{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.016160440856826573,
  "eval_steps": 500,
  "global_step": 2000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 8.080220428413287e-05,
      "grad_norm": 4661457.0,
      "learning_rate": 4.040404040404041e-08,
      "loss": 678012.1,
      "step": 10
    },
    {
      "epoch": 0.00016160440856826573,
      "grad_norm": 9865970.0,
      "learning_rate": 8.080808080808082e-08,
      "loss": 841980.1,
      "step": 20
    },
    {
      "epoch": 0.00024240661285239863,
      "grad_norm": 10027139.0,
      "learning_rate": 1.2121212121212122e-07,
      "loss": 816764.25,
      "step": 30
    },
    {
      "epoch": 0.00032320881713653147,
      "grad_norm": 16910428.0,
      "learning_rate": 1.6161616161616163e-07,
      "loss": 961656.7,
      "step": 40
    },
    {
      "epoch": 0.00040401102142066436,
      "grad_norm": 5068422.0,
      "learning_rate": 2.0202020202020202e-07,
      "loss": 679086.95,
      "step": 50
    },
    {
      "epoch": 0.00048481322570479725,
      "grad_norm": 2293057.5,
      "learning_rate": 2.4242424242424244e-07,
      "loss": 949613.2,
      "step": 60
    },
    {
      "epoch": 0.0005656154299889301,
      "grad_norm": 7849363.5,
      "learning_rate": 2.8282828282828283e-07,
      "loss": 859798.0,
      "step": 70
    },
    {
      "epoch": 0.0006464176342730629,
      "grad_norm": 2669786.75,
      "learning_rate": 3.2323232323232327e-07,
      "loss": 863443.9,
      "step": 80
    },
    {
      "epoch": 0.0007272198385571959,
      "grad_norm": 6021406.5,
      "learning_rate": 3.6363636363636366e-07,
      "loss": 734401.3,
      "step": 90
    },
    {
      "epoch": 0.0008080220428413287,
      "grad_norm": 5000674.5,
      "learning_rate": 4.0404040404040405e-07,
      "loss": 526539.8,
      "step": 100
    },
    {
      "epoch": 0.0008888242471254616,
      "grad_norm": 4362635.5,
      "learning_rate": 4.444444444444445e-07,
      "loss": 422810.45,
      "step": 110
    },
    {
      "epoch": 0.0009696264514095945,
      "grad_norm": 5343355.0,
      "learning_rate": 4.848484848484849e-07,
      "loss": 490438.55,
      "step": 120
    },
    {
      "epoch": 0.0010504286556937273,
      "grad_norm": 11366241.0,
      "learning_rate": 5.252525252525253e-07,
      "loss": 704809.3,
      "step": 130
    },
    {
      "epoch": 0.0011312308599778602,
      "grad_norm": 5520147.0,
      "learning_rate": 5.656565656565657e-07,
      "loss": 526687.5,
      "step": 140
    },
    {
      "epoch": 0.001212033064261993,
      "grad_norm": 3778741.0,
      "learning_rate": 6.060606060606061e-07,
      "loss": 322092.525,
      "step": 150
    },
    {
      "epoch": 0.0012928352685461259,
      "grad_norm": 3378473.0,
      "learning_rate": 6.464646464646465e-07,
      "loss": 516098.85,
      "step": 160
    },
    {
      "epoch": 0.0013736374728302587,
      "grad_norm": 4408531.0,
      "learning_rate": 6.868686868686869e-07,
      "loss": 410817.35,
      "step": 170
    },
    {
      "epoch": 0.0014544396771143918,
      "grad_norm": 766378.375,
      "learning_rate": 7.272727272727273e-07,
      "loss": 197623.1625,
      "step": 180
    },
    {
      "epoch": 0.0015352418813985246,
      "grad_norm": 5468394.5,
      "learning_rate": 7.676767676767678e-07,
      "loss": 238133.6,
      "step": 190
    },
    {
      "epoch": 0.0016160440856826574,
      "grad_norm": 7980194.5,
      "learning_rate": 8.080808080808081e-07,
      "loss": 182498.725,
      "step": 200
    },
    {
      "epoch": 0.0016968462899667903,
      "grad_norm": 737060.125,
      "learning_rate": 8.484848484848486e-07,
      "loss": 122880.075,
      "step": 210
    },
    {
      "epoch": 0.0017776484942509231,
      "grad_norm": 476725.8125,
      "learning_rate": 8.88888888888889e-07,
      "loss": 100649.7563,
      "step": 220
    },
    {
      "epoch": 0.001858450698535056,
      "grad_norm": 353791.875,
      "learning_rate": 9.292929292929294e-07,
      "loss": 58150.4563,
      "step": 230
    },
    {
      "epoch": 0.001939252902819189,
      "grad_norm": 684820.5625,
      "learning_rate": 9.696969696969698e-07,
      "loss": 44040.125,
      "step": 240
    },
    {
      "epoch": 0.0020200551071033216,
      "grad_norm": 228699.796875,
      "learning_rate": 1.0101010101010103e-06,
      "loss": 21505.6156,
      "step": 250
    },
    {
