{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 1624,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0006157635467980296,
      "grad_norm": 4.493983969665541,
      "learning_rate": 1.226993865030675e-07,
      "loss": 1.3473,
      "step": 1
    },
    {
      "epoch": 0.003078817733990148,
      "grad_norm": 4.4474323816099215,
      "learning_rate": 6.134969325153375e-07,
      "loss": 1.3469,
      "step": 5
    },
    {
      "epoch": 0.006157635467980296,
      "grad_norm": 3.6161265114493393,
      "learning_rate": 1.226993865030675e-06,
      "loss": 1.3111,
      "step": 10
    },
    {
      "epoch": 0.009236453201970444,
      "grad_norm": 3.4388426002995027,
      "learning_rate": 1.8404907975460124e-06,
      "loss": 1.1892,
      "step": 15
    },
    {
      "epoch": 0.012315270935960592,
      "grad_norm": 2.969738133217925,
      "learning_rate": 2.45398773006135e-06,
      "loss": 1.182,
      "step": 20
    },
    {
      "epoch": 0.01539408866995074,
      "grad_norm": 1.8184809758120228,
      "learning_rate": 3.0674846625766875e-06,
      "loss": 1.1923,
      "step": 25
    },
    {
      "epoch": 0.01847290640394089,
      "grad_norm": 1.7700574611447348,
      "learning_rate": 3.680981595092025e-06,
      "loss": 1.1143,
      "step": 30
    },
    {
      "epoch": 0.021551724137931036,
      "grad_norm": 1.4493848775199207,
      "learning_rate": 4.294478527607362e-06,
      "loss": 1.0562,
      "step": 35
    },
    {
      "epoch": 0.024630541871921183,
      "grad_norm": 1.6023969723287563,
      "learning_rate": 4.9079754601227e-06,
      "loss": 1.176,
      "step": 40
    },
    {
      "epoch": 0.02770935960591133,
      "grad_norm": 1.775925874797147,
      "learning_rate": 5.521472392638038e-06,
      "loss": 1.0501,
      "step": 45
    },
    {
      "epoch": 0.03078817733990148,
      "grad_norm": 1.7025367108699954,
      "learning_rate": 6.134969325153375e-06,
      "loss": 1.0873,
      "step": 50
    },
    {
      "epoch": 0.033866995073891626,
      "grad_norm": 1.6351772132021833,
      "learning_rate": 6.748466257668712e-06,
      "loss": 1.0488,
      "step": 55
    },
    {
      "epoch": 0.03694581280788178,
      "grad_norm": 1.8466156105694456,
      "learning_rate": 7.36196319018405e-06,
      "loss": 1.1915,
      "step": 60
    },
    {
      "epoch": 0.04002463054187192,
      "grad_norm": 1.7123950712488791,
      "learning_rate": 7.975460122699386e-06,
      "loss": 1.0874,
      "step": 65
    },
    {
      "epoch": 0.04310344827586207,
      "grad_norm": 1.729033172595946,
      "learning_rate": 8.588957055214725e-06,
      "loss": 1.126,
      "step": 70
    },
    {
      "epoch": 0.046182266009852216,
      "grad_norm": 1.6440833304631188,
      "learning_rate": 9.202453987730062e-06,
      "loss": 1.151,
      "step": 75
    },
    {
      "epoch": 0.04926108374384237,
      "grad_norm": 1.5971801295704615,
      "learning_rate": 9.8159509202454e-06,
      "loss": 1.1442,
      "step": 80
    },
    {
      "epoch": 0.05233990147783251,
      "grad_norm": 1.7549627000952002,
      "learning_rate": 1.0429447852760737e-05,
      "loss": 1.0829,
      "step": 85
    },
    {
      "epoch": 0.05541871921182266,
      "grad_norm": 1.6685588574422048,
      "learning_rate": 1.1042944785276076e-05,
      "loss": 1.1239,
      "step": 90
    },
    {
      "epoch": 0.058497536945812806,
      "grad_norm": 1.6105404978451945,
      "learning_rate": 1.1656441717791411e-05,
      "loss": 1.1161,
      "step": 95
    },
    {
      "epoch": 0.06157635467980296,
      "grad_norm": 1.5812947402510362,
      "learning_rate": 1.226993865030675e-05,
      "loss": 1.091,
      "step": 100
    },
    {
      "epoch": 0.06465517241379311,
      "grad_norm": 1.6644800851679993,
      "learning_rate": 1.2883435582822085e-05,
      "loss": 1.1222,
      "step": 105
    },
    {
      "epoch": 0.06773399014778325,
      "grad_norm": 1.8215702556323021,
      "learning_rate": 1.3496932515337424e-05,
      "loss": 1.1031,
      "step": 110
    },
    {
      "epoch": 0.0708128078817734,
      "grad_norm": 1.885303341592245,
      "learning_rate": 1.4110429447852763e-05,
      "loss": 1.1152,
      "step": 115
    },
    {
      "epoch": 0.07389162561576355,
      "grad_norm": 1.7379436755246627,
      "learning_rate": 1.47239263803681e-05,
      "loss": 1.0507,
      "step": 120
    },
    {
      "epoch": 0.0769704433497537,
      "grad_norm": 1.5189624722027886,
      "learning_rate": 1.5337423312883436e-05,
      "loss": 1.0961,
      "step": 125
    },
    {
      "epoch": 0.08004926108374384,
      "grad_norm": 1.73600600504045,
      "learning_rate": 1.5950920245398772e-05,
      "loss": 1.076,
      "step": 130
    },
    {
      "epoch": 0.08312807881773399,
      "grad_norm": 1.6622222602556214,
      "learning_rate": 1.656441717791411e-05,
      "loss": 1.1508,
      "step": 135
    },
    {
      "epoch": 0.08620689655172414,
      "grad_norm": 1.697419156661559,
      "learning_rate": 1.717791411042945e-05,
      "loss": 1.1383,
      "step": 140
    },
    {
      "epoch": 0.08928571428571429,
      "grad_norm": 1.7302277908525394,
      "learning_rate": 1.7791411042944788e-05,
      "loss": 1.1701,
      "step": 145
    },
    {
      "epoch": 0.09236453201970443,
      "grad_norm": 1.7733831911368454,
      "learning_rate": 1.8404907975460123e-05,
      "loss": 1.1614,
      "step": 150
    },
    {
      "epoch": 0.09544334975369458,
      "grad_norm": 1.6433040609724454,
      "learning_rate": 1.9018404907975462e-05,
      "loss": 1.0852,
      "step": 155
    },
    {
      "epoch": 0.09852216748768473,
      "grad_norm": 1.736230847219382,
      "learning_rate": 1.96319018404908e-05,
      "loss": 1.1285,
      "step": 160
    },
    {
      "epoch": 0.10160098522167488,
      "grad_norm": 1.8542590850595835,
      "learning_rate": 1.999990752408443e-05,
      "loss": 1.1887,
      "step": 165
    },
    {
      "epoch": 0.10467980295566502,
      "grad_norm": 1.8553635052067512,
      "learning_rate": 1.9998867189676517e-05,
      "loss": 1.1622,
      "step": 170
    },
    {
      "epoch": 0.10775862068965517,
      "grad_norm": 1.598964708297123,
      "learning_rate": 1.999667104662347e-05,
      "loss": 1.1521,
      "step": 175
    },
    {
      "epoch": 0.11083743842364532,
      "grad_norm": 1.614486786863126,
      "learning_rate": 1.9993319348786157e-05,
      "loss": 1.1895,
      "step": 180
    },
    {
      "epoch": 0.11391625615763547,
      "grad_norm": 1.6098251649149626,
      "learning_rate": 1.9988812483600597e-05,
      "loss": 1.1009,
      "step": 185
    },
    {
      "epoch": 0.11699507389162561,
      "grad_norm": 1.619405066121508,
      "learning_rate": 1.9983150972033186e-05,
      "loss": 1.1517,
      "step": 190
    },
    {
      "epoch": 0.12007389162561577,
      "grad_norm": 1.962687292010384,
      "learning_rate": 1.9976335468520452e-05,
      "loss": 1.1811,
      "step": 195
    },
    {
      "epoch": 0.12315270935960591,
      "grad_norm": 1.8705929596135908,
      "learning_rate": 1.9968366760893437e-05,
