_norm": 0.3986489474773407,
      "learning_rate": 5.1844014840393393e-05,
      "loss": 1.144,
      "step": 1710
    },
    {
      "epoch": 0.05426723457958669,
      "grad_norm": 0.48152726888656616,
      "learning_rate": 5.1293711675107705e-05,
      "loss": 1.1085,
      "step": 1720
    },
    {
      "epoch": 0.05458274175737498,
      "grad_norm": 0.4110780358314514,
      "learning_rate": 5.074325164364548e-05,
      "loss": 1.1274,
      "step": 1730
    },
    {
      "epoch": 0.05489824893516328,
      "grad_norm": 0.6647006273269653,
      "learning_rate": 5.019270149083256e-05,
      "loss": 1.1762,
      "step": 1740
    },
    {
      "epoch": 0.05521375611295157,
      "grad_norm": 0.4578000605106354,
      "learning_rate": 4.9642127972422295e-05,
      "loss": 1.1406,
      "step": 1750
    },
    {
      "epoch": 0.055529263290739865,
      "grad_norm": 0.6286119222640991,
      "learning_rate": 4.9091597847001175e-05,
      "loss": 0.958,
      "step": 1760
    },
    {
      "epoch": 0.055844770468528156,
      "grad_norm": 0.4872972369194031,
      "learning_rate": 4.854117786789413e-05,
      "loss": 1.1913,
      "step": 1770
    },
    {
      "epoch": 0.05616027764631645,
      "grad_norm": 0.4607793390750885,
      "learning_rate": 4.799093477507058e-05,
      "loss": 1.2621,
      "step": 1780
    },
    {
      "epoch": 0.05647578482410475,
      "grad_norm": 0.5443147420883179,
      "learning_rate": 4.7440935287051996e-05,
      "loss": 1.1725,
      "step": 1790
    },
    {
      "epoch": 0.05679129200189304,
      "grad_norm": 0.7406758069992065,
      "learning_rate": 4.689124609282202e-05,
      "loss": 1.1468,
      "step": 1800
    },
    {
      "epoch": 0.05710679917968134,
      "grad_norm": 0.6841813921928406,
      "learning_rate": 4.634193384374038e-05,
      "loss": 1.1332,
      "step": 1810
    },
    {
      "epoch": 0.05742230635746963,
      "grad_norm": 0.41650551557540894,
      "learning_rate": 4.579306514546107e-05,
      "loss": 1.1002,
      "step": 1820
    },
    {
      "epoch": 0.05773781353525793,
      "grad_norm": 0.5348483324050903,
      "learning_rate": 4.524470654985637e-05,
      "loss": 1.1698,
      "step": 1830
    },
    {
      "epoch": 0.058053320713046225,
      "grad_norm": 0.4644758403301239,
      "learning_rate": 4.4696924546947105e-05,
      "loss": 1.1359,
      "step": 1840
    },
    {
      "epoch": 0.058368827890834515,
      "grad_norm": 0.49113455414772034,
      "learning_rate": 4.414978555684069e-05,
      "loss": 1.1629,
      "step": 1850
    },
    {
      "epoch": 0.05868433506862281,
      "grad_norm": 0.5527147650718689,
      "learning_rate": 4.360335592167737e-05,
      "loss": 1.3023,
      "step": 1860
    },
    {
      "epoch": 0.058999842246411104,
      "grad_norm": 0.6101669073104858,
      "learning_rate": 4.305770189758613e-05,
      "loss": 1.0331,
      "step": 1870
    },
    {
      "epoch": 0.0593153494241994,
      "grad_norm": 0.8309857845306396,
      "learning_rate": 4.251288964665098e-05,
      "loss": 1.24,
      "step": 1880
    },
    {
      "epoch": 0.05963085660198769,
      "grad_norm": 0.5306587815284729,
      "learning_rate": 4.196898522888848e-05,
      "loss": 1.1996,
      "step": 1890