      "epoch": 0.0021008573113874547,
      "grad_norm": 91799.625,
      "learning_rate": 1.0505050505050506e-06,
      "loss": 11209.5648,
      "step": 260
    },
    {
      "epoch": 0.0021816595156715873,
      "grad_norm": 64077.56640625,
      "learning_rate": 1.090909090909091e-06,
      "loss": 5405.107,
      "step": 270
    },
    {
      "epoch": 0.0022624617199557204,
      "grad_norm": 68140.5078125,
      "learning_rate": 1.1313131313131313e-06,
      "loss": 2509.3971,
      "step": 280
    },
    {
      "epoch": 0.0023432639242398534,
      "grad_norm": 38560.15625,
      "learning_rate": 1.1717171717171719e-06,
      "loss": 1295.3724,
      "step": 290
    },
    {
      "epoch": 0.002424066128523986,
      "grad_norm": 2864.86474609375,
      "learning_rate": 1.2121212121212122e-06,
      "loss": 923.2405,
      "step": 300
    },
    {
      "epoch": 0.002504868332808119,
      "grad_norm": 2044.932861328125,
      "learning_rate": 1.2525252525252527e-06,
      "loss": 950.31,
      "step": 310
    },
    {
      "epoch": 0.0025856705370922517,
      "grad_norm": 2294.186279296875,
      "learning_rate": 1.292929292929293e-06,
      "loss": 830.1033,
      "step": 320
    },
    {
      "epoch": 0.002666472741376385,
      "grad_norm": 941.1571044921875,
      "learning_rate": 1.3333333333333334e-06,
      "loss": 711.3893,
      "step": 330
    },
    {
      "epoch": 0.0027472749456605174,
      "grad_norm": 6470.8056640625,
      "learning_rate": 1.3737373737373738e-06,
      "loss": 746.1014,
      "step": 340
    },
    {
      "epoch": 0.0028280771499446505,
      "grad_norm": 2412.6904296875,
      "learning_rate": 1.4141414141414143e-06,
      "loss": 774.8737,
      "step": 350
    },
    {
      "epoch": 0.0029088793542287835,
      "grad_norm": 2536.60791015625,
      "learning_rate": 1.4545454545454546e-06,
      "loss": 832.4265,
      "step": 360
    },
    {
      "epoch": 0.002989681558512916,
      "grad_norm": 5404.72265625,
      "learning_rate": 1.4949494949494952e-06,
      "loss": 676.5023,
      "step": 370
    },
    {
      "epoch": 0.003070483762797049,
      "grad_norm": 1271.8477783203125,
      "learning_rate": 1.5353535353535355e-06,
      "loss": 518.433,
      "step": 380
    },
    {
      "epoch": 0.003151285967081182,
      "grad_norm": 736.135009765625,
      "learning_rate": 1.5757575757575759e-06,
      "loss": 768.4572,
      "step": 390
    },
    {
      "epoch": 0.003232088171365315,
      "grad_norm": 1238.1878662109375,
      "learning_rate": 1.6161616161616162e-06,
      "loss": 696.9558,
      "step": 400
    },
    {
      "epoch": 0.0033128903756494475,
      "grad_norm": 1593.6436767578125,
      "learning_rate": 1.6565656565656565e-06,
      "loss": 670.7958,
      "step": 410
    },
    {
      "epoch": 0.0033936925799335806,
      "grad_norm": 819.6708374023438,
      "learning_rate": 1.6969696969696973e-06,
      "loss": 581.9092,
      "step": 420
    },
    {
      "epoch": 0.0034744947842177136,
      "grad_norm": 775.354248046875,
      "learning_rate": 1.7373737373737376e-06,
      "loss": 744.2987,
      "step": 430
    },
    {
      "epoch": 0.0035552969885018462,
      "grad_norm": 744.1176147460938,
      "learning_rate": 1.777777777777778e-06,
      "loss": 665.0154,
      "step": 440
    },
    {
      "epoch": 0.0036360991927859793,
      "grad_norm": 1123.35986328125,
      "learning_rate": 1.818181818181818e-06,
      "loss": 681.7526,
      "step": 450
    },
    {
      "epoch": 0.003716901397070112,
      "grad_norm": 2373.274658203125,
      "learning_rate": 1.8585858585858588e-06,
      "loss": 654.0196,
      "step": 460
    },
    {
      "epoch": 0.003797703601354245,
      "grad_norm": 1098.4403076171875,
      "learning_rate": 1.8989898989898992e-06,
      "loss": 657.3389,
      "step": 470
    },
    {
      "epoch": 0.003878505805638378,
      "grad_norm": 1008.40966796875,
      "learning_rate": 1.9393939393939395e-06,
      "loss": 564.041,
      "step": 480
    },
    {
      "epoch": 0.003959308009922511,
      "grad_norm": 1227.1414794921875,
      "learning_rate": 1.9797979797979796e-06,
      "loss": 614.5699,
      "step": 490
    },
    {
      "epoch": 0.004040110214206643,
      "grad_norm": 57328.9609375,
      "learning_rate": 2.0202020202020206e-06,
      "loss": 596.2117,
      "step": 500
    },
    {
