{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 6.668896321070234,
  "eval_steps": 500,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.013377926421404682,
      "grad_norm": 0.5197079181671143,
      "learning_rate": 0.0,
      "loss": 4.2636,
      "step": 1
    },
    {
      "epoch": 0.026755852842809364,
      "grad_norm": 0.5626901984214783,
      "learning_rate": 4e-05,
      "loss": 4.3971,
      "step": 2
    },
    {
      "epoch": 0.04013377926421405,
      "grad_norm": 0.5167903304100037,
      "learning_rate": 8e-05,
      "loss": 4.3249,
      "step": 3
    },
    {
      "epoch": 0.05351170568561873,
      "grad_norm": 0.4764951169490814,
      "learning_rate": 0.00012,
      "loss": 4.2031,
      "step": 4
    },
    {
      "epoch": 0.06688963210702341,
      "grad_norm": 0.45488491654396057,
      "learning_rate": 0.00016,
      "loss": 4.3914,
      "step": 5
    },
    {
      "epoch": 0.0802675585284281,
      "grad_norm": 0.568274736404419,
      "learning_rate": 0.0002,
      "loss": 4.2346,
      "step": 6
    },
    {
      "epoch": 0.09364548494983277,
      "grad_norm": 0.5974003076553345,
      "learning_rate": 0.0001999555061179088,
      "loss": 4.131,
      "step": 7
    },
    {
      "epoch": 0.10702341137123746,
      "grad_norm": 0.6204471588134766,
      "learning_rate": 0.00019991101223581757,
      "loss": 4.2256,
      "step": 8
    },
    {
      "epoch": 0.12040133779264214,
      "grad_norm": 0.7143808603286743,
      "learning_rate": 0.00019986651835372636,
      "loss": 3.8449,
      "step": 9
    },
    {
      "epoch": 0.13377926421404682,
      "grad_norm": 0.7799420356750488,
      "learning_rate": 0.00019982202447163517,
      "loss": 4.4301,
      "step": 10
    },
    {
      "epoch": 0.14715719063545152,
      "grad_norm": 0.8880407214164734,
      "learning_rate": 0.00019977753058954395,
      "loss": 4.2266,
      "step": 11
    },
    {
      "epoch": 0.1605351170568562,
      "grad_norm": 0.7776209712028503,
      "learning_rate": 0.00019973303670745273,
      "loss": 4.3208,
      "step": 12
    },
    {
      "epoch": 0.17391304347826086,
      "grad_norm": 0.9125858545303345,
      "learning_rate": 0.0001996885428253615,
      "loss": 4.4363,
      "step": 13
    },
    {
      "epoch": 0.18729096989966554,
      "grad_norm": 0.9000256657600403,
      "learning_rate": 0.00019964404894327032,
      "loss": 4.2917,
      "step": 14
    },
    {
      "epoch": 0.20066889632107024,
      "grad_norm": 0.9995108246803284,
      "learning_rate": 0.00019959955506117908,
      "loss": 4.1784,
      "step": 15
    },
    {
      "epoch": 0.2140468227424749,
      "grad_norm": 0.9209024310112,
      "learning_rate": 0.0001995550611790879,
      "loss": 4.7852,
      "step": 16
    },
    {
      "epoch": 0.22742474916387959,
      "grad_norm": 0.9421981573104858,
      "learning_rate": 0.00019951056729699667,
      "loss": 4.8501,
      "step": 17
    },
    {
      "epoch": 0.2408026755852843,
      "grad_norm": 0.9213201403617859,
      "learning_rate": 0.00019946607341490545,
      "loss": 4.7923,
      "step": 18
    },
    {
      "epoch": 0.25418060200668896,
      "grad_norm": 0.9378194212913513,
      "learning_rate": 0.00019942157953281423,
      "loss": 4.9593,
      "step": 19
    },
    {
      "epoch": 0.26755852842809363,
      "grad_norm": 1.0096492767333984,
      "learning_rate": 0.00019937708565072304,
      "loss": 4.7099,
      "step": 20
    },
    {
      "epoch": 0.2809364548494983,
      "grad_norm": 0.8903587460517883,
      "learning_rate": 0.00019933259176863183,
      "loss": 4.3746,
      "step": 21
    },
    {
      "epoch": 0.29431438127090304,
      "grad_norm": 0.7808490991592407,
      "learning_rate": 0.0001992880978865406,
      "loss": 4.5873,
      "step": 22
    },
    {
      "epoch": 0.3076923076923077,
      "grad_norm": 0.8145670294761658,
      "learning_rate": 0.0001992436040044494,
      "loss": 4.7924,
      "step": 23
    },
    {
      "epoch": 0.3210702341137124,
      "grad_norm": 0.7945849299430847,
      "learning_rate": 0.0001991991101223582,
      "loss": 4.8881,
      "step": 24
    },
    {
      "epoch": 0.33444816053511706,
      "grad_norm": 0.7871395349502563,
      "learning_rate": 0.00019915461624026696,
      "loss": 4.6922,
      "step": 25
    },
    {
      "epoch": 0.34782608695652173,
      "grad_norm": 0.9111238718032837,
      "learning_rate": 0.00019911012235817577,
      "loss": 4.9982,
      "step": 26
    },
    {
      "epoch": 0.3612040133779264,
      "grad_norm": 0.7121369242668152,
      "learning_rate": 0.00019906562847608455,
      "loss": 4.4756,
      "step": 27
    },
    {
      "epoch": 0.3745819397993311,
      "grad_norm": 0.7118422389030457,
      "learning_rate": 0.00019902113459399333,
      "loss": 5.1389,
      "step": 28
    },
    {
      "epoch": 0.3879598662207358,
      "grad_norm": 0.7100292444229126,
      "learning_rate": 0.0001989766407119021,
      "loss": 4.7691,
      "step": 29
    },
    {
      "epoch": 0.4013377926421405,
      "grad_norm": 0.708591639995575,
      "learning_rate": 0.00019893214682981092,
      "loss": 4.8721,
      "step": 30
    },
    {
      "epoch": 0.41471571906354515,
      "grad_norm": 0.6711616516113281,
      "learning_rate": 0.0001988876529477197,
      "loss": 4.9152,
      "step": 31
    },
    {
      "epoch": 0.4280936454849498,
      "grad_norm": 0.7158232927322388,
      "learning_rate": 0.0001988431590656285,
      "loss": 4.828,
      "step": 32
    },
    {
      "epoch": 0.4414715719063545,
      "grad_norm": 0.6246087551116943,
      "learning_rate": 0.00019879866518353727,
      "loss": 4.8452,
      "step": 33
    },
    {
      "epoch": 0.45484949832775917,
      "grad_norm": 0.6088873744010925,
      "learning_rate": 0.00019875417130144608,
      "loss": 4.9702,
      "step": 34
    },
    {
      "epoch": 0.4682274247491639,
      "grad_norm": 0.5798126459121704,
      "learning_rate": 0.00019870967741935483,
      "loss": 4.9838,
      "step": 35
    },
    {
      "epoch": 0.4816053511705686,
      "grad_norm": 0.6268919706344604,
      "learning_rate": 0.00019866518353726364,
      "loss": 4.7636,
      "step": 36
    },
    {
      "epoch": 0.49498327759197325,
      "grad_norm": 0.5649904012680054,
      "learning_rate": 0.00019862068965517243,
      "loss": 4.506,
      "step": 37
    },
    {
      "epoch": 0.5083612040133779,
      "grad_norm": 0.5947792530059814,
      "learning_rate": 0.0001985761957730812,
      "loss": 4.8057,
      "step": 38
    },
    {
      "epoch": 0.5217391304347826,
      "grad_norm": 0.6204257011413574,
      "learning_rate": 0.00019853170189099,
      "loss": 5.0511,
      "step": 39
    },
    {
      "epoch": 0.5351170568561873,
      "grad_norm": 0.5972265601158142,
      "learning_rate": 0.0001984872080088988,
      "loss": 4.924,
      "step": 40
    },
    {
      "epoch": 0.5484949832775919,
      "grad_norm": 0.6117077469825745,
      "learning_rate": 0.00019844271412680758,
      "loss": 4.8729,
      "step": 41
    },
    {
      "epoch": 0.5618729096989966,
      "grad_norm": 0.5085508823394775,
      "learning_rate": 0.00019839822024471637,
      "loss": 4.3616,
      "step": 42
    },
    {
      "epoch": 0.5752508361204013,
      "grad_norm": 0.550647497177124,
      "learning_rate": 0.00019835372636262515,
      "loss": 5.2512,
      "step": 43
    },
    {
      "epoch": 0.5886287625418061,
      "grad_norm": 0.48329588770866394,
      "learning_rate": 0.00019830923248053396,
      "loss": 4.9501,
      "step": 44
    },
    {
      "epoch": 0.6020066889632107,
      "grad_norm": 0.6313246488571167,
      "learning_rate": 0.0001982647385984427,
      "loss": 4.5767,
      "step": 45
    },
    {
      "epoch": 0.6153846153846154,
      "grad_norm": 0.5111928582191467,
      "learning_rate": 0.00019822024471635152,
      "loss": 4.5586,
      "step": 46
    },
    {
      "epoch": 0.6287625418060201,
      "grad_norm": 0.5264492630958557,
      "learning_rate": 0.0001981757508342603,
      "loss": 4.7033,
      "step": 47
    },
    {
      "epoch": 0.6421404682274248,
      "grad_norm": 0.5058289170265198,
      "learning_rate": 0.0001981312569521691,
      "loss": 4.8396,
      "step": 48
    },
    {
      "epoch": 0.6555183946488294,
      "grad_norm": 0.5688439607620239,
      "learning_rate": 0.00019808676307007787,
      "loss": 5.1887,
      "step": 49
    },
    {
      "epoch": 0.6688963210702341,
      "grad_norm": 0.5488842129707336,
      "learning_rate": 0.00019804226918798665,
      "loss": 4.6075,
      "step": 50
    },
    {
      "epoch": 0.6822742474916388,
      "grad_norm": 0.5358632206916809,
      "learning_rate": 0.00019799777530589546,
      "loss": 5.0205,
      "step": 51
    },
    {
      "epoch": 0.6956521739130435,
      "grad_norm": 0.47869494557380676,
      "learning_rate": 0.00019795328142380422,
      "loss": 4.8495,
      "step": 52
    },
    {
      "epoch": 0.7090301003344481,
      "grad_norm": 0.49378660321235657,
      "learning_rate": 0.00019790878754171303,
      "loss": 4.6563,
      "step": 53
    },
    {
      "epoch": 0.7224080267558528,
      "grad_norm": 0.5167868733406067,
      "learning_rate": 0.0001978642936596218,
      "loss": 5.2558,
      "step": 54
    },
    {
      "epoch": 0.7357859531772575,
      "grad_norm": 0.5230040550231934,
      "learning_rate": 0.0001978197997775306,
      "loss": 4.7769,
      "step": 55
    },
    {
      "epoch": 0.7491638795986622,
      "grad_norm": 0.4822310507297516,
      "learning_rate": 0.00019777530589543937,
      "loss": 4.9282,
      "step": 56
    },
    {
      "epoch": 0.7625418060200669,
      "grad_norm": 0.500045895576477,
      "learning_rate": 0.00019773081201334818,
      "loss": 5.0399,
      "step": 57
    },
    {
      "epoch": 0.7759197324414716,
      "grad_norm": 0.4740642309188843,
      "learning_rate": 0.00019768631813125696,
      "loss": 4.8041,
      "step": 58
    },
    {
      "epoch": 0.7892976588628763,
      "grad_norm": 0.45918184518814087,
      "learning_rate": 0.00019764182424916575,
      "loss": 4.6304,
      "step": 59
    },
    {
      "epoch": 0.802675585284281,
      "grad_norm": 0.53122878074646,
      "learning_rate": 0.00019759733036707453,
      "loss": 4.8377,
      "step": 60
    },
    {
      "epoch": 0.8160535117056856,
      "grad_norm": 0.4925791919231415,
      "learning_rate": 0.00019755283648498334,
      "loss": 5.0919,
      "step": 61
    },
    {
      "epoch": 0.8294314381270903,
      "grad_norm": 0.4777262806892395,
      "learning_rate": 0.0001975083426028921,
      "loss": 4.8379,
      "step": 62
    },
    {
      "epoch": 0.842809364548495,
      "grad_norm": 0.49119675159454346,
