{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9943851768669285,
  "eval_steps": 500,
  "global_step": 333,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.008983717012914094,
      "grad_norm": 5.839109932109842,
      "learning_rate": 1.663780588235294e-06,
      "loss": 0.851,
      "step": 1
    },
    {
      "epoch": 0.017967434025828188,
      "grad_norm": 5.796557129672097,
      "learning_rate": 3.327561176470588e-06,
      "loss": 0.8589,
      "step": 2
    },
    {
      "epoch": 0.02695115103874228,
      "grad_norm": 5.7589149287471075,
      "learning_rate": 4.991341764705883e-06,
      "loss": 0.8612,
      "step": 3
    },
    {
      "epoch": 0.035934868051656375,
      "grad_norm": 4.220444230457979,
      "learning_rate": 6.655122352941176e-06,
      "loss": 0.8097,
      "step": 4
    },
    {
      "epoch": 0.044918585064570464,
      "grad_norm": 2.2279497783130267,
      "learning_rate": 8.318902941176471e-06,
      "loss": 0.7672,
      "step": 5
    },
    {
      "epoch": 0.05390230207748456,
      "grad_norm": 3.814533547249801,
      "learning_rate": 9.982683529411766e-06,
      "loss": 0.7561,
      "step": 6
    },
    {
      "epoch": 0.06288601909039865,
      "grad_norm": 4.0811292825217835,
      "learning_rate": 1.1646464117647058e-05,
      "loss": 0.76,
      "step": 7
    },
    {
      "epoch": 0.07186973610331275,
      "grad_norm": 3.672264516349975,
      "learning_rate": 1.3310244705882352e-05,
      "loss": 0.7142,
      "step": 8
    },
    {
      "epoch": 0.08085345311622684,
      "grad_norm": 3.4788052125571918,
      "learning_rate": 1.4974025294117648e-05,
      "loss": 0.7167,
      "step": 9
    },
    {
      "epoch": 0.08983717012914093,
      "grad_norm": 2.107646195472596,
      "learning_rate": 1.6637805882352943e-05,
      "loss": 0.6916,
      "step": 10
    },
    {
      "epoch": 0.09882088714205503,
      "grad_norm": 2.3107123413459094,
      "learning_rate": 1.8301586470588235e-05,
      "loss": 0.6619,
      "step": 11
    },
    {
      "epoch": 0.10780460415496912,
      "grad_norm": 1.97063494813433,
      "learning_rate": 1.996536705882353e-05,
      "loss": 0.651,
      "step": 12
    },
    {
      "epoch": 0.11678832116788321,
      "grad_norm": 1.172538415212184,
      "learning_rate": 2.1629147647058824e-05,
      "loss": 0.6292,
      "step": 13
    },
    {
      "epoch": 0.1257720381807973,
      "grad_norm": 1.425434087428503,
      "learning_rate": 2.3292928235294116e-05,
      "loss": 0.624,
      "step": 14
    },
    {
      "epoch": 0.13475575519371139,
      "grad_norm": 1.0298093152751933,
      "learning_rate": 2.4956708823529412e-05,
      "loss": 0.6177,
      "step": 15
    },
    {
      "epoch": 0.1437394722066255,
      "grad_norm": 0.9184264462076409,
      "learning_rate": 2.6620489411764705e-05,
      "loss": 0.6094,
      "step": 16
    },
    {
      "epoch": 0.1527231892195396,
      "grad_norm": 0.9876447317203728,
      "learning_rate": 2.828427e-05,
      "loss": 0.5913,
      "step": 17
    },
    {
      "epoch": 0.16170690623245368,
      "grad_norm": 0.9186614069420004,
      "learning_rate": 2.9948050588235297e-05,
      "loss": 0.5853,
      "step": 18
    },
    {
      "epoch": 0.17069062324536777,
      "grad_norm": 0.9898980015818278,
      "learning_rate": 3.161183117647059e-05,
      "loss": 0.5837,
      "step": 19
    },
    {
      "epoch": 0.17967434025828186,
      "grad_norm": 1.0813008236428436,
      "learning_rate": 3.3275611764705885e-05,
      "loss": 0.5826,
      "step": 20
    },
    {
      "epoch": 0.18865805727119594,
      "grad_norm": 1.1477562316475312,
      "learning_rate": 3.493939235294118e-05,
      "loss": 0.5722,
      "step": 21
    },
    {
      "epoch": 0.19764177428411006,
      "grad_norm": 0.8464148409927464,
      "learning_rate": 3.660317294117647e-05,
      "loss": 0.5673,
      "step": 22
    },
    {
      "epoch": 0.20662549129702415,
      "grad_norm": 0.7702840400962941,
      "learning_rate": 3.826695352941177e-05,
      "loss": 0.5588,
      "step": 23
    },
    {
      "epoch": 0.21560920830993824,
      "grad_norm": 1.1144594493299889,
      "learning_rate": 3.993073411764706e-05,
      "loss": 0.5584,
      "step": 24
    },
    {
      "epoch": 0.22459292532285233,
      "grad_norm": 1.1476160682417311,
      "learning_rate": 4.1594514705882355e-05,
      "loss": 0.5569,
      "step": 25
    },
    {
      "epoch": 0.23357664233576642,
      "grad_norm": 1.1535268292624339,
      "learning_rate": 4.325829529411765e-05,
      "loss": 0.5553,
      "step": 26
    },
    {
      "epoch": 0.2425603593486805,
      "grad_norm": 1.5110809161997512,
      "learning_rate": 4.492207588235294e-05,
      "loss": 0.554,
      "step": 27
    },
    {
      "epoch": 0.2515440763615946,
      "grad_norm": 1.1832182446871022,
      "learning_rate": 4.658585647058823e-05,
      "loss": 0.5421,
      "step": 28
    },
    {
      "epoch": 0.2605277933745087,
      "grad_norm": 1.1859526315720494,
      "learning_rate": 4.824963705882353e-05,
      "loss": 0.5413,
      "step": 29
    },
    {
      "epoch": 0.26951151038742277,
      "grad_norm": 1.5427900149672942,
      "learning_rate": 4.9913417647058825e-05,
      "loss": 0.5447,
      "step": 30
    },
    {
      "epoch": 0.2784952274003369,
      "grad_norm": 1.1712836835670162,
      "learning_rate": 5.157719823529412e-05,
      "loss": 0.5346,
      "step": 31
    },
    {
      "epoch": 0.287478944413251,
      "grad_norm": 1.3657285701127917,
      "learning_rate": 5.324097882352941e-05,
      "loss": 0.54,
      "step": 32
    },
    {
      "epoch": 0.29646266142616506,
      "grad_norm": 1.401113516783269,
      "learning_rate": 5.490475941176471e-05,
      "loss": 0.5272,
      "step": 33
    },
    {
      "epoch": 0.3054463784390792,
      "grad_norm": 1.283015659035394,
      "learning_rate": 5.656854e-05,
      "loss": 0.5269,
      "step": 34
    },
    {
      "epoch": 0.31443009545199324,
      "grad_norm": 1.1625204652087118,
      "learning_rate": 5.65669787647368e-05,
      "loss": 0.5254,
      "step": 35
    },
    {
      "epoch": 0.32341381246490736,
      "grad_norm": 1.9389962167077017,
      "learning_rate": 5.6562295231301345e-05,
      "loss": 0.5325,
      "step": 36
    },
    {
      "epoch": 0.3323975294778215,
      "grad_norm": 1.2095439312643286,
      "learning_rate": 5.655448991673705e-05,
      "loss": 0.5247,
      "step": 37
    },
    {
      "epoch": 0.34138124649073553,
      "grad_norm": 2.1581962123054512,
      "learning_rate": 5.6543563682719456e-05,
      "loss": 0.5301,
      "step": 38
    },
    {
      "epoch": 0.35036496350364965,
      "grad_norm": 1.645168282852934,
      "learning_rate": 5.652951773546123e-05,
      "loss": 0.5251,
      "step": 39
    },
    {
      "epoch": 0.3593486805165637,
      "grad_norm": 1.7810864857463629,
      "learning_rate": 5.651235362557887e-05,
      "loss": 0.529,
      "step": 40
    },
    {
      "epoch": 0.36833239752947783,
      "grad_norm": 1.3721422433398012,
      "learning_rate": 5.649207324792161e-05,
      "loss": 0.5263,
      "step": 41
    },
    {
      "epoch": 0.3773161145423919,
      "grad_norm": 1.7315681950487964,