      "epoch": 0.004120912418490777,
      "grad_norm": 2744.44873046875,
      "learning_rate": 2.0606060606060607e-06,
      "loss": 676.9374,
      "step": 510
    },
    {
      "epoch": 0.004201714622774909,
      "grad_norm": 667.0958862304688,
      "learning_rate": 2.1010101010101013e-06,
      "loss": 812.1784,
      "step": 520
    },
    {
      "epoch": 0.004282516827059042,
      "grad_norm": 1447.203857421875,
      "learning_rate": 2.1414141414141414e-06,
      "loss": 640.1749,
      "step": 530
    },
    {
      "epoch": 0.004363319031343175,
      "grad_norm": 1186.56982421875,
      "learning_rate": 2.181818181818182e-06,
      "loss": 665.659,
      "step": 540
    },
    {
      "epoch": 0.004444121235627308,
      "grad_norm": 1263.9395751953125,
      "learning_rate": 2.2222222222222225e-06,
      "loss": 696.9437,
      "step": 550
    },
    {
      "epoch": 0.004524923439911441,
      "grad_norm": 1353.597412109375,
      "learning_rate": 2.2626262626262626e-06,
      "loss": 654.0818,
      "step": 560
    },
    {
      "epoch": 0.004605725644195573,
      "grad_norm": 876.721923828125,
      "learning_rate": 2.303030303030303e-06,
      "loss": 683.512,
      "step": 570
    },
    {
      "epoch": 0.004686527848479707,
      "grad_norm": 1866.14697265625,
      "learning_rate": 2.3434343434343437e-06,
      "loss": 727.845,
      "step": 580
    },
    {
      "epoch": 0.0047673300527638395,
      "grad_norm": 667.1629028320312,
      "learning_rate": 2.383838383838384e-06,
      "loss": 625.2918,
      "step": 590
    },
    {
      "epoch": 0.004848132257047972,
      "grad_norm": 614.8545532226562,
      "learning_rate": 2.4242424242424244e-06,
      "loss": 600.5201,
      "step": 600
    },
    {
      "epoch": 0.004928934461332105,
      "grad_norm": 879.8363647460938,
      "learning_rate": 2.4646464646464645e-06,
      "loss": 509.5772,
      "step": 610
    },
    {
      "epoch": 0.005009736665616238,
      "grad_norm": 1344.7303466796875,
      "learning_rate": 2.5050505050505055e-06,
      "loss": 639.9966,
      "step": 620
    },
    {
      "epoch": 0.005090538869900371,
      "grad_norm": 1223.520751953125,
      "learning_rate": 2.5454545454545456e-06,
      "loss": 706.7377,
      "step": 630
    },
    {
      "epoch": 0.0051713410741845035,
      "grad_norm": 1871.1324462890625,
      "learning_rate": 2.585858585858586e-06,
      "loss": 712.2218,
      "step": 640
    },
    {
      "epoch": 0.005252143278468637,
      "grad_norm": 804.082763671875,
      "learning_rate": 2.6262626262626263e-06,
      "loss": 666.623,
      "step": 650
    },
    {
      "epoch": 0.00533294548275277,
      "grad_norm": 10994.2314453125,
      "learning_rate": 2.666666666666667e-06,
      "loss": 726.5861,
      "step": 660
    },
    {
      "epoch": 0.005413747687036902,
      "grad_norm": 892.67919921875,
      "learning_rate": 2.7070707070707074e-06,
      "loss": 596.617,
      "step": 670
    },
    {
      "epoch": 0.005494549891321035,
      "grad_norm": 681.8887939453125,
      "learning_rate": 2.7474747474747475e-06,
      "loss": 496.8398,
      "step": 680
    },
    {
      "epoch": 0.005575352095605168,
      "grad_norm": 1190.68310546875,
      "learning_rate": 2.787878787878788e-06,
      "loss": 684.7522,
      "step": 690
    },
    {
      "epoch": 0.005656154299889301,
      "grad_norm": 1649.9376220703125,
      "learning_rate": 2.8282828282828286e-06,
      "loss": 717.8954,
      "step": 700
    },
    {
      "epoch": 0.0057369565041734336,
      "grad_norm": 2140.240234375,
      "learning_rate": 2.8686868686868687e-06,
      "loss": 633.3918,
      "step": 710
    },
    {
      "epoch": 0.005817758708457567,
      "grad_norm": 1765.3192138671875,
      "learning_rate": 2.9090909090909093e-06,
      "loss": 656.2145,
      "step": 720
    },
    {
      "epoch": 0.0058985609127417,
      "grad_norm": 2076.052001953125,
      "learning_rate": 2.9494949494949494e-06,
      "loss": 620.9487,
      "step": 730
    },
    {
      "epoch": 0.005979363117025832,
      "grad_norm": 817.6283569335938,
      "learning_rate": 2.9898989898989904e-06,
      "loss": 427.1178,
      "step": 740
    },
    {
      "epoch": 0.006060165321309965,
      "grad_norm": 1084.86767578125,
      "learning_rate": 3.0303030303030305e-06,
      "loss": 560.9343,
      "step": 750
    },
    {
      "epoch": 0.006140967525594098,