      "learning_rate": 0.0001974638487208009,
      "loss": 5.0819,
      "step": 63
    },
    {
      "epoch": 0.8561872909698997,
      "grad_norm": 0.4732685089111328,
      "learning_rate": 0.00019741935483870969,
      "loss": 4.8948,
      "step": 64
    },
    {
      "epoch": 0.8695652173913043,
      "grad_norm": 0.46269145607948303,
      "learning_rate": 0.00019737486095661847,
      "loss": 4.824,
      "step": 65
    },
    {
      "epoch": 0.882943143812709,
      "grad_norm": 0.49532708525657654,
      "learning_rate": 0.00019733036707452725,
      "loss": 4.8986,
      "step": 66
    },
    {
      "epoch": 0.8963210702341137,
      "grad_norm": 0.5253002643585205,
      "learning_rate": 0.00019728587319243606,
      "loss": 4.9073,
      "step": 67
    },
    {
      "epoch": 0.9096989966555183,
      "grad_norm": 0.5069419145584106,
      "learning_rate": 0.00019724137931034484,
      "loss": 4.8962,
      "step": 68
    },
    {
      "epoch": 0.9230769230769231,
      "grad_norm": 0.5038817524909973,
      "learning_rate": 0.00019719688542825363,
      "loss": 4.8711,
      "step": 69
    },
    {
      "epoch": 0.9364548494983278,
      "grad_norm": 0.4987100064754486,
      "learning_rate": 0.0001971523915461624,
      "loss": 4.8816,
      "step": 70
    },
    {
      "epoch": 0.9498327759197325,
      "grad_norm": 0.47370976209640503,
      "learning_rate": 0.00019710789766407122,
      "loss": 4.9675,
      "step": 71
    },
    {
      "epoch": 0.9632107023411371,
      "grad_norm": 0.5081727504730225,
      "learning_rate": 0.00019706340378197997,
      "loss": 4.2768,
      "step": 72
    },
    {
      "epoch": 0.9765886287625418,
      "grad_norm": 0.45571258664131165,
      "learning_rate": 0.00019701890989988878,
      "loss": 4.6182,
      "step": 73
    },
    {
      "epoch": 0.9899665551839465,
      "grad_norm": 0.5216127634048462,
      "learning_rate": 0.00019697441601779756,
      "loss": 4.7126,
      "step": 74
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.5393329858779907,
      "learning_rate": 0.00019692992213570635,
      "loss": 4.4919,
      "step": 75
    },
    {
      "epoch": 1.0133779264214047,
      "grad_norm": 0.4506986737251282,
      "learning_rate": 0.00019688542825361513,
      "loss": 4.5089,
      "step": 76
    },
    {
      "epoch": 1.0267558528428093,
      "grad_norm": 0.4328899085521698,
      "learning_rate": 0.00019684093437152394,
      "loss": 4.7518,
      "step": 77
    },
    {
      "epoch": 1.040133779264214,
      "grad_norm": 0.4397362470626831,
      "learning_rate": 0.00019679644048943272,
      "loss": 4.5069,
      "step": 78
    },
    {
      "epoch": 1.0535117056856187,
      "grad_norm": 0.4604664146900177,
      "learning_rate": 0.0001967519466073415,
      "loss": 4.7054,
      "step": 79
    },
    {
      "epoch": 1.0668896321070234,
      "grad_norm": 0.4398234784603119,
      "learning_rate": 0.00019670745272525029,
      "loss": 4.2743,
      "step": 80
    },
    {
      "epoch": 1.080267558528428,
      "grad_norm": 0.4570735692977905,
      "learning_rate": 0.0001966629588431591,
      "loss": 4.8012,
      "step": 81
    },
    {
      "epoch": 1.0936454849498327,
      "grad_norm": 0.4814144968986511,
      "learning_rate": 0.00019661846496106785,
      "loss": 4.6449,
      "step": 82
    },
    {
      "epoch": 1.1070234113712374,
      "grad_norm": 0.4526231288909912,
      "learning_rate": 0.00019657397107897666,
      "loss": 4.5546,
      "step": 83
    },
    {
      "epoch": 1.120401337792642,
      "grad_norm": 0.4847906827926636,
      "learning_rate": 0.00019652947719688544,
      "loss": 4.4421,
      "step": 84
    },
    {
      "epoch": 1.1337792642140467,
      "grad_norm": 0.5136271715164185,
      "learning_rate": 0.00019648498331479422,
      "loss": 4.7136,
      "step": 85
    },
    {
      "epoch": 1.1471571906354514,
      "grad_norm": 0.49209895730018616,
      "learning_rate": 0.000196440489432703,
      "loss": 4.3145,
      "step": 86
    },
    {
      "epoch": 1.160535117056856,
      "grad_norm": 0.4972032904624939,
      "learning_rate": 0.00019639599555061182,
      "loss": 4.0408,
      "step": 87
    },
    {
      "epoch": 1.1739130434782608,
      "grad_norm": 0.5077862739562988,
      "learning_rate": 0.0001963515016685206,
      "loss": 4.4074,
      "step": 88
    },
    {
      "epoch": 1.1872909698996654,
      "grad_norm": 0.5293861031532288,
      "learning_rate": 0.00019630700778642935,
      "loss": 4.5385,
      "step": 89
    },
    {
      "epoch": 1.2006688963210703,
      "grad_norm": 0.5062645673751831,
      "learning_rate": 0.00019626251390433816,
      "loss": 4.5141,
      "step": 90
    },
    {
      "epoch": 1.214046822742475,
      "grad_norm": 0.49655866622924805,
      "learning_rate": 0.00019621802002224695,
      "loss": 4.4765,
      "step": 91
    },
    {
      "epoch": 1.2274247491638797,
      "grad_norm": 0.6059755086898804,
      "learning_rate": 0.00019617352614015573,
      "loss": 4.568,
      "step": 92
    },
    {
      "epoch": 1.2408026755852843,
      "grad_norm": 0.5442761778831482,
      "learning_rate": 0.0001961290322580645,
      "loss": 4.7724,
      "step": 93
    },
    {
      "epoch": 1.254180602006689,
      "grad_norm": 0.5426056385040283,
      "learning_rate": 0.00019608453837597332,
      "loss": 4.5308,
      "step": 94
    },
    {
      "epoch": 1.2675585284280937,
      "grad_norm": 0.525372326374054,
      "learning_rate": 0.0001960400444938821,
      "loss": 4.394,
      "step": 95
    },
    {
      "epoch": 1.2809364548494984,
      "grad_norm": 0.5407588481903076,
      "learning_rate": 0.00019599555061179089,
      "loss": 4.7347,
      "step": 96
    },
    {
      "epoch": 1.294314381270903,
      "grad_norm": 0.5726659893989563,
      "learning_rate": 0.00019595105672969967,
      "loss": 4.9446,
      "step": 97
    },
    {
      "epoch": 1.3076923076923077,
      "grad_norm": 0.6211283206939697,
      "learning_rate": 0.00019590656284760848,
      "loss": 4.697,
      "step": 98
    },
    {
      "epoch": 1.3210702341137124,
      "grad_norm": 0.5627567172050476,
      "learning_rate": 0.00019586206896551723,
      "loss": 4.4892,
      "step": 99
    },
    {
      "epoch": 1.334448160535117,
      "grad_norm": 0.6174790859222412,
      "learning_rate": 0.00019581757508342604,
      "loss": 4.5686,
      "step": 100
    },
    {
      "epoch": 1.3478260869565217,
      "grad_norm": 0.5586990118026733,
      "learning_rate": 0.00019577308120133482,
      "loss": 4.3916,
      "step": 101
    },
    {
      "epoch": 1.3612040133779264,
      "grad_norm": 0.5655365586280823,
      "learning_rate": 0.0001957285873192436,
      "loss": 4.1789,
      "step": 102
    },
    {
      "epoch": 1.374581939799331,
      "grad_norm": 0.5834594964981079,
      "learning_rate": 0.0001956840934371524,
      "loss": 4.3316,
      "step": 103
    },
    {
      "epoch": 1.3879598662207357,
      "grad_norm": 0.6065447926521301,
      "learning_rate": 0.0001956395995550612,
      "loss": 4.5167,
      "step": 104
    },
    {
      "epoch": 1.4013377926421404,
      "grad_norm": 0.5250216722488403,
      "learning_rate": 0.00019559510567296998,
      "loss": 4.1718,
      "step": 105
    },
    {
      "epoch": 1.414715719063545,
      "grad_norm": 0.5861116051673889,
      "learning_rate": 0.00019555061179087876,
      "loss": 4.3077,
      "step": 106
    },
    {
      "epoch": 1.4280936454849498,
      "grad_norm": 0.6138104796409607,
      "learning_rate": 0.00019550611790878755,
      "loss": 4.4748,
      "step": 107
    },
    {
      "epoch": 1.4414715719063544,
      "grad_norm": 0.6742071509361267,
      "learning_rate": 0.00019546162402669636,
      "loss": 4.8769,
      "step": 108
    },
    {
      "epoch": 1.4548494983277591,
      "grad_norm": 0.6634951233863831,
      "learning_rate": 0.0001954171301446051,
      "loss": 4.6423,
      "step": 109
    },
    {
      "epoch": 1.468227424749164,
      "grad_norm": 0.626646876335144,
      "learning_rate": 0.00019537263626251392,
      "loss": 4.4654,
      "step": 110
    },
    {
      "epoch": 1.4816053511705687,
      "grad_norm": 0.6306963562965393,
      "learning_rate": 0.0001953281423804227,
      "loss": 4.7021,
      "step": 111
    },
    {
      "epoch": 1.4949832775919734,
      "grad_norm": 0.620370626449585,
      "learning_rate": 0.00019528364849833149,
      "loss": 4.587,
      "step": 112
    },
    {
      "epoch": 1.508361204013378,
      "grad_norm": 0.6410287618637085,
      "learning_rate": 0.00019523915461624027,
      "loss": 4.8089,
      "step": 113
    },
    {
      "epoch": 1.5217391304347827,
      "grad_norm": 0.676434338092804,
      "learning_rate": 0.00019519466073414908,
      "loss": 4.668,
      "step": 114
    },
    {
      "epoch": 1.5351170568561874,
      "grad_norm": 0.5756319761276245,
      "learning_rate": 0.00019515016685205786,
      "loss": 4.3223,
      "step": 115
    },
    {
      "epoch": 1.548494983277592,
      "grad_norm": 0.5850693583488464,
      "learning_rate": 0.00019510567296996664,
      "loss": 4.2343,
      "step": 116
    },
    {
      "epoch": 1.5618729096989967,
      "grad_norm": 0.6172360777854919,
      "learning_rate": 0.00019506117908787542,
      "loss": 4.6102,
      "step": 117
    },
    {
      "epoch": 1.5752508361204014,
      "grad_norm": 0.5887568593025208,
      "learning_rate": 0.00019501668520578423,
      "loss": 4.8097,
      "step": 118
    },
    {
      "epoch": 1.588628762541806,
      "grad_norm": 0.5763369798660278,
      "learning_rate": 0.000194972191323693,
      "loss": 4.2001,
      "step": 119
    },
    {
      "epoch": 1.6020066889632107,
      "grad_norm": 0.6158986687660217,
      "learning_rate": 0.0001949276974416018,
      "loss": 4.7075,
      "step": 120
    },
    {
      "epoch": 1.6153846153846154,
      "grad_norm": 0.5540957450866699,
      "learning_rate": 0.00019488320355951058,
      "loss": 4.452,
      "step": 121
    },
    {
      "epoch": 1.62876254180602,
      "grad_norm": 0.6193795204162598,
      "learning_rate": 0.00019483870967741936,
      "loss": 4.4583,
      "step": 122
    },
    {
      "epoch": 1.6421404682274248,
      "grad_norm": 0.6699966788291931,
      "learning_rate": 0.00019479421579532815,
      "loss": 4.3728,
      "step": 123
    },
    {
      "epoch": 1.6555183946488294,
      "grad_norm": 0.5904677510261536,
      "learning_rate": 0.00019474972191323696,
      "loss": 4.5452,
      "step": 124
    },
    {
      "epoch": 1.6688963210702341,
      "grad_norm": 0.6137760281562805,
      "learning_rate": 0.00019470522803114574,
      "loss": 4.2853,
      "step": 125
    },
    {
      "epoch": 1.6822742474916388,