      "loss": 1.1795,
      "step": 200
    },
    {
      "epoch": 0.12623152709359606,
      "grad_norm": 2.3424285153646536,
      "learning_rate": 1.9959245770286602e-05,
      "loss": 1.1681,
      "step": 205
    },
    {
      "epoch": 0.12931034482758622,
      "grad_norm": 1.663367679614271,
      "learning_rate": 1.994897355103136e-05,
      "loss": 1.1487,
      "step": 210
    },
    {
      "epoch": 0.13238916256157635,
      "grad_norm": 1.5642792092477233,
      "learning_rate": 1.9937551290534208e-05,
      "loss": 1.0976,
      "step": 215
    },
    {
      "epoch": 0.1354679802955665,
      "grad_norm": 1.6347816583496555,
      "learning_rate": 1.9924980309139455e-05,
      "loss": 1.1371,
      "step": 220
    },
    {
      "epoch": 0.13854679802955666,
      "grad_norm": 1.5727985206849546,
      "learning_rate": 1.9911262059976614e-05,
      "loss": 1.1688,
      "step": 225
    },
    {
      "epoch": 0.1416256157635468,
      "grad_norm": 1.6289348339260017,
      "learning_rate": 1.9896398128792413e-05,
      "loss": 1.1559,
      "step": 230
    },
    {
      "epoch": 0.14470443349753695,
      "grad_norm": 1.5917015565995545,
      "learning_rate": 1.988039023376751e-05,
      "loss": 1.1911,
      "step": 235
    },
    {
      "epoch": 0.1477832512315271,
      "grad_norm": 1.6761734057281337,
      "learning_rate": 1.9863240225317868e-05,
      "loss": 1.2299,
      "step": 240
    },
    {
      "epoch": 0.15086206896551724,
      "grad_norm": 1.6727859028630858,
      "learning_rate": 1.984495008588086e-05,
      "loss": 1.1299,
      "step": 245
    },
    {
      "epoch": 0.1539408866995074,
      "grad_norm": 1.5322125785420966,
      "learning_rate": 1.982552192968612e-05,
      "loss": 1.2231,
      "step": 250
    },
    {
      "epoch": 0.15701970443349753,
      "grad_norm": 1.5909980815013838,
      "learning_rate": 1.9804958002511137e-05,
      "loss": 1.1664,
      "step": 255
    },
    {
      "epoch": 0.16009852216748768,
      "grad_norm": 1.7819228785267296,
      "learning_rate": 1.9783260681421667e-05,
      "loss": 1.185,
      "step": 260
    },
    {
      "epoch": 0.16317733990147784,
      "grad_norm": 1.4865664466326194,
      "learning_rate": 1.9760432474496963e-05,
      "loss": 1.127,
      "step": 265
    },
    {
      "epoch": 0.16625615763546797,
      "grad_norm": 1.497755678021967,
      "learning_rate": 1.973647602053984e-05,
      "loss": 1.1611,
      "step": 270
    },
    {
      "epoch": 0.16933497536945813,
      "grad_norm": 1.5668865592327674,
      "learning_rate": 1.9711394088771658e-05,
      "loss": 1.186,
      "step": 275
    },
    {
      "epoch": 0.1724137931034483,
      "grad_norm": 1.5192898195409086,
      "learning_rate": 1.9685189578512206e-05,
      "loss": 1.1809,
      "step": 280
    },
    {
      "epoch": 0.17549261083743842,
      "grad_norm": 1.6754130450725508,
      "learning_rate": 1.9657865518844578e-05,
      "loss": 1.1528,
      "step": 285
    },
    {
      "epoch": 0.17857142857142858,
      "grad_norm": 1.5004241611119362,
      "learning_rate": 1.962942506826501e-05,
      "loss": 1.1289,
      "step": 290
    },
    {
      "epoch": 0.1816502463054187,
      "grad_norm": 1.5035121947688075,
      "learning_rate": 1.9599871514317785e-05,
      "loss": 1.1685,
      "step": 295
    },
    {
      "epoch": 0.18472906403940886,
      "grad_norm": 1.8654655404134601,
      "learning_rate": 1.9569208273215204e-05,
      "loss": 1.1035,
      "step": 300
    },
    {
      "epoch": 0.18780788177339902,
      "grad_norm": 1.8887573081817965,
      "learning_rate": 1.953743888944271e-05,
      "loss": 1.112,
      "step": 305
    },
    {
      "epoch": 0.19088669950738915,
      "grad_norm": 1.9277821447823358,
      "learning_rate": 1.950456703534915e-05,
      "loss": 1.2013,
      "step": 310
    },
    {
      "epoch": 0.1939655172413793,
      "grad_norm": 1.6134100147163304,
      "learning_rate": 1.9470596510722285e-05,
      "loss": 1.1093,
      "step": 315
    },
    {
      "epoch": 0.19704433497536947,
      "grad_norm": 1.6183814211344283,
      "learning_rate": 1.9435531242349545e-05,
      "loss": 1.2547,
      "step": 320
    },
    {
      "epoch": 0.2001231527093596,
      "grad_norm": 1.595453910878586,
      "learning_rate": 1.9399375283564134e-05,
      "loss": 1.1685,
      "step": 325
    },
    {
      "epoch": 0.20320197044334976,
      "grad_norm": 1.6440747941155283,
      "learning_rate": 1.9362132813776472e-05,
      "loss": 1.1496,
      "step": 330
    },
    {
      "epoch": 0.2062807881773399,
      "grad_norm": 1.5430948828690219,
      "learning_rate": 1.9323808137991084e-05,
      "loss": 1.1606,
      "step": 335
    },
    {
      "epoch": 0.20935960591133004,
      "grad_norm": 1.623429730881881,
      "learning_rate": 1.9284405686308982e-05,
      "loss": 1.1311,
      "step": 340
    },
    {
      "epoch": 0.2124384236453202,
      "grad_norm": 1.5507598693908542,
      "learning_rate": 1.924393001341555e-05,
      "loss": 1.1577,
      "step": 345
    },
    {
      "epoch": 0.21551724137931033,
      "grad_norm": 1.5550781702028158,
      "learning_rate": 1.9202385798054073e-05,
      "loss": 1.1284,
      "step": 350
    },
    {
      "epoch": 0.2185960591133005,
      "grad_norm": 1.5691583160849498,
      "learning_rate": 1.9159777842484878e-05,
      "loss": 1.1535,
      "step": 355
    },
    {
      "epoch": 0.22167487684729065,
      "grad_norm": 1.5879992236192328,
      "learning_rate": 1.911611107193024e-05,
      "loss": 1.1601,
      "step": 360
    },
    {
      "epoch": 0.22475369458128078,
      "grad_norm": 1.7627218946087504,
      "learning_rate": 1.9071390534005045e-05,
      "loss": 1.2033,
      "step": 365
    },
    {
      "epoch": 0.22783251231527094,
      "grad_norm": 1.8361097903573524,
      "learning_rate": 1.9025621398133333e-05,
      "loss": 1.1883,
      "step": 370
    },
    {
      "epoch": 0.2309113300492611,
      "grad_norm": 1.8944857841813638,
      "learning_rate": 1.8978808954950722e-05,
      "loss": 1.1639,
      "step": 375
    },
    {
      "epoch": 0.23399014778325122,
      "grad_norm": 1.594449482443384,
      "learning_rate": 1.8930958615692854e-05,
      "loss": 1.2107,
      "step": 380
    },
    {
      "epoch": 0.23706896551724138,
      "grad_norm": 1.5501019352648453,
      "learning_rate": 1.8882075911569887e-05,
      "loss": 1.0791,
      "step": 385
    },
    {
      "epoch": 0.24014778325123154,
      "grad_norm": 1.5055724087289595,
      "learning_rate": 1.8832166493127128e-05,
      "loss": 1.1428,
      "step": 390
    },
    {
      "epoch": 0.24322660098522167,
      "grad_norm": 1.5638224613523255,
      "learning_rate": 1.8781236129591847e-05,
      "loss": 1.1173,
      "step": 395
    },
    {
      "epoch": 0.24630541871921183,
      "grad_norm": 1.674524072809073,
      "learning_rate": 1.8729290708206412e-05,
      "loss": 1.1833,
      "step": 400
    },
    {
      "epoch": 0.24938423645320196,
      "grad_norm": 1.5225291068444973,
      "learning_rate": 1.867633623354773e-05,