    },
    {
      "epoch": 0.05994636377977599,
      "grad_norm": 0.5127861499786377,
      "learning_rate": 4.142605459423795e-05,
      "loss": 1.1979,
      "step": 1900
    },
    {
      "epoch": 0.06026187095756429,
      "grad_norm": 0.4082748591899872,
      "learning_rate": 4.088416357456471e-05,
      "loss": 1.1738,
      "step": 1910
    },
    {
      "epoch": 0.06057737813535258,
      "grad_norm": 0.41917574405670166,
      "learning_rate": 4.0343377875677854e-05,
      "loss": 1.2224,
      "step": 1920
    },
    {
      "epoch": 0.060892885313140875,
      "grad_norm": 0.4594613015651703,
      "learning_rate": 3.9803763069363256e-05,
      "loss": 1.2116,
      "step": 1930
    },
    {
      "epoch": 0.061208392490929166,
      "grad_norm": 0.5822821259498596,
      "learning_rate": 3.926538458543275e-05,
      "loss": 1.0644,
      "step": 1940
    },
    {
      "epoch": 0.06152389966871746,
      "grad_norm": 0.5038664937019348,
      "learning_rate": 3.8728307703790615e-05,
      "loss": 1.0946,
      "step": 1950
    },
    {
      "epoch": 0.06183940684650576,
      "grad_norm": 0.4776921272277832,
      "learning_rate": 3.819259754651819e-05,
      "loss": 1.0555,
      "step": 1960
    },
    {
      "epoch": 0.06215491402429405,
      "grad_norm": 0.37600579857826233,
      "learning_rate": 3.765831906997765e-05,
      "loss": 1.1422,
      "step": 1970
    },
    {
      "epoch": 0.06247042120208235,
      "grad_norm": 0.4422357380390167,
      "learning_rate": 3.7125537056935823e-05,
      "loss": 1.1688,
      "step": 1980
    },
    {
      "epoch": 0.06278592837987064,
      "grad_norm": 0.4450276792049408,
      "learning_rate": 3.659431610870918e-05,
      "loss": 1.2122,
      "step": 1990
    },
    {
      "epoch": 0.06310143555765893,
      "grad_norm": 0.48132121562957764,
      "learning_rate": 3.606472063733067e-05,
      "loss": 1.1017,
      "step": 2000
    },
    {
      "epoch": 0.06341694273544723,
      "grad_norm": 0.5494374632835388,
      "learning_rate": 3.553681485773962e-05,
      "loss": 1.1585,
      "step": 2010
    },
    {
      "epoch": 0.06373244991323553,
      "grad_norm": 0.7255235314369202,
      "learning_rate": 3.5010662779995476e-05,
      "loss": 1.1971,
      "step": 2020
    },
    {
      "epoch": 0.06404795709102382,
      "grad_norm": 0.44452548027038574,
      "learning_rate": 3.4486328201516374e-05,
      "loss": 1.1138,
      "step": 2030
    },
    {
      "epoch": 0.06436346426881212,
      "grad_norm": 0.45132672786712646,
      "learning_rate": 3.396387469934362e-05,
      "loss": 1.1491,
      "step": 2040
    },
    {
      "epoch": 0.06467897144660041,
      "grad_norm": 0.5831072330474854,
      "learning_rate": 3.3443365622432664e-05,
      "loss": 1.2272,
      "step": 2050
    },
    {
      "epoch": 0.0649944786243887,
      "grad_norm": 0.45740213990211487,
      "learning_rate": 3.2924864083971975e-05,
      "loss": 1.0365,
      "step": 2060
    },
    {
      "epoch": 0.065309985802177,
      "grad_norm": 0.4249660074710846,
      "learning_rate": 3.240843295373031e-05,
      "loss": 1.0816,
      "step": 2070
    },
    {
      "epoch": 0.0656254929799653,
      "grad_norm": 0.43736574053764343,