      "grad_norm": 786.8311767578125,
      "learning_rate": 3.070707070707071e-06,
      "loss": 520.7236,
      "step": 760
    },
    {
      "epoch": 0.006221769729878231,
      "grad_norm": 4558.533203125,
      "learning_rate": 3.111111111111111e-06,
      "loss": 555.48,
      "step": 770
    },
    {
      "epoch": 0.006302571934162364,
      "grad_norm": 785.0099487304688,
      "learning_rate": 3.1515151515151517e-06,
      "loss": 514.0267,
      "step": 780
    },
    {
      "epoch": 0.006383374138446497,
      "grad_norm": 1063.466796875,
      "learning_rate": 3.191919191919192e-06,
      "loss": 592.0211,
      "step": 790
    },
    {
      "epoch": 0.00646417634273063,
      "grad_norm": 2228.42041015625,
      "learning_rate": 3.2323232323232324e-06,
      "loss": 613.0331,
      "step": 800
    },
    {
      "epoch": 0.006544978547014762,
      "grad_norm": 1026.750732421875,
      "learning_rate": 3.2727272727272733e-06,
      "loss": 639.7108,
      "step": 810
    },
    {
      "epoch": 0.006625780751298895,
      "grad_norm": 2451.453369140625,
      "learning_rate": 3.313131313131313e-06,
      "loss": 659.9323,
      "step": 820
    },
    {
      "epoch": 0.0067065829555830285,
      "grad_norm": 1183.8045654296875,
      "learning_rate": 3.3535353535353536e-06,
      "loss": 601.1871,
      "step": 830
    },
    {
      "epoch": 0.006787385159867161,
      "grad_norm": 1861.18701171875,
      "learning_rate": 3.3939393939393946e-06,
      "loss": 589.3404,
      "step": 840
    },
    {
      "epoch": 0.006868187364151294,
      "grad_norm": 1340.5020751953125,
      "learning_rate": 3.4343434343434343e-06,
      "loss": 588.7653,
      "step": 850
    },
    {
      "epoch": 0.006948989568435427,
      "grad_norm": 8932.84375,
      "learning_rate": 3.4747474747474752e-06,
      "loss": 667.0873,
      "step": 860
    },
    {
      "epoch": 0.00702979177271956,
      "grad_norm": 1353.702392578125,
      "learning_rate": 3.515151515151515e-06,
      "loss": 574.4604,
      "step": 870
    },
    {
      "epoch": 0.0071105939770036925,
      "grad_norm": 1281.541748046875,
      "learning_rate": 3.555555555555556e-06,
      "loss": 683.6188,
      "step": 880
    },
    {
      "epoch": 0.007191396181287825,
      "grad_norm": 3347.4111328125,
      "learning_rate": 3.5959595959595965e-06,
      "loss": 591.1179,
      "step": 890
    },
    {
      "epoch": 0.007272198385571959,
      "grad_norm": 832.118896484375,
      "learning_rate": 3.636363636363636e-06,
      "loss": 509.1054,
      "step": 900
    },
    {
      "epoch": 0.007353000589856091,
      "grad_norm": 3215.1875,
      "learning_rate": 3.676767676767677e-06,
      "loss": 634.5958,
      "step": 910
    },
    {
      "epoch": 0.007433802794140224,
      "grad_norm": 1072.3865966796875,
      "learning_rate": 3.7171717171717177e-06,
      "loss": 525.3159,
      "step": 920
    },
    {
      "epoch": 0.007514604998424357,
      "grad_norm": 1002.2363891601562,
      "learning_rate": 3.757575757575758e-06,
      "loss": 659.7474,
      "step": 930
    },
    {
      "epoch": 0.00759540720270849,
      "grad_norm": 6355.693359375,
      "learning_rate": 3.7979797979797984e-06,
      "loss": 636.0396,
      "step": 940
    },
    {
      "epoch": 0.007676209406992623,
      "grad_norm": 1635.6080322265625,
      "learning_rate": 3.8383838383838385e-06,
      "loss": 624.7128,
      "step": 950
    },
    {
      "epoch": 0.007757011611276756,
      "grad_norm": 1284.3531494140625,
      "learning_rate": 3.878787878787879e-06,
      "loss": 637.0934,
      "step": 960
    },
    {
      "epoch": 0.007837813815560889,
      "grad_norm": 2921.176025390625,
      "learning_rate": 3.9191919191919196e-06,
      "loss": 648.8942,
      "step": 970
    },
    {
      "epoch": 0.007918616019845021,
      "grad_norm": 845.0542602539062,
      "learning_rate": 3.959595959595959e-06,
      "loss": 536.0546,
      "step": 980
    },
    {
      "epoch": 0.007999418224129154,
      "grad_norm": 986.8812866210938,
      "learning_rate": 4.000000000000001e-06,
      "loss": 643.056,
      "step": 990
    },
    {
      "epoch": 0.008080220428413287,
      "grad_norm": 748.7238159179688,
      "learning_rate": 4.040404040404041e-06,
      "loss": 587.7603,
      "step": 1000
    },
    {
      "epoch": 0.00816102263269742,
      "grad_norm": 4340.18017578125,
