{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9996613613274636,
  "eval_steps": 500,
  "global_step": 1476,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0006772773450728074,
      "grad_norm": 11.0625,
      "learning_rate": 2.702702702702703e-07,
      "loss": 2.4477,
      "step": 1
    },
    {
      "epoch": 0.0033863867253640365,
      "grad_norm": 11.8125,
      "learning_rate": 1.3513513513513515e-06,
      "loss": 2.4737,
      "step": 5
    },
    {
      "epoch": 0.006772773450728073,
      "grad_norm": 11.75,
      "learning_rate": 2.702702702702703e-06,
      "loss": 2.457,
      "step": 10
    },
    {
      "epoch": 0.01015916017609211,
      "grad_norm": 5.46875,
      "learning_rate": 4.0540540540540545e-06,
      "loss": 2.4211,
      "step": 15
    },
    {
      "epoch": 0.013545546901456146,
      "grad_norm": 2.875,
      "learning_rate": 5.405405405405406e-06,
      "loss": 2.3635,
      "step": 20
    },
    {
      "epoch": 0.016931933626820182,
      "grad_norm": 2.109375,
      "learning_rate": 6.7567567567567575e-06,
      "loss": 2.3577,
      "step": 25
    },
    {
      "epoch": 0.02031832035218422,
      "grad_norm": 1.7265625,
      "learning_rate": 8.108108108108109e-06,
      "loss": 2.2991,
      "step": 30
    },
    {
      "epoch": 0.023704707077548254,
      "grad_norm": 1.671875,
      "learning_rate": 9.45945945945946e-06,
      "loss": 2.2859,
      "step": 35
    },
    {
      "epoch": 0.027091093802912292,
      "grad_norm": 1.7421875,
      "learning_rate": 1.0810810810810812e-05,
      "loss": 2.2701,
      "step": 40
    },
    {
      "epoch": 0.03047748052827633,
      "grad_norm": 1.375,
      "learning_rate": 1.2162162162162164e-05,
      "loss": 2.2274,
      "step": 45
    },
    {
      "epoch": 0.033863867253640365,
      "grad_norm": 1.3203125,
      "learning_rate": 1.3513513513513515e-05,
      "loss": 2.2163,
      "step": 50
    },
    {
      "epoch": 0.0372502539790044,
      "grad_norm": 1.3046875,
      "learning_rate": 1.4864864864864865e-05,
      "loss": 2.2257,
      "step": 55
    },
    {
      "epoch": 0.04063664070436844,
      "grad_norm": 1.3046875,
      "learning_rate": 1.6216216216216218e-05,
      "loss": 2.2233,
      "step": 60
    },
    {
      "epoch": 0.04402302742973248,
      "grad_norm": 1.28125,
      "learning_rate": 1.756756756756757e-05,
      "loss": 2.2121,
      "step": 65
    },
    {
      "epoch": 0.04740941415509651,
      "grad_norm": 1.2421875,
      "learning_rate": 1.891891891891892e-05,
      "loss": 2.1914,
      "step": 70
    },
    {
      "epoch": 0.05079580088046055,
      "grad_norm": 1.2734375,
      "learning_rate": 2.0270270270270273e-05,
      "loss": 2.2119,
      "step": 75
    },
    {
      "epoch": 0.054182187605824585,
      "grad_norm": 1.21875,
      "learning_rate": 2.1621621621621624e-05,
      "loss": 2.1958,
      "step": 80
    },
    {
      "epoch": 0.05756857433118862,
      "grad_norm": 1.28125,
      "learning_rate": 2.2972972972972976e-05,
      "loss": 2.1629,
      "step": 85
    },
    {
      "epoch": 0.06095496105655266,
      "grad_norm": 1.21875,
      "learning_rate": 2.4324324324324327e-05,
      "loss": 2.1922,
      "step": 90
    },
    {
      "epoch": 0.06434134778191669,
      "grad_norm": 1.21875,
      "learning_rate": 2.567567567567568e-05,
      "loss": 2.1729,
      "step": 95
    },
    {
      "epoch": 0.06772773450728073,
      "grad_norm": 1.3203125,
      "learning_rate": 2.702702702702703e-05,
      "loss": 2.1695,
      "step": 100
    },
    {
      "epoch": 0.07111412123264477,
      "grad_norm": 1.21875,
      "learning_rate": 2.8378378378378378e-05,
      "loss": 2.1656,
      "step": 105
    },
    {
      "epoch": 0.0745005079580088,
      "grad_norm": 1.265625,
      "learning_rate": 2.972972972972973e-05,
      "loss": 2.1775,
      "step": 110
    },
    {
      "epoch": 0.07788689468337284,
      "grad_norm": 1.265625,
      "learning_rate": 3.108108108108108e-05,
      "loss": 2.1819,
      "step": 115
    },
    {
      "epoch": 0.08127328140873688,
      "grad_norm": 1.2421875,
      "learning_rate": 3.2432432432432436e-05,
      "loss": 2.1626,
      "step": 120
    },
    {
      "epoch": 0.08465966813410092,
      "grad_norm": 1.1953125,
      "learning_rate": 3.378378378378379e-05,
      "loss": 2.1499,
      "step": 125
    },
    {
      "epoch": 0.08804605485946496,
      "grad_norm": 1.2734375,
      "learning_rate": 3.513513513513514e-05,
      "loss": 2.191,
      "step": 130
    },
    {
      "epoch": 0.091432441584829,
      "grad_norm": 1.234375,
      "learning_rate": 3.648648648648649e-05,
      "loss": 2.1903,
      "step": 135
    },
    {
      "epoch": 0.09481882831019302,
      "grad_norm": 1.265625,
      "learning_rate": 3.783783783783784e-05,
      "loss": 2.1469,
      "step": 140
    },
    {
      "epoch": 0.09820521503555706,
      "grad_norm": 1.2265625,
      "learning_rate": 3.918918918918919e-05,
      "loss": 2.1473,
      "step": 145
    },
    {
      "epoch": 0.1015916017609211,
      "grad_norm": 1.2109375,
      "learning_rate": 3.9999776147073465e-05,
      "loss": 2.148,
      "step": 150
    },
    {
      "epoch": 0.10497798848628513,
      "grad_norm": 1.234375,
      "learning_rate": 3.999725785919791e-05,
      "loss": 2.1608,
      "step": 155
    },
    {
      "epoch": 0.10836437521164917,
      "grad_norm": 1.2109375,
      "learning_rate": 3.9991941820788456e-05,
      "loss": 2.1793,
      "step": 160
    },
    {
      "epoch": 0.11175076193701321,
      "grad_norm": 1.2265625,
      "learning_rate": 3.998382877559453e-05,
      "loss": 2.1449,
      "step": 165
    },
    {
      "epoch": 0.11513714866237724,
      "grad_norm": 1.1953125,
      "learning_rate": 3.99729198586856e-05,
      "loss": 2.146,
      "step": 170
    },
    {
      "epoch": 0.11852353538774128,
      "grad_norm": 1.2109375,
      "learning_rate": 3.995921659629232e-05,
      "loss": 2.1624,
      "step": 175
    },
    {
      "epoch": 0.12190992211310532,
      "grad_norm": 1.2265625,
      "learning_rate": 3.9942720905593045e-05,