      "learning_rate": 3.1894134850433625e-05,
      "loss": 1.0117,
      "step": 2080
    },
    {
      "epoch": 0.06594100015775359,
      "grad_norm": 0.41917362809181213,
      "learning_rate": 3.13820321341724e-05,
      "loss": 1.0575,
      "step": 2090
    },
    {
      "epoch": 0.06625650733554188,
      "grad_norm": 0.6465585231781006,
      "learning_rate": 3.0872186898840193e-05,
      "loss": 1.0389,
      "step": 2100
    },
    {
      "epoch": 0.06657201451333018,
      "grad_norm": 0.5854601860046387,
      "learning_rate": 3.036466096460472e-05,
      "loss": 1.0875,
      "step": 2110
    },
    {
      "epoch": 0.06688752169111847,
      "grad_norm": 0.7354947924613953,
      "learning_rate": 2.9859515870411902e-05,
      "loss": 1.1608,
      "step": 2120
    },
    {
      "epoch": 0.06720302886890676,
      "grad_norm": 0.520214855670929,
      "learning_rate": 2.9356812866524076e-05,
      "loss": 1.1178,
      "step": 2130
    },
    {
      "epoch": 0.06751853604669507,
      "grad_norm": 0.5355998277664185,
      "learning_rate": 2.88566129070933e-05,
      "loss": 1.1249,
      "step": 2140
    },
    {
      "epoch": 0.06783404322448336,
      "grad_norm": 0.49371784925460815,
      "learning_rate": 2.8358976642770486e-05,
      "loss": 1.1029,
      "step": 2150
    },
    {
      "epoch": 0.06814955040227165,
      "grad_norm": 0.40560880303382874,
      "learning_rate": 2.7863964413351253e-05,
      "loss": 1.1094,
      "step": 2160
    },
    {
      "epoch": 0.06846505758005994,
      "grad_norm": 0.4687036871910095,
      "learning_rate": 2.737163624045962e-05,
      "loss": 1.0718,
      "step": 2170
    },
    {
      "epoch": 0.06878056475784824,
      "grad_norm": 0.4389897584915161,
      "learning_rate": 2.688205182027026e-05,
      "loss": 1.0652,
      "step": 2180
    },
    {
      "epoch": 0.06909607193563654,
      "grad_norm": 0.46042123436927795,
      "learning_rate": 2.6395270516270077e-05,
      "loss": 1.0748,
      "step": 2190
    },
    {
      "epoch": 0.06941157911342483,
      "grad_norm": 0.5048350691795349,
      "learning_rate": 2.591135135206026e-05,
      "loss": 1.1513,
      "step": 2200
    },
    {
      "epoch": 0.06972708629121313,
      "grad_norm": 0.6022383570671082,
      "learning_rate": 2.543035300419951e-05,
      "loss": 1.0872,
      "step": 2210
    },
    {
      "epoch": 0.07004259346900142,
      "grad_norm": 0.9275935292243958,
      "learning_rate": 2.4952333795089338e-05,
      "loss": 1.2373,
      "step": 2220
    },
    {
      "epoch": 0.07035810064678971,
      "grad_norm": 0.4814078211784363,
      "learning_rate": 2.4477351685902293e-05,
      "loss": 1.2757,
      "step": 2230
    },
    {
      "epoch": 0.070673607824578,
      "grad_norm": 0.48762601613998413,
      "learning_rate": 2.4005464269554077e-05,
      "loss": 1.0695,
      "step": 2240
    },
    {
      "epoch": 0.07098911500236631,
      "grad_norm": 0.7237532138824463,
      "learning_rate": 2.3536728763720133e-05,
      "loss": 1.1084,
      "step": 2250
    },
    {
      "epoch": 0.0713046221801546,
      "grad_norm": 0.35012200474739075,
      "learning_rate": 2.3071202003897902e-05,
      "loss": 1.086,
      "step": 2260
    },
    {
      "epoch": 0.07162012935794289,