      "learning_rate": 4.080808080808081e-06,
      "loss": 545.2501,
      "step": 1010
    },
    {
      "epoch": 0.008241824836981554,
      "grad_norm": 1614.0966796875,
      "learning_rate": 4.1212121212121215e-06,
      "loss": 652.8924,
      "step": 1020
    },
    {
      "epoch": 0.008322627041265686,
      "grad_norm": 10709.900390625,
      "learning_rate": 4.161616161616161e-06,
      "loss": 598.342,
      "step": 1030
    },
    {
      "epoch": 0.008403429245549819,
      "grad_norm": 870.67578125,
      "learning_rate": 4.2020202020202026e-06,
      "loss": 612.1897,
      "step": 1040
    },
    {
      "epoch": 0.008484231449833951,
      "grad_norm": 686.4441528320312,
      "learning_rate": 4.242424242424243e-06,
      "loss": 564.3604,
      "step": 1050
    },
    {
      "epoch": 0.008565033654118084,
      "grad_norm": 800.3753051757812,
      "learning_rate": 4.282828282828283e-06,
      "loss": 515.465,
      "step": 1060
    },
    {
      "epoch": 0.008645835858402217,
      "grad_norm": 1612.2467041015625,
      "learning_rate": 4.323232323232323e-06,
      "loss": 615.4043,
      "step": 1070
    },
    {
      "epoch": 0.00872663806268635,
      "grad_norm": 1488.6268310546875,
      "learning_rate": 4.363636363636364e-06,
      "loss": 584.7924,
      "step": 1080
    },
    {
      "epoch": 0.008807440266970484,
      "grad_norm": 1085.609619140625,
      "learning_rate": 4.4040404040404044e-06,
      "loss": 568.4938,
      "step": 1090
    },
    {
      "epoch": 0.008888242471254616,
      "grad_norm": 1548.931884765625,
      "learning_rate": 4.444444444444445e-06,
      "loss": 515.5535,
      "step": 1100
    },
    {
      "epoch": 0.008969044675538749,
      "grad_norm": 1010.5916137695312,
      "learning_rate": 4.484848484848485e-06,
      "loss": 496.3722,
      "step": 1110
    },
    {
      "epoch": 0.009049846879822881,
      "grad_norm": 1959.3551025390625,
      "learning_rate": 4.525252525252525e-06,
      "loss": 697.1435,
      "step": 1120
    },
    {
      "epoch": 0.009130649084107014,
      "grad_norm": 2370.032470703125,
      "learning_rate": 4.565656565656566e-06,
      "loss": 541.7345,
      "step": 1130
    },
    {
      "epoch": 0.009211451288391147,
      "grad_norm": 1848.2596435546875,
      "learning_rate": 4.606060606060606e-06,
      "loss": 539.6271,
      "step": 1140
    },
    {
      "epoch": 0.00929225349267528,
      "grad_norm": 2245.12548828125,
      "learning_rate": 4.646464646464647e-06,
      "loss": 485.2028,
      "step": 1150
    },
    {
      "epoch": 0.009373055696959414,
      "grad_norm": 1303.244140625,
      "learning_rate": 4.6868686868686874e-06,
      "loss": 552.4869,
      "step": 1160
    },
    {
      "epoch": 0.009453857901243546,
      "grad_norm": 813.8889770507812,
      "learning_rate": 4.727272727272727e-06,
      "loss": 566.426,
      "step": 1170
    },
    {
      "epoch": 0.009534660105527679,
      "grad_norm": 2155.474609375,
      "learning_rate": 4.767676767676768e-06,
      "loss": 694.8829,
      "step": 1180
    },
    {
      "epoch": 0.009615462309811812,
      "grad_norm": 3313.629150390625,
      "learning_rate": 4.808080808080808e-06,
      "loss": 541.8082,
      "step": 1190
    },
    {
      "epoch": 0.009696264514095944,
      "grad_norm": 830.4761352539062,
      "learning_rate": 4.848484848484849e-06,
      "loss": 633.2839,
      "step": 1200
    },
    {
      "epoch": 0.009777066718380077,
      "grad_norm": 2314.525634765625,
      "learning_rate": 4.888888888888889e-06,
      "loss": 465.8495,
      "step": 1210
    },
    {
      "epoch": 0.00985786892266421,
      "grad_norm": 1821.2579345703125,
      "learning_rate": 4.929292929292929e-06,
      "loss": 543.8752,
      "step": 1220
    },
    {
      "epoch": 0.009938671126948344,
      "grad_norm": 471.08380126953125,
      "learning_rate": 4.96969696969697e-06,
      "loss": 405.759,
      "step": 1230
    },
    {
      "epoch": 0.010019473331232476,
      "grad_norm": 1476.8787841796875,
      "learning_rate": 5.010101010101011e-06,
      "loss": 617.1836,
      "step": 1240
    },
    {
      "epoch": 0.010100275535516609,
      "grad_norm": 667.0234985351562,
      "learning_rate": 5.050505050505051e-06,
      "loss": 513.511,
      "step": 1250
    },
    {
      "epoch": 0.010181077739800742,
      "grad_norm": 613.2576904296875,