      "loss": 1.1953,
      "step": 405
    },
    {
      "epoch": 0.2524630541871921,
      "grad_norm": 1.5790532812554834,
      "learning_rate": 1.8622378826833186e-05,
      "loss": 1.1855,
      "step": 410
    },
    {
      "epoch": 0.2555418719211823,
      "grad_norm": 1.575820966011815,
      "learning_rate": 1.856742472521304e-05,
      "loss": 1.2015,
      "step": 415
    },
    {
      "epoch": 0.25862068965517243,
      "grad_norm": 1.702462273211937,
      "learning_rate": 1.8511480281049475e-05,
      "loss": 1.1835,
      "step": 420
    },
    {
      "epoch": 0.2616995073891626,
      "grad_norm": 1.514583411614364,
      "learning_rate": 1.8454551961182276e-05,
      "loss": 1.1429,
      "step": 425
    },
    {
      "epoch": 0.2647783251231527,
      "grad_norm": 1.4164788658237368,
      "learning_rate": 1.8396646346181327e-05,
      "loss": 1.1677,
      "step": 430
    },
    {
      "epoch": 0.26785714285714285,
      "grad_norm": 1.4757351202467979,
      "learning_rate": 1.8337770129585918e-05,
      "loss": 1.2154,
      "step": 435
    },
    {
      "epoch": 0.270935960591133,
      "grad_norm": 1.4867067667008775,
      "learning_rate": 1.8277930117131025e-05,
      "loss": 1.2043,
      "step": 440
    },
    {
      "epoch": 0.27401477832512317,
      "grad_norm": 1.5564768383648129,
      "learning_rate": 1.8217133225960597e-05,
      "loss": 1.1302,
      "step": 445
    },
    {
      "epoch": 0.2770935960591133,
      "grad_norm": 1.349915616217698,
      "learning_rate": 1.8155386483827995e-05,
      "loss": 1.165,
      "step": 450
    },
    {
      "epoch": 0.2801724137931034,
      "grad_norm": 1.7036931496180605,
      "learning_rate": 1.8092697028283598e-05,
      "loss": 1.1482,
      "step": 455
    },
    {
      "epoch": 0.2832512315270936,
      "grad_norm": 1.4702416496008561,
      "learning_rate": 1.8029072105849767e-05,
      "loss": 1.1427,
      "step": 460
    },
    {
      "epoch": 0.28633004926108374,
      "grad_norm": 1.5263944396321798,
      "learning_rate": 1.7964519071183188e-05,
      "loss": 1.1609,
      "step": 465
    },
    {
      "epoch": 0.2894088669950739,
      "grad_norm": 1.474253606234052,
      "learning_rate": 1.789904538622471e-05,
      "loss": 1.1558,
      "step": 470
    },
    {
      "epoch": 0.29248768472906406,
      "grad_norm": 1.4856503525131113,
      "learning_rate": 1.7832658619336794e-05,
      "loss": 1.1912,
      "step": 475
    },
    {
      "epoch": 0.2955665024630542,
      "grad_norm": 1.2824802514930713,
      "learning_rate": 1.7765366444428655e-05,
      "loss": 1.1649,
      "step": 480
    },
    {
      "epoch": 0.2986453201970443,
      "grad_norm": 1.4817304449242725,
      "learning_rate": 1.7697176640069217e-05,
      "loss": 1.1221,
      "step": 485
    },
    {
      "epoch": 0.3017241379310345,
      "grad_norm": 1.473031674973397,
      "learning_rate": 1.762809708858793e-05,
      "loss": 1.1696,
      "step": 490
    },
    {
      "epoch": 0.30480295566502463,
      "grad_norm": 1.4475363395188228,
      "learning_rate": 1.7558135775163645e-05,
      "loss": 1.1658,
      "step": 495
    },
    {
      "epoch": 0.3078817733990148,
      "grad_norm": 1.5052683410762635,
      "learning_rate": 1.7487300786901568e-05,
      "loss": 1.0746,
      "step": 500
    },
    {
      "epoch": 0.31096059113300495,
      "grad_norm": 1.4693302404038355,
      "learning_rate": 1.7415600311898436e-05,
      "loss": 1.2215,
      "step": 505
    },
    {
      "epoch": 0.31403940886699505,
      "grad_norm": 1.5571878160156238,
      "learning_rate": 1.734304263829602e-05,
      "loss": 1.1811,
      "step": 510
    },
    {
      "epoch": 0.3171182266009852,
      "grad_norm": 1.5063917088394518,
      "learning_rate": 1.726963615332308e-05,
      "loss": 1.212,
      "step": 515
    },
    {
      "epoch": 0.32019704433497537,
      "grad_norm": 1.4171304479384577,
      "learning_rate": 1.7195389342325843e-05,
      "loss": 1.1679,
      "step": 520
    },
    {
      "epoch": 0.3232758620689655,
      "grad_norm": 1.4144994910079824,
      "learning_rate": 1.7120310787787136e-05,
      "loss": 1.1365,
      "step": 525
    },
    {
      "epoch": 0.3263546798029557,
      "grad_norm": 1.5705003736463778,
      "learning_rate": 1.7044409168334327e-05,
      "loss": 1.1522,
      "step": 530
    },
    {
      "epoch": 0.3294334975369458,
      "grad_norm": 1.6141458336083276,
      "learning_rate": 1.696769325773611e-05,
      "loss": 1.1262,
      "step": 535
    },
    {
      "epoch": 0.33251231527093594,
      "grad_norm": 1.6899457322868179,
      "learning_rate": 1.6890171923888323e-05,
      "loss": 1.173,
      "step": 540
    },
    {
      "epoch": 0.3355911330049261,
      "grad_norm": 1.3568624754623695,
      "learning_rate": 1.6811854127788857e-05,
      "loss": 1.237,
      "step": 545
    },
    {
      "epoch": 0.33866995073891626,
      "grad_norm": 1.384883985302723,
      "learning_rate": 1.6732748922501832e-05,
      "loss": 1.234,
      "step": 550
    },
    {
      "epoch": 0.3417487684729064,
      "grad_norm": 1.4616266982830617,
      "learning_rate": 1.6652865452111115e-05,
      "loss": 1.1097,
      "step": 555
    },
    {
      "epoch": 0.3448275862068966,
      "grad_norm": 1.464277732440631,
      "learning_rate": 1.657221295066332e-05,
      "loss": 1.1815,
      "step": 560
    },
    {
      "epoch": 0.3479064039408867,
      "grad_norm": 1.459204222659814,
      "learning_rate": 1.6490800741100396e-05,
      "loss": 1.1754,
      "step": 565
    },
    {
      "epoch": 0.35098522167487683,
      "grad_norm": 1.4690595976889564,
      "learning_rate": 1.6408638234181975e-05,
      "loss": 1.1806,
      "step": 570
    },
    {
      "epoch": 0.354064039408867,
      "grad_norm": 1.5038369109700545,
      "learning_rate": 1.6325734927397514e-05,
      "loss": 1.1494,
      "step": 575
    },
    {
      "epoch": 0.35714285714285715,
      "grad_norm": 1.3868081069957796,
      "learning_rate": 1.6242100403868472e-05,
      "loss": 1.1521,
      "step": 580
    },
    {
      "epoch": 0.3602216748768473,
      "grad_norm": 1.461401339299667,
      "learning_rate": 1.615774433124054e-05,
      "loss": 1.1438,
      "step": 585
    },
    {
      "epoch": 0.3633004926108374,
      "grad_norm": 1.5110859463387252,
      "learning_rate": 1.6072676460566136e-05,
      "loss": 1.161,
      "step": 590
    },
    {
      "epoch": 0.36637931034482757,
      "grad_norm": 1.5971481491166415,
      "learning_rate": 1.5986906625177215e-05,
      "loss": 1.1723,
      "step": 595
    },
    {
      "epoch": 0.3694581280788177,
      "grad_norm": 1.4286268161082303,
      "learning_rate": 1.590044473954863e-05,
      "loss": 1.2608,
      "step": 600
    },
    {
      "epoch": 0.3725369458128079,
      "grad_norm": 1.4373833800913127,
      "learning_rate": 1.5813300798152048e-05,
      "loss": 1.1513,
      "step": 605
    },
    {
      "epoch": 0.37561576354679804,
      "grad_norm": 1.3341831212814368,
      "learning_rate": 1.5725484874300673e-05,