      "grad_norm": 0.5497077107429504,
      "learning_rate": 2.260894043651537e-05,
      "loss": 1.0629,
      "step": 2270
    },
    {
      "epoch": 0.07193563653573119,
      "grad_norm": 0.4760759770870209,
      "learning_rate": 2.2150000112086755e-05,
      "loss": 1.0076,
      "step": 2280
    },
    {
      "epoch": 0.07225114371351948,
      "grad_norm": 0.49443313479423523,
      "learning_rate": 2.1694436678416215e-05,
      "loss": 1.0282,
      "step": 2290
    },
    {
      "epoch": 0.07256665089130777,
      "grad_norm": 0.508001446723938,
      "learning_rate": 2.1242305373850468e-05,
      "loss": 1.211,
      "step": 2300
    },
    {
      "epoch": 0.07288215806909608,
      "grad_norm": 0.6970950961112976,
      "learning_rate": 2.079366102058088e-05,
      "loss": 1.2444,
      "step": 2310
    },
    {
      "epoch": 0.07319766524688437,
      "grad_norm": 0.8585352897644043,
      "learning_rate": 2.0348558017996162e-05,
      "loss": 1.0957,
      "step": 2320
    },
    {
      "epoch": 0.07351317242467266,
      "grad_norm": 0.417327880859375,
      "learning_rate": 1.9907050336086368e-05,
      "loss": 1.1682,
      "step": 2330
    },
    {
      "epoch": 0.07382867960246095,
      "grad_norm": 1.1333324909210205,
      "learning_rate": 1.946919150889876e-05,
      "loss": 1.08,
      "step": 2340
    },
    {
      "epoch": 0.07414418678024925,
      "grad_norm": 0.5783224105834961,
      "learning_rate": 1.9035034628046727e-05,
      "loss": 1.0406,
      "step": 2350
    },
    {
      "epoch": 0.07445969395803755,
      "grad_norm": 0.505717933177948,
      "learning_rate": 1.860463233627225e-05,
      "loss": 1.2299,
      "step": 2360
    },
    {
      "epoch": 0.07477520113582584,
      "grad_norm": 0.7536906599998474,
      "learning_rate": 1.817803682106285e-05,
      "loss": 1.1869,
      "step": 2370
    },
    {
      "epoch": 0.07509070831361414,
      "grad_norm": 0.6558195948600769,
      "learning_rate": 1.7755299808323657e-05,
      "loss": 1.2183,
      "step": 2380
    },
    {
      "epoch": 0.07540621549140243,
      "grad_norm": 0.48988932371139526,
      "learning_rate": 1.7336472556105593e-05,
      "loss": 1.0805,
      "step": 2390
    },
    {
      "epoch": 0.07572172266919072,
      "grad_norm": 0.46199411153793335,
      "learning_rate": 1.6921605848390075e-05,
      "loss": 1.0771,
      "step": 2400
    },
    {
      "epoch": 0.07603722984697901,
      "grad_norm": 0.5149940848350525,
      "learning_rate": 1.651074998893139e-05,
      "loss": 1.1492,
      "step": 2410
    },
    {
      "epoch": 0.07635273702476732,
      "grad_norm": 0.38356372714042664,
      "learning_rate": 1.6103954795157188e-05,
      "loss": 1.1467,
      "step": 2420
    },
    {
      "epoch": 0.07666824420255561,
      "grad_norm": 0.4538305997848511,
      "learning_rate": 1.570126959212801e-05,
      "loss": 0.9464,
      "step": 2430
    },
    {
      "epoch": 0.0769837513803439,
      "grad_norm": 0.5672346949577332,
      "learning_rate": 1.530274320655644e-05,
      "loss": 1.1144,
      "step": 2440
    },
    {
      "epoch": 0.0772992585581322,
      "grad_norm": 0.6441534161567688,
      "learning_rate": 1.4908423960886809e-05,
      "loss": 1.2172,