      "loss": 2.1611,
      "step": 180
    },
    {
      "epoch": 0.12529630883846934,
      "grad_norm": 1.171875,
      "learning_rate": 3.992343509444555e-05,
      "loss": 2.1585,
      "step": 185
    },
    {
      "epoch": 0.12868269556383338,
      "grad_norm": 1.2578125,
      "learning_rate": 3.9901361861064215e-05,
      "loss": 2.1458,
      "step": 190
    },
    {
      "epoch": 0.13206908228919742,
      "grad_norm": 1.2421875,
      "learning_rate": 3.987650429364247e-05,
      "loss": 2.1729,
      "step": 195
    },
    {
      "epoch": 0.13545546901456146,
      "grad_norm": 1.2109375,
      "learning_rate": 3.984886586992077e-05,
      "loss": 2.1515,
      "step": 200
    },
    {
      "epoch": 0.1388418557399255,
      "grad_norm": 1.140625,
      "learning_rate": 3.9818450456700014e-05,
      "loss": 2.1697,
      "step": 205
    },
    {
      "epoch": 0.14222824246528953,
      "grad_norm": 1.1796875,
      "learning_rate": 3.978526230930056e-05,
      "loss": 2.155,
      "step": 210
    },
    {
      "epoch": 0.14561462919065357,
      "grad_norm": 1.1796875,
      "learning_rate": 3.97493060709669e-05,
      "loss": 2.1506,
      "step": 215
    },
    {
      "epoch": 0.1490010159160176,
      "grad_norm": 1.15625,
      "learning_rate": 3.971058677221799e-05,
      "loss": 2.1427,
      "step": 220
    },
    {
      "epoch": 0.15238740264138165,
      "grad_norm": 1.2578125,
      "learning_rate": 3.966910983014349e-05,
      "loss": 2.1601,
      "step": 225
    },
    {
      "epoch": 0.15577378936674569,
      "grad_norm": 1.1953125,
      "learning_rate": 3.962488104764586e-05,
      "loss": 2.1503,
      "step": 230
    },
    {
      "epoch": 0.15916017609210972,
      "grad_norm": 1.171875,
      "learning_rate": 3.95779066126285e-05,
      "loss": 2.1404,
      "step": 235
    },
    {
      "epoch": 0.16254656281747376,
      "grad_norm": 1.1875,
      "learning_rate": 3.952819309713002e-05,
      "loss": 2.1468,
      "step": 240
    },
    {
      "epoch": 0.1659329495428378,
      "grad_norm": 1.1875,
      "learning_rate": 3.947574745640475e-05,
      "loss": 2.1482,
      "step": 245
    },
    {
      "epoch": 0.16931933626820184,
      "grad_norm": 1.171875,
      "learning_rate": 3.942057702794969e-05,
      "loss": 2.1544,
      "step": 250
    },
    {
      "epoch": 0.17270572299356587,
      "grad_norm": 1.1796875,
      "learning_rate": 3.9362689530477915e-05,
      "loss": 2.1575,
      "step": 255
    },
    {
      "epoch": 0.1760921097189299,
      "grad_norm": 1.1953125,
      "learning_rate": 3.930209306283867e-05,
      "loss": 2.1607,
      "step": 260
    },
    {
      "epoch": 0.17947849644429395,
      "grad_norm": 1.1796875,
      "learning_rate": 3.923879610288432e-05,
      "loss": 2.1627,
      "step": 265
    },
    {
      "epoch": 0.182864883169658,
      "grad_norm": 1.203125,
      "learning_rate": 3.917280750628421e-05,
      "loss": 2.1424,
      "step": 270
    },
    {
      "epoch": 0.186251269895022,
      "grad_norm": 1.1640625,
      "learning_rate": 3.9104136505285716e-05,
      "loss": 2.1607,
      "step": 275
    },
    {
      "epoch": 0.18963765662038604,
      "grad_norm": 1.1875,
      "learning_rate": 3.903279270742259e-05,
      "loss": 2.136,
      "step": 280
    },
    {
      "epoch": 0.19302404334575007,
      "grad_norm": 1.1953125,
      "learning_rate": 3.8958786094170784e-05,
      "loss": 2.1558,
      "step": 285
    },
    {
      "epoch": 0.1964104300711141,
      "grad_norm": 1.1953125,
      "learning_rate": 3.8882127019552006e-05,
      "loss": 2.1368,
      "step": 290
    },
    {
      "epoch": 0.19979681679647815,
      "grad_norm": 1.171875,
      "learning_rate": 3.880282620868507e-05,
      "loss": 2.1243,
      "step": 295
    },
    {
      "epoch": 0.2031832035218422,
      "grad_norm": 1.1796875,
      "learning_rate": 3.872089475628546e-05,
      "loss": 2.1518,
      "step": 300
    },
    {
      "epoch": 0.20656959024720623,
      "grad_norm": 1.15625,
      "learning_rate": 3.863634412511302e-05,
      "loss": 2.12,
      "step": 305
    },
    {
      "epoch": 0.20995597697257026,
      "grad_norm": 1.1484375,
      "learning_rate": 3.8549186144368304e-05,
      "loss": 2.172,
      "step": 310
    },
    {
      "epoch": 0.2133423636979343,
      "grad_norm": 1.15625,
      "learning_rate": 3.845943300803754e-05,
      "loss": 2.1216,
      "step": 315
    },
    {
      "epoch": 0.21672875042329834,
      "grad_norm": 1.140625,
      "learning_rate": 3.8367097273186644e-05,
      "loss": 2.1387,
      "step": 320
    },
    {
      "epoch": 0.22011513714866238,
      "grad_norm": 1.1875,
      "learning_rate": 3.827219185820441e-05,
      "loss": 2.1338,
      "step": 325
    },
    {
      "epoch": 0.22350152387402641,
      "grad_norm": 1.1796875,
      "learning_rate": 3.817473004099508e-05,
      "loss": 2.134,
      "step": 330
    },
    {
      "epoch": 0.22688791059939045,
      "grad_norm": 1.1640625,
      "learning_rate": 3.807472545712076e-05,
      "loss": 2.1619,
      "step": 335
    },
    {
      "epoch": 0.2302742973247545,
      "grad_norm": 1.1640625,
      "learning_rate": 3.797219209789365e-05,
      "loss": 2.129,
      "step": 340
    },
    {
      "epoch": 0.23366068405011853,
      "grad_norm": 1.1484375,
      "learning_rate": 3.786714430841858e-05,
      "loss": 2.1493,
      "step": 345
    },
    {
      "epoch": 0.23704707077548257,
      "grad_norm": 1.1796875,
      "learning_rate": 3.7759596785586066e-05,
      "loss": 2.1453,
      "step": 350
    },
    {
      "epoch": 0.2404334575008466,
      "grad_norm": 1.1328125,
      "learning_rate": 3.7649564576016076e-05,
      "loss": 2.1411,
      "step": 355
    },
    {
      "epoch": 0.24381984422621064,
      "grad_norm": 1.2109375,
      "learning_rate": 3.7537063073952904e-05,
      "loss": 2.1367,
      "step": 360
    },
    {
      "epoch": 0.24720623095157468,
      "grad_norm": 1.1953125,
      "learning_rate": 3.742210801911147e-05,
      "loss": 2.1542,
      "step": 365
    },
    {
      "epoch": 0.2505926176769387,
