{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.40210627094303497,
  "eval_steps": 500,
  "global_step": 210,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0019147917663954045,
      "grad_norm": 1.17178213596344,
      "learning_rate": 0.0,
      "loss": 0.5878,
      "step": 1
    },
    {
      "epoch": 0.003829583532790809,
      "grad_norm": 1.167578101158142,
      "learning_rate": 3.846153846153847e-07,
      "loss": 0.5815,
      "step": 2
    },
    {
      "epoch": 0.0057443752991862135,
      "grad_norm": 1.2005791664123535,
      "learning_rate": 7.692307692307694e-07,
      "loss": 0.6327,
      "step": 3
    },
    {
      "epoch": 0.007659167065581618,
      "grad_norm": 1.1366013288497925,
      "learning_rate": 1.153846153846154e-06,
      "loss": 0.5869,
      "step": 4
    },
    {
      "epoch": 0.009573958831977022,
      "grad_norm": 1.0155231952667236,
      "learning_rate": 1.5384615384615387e-06,
      "loss": 0.5233,
      "step": 5
    },
    {
      "epoch": 0.011488750598372427,
      "grad_norm": 1.2220700979232788,
      "learning_rate": 1.9230769230769234e-06,
      "loss": 0.5699,
      "step": 6
    },
    {
      "epoch": 0.013403542364767831,
      "grad_norm": 1.1095085144042969,
      "learning_rate": 2.307692307692308e-06,
      "loss": 0.5378,
      "step": 7
    },
    {
      "epoch": 0.015318334131163236,
      "grad_norm": 1.0518015623092651,
      "learning_rate": 2.6923076923076923e-06,
      "loss": 0.5432,
      "step": 8
    },
    {
      "epoch": 0.01723312589755864,
      "grad_norm": 1.0267938375473022,
      "learning_rate": 3.0769230769230774e-06,
      "loss": 0.5702,
      "step": 9
    },
    {
      "epoch": 0.019147917663954045,
      "grad_norm": 1.0278018712997437,
      "learning_rate": 3.4615384615384617e-06,
      "loss": 0.5411,
      "step": 10
    },
    {
      "epoch": 0.02106270943034945,
      "grad_norm": 0.9818478226661682,
      "learning_rate": 3.846153846153847e-06,
      "loss": 0.5214,
      "step": 11
    },
    {
      "epoch": 0.022977501196744854,
      "grad_norm": 0.9112760424613953,
      "learning_rate": 4.230769230769231e-06,
      "loss": 0.5235,
      "step": 12
    },
    {
      "epoch": 0.02489229296314026,
      "grad_norm": 1.1367770433425903,
      "learning_rate": 4.615384615384616e-06,
      "loss": 0.5689,
      "step": 13
    },
    {
      "epoch": 0.026807084729535663,
      "grad_norm": 0.887293815612793,
      "learning_rate": 5e-06,
      "loss": 0.5062,
      "step": 14
    },
    {
      "epoch": 0.028721876495931067,
      "grad_norm": 0.8234561085700989,
      "learning_rate": 5.384615384615385e-06,
      "loss": 0.4938,
      "step": 15
    },
    {
      "epoch": 0.030636668262326472,
      "grad_norm": 0.8301621079444885,
      "learning_rate": 5.769230769230769e-06,
      "loss": 0.4955,
      "step": 16
    },
    {
      "epoch": 0.032551460028721876,
      "grad_norm": 0.7201998829841614,
      "learning_rate": 6.153846153846155e-06,
      "loss": 0.4481,
      "step": 17
    },
    {
      "epoch": 0.03446625179511728,
      "grad_norm": 0.7492491602897644,
      "learning_rate": 6.538461538461539e-06,
      "loss": 0.4924,
      "step": 18
    },
    {
      "epoch": 0.036381043561512685,
      "grad_norm": 0.8051350712776184,
      "learning_rate": 6.923076923076923e-06,
      "loss": 0.5109,
      "step": 19
    },
    {