      "step": 2450
    },
    {
      "epoch": 0.0776147657359205,
      "grad_norm": 0.4857594072818756,
      "learning_rate": 1.451835966743586e-05,
      "loss": 1.1294,
      "step": 2460
    },
    {
      "epoch": 0.07793027291370878,
      "grad_norm": 0.5527583360671997,
      "learning_rate": 1.4132597622595423e-05,
      "loss": 1.0536,
      "step": 2470
    },
    {
      "epoch": 0.07824578009149707,
      "grad_norm": 0.8115904927253723,
      "learning_rate": 1.375118460109766e-05,
      "loss": 1.1606,
      "step": 2480
    },
    {
      "epoch": 0.07856128726928538,
      "grad_norm": 0.6380918622016907,
      "learning_rate": 1.3374166850343328e-05,
      "loss": 1.1524,
      "step": 2490
    },
    {
      "epoch": 0.07887679444707367,
      "grad_norm": 0.44737401604652405,
      "learning_rate": 1.3001590084794307e-05,
      "loss": 1.1865,
      "step": 2500
    },
    {
      "epoch": 0.07919230162486196,
      "grad_norm": 0.44319379329681396,
      "learning_rate": 1.2633499480430566e-05,
      "loss": 1.2385,
      "step": 2510
    },
    {
      "epoch": 0.07950780880265026,
      "grad_norm": 0.48288267850875854,
      "learning_rate": 1.226993966927238e-05,
      "loss": 1.0418,
      "step": 2520
    },
    {
      "epoch": 0.07982331598043856,
      "grad_norm": 0.4171326160430908,
      "learning_rate": 1.191095473396861e-05,
      "loss": 1.1814,
      "step": 2530
    },
    {
      "epoch": 0.08013882315822685,
      "grad_norm": 0.459581196308136,
      "learning_rate": 1.1556588202451612e-05,
      "loss": 1.0307,
      "step": 2540
    },
    {
      "epoch": 0.08045433033601515,
      "grad_norm": 0.5369367599487305,
      "learning_rate": 1.120688304265925e-05,
      "loss": 1.1631,
      "step": 2550
    },
    {
      "epoch": 0.08076983751380344,
      "grad_norm": 0.6328393220901489,
      "learning_rate": 1.0861881657324985e-05,
      "loss": 1.1548,
      "step": 2560
    },
    {
      "epoch": 0.08108534469159173,
      "grad_norm": 0.46232977509498596,
      "learning_rate": 1.0521625878836416e-05,
      "loss": 1.1935,
      "step": 2570
    },
    {
      "epoch": 0.08140085186938002,
      "grad_norm": 0.46396756172180176,
      "learning_rate": 1.0186156964162957e-05,
      "loss": 1.108,
      "step": 2580
    },
    {
      "epoch": 0.08171635904716833,
      "grad_norm": 0.4770895838737488,
      "learning_rate": 9.855515589853326e-06,
      "loss": 1.1103,
      "step": 2590
    },
    {
      "epoch": 0.08203186622495662,
      "grad_norm": 0.5198113322257996,
      "learning_rate": 9.52974184710344e-06,
      "loss": 1.1943,
      "step": 2600
    },
    {
      "epoch": 0.08234737340274491,
      "grad_norm": 0.45006701350212097,
      "learning_rate": 9.208875236895182e-06,
      "loss": 1.1236,
      "step": 2610
    },
    {
      "epoch": 0.08266288058053321,
      "grad_norm": 0.7262599468231201,
      "learning_rate": 8.892954665206804e-06,
      "loss": 1.2231,
      "step": 2620
    },
    {
      "epoch": 0.0829783877583215,
      "grad_norm": 0.5783856511116028,
      "learning_rate": 8.582018438295553e-06,
      "loss": 1.0805,
      "step": 2630
    },
    {
      "epoch": 0.0832938949361098,
      "grad_norm": 0.5482893586158752,