      "grad_norm": 1.2109375,
      "learning_rate": 3.730471549447516e-05,
      "loss": 2.1327,
      "step": 370
    },
    {
      "epoch": 0.25397900440230275,
      "grad_norm": 1.1328125,
      "learning_rate": 3.7184901924045745e-05,
      "loss": 2.1263,
      "step": 375
    },
    {
      "epoch": 0.25736539112766676,
      "grad_norm": 1.15625,
      "learning_rate": 3.7062684070545535e-05,
      "loss": 2.1543,
      "step": 380
    },
    {
      "epoch": 0.26075177785303083,
      "grad_norm": 1.2109375,
      "learning_rate": 3.69380790330722e-05,
      "loss": 2.1336,
      "step": 385
    },
    {
      "epoch": 0.26413816457839484,
      "grad_norm": 1.1796875,
      "learning_rate": 3.6811104244706424e-05,
      "loss": 2.148,
      "step": 390
    },
    {
      "epoch": 0.2675245513037589,
      "grad_norm": 1.203125,
      "learning_rate": 3.668177747007297e-05,
      "loss": 2.1161,
      "step": 395
    },
    {
      "epoch": 0.2709109380291229,
      "grad_norm": 1.1875,
      "learning_rate": 3.6550116802855244e-05,
      "loss": 2.1428,
      "step": 400
    },
    {
      "epoch": 0.274297324754487,
      "grad_norm": 1.171875,
      "learning_rate": 3.6416140663263885e-05,
      "loss": 2.1307,
      "step": 405
    },
    {
      "epoch": 0.277683711479851,
      "grad_norm": 1.1640625,
      "learning_rate": 3.627986779545967e-05,
      "loss": 2.1407,
      "step": 410
    },
    {
      "epoch": 0.28107009820521506,
      "grad_norm": 1.125,
      "learning_rate": 3.614131726493102e-05,
      "loss": 2.1198,
      "step": 415
    },
    {
      "epoch": 0.28445648493057907,
      "grad_norm": 1.171875,
      "learning_rate": 3.600050845582669e-05,
      "loss": 2.1339,
      "step": 420
    },
    {
      "epoch": 0.28784287165594313,
      "grad_norm": 1.09375,
      "learning_rate": 3.5857461068243744e-05,
      "loss": 2.1346,
      "step": 425
    },
    {
      "epoch": 0.29122925838130714,
      "grad_norm": 1.125,
      "learning_rate": 3.5712195115471394e-05,
      "loss": 2.1184,
      "step": 430
    },
    {
      "epoch": 0.2946156451066712,
      "grad_norm": 1.1953125,
      "learning_rate": 3.5564730921191e-05,
      "loss": 2.1343,
      "step": 435
    },
    {
      "epoch": 0.2980020318320352,
      "grad_norm": 1.1328125,
      "learning_rate": 3.5415089116632665e-05,
      "loss": 2.1285,
      "step": 440
    },
    {
      "epoch": 0.30138841855739923,
      "grad_norm": 1.171875,
      "learning_rate": 3.526329063768878e-05,
      "loss": 2.1209,
      "step": 445
    },
    {
      "epoch": 0.3047748052827633,
      "grad_norm": 1.125,
      "learning_rate": 3.510935672198495e-05,
      "loss": 2.1302,
      "step": 450
    },
    {
      "epoch": 0.3081611920081273,
      "grad_norm": 1.171875,
      "learning_rate": 3.495330890590871e-05,
      "loss": 2.1075,
      "step": 455
    },
    {
      "epoch": 0.31154757873349137,
      "grad_norm": 1.140625,
      "learning_rate": 3.4795169021596417e-05,
      "loss": 2.1173,
      "step": 460
    },
    {
      "epoch": 0.3149339654588554,
      "grad_norm": 1.1640625,
      "learning_rate": 3.463495919387885e-05,
      "loss": 2.1284,
      "step": 465
    },
    {
      "epoch": 0.31832035218421945,
      "grad_norm": 1.1640625,
      "learning_rate": 3.447270183718574e-05,
      "loss": 2.1303,
      "step": 470
    },
    {
      "epoch": 0.32170673890958346,
      "grad_norm": 1.1796875,
      "learning_rate": 3.430841965240983e-05,
      "loss": 2.1401,
      "step": 475
    },
    {
      "epoch": 0.3250931256349475,
      "grad_norm": 1.140625,
      "learning_rate": 3.4142135623730954e-05,
      "loss": 2.1344,
      "step": 480
    },
    {
      "epoch": 0.32847951236031153,
      "grad_norm": 1.140625,
      "learning_rate": 3.397387301540028e-05,
      "loss": 2.1269,
      "step": 485
    },
    {
      "epoch": 0.3318658990856756,
      "grad_norm": 1.1171875,
      "learning_rate": 3.380365536848558e-05,
      "loss": 2.1308,
      "step": 490
    },
    {
      "epoch": 0.3352522858110396,
      "grad_norm": 1.0859375,
      "learning_rate": 3.363150649757763e-05,
      "loss": 2.1456,
      "step": 495
    },
    {
      "epoch": 0.3386386725364037,
      "grad_norm": 1.1640625,
      "learning_rate": 3.345745048745838e-05,
      "loss": 2.1163,
      "step": 500
    },
    {
      "epoch": 0.3420250592617677,
      "grad_norm": 1.1328125,
      "learning_rate": 3.328151168973139e-05,
      "loss": 2.1115,
      "step": 505
    },
    {
      "epoch": 0.34541144598713175,
      "grad_norm": 1.1640625,
      "learning_rate": 3.3103714719414814e-05,
      "loss": 2.1203,
      "step": 510
    },
    {
      "epoch": 0.34879783271249576,
      "grad_norm": 1.15625,
      "learning_rate": 3.2924084451497606e-05,
      "loss": 2.1226,
      "step": 515
    },
    {
      "epoch": 0.3521842194378598,
      "grad_norm": 1.1328125,
      "learning_rate": 3.274264601745938e-05,
      "loss": 2.1006,
      "step": 520
    },
    {
      "epoch": 0.35557060616322383,
      "grad_norm": 1.140625,
      "learning_rate": 3.25594248017543e-05,
      "loss": 2.1031,
      "step": 525
    },
    {
      "epoch": 0.3589569928885879,
      "grad_norm": 1.171875,
      "learning_rate": 3.237444643825965e-05,
      "loss": 2.128,
      "step": 530
    },
    {
      "epoch": 0.3623433796139519,
      "grad_norm": 1.109375,
      "learning_rate": 3.21877368066895e-05,
      "loss": 2.1303,
      "step": 535
    },
    {
      "epoch": 0.365729766339316,
      "grad_norm": 1.1484375,
      "learning_rate": 3.199932202897391e-05,
      "loss": 2.1269,
      "step": 540
    },
    {
      "epoch": 0.36911615306468,
      "grad_norm": 1.140625,
      "learning_rate": 3.1809228465604345e-05,
      "loss": 2.1526,
      "step": 545
    },
    {
      "epoch": 0.372502539790044,
      "grad_norm": 1.1015625,
      "learning_rate": 3.1617482711945616e-05,
      "loss": 2.1275,
      "step": 550
    },
    {
      "epoch": 0.37588892651540806,
      "grad_norm": 1.109375,
      "learning_rate": 3.1424111594515085e-05,
      "loss": 2.1329,