      "grad_norm": 0.6396192908287048,
      "learning_rate": 0.00019466073414905452,
      "loss": 4.4258,
      "step": 126
    },
    {
      "epoch": 1.6956521739130435,
      "grad_norm": 0.6190487742424011,
      "learning_rate": 0.0001946162402669633,
      "loss": 4.9866,
      "step": 127
    },
    {
      "epoch": 1.7090301003344481,
      "grad_norm": 0.6971675157546997,
      "learning_rate": 0.0001945717463848721,
      "loss": 4.2126,
      "step": 128
    },
    {
      "epoch": 1.7224080267558528,
      "grad_norm": 0.6245931386947632,
      "learning_rate": 0.00019452725250278087,
      "loss": 4.8477,
      "step": 129
    },
    {
      "epoch": 1.7357859531772575,
      "grad_norm": 0.5675052404403687,
      "learning_rate": 0.00019448275862068965,
      "loss": 4.4097,
      "step": 130
    },
    {
      "epoch": 1.7491638795986622,
      "grad_norm": 0.6594040393829346,
      "learning_rate": 0.00019443826473859846,
      "loss": 4.3747,
      "step": 131
    },
    {
      "epoch": 1.7625418060200668,
      "grad_norm": 0.6377655267715454,
      "learning_rate": 0.00019439377085650724,
      "loss": 4.2733,
      "step": 132
    },
    {
      "epoch": 1.7759197324414715,
      "grad_norm": 0.6167862415313721,
      "learning_rate": 0.00019434927697441602,
      "loss": 4.5694,
      "step": 133
    },
    {
      "epoch": 1.7892976588628762,
      "grad_norm": 0.577671468257904,
      "learning_rate": 0.0001943047830923248,
      "loss": 4.5006,
      "step": 134
    },
    {
      "epoch": 1.8026755852842808,
      "grad_norm": 0.6361016035079956,
      "learning_rate": 0.00019426028921023362,
      "loss": 4.9907,
      "step": 135
    },
    {
      "epoch": 1.8160535117056855,
      "grad_norm": 0.6445321440696716,
      "learning_rate": 0.00019421579532814237,
      "loss": 4.779,
      "step": 136
    },
    {
      "epoch": 1.8294314381270902,
      "grad_norm": 0.5955402851104736,
      "learning_rate": 0.00019417130144605118,
      "loss": 4.6026,
      "step": 137
    },
    {
      "epoch": 1.8428093645484949,
      "grad_norm": 0.6807080507278442,
      "learning_rate": 0.00019412680756395996,
      "loss": 4.7124,
      "step": 138
    },
    {
      "epoch": 1.8561872909698995,
      "grad_norm": 0.5799288153648376,
      "learning_rate": 0.00019408231368186875,
      "loss": 4.0701,
      "step": 139
    },
    {
      "epoch": 1.8695652173913042,
      "grad_norm": 0.6187757253646851,
      "learning_rate": 0.00019403781979977753,
      "loss": 4.705,
      "step": 140
    },
    {
      "epoch": 1.8829431438127089,
      "grad_norm": 0.6614826917648315,
      "learning_rate": 0.00019399332591768634,
      "loss": 4.8146,
      "step": 141
    },
    {
      "epoch": 1.8963210702341136,
      "grad_norm": 0.6204859614372253,
      "learning_rate": 0.00019394883203559512,
      "loss": 4.3041,
      "step": 142
    },
    {
      "epoch": 1.9096989966555182,
      "grad_norm": 0.6527450680732727,
      "learning_rate": 0.0001939043381535039,
      "loss": 4.4493,
      "step": 143
    },
    {
      "epoch": 1.9230769230769231,
      "grad_norm": 0.6470615267753601,
      "learning_rate": 0.00019385984427141268,
      "loss": 4.7771,
      "step": 144
    },
    {
      "epoch": 1.9364548494983278,
      "grad_norm": 0.5642555952072144,
      "learning_rate": 0.0001938153503893215,
      "loss": 4.3344,
      "step": 145
    },
    {
      "epoch": 1.9498327759197325,
      "grad_norm": 0.6206467151641846,
      "learning_rate": 0.00019377085650723025,
      "loss": 4.2191,
      "step": 146
    },
    {
      "epoch": 1.9632107023411371,
      "grad_norm": 0.6079016923904419,
      "learning_rate": 0.00019372636262513906,
      "loss": 4.7397,
      "step": 147
    },
    {
      "epoch": 1.9765886287625418,
      "grad_norm": 0.6197662353515625,
      "learning_rate": 0.00019368186874304784,
      "loss": 4.5342,
      "step": 148
    },
    {
      "epoch": 1.9899665551839465,
      "grad_norm": 0.6556297540664673,
      "learning_rate": 0.00019363737486095662,
      "loss": 4.6709,
      "step": 149
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.7837930917739868,
      "learning_rate": 0.0001935928809788654,
      "loss": 4.6215,
      "step": 150
    },
    {
      "epoch": 2.0133779264214047,
      "grad_norm": 0.5267267227172852,
      "learning_rate": 0.00019354838709677422,
      "loss": 4.2695,
      "step": 151
    },
    {
      "epoch": 2.0267558528428093,
      "grad_norm": 0.5862157344818115,
      "learning_rate": 0.000193503893214683,
      "loss": 4.3702,
      "step": 152
    },
    {
      "epoch": 2.040133779264214,
      "grad_norm": 0.538254976272583,
      "learning_rate": 0.00019345939933259178,
      "loss": 4.3953,
      "step": 153
    },
    {
      "epoch": 2.0535117056856187,
      "grad_norm": 0.5977053642272949,
      "learning_rate": 0.00019341490545050056,
      "loss": 4.2156,
      "step": 154
    },
    {
      "epoch": 2.0668896321070234,
      "grad_norm": 0.606006383895874,
      "learning_rate": 0.00019337041156840937,
      "loss": 4.2802,
      "step": 155
    },
    {
      "epoch": 2.080267558528428,
      "grad_norm": 0.6071277856826782,
      "learning_rate": 0.00019332591768631813,
      "loss": 4.5545,
      "step": 156
    },
    {
      "epoch": 2.0936454849498327,
      "grad_norm": 0.6281546354293823,
      "learning_rate": 0.00019328142380422694,
      "loss": 4.6105,
      "step": 157
    },
    {
      "epoch": 2.1070234113712374,
      "grad_norm": 0.5703116655349731,
      "learning_rate": 0.00019323692992213572,
      "loss": 4.2751,
      "step": 158
    },
    {
      "epoch": 2.120401337792642,
      "grad_norm": 0.6587452292442322,
      "learning_rate": 0.0001931924360400445,
      "loss": 4.6342,
      "step": 159
    },
    {
      "epoch": 2.1337792642140467,
      "grad_norm": 0.6141905784606934,
      "learning_rate": 0.00019314794215795328,
      "loss": 4.4345,
      "step": 160
    },
    {
      "epoch": 2.1471571906354514,
      "grad_norm": 0.6741939187049866,
      "learning_rate": 0.0001931034482758621,
      "loss": 4.0257,
      "step": 161
    },
    {
      "epoch": 2.160535117056856,
      "grad_norm": 0.6468759179115295,
      "learning_rate": 0.00019305895439377088,
      "loss": 4.2313,
      "step": 162
    },
    {
      "epoch": 2.1739130434782608,
      "grad_norm": 0.6703383326530457,
      "learning_rate": 0.00019301446051167966,
      "loss": 4.2164,
      "step": 163
    },
    {
      "epoch": 2.1872909698996654,
      "grad_norm": 0.710967481136322,
      "learning_rate": 0.00019296996662958844,
      "loss": 4.3398,
      "step": 164
    },
    {
      "epoch": 2.20066889632107,
      "grad_norm": 0.6862124800682068,
      "learning_rate": 0.00019292547274749725,
      "loss": 4.3379,
      "step": 165
    },
    {
      "epoch": 2.2140468227424748,
      "grad_norm": 0.6288430690765381,
      "learning_rate": 0.000192880978865406,
      "loss": 4.3487,
      "step": 166
    },
    {
      "epoch": 2.2274247491638794,
      "grad_norm": 0.6358796954154968,
      "learning_rate": 0.00019283648498331481,
      "loss": 4.1656,
      "step": 167
    },
    {
      "epoch": 2.240802675585284,
      "grad_norm": 0.6818917393684387,
      "learning_rate": 0.0001927919911012236,
      "loss": 4.5363,
      "step": 168
    },
    {
      "epoch": 2.254180602006689,
      "grad_norm": 0.6996105313301086,
      "learning_rate": 0.00019274749721913238,
      "loss": 4.3208,
      "step": 169
    },
    {
      "epoch": 2.2675585284280935,
      "grad_norm": 0.6730326414108276,
      "learning_rate": 0.00019270300333704116,
      "loss": 4.1401,
      "step": 170
    },
    {
      "epoch": 2.280936454849498,
      "grad_norm": 0.7022603750228882,
      "learning_rate": 0.00019265850945494994,
      "loss": 4.5761,
      "step": 171
    },
    {
      "epoch": 2.294314381270903,
      "grad_norm": 0.6525995135307312,
      "learning_rate": 0.00019261401557285875,
      "loss": 4.4017,
      "step": 172
    },
    {
      "epoch": 2.3076923076923075,
      "grad_norm": 0.7066033482551575,
      "learning_rate": 0.0001925695216907675,
      "loss": 4.0037,
      "step": 173
    },
    {
      "epoch": 2.321070234113712,
      "grad_norm": 0.6708059310913086,
      "learning_rate": 0.00019252502780867632,
      "loss": 4.1947,
      "step": 174
    },
    {
      "epoch": 2.334448160535117,
      "grad_norm": 0.8711172342300415,
      "learning_rate": 0.0001924805339265851,
      "loss": 3.9958,
      "step": 175
    },
    {
      "epoch": 2.3478260869565215,
      "grad_norm": 0.7258634567260742,
      "learning_rate": 0.00019243604004449388,
      "loss": 4.4682,
      "step": 176
    },
    {
      "epoch": 2.361204013377926,
      "grad_norm": 0.7693021893501282,
      "learning_rate": 0.00019239154616240267,
      "loss": 4.54,
      "step": 177
    },
    {
      "epoch": 2.374581939799331,
      "grad_norm": 0.7271276116371155,
      "learning_rate": 0.00019234705228031148,
      "loss": 4.2942,
      "step": 178
    },
    {
      "epoch": 2.387959866220736,
      "grad_norm": 0.6836609244346619,
      "learning_rate": 0.00019230255839822026,
      "loss": 4.3099,
      "step": 179
    },
    {
      "epoch": 2.4013377926421406,
      "grad_norm": 0.731164813041687,
      "learning_rate": 0.00019225806451612904,
      "loss": 4.4077,
      "step": 180
    },
    {
      "epoch": 2.4147157190635453,
      "grad_norm": 0.7575274109840393,
      "learning_rate": 0.00019221357063403782,
      "loss": 4.6572,
      "step": 181
    },
    {
      "epoch": 2.42809364548495,
      "grad_norm": 0.8461325168609619,
      "learning_rate": 0.00019216907675194663,
      "loss": 4.4922,
      "step": 182
    },
    {
      "epoch": 2.4414715719063547,
      "grad_norm": 0.7225251197814941,
      "learning_rate": 0.0001921245828698554,
      "loss": 4.0372,
      "step": 183
    },
    {
      "epoch": 2.4548494983277593,
      "grad_norm": 3.563720703125,
      "learning_rate": 0.0001920800889877642,
      "loss": 4.5412,
      "step": 184
    },
    {
      "epoch": 2.468227424749164,
      "grad_norm": 0.8452121019363403,
      "learning_rate": 0.00019203559510567298,
      "loss": 4.4961,
      "step": 185
    },
    {
      "epoch": 2.4816053511705687,
      "grad_norm": 0.8734024167060852,
      "learning_rate": 0.00019199110122358176,
      "loss": 4.0884,
      "step": 186
    },
    {
      "epoch": 2.4949832775919734,
      "grad_norm": 1.1765823364257812,
      "learning_rate": 0.00019194660734149054,
      "loss": 4.2228,
      "step": 187
    },
    {
      "epoch": 2.508361204013378,
      "grad_norm": 0.750206708908081,
      "learning_rate": 0.00019190211345939935,