      "learning_rate": 5.646867884136221e-05,
      "loss": 0.5233,
      "step": 42
    },
    {
      "epoch": 0.386299831555306,
      "grad_norm": 1.2328449836692366,
      "learning_rate": 5.644217298854978e-05,
      "loss": 0.5121,
      "step": 43
    },
    {
      "epoch": 0.3952835485682201,
      "grad_norm": 1.688017129833438,
      "learning_rate": 5.641255861562469e-05,
      "loss": 0.5076,
      "step": 44
    },
    {
      "epoch": 0.4042672655811342,
      "grad_norm": 1.3793575135144216,
      "learning_rate": 5.637983899189553e-05,
      "loss": 0.5178,
      "step": 45
    },
    {
      "epoch": 0.4132509825940483,
      "grad_norm": 1.1818050497678425,
      "learning_rate": 5.634401772947817e-05,
      "loss": 0.5099,
      "step": 46
    },
    {
      "epoch": 0.42223469960696236,
      "grad_norm": 1.334664385713327,
      "learning_rate": 5.630509878289701e-05,
      "loss": 0.5125,
      "step": 47
    },
    {
      "epoch": 0.4312184166198765,
      "grad_norm": 1.1824189144503987,
      "learning_rate": 5.626308644864844e-05,
      "loss": 0.5035,
      "step": 48
    },
    {
      "epoch": 0.4402021336327906,
      "grad_norm": 1.8310997335960193,
      "learning_rate": 5.621798536472649e-05,
      "loss": 0.5129,
      "step": 49
    },
    {
      "epoch": 0.44918585064570465,
      "grad_norm": 0.9498119984179156,
      "learning_rate": 5.616980051011082e-05,
      "loss": 0.4933,
      "step": 50
    },
    {
      "epoch": 0.45816956765861877,
      "grad_norm": 1.3738927247826827,
      "learning_rate": 5.611853720421709e-05,
      "loss": 0.5114,
      "step": 51
    },
    {
      "epoch": 0.46715328467153283,
      "grad_norm": 1.3239245607536279,
      "learning_rate": 5.606420110630966e-05,
      "loss": 0.5051,
      "step": 52
    },
    {
      "epoch": 0.47613700168444695,
      "grad_norm": 1.3491292246681823,
      "learning_rate": 5.6006798214876905e-05,
      "loss": 0.4984,
      "step": 53
    },
    {
      "epoch": 0.485120718697361,
      "grad_norm": 1.556526288137193,
      "learning_rate": 5.5946334866968935e-05,
      "loss": 0.5057,
      "step": 54
    },
    {
      "epoch": 0.4941044357102751,
      "grad_norm": 0.8396609232112651,
      "learning_rate": 5.5882817737498054e-05,
      "loss": 0.5067,
      "step": 55
    },
    {
      "epoch": 0.5030881527231892,
      "grad_norm": 1.3086549421559865,
      "learning_rate": 5.581625383850187e-05,
      "loss": 0.5016,
      "step": 56
    },
    {
      "epoch": 0.5120718697361033,
      "grad_norm": 1.6438518323894828,
      "learning_rate": 5.5746650518369164e-05,
      "loss": 0.5105,
      "step": 57
    },
    {
      "epoch": 0.5210555867490174,
      "grad_norm": 0.9704759451995054,
      "learning_rate": 5.5674015461028693e-05,
      "loss": 0.501,
      "step": 58
    },
    {
      "epoch": 0.5300393037619315,
      "grad_norm": 1.520513804956881,
      "learning_rate": 5.559835668510092e-05,
      "loss": 0.5055,
      "step": 59
    },
    {
      "epoch": 0.5390230207748455,
      "grad_norm": 1.4886814176870304,
      "learning_rate": 5.5519682543012745e-05,
      "loss": 0.4997,
      "step": 60
    },
    {
      "epoch": 0.5480067377877597,
      "grad_norm": 0.9983783244132893,
      "learning_rate": 5.5438001720075464e-05,
      "loss": 0.4999,
      "step": 61
    },
    {
      "epoch": 0.5569904548006738,
      "grad_norm": 1.8027514771648614,
      "learning_rate": 5.535332323352595e-05,
      "loss": 0.4973,
      "step": 62
    },
    {
      "epoch": 0.5659741718135879,
      "grad_norm": 1.5234595633592394,
      "learning_rate": 5.526565643153117e-05,
      "loss": 0.5027,
      "step": 63
    },
    {
      "epoch": 0.574957888826502,
      "grad_norm": 1.5395421556206257,
      "learning_rate": 5.517501099215618e-05,
      "loss": 0.4927,
      "step": 64
    },
    {
      "epoch": 0.583941605839416,
      "grad_norm": 1.2783343984201738,
      "learning_rate": 5.5081396922295734e-05,
      "loss": 0.4927,
      "step": 65
    },
    {
      "epoch": 0.5929253228523301,
      "grad_norm": 1.3855491947405612,
      "learning_rate": 5.498482455656953e-05,
      "loss": 0.4922,
      "step": 66
    },
    {
      "epoch": 0.6019090398652442,
      "grad_norm": 1.1582514991567736,
      "learning_rate": 5.488530455618133e-05,
      "loss": 0.4911,
      "step": 67
    },
    {
      "epoch": 0.6108927568781584,
      "grad_norm": 1.4059224870864222,
      "learning_rate": 5.4782847907741996e-05,
      "loss": 0.4947,
      "step": 68
    },
    {
      "epoch": 0.6198764738910725,
      "grad_norm": 1.1973724677392805,
      "learning_rate": 5.4677465922056604e-05,
      "loss": 0.4891,
      "step": 69
    },
    {
      "epoch": 0.6288601909039865,
      "grad_norm": 1.3369194820930967,
      "learning_rate": 5.456917023287581e-05,
      "loss": 0.4877,
      "step": 70
    },
    {
      "epoch": 0.6378439079169006,
      "grad_norm": 1.2691892558171078,
      "learning_rate": 5.445797279561149e-05,
      "loss": 0.4895,
      "step": 71
    },
    {
      "epoch": 0.6468276249298147,
      "grad_norm": 1.0922464334569515,
      "learning_rate": 5.434388588601693e-05,
      "loss": 0.4803,
      "step": 72
    },
    {
      "epoch": 0.6558113419427288,
      "grad_norm": 0.8753013887933451,
      "learning_rate": 5.422692209883164e-05,
      "loss": 0.4914,
      "step": 73
    },
    {
      "epoch": 0.664795058955643,
      "grad_norm": 0.9796697285671012,
      "learning_rate": 5.4107094346390925e-05,
      "loss": 0.4869,
      "step": 74
    },
    {
      "epoch": 0.673778775968557,
      "grad_norm": 1.2844167323369493,
      "learning_rate": 5.398441585720041e-05,
      "loss": 0.4867,
      "step": 75
    },
    {
      "epoch": 0.6827624929814711,
      "grad_norm": 1.4887584341619542,
      "learning_rate": 5.3858900174475716e-05,
      "loss": 0.4991,
      "step": 76
    },
    {
      "epoch": 0.6917462099943852,
      "grad_norm": 0.8921558529768149,
      "learning_rate": 5.373056115464729e-05,
      "loss": 0.4867,
      "step": 77
    },
    {
      "epoch": 0.7007299270072993,
      "grad_norm": 1.2859329068604128,
      "learning_rate": 5.359941296583069e-05,
      "loss": 0.4882,
      "step": 78
    },
    {
      "epoch": 0.7097136440202133,
      "grad_norm": 0.7544133782391679,
      "learning_rate": 5.346547008626259e-05,
      "loss": 0.4804,
      "step": 79
    },
    {
      "epoch": 0.7186973610331274,
      "grad_norm": 1.2158913130951334,
      "learning_rate": 5.332874730270231e-05,
      "loss": 0.4965,
      "step": 80
    },
    {
      "epoch": 0.7276810780460415,
      "grad_norm": 1.5397019534433478,
      "learning_rate": 5.3189259708799525e-05,
      "loss": 0.4863,
      "step": 81
    },
    {
      "epoch": 0.7366647950589557,
      "grad_norm": 1.0368691373073795,
      "learning_rate": 5.304702270342788e-05,
      "loss": 0.4709,
      "step": 82
    },
    {
      "epoch": 0.7456485120718698,
      "grad_norm": 1.6327070593356623,
      "learning_rate": 5.290205198898512e-05,
      "loss": 0.4935,
      "step": 83
    },
    {
      "epoch": 0.7546322290847838,
      "grad_norm": 0.7996850897364002,
      "learning_rate": 5.275436356965955e-05,