      "step": 555
    },
    {
      "epoch": 0.37927531324077207,
      "grad_norm": 1.1328125,
      "learning_rate": 3.1229142167229366e-05,
      "loss": 2.1363,
      "step": 560
    },
    {
      "epoch": 0.38266169996613614,
      "grad_norm": 1.140625,
      "learning_rate": 3.1032601707619375e-05,
      "loss": 2.1249,
      "step": 565
    },
    {
      "epoch": 0.38604808669150015,
      "grad_norm": 1.125,
      "learning_rate": 3.083451771301399e-05,
      "loss": 2.1323,
      "step": 570
    },
    {
      "epoch": 0.3894344734168642,
      "grad_norm": 1.1171875,
      "learning_rate": 3.0634917896692965e-05,
      "loss": 2.1328,
      "step": 575
    },
    {
      "epoch": 0.3928208601422282,
      "grad_norm": 1.15625,
      "learning_rate": 3.0433830184009694e-05,
      "loss": 2.1177,
      "step": 580
    },
    {
      "epoch": 0.3962072468675923,
      "grad_norm": 1.1484375,
      "learning_rate": 3.023128270848427e-05,
      "loss": 2.1293,
      "step": 585
    },
    {
      "epoch": 0.3995936335929563,
      "grad_norm": 1.09375,
      "learning_rate": 3.002730380786737e-05,
      "loss": 2.1073,
      "step": 590
    },
    {
      "epoch": 0.40298002031832036,
      "grad_norm": 1.125,
      "learning_rate": 2.982192202017568e-05,
      "loss": 2.1186,
      "step": 595
    },
    {
      "epoch": 0.4063664070436844,
      "grad_norm": 1.140625,
      "learning_rate": 2.9615166079699178e-05,
      "loss": 2.1191,
      "step": 600
    },
    {
      "epoch": 0.40975279376904844,
      "grad_norm": 1.125,
      "learning_rate": 2.9407064912981035e-05,
      "loss": 2.1293,
      "step": 605
    },
    {
      "epoch": 0.41313918049441245,
      "grad_norm": 1.125,
      "learning_rate": 2.91976476347706e-05,
      "loss": 2.1106,
      "step": 610
    },
    {
      "epoch": 0.4165255672197765,
      "grad_norm": 1.109375,
      "learning_rate": 2.898694354395006e-05,
      "loss": 2.1094,
      "step": 615
    },
    {
      "epoch": 0.4199119539451405,
      "grad_norm": 1.1484375,
      "learning_rate": 2.8774982119435307e-05,
      "loss": 2.1175,
      "step": 620
    },
    {
      "epoch": 0.4232983406705046,
      "grad_norm": 1.1484375,
      "learning_rate": 2.856179301605165e-05,
      "loss": 2.1291,
      "step": 625
    },
    {
      "epoch": 0.4266847273958686,
      "grad_norm": 1.109375,
      "learning_rate": 2.8347406060384934e-05,
      "loss": 2.1246,
      "step": 630
    },
    {
      "epoch": 0.43007111412123267,
      "grad_norm": 1.1640625,
      "learning_rate": 2.8131851246608533e-05,
      "loss": 2.1263,
      "step": 635
    },
    {
      "epoch": 0.4334575008465967,
      "grad_norm": 1.1328125,
      "learning_rate": 2.7915158732287042e-05,
      "loss": 2.1355,
      "step": 640
    },
    {
      "epoch": 0.43684388757196074,
      "grad_norm": 1.109375,
      "learning_rate": 2.7697358834156974e-05,
      "loss": 2.124,
      "step": 645
    },
    {
      "epoch": 0.44023027429732475,
      "grad_norm": 1.1640625,
      "learning_rate": 2.747848202388528e-05,
      "loss": 2.107,
      "step": 650
    },
    {
      "epoch": 0.4436166610226888,
      "grad_norm": 1.1328125,
      "learning_rate": 2.725855892380613e-05,
      "loss": 2.1343,
      "step": 655
    },
    {
      "epoch": 0.44700304774805283,
      "grad_norm": 1.140625,
      "learning_rate": 2.703762030263666e-05,
      "loss": 2.1117,
      "step": 660
    },
    {
      "epoch": 0.45038943447341684,
      "grad_norm": 1.1015625,
      "learning_rate": 2.6815697071172245e-05,
      "loss": 2.1107,
      "step": 665
    },
    {
      "epoch": 0.4537758211987809,
      "grad_norm": 1.1640625,
      "learning_rate": 2.6592820277961805e-05,
      "loss": 2.1012,
      "step": 670
    },
    {
      "epoch": 0.4571622079241449,
      "grad_norm": 1.140625,
      "learning_rate": 2.6369021104963976e-05,
      "loss": 2.1291,
      "step": 675
    },
    {
      "epoch": 0.460548594649509,
      "grad_norm": 1.1484375,
      "learning_rate": 2.61443308631845e-05,
      "loss": 2.1062,
      "step": 680
    },
    {
      "epoch": 0.463934981374873,
      "grad_norm": 1.109375,
      "learning_rate": 2.591878098829563e-05,
      "loss": 2.1077,
      "step": 685
    },
    {
      "epoch": 0.46732136810023706,
      "grad_norm": 1.140625,
      "learning_rate": 2.5692403036238038e-05,
      "loss": 2.1231,
      "step": 690
    },
    {
      "epoch": 0.47070775482560107,
      "grad_norm": 1.1328125,
      "learning_rate": 2.546522867880598e-05,
      "loss": 2.1441,
      "step": 695
    },
    {
      "epoch": 0.47409414155096513,
      "grad_norm": 1.140625,
      "learning_rate": 2.5237289699216136e-05,
      "loss": 2.109,
      "step": 700
    },
    {
      "epoch": 0.47748052827632914,
      "grad_norm": 1.140625,
      "learning_rate": 2.5008617987660975e-05,
      "loss": 2.1384,
      "step": 705
    },
    {
      "epoch": 0.4808669150016932,
      "grad_norm": 1.125,
      "learning_rate": 2.4779245536847074e-05,
      "loss": 2.1083,
      "step": 710
    },
    {
      "epoch": 0.4842533017270572,
      "grad_norm": 1.1328125,
      "learning_rate": 2.4549204437519137e-05,
      "loss": 2.1149,
      "step": 715
    },
    {
      "epoch": 0.4876396884524213,
      "grad_norm": 1.09375,
      "learning_rate": 2.4318526873970278e-05,
      "loss": 2.1059,
      "step": 720
    },
    {
      "epoch": 0.4910260751777853,
      "grad_norm": 1.125,
      "learning_rate": 2.4087245119539227e-05,
      "loss": 2.1218,
      "step": 725
    },
    {
      "epoch": 0.49441246190314936,
      "grad_norm": 1.109375,
      "learning_rate": 2.385539153209508e-05,
      "loss": 2.112,
      "step": 730
    },
    {
      "epoch": 0.49779884862851337,
      "grad_norm": 1.15625,
      "learning_rate": 2.3622998549510197e-05,
      "loss": 2.1128,
      "step": 735
    },
    {
      "epoch": 0.5011852353538774,
      "grad_norm": 1.1328125,
      "learning_rate": 2.3390098685121938e-05,
      "loss": 2.1269,
      "step": 740
    },
    {
      "epoch": 0.5045716220792414,
      "grad_norm": 1.1484375,