      "learning_rate": 8.276104258052786e-06,
      "loss": 1.1554,
      "step": 2640
    },
    {
      "epoch": 0.08360940211389808,
      "grad_norm": 0.4669418931007385,
      "learning_rate": 7.975249217432612e-06,
      "loss": 1.0977,
      "step": 2650
    },
    {
      "epoch": 0.08392490929168639,
      "grad_norm": 0.6652015447616577,
      "learning_rate": 7.67948979595428e-06,
      "loss": 1.0844,
      "step": 2660
    },
    {
      "epoch": 0.08424041646947468,
      "grad_norm": 0.5294978022575378,
      "learning_rate": 7.388861855278861e-06,
      "loss": 1.095,
      "step": 2670
    },
    {
      "epoch": 0.08455592364726297,
      "grad_norm": 0.45833665132522583,
      "learning_rate": 7.103400634860946e-06,
      "loss": 1.1586,
      "step": 2680
    },
    {
      "epoch": 0.08487143082505128,
      "grad_norm": 0.5748533606529236,
      "learning_rate": 6.823140747675816e-06,
      "loss": 1.3103,
      "step": 2690
    },
    {
      "epoch": 0.08518693800283957,
      "grad_norm": 0.546541690826416,
      "learning_rate": 6.548116176022456e-06,
      "loss": 1.1564,
      "step": 2700
    },
    {
      "epoch": 0.08550244518062786,
      "grad_norm": 0.7005789279937744,
      "learning_rate": 6.2783602674030935e-06,
      "loss": 1.1113,
      "step": 2710
    },
    {
      "epoch": 0.08581795235841616,
      "grad_norm": 0.5347318649291992,
      "learning_rate": 6.013905730479824e-06,
      "loss": 1.2208,
      "step": 2720
    },
    {
      "epoch": 0.08613345953620445,
      "grad_norm": 0.47552213072776794,
      "learning_rate": 5.754784631108484e-06,
      "loss": 1.1326,
      "step": 2730
    },
    {
      "epoch": 0.08644896671399274,
      "grad_norm": 0.47632184624671936,
      "learning_rate": 5.5010283884506135e-06,
      "loss": 1.0831,
      "step": 2740
    },
    {
      "epoch": 0.08676447389178103,
      "grad_norm": 0.5365753173828125,
      "learning_rate": 5.252667771163827e-06,
      "loss": 1.2979,
      "step": 2750
    },
    {
      "epoch": 0.08707998106956934,
      "grad_norm": 0.4148613512516022,
      "learning_rate": 5.009732893670971e-06,
      "loss": 1.0888,
      "step": 2760
    },
    {
      "epoch": 0.08739548824735763,
      "grad_norm": 0.44592395424842834,
      "learning_rate": 4.7722532125087095e-06,
      "loss": 1.0999,
      "step": 2770
    },
    {
      "epoch": 0.08771099542514592,
      "grad_norm": 0.7864736914634705,
      "learning_rate": 4.540257522755809e-06,
      "loss": 1.0966,
      "step": 2780
    },
    {
      "epoch": 0.08802650260293422,
      "grad_norm": 0.6768621206283569,
      "learning_rate": 4.313773954541672e-06,
      "loss": 1.1897,
      "step": 2790
    },
    {
      "epoch": 0.08834200978072251,
      "grad_norm": 0.5477901101112366,
      "learning_rate": 4.092829969635464e-06,
      "loss": 1.1737,
      "step": 2800
    },
    {
      "epoch": 0.0886575169585108,
      "grad_norm": 0.5500572323799133,
      "learning_rate": 3.877452358116324e-06,
      "loss": 1.1163,
      "step": 2810
    },
    {
      "epoch": 0.0889730241362991,
      "grad_norm": 0.7095085978507996,
      "learning_rate": 3.6676672351249495e-06,
      "loss": 1.088,
      "step": 2820
    },
    {