      "loss": 0.4833,
      "step": 84
    },
    {
      "epoch": 0.7636159460976979,
      "grad_norm": 1.3138953152120028,
      "learning_rate": 5.260397374966324e-05,
      "loss": 0.4831,
      "step": 85
    },
    {
      "epoch": 0.772599663110612,
      "grad_norm": 1.0413080501496228,
      "learning_rate": 5.245089913143211e-05,
      "loss": 0.4857,
      "step": 86
    },
    {
      "epoch": 0.7815833801235261,
      "grad_norm": 1.2505485373744896,
      "learning_rate": 5.229515661379311e-05,
      "loss": 0.4783,
      "step": 87
    },
    {
      "epoch": 0.7905670971364402,
      "grad_norm": 0.8250922322384497,
      "learning_rate": 5.213676339009861e-05,
      "loss": 0.4888,
      "step": 88
    },
    {
      "epoch": 0.7995508141493542,
      "grad_norm": 0.930860026309023,
      "learning_rate": 5.197573694632837e-05,
      "loss": 0.4722,
      "step": 89
    },
    {
      "epoch": 0.8085345311622684,
      "grad_norm": 1.1999339483974405,
      "learning_rate": 5.181209505915914e-05,
      "loss": 0.4844,
      "step": 90
    },
    {
      "epoch": 0.8175182481751825,
      "grad_norm": 0.9824068299768655,
      "learning_rate": 5.164585579400215e-05,
      "loss": 0.4835,
      "step": 91
    },
    {
      "epoch": 0.8265019651880966,
      "grad_norm": 1.506768206352777,
      "learning_rate": 5.1477037503008845e-05,
      "loss": 0.4795,
      "step": 92
    },
    {
      "epoch": 0.8354856822010107,
      "grad_norm": 0.7471067724218583,
      "learning_rate": 5.130565882304478e-05,
      "loss": 0.4752,
      "step": 93
    },
    {
      "epoch": 0.8444693992139247,
      "grad_norm": 1.231448677263746,
      "learning_rate": 5.113173867363228e-05,
      "loss": 0.4789,
      "step": 94
    },
    {
      "epoch": 0.8534531162268388,
      "grad_norm": 1.3024759491093874,
      "learning_rate": 5.095529625486171e-05,
      "loss": 0.4805,
      "step": 95
    },
    {
      "epoch": 0.862436833239753,
      "grad_norm": 0.9214289630598883,
      "learning_rate": 5.0776351045271936e-05,
      "loss": 0.486,
      "step": 96
    },
    {
      "epoch": 0.8714205502526671,
      "grad_norm": 1.0404626244101485,
      "learning_rate": 5.0594922799699925e-05,
      "loss": 0.4845,
      "step": 97
    },
    {
      "epoch": 0.8804042672655812,
      "grad_norm": 0.849414300185437,
      "learning_rate": 5.0411031547099916e-05,
      "loss": 0.467,
      "step": 98
    },
    {
      "epoch": 0.8893879842784952,
      "grad_norm": 0.6937685475658341,
      "learning_rate": 5.0224697588332306e-05,
      "loss": 0.4808,
      "step": 99
    },
    {
      "epoch": 0.8983717012914093,
      "grad_norm": 0.9214456229044827,
      "learning_rate": 5.003594149392247e-05,
      "loss": 0.4686,
      "step": 100
    },
    {
      "epoch": 0.9073554183043234,
      "grad_norm": 0.8633442974924467,
      "learning_rate": 4.984478410178992e-05,
      "loss": 0.481,
      "step": 101
    },
    {
      "epoch": 0.9163391353172375,
      "grad_norm": 1.0904707694598144,
      "learning_rate": 4.965124651494785e-05,
      "loss": 0.4677,
      "step": 102
    },
    {
      "epoch": 0.9253228523301515,
      "grad_norm": 0.8481076014465915,
      "learning_rate": 4.9455350099173434e-05,
      "loss": 0.4802,
      "step": 103
    },
    {
      "epoch": 0.9343065693430657,
      "grad_norm": 0.6006636997635961,
      "learning_rate": 4.925711648064916e-05,
      "loss": 0.4694,
      "step": 104
    },
    {
      "epoch": 0.9432902863559798,
      "grad_norm": 0.7238908358412791,
      "learning_rate": 4.9056567543575374e-05,
      "loss": 0.4809,
      "step": 105
    },
    {
      "epoch": 0.9522740033688939,
      "grad_norm": 0.719964931081471,
      "learning_rate": 4.885372542775435e-05,
      "loss": 0.4677,
      "step": 106
    },
    {
      "epoch": 0.961257720381808,
      "grad_norm": 0.6462909305498641,
      "learning_rate": 4.864861252614612e-05,
      "loss": 0.47,
      "step": 107
    },
    {
      "epoch": 0.970241437394722,
      "grad_norm": 0.605236737922692,
      "learning_rate": 4.844125148239645e-05,
      "loss": 0.4604,
      "step": 108
    },
    {
      "epoch": 0.9792251544076361,
      "grad_norm": 0.6196855103029525,
      "learning_rate": 4.823166518833697e-05,
      "loss": 0.4686,
      "step": 109
    },
    {
      "epoch": 0.9882088714205502,
      "grad_norm": 0.5910630181346482,
      "learning_rate": 4.801987678145811e-05,
      "loss": 0.4771,
      "step": 110
    },
    {
      "epoch": 0.9971925884334644,
      "grad_norm": 0.5265183055483387,
      "learning_rate": 4.7805909642354734e-05,
      "loss": 0.4701,
      "step": 111
    },
    {
      "epoch": 1.0075800112296462,
      "grad_norm": 0.7416855722204436,
      "learning_rate": 4.7589787392145085e-05,
      "loss": 0.8384,
      "step": 112
    },
    {
      "epoch": 1.0165637282425604,
      "grad_norm": 0.9235671375653838,
      "learning_rate": 4.737153388986303e-05,
      "loss": 0.4427,
      "step": 113
    },
    {
      "epoch": 1.0255474452554745,
      "grad_norm": 1.2985752147785057,
      "learning_rate": 4.7151173229824185e-05,
      "loss": 0.4674,
      "step": 114
    },
    {
      "epoch": 1.0345311622683886,
      "grad_norm": 0.8686104223850358,
      "learning_rate": 4.6928729738965966e-05,
      "loss": 0.4585,
      "step": 115
    },
    {
      "epoch": 1.0435148792813027,
      "grad_norm": 1.4248620564884518,
      "learning_rate": 4.6704227974162e-05,
      "loss": 0.4664,
      "step": 116
    },
    {
      "epoch": 1.0524985962942168,
      "grad_norm": 0.8090345615734776,
      "learning_rate": 4.647769271951114e-05,
      "loss": 0.4526,
      "step": 117
    },
    {
      "epoch": 1.0614823133071307,
      "grad_norm": 1.8942652592210458,
      "learning_rate": 4.624914898360141e-05,
      "loss": 0.4616,
      "step": 118
    },
    {
      "epoch": 1.0704660303200448,
      "grad_norm": 1.176745396077979,
      "learning_rate": 4.601862199674913e-05,
      "loss": 0.4517,
      "step": 119
    },
    {
      "epoch": 1.079449747332959,
      "grad_norm": 1.9084152203109792,
      "learning_rate": 4.5786137208213634e-05,
      "loss": 0.4512,
      "step": 120
    },
    {
      "epoch": 1.088433464345873,
      "grad_norm": 1.837711635191509,
      "learning_rate": 4.555172028338775e-05,
      "loss": 0.4642,
      "step": 121
    },
    {
      "epoch": 1.0974171813587872,
      "grad_norm": 0.8682381358770546,
      "learning_rate": 4.531539710096439e-05,
      "loss": 0.4458,
      "step": 122
    },
    {
      "epoch": 1.1064008983717013,
      "grad_norm": 1.2949713196196633,
      "learning_rate": 4.507719375007978e-05,
      "loss": 0.4659,
      "step": 123
    },
    {
      "epoch": 1.1153846153846154,
      "grad_norm": 0.6820031143942866,
      "learning_rate": 4.483713652743316e-05,
      "loss": 0.4422,
      "step": 124
    },
    {
      "epoch": 1.1243683323975295,
      "grad_norm": 0.9015249547203725,
      "learning_rate": 4.459525193438388e-05,
      "loss": 0.4674,
      "step": 125
    },
    {
      "epoch": 1.1333520494104437,
      "grad_norm": 0.7596159903360272,
      "learning_rate": 4.4351566674025625e-05,
      "loss": 0.4397,