      "learning_rate": 2.3156724523183832e-05,
      "loss": 2.1233,
      "step": 745
    },
    {
      "epoch": 0.5079580088046055,
      "grad_norm": 1.109375,
      "learning_rate": 2.2922908714306814e-05,
      "loss": 2.1443,
      "step": 750
    },
    {
      "epoch": 0.5113443955299696,
      "grad_norm": 1.078125,
      "learning_rate": 2.268868397089119e-05,
      "loss": 2.0891,
      "step": 755
    },
    {
      "epoch": 0.5147307822553335,
      "grad_norm": 1.1015625,
      "learning_rate": 2.2454083062549957e-05,
      "loss": 2.112,
      "step": 760
    },
    {
      "epoch": 0.5181171689806976,
      "grad_norm": 1.1171875,
      "learning_rate": 2.2219138811524102e-05,
      "loss": 2.1152,
      "step": 765
    },
    {
      "epoch": 0.5215035557060617,
      "grad_norm": 1.1328125,
      "learning_rate": 2.1983884088090553e-05,
      "loss": 2.1255,
      "step": 770
    },
    {
      "epoch": 0.5248899424314256,
      "grad_norm": 1.109375,
      "learning_rate": 2.1748351805963422e-05,
      "loss": 2.1058,
      "step": 775
    },
    {
      "epoch": 0.5282763291567897,
      "grad_norm": 1.125,
      "learning_rate": 2.151257491768914e-05,
      "loss": 2.1231,
      "step": 780
    },
    {
      "epoch": 0.5316627158821537,
      "grad_norm": 1.109375,
      "learning_rate": 2.127658641003617e-05,
      "loss": 2.1258,
      "step": 785
    },
    {
      "epoch": 0.5350491026075178,
      "grad_norm": 1.09375,
      "learning_rate": 2.1040419299379958e-05,
      "loss": 2.1146,
      "step": 790
    },
    {
      "epoch": 0.5384354893328818,
      "grad_norm": 1.1171875,
      "learning_rate": 2.0804106627083708e-05,
      "loss": 2.1032,
      "step": 795
    },
    {
      "epoch": 0.5418218760582458,
      "grad_norm": 1.1328125,
      "learning_rate": 2.0567681454875664e-05,
      "loss": 2.117,
      "step": 800
    },
    {
      "epoch": 0.5452082627836099,
      "grad_norm": 1.125,
      "learning_rate": 2.0331176860223575e-05,
      "loss": 2.1131,
      "step": 805
    },
    {
      "epoch": 0.548594649508974,
      "grad_norm": 1.09375,
      "learning_rate": 2.009462593170691e-05,
      "loss": 2.1134,
      "step": 810
    },
    {
      "epoch": 0.5519810362343379,
      "grad_norm": 1.1328125,
      "learning_rate": 1.985806176438756e-05,
      "loss": 2.126,
      "step": 815
    },
    {
      "epoch": 0.555367422959702,
      "grad_norm": 1.1171875,
      "learning_rate": 1.9621517455179627e-05,
      "loss": 2.1293,
      "step": 820
    },
    {
      "epoch": 0.558753809685066,
      "grad_norm": 1.140625,
      "learning_rate": 1.93850260982189e-05,
      "loss": 2.1059,
      "step": 825
    },
    {
      "epoch": 0.5621401964104301,
      "grad_norm": 1.15625,
      "learning_rate": 1.9148620780232842e-05,
      "loss": 2.1121,
      "step": 830
    },
    {
      "epoch": 0.5655265831357941,
      "grad_norm": 1.1171875,
      "learning_rate": 1.891233457591142e-05,
      "loss": 2.1178,
      "step": 835
    },
    {
      "epoch": 0.5689129698611581,
      "grad_norm": 1.109375,
      "learning_rate": 1.8676200543279864e-05,
      "loss": 2.105,
      "step": 840
    },
    {
      "epoch": 0.5722993565865222,
      "grad_norm": 1.109375,
      "learning_rate": 1.8440251719073533e-05,
      "loss": 2.1097,
      "step": 845
    },
    {
      "epoch": 0.5756857433118863,
      "grad_norm": 1.1015625,
      "learning_rate": 1.8204521114115872e-05,
      "loss": 2.1136,
      "step": 850
    },
    {
      "epoch": 0.5790721300372502,
      "grad_norm": 1.1015625,
      "learning_rate": 1.7969041708700028e-05,
      "loss": 2.087,
      "step": 855
    },
    {
      "epoch": 0.5824585167626143,
      "grad_norm": 1.109375,
      "learning_rate": 1.7733846447974606e-05,
      "loss": 2.115,
      "step": 860
    },
    {
      "epoch": 0.5858449034879784,
      "grad_norm": 1.140625,
      "learning_rate": 1.7498968237334484e-05,
      "loss": 2.1076,
      "step": 865
    },
    {
      "epoch": 0.5892312902133424,
      "grad_norm": 1.15625,
      "learning_rate": 1.7264439937817112e-05,
      "loss": 2.1182,
      "step": 870
    },
    {
      "epoch": 0.5926176769387064,
      "grad_norm": 1.1171875,
      "learning_rate": 1.7030294361505007e-05,
      "loss": 2.1157,
      "step": 875
    },
    {
      "epoch": 0.5960040636640704,
      "grad_norm": 1.125,
      "learning_rate": 1.6796564266935158e-05,
      "loss": 2.1072,
      "step": 880
    },
    {
      "epoch": 0.5993904503894345,
      "grad_norm": 1.09375,
      "learning_rate": 1.6563282354515857e-05,
      "loss": 2.1034,
      "step": 885
    },
    {
      "epoch": 0.6027768371147985,
      "grad_norm": 1.078125,
      "learning_rate": 1.6330481261951726e-05,
      "loss": 2.0906,
      "step": 890
    },
    {
      "epoch": 0.6061632238401625,
      "grad_norm": 1.1328125,
      "learning_rate": 1.609819355967744e-05,
      "loss": 2.1431,
      "step": 895
    },
    {
      "epoch": 0.6095496105655266,
      "grad_norm": 1.1015625,
      "learning_rate": 1.586645174630094e-05,
      "loss": 2.0905,
      "step": 900
    },
    {
      "epoch": 0.6129359972908907,
      "grad_norm": 1.0859375,
      "learning_rate": 1.563528824405666e-05,
      "loss": 2.1059,
      "step": 905
    },
    {
      "epoch": 0.6163223840162546,
      "grad_norm": 1.0703125,
      "learning_rate": 1.5404735394269403e-05,
      "loss": 2.1163,
      "step": 910
    },
    {
      "epoch": 0.6197087707416187,
      "grad_norm": 1.0703125,
      "learning_rate": 1.5174825452829615e-05,
      "loss": 2.1211,
      "step": 915
    },
    {
      "epoch": 0.6230951574669827,
      "grad_norm": 1.09375,
      "learning_rate": 1.4945590585680539e-05,
      "loss": 2.1069,
      "step": 920
    },
    {
      "epoch": 0.6264815441923468,
      "grad_norm": 1.09375,
      "learning_rate": 1.4717062864318004e-05,
      "loss": 2.0974,
      "step": 925
    },
    {
      "epoch": 0.6298679309177108,
      "grad_norm": 1.09375,
      "learning_rate": 1.4489274261303387e-05,
      "loss": 2.1034,
      "step": 930
    },
    {
      "epoch": 0.6332543176430748,