      "loss": 4.4305,
      "step": 188
    },
    {
      "epoch": 2.5217391304347827,
      "grad_norm": 0.7574430704116821,
      "learning_rate": 0.00019185761957730814,
      "loss": 4.1511,
      "step": 189
    },
    {
      "epoch": 2.5351170568561874,
      "grad_norm": 0.7105517387390137,
      "learning_rate": 0.00019181312569521692,
      "loss": 4.4793,
      "step": 190
    },
    {
      "epoch": 2.548494983277592,
      "grad_norm": 0.7495557069778442,
      "learning_rate": 0.0001917686318131257,
      "loss": 4.1335,
      "step": 191
    },
    {
      "epoch": 2.5618729096989967,
      "grad_norm": 0.8001168966293335,
      "learning_rate": 0.0001917241379310345,
      "loss": 4.7898,
      "step": 192
    },
    {
      "epoch": 2.5752508361204014,
      "grad_norm": 0.7402104735374451,
      "learning_rate": 0.00019167964404894327,
      "loss": 4.4482,
      "step": 193
    },
    {
      "epoch": 2.588628762541806,
      "grad_norm": 0.748267650604248,
      "learning_rate": 0.00019163515016685207,
      "loss": 4.3167,
      "step": 194
    },
    {
      "epoch": 2.6020066889632107,
      "grad_norm": 0.8291250467300415,
      "learning_rate": 0.00019159065628476086,
      "loss": 4.058,
      "step": 195
    },
    {
      "epoch": 2.6153846153846154,
      "grad_norm": 0.6945542693138123,
      "learning_rate": 0.00019154616240266964,
      "loss": 3.9751,
      "step": 196
    },
    {
      "epoch": 2.62876254180602,
      "grad_norm": 0.7307319045066833,
      "learning_rate": 0.00019150166852057842,
      "loss": 4.2736,
      "step": 197
    },
    {
      "epoch": 2.6421404682274248,
      "grad_norm": 0.7489168047904968,
      "learning_rate": 0.00019145717463848723,
      "loss": 4.3075,
      "step": 198
    },
    {
      "epoch": 2.6555183946488294,
      "grad_norm": 0.9727582931518555,
      "learning_rate": 0.00019141268075639601,
      "loss": 4.6474,
      "step": 199
    },
    {
      "epoch": 2.668896321070234,
      "grad_norm": 0.6776256561279297,
      "learning_rate": 0.0001913681868743048,
      "loss": 4.4217,
      "step": 200
    },
    {
      "epoch": 2.682274247491639,
      "grad_norm": 0.7305111885070801,
      "learning_rate": 0.00019132369299221358,
      "loss": 4.2804,
      "step": 201
    },
    {
      "epoch": 2.6956521739130435,
      "grad_norm": 0.7196978330612183,
      "learning_rate": 0.0001912791991101224,
      "loss": 4.3941,
      "step": 202
    },
    {
      "epoch": 2.709030100334448,
      "grad_norm": 0.7988458871841431,
      "learning_rate": 0.00019123470522803114,
      "loss": 4.437,
      "step": 203
    },
    {
      "epoch": 2.722408026755853,
      "grad_norm": 0.7004797458648682,
      "learning_rate": 0.00019119021134593995,
      "loss": 4.4986,
      "step": 204
    },
    {
      "epoch": 2.7357859531772575,
      "grad_norm": 0.677796483039856,
      "learning_rate": 0.00019114571746384874,
      "loss": 4.0851,
      "step": 205
    },
    {
      "epoch": 2.749163879598662,
      "grad_norm": 0.7527475357055664,
      "learning_rate": 0.00019110122358175752,
      "loss": 4.4469,
      "step": 206
    },
    {
      "epoch": 2.762541806020067,
      "grad_norm": 1.1659115552902222,
      "learning_rate": 0.0001910567296996663,
      "loss": 4.3284,
      "step": 207
    },
    {
      "epoch": 2.7759197324414715,
      "grad_norm": 0.7238364815711975,
      "learning_rate": 0.0001910122358175751,
      "loss": 4.2605,
      "step": 208
    },
    {
      "epoch": 2.789297658862876,
      "grad_norm": 0.7537760734558105,
      "learning_rate": 0.0001909677419354839,
      "loss": 4.3775,
      "step": 209
    },
    {
      "epoch": 2.802675585284281,
      "grad_norm": 0.6874127388000488,
      "learning_rate": 0.00019092324805339267,
      "loss": 4.3404,
      "step": 210
    },
    {
      "epoch": 2.8160535117056855,
      "grad_norm": 0.7045959830284119,
      "learning_rate": 0.00019087875417130146,
      "loss": 4.1568,
      "step": 211
    },
    {
      "epoch": 2.82943143812709,
      "grad_norm": 0.7249194383621216,
      "learning_rate": 0.00019083426028921027,
      "loss": 4.1969,
      "step": 212
    },
    {
      "epoch": 2.842809364548495,
      "grad_norm": 0.8331268429756165,
      "learning_rate": 0.00019078976640711902,
      "loss": 4.3169,
      "step": 213
    },
    {
      "epoch": 2.8561872909698995,
      "grad_norm": 0.7171936631202698,
      "learning_rate": 0.0001907452725250278,
      "loss": 4.5123,
      "step": 214
    },
    {
      "epoch": 2.869565217391304,
      "grad_norm": 0.759919285774231,
      "learning_rate": 0.0001907007786429366,
      "loss": 4.5412,
      "step": 215
    },
    {
      "epoch": 2.882943143812709,
      "grad_norm": 0.7451274991035461,
      "learning_rate": 0.0001906562847608454,
      "loss": 4.5253,
      "step": 216
    },
    {
      "epoch": 2.8963210702341136,
      "grad_norm": 0.6564481258392334,
      "learning_rate": 0.00019061179087875418,
      "loss": 4.1092,
      "step": 217
    },
    {
      "epoch": 2.9096989966555182,
      "grad_norm": 0.7339865565299988,
      "learning_rate": 0.00019056729699666296,
      "loss": 4.5092,
      "step": 218
    },
    {
      "epoch": 2.9230769230769234,
      "grad_norm": 0.7113937735557556,
      "learning_rate": 0.00019052280311457177,
      "loss": 4.3355,
      "step": 219
    },
    {
      "epoch": 2.936454849498328,
      "grad_norm": 0.7306456565856934,
      "learning_rate": 0.00019047830923248053,
      "loss": 4.5745,
      "step": 220
    },
    {
      "epoch": 2.9498327759197327,
      "grad_norm": 0.7971818447113037,
      "learning_rate": 0.00019043381535038933,
      "loss": 4.2903,
      "step": 221
    },
    {
      "epoch": 2.9632107023411374,
      "grad_norm": 0.7757331728935242,
      "learning_rate": 0.00019038932146829812,
      "loss": 4.2832,
      "step": 222
    },
    {
      "epoch": 2.976588628762542,
      "grad_norm": 0.7326288223266602,
      "learning_rate": 0.0001903448275862069,
      "loss": 4.2444,
      "step": 223
    },
    {
      "epoch": 2.9899665551839467,
      "grad_norm": 0.7363834381103516,
      "learning_rate": 0.00019030033370411568,
      "loss": 4.6744,
      "step": 224
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.8835271596908569,
      "learning_rate": 0.0001902558398220245,
      "loss": 4.432,
      "step": 225
    },
    {
      "epoch": 3.0133779264214047,
      "grad_norm": 0.6591921448707581,
      "learning_rate": 0.00019021134593993327,
      "loss": 4.1353,
      "step": 226
    },
    {
      "epoch": 3.0267558528428093,
      "grad_norm": 0.6895263195037842,
      "learning_rate": 0.00019016685205784206,
      "loss": 4.1253,
      "step": 227
    },
    {
      "epoch": 3.040133779264214,
      "grad_norm": 0.6476898789405823,
      "learning_rate": 0.00019012235817575084,
      "loss": 4.0354,
      "step": 228
    },
    {
      "epoch": 3.0535117056856187,
      "grad_norm": 0.6398957967758179,
      "learning_rate": 0.00019007786429365965,
      "loss": 4.024,
      "step": 229
    },
    {
      "epoch": 3.0668896321070234,
      "grad_norm": 0.7483389973640442,
      "learning_rate": 0.0001900333704115684,
      "loss": 4.1405,
      "step": 230
    },
    {
      "epoch": 3.080267558528428,
      "grad_norm": 0.7003724575042725,
      "learning_rate": 0.0001899888765294772,
      "loss": 4.3593,
      "step": 231
    },
    {
      "epoch": 3.0936454849498327,
      "grad_norm": 0.7426732182502747,
      "learning_rate": 0.000189944382647386,
      "loss": 4.485,
      "step": 232
    },
    {
      "epoch": 3.1070234113712374,
      "grad_norm": 0.6957541108131409,
      "learning_rate": 0.00018989988876529478,
      "loss": 4.1017,
      "step": 233
    },
    {
      "epoch": 3.120401337792642,
      "grad_norm": 0.8613067865371704,
      "learning_rate": 0.00018985539488320356,
      "loss": 4.3038,
      "step": 234
    },
    {
      "epoch": 3.1337792642140467,
      "grad_norm": 0.8375754952430725,
      "learning_rate": 0.00018981090100111237,
      "loss": 4.4356,
      "step": 235
    },
    {
      "epoch": 3.1471571906354514,
      "grad_norm": 0.7878522872924805,
      "learning_rate": 0.00018976640711902115,
      "loss": 3.9916,
      "step": 236
    },
    {
      "epoch": 3.160535117056856,
      "grad_norm": 0.7463901042938232,
      "learning_rate": 0.00018972191323692993,
      "loss": 3.6761,
      "step": 237
    },
    {
      "epoch": 3.1739130434782608,
      "grad_norm": 0.7360939979553223,
      "learning_rate": 0.00018967741935483872,
      "loss": 3.9573,
      "step": 238
    },
    {
      "epoch": 3.1872909698996654,
      "grad_norm": 0.891861081123352,
      "learning_rate": 0.00018963292547274753,
      "loss": 4.1853,
      "step": 239
    },
    {
      "epoch": 3.20066889632107,
      "grad_norm": 0.8589549660682678,
      "learning_rate": 0.00018958843159065628,
      "loss": 4.0679,
      "step": 240
    },
    {
      "epoch": 3.2140468227424748,
      "grad_norm": 0.9534163475036621,
      "learning_rate": 0.0001895439377085651,
      "loss": 3.732,
      "step": 241
    },
    {
      "epoch": 3.2274247491638794,
      "grad_norm": 0.8968185186386108,
      "learning_rate": 0.00018949944382647387,
      "loss": 4.2217,
      "step": 242
    },
    {
      "epoch": 3.240802675585284,
      "grad_norm": 0.81589275598526,
      "learning_rate": 0.00018945494994438266,
      "loss": 4.428,
      "step": 243
    },
    {
      "epoch": 3.254180602006689,
      "grad_norm": 0.929050862789154,
      "learning_rate": 0.00018941045606229144,
      "loss": 4.3468,
      "step": 244
    },
    {
      "epoch": 3.2675585284280935,
      "grad_norm": 0.8535035252571106,
      "learning_rate": 0.00018936596218020025,
      "loss": 3.8489,
      "step": 245
    },
    {
      "epoch": 3.280936454849498,
      "grad_norm": 0.9484681487083435,
      "learning_rate": 0.00018932146829810903,
      "loss": 4.0132,
      "step": 246
    },
    {
      "epoch": 3.294314381270903,
      "grad_norm": 0.8190047144889832,
      "learning_rate": 0.0001892769744160178,
      "loss": 4.3574,
      "step": 247
    },
    {
      "epoch": 3.3076923076923075,
      "grad_norm": 0.8764749765396118,
      "learning_rate": 0.0001892324805339266,
      "loss": 4.3103,
      "step": 248
    },
    {
      "epoch": 3.321070234113712,
      "grad_norm": 0.8929185271263123,
      "learning_rate": 0.0001891879866518354,
      "loss": 4.3606,
      "step": 249
    },
    {
      "epoch": 3.334448160535117,
      "grad_norm": 0.9096692204475403,
      "learning_rate": 0.00018914349276974416,
      "loss": 4.0047,
      "step": 250
    },
    {
      "epoch": 3.3478260869565215,