      "step": 126
    },
    {
      "epoch": 1.1423357664233578,
      "grad_norm": 0.8048846249914353,
      "learning_rate": 4.410610764823863e-05,
      "loss": 0.4537,
      "step": 127
    },
    {
      "epoch": 1.1513194834362717,
      "grad_norm": 0.5674553882616571,
      "learning_rate": 4.3858901954719706e-05,
      "loss": 0.4431,
      "step": 128
    },
    {
      "epoch": 1.1603032004491858,
      "grad_norm": 0.7662516848679161,
      "learning_rate": 4.3609976883990836e-05,
      "loss": 0.4564,
      "step": 129
    },
    {
      "epoch": 1.1692869174621,
      "grad_norm": 0.5438455650436745,
      "learning_rate": 4.335935991638637e-05,
      "loss": 0.4461,
      "step": 130
    },
    {
      "epoch": 1.178270634475014,
      "grad_norm": 0.659318129425458,
      "learning_rate": 4.310707871901933e-05,
      "loss": 0.4489,
      "step": 131
    },
    {
      "epoch": 1.1872543514879281,
      "grad_norm": 0.603233928019015,
      "learning_rate": 4.2853161142727056e-05,
      "loss": 0.4463,
      "step": 132
    },
    {
      "epoch": 1.1962380685008422,
      "grad_norm": 0.5437863962201774,
      "learning_rate": 4.25976352189966e-05,
      "loss": 0.4476,
      "step": 133
    },
    {
      "epoch": 1.2052217855137564,
      "grad_norm": 0.44198003824815413,
      "learning_rate": 4.234052915687014e-05,
      "loss": 0.4417,
      "step": 134
    },
    {
      "epoch": 1.2142055025266705,
      "grad_norm": 0.4940269180275878,
      "learning_rate": 4.208187133983084e-05,
      "loss": 0.4437,
      "step": 135
    },
    {
      "epoch": 1.2231892195395846,
      "grad_norm": 0.4676186442195625,
      "learning_rate": 4.18216903226694e-05,
      "loss": 0.4429,
      "step": 136
    },
    {
      "epoch": 1.2321729365524985,
      "grad_norm": 0.557497711331083,
      "learning_rate": 4.156001482833174e-05,
      "loss": 0.4464,
      "step": 137
    },
    {
      "epoch": 1.2411566535654126,
      "grad_norm": 0.47586024576684477,
      "learning_rate": 4.1296873744748095e-05,
      "loss": 0.4426,
      "step": 138
    },
    {
      "epoch": 1.2501403705783267,
      "grad_norm": 0.32603863165308805,
      "learning_rate": 4.103229612164391e-05,
      "loss": 0.4416,
      "step": 139
    },
    {
      "epoch": 1.2591240875912408,
      "grad_norm": 0.4351165089036505,
      "learning_rate": 4.076631116733286e-05,
      "loss": 0.4401,
      "step": 140
    },
    {
      "epoch": 1.268107804604155,
      "grad_norm": 0.4958952788514796,
      "learning_rate": 4.0498948245492365e-05,
      "loss": 0.4452,
      "step": 141
    },
    {
      "epoch": 1.277091521617069,
      "grad_norm": 0.5509848964383834,
      "learning_rate": 4.023023687192194e-05,
      "loss": 0.4411,
      "step": 142
    },
    {
      "epoch": 1.2860752386299832,
      "grad_norm": 0.3808831227241545,
      "learning_rate": 3.996020671128483e-05,
      "loss": 0.4443,
      "step": 143
    },
    {
      "epoch": 1.2950589556428973,
      "grad_norm": 0.36241744805641257,
      "learning_rate": 3.9688887573833065e-05,
      "loss": 0.45,
      "step": 144
    },
    {
      "epoch": 1.3040426726558114,
      "grad_norm": 0.36700641622974883,
      "learning_rate": 3.941630941211662e-05,
      "loss": 0.4307,
      "step": 145
    },
    {
      "epoch": 1.3130263896687255,
      "grad_norm": 0.3136123844114524,
      "learning_rate": 3.914250231767668e-05,
      "loss": 0.4428,
      "step": 146
    },
    {
      "epoch": 1.3220101066816397,
      "grad_norm": 0.30609996444955184,
      "learning_rate": 3.886749651772372e-05,
      "loss": 0.4433,
      "step": 147
    },
    {
      "epoch": 1.3309938236945535,
      "grad_norm": 0.35148333104030377,
      "learning_rate": 3.8591322371800516e-05,
      "loss": 0.448,
      "step": 148
    },
    {
      "epoch": 1.3399775407074677,
      "grad_norm": 0.3406382310678225,
      "learning_rate": 3.831401036843058e-05,
      "loss": 0.4356,
      "step": 149
    },
    {
      "epoch": 1.3489612577203818,
      "grad_norm": 0.32870885630433616,
      "learning_rate": 3.8035591121752334e-05,
      "loss": 0.4433,
      "step": 150
    },
    {
      "epoch": 1.357944974733296,
      "grad_norm": 0.40610208793950014,
      "learning_rate": 3.7756095368139454e-05,
      "loss": 0.4419,
      "step": 151
    },
    {
      "epoch": 1.36692869174621,
      "grad_norm": 0.3410029435891093,
      "learning_rate": 3.747555396280769e-05,
      "loss": 0.4284,
      "step": 152
    },
    {
      "epoch": 1.3759124087591241,
      "grad_norm": 0.30509070420240164,
      "learning_rate": 3.719399787640854e-05,
      "loss": 0.4346,
      "step": 153
    },
    {
      "epoch": 1.3848961257720382,
      "grad_norm": 0.30952788069600484,
      "learning_rate": 3.691145819161026e-05,
      "loss": 0.4366,
      "step": 154
    },
    {
      "epoch": 1.3938798427849521,
      "grad_norm": 0.28009014545926486,
      "learning_rate": 3.6627966099666466e-05,
      "loss": 0.4322,
      "step": 155
    },
    {
      "epoch": 1.4028635597978663,
      "grad_norm": 0.255197994228949,
      "learning_rate": 3.6343552896972686e-05,
      "loss": 0.4297,
      "step": 156
    },
    {
      "epoch": 1.4118472768107804,
      "grad_norm": 0.26775780849454006,
      "learning_rate": 3.605824998161141e-05,
      "loss": 0.4423,
      "step": 157
    },
    {
      "epoch": 1.4208309938236945,
      "grad_norm": 0.2748112146974553,
      "learning_rate": 3.5772088849885886e-05,
      "loss": 0.4316,
      "step": 158
    },
    {
      "epoch": 1.4298147108366086,
      "grad_norm": 0.272040591675239,
      "learning_rate": 3.548510109284296e-05,
      "loss": 0.4464,
      "step": 159
    },
    {
      "epoch": 1.4387984278495227,
      "grad_norm": 0.23246881072840514,
      "learning_rate": 3.519731839278567e-05,
      "loss": 0.4409,
      "step": 160
    },
    {
      "epoch": 1.4477821448624368,
      "grad_norm": 0.24504527703462542,
      "learning_rate": 3.4908772519775565e-05,
      "loss": 0.4419,
      "step": 161
    },
    {
      "epoch": 1.456765861875351,
      "grad_norm": 0.24765234500656855,
      "learning_rate": 3.461949532812546e-05,
      "loss": 0.4478,
      "step": 162
    },
    {
      "epoch": 1.465749578888265,
      "grad_norm": 0.327477057941263,
      "learning_rate": 3.43295187528828e-05,
      "loss": 0.4411,
      "step": 163
    },
    {
      "epoch": 1.4747332959011792,
      "grad_norm": 0.27408220536723477,
      "learning_rate": 3.403887480630422e-05,
      "loss": 0.4372,
      "step": 164
    },
    {
      "epoch": 1.4837170129140933,
      "grad_norm": 0.2479972152670547,
      "learning_rate": 3.374759557432146e-05,
      "loss": 0.44,
      "step": 165
    },
    {
      "epoch": 1.4927007299270074,
      "grad_norm": 0.2576799538776205,
      "learning_rate": 3.345571321299926e-05,
      "loss": 0.4345,
      "step": 166
    },
    {
      "epoch": 1.5016844469399215,
      "grad_norm": 0.2949577781981658,
      "learning_rate": 3.316325994498539e-05,
      "loss": 0.4426,
      "step": 167
    },
    {
      "epoch": 1.5106681639528357,
      "grad_norm": 0.25387337797793086,
      "learning_rate": 3.28702680559535e-05,
      "loss": 0.4414,