      "loss": 1.2113,
      "step": 610
    },
    {
      "epoch": 0.3786945812807882,
      "grad_norm": 1.5190206942826225,
      "learning_rate": 1.5637007118984814e-05,
      "loss": 1.1709,
      "step": 615
    },
    {
      "epoch": 0.3817733990147783,
      "grad_norm": 1.5185520660122134,
      "learning_rate": 1.5547877759698498e-05,
      "loss": 1.1738,
      "step": 620
    },
    {
      "epoch": 0.38485221674876846,
      "grad_norm": 1.4548778306729604,
      "learning_rate": 1.5458107099257245e-05,
      "loss": 1.1844,
      "step": 625
    },
    {
      "epoch": 0.3879310344827586,
      "grad_norm": 1.4961909017751673,
      "learning_rate": 1.5367705514607107e-05,
      "loss": 1.1339,
      "step": 630
    },
    {
      "epoch": 0.3910098522167488,
      "grad_norm": 1.5579155399686744,
      "learning_rate": 1.527668345562516e-05,
      "loss": 1.1429,
      "step": 635
    },
    {
      "epoch": 0.39408866995073893,
      "grad_norm": 1.5105103000174038,
      "learning_rate": 1.518505144391157e-05,
      "loss": 1.161,
      "step": 640
    },
    {
      "epoch": 0.39716748768472904,
      "grad_norm": 1.6549951176824607,
      "learning_rate": 1.5092820071573358e-05,
      "loss": 1.1872,
      "step": 645
    },
    {
      "epoch": 0.4002463054187192,
      "grad_norm": 1.5806601399836648,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 1.1554,
      "step": 650
    },
    {
      "epoch": 0.40332512315270935,
      "grad_norm": 1.4584931429544372,
      "learning_rate": 1.490660195863106e-05,
      "loss": 1.1265,
      "step": 655
    },
    {
      "epoch": 0.4064039408866995,
      "grad_norm": 1.458033700328426,
      "learning_rate": 1.4812636743715912e-05,
      "loss": 1.1269,
      "step": 660
    },
    {
      "epoch": 0.40948275862068967,
      "grad_norm": 1.4685161359205865,
      "learning_rate": 1.4718115217065766e-05,
      "loss": 1.1796,
      "step": 665
    },
    {
      "epoch": 0.4125615763546798,
      "grad_norm": 1.5238321795998173,
      "learning_rate": 1.462304830479811e-05,
      "loss": 1.1195,
      "step": 670
    },
    {
      "epoch": 0.41564039408866993,
      "grad_norm": 1.6614414988921782,
      "learning_rate": 1.4527446996073714e-05,
      "loss": 1.1401,
      "step": 675
    },
    {
      "epoch": 0.4187192118226601,
      "grad_norm": 1.4097240492231076,
      "learning_rate": 1.4431322341826348e-05,
      "loss": 1.1393,
      "step": 680
    },
    {
      "epoch": 0.42179802955665024,
      "grad_norm": 1.4228625711786493,
      "learning_rate": 1.433468545348537e-05,
      "loss": 1.1478,
      "step": 685
    },
    {
      "epoch": 0.4248768472906404,
      "grad_norm": 1.447499449584921,
      "learning_rate": 1.4237547501691298e-05,
      "loss": 1.1677,
      "step": 690
    },
    {
      "epoch": 0.42795566502463056,
      "grad_norm": 1.4987316333038885,
      "learning_rate": 1.4139919715004558e-05,
      "loss": 1.1561,
      "step": 695
    },
    {
      "epoch": 0.43103448275862066,
      "grad_norm": 1.4198772691620967,
      "learning_rate": 1.4041813378607534e-05,
      "loss": 1.1893,
      "step": 700
    },
    {
      "epoch": 0.4341133004926108,
      "grad_norm": 1.5954311132303836,
      "learning_rate": 1.3943239833000068e-05,
      "loss": 1.1346,
      "step": 705
    },
    {
      "epoch": 0.437192118226601,
      "grad_norm": 1.4365762676090503,
      "learning_rate": 1.3844210472688557e-05,
      "loss": 1.1279,
      "step": 710
    },
    {
      "epoch": 0.44027093596059114,
      "grad_norm": 1.363072566084107,
      "learning_rate": 1.3744736744868832e-05,
      "loss": 1.1251,
      "step": 715
    },
    {
      "epoch": 0.4433497536945813,
      "grad_norm": 1.5554328905759325,
      "learning_rate": 1.3644830148102915e-05,
      "loss": 1.195,
      "step": 720
    },
    {
      "epoch": 0.44642857142857145,
      "grad_norm": 1.4696304863700063,
      "learning_rate": 1.3544502230989868e-05,
      "loss": 1.164,
      "step": 725
    },
    {
      "epoch": 0.44950738916256155,
      "grad_norm": 1.4330533709390259,
      "learning_rate": 1.344376459083085e-05,
      "loss": 1.1547,
      "step": 730
    },
    {
      "epoch": 0.4525862068965517,
      "grad_norm": 1.6247690863462712,
      "learning_rate": 1.3342628872288505e-05,
      "loss": 1.1644,
      "step": 735
    },
    {
      "epoch": 0.45566502463054187,
      "grad_norm": 1.440406901583673,
      "learning_rate": 1.3241106766040957e-05,
      "loss": 1.1286,
      "step": 740
    },
    {
      "epoch": 0.45874384236453203,
      "grad_norm": 1.6021600392760522,
      "learning_rate": 1.3139210007430404e-05,
      "loss": 1.1378,
      "step": 745
    },
    {
      "epoch": 0.4618226600985222,
      "grad_norm": 1.3982682278709635,
      "learning_rate": 1.3036950375106588e-05,
      "loss": 1.1526,
      "step": 750
    },
    {
      "epoch": 0.4649014778325123,
      "grad_norm": 1.4721568383087724,
      "learning_rate": 1.2934339689665274e-05,
      "loss": 1.1229,
      "step": 755
    },
    {
      "epoch": 0.46798029556650245,
      "grad_norm": 1.5355186977371336,
      "learning_rate": 1.283138981228184e-05,
      "loss": 1.1524,
      "step": 760
    },
    {
      "epoch": 0.4710591133004926,
      "grad_norm": 1.5240333624023144,
      "learning_rate": 1.2728112643340201e-05,
      "loss": 1.1636,
      "step": 765
    },
    {
      "epoch": 0.47413793103448276,
      "grad_norm": 1.6451180501778482,
      "learning_rate": 1.26245201210572e-05,
      "loss": 1.1368,
      "step": 770
    },
    {
      "epoch": 0.4772167487684729,
      "grad_norm": 1.6989688372646368,
      "learning_rate": 1.2520624220102623e-05,
      "loss": 1.167,
      "step": 775
    },
    {
      "epoch": 0.4802955665024631,
      "grad_norm": 1.6795326748629802,
      "learning_rate": 1.2416436950215001e-05,
      "loss": 1.164,
      "step": 780
    },
    {
      "epoch": 0.4833743842364532,
      "grad_norm": 1.488500008065555,
      "learning_rate": 1.2311970354813345e-05,
      "loss": 1.2394,
      "step": 785
    },
    {
      "epoch": 0.48645320197044334,
      "grad_norm": 1.3975868463834171,
      "learning_rate": 1.220723650960502e-05,
      "loss": 1.1393,
      "step": 790
    },
    {
      "epoch": 0.4895320197044335,
      "grad_norm": 1.4325386576318664,
      "learning_rate": 1.2102247521189838e-05,
      "loss": 1.1086,
      "step": 795
    },
    {
      "epoch": 0.49261083743842365,
      "grad_norm": 1.4566996899637203,
      "learning_rate": 1.199701552566064e-05,
      "loss": 1.1061,
      "step": 800
    },
    {
      "epoch": 0.4956896551724138,
      "grad_norm": 1.5040441523115922,
      "learning_rate": 1.1891552687200414e-05,
      "loss": 1.194,
      "step": 805
    },
    {
      "epoch": 0.4987684729064039,
      "grad_norm": 1.5136962238950937,
      "learning_rate": 1.1785871196676196e-05,
      "loss": 1.2158,
      "step": 810
    },
    {
      "epoch": 0.5018472906403941,
      "grad_norm": 1.43203711840622,
      "learning_rate": 1.167998327022988e-05,
      "loss": 1.2253,