      "learning_rate": 5.090909090909091e-06,
      "loss": 468.2697,
      "step": 1260
    },
    {
      "epoch": 0.010261879944084874,
      "grad_norm": 1321.9991455078125,
      "learning_rate": 5.131313131313131e-06,
      "loss": 476.6512,
      "step": 1270
    },
    {
      "epoch": 0.010342682148369007,
      "grad_norm": 840.0568237304688,
      "learning_rate": 5.171717171717172e-06,
      "loss": 607.1966,
      "step": 1280
    },
    {
      "epoch": 0.01042348435265314,
      "grad_norm": 1686.292724609375,
      "learning_rate": 5.212121212121213e-06,
      "loss": 538.6932,
      "step": 1290
    },
    {
      "epoch": 0.010504286556937274,
      "grad_norm": 2623.5302734375,
      "learning_rate": 5.2525252525252526e-06,
      "loss": 511.9114,
      "step": 1300
    },
    {
      "epoch": 0.010585088761221407,
      "grad_norm": 1300.63671875,
      "learning_rate": 5.292929292929293e-06,
      "loss": 573.5399,
      "step": 1310
    },
    {
      "epoch": 0.01066589096550554,
      "grad_norm": 828.8121948242188,
      "learning_rate": 5.333333333333334e-06,
      "loss": 617.9385,
      "step": 1320
    },
    {
      "epoch": 0.010746693169789672,
      "grad_norm": 965.5888671875,
      "learning_rate": 5.373737373737374e-06,
      "loss": 428.5168,
      "step": 1330
    },
    {
      "epoch": 0.010827495374073804,
      "grad_norm": 1123.3818359375,
      "learning_rate": 5.414141414141415e-06,
      "loss": 582.797,
      "step": 1340
    },
    {
      "epoch": 0.010908297578357937,
      "grad_norm": 4098.775390625,
      "learning_rate": 5.4545454545454545e-06,
      "loss": 633.2868,
      "step": 1350
    },
    {
      "epoch": 0.01098909978264207,
      "grad_norm": 792.64306640625,
      "learning_rate": 5.494949494949495e-06,
      "loss": 688.7412,
      "step": 1360
    },
    {
      "epoch": 0.011069901986926204,
      "grad_norm": 1111.2113037109375,
      "learning_rate": 5.5353535353535355e-06,
      "loss": 473.5282,
      "step": 1370
    },
    {
      "epoch": 0.011150704191210337,
      "grad_norm": 5420.9765625,
      "learning_rate": 5.575757575757576e-06,
      "loss": 516.3693,
      "step": 1380
    },
    {
      "epoch": 0.01123150639549447,
      "grad_norm": 1263.883056640625,
      "learning_rate": 5.616161616161617e-06,
      "loss": 457.2747,
      "step": 1390
    },
    {
      "epoch": 0.011312308599778602,
      "grad_norm": 944.7489624023438,
      "learning_rate": 5.656565656565657e-06,
      "loss": 539.8824,
      "step": 1400
    },
    {
      "epoch": 0.011393110804062734,
      "grad_norm": 562.0636596679688,
      "learning_rate": 5.696969696969697e-06,
      "loss": 455.684,
      "step": 1410
    },
    {
      "epoch": 0.011473913008346867,
      "grad_norm": 1293.833984375,
      "learning_rate": 5.7373737373737374e-06,
      "loss": 527.1053,
      "step": 1420
    },
    {
      "epoch": 0.011554715212631,
      "grad_norm": 1857.3572998046875,
      "learning_rate": 5.777777777777778e-06,
      "loss": 519.1355,
      "step": 1430
    },
    {
      "epoch": 0.011635517416915134,
      "grad_norm": 1110.5677490234375,
      "learning_rate": 5.8181818181818185e-06,
      "loss": 470.4913,
      "step": 1440
    },
    {
      "epoch": 0.011716319621199267,
      "grad_norm": 1656.9898681640625,
      "learning_rate": 5.858585858585859e-06,
      "loss": 570.9559,
      "step": 1450
    },
    {
      "epoch": 0.0117971218254834,
      "grad_norm": 830.130126953125,
      "learning_rate": 5.898989898989899e-06,
      "loss": 551.6251,
      "step": 1460
    },
    {
      "epoch": 0.011877924029767532,
      "grad_norm": 718.1753540039062,
      "learning_rate": 5.93939393939394e-06,
      "loss": 495.0048,
      "step": 1470
    },
    {
      "epoch": 0.011958726234051665,
      "grad_norm": 2903.817138671875,
      "learning_rate": 5.979797979797981e-06,
      "loss": 742.8938,
      "step": 1480
    },
    {
      "epoch": 0.012039528438335797,
      "grad_norm": 893.6466674804688,
      "learning_rate": 6.0202020202020204e-06,
      "loss": 542.4396,
      "step": 1490
    },
    {
      "epoch": 0.01212033064261993,
      "grad_norm": 1576.75390625,
      "learning_rate": 6.060606060606061e-06,
      "loss": 513.3051,
      "step": 1500
    },
    {
      "epoch": 0.012201132846904064,
      "grad_norm": 1925.69140625,