      "epoch": 0.03829583532790809,
      "grad_norm": 0.6143934726715088,
      "learning_rate": 7.307692307692308e-06,
      "loss": 0.4846,
      "step": 20
    },
    {
      "epoch": 0.040210627094303494,
      "grad_norm": 0.7769045233726501,
      "learning_rate": 7.692307692307694e-06,
      "loss": 0.4761,
      "step": 21
    },
    {
      "epoch": 0.0421254188606989,
      "grad_norm": 0.5961856842041016,
      "learning_rate": 8.076923076923077e-06,
      "loss": 0.467,
      "step": 22
    },
    {
      "epoch": 0.0440402106270943,
      "grad_norm": 0.683874785900116,
      "learning_rate": 8.461538461538462e-06,
      "loss": 0.4847,
      "step": 23
    },
    {
      "epoch": 0.04595500239348971,
      "grad_norm": 0.5849653482437134,
      "learning_rate": 8.846153846153847e-06,
      "loss": 0.4514,
      "step": 24
    },
    {
      "epoch": 0.04786979415988511,
      "grad_norm": 0.6217619180679321,
      "learning_rate": 9.230769230769232e-06,
      "loss": 0.4733,
      "step": 25
    },
    {
      "epoch": 0.04978458592628052,
      "grad_norm": 0.5796915888786316,
      "learning_rate": 9.615384615384616e-06,
      "loss": 0.4735,
      "step": 26
    },
    {
      "epoch": 0.05169937769267592,
      "grad_norm": 0.5245983600616455,
      "learning_rate": 1e-05,
      "loss": 0.4398,
      "step": 27
    },
    {
      "epoch": 0.053614169459071326,
      "grad_norm": 0.6732852458953857,
      "learning_rate": 9.999899706000774e-06,
      "loss": 0.4907,
      "step": 28
    },
    {
      "epoch": 0.05552896122546673,
      "grad_norm": 0.5512049794197083,
      "learning_rate": 9.999598828026644e-06,
      "loss": 0.4723,
      "step": 29
    },
    {
      "epoch": 0.057443752991862135,
      "grad_norm": 0.6110374331474304,
      "learning_rate": 9.999097378148116e-06,
      "loss": 0.4414,
      "step": 30
    },
    {
      "epoch": 0.05935854475825754,
      "grad_norm": 0.5515976548194885,
      "learning_rate": 9.998395376482152e-06,
      "loss": 0.4314,
      "step": 31
    },
    {
      "epoch": 0.061273336524652944,
      "grad_norm": 0.5074656009674072,
      "learning_rate": 9.99749285119138e-06,
      "loss": 0.4149,
      "step": 32
    },
    {
      "epoch": 0.06318812829104835,
      "grad_norm": 0.6015526056289673,
      "learning_rate": 9.996389838482942e-06,
      "loss": 0.517,
      "step": 33
    },
    {
      "epoch": 0.06510292005744375,
      "grad_norm": 0.5998480916023254,
      "learning_rate": 9.995086382607064e-06,
      "loss": 0.4695,
      "step": 34
    },
    {
      "epoch": 0.06701771182383916,
      "grad_norm": 0.5995810627937317,
      "learning_rate": 9.993582535855265e-06,
      "loss": 0.4453,
      "step": 35
    },
    {
      "epoch": 0.06893250359023456,
      "grad_norm": 0.5346629023551941,
      "learning_rate": 9.991878358558267e-06,
      "loss": 0.4667,
      "step": 36
    },
    {
      "epoch": 0.07084729535662997,
      "grad_norm": 0.5550440549850464,
      "learning_rate": 9.989973919083576e-06,
      "loss": 0.4488,
      "step": 37
    },
    {
      "epoch": 0.07276208712302537,
      "grad_norm": 0.5914222598075867,
      "learning_rate": 9.987869293832727e-06,
      "loss": 0.4473,
      "step": 38
    },
    {
      "epoch": 0.07467687888942078,
      "grad_norm": 0.46610352396965027,
      "learning_rate": 9.985564567238237e-06,
      "loss": 0.4265,
      "step": 39