      "step": 815
    },
    {
      "epoch": 0.5049261083743842,
      "grad_norm": 1.42167427549154,
      "learning_rate": 1.1573901147866108e-05,
      "loss": 1.156,
      "step": 820
    },
    {
      "epoch": 0.5080049261083743,
      "grad_norm": 1.5113417816861492,
      "learning_rate": 1.1467637092037399e-05,
      "loss": 1.1596,
      "step": 825
    },
    {
      "epoch": 0.5110837438423645,
      "grad_norm": 1.5186319704514581,
      "learning_rate": 1.1361203386226672e-05,
      "loss": 1.1611,
      "step": 830
    },
    {
      "epoch": 0.5141625615763546,
      "grad_norm": 1.4651361095424547,
      "learning_rate": 1.1254612333527368e-05,
      "loss": 1.1309,
      "step": 835
    },
    {
      "epoch": 0.5172413793103449,
      "grad_norm": 1.425549320162281,
      "learning_rate": 1.1147876255221274e-05,
      "loss": 1.1368,
      "step": 840
    },
    {
      "epoch": 0.520320197044335,
      "grad_norm": 1.411434259697471,
      "learning_rate": 1.1041007489354263e-05,
      "loss": 1.1414,
      "step": 845
    },
    {
      "epoch": 0.5233990147783252,
      "grad_norm": 1.431239256844576,
      "learning_rate": 1.093401838931009e-05,
      "loss": 1.1497,
      "step": 850
    },
    {
      "epoch": 0.5264778325123153,
      "grad_norm": 1.6424648675223965,
      "learning_rate": 1.0826921322382407e-05,
      "loss": 1.1465,
      "step": 855
    },
    {
      "epoch": 0.5295566502463054,
      "grad_norm": 1.5189024926741261,
      "learning_rate": 1.071972866834519e-05,
      "loss": 1.0773,
      "step": 860
    },
    {
      "epoch": 0.5326354679802956,
      "grad_norm": 1.529584232497628,
      "learning_rate": 1.061245281802171e-05,
      "loss": 1.1686,
      "step": 865
    },
    {
      "epoch": 0.5357142857142857,
      "grad_norm": 1.7480556642680698,
      "learning_rate": 1.0505106171852226e-05,
      "loss": 1.0471,
      "step": 870
    },
    {
      "epoch": 0.5387931034482759,
      "grad_norm": 1.6628702754718676,
      "learning_rate": 1.039770113846056e-05,
      "loss": 1.096,
      "step": 875
    },
    {
      "epoch": 0.541871921182266,
      "grad_norm": 1.49343078733537,
      "learning_rate": 1.0290250133219755e-05,
      "loss": 1.1429,
      "step": 880
    },
    {
      "epoch": 0.5449507389162561,
      "grad_norm": 1.5415718025720817,
      "learning_rate": 1.0182765576816916e-05,
      "loss": 1.2058,
      "step": 885
    },
    {
      "epoch": 0.5480295566502463,
      "grad_norm": 1.4455552338300535,
      "learning_rate": 1.0075259893817465e-05,
      "loss": 1.1419,
      "step": 890
    },
    {
      "epoch": 0.5511083743842364,
      "grad_norm": 1.4061212767046805,
      "learning_rate": 9.967745511228922e-06,
      "loss": 1.1221,
      "step": 895
    },
    {
      "epoch": 0.5541871921182266,
      "grad_norm": 1.4302377392819847,
      "learning_rate": 9.860234857064439e-06,
      "loss": 1.1501,
      "step": 900
    },
    {
      "epoch": 0.5572660098522167,
      "grad_norm": 1.4222398778968683,
      "learning_rate": 9.752740358906168e-06,
      "loss": 1.1561,
      "step": 905
    },
    {
      "epoch": 0.5603448275862069,
      "grad_norm": 1.500530584874062,
      "learning_rate": 9.645274442468746e-06,
      "loss": 1.1471,
      "step": 910
    },
    {
      "epoch": 0.5634236453201971,
      "grad_norm": 1.4937446987241654,
      "learning_rate": 9.537849530162915e-06,
      "loss": 1.1548,
      "step": 915
    },
    {
      "epoch": 0.5665024630541872,
      "grad_norm": 1.4491765655122129,
      "learning_rate": 9.430478039659604e-06,
      "loss": 1.1309,
      "step": 920
    },
    {
      "epoch": 0.5695812807881774,
      "grad_norm": 1.3759912913105639,
      "learning_rate": 9.323172382454494e-06,
      "loss": 1.1566,
      "step": 925
    },
    {
      "epoch": 0.5726600985221675,
      "grad_norm": 1.5923808044967127,
      "learning_rate": 9.215944962433329e-06,
      "loss": 1.0795,
      "step": 930
    },
    {
      "epoch": 0.5757389162561576,
      "grad_norm": 1.5033947807003292,
      "learning_rate": 9.108808174438128e-06,
      "loss": 1.1606,
      "step": 935
    },
    {
      "epoch": 0.5788177339901478,
      "grad_norm": 1.4510474090120615,
      "learning_rate": 9.001774402834375e-06,
      "loss": 1.1091,
      "step": 940
    },
    {
      "epoch": 0.5818965517241379,
      "grad_norm": 1.5670716848759052,
      "learning_rate": 8.89485602007949e-06,
      "loss": 1.1601,
      "step": 945
    },
    {
      "epoch": 0.5849753694581281,
      "grad_norm": 1.532125612212799,
      "learning_rate": 8.788065385292637e-06,
      "loss": 1.1274,
      "step": 950
    },
    {
      "epoch": 0.5880541871921182,
      "grad_norm": 1.499447401956665,
      "learning_rate": 8.68141484282609e-06,
      "loss": 1.1624,
      "step": 955
    },
    {
      "epoch": 0.5911330049261084,
      "grad_norm": 1.5923150835068045,
      "learning_rate": 8.574916720838293e-06,
      "loss": 1.1827,
      "step": 960
    },
    {
      "epoch": 0.5942118226600985,
      "grad_norm": 1.4580558578441905,
      "learning_rate": 8.468583329868805e-06,
      "loss": 1.1696,
      "step": 965
    },
    {
      "epoch": 0.5972906403940886,
      "grad_norm": 1.5492361872355094,
      "learning_rate": 8.362426961415274e-06,
      "loss": 1.15,
      "step": 970
    },
    {
      "epoch": 0.6003694581280788,
      "grad_norm": 1.5036727692923915,
      "learning_rate": 8.256459886512618e-06,
      "loss": 1.1198,
      "step": 975
    },
    {
      "epoch": 0.603448275862069,
      "grad_norm": 1.64406664867369,
      "learning_rate": 8.150694354314556e-06,
      "loss": 1.199,
      "step": 980
    },
    {
      "epoch": 0.6065270935960592,
      "grad_norm": 1.585130722013942,
      "learning_rate": 8.0451425906777e-06,
      "loss": 1.157,
      "step": 985
    },
    {
      "epoch": 0.6096059113300493,
      "grad_norm": 1.404711489070082,
      "learning_rate": 7.939816796748296e-06,
      "loss": 1.1096,
      "step": 990
    },
    {
      "epoch": 0.6126847290640394,
      "grad_norm": 1.49206869163201,
      "learning_rate": 7.834729147551858e-06,
      "loss": 1.1409,
      "step": 995
    },
    {
      "epoch": 0.6157635467980296,
      "grad_norm": 1.4929755483685077,
      "learning_rate": 7.729891790585817e-06,
      "loss": 1.0844,
      "step": 1000
    },
    {
      "epoch": 0.6188423645320197,
      "grad_norm": 1.556778260441799,
      "learning_rate": 7.625316844415327e-06,
      "loss": 1.0761,
      "step": 1005
    },
    {
      "epoch": 0.6219211822660099,
      "grad_norm": 1.424739949173979,
      "learning_rate": 7.521016397272436e-06,
      "loss": 1.1343,
      "step": 1010
    },
    {
      "epoch": 0.625,
      "grad_norm": 1.6016301411568759,
      "learning_rate": 7.417002505658773e-06,
      "loss": 1.1936,
      "step": 1015
    },
    {
      "epoch": 0.6280788177339901,
      "grad_norm": 1.4544270434347801,
      "learning_rate": 7.313287192951866e-06,
      "loss": 1.1491,
      "step": 1020
    },
    {
      "epoch": 0.6311576354679803,