      "learning_rate": 6.101010101010101e-06,
      "loss": 450.3941,
      "step": 1510
    },
    {
      "epoch": 0.012281935051188197,
      "grad_norm": 1998.0458984375,
      "learning_rate": 6.141414141414142e-06,
      "loss": 635.4773,
      "step": 1520
    },
    {
      "epoch": 0.01236273725547233,
      "grad_norm": 1258.9659423828125,
      "learning_rate": 6.181818181818183e-06,
      "loss": 595.3451,
      "step": 1530
    },
    {
      "epoch": 0.012443539459756462,
      "grad_norm": 1726.843017578125,
      "learning_rate": 6.222222222222222e-06,
      "loss": 556.3297,
      "step": 1540
    },
    {
      "epoch": 0.012524341664040595,
      "grad_norm": 2486.63134765625,
      "learning_rate": 6.262626262626263e-06,
      "loss": 647.7714,
      "step": 1550
    },
    {
      "epoch": 0.012605143868324727,
      "grad_norm": 1644.4110107421875,
      "learning_rate": 6.303030303030303e-06,
      "loss": 611.4104,
      "step": 1560
    },
    {
      "epoch": 0.01268594607260886,
      "grad_norm": 2169.43408203125,
      "learning_rate": 6.343434343434344e-06,
      "loss": 445.9718,
      "step": 1570
    },
    {
      "epoch": 0.012766748276892994,
      "grad_norm": 1554.9425048828125,
      "learning_rate": 6.383838383838384e-06,
      "loss": 561.5079,
      "step": 1580
    },
    {
      "epoch": 0.012847550481177127,
      "grad_norm": 891.0722045898438,
      "learning_rate": 6.424242424242424e-06,
      "loss": 510.2931,
      "step": 1590
    },
    {
      "epoch": 0.01292835268546126,
      "grad_norm": 22828.455078125,
      "learning_rate": 6.464646464646465e-06,
      "loss": 552.862,
      "step": 1600
    },
    {
      "epoch": 0.013009154889745392,
      "grad_norm": 722.2283325195312,
      "learning_rate": 6.505050505050505e-06,
      "loss": 512.3786,
      "step": 1610
    },
    {
      "epoch": 0.013089957094029525,
      "grad_norm": 1105.0716552734375,
      "learning_rate": 6.545454545454547e-06,
      "loss": 532.3522,
      "step": 1620
    },
    {
      "epoch": 0.013170759298313657,
      "grad_norm": 1146.0550537109375,
      "learning_rate": 6.5858585858585856e-06,
      "loss": 647.0639,
      "step": 1630
    },
    {
      "epoch": 0.01325156150259779,
      "grad_norm": 2632.366943359375,
      "learning_rate": 6.626262626262626e-06,
      "loss": 613.0408,
      "step": 1640
    },
    {
      "epoch": 0.013332363706881924,
      "grad_norm": 1236.950439453125,
      "learning_rate": 6.666666666666667e-06,
      "loss": 614.0946,
      "step": 1650
    },
    {
      "epoch": 0.013413165911166057,
      "grad_norm": 973.5294799804688,
      "learning_rate": 6.707070707070707e-06,
      "loss": 383.6397,
      "step": 1660
    },
    {
      "epoch": 0.01349396811545019,
      "grad_norm": 1101.2989501953125,
      "learning_rate": 6.747474747474749e-06,
      "loss": 556.5678,
      "step": 1670
    },
    {
      "epoch": 0.013574770319734322,
      "grad_norm": 694.1087646484375,
      "learning_rate": 6.787878787878789e-06,
      "loss": 415.0727,
      "step": 1680
    },
    {
      "epoch": 0.013655572524018455,
      "grad_norm": 1031.21240234375,
      "learning_rate": 6.828282828282828e-06,
      "loss": 508.7032,
      "step": 1690
    },
    {
      "epoch": 0.013736374728302587,
      "grad_norm": 734.2860717773438,
      "learning_rate": 6.8686868686868685e-06,
      "loss": 482.0532,
      "step": 1700
    },
    {
      "epoch": 0.01381717693258672,
      "grad_norm": 879.1790771484375,
      "learning_rate": 6.909090909090909e-06,
      "loss": 655.4877,
      "step": 1710
    },
    {
      "epoch": 0.013897979136870854,
      "grad_norm": 1699.49560546875,
      "learning_rate": 6.9494949494949505e-06,
      "loss": 400.7212,
      "step": 1720
    },
    {
      "epoch": 0.013978781341154987,
      "grad_norm": 1901.2413330078125,
      "learning_rate": 6.989898989898991e-06,
      "loss": 496.4101,
      "step": 1730
    },
    {
      "epoch": 0.01405958354543912,
      "grad_norm": 1368.836669921875,
      "learning_rate": 7.03030303030303e-06,
      "loss": 520.5295,
      "step": 1740
    },
    {
      "epoch": 0.014140385749723252,
      "grad_norm": 1587.8297119140625,
      "learning_rate": 7.0707070707070704e-06,
      "loss": 397.7875,
      "step": 1750
    },
    {
      "epoch": 0.014221187954007385,
      "grad_norm": 1054.773681640625,