      "step": 168
    },
    {
      "epoch": 1.5196518809657495,
      "grad_norm": 0.29898671970481416,
      "learning_rate": 3.2576769891038794e-05,
      "loss": 0.4295,
      "step": 169
    },
    {
      "epoch": 1.5286355979786637,
      "grad_norm": 0.22913788989001677,
      "learning_rate": 3.2282797851267353e-05,
      "loss": 0.4402,
      "step": 170
    },
    {
      "epoch": 1.5376193149915778,
      "grad_norm": 0.3095695481139338,
      "learning_rate": 3.198838438997912e-05,
      "loss": 0.4367,
      "step": 171
    },
    {
      "epoch": 1.546603032004492,
      "grad_norm": 0.2942100195557735,
      "learning_rate": 3.169356200924522e-05,
      "loss": 0.441,
      "step": 172
    },
    {
      "epoch": 1.5555867490174058,
      "grad_norm": 0.2009760877564207,
      "learning_rate": 3.1398363256279894e-05,
      "loss": 0.4361,
      "step": 173
    },
    {
      "epoch": 1.56457046603032,
      "grad_norm": 0.22271313757898464,
      "learning_rate": 3.110282071984731e-05,
      "loss": 0.429,
      "step": 174
    },
    {
      "epoch": 1.573554183043234,
      "grad_norm": 0.23273324004009835,
      "learning_rate": 3.080696702666401e-05,
      "loss": 0.444,
      "step": 175
    },
    {
      "epoch": 1.5825379000561481,
      "grad_norm": 0.23105576612308468,
      "learning_rate": 3.051083483779696e-05,
      "loss": 0.4415,
      "step": 176
    },
    {
      "epoch": 1.5915216170690623,
      "grad_norm": 0.2707194719098357,
      "learning_rate": 3.0214456845057964e-05,
      "loss": 0.4337,
      "step": 177
    },
    {
      "epoch": 1.6005053340819764,
      "grad_norm": 0.19401437496621246,
      "learning_rate": 2.9917865767394592e-05,
      "loss": 0.4293,
      "step": 178
    },
    {
      "epoch": 1.6094890510948905,
      "grad_norm": 0.23447424010049614,
      "learning_rate": 2.9621094347278115e-05,
      "loss": 0.4381,
      "step": 179
    },
    {
      "epoch": 1.6184727681078046,
      "grad_norm": 0.24721825324120247,
      "learning_rate": 2.9324175347088936e-05,
      "loss": 0.4347,
      "step": 180
    },
    {
      "epoch": 1.6274564851207187,
      "grad_norm": 0.18188660812700427,
      "learning_rate": 2.9027141545499668e-05,
      "loss": 0.4393,
      "step": 181
    },
    {
      "epoch": 1.6364402021336328,
      "grad_norm": 0.190718000775263,
      "learning_rate": 2.873002573385654e-05,
      "loss": 0.4302,
      "step": 182
    },
    {
      "epoch": 1.645423919146547,
      "grad_norm": 0.20646822886164018,
      "learning_rate": 2.84328607125594e-05,
      "loss": 0.4265,
      "step": 183
    },
    {
      "epoch": 1.654407636159461,
      "grad_norm": 0.18501972197639874,
      "learning_rate": 2.813567928744061e-05,
      "loss": 0.4264,
      "step": 184
    },
    {
      "epoch": 1.6633913531723752,
      "grad_norm": 0.2131469646078243,
      "learning_rate": 2.7838514266143464e-05,
      "loss": 0.4353,
      "step": 185
    },
    {
      "epoch": 1.6723750701852893,
      "grad_norm": 0.20765635355407858,
      "learning_rate": 2.754139845450034e-05,
      "loss": 0.4358,
      "step": 186
    },
    {
      "epoch": 1.6813587871982034,
      "grad_norm": 0.21408480102506994,
      "learning_rate": 2.7244364652911073e-05,
      "loss": 0.436,
      "step": 187
    },
    {
      "epoch": 1.6903425042111173,
      "grad_norm": 0.20394614955447768,
      "learning_rate": 2.6947445652721887e-05,
      "loss": 0.4317,
      "step": 188
    },
    {
      "epoch": 1.6993262212240314,
      "grad_norm": 0.2462732253149224,
      "learning_rate": 2.6650674232605416e-05,
      "loss": 0.4381,
      "step": 189
    },
    {
      "epoch": 1.7083099382369455,
      "grad_norm": 0.19402664296982586,
      "learning_rate": 2.635408315494204e-05,
      "loss": 0.4289,
      "step": 190
    },
    {
      "epoch": 1.7172936552498597,
      "grad_norm": 0.25323884789633694,
      "learning_rate": 2.6057705162203045e-05,
      "loss": 0.4387,
      "step": 191
    },
    {
      "epoch": 1.7262773722627736,
      "grad_norm": 0.17952016249190222,
      "learning_rate": 2.5761572973335996e-05,
      "loss": 0.4392,
      "step": 192
    },
    {
      "epoch": 1.7352610892756877,
      "grad_norm": 0.21550956031191665,
      "learning_rate": 2.5465719280152693e-05,
      "loss": 0.4385,
      "step": 193
    },
    {
      "epoch": 1.7442448062886018,
      "grad_norm": 0.2260946384930941,
      "learning_rate": 2.5170176743720114e-05,
      "loss": 0.4297,
      "step": 194
    },
    {
      "epoch": 1.753228523301516,
      "grad_norm": 0.189354294580711,
      "learning_rate": 2.487497799075478e-05,
      "loss": 0.4402,
      "step": 195
    },
    {
      "epoch": 1.76221224031443,
      "grad_norm": 0.18347107243203223,
      "learning_rate": 2.4580155610020893e-05,
      "loss": 0.4436,
      "step": 196
    },
    {
      "epoch": 1.7711959573273441,
      "grad_norm": 0.23718361693873785,
      "learning_rate": 2.4285742148732662e-05,
      "loss": 0.431,
      "step": 197
    },
    {
      "epoch": 1.7801796743402583,
      "grad_norm": 0.15875452261920012,
      "learning_rate": 2.399177010896121e-05,
      "loss": 0.4227,
      "step": 198
    },
    {
      "epoch": 1.7891633913531724,
      "grad_norm": 0.20090739499698712,
      "learning_rate": 2.3698271944046514e-05,
      "loss": 0.4287,
      "step": 199
    },
    {
      "epoch": 1.7981471083660865,
      "grad_norm": 0.15078862855529346,
      "learning_rate": 2.3405280055014613e-05,
      "loss": 0.4293,
      "step": 200
    },
    {
      "epoch": 1.8071308253790006,
      "grad_norm": 0.16864941614937132,
      "learning_rate": 2.3112826787000755e-05,
      "loss": 0.433,
      "step": 201
    },
    {
      "epoch": 1.8161145423919147,
      "grad_norm": 0.1640970390638266,
      "learning_rate": 2.2820944425678543e-05,
      "loss": 0.4465,
      "step": 202
    },
    {
      "epoch": 1.8250982594048288,
      "grad_norm": 0.181930840073729,
      "learning_rate": 2.2529665193695787e-05,
      "loss": 0.4311,
      "step": 203
    },
    {
      "epoch": 1.834081976417743,
      "grad_norm": 0.1618511339670659,
      "learning_rate": 2.2239021247117203e-05,
      "loss": 0.425,
      "step": 204
    },
    {
      "epoch": 1.843065693430657,
      "grad_norm": 0.1608105810202212,
      "learning_rate": 2.1949044671874553e-05,
      "loss": 0.4372,
      "step": 205
    },
    {
      "epoch": 1.8520494104435712,
      "grad_norm": 0.17614884968053077,
      "learning_rate": 2.165976748022444e-05,
      "loss": 0.4242,
      "step": 206
    },
    {
      "epoch": 1.861033127456485,
      "grad_norm": 0.18688103883770055,
      "learning_rate": 2.1371221607214342e-05,
      "loss": 0.4329,
      "step": 207
    },
    {
      "epoch": 1.8700168444693992,
      "grad_norm": 0.1888649411522682,
      "learning_rate": 2.108343890715705e-05,
      "loss": 0.4421,
      "step": 208
    },
    {
      "epoch": 1.8790005614823133,
      "grad_norm": 0.19568546233660988,
      "learning_rate": 2.0796451150114122e-05,
      "loss": 0.4292,
      "step": 209
    },
    {
      "epoch": 1.8879842784952274,
      "grad_norm": 0.20130506786491167,
      "learning_rate": 2.0510290018388582e-05,