      "grad_norm": 0.885143518447876,
      "learning_rate": 0.00018909899888765297,
      "loss": 4.182,
      "step": 251
    },
    {
      "epoch": 3.361204013377926,
      "grad_norm": 0.7724215984344482,
      "learning_rate": 0.00018905450500556175,
      "loss": 3.9529,
      "step": 252
    },
    {
      "epoch": 3.374581939799331,
      "grad_norm": 0.8351865410804749,
      "learning_rate": 0.00018901001112347053,
      "loss": 3.9533,
      "step": 253
    },
    {
      "epoch": 3.387959866220736,
      "grad_norm": 0.8684999942779541,
      "learning_rate": 0.00018896551724137932,
      "loss": 3.8594,
      "step": 254
    },
    {
      "epoch": 3.4013377926421406,
      "grad_norm": 0.8903334736824036,
      "learning_rate": 0.0001889210233592881,
      "loss": 3.9248,
      "step": 255
    },
    {
      "epoch": 3.4147157190635453,
      "grad_norm": 0.826690137386322,
      "learning_rate": 0.0001888765294771969,
      "loss": 4.0389,
      "step": 256
    },
    {
      "epoch": 3.42809364548495,
      "grad_norm": 0.8306142687797546,
      "learning_rate": 0.00018883203559510566,
      "loss": 3.8168,
      "step": 257
    },
    {
      "epoch": 3.4414715719063547,
      "grad_norm": 0.9032199382781982,
      "learning_rate": 0.00018878754171301447,
      "loss": 4.178,
      "step": 258
    },
    {
      "epoch": 3.4548494983277593,
      "grad_norm": 0.9081966280937195,
      "learning_rate": 0.00018874304783092326,
      "loss": 4.2583,
      "step": 259
    },
    {
      "epoch": 3.468227424749164,
      "grad_norm": 0.8424077033996582,
      "learning_rate": 0.00018869855394883204,
      "loss": 4.3285,
      "step": 260
    },
    {
      "epoch": 3.4816053511705687,
      "grad_norm": 0.8302170038223267,
      "learning_rate": 0.00018865406006674082,
      "loss": 4.1346,
      "step": 261
    },
    {
      "epoch": 3.4949832775919734,
      "grad_norm": 0.8747193217277527,
      "learning_rate": 0.00018860956618464963,
      "loss": 4.0747,
      "step": 262
    },
    {
      "epoch": 3.508361204013378,
      "grad_norm": 0.8613927364349365,
      "learning_rate": 0.0001885650723025584,
      "loss": 4.2346,
      "step": 263
    },
    {
      "epoch": 3.5217391304347827,
      "grad_norm": 0.8321558833122253,
      "learning_rate": 0.0001885205784204672,
      "loss": 3.9781,
      "step": 264
    },
    {
      "epoch": 3.5351170568561874,
      "grad_norm": 0.8961741328239441,
      "learning_rate": 0.00018847608453837598,
      "loss": 4.311,
      "step": 265
    },
    {
      "epoch": 3.548494983277592,
      "grad_norm": 0.7703898549079895,
      "learning_rate": 0.00018843159065628479,
      "loss": 4.1163,
      "step": 266
    },
    {
      "epoch": 3.5618729096989967,
      "grad_norm": 0.880051851272583,
      "learning_rate": 0.00018838709677419354,
      "loss": 3.8032,
      "step": 267
    },
    {
      "epoch": 3.5752508361204014,
      "grad_norm": 0.8287038207054138,
      "learning_rate": 0.00018834260289210235,
      "loss": 4.1627,
      "step": 268
    },
    {
      "epoch": 3.588628762541806,
      "grad_norm": 0.9726569652557373,
      "learning_rate": 0.00018829810901001113,
      "loss": 4.4055,
      "step": 269
    },
    {
      "epoch": 3.6020066889632107,
      "grad_norm": 0.8071132898330688,
      "learning_rate": 0.00018825361512791992,
      "loss": 4.1709,
      "step": 270
    },
    {
      "epoch": 3.6153846153846154,
      "grad_norm": 0.8310988545417786,
      "learning_rate": 0.0001882091212458287,
      "loss": 4.2359,
      "step": 271
    },
    {
      "epoch": 3.62876254180602,
      "grad_norm": 0.8713561296463013,
      "learning_rate": 0.0001881646273637375,
      "loss": 4.1247,
      "step": 272
    },
    {
      "epoch": 3.6421404682274248,
      "grad_norm": 0.8964342474937439,
      "learning_rate": 0.0001881201334816463,
      "loss": 4.0794,
      "step": 273
    },
    {
      "epoch": 3.6555183946488294,
      "grad_norm": 0.9901681542396545,
      "learning_rate": 0.00018807563959955507,
      "loss": 4.0217,
      "step": 274
    },
    {
      "epoch": 3.668896321070234,
      "grad_norm": 0.9279042482376099,
      "learning_rate": 0.00018803114571746385,
      "loss": 4.3244,
      "step": 275
    },
    {
      "epoch": 3.682274247491639,
      "grad_norm": 0.8105964660644531,
      "learning_rate": 0.00018798665183537266,
      "loss": 3.9041,
      "step": 276
    },
    {
      "epoch": 3.6956521739130435,
      "grad_norm": 0.8511622548103333,
      "learning_rate": 0.00018794215795328142,
      "loss": 3.8969,
      "step": 277
    },
    {
      "epoch": 3.709030100334448,
      "grad_norm": 0.9072037935256958,
      "learning_rate": 0.00018789766407119023,
      "loss": 4.2185,
      "step": 278
    },
    {
      "epoch": 3.722408026755853,
      "grad_norm": 0.9792962670326233,
      "learning_rate": 0.000187853170189099,
      "loss": 4.1915,
      "step": 279
    },
    {
      "epoch": 3.7357859531772575,
      "grad_norm": 0.8579828143119812,
      "learning_rate": 0.0001878086763070078,
      "loss": 3.8903,
      "step": 280
    },
    {
      "epoch": 3.749163879598662,
      "grad_norm": 0.9866719841957092,
      "learning_rate": 0.00018776418242491658,
      "loss": 4.2022,
      "step": 281
    },
    {
      "epoch": 3.762541806020067,
      "grad_norm": 0.9251964688301086,
      "learning_rate": 0.00018771968854282539,
      "loss": 3.9536,
      "step": 282
    },
    {
      "epoch": 3.7759197324414715,
      "grad_norm": 1.0300836563110352,
      "learning_rate": 0.00018767519466073417,
      "loss": 4.2908,
      "step": 283
    },
    {
      "epoch": 3.789297658862876,
      "grad_norm": 1.0194575786590576,
      "learning_rate": 0.00018763070077864295,
      "loss": 4.1851,
      "step": 284
    },
    {
      "epoch": 3.802675585284281,
      "grad_norm": 0.8165330290794373,
      "learning_rate": 0.00018758620689655173,
      "loss": 4.138,
      "step": 285
    },
    {
      "epoch": 3.8160535117056855,
      "grad_norm": 1.0104280710220337,
      "learning_rate": 0.00018754171301446054,
      "loss": 3.9481,
      "step": 286
    },
    {
      "epoch": 3.82943143812709,
      "grad_norm": 0.9972538352012634,
      "learning_rate": 0.0001874972191323693,
      "loss": 4.3932,
      "step": 287
    },
    {
      "epoch": 3.842809364548495,
      "grad_norm": 0.96323162317276,
      "learning_rate": 0.0001874527252502781,
      "loss": 4.1133,
      "step": 288
    },
    {
      "epoch": 3.8561872909698995,
      "grad_norm": 0.8500615954399109,
      "learning_rate": 0.0001874082313681869,
      "loss": 4.2205,
      "step": 289
    },
    {
      "epoch": 3.869565217391304,
      "grad_norm": 0.8451250195503235,
      "learning_rate": 0.00018736373748609567,
      "loss": 4.1371,
      "step": 290
    },
    {
      "epoch": 3.882943143812709,
      "grad_norm": 0.9399815201759338,
      "learning_rate": 0.00018731924360400445,
      "loss": 4.5237,
      "step": 291
    },
    {
      "epoch": 3.8963210702341136,
      "grad_norm": 0.8061622977256775,
      "learning_rate": 0.00018727474972191326,
      "loss": 4.1033,
      "step": 292
    },
    {
      "epoch": 3.9096989966555182,
      "grad_norm": 0.7987121343612671,
      "learning_rate": 0.00018723025583982205,
      "loss": 3.9311,
      "step": 293
    },
    {
      "epoch": 3.9230769230769234,
      "grad_norm": 0.9041138291358948,
      "learning_rate": 0.00018718576195773083,
      "loss": 4.0252,
      "step": 294
    },
    {
      "epoch": 3.936454849498328,
      "grad_norm": 1.0002484321594238,
      "learning_rate": 0.0001871412680756396,
      "loss": 4.4605,
      "step": 295
    },
    {
      "epoch": 3.9498327759197327,
      "grad_norm": 0.9991098046302795,
      "learning_rate": 0.0001870967741935484,
      "loss": 4.1528,
      "step": 296
    },
    {
      "epoch": 3.9632107023411374,
      "grad_norm": 1.2179397344589233,
      "learning_rate": 0.00018705228031145718,
      "loss": 4.5224,
      "step": 297
    },
    {
      "epoch": 3.976588628762542,
      "grad_norm": 0.8279774785041809,
      "learning_rate": 0.00018700778642936596,
      "loss": 3.9464,
      "step": 298
    },
    {
      "epoch": 3.9899665551839467,
      "grad_norm": 0.8012803792953491,
      "learning_rate": 0.00018696329254727477,
      "loss": 4.0139,
      "step": 299
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.9700272083282471,
      "learning_rate": 0.00018691879866518355,
      "loss": 3.8306,
      "step": 300
    },
    {
      "epoch": 4.013377926421405,
      "grad_norm": 0.7136749625205994,
      "learning_rate": 0.00018687430478309233,
      "loss": 3.9253,
      "step": 301
    },
    {
      "epoch": 4.026755852842809,
      "grad_norm": 0.7885096669197083,
      "learning_rate": 0.00018682981090100111,
      "loss": 3.927,
      "step": 302
    },
    {
      "epoch": 4.040133779264214,
      "grad_norm": 0.7801666855812073,
      "learning_rate": 0.00018678531701890992,
      "loss": 3.6482,
      "step": 303
    },
    {
      "epoch": 4.053511705685619,
      "grad_norm": 0.7857955098152161,
      "learning_rate": 0.00018674082313681868,
      "loss": 4.0665,
      "step": 304
    },
    {
      "epoch": 4.066889632107023,
      "grad_norm": 0.707421064376831,
      "learning_rate": 0.0001866963292547275,
      "loss": 3.9142,
      "step": 305
    },
    {
      "epoch": 4.080267558528428,
      "grad_norm": 0.7936912775039673,
      "learning_rate": 0.00018665183537263627,
      "loss": 4.1227,
      "step": 306
    },
    {
      "epoch": 4.093645484949833,
      "grad_norm": 0.8899754881858826,
      "learning_rate": 0.00018660734149054505,
      "loss": 3.7661,
      "step": 307
    },
    {
      "epoch": 4.107023411371237,
      "grad_norm": 0.7760347723960876,
      "learning_rate": 0.00018656284760845384,
      "loss": 3.8921,
      "step": 308
    },
    {
      "epoch": 4.120401337792642,
      "grad_norm": 0.8672968745231628,
      "learning_rate": 0.00018651835372636265,
      "loss": 3.6037,
      "step": 309
    },
    {
      "epoch": 4.133779264214047,
      "grad_norm": 0.8046863675117493,
      "learning_rate": 0.0001864738598442714,
      "loss": 3.9117,
      "step": 310
    },
    {
      "epoch": 4.147157190635451,
      "grad_norm": 0.9172897934913635,
      "learning_rate": 0.0001864293659621802,
      "loss": 3.7229,
      "step": 311
    },
    {
      "epoch": 4.160535117056856,
      "grad_norm": 0.9616653919219971,
      "learning_rate": 0.000186384872080089,
      "loss": 3.8851,
      "step": 312
    },
    {
      "epoch": 4.173913043478261,
      "grad_norm": 0.9659278988838196,
      "learning_rate": 0.0001863403781979978,
      "loss": 4.005,
      "step": 313
    },