      "learning_rate": 7.111111111111112e-06,
      "loss": 456.9818,
      "step": 1760
    },
    {
      "epoch": 0.014301990158291518,
      "grad_norm": 3368.4013671875,
      "learning_rate": 7.151515151515152e-06,
      "loss": 397.1987,
      "step": 1770
    },
    {
      "epoch": 0.01438279236257565,
      "grad_norm": 1371.4473876953125,
      "learning_rate": 7.191919191919193e-06,
      "loss": 433.1569,
      "step": 1780
    },
    {
      "epoch": 0.014463594566859785,
      "grad_norm": 1193.6456298828125,
      "learning_rate": 7.232323232323232e-06,
      "loss": 368.5404,
      "step": 1790
    },
    {
      "epoch": 0.014544396771143917,
      "grad_norm": 976.5448608398438,
      "learning_rate": 7.272727272727272e-06,
      "loss": 427.9085,
      "step": 1800
    },
    {
      "epoch": 0.01462519897542805,
      "grad_norm": 665.8095092773438,
      "learning_rate": 7.313131313131314e-06,
      "loss": 461.1592,
      "step": 1810
    },
    {
      "epoch": 0.014706001179712182,
      "grad_norm": 1352.773681640625,
      "learning_rate": 7.353535353535354e-06,
      "loss": 421.6953,
      "step": 1820
    },
    {
      "epoch": 0.014786803383996315,
      "grad_norm": 1284.1273193359375,
      "learning_rate": 7.393939393939395e-06,
      "loss": 486.5889,
      "step": 1830
    },
    {
      "epoch": 0.014867605588280448,
      "grad_norm": 3249.41650390625,
      "learning_rate": 7.434343434343435e-06,
      "loss": 534.9857,
      "step": 1840
    },
    {
      "epoch": 0.01494840779256458,
      "grad_norm": 1162.9168701171875,
      "learning_rate": 7.474747474747475e-06,
      "loss": 684.8136,
      "step": 1850
    },
    {
      "epoch": 0.015029209996848715,
      "grad_norm": 2700.0771484375,
      "learning_rate": 7.515151515151516e-06,
      "loss": 438.9466,
      "step": 1860
    },
    {
      "epoch": 0.015110012201132847,
      "grad_norm": 1472.51171875,
      "learning_rate": 7.555555555555556e-06,
      "loss": 541.1812,
      "step": 1870
    },
    {
      "epoch": 0.01519081440541698,
      "grad_norm": 763.5363159179688,
      "learning_rate": 7.595959595959597e-06,
      "loss": 574.8496,
      "step": 1880
    },
    {
      "epoch": 0.015271616609701113,
      "grad_norm": 1097.047607421875,
      "learning_rate": 7.636363636363638e-06,
      "loss": 495.2874,
      "step": 1890
    },
    {
      "epoch": 0.015352418813985245,
      "grad_norm": 1086.6981201171875,
      "learning_rate": 7.676767676767677e-06,
      "loss": 494.5146,
      "step": 1900
    },
    {
      "epoch": 0.015433221018269378,
      "grad_norm": 1003.9575805664062,
      "learning_rate": 7.717171717171717e-06,
      "loss": 493.7404,
      "step": 1910
    },
    {
      "epoch": 0.015514023222553512,
      "grad_norm": 1008.3009643554688,
      "learning_rate": 7.757575757575758e-06,
      "loss": 507.6875,
      "step": 1920
    },
    {
      "epoch": 0.015594825426837645,
      "grad_norm": 2552.92626953125,
      "learning_rate": 7.797979797979799e-06,
      "loss": 420.8823,
      "step": 1930
    },
    {
      "epoch": 0.015675627631121777,
      "grad_norm": 735.5524291992188,
      "learning_rate": 7.838383838383839e-06,
      "loss": 519.3292,
      "step": 1940
    },
    {
      "epoch": 0.01575642983540591,
      "grad_norm": 1559.79296875,
      "learning_rate": 7.878787878787878e-06,
      "loss": 468.1747,
      "step": 1950
    },
    {
      "epoch": 0.015837232039690043,
      "grad_norm": 1734.436767578125,
      "learning_rate": 7.919191919191919e-06,
      "loss": 417.3558,
      "step": 1960
    },
    {
      "epoch": 0.015918034243974175,
      "grad_norm": 2985.763671875,
      "learning_rate": 7.959595959595959e-06,
      "loss": 427.185,
      "step": 1970
    },
    {
      "epoch": 0.015998836448258308,
      "grad_norm": 1284.0130615234375,
      "learning_rate": 8.000000000000001e-06,
      "loss": 523.1269,
      "step": 1980
    },
    {
      "epoch": 0.01607963865254244,
      "grad_norm": 1079.0145263671875,
      "learning_rate": 8.040404040404042e-06,
      "loss": 497.3142,
      "step": 1990
    },
    {
      "epoch": 0.016160440856826573,
      "grad_norm": 1260.9786376953125,
      "learning_rate": 8.080808080808082e-06,
      "loss": 495.9584,
      "step": 2000
    }
  ],
  "logging_steps": 10,
  "max_steps": 123750,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}