      "grad_norm": 1.1015625,
      "learning_rate": 1.4262256645790469e-05,
      "loss": 2.1126,
      "step": 935
    },
    {
      "epoch": 0.6366407043684389,
      "grad_norm": 1.109375,
      "learning_rate": 1.4036041779066696e-05,
      "loss": 2.1133,
      "step": 940
    },
    {
      "epoch": 0.640027091093803,
      "grad_norm": 1.0625,
      "learning_rate": 1.3810661310109565e-05,
      "loss": 2.1165,
      "step": 945
    },
    {
      "epoch": 0.6434134778191669,
      "grad_norm": 1.0859375,
      "learning_rate": 1.3586146771158746e-05,
      "loss": 2.1116,
      "step": 950
    },
    {
      "epoch": 0.646799864544531,
      "grad_norm": 1.1328125,
      "learning_rate": 1.3362529573304473e-05,
      "loss": 2.0953,
      "step": 955
    },
    {
      "epoch": 0.650186251269895,
      "grad_norm": 1.125,
      "learning_rate": 1.3139841002092949e-05,
      "loss": 2.1109,
      "step": 960
    },
    {
      "epoch": 0.6535726379952591,
      "grad_norm": 1.109375,
      "learning_rate": 1.2918112213149292e-05,
      "loss": 2.0961,
      "step": 965
    },
    {
      "epoch": 0.6569590247206231,
      "grad_norm": 1.1015625,
      "learning_rate": 1.2697374227818609e-05,
      "loss": 2.1,
      "step": 970
    },
    {
      "epoch": 0.6603454114459871,
      "grad_norm": 1.1171875,
      "learning_rate": 1.2477657928825977e-05,
      "loss": 2.122,
      "step": 975
    },
    {
      "epoch": 0.6637317981713512,
      "grad_norm": 1.125,
      "learning_rate": 1.2258994055955658e-05,
      "loss": 2.0896,
      "step": 980
    },
    {
      "epoch": 0.6671181848967151,
      "grad_norm": 1.1015625,
      "learning_rate": 1.2041413201750473e-05,
      "loss": 2.0827,
      "step": 985
    },
    {
      "epoch": 0.6705045716220792,
      "grad_norm": 1.0703125,
      "learning_rate": 1.1824945807231642e-05,
      "loss": 2.1213,
      "step": 990
    },
    {
      "epoch": 0.6738909583474433,
      "grad_norm": 1.0859375,
      "learning_rate": 1.1609622157639913e-05,
      "loss": 2.113,
      "step": 995
    },
    {
      "epoch": 0.6772773450728073,
      "grad_norm": 1.125,
      "learning_rate": 1.139547237819846e-05,
      "loss": 2.1118,
      "step": 1000
    },
    {
      "epoch": 0.6806637317981713,
      "grad_norm": 1.109375,
      "learning_rate": 1.1182526429898118e-05,
      "loss": 2.1255,
      "step": 1005
    },
    {
      "epoch": 0.6840501185235354,
      "grad_norm": 1.1328125,
      "learning_rate": 1.0970814105305689e-05,
      "loss": 2.116,
      "step": 1010
    },
    {
      "epoch": 0.6874365052488994,
      "grad_norm": 1.1015625,
      "learning_rate": 1.0760365024395745e-05,
      "loss": 2.0958,
      "step": 1015
    },
    {
      "epoch": 0.6908228919742635,
      "grad_norm": 1.1015625,
      "learning_rate": 1.0551208630406587e-05,
      "loss": 2.0891,
      "step": 1020
    },
    {
      "epoch": 0.6942092786996275,
      "grad_norm": 1.1328125,
      "learning_rate": 1.0343374185720927e-05,
      "loss": 2.1199,
      "step": 1025
    },
    {
      "epoch": 0.6975956654249915,
      "grad_norm": 1.109375,
      "learning_rate": 1.0136890767771923e-05,
      "loss": 2.1115,
      "step": 1030
    },
    {
      "epoch": 0.7009820521503556,
      "grad_norm": 1.109375,
      "learning_rate": 9.931787264975021e-06,
      "loss": 2.1147,
      "step": 1035
    },
    {
      "epoch": 0.7043684388757196,
      "grad_norm": 1.1015625,
      "learning_rate": 9.72809237268628e-06,
      "loss": 2.1142,
      "step": 1040
    },
    {
      "epoch": 0.7077548256010836,
      "grad_norm": 1.09375,
      "learning_rate": 9.525834589187701e-06,
      "loss": 2.1012,
      "step": 1045
    },
    {
      "epoch": 0.7111412123264477,
      "grad_norm": 1.1015625,
      "learning_rate": 9.325042211700111e-06,
      "loss": 2.1048,
      "step": 1050
    },
    {
      "epoch": 0.7145275990518117,
      "grad_norm": 1.1171875,
      "learning_rate": 9.125743332424213e-06,
      "loss": 2.1197,
      "step": 1055
    },
    {
      "epoch": 0.7179139857771758,
      "grad_norm": 1.078125,
      "learning_rate": 8.92796583461031e-06,
      "loss": 2.0983,
      "step": 1060
    },
    {
      "epoch": 0.7213003725025398,
      "grad_norm": 1.1171875,
      "learning_rate": 8.731737388657198e-06,
      "loss": 2.1128,
      "step": 1065
    },
    {
      "epoch": 0.7246867592279038,
      "grad_norm": 1.1015625,
      "learning_rate": 8.537085448240951e-06,
      "loss": 2.1188,
      "step": 1070
    },
    {
      "epoch": 0.7280731459532679,
      "grad_norm": 1.1171875,
      "learning_rate": 8.344037246473919e-06,
      "loss": 2.1098,
      "step": 1075
    },
    {
      "epoch": 0.731459532678632,
      "grad_norm": 1.1171875,
      "learning_rate": 8.152619792094674e-06,
      "loss": 2.1052,
      "step": 1080
    },
    {
      "epoch": 0.7348459194039959,
      "grad_norm": 1.125,
      "learning_rate": 7.962859865689282e-06,
      "loss": 2.1407,
      "step": 1085
    },
    {
      "epoch": 0.73823230612936,
      "grad_norm": 1.15625,
      "learning_rate": 7.77478401594453e-06,
      "loss": 2.1314,
      "step": 1090
    },
    {
      "epoch": 0.741618692854724,
      "grad_norm": 1.1015625,
      "learning_rate": 7.588418555933581e-06,
      "loss": 2.101,
      "step": 1095
    },
    {
      "epoch": 0.745005079580088,
      "grad_norm": 1.109375,
      "learning_rate": 7.403789559434573e-06,
      "loss": 2.1036,
      "step": 1100
    },
    {
      "epoch": 0.7483914663054521,
      "grad_norm": 1.09375,
      "learning_rate": 7.220922857282804e-06,
      "loss": 2.0901,
      "step": 1105
    },
    {
      "epoch": 0.7517778530308161,
      "grad_norm": 1.09375,
      "learning_rate": 7.039844033756713e-06,
      "loss": 2.088,
      "step": 1110
    },
    {
      "epoch": 0.7551642397561802,
      "grad_norm": 1.140625,
      "learning_rate": 6.860578422998563e-06,
      "loss": 2.1146,
      "step": 1115
    },
    {
      "epoch": 0.7585506264815441,
      "grad_norm": 1.1328125,
      "learning_rate": 6.683151105469956e-06,
      "loss": 2.1177,