    {
      "epoch": 4.187290969899665,
      "grad_norm": 0.9171205163002014,
      "learning_rate": 0.00018629588431590656,
      "loss": 3.8634,
      "step": 314
    },
    {
      "epoch": 4.20066889632107,
      "grad_norm": 0.9968683123588562,
      "learning_rate": 0.00018625139043381537,
      "loss": 3.7321,
      "step": 315
    },
    {
      "epoch": 4.214046822742475,
      "grad_norm": 0.8762083053588867,
      "learning_rate": 0.00018620689655172415,
      "loss": 3.931,
      "step": 316
    },
    {
      "epoch": 4.2274247491638794,
      "grad_norm": 0.9815887212753296,
      "learning_rate": 0.00018616240266963293,
      "loss": 3.9975,
      "step": 317
    },
    {
      "epoch": 4.240802675585284,
      "grad_norm": 1.0065505504608154,
      "learning_rate": 0.00018611790878754171,
      "loss": 3.8364,
      "step": 318
    },
    {
      "epoch": 4.254180602006689,
      "grad_norm": 0.9785431623458862,
      "learning_rate": 0.00018607341490545052,
      "loss": 3.8822,
      "step": 319
    },
    {
      "epoch": 4.2675585284280935,
      "grad_norm": 1.077799677848816,
      "learning_rate": 0.00018602892102335928,
      "loss": 3.8299,
      "step": 320
    },
    {
      "epoch": 4.280936454849498,
      "grad_norm": 0.8109619617462158,
      "learning_rate": 0.0001859844271412681,
      "loss": 3.8096,
      "step": 321
    },
    {
      "epoch": 4.294314381270903,
      "grad_norm": 0.967856764793396,
      "learning_rate": 0.00018593993325917687,
      "loss": 3.8639,
      "step": 322
    },
    {
      "epoch": 4.3076923076923075,
      "grad_norm": 0.8657905459403992,
      "learning_rate": 0.00018589543937708568,
      "loss": 3.7556,
      "step": 323
    },
    {
      "epoch": 4.321070234113712,
      "grad_norm": 0.9641517400741577,
      "learning_rate": 0.00018585094549499444,
      "loss": 3.9702,
      "step": 324
    },
    {
      "epoch": 4.334448160535117,
      "grad_norm": 0.9664435982704163,
      "learning_rate": 0.00018580645161290325,
      "loss": 3.8754,
      "step": 325
    },
    {
      "epoch": 4.3478260869565215,
      "grad_norm": 0.8322617411613464,
      "learning_rate": 0.00018576195773081203,
      "loss": 3.83,
      "step": 326
    },
    {
      "epoch": 4.361204013377926,
      "grad_norm": 1.0363450050354004,
      "learning_rate": 0.0001857174638487208,
      "loss": 3.9825,
      "step": 327
    },
    {
      "epoch": 4.374581939799331,
      "grad_norm": 1.0125840902328491,
      "learning_rate": 0.0001856729699666296,
      "loss": 3.6525,
      "step": 328
    },
    {
      "epoch": 4.3879598662207355,
      "grad_norm": 0.9922601580619812,
      "learning_rate": 0.0001856284760845384,
      "loss": 4.2373,
      "step": 329
    },
    {
      "epoch": 4.40133779264214,
      "grad_norm": 0.9070426225662231,
      "learning_rate": 0.00018558398220244716,
      "loss": 3.9623,
      "step": 330
    },
    {
      "epoch": 4.414715719063545,
      "grad_norm": 0.9369637370109558,
      "learning_rate": 0.00018553948832035597,
      "loss": 3.9297,
      "step": 331
    },
    {
      "epoch": 4.4280936454849495,
      "grad_norm": 1.108876347541809,
      "learning_rate": 0.00018549499443826475,
      "loss": 3.7325,
      "step": 332
    },
    {
      "epoch": 4.441471571906354,
      "grad_norm": 0.9405660629272461,
      "learning_rate": 0.00018545050055617356,
      "loss": 3.8615,
      "step": 333
    },
    {
      "epoch": 4.454849498327759,
      "grad_norm": 0.9730128645896912,
      "learning_rate": 0.00018540600667408231,
      "loss": 4.1794,
      "step": 334
    },
    {
      "epoch": 4.468227424749164,
      "grad_norm": 0.9341335892677307,
      "learning_rate": 0.00018536151279199112,
      "loss": 3.9422,
      "step": 335
    },
    {
      "epoch": 4.481605351170568,
      "grad_norm": 0.9262625575065613,
      "learning_rate": 0.0001853170189098999,
      "loss": 3.9819,
      "step": 336
    },
    {
      "epoch": 4.494983277591973,
      "grad_norm": 1.0419141054153442,
      "learning_rate": 0.00018527252502780866,
      "loss": 3.7481,
      "step": 337
    },
    {
      "epoch": 4.508361204013378,
      "grad_norm": 0.8986826539039612,
      "learning_rate": 0.00018522803114571747,
      "loss": 4.1195,
      "step": 338
    },
    {
      "epoch": 4.521739130434782,
      "grad_norm": 0.9502431154251099,
      "learning_rate": 0.00018518353726362625,
      "loss": 3.8521,
      "step": 339
    },
    {
      "epoch": 4.535117056856187,
      "grad_norm": 0.8936267495155334,
      "learning_rate": 0.00018513904338153504,
      "loss": 3.647,
      "step": 340
    },
    {
      "epoch": 4.548494983277592,
      "grad_norm": 0.8870158195495605,
      "learning_rate": 0.00018509454949944382,
      "loss": 3.7624,
      "step": 341
    },
    {
      "epoch": 4.561872909698996,
      "grad_norm": 0.9030978679656982,
      "learning_rate": 0.00018505005561735263,
      "loss": 3.8018,
      "step": 342
    },
    {
      "epoch": 4.575250836120401,
      "grad_norm": 0.8690946698188782,
      "learning_rate": 0.0001850055617352614,
      "loss": 3.6183,
      "step": 343
    },
    {
      "epoch": 4.588628762541806,
      "grad_norm": 0.9812071323394775,
      "learning_rate": 0.0001849610678531702,
      "loss": 4.1231,
      "step": 344
    },
    {
      "epoch": 4.602006688963211,
      "grad_norm": 0.9404383301734924,
      "learning_rate": 0.00018491657397107897,
      "loss": 3.9645,
      "step": 345
    },
    {
      "epoch": 4.615384615384615,
      "grad_norm": 1.0422123670578003,
      "learning_rate": 0.00018487208008898778,
      "loss": 3.9031,
      "step": 346
    },
    {
      "epoch": 4.6287625418060205,
      "grad_norm": 0.9838129281997681,
      "learning_rate": 0.00018482758620689654,
      "loss": 3.9985,
      "step": 347
    },
    {
      "epoch": 4.642140468227424,
      "grad_norm": 0.9232532978057861,
      "learning_rate": 0.00018478309232480535,
      "loss": 4.0343,
      "step": 348
    },
    {
      "epoch": 4.65551839464883,
      "grad_norm": 0.9242956042289734,
      "learning_rate": 0.00018473859844271413,
      "loss": 4.0669,
      "step": 349
    },
    {
      "epoch": 4.668896321070234,
      "grad_norm": 0.919269859790802,
      "learning_rate": 0.0001846941045606229,
      "loss": 4.0549,
      "step": 350
    },
    {
      "epoch": 4.682274247491639,
      "grad_norm": 0.93565833568573,
      "learning_rate": 0.0001846496106785317,
      "loss": 4.1306,
      "step": 351
    },
    {
      "epoch": 4.695652173913043,
      "grad_norm": 0.9001899361610413,
      "learning_rate": 0.0001846051167964405,
      "loss": 3.8916,
      "step": 352
    },
    {
      "epoch": 4.709030100334449,
      "grad_norm": 0.8896821737289429,
      "learning_rate": 0.0001845606229143493,
      "loss": 3.8377,
      "step": 353
    },
    {
      "epoch": 4.722408026755852,
      "grad_norm": 1.0137807130813599,
      "learning_rate": 0.00018451612903225807,
      "loss": 3.9923,
      "step": 354
    },
    {
      "epoch": 4.735785953177258,
      "grad_norm": 1.075823426246643,
      "learning_rate": 0.00018447163515016685,
      "loss": 4.0706,
      "step": 355
    },
    {
      "epoch": 4.749163879598662,
      "grad_norm": 1.0076895952224731,
      "learning_rate": 0.00018442714126807566,
      "loss": 4.0759,
      "step": 356
    },
    {
      "epoch": 4.762541806020067,
      "grad_norm": 0.9387428164482117,
      "learning_rate": 0.00018438264738598442,
      "loss": 3.6959,
      "step": 357
    },
    {
      "epoch": 4.775919732441472,
      "grad_norm": 0.8920648097991943,
      "learning_rate": 0.00018433815350389323,
      "loss": 3.9213,
      "step": 358
    },
    {
      "epoch": 4.789297658862877,
      "grad_norm": 1.0252491235733032,
      "learning_rate": 0.000184293659621802,
      "loss": 3.9118,
      "step": 359
    },
    {
      "epoch": 4.802675585284281,
      "grad_norm": 1.0382707118988037,
      "learning_rate": 0.0001842491657397108,
      "loss": 4.0172,
      "step": 360
    },
    {
      "epoch": 4.816053511705686,
      "grad_norm": 1.07838773727417,
      "learning_rate": 0.00018420467185761957,
      "loss": 3.8531,
      "step": 361
    },
    {
      "epoch": 4.829431438127091,
      "grad_norm": 0.9974546432495117,
      "learning_rate": 0.00018416017797552838,
      "loss": 4.0387,
      "step": 362
    },
    {
      "epoch": 4.842809364548495,
      "grad_norm": 1.024491548538208,
      "learning_rate": 0.00018411568409343717,
      "loss": 3.9504,
      "step": 363
    },
    {
      "epoch": 4.8561872909699,
      "grad_norm": 0.9236369132995605,
      "learning_rate": 0.00018407119021134595,
      "loss": 3.7119,
      "step": 364
    },
    {
      "epoch": 4.869565217391305,
      "grad_norm": 0.935644268989563,
      "learning_rate": 0.00018402669632925473,
      "loss": 4.0077,
      "step": 365
    },
    {
      "epoch": 4.882943143812709,
      "grad_norm": 0.9328681230545044,
      "learning_rate": 0.00018398220244716354,
      "loss": 3.9133,
      "step": 366
    },
    {
      "epoch": 4.896321070234114,
      "grad_norm": 0.9596607089042664,
      "learning_rate": 0.0001839377085650723,
      "loss": 3.8003,
      "step": 367
    },
    {
      "epoch": 4.909698996655519,
      "grad_norm": 0.9878052473068237,
      "learning_rate": 0.0001838932146829811,
      "loss": 3.8805,
      "step": 368
    },
    {
      "epoch": 4.923076923076923,
      "grad_norm": 1.00381600856781,
      "learning_rate": 0.0001838487208008899,
      "loss": 4.0264,
      "step": 369
    },
    {
      "epoch": 4.936454849498328,
      "grad_norm": 1.024754524230957,
      "learning_rate": 0.00018380422691879867,
      "loss": 3.7291,
      "step": 370
    },
    {
      "epoch": 4.949832775919733,
      "grad_norm": 0.9670823812484741,
      "learning_rate": 0.00018375973303670745,
      "loss": 3.9418,
      "step": 371
    },
    {
      "epoch": 4.963210702341137,
      "grad_norm": 0.9736581444740295,
      "learning_rate": 0.00018371523915461626,
      "loss": 3.8813,
      "step": 372
    },
    {
      "epoch": 4.976588628762542,
      "grad_norm": 0.9752672910690308,
      "learning_rate": 0.00018367074527252504,
      "loss": 3.6717,
      "step": 373
    },
    {
      "epoch": 4.989966555183947,
      "grad_norm": 1.1268304586410522,
      "learning_rate": 0.00018362625139043383,
      "loss": 3.9782,
      "step": 374
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.7933701276779175,
      "learning_rate": 0.0001835817575083426,
      "loss": 3.001,
      "step": 375
    },
    {
      "epoch": 5.013377926421405,
      "grad_norm": 0.8035010099411011,
      "learning_rate": 0.00018353726362625142,
      "loss": 3.7943,