      "grad_norm": 1.5188184689956181,
      "learning_rate": 7.209882448015338e-06,
      "loss": 1.119,
      "step": 1025
    },
    {
      "epoch": 0.6342364532019704,
      "grad_norm": 1.496611392851464,
      "learning_rate": 7.1068002238130465e-06,
      "loss": 1.1512,
      "step": 1030
    },
    {
      "epoch": 0.6373152709359606,
      "grad_norm": 1.5873376562589607,
      "learning_rate": 7.004052436027397e-06,
      "loss": 1.1361,
      "step": 1035
    },
    {
      "epoch": 0.6403940886699507,
      "grad_norm": 1.4613901474576816,
      "learning_rate": 6.901650961681976e-06,
      "loss": 1.1297,
      "step": 1040
    },
    {
      "epoch": 0.6434729064039408,
      "grad_norm": 1.4155799553520878,
      "learning_rate": 6.799607637768621e-06,
      "loss": 1.1453,
      "step": 1045
    },
    {
      "epoch": 0.646551724137931,
      "grad_norm": 1.3952525842081505,
      "learning_rate": 6.6979342598791395e-06,
      "loss": 1.1517,
      "step": 1050
    },
    {
      "epoch": 0.6496305418719212,
      "grad_norm": 1.5075362520782292,
      "learning_rate": 6.596642580841827e-06,
      "loss": 1.1388,
      "step": 1055
    },
    {
      "epoch": 0.6527093596059114,
      "grad_norm": 1.562261894263839,
      "learning_rate": 6.495744309362879e-06,
      "loss": 1.0782,
      "step": 1060
    },
    {
      "epoch": 0.6557881773399015,
      "grad_norm": 1.5347470540813357,
      "learning_rate": 6.395251108672975e-06,
      "loss": 1.1215,
      "step": 1065
    },
    {
      "epoch": 0.6588669950738916,
      "grad_norm": 1.5421244523828281,
      "learning_rate": 6.295174595179041e-06,
      "loss": 1.1457,
      "step": 1070
    },
    {
      "epoch": 0.6619458128078818,
      "grad_norm": 1.5932458205065625,
      "learning_rate": 6.195526337121483e-06,
      "loss": 1.1684,
      "step": 1075
    },
    {
      "epoch": 0.6650246305418719,
      "grad_norm": 1.4273460422681852,
      "learning_rate": 6.096317853236975e-06,
      "loss": 1.0596,
      "step": 1080
    },
    {
      "epoch": 0.6681034482758621,
      "grad_norm": 1.4842226246708523,
      "learning_rate": 5.997560611426947e-06,
      "loss": 1.1235,
      "step": 1085
    },
    {
      "epoch": 0.6711822660098522,
      "grad_norm": 1.650992676115009,
      "learning_rate": 5.899266027431965e-06,
      "loss": 1.1626,
      "step": 1090
    },
    {
      "epoch": 0.6742610837438424,
      "grad_norm": 1.3840593858951986,
      "learning_rate": 5.80144546351216e-06,
      "loss": 1.0951,
      "step": 1095
    },
    {
      "epoch": 0.6773399014778325,
      "grad_norm": 1.4551292638519857,
      "learning_rate": 5.704110227133792e-06,
      "loss": 1.1375,
      "step": 1100
    },
    {
      "epoch": 0.6804187192118226,
      "grad_norm": 1.4453579425374732,
      "learning_rate": 5.607271569662203e-06,
      "loss": 1.1044,
      "step": 1105
    },
    {
      "epoch": 0.6834975369458128,
      "grad_norm": 1.545596299102939,
      "learning_rate": 5.510940685061202e-06,
      "loss": 1.1272,
      "step": 1110
    },
    {
      "epoch": 0.6865763546798029,
      "grad_norm": 1.6201552405250887,
      "learning_rate": 5.41512870859912e-06,
      "loss": 1.1381,
      "step": 1115
    },
    {
      "epoch": 0.6896551724137931,
      "grad_norm": 1.6253979905239055,
      "learning_rate": 5.319846715561656e-06,
      "loss": 1.1521,
      "step": 1120
    },
    {
      "epoch": 0.6927339901477833,
      "grad_norm": 1.4123662793473868,
      "learning_rate": 5.225105719971615e-06,
      "loss": 1.0866,
      "step": 1125
    },
    {
      "epoch": 0.6958128078817734,
      "grad_norm": 1.4441517484799031,
      "learning_rate": 5.130916673315762e-06,
      "loss": 1.1759,
      "step": 1130
    },
    {
      "epoch": 0.6988916256157636,
      "grad_norm": 1.5221193825774655,
      "learning_rate": 5.037290463278914e-06,
      "loss": 1.1203,
      "step": 1135
    },
    {
      "epoch": 0.7019704433497537,
      "grad_norm": 1.4677457356736134,
      "learning_rate": 4.94423791248536e-06,
      "loss": 1.1251,
      "step": 1140
    },
    {
      "epoch": 0.7050492610837439,
      "grad_norm": 1.6141381215936579,
      "learning_rate": 4.851769777247857e-06,
      "loss": 1.217,
      "step": 1145
    },
    {
      "epoch": 0.708128078817734,
      "grad_norm": 1.6103071927467703,
      "learning_rate": 4.759896746324247e-06,
      "loss": 1.1222,
      "step": 1150
    },
    {
      "epoch": 0.7112068965517241,
      "grad_norm": 1.5166398739531457,
      "learning_rate": 4.668629439681907e-06,
      "loss": 1.1379,
      "step": 1155
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 1.420621169931162,
      "learning_rate": 4.577978407270156e-06,
      "loss": 1.1476,
      "step": 1160
    },
    {
      "epoch": 0.7173645320197044,
      "grad_norm": 1.3864755314226962,
      "learning_rate": 4.487954127800726e-06,
      "loss": 1.0983,
      "step": 1165
    },
    {
      "epoch": 0.7204433497536946,
      "grad_norm": 1.4166642239072755,
      "learning_rate": 4.398567007536493e-06,
      "loss": 1.2139,
      "step": 1170
    },
    {
      "epoch": 0.7235221674876847,
      "grad_norm": 1.4408363764592564,
      "learning_rate": 4.309827379088589e-06,
      "loss": 1.1398,
      "step": 1175
    },
    {
      "epoch": 0.7266009852216748,
      "grad_norm": 1.460803505844183,
      "learning_rate": 4.221745500221992e-06,
      "loss": 1.07,
      "step": 1180
    },
    {
      "epoch": 0.729679802955665,
      "grad_norm": 1.5481395153546555,
      "learning_rate": 4.134331552669812e-06,
      "loss": 1.1665,
      "step": 1185
    },
    {
      "epoch": 0.7327586206896551,
      "grad_norm": 1.4696030221624719,
      "learning_rate": 4.047595640956326e-06,
      "loss": 1.1585,
      "step": 1190
    },
    {
      "epoch": 0.7358374384236454,
      "grad_norm": 1.4757749787371979,
      "learning_rate": 3.961547791228963e-06,
      "loss": 1.0988,
      "step": 1195
    },
    {
      "epoch": 0.7389162561576355,
      "grad_norm": 1.3709822252345534,
      "learning_rate": 3.876197950099351e-06,
      "loss": 1.1042,
      "step": 1200
    },
    {
      "epoch": 0.7419950738916257,
      "grad_norm": 1.4662944463259866,
      "learning_rate": 3.7915559834935355e-06,
      "loss": 1.1414,
      "step": 1205
    },
    {
      "epoch": 0.7450738916256158,
      "grad_norm": 1.4183955796490233,
      "learning_rate": 3.7076316755115407e-06,
      "loss": 1.1067,
      "step": 1210
    },
    {
      "epoch": 0.7481527093596059,
      "grad_norm": 1.5601667292176937,
      "learning_rate": 3.6244347272963974e-06,
      "loss": 1.1339,
      "step": 1215
    },
    {
      "epoch": 0.7512315270935961,
      "grad_norm": 1.5397349254314066,
      "learning_rate": 3.5419747559127294e-06,
      "loss": 1.154,
      "step": 1220
    },
    {
      "epoch": 0.7543103448275862,
      "grad_norm": 1.4851545513675735,
      "learning_rate": 3.4602612932351065e-06,
      "loss": 1.1691,
      "step": 1225
    },
    {
      "epoch": 0.7573891625615764,
      "grad_norm": 1.5776045618519325,