      "step": 1120
    },
    {
      "epoch": 0.7619370132069082,
      "grad_norm": 1.125,
      "learning_rate": 6.50758690444297e-06,
      "loss": 2.1109,
      "step": 1125
    },
    {
      "epoch": 0.7653233999322723,
      "grad_norm": 1.1171875,
      "learning_rate": 6.333910382527175e-06,
      "loss": 2.1356,
      "step": 1130
    },
    {
      "epoch": 0.7687097866576363,
      "grad_norm": 1.09375,
      "learning_rate": 6.1621458382331444e-06,
      "loss": 2.1003,
      "step": 1135
    },
    {
      "epoch": 0.7720961733830003,
      "grad_norm": 1.1015625,
      "learning_rate": 5.9923173025729895e-06,
      "loss": 2.1078,
      "step": 1140
    },
    {
      "epoch": 0.7754825601083644,
      "grad_norm": 1.140625,
      "learning_rate": 5.824448535698195e-06,
      "loss": 2.0915,
      "step": 1145
    },
    {
      "epoch": 0.7788689468337284,
      "grad_norm": 1.140625,
      "learning_rate": 5.658563023575478e-06,
      "loss": 2.1269,
      "step": 1150
    },
    {
      "epoch": 0.7822553335590925,
      "grad_norm": 1.09375,
      "learning_rate": 5.494683974700878e-06,
      "loss": 2.1221,
      "step": 1155
    },
    {
      "epoch": 0.7856417202844564,
      "grad_norm": 1.1015625,
      "learning_rate": 5.332834316852786e-06,
      "loss": 2.0986,
      "step": 1160
    },
    {
      "epoch": 0.7890281070098205,
      "grad_norm": 1.125,
      "learning_rate": 5.173036693884164e-06,
      "loss": 2.1099,
      "step": 1165
    },
    {
      "epoch": 0.7924144937351846,
      "grad_norm": 1.1484375,
      "learning_rate": 5.015313462554527e-06,
      "loss": 2.1086,
      "step": 1170
    },
    {
      "epoch": 0.7958008804605486,
      "grad_norm": 1.125,
      "learning_rate": 4.859686689402099e-06,
      "loss": 2.1064,
      "step": 1175
    },
    {
      "epoch": 0.7991872671859126,
      "grad_norm": 1.125,
      "learning_rate": 4.70617814765651e-06,
      "loss": 2.1341,
      "step": 1180
    },
    {
      "epoch": 0.8025736539112767,
      "grad_norm": 1.1015625,
      "learning_rate": 4.554809314192629e-06,
      "loss": 2.1181,
      "step": 1185
    },
    {
      "epoch": 0.8059600406366407,
      "grad_norm": 1.1171875,
      "learning_rate": 4.405601366525776e-06,
      "loss": 2.1068,
      "step": 1190
    },
    {
      "epoch": 0.8093464273620048,
      "grad_norm": 1.1328125,
      "learning_rate": 4.258575179848847e-06,
      "loss": 2.0982,
      "step": 1195
    },
    {
      "epoch": 0.8127328140873687,
      "grad_norm": 1.09375,
      "learning_rate": 4.113751324111748e-06,
      "loss": 2.1218,
      "step": 1200
    },
    {
      "epoch": 0.8161192008127328,
      "grad_norm": 1.0859375,
      "learning_rate": 3.971150061143492e-06,
      "loss": 2.1166,
      "step": 1205
    },
    {
      "epoch": 0.8195055875380969,
      "grad_norm": 1.1015625,
      "learning_rate": 3.830791341817468e-06,
      "loss": 2.1201,
      "step": 1210
    },
    {
      "epoch": 0.8228919742634608,
      "grad_norm": 1.109375,
      "learning_rate": 3.6926948032601663e-06,
      "loss": 2.1185,
      "step": 1215
    },
    {
      "epoch": 0.8262783609888249,
      "grad_norm": 1.0703125,
      "learning_rate": 3.5568797661038004e-06,
      "loss": 2.1061,
      "step": 1220
    },
    {
      "epoch": 0.829664747714189,
      "grad_norm": 1.125,
      "learning_rate": 3.4233652317832424e-06,
      "loss": 2.1188,
      "step": 1225
    },
    {
      "epoch": 0.833051134439553,
      "grad_norm": 1.1015625,
      "learning_rate": 3.292169879877569e-06,
      "loss": 2.1077,
      "step": 1230
    },
    {
      "epoch": 0.836437521164917,
      "grad_norm": 1.09375,
      "learning_rate": 3.1633120654966932e-06,
      "loss": 2.124,
      "step": 1235
    },
    {
      "epoch": 0.839823907890281,
      "grad_norm": 1.109375,
      "learning_rate": 3.0368098167133376e-06,
      "loss": 2.1057,
      "step": 1240
    },
    {
      "epoch": 0.8432102946156451,
      "grad_norm": 1.09375,
      "learning_rate": 2.9126808320408016e-06,
      "loss": 2.1011,
      "step": 1245
    },
    {
      "epoch": 0.8465966813410092,
      "grad_norm": 1.09375,
      "learning_rate": 2.7909424779567996e-06,
      "loss": 2.0859,
      "step": 1250
    },
    {
      "epoch": 0.8499830680663731,
      "grad_norm": 1.1015625,
      "learning_rate": 2.6716117864738024e-06,
      "loss": 2.1224,
      "step": 1255
    },
    {
      "epoch": 0.8533694547917372,
      "grad_norm": 1.140625,
      "learning_rate": 2.5547054527561253e-06,
      "loss": 2.108,
      "step": 1260
    },
    {
      "epoch": 0.8567558415171013,
      "grad_norm": 1.0859375,
      "learning_rate": 2.4402398327841658e-06,
      "loss": 2.1146,
      "step": 1265
    },
    {
      "epoch": 0.8601422282424653,
      "grad_norm": 1.125,
      "learning_rate": 2.3282309410661007e-06,
      "loss": 2.1147,
      "step": 1270
    },
    {
      "epoch": 0.8635286149678293,
      "grad_norm": 1.0859375,
      "learning_rate": 2.2186944483973273e-06,
      "loss": 2.087,
      "step": 1275
    },
    {
      "epoch": 0.8669150016931934,
      "grad_norm": 1.125,
      "learning_rate": 2.1116456796680376e-06,
      "loss": 2.0946,
      "step": 1280
    },
    {
      "epoch": 0.8703013884185574,
      "grad_norm": 1.09375,
      "learning_rate": 2.0070996117191676e-06,
      "loss": 2.115,
      "step": 1285
    },
    {
      "epoch": 0.8736877751439215,
      "grad_norm": 1.125,
      "learning_rate": 1.9050708712470012e-06,
      "loss": 2.1077,
      "step": 1290
    },
    {
      "epoch": 0.8770741618692854,
      "grad_norm": 1.1171875,
      "learning_rate": 1.8055737327568378e-06,
      "loss": 2.1076,
      "step": 1295
    },
    {
      "epoch": 0.8804605485946495,
      "grad_norm": 1.0859375,
      "learning_rate": 1.7086221165658544e-06,
      "loss": 2.102,
      "step": 1300
    },
    {
      "epoch": 0.8838469353200136,
      "grad_norm": 1.1171875,
      "learning_rate": 1.614229586855609e-06,
      "loss": 2.1115,
      "step": 1305
    },
    {
      "epoch": 0.8872333220453776,
      "grad_norm": 1.09375,