      "loss": 0.4305,
      "step": 210
    },
    {
      "epoch": 1.8969679955081415,
      "grad_norm": 0.16622134218898874,
      "learning_rate": 2.0224987103027312e-05,
      "loss": 0.4304,
      "step": 211
    },
    {
      "epoch": 1.9059517125210554,
      "grad_norm": 0.1714373409659067,
      "learning_rate": 1.9940573900333532e-05,
      "loss": 0.4388,
      "step": 212
    },
    {
      "epoch": 1.9149354295339696,
      "grad_norm": 0.16165865567699295,
      "learning_rate": 1.9657081808389732e-05,
      "loss": 0.4199,
      "step": 213
    },
    {
      "epoch": 1.9239191465468837,
      "grad_norm": 0.1588217221931874,
      "learning_rate": 1.9374542123591462e-05,
      "loss": 0.4391,
      "step": 214
    },
    {
      "epoch": 1.9329028635597978,
      "grad_norm": 0.16789101350762428,
      "learning_rate": 1.9092986037192315e-05,
      "loss": 0.429,
      "step": 215
    },
    {
      "epoch": 1.941886580572712,
      "grad_norm": 0.14933213528792,
      "learning_rate": 1.881244463186054e-05,
      "loss": 0.4327,
      "step": 216
    },
    {
      "epoch": 1.950870297585626,
      "grad_norm": 0.15634986800174527,
      "learning_rate": 1.8532948878247664e-05,
      "loss": 0.4222,
      "step": 217
    },
    {
      "epoch": 1.9598540145985401,
      "grad_norm": 0.16965600317101812,
      "learning_rate": 1.825452963156942e-05,
      "loss": 0.4422,
      "step": 218
    },
    {
      "epoch": 1.9688377316114543,
      "grad_norm": 0.13903361538978226,
      "learning_rate": 1.7977217628199486e-05,
      "loss": 0.4324,
      "step": 219
    },
    {
      "epoch": 1.9778214486243684,
      "grad_norm": 0.18809988345643205,
      "learning_rate": 1.770104348227628e-05,
      "loss": 0.4284,
      "step": 220
    },
    {
      "epoch": 1.9868051656372825,
      "grad_norm": 0.16274715282926117,
      "learning_rate": 1.742603768232333e-05,
      "loss": 0.417,
      "step": 221
    },
    {
      "epoch": 1.9957888826501966,
      "grad_norm": 0.18587407946726342,
      "learning_rate": 1.7152230587883387e-05,
      "loss": 0.4419,
      "step": 222
    },
    {
      "epoch": 2.0061763054463784,
      "grad_norm": 0.35271858704421305,
      "learning_rate": 1.6879652426166937e-05,
      "loss": 0.7668,
      "step": 223
    },
    {
      "epoch": 2.0151600224592925,
      "grad_norm": 0.18863183688272459,
      "learning_rate": 1.660833328871518e-05,
      "loss": 0.3965,
      "step": 224
    },
    {
      "epoch": 2.0241437394722066,
      "grad_norm": 0.2508766464598683,
      "learning_rate": 1.6338303128078067e-05,
      "loss": 0.4199,
      "step": 225
    },
    {
      "epoch": 2.0331274564851207,
      "grad_norm": 0.1846517351611734,
      "learning_rate": 1.6069591754507644e-05,
      "loss": 0.4059,
      "step": 226
    },
    {
      "epoch": 2.042111173498035,
      "grad_norm": 0.20660957505878264,
      "learning_rate": 1.5802228832667142e-05,
      "loss": 0.4115,
      "step": 227
    },
    {
      "epoch": 2.051094890510949,
      "grad_norm": 0.20768143407603465,
      "learning_rate": 1.553624387835609e-05,
      "loss": 0.4033,
      "step": 228
    },
    {
      "epoch": 2.060078607523863,
      "grad_norm": 0.23623799408278454,
      "learning_rate": 1.5271666255251907e-05,
      "loss": 0.419,
      "step": 229
    },
    {
      "epoch": 2.069062324536777,
      "grad_norm": 0.19328378884342695,
      "learning_rate": 1.5008525171668266e-05,
      "loss": 0.4045,
      "step": 230
    },
    {
      "epoch": 2.0780460415496913,
      "grad_norm": 0.20354843925554086,
      "learning_rate": 1.474684967733061e-05,
      "loss": 0.4063,
      "step": 231
    },
    {
      "epoch": 2.0870297585626054,
      "grad_norm": 0.1962670346347417,
      "learning_rate": 1.4486668660169169e-05,
      "loss": 0.4098,
      "step": 232
    },
    {
      "epoch": 2.0960134755755195,
      "grad_norm": 0.1697063402554285,
      "learning_rate": 1.4228010843129864e-05,
      "loss": 0.3947,
      "step": 233
    },
    {
      "epoch": 2.1049971925884337,
      "grad_norm": 0.21767004866396167,
      "learning_rate": 1.3970904781003401e-05,
      "loss": 0.4114,
      "step": 234
    },
    {
      "epoch": 2.1139809096013478,
      "grad_norm": 0.16628750228694827,
      "learning_rate": 1.3715378857272944e-05,
      "loss": 0.4089,
      "step": 235
    },
    {
      "epoch": 2.1229646266142614,
      "grad_norm": 0.17079039606673918,
      "learning_rate": 1.3461461280980681e-05,
      "loss": 0.4047,
      "step": 236
    },
    {
      "epoch": 2.1319483436271756,
      "grad_norm": 0.1984666308638168,
      "learning_rate": 1.3209180083613638e-05,
      "loss": 0.4126,
      "step": 237
    },
    {
      "epoch": 2.1409320606400897,
      "grad_norm": 0.15844397108776676,
      "learning_rate": 1.2958563116009172e-05,
      "loss": 0.401,
      "step": 238
    },
    {
      "epoch": 2.149915777653004,
      "grad_norm": 0.1682087305867551,
      "learning_rate": 1.27096380452803e-05,
      "loss": 0.404,
      "step": 239
    },
    {
      "epoch": 2.158899494665918,
      "grad_norm": 0.15876022024115735,
      "learning_rate": 1.2462432351761374e-05,
      "loss": 0.406,
      "step": 240
    },
    {
      "epoch": 2.167883211678832,
      "grad_norm": 0.13010884230934192,
      "learning_rate": 1.2216973325974375e-05,
      "loss": 0.3955,
      "step": 241
    },
    {
      "epoch": 2.176866928691746,
      "grad_norm": 0.15194484720161236,
      "learning_rate": 1.1973288065616129e-05,
      "loss": 0.4204,
      "step": 242
    },
    {
      "epoch": 2.1858506457046603,
      "grad_norm": 0.146927657510162,
      "learning_rate": 1.1731403472566841e-05,
      "loss": 0.4259,
      "step": 243
    },
    {
      "epoch": 2.1948343627175744,
      "grad_norm": 0.1357980790994208,
      "learning_rate": 1.1491346249920226e-05,
      "loss": 0.3998,
      "step": 244
    },
    {
      "epoch": 2.2038180797304885,
      "grad_norm": 0.14520203880695473,
      "learning_rate": 1.1253142899035609e-05,
      "loss": 0.4038,
      "step": 245
    },
    {
      "epoch": 2.2128017967434026,
      "grad_norm": 0.12180527212805993,
      "learning_rate": 1.101681971661226e-05,
      "loss": 0.3995,
      "step": 246
    },
    {
      "epoch": 2.2217855137563167,
      "grad_norm": 0.16678443391851008,
      "learning_rate": 1.0782402791786366e-05,
      "loss": 0.4093,
      "step": 247
    },
    {
      "epoch": 2.230769230769231,
      "grad_norm": 0.14169624933871403,
      "learning_rate": 1.054991800325088e-05,
      "loss": 0.3945,
      "step": 248
    },
    {
      "epoch": 2.239752947782145,
      "grad_norm": 0.12385040853111917,
      "learning_rate": 1.0319391016398607e-05,
      "loss": 0.4045,
      "step": 249
    },
    {
      "epoch": 2.248736664795059,
      "grad_norm": 0.13012596569465296,
      "learning_rate": 1.009084728048887e-05,
      "loss": 0.4037,
      "step": 250
    },
    {
      "epoch": 2.257720381807973,
      "grad_norm": 0.12678805353118078,
      "learning_rate": 9.864312025838009e-06,
      "loss": 0.4085,
      "step": 251
    },
    {
      "epoch": 2.2667040988208873,
      "grad_norm": 0.13122317737854305,