      "epoch": 0.0892885313140874,
      "grad_norm": 0.6136842966079712,
      "learning_rate": 3.463500037697104e-06,
      "loss": 1.1441,
      "step": 2830
    },
    {
      "epoch": 0.08960403849187569,
      "grad_norm": 0.48445186018943787,
      "learning_rate": 3.2649755216792867e-06,
      "loss": 1.1387,
      "step": 2840
    },
    {
      "epoch": 0.08991954566966398,
      "grad_norm": 0.5624507665634155,
      "learning_rate": 3.0721177587270224e-06,
      "loss": 1.0582,
      "step": 2850
    },
    {
      "epoch": 0.09023505284745229,
      "grad_norm": 0.533638060092926,
      "learning_rate": 2.884950133386055e-06,
      "loss": 1.151,
      "step": 2860
    },
    {
      "epoch": 0.09055056002524058,
      "grad_norm": 0.5064953565597534,
      "learning_rate": 2.7034953402570174e-06,
      "loss": 1.0153,
      "step": 2870
    },
    {
      "epoch": 0.09086606720302887,
      "grad_norm": 0.4341917932033539,
      "learning_rate": 2.5277753812435467e-06,
      "loss": 1.094,
      "step": 2880
    },
    {
      "epoch": 0.09118157438081716,
      "grad_norm": 0.3772113025188446,
      "learning_rate": 2.35781156288451e-06,
      "loss": 1.0971,
      "step": 2890
    },
    {
      "epoch": 0.09149708155860546,
      "grad_norm": 0.43354088068008423,
      "learning_rate": 2.1936244937705906e-06,
      "loss": 1.1798,
      "step": 2900
    },
    {
      "epoch": 0.09181258873639375,
      "grad_norm": 0.3896254599094391,
      "learning_rate": 2.0352340820453453e-06,
      "loss": 1.0737,
      "step": 2910
    },
    {
      "epoch": 0.09212809591418204,
      "grad_norm": 0.36763468384742737,
      "learning_rate": 1.8826595329913489e-06,
      "loss": 1.0351,
      "step": 2920
    },
    {
      "epoch": 0.09244360309197035,
      "grad_norm": 0.5166642069816589,
      "learning_rate": 1.7359193467014812e-06,
      "loss": 1.1765,
      "step": 2930
    },
    {
      "epoch": 0.09275911026975864,
      "grad_norm": 0.5475133061408997,
      "learning_rate": 1.5950313158357432e-06,
      "loss": 1.127,
      "step": 2940
    },
    {
      "epoch": 0.09307461744754693,
      "grad_norm": 0.46491655707359314,
      "learning_rate": 1.460012523463833e-06,
      "loss": 1.0714,
      "step": 2950
    },
    {
      "epoch": 0.09339012462533523,
      "grad_norm": 0.42786383628845215,
      "learning_rate": 1.3308793409937959e-06,
      "loss": 1.1625,
      "step": 2960
    },
    {
      "epoch": 0.09370563180312352,
      "grad_norm": 0.5214555263519287,
      "learning_rate": 1.2076474261869363e-06,
      "loss": 1.2349,
      "step": 2970
    },
    {
      "epoch": 0.09402113898091181,
      "grad_norm": 0.6581803560256958,
      "learning_rate": 1.0903317212592579e-06,
      "loss": 1.0605,
      "step": 2980
    },
    {
      "epoch": 0.0943366461587001,
      "grad_norm": 0.3951626121997833,
      "learning_rate": 9.78946451069701e-07,
      "loss": 1.2178,
      "step": 2990
    },
    {
      "epoch": 0.09465215333648841,
      "grad_norm": 0.651982307434082,
      "learning_rate": 8.735051213953182e-07,
      "loss": 1.2502,
      "step": 3000
    },
    {
      "epoch": 0.0949676605142767,
      "grad_norm": 0.5281954407691956,
      "learning_rate": 7.740205172936665e-07,