      "learning_rate": 1.5224093497742654e-06,
      "loss": 2.1187,
      "step": 1310
    },
    {
      "epoch": 0.8906197087707416,
      "grad_norm": 1.1015625,
      "learning_rate": 1.4331742515890091e-06,
      "loss": 2.1099,
      "step": 1315
    },
    {
      "epoch": 0.8940060954961057,
      "grad_norm": 1.078125,
      "learning_rate": 1.3465367768887471e-06,
      "loss": 2.11,
      "step": 1320
    },
    {
      "epoch": 0.8973924822214697,
      "grad_norm": 1.0625,
      "learning_rate": 1.2625090468374212e-06,
      "loss": 2.1088,
      "step": 1325
    },
    {
      "epoch": 0.9007788689468337,
      "grad_norm": 1.1171875,
      "learning_rate": 1.1811028174782102e-06,
      "loss": 2.1239,
      "step": 1330
    },
    {
      "epoch": 0.9041652556721977,
      "grad_norm": 1.078125,
      "learning_rate": 1.1023294780887394e-06,
      "loss": 2.1018,
      "step": 1335
    },
    {
      "epoch": 0.9075516423975618,
      "grad_norm": 1.1328125,
      "learning_rate": 1.0262000495876735e-06,
      "loss": 2.12,
      "step": 1340
    },
    {
      "epoch": 0.9109380291229259,
      "grad_norm": 1.09375,
      "learning_rate": 9.527251829927997e-07,
      "loss": 2.1071,
      "step": 1345
    },
    {
      "epoch": 0.9143244158482898,
      "grad_norm": 1.09375,
      "learning_rate": 8.819151579308949e-07,
      "loss": 2.1099,
      "step": 1350
    },
    {
      "epoch": 0.9177108025736539,
      "grad_norm": 1.109375,
      "learning_rate": 8.137798811995235e-07,
      "loss": 2.1087,
      "step": 1355
    },
    {
      "epoch": 0.921097189299018,
      "grad_norm": 1.109375,
      "learning_rate": 7.483288853810067e-07,
      "loss": 2.1193,
      "step": 1360
    },
    {
      "epoch": 0.924483576024382,
      "grad_norm": 1.1171875,
      "learning_rate": 6.85571327508765e-07,
      "loss": 2.1228,
      "step": 1365
    },
    {
      "epoch": 0.927869962749746,
      "grad_norm": 1.1015625,
      "learning_rate": 6.255159877861782e-07,
      "loss": 2.1164,
      "step": 1370
    },
    {
      "epoch": 0.93125634947511,
      "grad_norm": 1.109375,
      "learning_rate": 5.681712683581775e-07,
      "loss": 2.1116,
      "step": 1375
    },
    {
      "epoch": 0.9346427362004741,
      "grad_norm": 1.1015625,
      "learning_rate": 5.135451921357337e-07,
      "loss": 2.0996,
      "step": 1380
    },
    {
      "epoch": 0.9380291229258382,
      "grad_norm": 1.140625,
      "learning_rate": 4.616454016733851e-07,
      "loss": 2.1326,
      "step": 1385
    },
    {
      "epoch": 0.9414155096512021,
      "grad_norm": 1.1171875,
      "learning_rate": 4.1247915810001205e-07,
      "loss": 2.1102,
      "step": 1390
    },
    {
      "epoch": 0.9448018963765662,
      "grad_norm": 1.0546875,
      "learning_rate": 3.660533401029387e-07,
      "loss": 2.1122,
      "step": 1395
    },
    {
      "epoch": 0.9481882831019303,
      "grad_norm": 1.1484375,
      "learning_rate": 3.2237444296557173e-07,
      "loss": 2.1179,
      "step": 1400
    },
    {
      "epoch": 0.9515746698272943,
      "grad_norm": 1.125,
      "learning_rate": 2.81448577658654e-07,
      "loss": 2.1075,
      "step": 1405
    },
    {
      "epoch": 0.9549610565526583,
      "grad_norm": 1.1015625,
      "learning_rate": 2.432814699853081e-07,
      "loss": 2.1126,
      "step": 1410
    },
    {
      "epoch": 0.9583474432780223,
      "grad_norm": 1.0859375,
      "learning_rate": 2.078784597799599e-07,
      "loss": 2.106,
      "step": 1415
    },
    {
      "epoch": 0.9617338300033864,
      "grad_norm": 1.109375,
      "learning_rate": 1.752445001612535e-07,
      "loss": 2.1234,
      "step": 1420
    },
    {
      "epoch": 0.9651202167287504,
      "grad_norm": 1.09375,
      "learning_rate": 1.4538415683907236e-07,
      "loss": 2.0946,
      "step": 1425
    },
    {
      "epoch": 0.9685066034541144,
      "grad_norm": 1.1171875,
      "learning_rate": 1.1830160747577924e-07,
      "loss": 2.0985,
      "step": 1430
    },
    {
      "epoch": 0.9718929901794785,
      "grad_norm": 1.1953125,
      "learning_rate": 9.400064110172358e-08,
      "loss": 2.1185,
      "step": 1435
    },
    {
      "epoch": 0.9752793769048426,
      "grad_norm": 1.109375,
      "learning_rate": 7.248465758513457e-08,
      "loss": 2.0938,
      "step": 1440
    },
    {
      "epoch": 0.9786657636302065,
      "grad_norm": 1.09375,
      "learning_rate": 5.375666715645489e-08,
      "loss": 2.109,
      "step": 1445
    },
    {
      "epoch": 0.9820521503555706,
      "grad_norm": 1.1015625,
      "learning_rate": 3.781928998718654e-08,
      "loss": 2.1146,
      "step": 1450
    },
    {
      "epoch": 0.9854385370809347,
      "grad_norm": 1.09375,
      "learning_rate": 2.4674755823321794e-08,
      "loss": 2.1029,
      "step": 1455
    },
    {
      "epoch": 0.9888249238062987,
      "grad_norm": 1.1015625,
      "learning_rate": 1.4324903673370583e-08,
      "loss": 2.1195,
      "step": 1460
    },
    {
      "epoch": 0.9922113105316627,
      "grad_norm": 1.0859375,
      "learning_rate": 6.771181551088468e-09,
      "loss": 2.1055,
      "step": 1465
    },
    {
      "epoch": 0.9955976972570267,
      "grad_norm": 1.125,
      "learning_rate": 2.014646272876508e-09,
      "loss": 2.1176,
      "step": 1470
    },
    {
      "epoch": 0.9989840839823908,
      "grad_norm": 1.078125,
      "learning_rate": 5.596330993284937e-11,
      "loss": 2.1044,
      "step": 1475
    },
    {
      "epoch": 0.9996613613274636,
      "eval_loss": 2.1087405681610107,
      "eval_runtime": 85.1679,
      "eval_samples_per_second": 15.323,
      "eval_steps_per_second": 1.926,
      "step": 1476
    },
    {
      "epoch": 0.9996613613274636,
      "step": 1476,
      "total_flos": 3.002086743311647e+17,
      "train_loss": 2.1322910175090883,
      "train_runtime": 6591.5154,
      "train_samples_per_second": 3.583,
      "train_steps_per_second": 0.224
    }
  ],
  "logging_steps": 5,
  "max_steps": 1476,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.002086743311647e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}