      "learning_rate": 9.63981026103404e-06,
      "loss": 0.415,
      "step": 252
    },
    {
      "epoch": 2.2756878158338014,
      "grad_norm": 0.1228277598641854,
      "learning_rate": 9.417366770175821e-06,
      "loss": 0.4017,
      "step": 253
    },
    {
      "epoch": 2.2846715328467155,
      "grad_norm": 0.1285711649465565,
      "learning_rate": 9.197006110136977e-06,
      "loss": 0.422,
      "step": 254
    },
    {
      "epoch": 2.293655249859629,
      "grad_norm": 0.13481029664375577,
      "learning_rate": 8.978752607854924e-06,
      "loss": 0.4046,
      "step": 255
    },
    {
      "epoch": 2.3026389668725433,
      "grad_norm": 0.11543534305483437,
      "learning_rate": 8.762630357645268e-06,
      "loss": 0.414,
      "step": 256
    },
    {
      "epoch": 2.3116226838854574,
      "grad_norm": 0.13900679103109587,
      "learning_rate": 8.548663218541897e-06,
      "loss": 0.4125,
      "step": 257
    },
    {
      "epoch": 2.3206064008983716,
      "grad_norm": 0.1259428788035887,
      "learning_rate": 8.336874811663032e-06,
      "loss": 0.3983,
      "step": 258
    },
    {
      "epoch": 2.3295901179112857,
      "grad_norm": 0.11085457797406689,
      "learning_rate": 8.127288517603557e-06,
      "loss": 0.4193,
      "step": 259
    },
    {
      "epoch": 2.3385738349242,
      "grad_norm": 0.11622214274255102,
      "learning_rate": 7.919927473853877e-06,
      "loss": 0.3937,
      "step": 260
    },
    {
      "epoch": 2.347557551937114,
      "grad_norm": 0.12462100200017505,
      "learning_rate": 7.714814572245652e-06,
      "loss": 0.4155,
      "step": 261
    },
    {
      "epoch": 2.356541268950028,
      "grad_norm": 0.12614101513377815,
      "learning_rate": 7.511972456424624e-06,
      "loss": 0.4121,
      "step": 262
    },
    {
      "epoch": 2.365524985962942,
      "grad_norm": 0.12461680064110435,
      "learning_rate": 7.31142351935084e-06,
      "loss": 0.4068,
      "step": 263
    },
    {
      "epoch": 2.3745087029758563,
      "grad_norm": 0.11950703045498604,
      "learning_rate": 7.113189900826568e-06,
      "loss": 0.4117,
      "step": 264
    },
    {
      "epoch": 2.3834924199887704,
      "grad_norm": 0.13056991708575583,
      "learning_rate": 6.917293485052153e-06,
      "loss": 0.4128,
      "step": 265
    },
    {
      "epoch": 2.3924761370016845,
      "grad_norm": 0.13745171444337193,
      "learning_rate": 6.723755898210081e-06,
      "loss": 0.4151,
      "step": 266
    },
    {
      "epoch": 2.4014598540145986,
      "grad_norm": 0.13033817774915013,
      "learning_rate": 6.53259850607753e-06,
      "loss": 0.4016,
      "step": 267
    },
    {
      "epoch": 2.4104435710275127,
      "grad_norm": 0.10688098668609738,
      "learning_rate": 6.343842411667697e-06,
      "loss": 0.4134,
      "step": 268
    },
    {
      "epoch": 2.419427288040427,
      "grad_norm": 0.1287888983585153,
      "learning_rate": 6.157508452900079e-06,
      "loss": 0.4174,
      "step": 269
    },
    {
      "epoch": 2.428411005053341,
      "grad_norm": 0.1248679388883586,
      "learning_rate": 5.973617200300082e-06,
      "loss": 0.3956,
      "step": 270
    },
    {
      "epoch": 2.437394722066255,
      "grad_norm": 0.11070702147294058,
      "learning_rate": 5.792188954728074e-06,
      "loss": 0.3993,
      "step": 271
    },
    {
      "epoch": 2.446378439079169,
      "grad_norm": 0.12330276650877879,
      "learning_rate": 5.6132437451382956e-06,
      "loss": 0.4085,
      "step": 272
    },
    {
      "epoch": 2.4553621560920833,
      "grad_norm": 0.1119180851449193,
      "learning_rate": 5.436801326367725e-06,
      "loss": 0.3998,
      "step": 273
    },
    {
      "epoch": 2.464345873104997,
      "grad_norm": 0.1328994320544872,
      "learning_rate": 5.262881176955216e-06,
      "loss": 0.419,
      "step": 274
    },
    {
      "epoch": 2.473329590117911,
      "grad_norm": 0.11619373835235447,
      "learning_rate": 5.09150249699116e-06,
      "loss": 0.4097,
      "step": 275
    },
    {
      "epoch": 2.482313307130825,
      "grad_norm": 0.11879193778970075,
      "learning_rate": 4.92268420599785e-06,
      "loss": 0.4154,
      "step": 276
    },
    {
      "epoch": 2.4912970241437393,
      "grad_norm": 0.11150882823633602,
      "learning_rate": 4.756444940840868e-06,
      "loss": 0.3938,
      "step": 277
    },
    {
      "epoch": 2.5002807411566534,
      "grad_norm": 0.10688470139111161,
      "learning_rate": 4.5928030536716305e-06,
      "loss": 0.3951,
      "step": 278
    },
    {
      "epoch": 2.5092644581695676,
      "grad_norm": 0.10966725943546395,
      "learning_rate": 4.431776609901392e-06,
      "loss": 0.4067,
      "step": 279
    },
    {
      "epoch": 2.5182481751824817,
      "grad_norm": 0.10615758090408767,
      "learning_rate": 4.273383386206893e-06,
      "loss": 0.4171,
      "step": 280
    },
    {
      "epoch": 2.527231892195396,
      "grad_norm": 0.10796581041782187,
      "learning_rate": 4.11764086856789e-06,
      "loss": 0.3986,
      "step": 281
    },
    {
      "epoch": 2.53621560920831,
      "grad_norm": 0.10704980127868732,
      "learning_rate": 3.964566250336768e-06,
      "loss": 0.4025,
      "step": 282
    },
    {
      "epoch": 2.545199326221224,
      "grad_norm": 0.11367865379296356,
      "learning_rate": 3.814176430340453e-06,
      "loss": 0.4189,
      "step": 283
    },
    {
      "epoch": 2.554183043234138,
      "grad_norm": 0.10114216928540735,
      "learning_rate": 3.6664880110148826e-06,
      "loss": 0.3926,
      "step": 284
    },
    {
      "epoch": 2.5631667602470523,
      "grad_norm": 0.11361742890474415,
      "learning_rate": 3.5215172965721247e-06,
      "loss": 0.4245,
      "step": 285
    },
    {
      "epoch": 2.5721504772599664,
      "grad_norm": 0.1084518684866817,
      "learning_rate": 3.3792802912004827e-06,
      "loss": 0.4107,
      "step": 286
    },
    {
      "epoch": 2.5811341942728805,
      "grad_norm": 0.10549395673412279,
      "learning_rate": 3.2397926972976876e-06,
      "loss": 0.4107,
      "step": 287
    },
    {
      "epoch": 2.5901179112857946,
      "grad_norm": 0.09969645074776698,
      "learning_rate": 3.1030699137374146e-06,
      "loss": 0.3994,
      "step": 288
    },
    {
      "epoch": 2.5991016282987087,
      "grad_norm": 0.1056666118347554,
      "learning_rate": 2.969127034169312e-06,
      "loss": 0.3998,
      "step": 289
    },
    {
      "epoch": 2.608085345311623,
      "grad_norm": 0.1035766308650711,
      "learning_rate": 2.837978845352723e-06,
      "loss": 0.417,
      "step": 290
    },
    {
      "epoch": 2.6170690623245365,
      "grad_norm": 0.10302107528120776,
      "learning_rate": 2.709639825524283e-06,
      "loss": 0.4109,
      "step": 291
    },
    {
      "epoch": 2.626052779337451,
      "grad_norm": 0.10759853807512873,
      "learning_rate": 2.5841241427995886e-06,
      "loss": 0.3955,
      "step": 292
    },
    {
      "epoch": 2.6350364963503647,
      "grad_norm": 0.10880577647272674,
      "learning_rate": 2.461445653609079e-06,
      "loss": 0.4147,
      "step": 293
    },
    {
      "epoch": 2.6440202133632793,
      "grad_norm": 0.09689142380499235,