      "step": 376
    },
    {
      "epoch": 5.026755852842809,
      "grad_norm": 0.8016420006752014,
      "learning_rate": 0.00018349276974416017,
      "loss": 3.7454,
      "step": 377
    },
    {
      "epoch": 5.040133779264214,
      "grad_norm": 0.6844643354415894,
      "learning_rate": 0.00018344827586206896,
      "loss": 3.699,
      "step": 378
    },
    {
      "epoch": 5.053511705685619,
      "grad_norm": 0.8649943470954895,
      "learning_rate": 0.00018340378197997777,
      "loss": 3.7197,
      "step": 379
    },
    {
      "epoch": 5.066889632107023,
      "grad_norm": 0.9685015678405762,
      "learning_rate": 0.00018335928809788655,
      "loss": 3.6952,
      "step": 380
    },
    {
      "epoch": 5.080267558528428,
      "grad_norm": 0.8728330135345459,
      "learning_rate": 0.00018331479421579533,
      "loss": 3.7164,
      "step": 381
    },
    {
      "epoch": 5.093645484949833,
      "grad_norm": 0.962504506111145,
      "learning_rate": 0.0001832703003337041,
      "loss": 3.6123,
      "step": 382
    },
    {
      "epoch": 5.107023411371237,
      "grad_norm": 0.9194462895393372,
      "learning_rate": 0.00018322580645161292,
      "loss": 3.275,
      "step": 383
    },
    {
      "epoch": 5.120401337792642,
      "grad_norm": 0.9851329326629639,
      "learning_rate": 0.00018318131256952168,
      "loss": 3.6222,
      "step": 384
    },
    {
      "epoch": 5.133779264214047,
      "grad_norm": 1.0702580213546753,
      "learning_rate": 0.0001831368186874305,
      "loss": 3.8728,
      "step": 385
    },
    {
      "epoch": 5.147157190635451,
      "grad_norm": 1.3237228393554688,
      "learning_rate": 0.00018309232480533927,
      "loss": 3.8948,
      "step": 386
    },
    {
      "epoch": 5.160535117056856,
      "grad_norm": 1.0076218843460083,
      "learning_rate": 0.00018304783092324805,
      "loss": 3.8894,
      "step": 387
    },
    {
      "epoch": 5.173913043478261,
      "grad_norm": 1.084722876548767,
      "learning_rate": 0.00018300333704115683,
      "loss": 3.7398,
      "step": 388
    },
    {
      "epoch": 5.187290969899665,
      "grad_norm": 0.9112711548805237,
      "learning_rate": 0.00018295884315906564,
      "loss": 3.5901,
      "step": 389
    },
    {
      "epoch": 5.20066889632107,
      "grad_norm": 0.9451406002044678,
      "learning_rate": 0.00018291434927697443,
      "loss": 3.6313,
      "step": 390
    },
    {
      "epoch": 5.214046822742475,
      "grad_norm": 0.8901047706604004,
      "learning_rate": 0.0001828698553948832,
      "loss": 3.3191,
      "step": 391
    },
    {
      "epoch": 5.2274247491638794,
      "grad_norm": 0.9838565587997437,
      "learning_rate": 0.000182825361512792,
      "loss": 3.882,
      "step": 392
    },
    {
      "epoch": 5.240802675585284,
      "grad_norm": 0.9839156866073608,
      "learning_rate": 0.0001827808676307008,
      "loss": 3.6068,
      "step": 393
    },
    {
      "epoch": 5.254180602006689,
      "grad_norm": 0.9328583478927612,
      "learning_rate": 0.00018273637374860956,
      "loss": 3.6856,
      "step": 394
    },
    {
      "epoch": 5.2675585284280935,
      "grad_norm": 0.8705796003341675,
      "learning_rate": 0.00018269187986651837,
      "loss": 3.7282,
      "step": 395
    },
    {
      "epoch": 5.280936454849498,
      "grad_norm": 0.9675374031066895,
      "learning_rate": 0.00018264738598442715,
      "loss": 3.6588,
      "step": 396
    },
    {
      "epoch": 5.294314381270903,
      "grad_norm": 1.145280361175537,
      "learning_rate": 0.00018260289210233593,
      "loss": 3.8843,
      "step": 397
    },
    {
      "epoch": 5.3076923076923075,
      "grad_norm": 0.9769694805145264,
      "learning_rate": 0.0001825583982202447,
      "loss": 3.7207,
      "step": 398
    },
    {
      "epoch": 5.321070234113712,
      "grad_norm": 0.9277816414833069,
      "learning_rate": 0.00018251390433815352,
      "loss": 3.712,
      "step": 399
    },
    {
      "epoch": 5.334448160535117,
      "grad_norm": 1.1015180349349976,
      "learning_rate": 0.0001824694104560623,
      "loss": 3.7941,
      "step": 400
    },
    {
      "epoch": 5.3478260869565215,
      "grad_norm": 1.2234200239181519,
      "learning_rate": 0.0001824249165739711,
      "loss": 3.6559,
      "step": 401
    },
    {
      "epoch": 5.361204013377926,
      "grad_norm": 0.9358471035957336,
      "learning_rate": 0.00018238042269187987,
      "loss": 3.7665,
      "step": 402
    },
    {
      "epoch": 5.374581939799331,
      "grad_norm": 0.8287034630775452,
      "learning_rate": 0.00018233592880978868,
      "loss": 3.8265,
      "step": 403
    },
    {
      "epoch": 5.3879598662207355,
      "grad_norm": 1.0219204425811768,
      "learning_rate": 0.00018229143492769743,
      "loss": 3.6829,
      "step": 404
    },
    {
      "epoch": 5.40133779264214,
      "grad_norm": 1.0601041316986084,
      "learning_rate": 0.00018224694104560624,
      "loss": 3.5879,
      "step": 405
    },
    {
      "epoch": 5.414715719063545,
      "grad_norm": 1.2221566438674927,
      "learning_rate": 0.00018220244716351503,
      "loss": 3.6023,
      "step": 406
    },
    {
      "epoch": 5.4280936454849495,
      "grad_norm": 0.9589087963104248,
      "learning_rate": 0.0001821579532814238,
      "loss": 3.9109,
      "step": 407
    },
    {
      "epoch": 5.441471571906354,
      "grad_norm": 1.088295340538025,
      "learning_rate": 0.0001821134593993326,
      "loss": 3.5695,
      "step": 408
    },
    {
      "epoch": 5.454849498327759,
      "grad_norm": 1.1284915208816528,
      "learning_rate": 0.0001820689655172414,
      "loss": 3.6004,
      "step": 409
    },
    {
      "epoch": 5.468227424749164,
      "grad_norm": 1.0108689069747925,
      "learning_rate": 0.00018202447163515018,
      "loss": 3.7797,
      "step": 410
    },
    {
      "epoch": 5.481605351170568,
      "grad_norm": 0.8550918102264404,
      "learning_rate": 0.00018197997775305896,
      "loss": 3.5758,
      "step": 411
    },
    {
      "epoch": 5.494983277591973,
      "grad_norm": 0.8765145540237427,
      "learning_rate": 0.00018193548387096775,
      "loss": 3.7625,
      "step": 412
    },
    {
      "epoch": 5.508361204013378,
      "grad_norm": 1.0253541469573975,
      "learning_rate": 0.00018189098998887656,
      "loss": 3.6033,
      "step": 413
    },
    {
      "epoch": 5.521739130434782,
      "grad_norm": 1.0475622415542603,
      "learning_rate": 0.0001818464961067853,
      "loss": 3.813,
      "step": 414
    },
    {
      "epoch": 5.535117056856187,
      "grad_norm": 1.053133249282837,
      "learning_rate": 0.00018180200222469412,
      "loss": 3.4779,
      "step": 415
    },
    {
      "epoch": 5.548494983277592,
      "grad_norm": 1.0151216983795166,
      "learning_rate": 0.0001817575083426029,
      "loss": 3.9038,
      "step": 416
    },
    {
      "epoch": 5.561872909698996,
      "grad_norm": 1.4666434526443481,
      "learning_rate": 0.00018171301446051169,
      "loss": 3.4735,
      "step": 417
    },
    {
      "epoch": 5.575250836120401,
      "grad_norm": 1.1043344736099243,
      "learning_rate": 0.00018166852057842047,
      "loss": 3.7449,
      "step": 418
    },
    {
      "epoch": 5.588628762541806,
      "grad_norm": 0.900745153427124,
      "learning_rate": 0.00018162402669632925,
      "loss": 3.7401,
      "step": 419
    },
    {
      "epoch": 5.602006688963211,
      "grad_norm": 0.9771101474761963,
      "learning_rate": 0.00018157953281423806,
      "loss": 3.8328,
      "step": 420
    },
    {
      "epoch": 5.615384615384615,
      "grad_norm": 0.9099516272544861,
      "learning_rate": 0.00018153503893214682,
      "loss": 3.6245,
      "step": 421
    },
    {
      "epoch": 5.6287625418060205,
      "grad_norm": 0.9844585657119751,
      "learning_rate": 0.00018149054505005563,
      "loss": 3.5776,
      "step": 422
    },
    {
      "epoch": 5.642140468227424,
      "grad_norm": 1.0481154918670654,
      "learning_rate": 0.0001814460511679644,
      "loss": 3.6304,
      "step": 423
    },
    {
      "epoch": 5.65551839464883,
      "grad_norm": 0.9971081614494324,
      "learning_rate": 0.0001814015572858732,
      "loss": 3.7863,
      "step": 424
    },
    {
      "epoch": 5.668896321070234,
      "grad_norm": 0.9247872829437256,
      "learning_rate": 0.00018135706340378197,
      "loss": 3.7319,
      "step": 425
    },
    {
      "epoch": 5.682274247491639,
      "grad_norm": 0.9895725846290588,
      "learning_rate": 0.00018131256952169078,
      "loss": 3.78,
      "step": 426
    },
    {
      "epoch": 5.695652173913043,
      "grad_norm": 1.0847641229629517,
      "learning_rate": 0.00018126807563959956,
      "loss": 3.8662,
      "step": 427
    },
    {
      "epoch": 5.709030100334449,
      "grad_norm": 0.986259937286377,
      "learning_rate": 0.00018122358175750835,
      "loss": 3.5621,
      "step": 428
    },
    {
      "epoch": 5.722408026755852,
      "grad_norm": 0.9166681170463562,
      "learning_rate": 0.00018117908787541713,
      "loss": 3.6153,
      "step": 429
    },
    {
      "epoch": 5.735785953177258,
      "grad_norm": 1.1331177949905396,
      "learning_rate": 0.00018113459399332594,
      "loss": 3.5976,
      "step": 430
    },
    {
      "epoch": 5.749163879598662,
      "grad_norm": 0.8743540644645691,
      "learning_rate": 0.0001810901001112347,
      "loss": 3.2511,
      "step": 431
    },
    {
      "epoch": 5.762541806020067,
      "grad_norm": 1.0700207948684692,
      "learning_rate": 0.0001810456062291435,
      "loss": 3.7634,
      "step": 432
    },
    {
      "epoch": 5.775919732441472,
      "grad_norm": 0.9412694573402405,
      "learning_rate": 0.00018100111234705229,
      "loss": 3.6264,
      "step": 433
    },
    {
      "epoch": 5.789297658862877,
      "grad_norm": 1.0398496389389038,
      "learning_rate": 0.00018095661846496107,
      "loss": 3.744,
      "step": 434
    },
    {
      "epoch": 5.802675585284281,
      "grad_norm": 0.9605004787445068,
      "learning_rate": 0.00018091212458286985,
      "loss": 3.5532,
      "step": 435
    },
    {
      "epoch": 5.816053511705686,
      "grad_norm": 1.0449095964431763,
      "learning_rate": 0.00018086763070077866,
      "loss": 4.0611,
      "step": 436
    },
    {
      "epoch": 5.829431438127091,
      "grad_norm": 0.9342606663703918,
      "learning_rate": 0.00018082313681868744,
      "loss": 3.9957,
      "step": 437
    },
    {
      "epoch": 5.842809364548495,
      "grad_norm": 0.9687880873680115,
      "learning_rate": 0.00018077864293659622,
      "loss": 3.9299,
      "step": 438
    },
    {
      "epoch": 5.8561872909699,
      "grad_norm": 1.1390576362609863,