      "loss": 1.107,
      "step": 3010
    },
    {
      "epoch": 0.09528316769206499,
      "grad_norm": 0.5843939185142517,
      "learning_rate": 6.805047015526089e-07,
      "loss": 1.1559,
      "step": 3020
    },
    {
      "epoch": 0.0955986748698533,
      "grad_norm": 0.6147716641426086,
      "learning_rate": 5.929690132276123e-07,
      "loss": 1.1583,
      "step": 3030
    },
    {
      "epoch": 0.09591418204764159,
      "grad_norm": 0.6916910409927368,
      "learning_rate": 5.114240662669023e-07,
      "loss": 1.2058,
      "step": 3040
    },
    {
      "epoch": 0.09622968922542988,
      "grad_norm": 0.4254242479801178,
      "learning_rate": 4.3587974822449364e-07,
      "loss": 1.0994,
      "step": 3050
    },
    {
      "epoch": 0.09654519640321817,
      "grad_norm": 0.6165900230407715,
      "learning_rate": 3.663452190612482e-07,
      "loss": 1.0987,
      "step": 3060
    },
    {
      "epoch": 0.09686070358100647,
      "grad_norm": 0.7261757850646973,
      "learning_rate": 3.028289100342585e-07,
      "loss": 1.1478,
      "step": 3070
    },
    {
      "epoch": 0.09717621075879476,
      "grad_norm": 0.5742084980010986,
      "learning_rate": 2.4533852267450974e-07,
      "loss": 1.1684,
      "step": 3080
    },
    {
      "epoch": 0.09749171793658305,
      "grad_norm": 0.6034279465675354,
      "learning_rate": 1.938810278530323e-07,
      "loss": 1.1457,
      "step": 3090
    },
    {
      "epoch": 0.09780722511437136,
      "grad_norm": 0.5837968587875366,
      "learning_rate": 1.4846266493570017e-07,
      "loss": 1.1902,
      "step": 3100
    },
    {
      "epoch": 0.09812273229215965,
      "grad_norm": 0.5142175555229187,
      "learning_rate": 1.0908894102666378e-07,
      "loss": 1.1009,
      "step": 3110
    },
    {
      "epoch": 0.09843823946994794,
      "grad_norm": 0.4157280921936035,
      "learning_rate": 7.57646303006121e-08,
      "loss": 1.1106,
      "step": 3120
    },
    {
      "epoch": 0.09875374664773623,
      "grad_norm": 0.41369205713272095,
      "learning_rate": 4.8493773423863295e-08,
      "loss": 1.1904,
      "step": 3130
    },
    {
      "epoch": 0.09906925382552453,
      "grad_norm": 0.6153455972671509,
      "learning_rate": 2.727967706447343e-08,
      "loss": 1.114,
      "step": 3140
    },
    {
      "epoch": 0.09938476100331282,
      "grad_norm": 0.4912586808204651,
      "learning_rate": 1.2124913491240453e-08,
      "loss": 1.0558,
      "step": 3150
    },
    {
      "epoch": 0.09970026818110111,
      "grad_norm": 0.5294234752655029,
      "learning_rate": 3.0313202618370562e-09,
      "loss": 1.1303,
      "step": 3160
    },
    {
      "epoch": 0.10001577535888942,
      "grad_norm": 0.5471630096435547,
      "learning_rate": 0.0,
      "loss": 1.1321,
      "step": 3170
    },
    {
      "epoch": 0.10001577535888942,
      "step": 3170,
      "total_flos": 7.093787845892506e+17,
      "train_loss": 1.1559843716185176,
      "train_runtime": 9587.5735,
      "train_samples_per_second": 2.645,
      "train_steps_per_second": 0.331
    }
  ],
  "logging_steps": 10,
  "max_steps": 3170,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7.093787845892506e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}