      "learning_rate": 3.3793037848461873e-06,
      "loss": 1.1247,
      "step": 1230
    },
    {
      "epoch": 0.7604679802955665,
      "grad_norm": 1.456176716263657,
      "learning_rate": 3.2991115889448877e-06,
      "loss": 1.1317,
      "step": 1235
    },
    {
      "epoch": 0.7635467980295566,
      "grad_norm": 1.5554417996449454,
      "learning_rate": 3.2196939752646183e-06,
      "loss": 1.1242,
      "step": 1240
    },
    {
      "epoch": 0.7666256157635468,
      "grad_norm": 1.52821516403833,
      "learning_rate": 3.141060124001776e-06,
      "loss": 1.1358,
      "step": 1245
    },
    {
      "epoch": 0.7697044334975369,
      "grad_norm": 1.5409063043253681,
      "learning_rate": 3.063219124754543e-06,
      "loss": 1.1329,
      "step": 1250
    },
    {
      "epoch": 0.7727832512315271,
      "grad_norm": 1.4384634943114656,
      "learning_rate": 2.9861799754722033e-06,
      "loss": 1.0958,
      "step": 1255
    },
    {
      "epoch": 0.7758620689655172,
      "grad_norm": 1.4620528648821034,
      "learning_rate": 2.9099515814150335e-06,
      "loss": 1.0788,
      "step": 1260
    },
    {
      "epoch": 0.7789408866995073,
      "grad_norm": 1.4542105514554136,
      "learning_rate": 2.8345427541248993e-06,
      "loss": 1.0791,
      "step": 1265
    },
    {
      "epoch": 0.7820197044334976,
      "grad_norm": 1.4558593967671596,
      "learning_rate": 2.7599622104066937e-06,
      "loss": 1.1828,
      "step": 1270
    },
    {
      "epoch": 0.7850985221674877,
      "grad_norm": 1.5321994639346959,
      "learning_rate": 2.6862185713207467e-06,
      "loss": 1.1391,
      "step": 1275
    },
    {
      "epoch": 0.7881773399014779,
      "grad_norm": 1.6770610614670527,
      "learning_rate": 2.6133203611862554e-06,
      "loss": 1.1285,
      "step": 1280
    },
    {
      "epoch": 0.791256157635468,
      "grad_norm": 1.3874238119481903,
      "learning_rate": 2.5412760065959386e-06,
      "loss": 1.1112,
      "step": 1285
    },
    {
      "epoch": 0.7943349753694581,
      "grad_norm": 1.427261231892756,
      "learning_rate": 2.4700938354419823e-06,
      "loss": 1.1246,
      "step": 1290
    },
    {
      "epoch": 0.7974137931034483,
      "grad_norm": 1.4509495050892607,
      "learning_rate": 2.3997820759533654e-06,
      "loss": 1.1514,
      "step": 1295
    },
    {
      "epoch": 0.8004926108374384,
      "grad_norm": 1.4238012977677232,
      "learning_rate": 2.3303488557447374e-06,
      "loss": 1.0915,
      "step": 1300
    },
    {
      "epoch": 0.8035714285714286,
      "grad_norm": 1.4554332844221027,
      "learning_rate": 2.26180220087692e-06,
      "loss": 1.1059,
      "step": 1305
    },
    {
      "epoch": 0.8066502463054187,
      "grad_norm": 1.5026437998930917,
      "learning_rate": 2.194150034929133e-06,
      "loss": 1.1326,
      "step": 1310
    },
    {
      "epoch": 0.8097290640394089,
      "grad_norm": 1.496022579048587,
      "learning_rate": 2.1274001780830776e-06,
      "loss": 1.1518,
      "step": 1315
    },
    {
      "epoch": 0.812807881773399,
      "grad_norm": 1.5081221072398265,
      "learning_rate": 2.0615603462189824e-06,
      "loss": 1.1587,
      "step": 1320
    },
    {
      "epoch": 0.8158866995073891,
      "grad_norm": 1.4299777929927682,
      "learning_rate": 1.9966381500236786e-06,
      "loss": 1.0735,
      "step": 1325
    },
    {
      "epoch": 0.8189655172413793,
      "grad_norm": 1.3998277385213755,
      "learning_rate": 1.932641094110855e-06,
      "loss": 1.1163,
      "step": 1330
    },
    {
      "epoch": 0.8220443349753694,
      "grad_norm": 1.3882987227650319,
      "learning_rate": 1.869576576153581e-06,
      "loss": 1.0641,
      "step": 1335
    },
    {
      "epoch": 0.8251231527093597,
      "grad_norm": 1.5119918997782806,
      "learning_rate": 1.8074518860291646e-06,
      "loss": 1.1125,
      "step": 1340
    },
    {
      "epoch": 0.8282019704433498,
      "grad_norm": 1.5078421997257327,
      "learning_rate": 1.746274204976498e-06,
      "loss": 1.1127,
      "step": 1345
    },
    {
      "epoch": 0.8312807881773399,
      "grad_norm": 1.4280729708147877,
      "learning_rate": 1.6860506047659442e-06,
      "loss": 1.1272,
      "step": 1350
    },
    {
      "epoch": 0.8343596059113301,
      "grad_norm": 1.5253523813470884,
      "learning_rate": 1.6267880468818787e-06,
      "loss": 1.0981,
      "step": 1355
    },
    {
      "epoch": 0.8374384236453202,
      "grad_norm": 1.4972346734462003,
      "learning_rate": 1.5684933817180014e-06,
      "loss": 1.1406,
      "step": 1360
    },
    {
      "epoch": 0.8405172413793104,
      "grad_norm": 1.5114578969478256,
      "learning_rate": 1.5111733477854507e-06,
      "loss": 1.144,
      "step": 1365
    },
    {
      "epoch": 0.8435960591133005,
      "grad_norm": 1.510922697559739,
      "learning_rate": 1.454834570933884e-06,
      "loss": 1.0911,
      "step": 1370
    },
    {
      "epoch": 0.8466748768472906,
      "grad_norm": 1.3971524947819978,
      "learning_rate": 1.399483563585573e-06,
      "loss": 1.1277,
      "step": 1375
    },
    {
      "epoch": 0.8497536945812808,
      "grad_norm": 1.515344246783894,
      "learning_rate": 1.345126723982594e-06,
      "loss": 1.1972,
      "step": 1380
    },
    {
      "epoch": 0.8528325123152709,
      "grad_norm": 1.5807183853866074,
      "learning_rate": 1.2917703354472467e-06,
      "loss": 1.1163,
      "step": 1385
    },
    {
      "epoch": 0.8559113300492611,
      "grad_norm": 1.4684701984206927,
      "learning_rate": 1.2394205656557224e-06,
      "loss": 1.0967,
      "step": 1390
    },
    {
      "epoch": 0.8589901477832512,
      "grad_norm": 1.473097270756067,
      "learning_rate": 1.1880834659251706e-06,
      "loss": 1.1412,
      "step": 1395
    },
    {
      "epoch": 0.8620689655172413,
      "grad_norm": 1.5801415877113048,
      "learning_rate": 1.1377649705142012e-06,
      "loss": 1.1271,
      "step": 1400
    },
    {
      "epoch": 0.8651477832512315,
      "grad_norm": 1.6395518223012362,
      "learning_rate": 1.0884708959369116e-06,
      "loss": 1.1615,
      "step": 1405
    },
    {
      "epoch": 0.8682266009852216,
      "grad_norm": 1.4471805497010097,
      "learning_rate": 1.040206940290547e-06,
      "loss": 1.097,
      "step": 1410
    },
    {
      "epoch": 0.8713054187192119,
      "grad_norm": 1.4139213372552049,
      "learning_rate": 9.929786825968213e-07,
      "loss": 1.0944,
      "step": 1415
    },
    {
      "epoch": 0.874384236453202,
      "grad_norm": 1.557966266246707,
      "learning_rate": 9.467915821570228e-07,
      "loss": 1.1476,
      "step": 1420
    },
    {
      "epoch": 0.8774630541871922,
      "grad_norm": 1.588720438616585,
      "learning_rate": 9.016509779209536e-07,
      "loss": 1.1287,
      "step": 1425
    },
    {
      "epoch": 0.8805418719211823,
      "grad_norm": 1.5819497563662563,
      "learning_rate": 8.575620878697744e-07,
      "loss": 1.1043,
      "step": 1430
    },
    {
      "epoch": 0.8836206896551724,
      "grad_norm": 1.5358289497842152,