    },
    {
      "epoch": 0.07659167065581618,
      "grad_norm": 0.6092692017555237,
      "learning_rate": 9.983059831760205e-06,
      "loss": 0.469,
      "step": 40
    },
    {
      "epoch": 0.07850646242221158,
      "grad_norm": 0.48001110553741455,
      "learning_rate": 9.980355187882606e-06,
      "loss": 0.4219,
      "step": 41
    },
    {
      "epoch": 0.08042125418860699,
      "grad_norm": 0.4854871332645416,
      "learning_rate": 9.977450744109258e-06,
      "loss": 0.3991,
      "step": 42
    },
    {
      "epoch": 0.0823360459550024,
      "grad_norm": 0.47407639026641846,
      "learning_rate": 9.974346616959476e-06,
      "loss": 0.4177,
      "step": 43
    },
    {
      "epoch": 0.0842508377213978,
      "grad_norm": 0.599115788936615,
      "learning_rate": 9.97104293096339e-06,
      "loss": 0.4401,
      "step": 44
    },
    {
      "epoch": 0.0861656294877932,
      "grad_norm": 0.5324896574020386,
      "learning_rate": 9.967539818656953e-06,
      "loss": 0.4265,
      "step": 45
    },
    {
      "epoch": 0.0880804212541886,
      "grad_norm": 0.7636860609054565,
      "learning_rate": 9.96383742057662e-06,
      "loss": 0.4957,
      "step": 46
    },
    {
      "epoch": 0.08999521302058401,
      "grad_norm": 0.4660010039806366,
      "learning_rate": 9.959935885253715e-06,
      "loss": 0.4236,
      "step": 47
    },
    {
      "epoch": 0.09191000478697942,
      "grad_norm": 0.4723781645298004,
      "learning_rate": 9.955835369208475e-06,
      "loss": 0.405,
      "step": 48
    },
    {
      "epoch": 0.09382479655337482,
      "grad_norm": 0.473716139793396,
      "learning_rate": 9.951536036943753e-06,
      "loss": 0.4054,
      "step": 49
    },
    {
      "epoch": 0.09573958831977022,
      "grad_norm": 0.8201298713684082,
      "learning_rate": 9.94703806093845e-06,
      "loss": 0.4437,
      "step": 50
    },
    {
      "epoch": 0.09765438008616563,
      "grad_norm": 0.5334436297416687,
      "learning_rate": 9.942341621640558e-06,
      "loss": 0.4112,
      "step": 51
    },
    {
      "epoch": 0.09956917185256103,
      "grad_norm": 0.4933352768421173,
      "learning_rate": 9.937446907459954e-06,
      "loss": 0.4303,
      "step": 52
    },
    {
      "epoch": 0.10148396361895644,
      "grad_norm": 0.44361305236816406,
      "learning_rate": 9.932354114760819e-06,
      "loss": 0.4039,
      "step": 53
    },
    {
      "epoch": 0.10339875538535184,
      "grad_norm": 0.5590924024581909,
      "learning_rate": 9.92706344785377e-06,
      "loss": 0.51,
      "step": 54
    },
    {
      "epoch": 0.10531354715174725,
      "grad_norm": 0.3887651264667511,
      "learning_rate": 9.921575118987672e-06,
      "loss": 0.3709,
      "step": 55
    },
    {
      "epoch": 0.10722833891814265,
      "grad_norm": 0.43217676877975464,
      "learning_rate": 9.915889348341098e-06,
      "loss": 0.4107,
      "step": 56
    },
    {
      "epoch": 0.10914313068453806,
      "grad_norm": 0.46679773926734924,
      "learning_rate": 9.910006364013522e-06,
      "loss": 0.3832,
      "step": 57
    },
    {
      "epoch": 0.11105792245093346,
      "grad_norm": 0.5174672603607178,
      "learning_rate": 9.903926402016153e-06,
      "loss": 0.397,
      "step": 58
    },
    {
      "epoch": 0.11297271421732887,
      "grad_norm": 0.5932629704475403,
      "learning_rate": 9.897649706262474e-06,