      "learning_rate": 0.000180734149054505,
      "loss": 3.6552,
      "step": 439
    },
    {
      "epoch": 5.869565217391305,
      "grad_norm": 0.9280988574028015,
      "learning_rate": 0.00018068965517241382,
      "loss": 3.7828,
      "step": 440
    },
    {
      "epoch": 5.882943143812709,
      "grad_norm": 1.2928193807601929,
      "learning_rate": 0.00018064516129032257,
      "loss": 3.6292,
      "step": 441
    },
    {
      "epoch": 5.896321070234114,
      "grad_norm": 1.0959875583648682,
      "learning_rate": 0.00018060066740823138,
      "loss": 3.4293,
      "step": 442
    },
    {
      "epoch": 5.909698996655519,
      "grad_norm": 1.0713289976119995,
      "learning_rate": 0.00018055617352614016,
      "loss": 3.7767,
      "step": 443
    },
    {
      "epoch": 5.923076923076923,
      "grad_norm": 0.9309440851211548,
      "learning_rate": 0.00018051167964404895,
      "loss": 3.5473,
      "step": 444
    },
    {
      "epoch": 5.936454849498328,
      "grad_norm": 1.0999056100845337,
      "learning_rate": 0.00018046718576195773,
      "loss": 3.9694,
      "step": 445
    },
    {
      "epoch": 5.949832775919733,
      "grad_norm": 1.1073781251907349,
      "learning_rate": 0.00018042269187986654,
      "loss": 3.6882,
      "step": 446
    },
    {
      "epoch": 5.963210702341137,
      "grad_norm": 1.0430257320404053,
      "learning_rate": 0.00018037819799777532,
      "loss": 3.4009,
      "step": 447
    },
    {
      "epoch": 5.976588628762542,
      "grad_norm": 1.1132690906524658,
      "learning_rate": 0.0001803337041156841,
      "loss": 3.8832,
      "step": 448
    },
    {
      "epoch": 5.989966555183947,
      "grad_norm": 1.0147771835327148,
      "learning_rate": 0.00018028921023359289,
      "loss": 3.7117,
      "step": 449
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.458959698677063,
      "learning_rate": 0.0001802447163515017,
      "loss": 3.5745,
      "step": 450
    },
    {
      "epoch": 6.013377926421405,
      "grad_norm": 0.8363592028617859,
      "learning_rate": 0.00018020022246941045,
      "loss": 3.5835,
      "step": 451
    },
    {
      "epoch": 6.026755852842809,
      "grad_norm": 0.8071937561035156,
      "learning_rate": 0.00018015572858731926,
      "loss": 3.5923,
      "step": 452
    },
    {
      "epoch": 6.040133779264214,
      "grad_norm": 0.7746313214302063,
      "learning_rate": 0.00018011123470522804,
      "loss": 3.5688,
      "step": 453
    },
    {
      "epoch": 6.053511705685619,
      "grad_norm": 0.689179539680481,
      "learning_rate": 0.00018006674082313682,
      "loss": 3.412,
      "step": 454
    },
    {
      "epoch": 6.066889632107023,
      "grad_norm": 0.8438050746917725,
      "learning_rate": 0.0001800222469410456,
      "loss": 3.403,
      "step": 455
    },
    {
      "epoch": 6.080267558528428,
      "grad_norm": 0.7670062780380249,
      "learning_rate": 0.00017997775305895442,
      "loss": 3.5029,
      "step": 456
    },
    {
      "epoch": 6.093645484949833,
      "grad_norm": 0.8185870051383972,
      "learning_rate": 0.0001799332591768632,
      "loss": 3.4584,
      "step": 457
    },
    {
      "epoch": 6.107023411371237,
      "grad_norm": 0.9618543386459351,
      "learning_rate": 0.00017988876529477198,
      "loss": 3.6538,
      "step": 458
    },
    {
      "epoch": 6.120401337792642,
      "grad_norm": 0.959724485874176,
      "learning_rate": 0.00017984427141268076,
      "loss": 3.5284,
      "step": 459
    },
    {
      "epoch": 6.133779264214047,
      "grad_norm": 0.8044765591621399,
      "learning_rate": 0.00017979977753058955,
      "loss": 3.2198,
      "step": 460
    },
    {
      "epoch": 6.147157190635451,
      "grad_norm": 0.8287092447280884,
      "learning_rate": 0.00017975528364849833,
      "loss": 3.4977,
      "step": 461
    },
    {
      "epoch": 6.160535117056856,
      "grad_norm": 0.8855329155921936,
      "learning_rate": 0.0001797107897664071,
      "loss": 3.5008,
      "step": 462
    },
    {
      "epoch": 6.173913043478261,
      "grad_norm": 0.8839483857154846,
      "learning_rate": 0.00017966629588431592,
      "loss": 3.6135,
      "step": 463
    },
    {
      "epoch": 6.187290969899665,
      "grad_norm": 0.963446319103241,
      "learning_rate": 0.0001796218020022247,
      "loss": 3.6156,
      "step": 464
    },
    {
      "epoch": 6.20066889632107,
      "grad_norm": 0.896743655204773,
      "learning_rate": 0.00017957730812013348,
      "loss": 3.6623,
      "step": 465
    },
    {
      "epoch": 6.214046822742475,
      "grad_norm": 0.9268617033958435,
      "learning_rate": 0.00017953281423804227,
      "loss": 3.4343,
      "step": 466
    },
    {
      "epoch": 6.2274247491638794,
      "grad_norm": 0.8335449695587158,
      "learning_rate": 0.00017948832035595108,
      "loss": 3.5716,
      "step": 467
    },
    {
      "epoch": 6.240802675585284,
      "grad_norm": 0.7771849036216736,
      "learning_rate": 0.00017944382647385983,
      "loss": 3.5191,
      "step": 468
    },
    {
      "epoch": 6.254180602006689,
      "grad_norm": 0.9157488346099854,
      "learning_rate": 0.00017939933259176864,
      "loss": 3.5583,
      "step": 469
    },
    {
      "epoch": 6.2675585284280935,
      "grad_norm": 0.9348477721214294,
      "learning_rate": 0.00017935483870967742,
      "loss": 3.3137,
      "step": 470
    },
    {
      "epoch": 6.280936454849498,
      "grad_norm": 0.8791135549545288,
      "learning_rate": 0.0001793103448275862,
      "loss": 3.5111,
      "step": 471
    },
    {
      "epoch": 6.294314381270903,
      "grad_norm": 0.9963672757148743,
      "learning_rate": 0.000179265850945495,
      "loss": 3.7518,
      "step": 472
    },
    {
      "epoch": 6.3076923076923075,
      "grad_norm": 0.9291539192199707,
      "learning_rate": 0.0001792213570634038,
      "loss": 3.4524,
      "step": 473
    },
    {
      "epoch": 6.321070234113712,
      "grad_norm": 0.9349279403686523,
      "learning_rate": 0.00017917686318131258,
      "loss": 3.4753,
      "step": 474
    },
    {
      "epoch": 6.334448160535117,
      "grad_norm": 0.8984476327896118,
      "learning_rate": 0.00017913236929922136,
      "loss": 3.7325,
      "step": 475
    },
    {
      "epoch": 6.3478260869565215,
      "grad_norm": 0.8452139496803284,
      "learning_rate": 0.00017908787541713015,
      "loss": 3.8021,
      "step": 476
    },
    {
      "epoch": 6.361204013377926,
      "grad_norm": 0.9418376088142395,
      "learning_rate": 0.00017904338153503895,
      "loss": 3.7426,
      "step": 477
    },
    {
      "epoch": 6.374581939799331,
      "grad_norm": 1.0661097764968872,
      "learning_rate": 0.0001789988876529477,
      "loss": 3.7556,
      "step": 478
    },
    {
      "epoch": 6.3879598662207355,
      "grad_norm": 0.9645984768867493,
      "learning_rate": 0.00017895439377085652,
      "loss": 3.3353,
      "step": 479
    },
    {
      "epoch": 6.40133779264214,
      "grad_norm": 0.9243470430374146,
      "learning_rate": 0.0001789098998887653,
      "loss": 3.5729,
      "step": 480
    },
    {
      "epoch": 6.414715719063545,
      "grad_norm": 0.885061502456665,
      "learning_rate": 0.00017886540600667408,
      "loss": 3.5699,
      "step": 481
    },
    {
      "epoch": 6.4280936454849495,
      "grad_norm": 0.9025402069091797,
      "learning_rate": 0.00017882091212458287,
      "loss": 3.4532,
      "step": 482
    },
    {
      "epoch": 6.441471571906354,
      "grad_norm": 0.9760842323303223,
      "learning_rate": 0.00017877641824249168,
      "loss": 3.7222,
      "step": 483
    },
    {
      "epoch": 6.454849498327759,
      "grad_norm": 1.2709609270095825,
      "learning_rate": 0.00017873192436040046,
      "loss": 3.756,
      "step": 484
    },
    {
      "epoch": 6.468227424749164,
      "grad_norm": 0.904513955116272,
      "learning_rate": 0.00017868743047830924,
      "loss": 3.657,
      "step": 485
    },
    {
      "epoch": 6.481605351170568,
      "grad_norm": 1.158915400505066,
      "learning_rate": 0.00017864293659621802,
      "loss": 3.5897,
      "step": 486
    },
    {
      "epoch": 6.494983277591973,
      "grad_norm": 0.9457879066467285,
      "learning_rate": 0.00017859844271412683,
      "loss": 3.4394,
      "step": 487
    },
    {
      "epoch": 6.508361204013378,
      "grad_norm": 1.04762601852417,
      "learning_rate": 0.0001785539488320356,
      "loss": 3.39,
      "step": 488
    },
    {
      "epoch": 6.521739130434782,
      "grad_norm": 0.9370948076248169,
      "learning_rate": 0.0001785094549499444,
      "loss": 3.4816,
      "step": 489
    },
    {
      "epoch": 6.535117056856187,
      "grad_norm": 1.1307988166809082,
      "learning_rate": 0.00017846496106785318,
      "loss": 3.5079,
      "step": 490
    },
    {
      "epoch": 6.548494983277592,
      "grad_norm": 1.0025054216384888,
      "learning_rate": 0.00017842046718576196,
      "loss": 3.3447,
      "step": 491
    },
    {
      "epoch": 6.561872909698996,
      "grad_norm": 0.9893412590026855,
      "learning_rate": 0.00017837597330367074,
      "loss": 3.4698,
      "step": 492
    },
    {
      "epoch": 6.575250836120401,
      "grad_norm": 1.0476347208023071,
      "learning_rate": 0.00017833147942157955,
      "loss": 3.4955,
      "step": 493
    },
    {
      "epoch": 6.588628762541806,
      "grad_norm": 1.0508891344070435,
      "learning_rate": 0.00017828698553948834,
      "loss": 3.6568,
      "step": 494
    },
    {
      "epoch": 6.602006688963211,
      "grad_norm": 1.0397465229034424,
      "learning_rate": 0.00017824249165739712,
      "loss": 3.5087,
      "step": 495
    },
    {
      "epoch": 6.615384615384615,
      "grad_norm": 0.9884181022644043,
      "learning_rate": 0.0001781979977753059,
      "loss": 3.6778,
      "step": 496
    },
    {
      "epoch": 6.6287625418060205,
      "grad_norm": 1.1187562942504883,
      "learning_rate": 0.0001781535038932147,
      "loss": 3.4345,
      "step": 497
    },
    {
      "epoch": 6.642140468227424,
      "grad_norm": 1.1133880615234375,
      "learning_rate": 0.00017810901001112347,
      "loss": 3.4959,
      "step": 498
    },
    {
      "epoch": 6.65551839464883,
      "grad_norm": 0.9368589520454407,
      "learning_rate": 0.00017806451612903228,
      "loss": 3.628,
      "step": 499
    },
    {
      "epoch": 6.668896321070234,
      "grad_norm": 1.0427212715148926,
      "learning_rate": 0.00017802002224694106,
      "loss": 3.6073,
      "step": 500
    }
  ],
  "logging_steps": 1,
  "max_steps": 4500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 60,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.54743987923712e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}