      "learning_rate": 2.3416179011683658e-06,
      "loss": 0.4106,
      "step": 294
    },
    {
      "epoch": 2.653003930376193,
      "grad_norm": 0.10685593466563702,
      "learning_rate": 2.2246541139830715e-06,
      "loss": 0.4014,
      "step": 295
    },
    {
      "epoch": 2.661987647389107,
      "grad_norm": 0.0983605426363825,
      "learning_rate": 2.1105672043885143e-06,
      "loss": 0.4116,
      "step": 296
    },
    {
      "epoch": 2.670971364402021,
      "grad_norm": 0.09655008561783156,
      "learning_rate": 1.9993697671241945e-06,
      "loss": 0.4081,
      "step": 297
    },
    {
      "epoch": 2.6799550814149353,
      "grad_norm": 0.11364035199790146,
      "learning_rate": 1.8910740779434006e-06,
      "loss": 0.4076,
      "step": 298
    },
    {
      "epoch": 2.6889387984278494,
      "grad_norm": 0.10418804191385413,
      "learning_rate": 1.7856920922580113e-06,
      "loss": 0.4011,
      "step": 299
    },
    {
      "epoch": 2.6979225154407636,
      "grad_norm": 0.09941508460951572,
      "learning_rate": 1.683235443818673e-06,
      "loss": 0.3984,
      "step": 300
    },
    {
      "epoch": 2.7069062324536777,
      "grad_norm": 0.09866363300713496,
      "learning_rate": 1.5837154434304704e-06,
      "loss": 0.4088,
      "step": 301
    },
    {
      "epoch": 2.715889949466592,
      "grad_norm": 0.10476593966969794,
      "learning_rate": 1.4871430777042698e-06,
      "loss": 0.4158,
      "step": 302
    },
    {
      "epoch": 2.724873666479506,
      "grad_norm": 0.1003488280191521,
      "learning_rate": 1.3935290078438229e-06,
      "loss": 0.4023,
      "step": 303
    },
    {
      "epoch": 2.73385738349242,
      "grad_norm": 0.09404490189909534,
      "learning_rate": 1.3028835684688378e-06,
      "loss": 0.3971,
      "step": 304
    },
    {
      "epoch": 2.742841100505334,
      "grad_norm": 0.10211715235272904,
      "learning_rate": 1.2152167664740558e-06,
      "loss": 0.4075,
      "step": 305
    },
    {
      "epoch": 2.7518248175182483,
      "grad_norm": 0.1049198084000181,
      "learning_rate": 1.1305382799245437e-06,
      "loss": 0.3984,
      "step": 306
    },
    {
      "epoch": 2.7608085345311624,
      "grad_norm": 0.09652931457838564,
      "learning_rate": 1.0488574569872655e-06,
      "loss": 0.4098,
      "step": 307
    },
    {
      "epoch": 2.7697922515440765,
      "grad_norm": 0.08905719728483828,
      "learning_rate": 9.70183314899084e-07,
      "loss": 0.3915,
      "step": 308
    },
    {
      "epoch": 2.7787759685569906,
      "grad_norm": 0.0938697467560363,
      "learning_rate": 8.945245389713029e-07,
      "loss": 0.4248,
      "step": 309
    },
    {
      "epoch": 2.7877596855699043,
      "grad_norm": 0.1012541455507019,
      "learning_rate": 8.218894816308398e-07,
      "loss": 0.3914,
      "step": 310
    },
    {
      "epoch": 2.796743402582819,
      "grad_norm": 0.09264156371198151,
      "learning_rate": 7.522861614981325e-07,
      "loss": 0.4122,
      "step": 311
    },
    {
      "epoch": 2.8057271195957325,
      "grad_norm": 0.0957623449129902,
      "learning_rate": 6.85722262501941e-07,
      "loss": 0.4102,
      "step": 312
    },
    {
      "epoch": 2.814710836608647,
      "grad_norm": 0.09713743297082324,
      "learning_rate": 6.222051330310619e-07,
      "loss": 0.4026,
      "step": 313
    },
    {
      "epoch": 2.8236945536215607,
      "grad_norm": 0.09186827414249998,
      "learning_rate": 5.617417851230931e-07,
      "loss": 0.4096,
      "step": 314
    },
    {
      "epoch": 2.832678270634475,
      "grad_norm": 0.09713737535786383,
      "learning_rate": 5.043388936903404e-07,
      "loss": 0.4032,
      "step": 315
    },
    {
      "epoch": 2.841661987647389,
      "grad_norm": 0.09342384537711217,
      "learning_rate": 4.500027957829165e-07,
      "loss": 0.4071,
      "step": 316
    },
    {
      "epoch": 2.850645704660303,
      "grad_norm": 0.09314756742095372,
      "learning_rate": 3.987394898891804e-07,
      "loss": 0.4137,
      "step": 317
    },
    {
      "epoch": 2.859629421673217,
      "grad_norm": 0.0887933631925566,
      "learning_rate": 3.50554635273513e-07,
      "loss": 0.4081,
      "step": 318
    },
    {
      "epoch": 2.8686131386861313,
      "grad_norm": 0.09565485915542679,
      "learning_rate": 3.0545355135156165e-07,
      "loss": 0.4025,
      "step": 319
    },
    {
      "epoch": 2.8775968556990454,
      "grad_norm": 0.0950394236491987,
      "learning_rate": 2.634412171029911e-07,
      "loss": 0.4118,
      "step": 320
    },
    {
      "epoch": 2.8865805727119596,
      "grad_norm": 0.08929953352976103,
      "learning_rate": 2.2452227052183503e-07,
      "loss": 0.4051,
      "step": 321
    },
    {
      "epoch": 2.8955642897248737,
      "grad_norm": 0.087031212892392,
      "learning_rate": 1.8870100810447e-07,
      "loss": 0.4037,
      "step": 322
    },
    {
      "epoch": 2.904548006737788,
      "grad_norm": 0.09107399363880533,
      "learning_rate": 1.559813843753062e-07,
      "loss": 0.4089,
      "step": 323
    },
    {
      "epoch": 2.913531723750702,
      "grad_norm": 0.09057801956608469,
      "learning_rate": 1.2636701145022035e-07,
      "loss": 0.4016,
      "step": 324
    },
    {
      "epoch": 2.922515440763616,
      "grad_norm": 0.09636376248734924,
      "learning_rate": 9.986115863779269e-08,
      "loss": 0.4139,
      "step": 325
    },
    {
      "epoch": 2.93149915777653,
      "grad_norm": 0.08743246679042953,
      "learning_rate": 7.646675207839336e-08,
      "loss": 0.3973,
      "step": 326
    },
    {
      "epoch": 2.9404828747894443,
      "grad_norm": 0.08981696851089446,
      "learning_rate": 5.618637442113561e-08,
      "loss": 0.3952,
      "step": 327
    },
    {
      "epoch": 2.9494665918023584,
      "grad_norm": 0.09021046410712788,
      "learning_rate": 3.9022264538775445e-08,
      "loss": 0.4095,
      "step": 328
    },
    {
      "epoch": 2.958450308815272,
      "grad_norm": 0.09125597188841794,
      "learning_rate": 2.4976317280541322e-08,
      "loss": 0.404,
      "step": 329
    },
    {
      "epoch": 2.9674340258281866,
      "grad_norm": 0.09031347282702284,
      "learning_rate": 1.4050083262957056e-08,
      "loss": 0.4065,
      "step": 330
    },
    {
      "epoch": 2.9764177428411003,
      "grad_norm": 0.09548041676594968,
      "learning_rate": 6.24476869865149e-09,
      "loss": 0.4085,
      "step": 331
    },
    {
      "epoch": 2.985401459854015,
      "grad_norm": 0.08990746175705872,
      "learning_rate": 1.5612352631989697e-09,
      "loss": 0.4029,
      "step": 332
    },
    {
      "epoch": 2.9943851768669285,
      "grad_norm": 0.09349234362985716,
      "learning_rate": 0.0,
      "loss": 0.407,
      "step": 333
    },
    {
      "epoch": 2.9943851768669285,
      "step": 333,
      "total_flos": 4607052149424128.0,
      "train_loss": 0.4632815238949773,
      "train_runtime": 115748.1353,
      "train_samples_per_second": 2.954,
      "train_steps_per_second": 0.003
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 333,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4607052149424128.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}