      "loss": 0.4584,
      "step": 59
    },
    {
      "epoch": 0.11488750598372427,
      "grad_norm": 0.5002132058143616,
      "learning_rate": 9.891176528558451e-06,
      "loss": 0.4115,
      "step": 60
    },
    {
      "epoch": 0.11680229775011967,
      "grad_norm": 0.44753795862197876,
      "learning_rate": 9.884507128592435e-06,
      "loss": 0.417,
      "step": 61
    },
    {
      "epoch": 0.11871708951651508,
      "grad_norm": 0.43308743834495544,
      "learning_rate": 9.877641773924748e-06,
      "loss": 0.3872,
      "step": 62
    },
    {
      "epoch": 0.12063188128291048,
      "grad_norm": 0.44310009479522705,
      "learning_rate": 9.870580739976936e-06,
      "loss": 0.3877,
      "step": 63
    },
    {
      "epoch": 0.12254667304930589,
      "grad_norm": 0.4362034797668457,
      "learning_rate": 9.863324310020735e-06,
      "loss": 0.3957,
      "step": 64
    },
    {
      "epoch": 0.12446146481570129,
      "grad_norm": 0.46891993284225464,
      "learning_rate": 9.855872775166696e-06,
      "loss": 0.427,
      "step": 65
    },
    {
      "epoch": 0.1263762565820967,
      "grad_norm": 0.47662439942359924,
      "learning_rate": 9.848226434352513e-06,
      "loss": 0.4041,
      "step": 66
    },
    {
      "epoch": 0.12829104834849211,
      "grad_norm": 0.5884003043174744,
      "learning_rate": 9.840385594331022e-06,
      "loss": 0.4487,
      "step": 67
    },
    {
      "epoch": 0.1302058401148875,
      "grad_norm": 0.5149067640304565,
      "learning_rate": 9.83235056965791e-06,
      "loss": 0.4567,
      "step": 68
    },
    {
      "epoch": 0.13212063188128292,
      "grad_norm": 0.4653552174568176,
      "learning_rate": 9.824121682679072e-06,
      "loss": 0.4024,
      "step": 69
    },
    {
      "epoch": 0.13403542364767831,
      "grad_norm": 0.47937148809432983,
      "learning_rate": 9.815699263517712e-06,
      "loss": 0.414,
      "step": 70
    },
    {
      "epoch": 0.13595021541407373,
      "grad_norm": 0.4588964581489563,
      "learning_rate": 9.807083650061063e-06,
      "loss": 0.4199,
      "step": 71
    },
    {
      "epoch": 0.13786500718046912,
      "grad_norm": 0.5369502902030945,
      "learning_rate": 9.798275187946859e-06,
      "loss": 0.4213,
      "step": 72
    },
    {
      "epoch": 0.13977979894686454,
      "grad_norm": 0.48553192615509033,
      "learning_rate": 9.789274230549456e-06,
      "loss": 0.3694,
      "step": 73
    },
    {
      "epoch": 0.14169459071325993,
      "grad_norm": 0.3765530586242676,
      "learning_rate": 9.780081138965663e-06,
      "loss": 0.3445,
      "step": 74
    },
    {
      "epoch": 0.14360938247965535,
      "grad_norm": 0.48994067311286926,
      "learning_rate": 9.770696282000245e-06,
      "loss": 0.4202,
      "step": 75
    },
    {
      "epoch": 0.14552417424605074,
      "grad_norm": 0.45693519711494446,
      "learning_rate": 9.761120036151138e-06,
      "loss": 0.384,
      "step": 76
    },
    {
      "epoch": 0.14743896601244616,
      "grad_norm": 0.4083925485610962,
      "learning_rate": 9.751352785594337e-06,
      "loss": 0.3618,
      "step": 77
    },
    {
      "epoch": 0.14935375777884155,
      "grad_norm": 0.5029312372207642,
      "learning_rate": 9.741394922168495e-06,
      "loss": 0.4446,
      "step": 78
    },
    {
      "epoch": 0.15126854954523697,
      "grad_norm": 0.4641478359699249,