      "learning_rate": 8.145300084128349e-07,
      "loss": 1.1285,
      "step": 1435
    },
    {
      "epoch": 0.8866995073891626,
      "grad_norm": 1.3701266308412228,
      "learning_rate": 7.725597137985741e-07,
      "loss": 1.0724,
      "step": 1440
    },
    {
      "epoch": 0.8897783251231527,
      "grad_norm": 1.5142386813925248,
      "learning_rate": 7.316560555395069e-07,
      "loss": 1.0832,
      "step": 1445
    },
    {
      "epoch": 0.8928571428571429,
      "grad_norm": 1.3918336722415168,
      "learning_rate": 6.918237618514378e-07,
      "loss": 1.0547,
      "step": 1450
    },
    {
      "epoch": 0.895935960591133,
      "grad_norm": 1.4611322575658334,
      "learning_rate": 6.530674371068946e-07,
      "loss": 1.119,
      "step": 1455
    },
    {
      "epoch": 0.8990147783251231,
      "grad_norm": 1.5505228144124603,
      "learning_rate": 6.153915613028915e-07,
      "loss": 1.1202,
      "step": 1460
    },
    {
      "epoch": 0.9020935960591133,
      "grad_norm": 1.445558891993411,
      "learning_rate": 5.788004895430799e-07,
      "loss": 1.0534,
      "step": 1465
    },
    {
      "epoch": 0.9051724137931034,
      "grad_norm": 1.475842859955036,
      "learning_rate": 5.43298451534312e-07,
      "loss": 1.1604,
      "step": 1470
    },
    {
      "epoch": 0.9082512315270936,
      "grad_norm": 1.4910920536611167,
      "learning_rate": 5.088895510977154e-07,
      "loss": 1.1112,
      "step": 1475
    },
    {
      "epoch": 0.9113300492610837,
      "grad_norm": 1.4558009812508625,
      "learning_rate": 4.755777656943239e-07,
      "loss": 1.1369,
      "step": 1480
    },
    {
      "epoch": 0.9144088669950738,
      "grad_norm": 1.581756224157198,
      "learning_rate": 4.433669459652945e-07,
      "loss": 1.0996,
      "step": 1485
    },
    {
      "epoch": 0.9174876847290641,
      "grad_norm": 1.351102512105557,
      "learning_rate": 4.1226081528680907e-07,
      "loss": 1.1564,
      "step": 1490
    },
    {
      "epoch": 0.9205665024630542,
      "grad_norm": 1.3920276021769993,
      "learning_rate": 3.822629693396651e-07,
      "loss": 1.0934,
      "step": 1495
    },
    {
      "epoch": 0.9236453201970444,
      "grad_norm": 1.5093480708629614,
      "learning_rate": 3.5337687569363734e-07,
      "loss": 1.1443,
      "step": 1500
    },
    {
      "epoch": 0.9267241379310345,
      "grad_norm": 1.518871812354284,
      "learning_rate": 3.2560587340665694e-07,
      "loss": 1.1272,
      "step": 1505
    },
    {
      "epoch": 0.9298029556650246,
      "grad_norm": 1.4737066145577384,
      "learning_rate": 2.989531726388262e-07,
      "loss": 1.1481,
      "step": 1510
    },
    {
      "epoch": 0.9328817733990148,
      "grad_norm": 1.6117607466517032,
      "learning_rate": 2.7342185428134604e-07,
      "loss": 1.1454,
      "step": 1515
    },
    {
      "epoch": 0.9359605911330049,
      "grad_norm": 1.3773319434896618,
      "learning_rate": 2.4901486960039025e-07,
      "loss": 1.1395,
      "step": 1520
    },
    {
      "epoch": 0.9390394088669951,
      "grad_norm": 1.5053964239987367,
      "learning_rate": 2.257350398959457e-07,
      "loss": 1.1042,
      "step": 1525
    },
    {
      "epoch": 0.9421182266009852,
      "grad_norm": 1.4719949014993627,
      "learning_rate": 2.035850561756969e-07,
      "loss": 1.12,
      "step": 1530
    },
    {
      "epoch": 0.9451970443349754,
      "grad_norm": 1.5068767379433818,
      "learning_rate": 1.8256747884395577e-07,
      "loss": 1.1165,
      "step": 1535
    },
    {
      "epoch": 0.9482758620689655,
      "grad_norm": 1.4951584035193635,
      "learning_rate": 1.6268473740569723e-07,
      "loss": 1.2019,
      "step": 1540
    },
    {
      "epoch": 0.9513546798029556,
      "grad_norm": 1.4953979801913249,
      "learning_rate": 1.4393913018572182e-07,
      "loss": 1.0962,
      "step": 1545
    },
    {
      "epoch": 0.9544334975369458,
      "grad_norm": 1.3741576802079731,
      "learning_rate": 1.2633282406298576e-07,
      "loss": 1.1159,
      "step": 1550
    },
    {
      "epoch": 0.9575123152709359,
      "grad_norm": 1.3447158004367004,
      "learning_rate": 1.0986785422011593e-07,
      "loss": 1.1011,
      "step": 1555
    },
    {
      "epoch": 0.9605911330049262,
      "grad_norm": 1.3750643633251347,
      "learning_rate": 9.454612390816686e-08,
      "loss": 1.0988,
      "step": 1560
    },
    {
      "epoch": 0.9636699507389163,
      "grad_norm": 1.4455853924361388,
      "learning_rate": 8.036940422660345e-08,
      "loss": 1.1191,
      "step": 1565
    },
    {
      "epoch": 0.9667487684729064,
      "grad_norm": 1.5044051468772355,
      "learning_rate": 6.733933391858238e-08,
      "loss": 1.1709,
      "step": 1570
    },
    {
      "epoch": 0.9698275862068966,
      "grad_norm": 1.4900863751566817,
      "learning_rate": 5.5457419181517145e-08,
      "loss": 1.11,
      "step": 1575
    },
    {
      "epoch": 0.9729064039408867,
      "grad_norm": 1.4196313395774074,
      "learning_rate": 4.472503349297497e-08,
      "loss": 1.0681,
      "step": 1580
    },
    {
      "epoch": 0.9759852216748769,
      "grad_norm": 1.4242212504411171,
      "learning_rate": 3.5143417451907195e-08,
      "loss": 1.0829,
      "step": 1585
    },
    {
      "epoch": 0.979064039408867,
      "grad_norm": 1.5207812508125576,
      "learning_rate": 2.671367863524732e-08,
      "loss": 1.1368,
      "step": 1590
    },
    {
      "epoch": 0.9821428571428571,
      "grad_norm": 1.5741176350001598,
      "learning_rate": 1.943679146988009e-08,
      "loss": 1.1599,
      "step": 1595
    },
    {
      "epoch": 0.9852216748768473,
      "grad_norm": 1.5272418292052072,
      "learning_rate": 1.3313597120002686e-08,
      "loss": 1.1514,
      "step": 1600
    },
    {
      "epoch": 0.9883004926108374,
      "grad_norm": 1.5621400273450403,
      "learning_rate": 8.34480338989141e-09,
      "loss": 1.1646,
      "step": 1605
    },
    {
      "epoch": 0.9913793103448276,
      "grad_norm": 1.400332249993296,
      "learning_rate": 4.530984642087122e-09,
      "loss": 1.0466,
      "step": 1610
    },
    {
      "epoch": 0.9944581280788177,
      "grad_norm": 1.413079680961053,
      "learning_rate": 1.872581730997247e-09,
      "loss": 1.1506,
      "step": 1615
    },
    {
      "epoch": 0.9975369458128078,
      "grad_norm": 1.5085064752676391,
      "learning_rate": 3.699019519387559e-10,
      "loss": 1.0614,
      "step": 1620
    },
    {
      "epoch": 1.0,
      "eval_loss": 1.1199536323547363,
      "eval_runtime": 952.2023,
      "eval_samples_per_second": 24.269,
      "eval_steps_per_second": 0.759,
      "step": 1624
    },
    {
      "epoch": 1.0,
      "step": 1624,
      "total_flos": 106409733357568.0,
      "train_loss": 1.1425106062971313,
      "train_runtime": 8990.4754,
      "train_samples_per_second": 5.78,
      "train_steps_per_second": 0.181
    }
  ],
  "logging_steps": 5,
  "max_steps": 1624,
  "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": 106409733357568.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}