      "learning_rate": 9.731246845359187e-06,
      "loss": 0.3837,
      "step": 79
    },
    {
      "epoch": 0.15318334131163236,
      "grad_norm": 0.42773452401161194,
      "learning_rate": 9.720908962282893e-06,
      "loss": 0.3747,
      "step": 80
    },
    {
      "epoch": 0.15509813307802778,
      "grad_norm": 0.45671483874320984,
      "learning_rate": 9.710381687670675e-06,
      "loss": 0.396,
      "step": 81
    },
    {
      "epoch": 0.15701292484442317,
      "grad_norm": 0.518532931804657,
      "learning_rate": 9.699665443851518e-06,
      "loss": 0.4207,
      "step": 82
    },
    {
      "epoch": 0.1589277166108186,
      "grad_norm": 0.42763078212738037,
      "learning_rate": 9.688760660735403e-06,
      "loss": 0.3617,
      "step": 83
    },
    {
      "epoch": 0.16084250837721398,
      "grad_norm": 0.4328898787498474,
      "learning_rate": 9.677667775796052e-06,
      "loss": 0.3612,
      "step": 84
    },
    {
      "epoch": 0.1627573001436094,
      "grad_norm": 0.5515812635421753,
      "learning_rate": 9.666387234053385e-06,
      "loss": 0.4151,
      "step": 85
    },
    {
      "epoch": 0.1646720919100048,
      "grad_norm": 0.4455530047416687,
      "learning_rate": 9.654919488055656e-06,
      "loss": 0.4048,
      "step": 86
    },
    {
      "epoch": 0.1665868836764002,
      "grad_norm": 0.43045058846473694,
      "learning_rate": 9.643264997861312e-06,
      "loss": 0.3769,
      "step": 87
    },
    {
      "epoch": 0.1685016754427956,
      "grad_norm": 0.44380709528923035,
      "learning_rate": 9.631424231020523e-06,
      "loss": 0.3963,
      "step": 88
    },
    {
      "epoch": 0.170416467209191,
      "grad_norm": 0.5282372832298279,
      "learning_rate": 9.619397662556434e-06,
      "loss": 0.418,
      "step": 89
    },
    {
      "epoch": 0.1723312589755864,
      "grad_norm": 0.5417309403419495,
      "learning_rate": 9.607185774946106e-06,
      "loss": 0.483,
      "step": 90
    },
    {
      "epoch": 0.17424605074198182,
      "grad_norm": 0.48023760318756104,
      "learning_rate": 9.594789058101154e-06,
      "loss": 0.4041,
      "step": 91
    },
    {
      "epoch": 0.1761608425083772,
      "grad_norm": 0.43183737993240356,
      "learning_rate": 9.582208009348104e-06,
      "loss": 0.3724,
      "step": 92
    },
    {
      "epoch": 0.17807563427477263,
      "grad_norm": 0.4684963822364807,
      "learning_rate": 9.569443133408434e-06,
      "loss": 0.4199,
      "step": 93
    },
    {
      "epoch": 0.17999042604116802,
      "grad_norm": 0.5076436400413513,
      "learning_rate": 9.556494942378328e-06,
      "loss": 0.4138,
      "step": 94
    },
    {
      "epoch": 0.18190521780756344,
      "grad_norm": 0.527299165725708,
      "learning_rate": 9.543363955708124e-06,
      "loss": 0.4089,
      "step": 95
    },
    {
      "epoch": 0.18382000957395883,
      "grad_norm": 0.4080921411514282,
      "learning_rate": 9.530050700181499e-06,
      "loss": 0.3337,
      "step": 96
    },
    {
      "epoch": 0.18573480134035425,
      "grad_norm": 0.42223554849624634,
      "learning_rate": 9.5165557098943e-06,
      "loss": 0.3758,
      "step": 97
    },
    {
      "epoch": 0.18764959310674964,
      "grad_norm": 0.4336399435997009,
      "learning_rate": 9.502879526233151e-06,
      "loss": 0.3664,
      "step": 98
    },
    {
      "epoch": 0.18956438487314506,
      "grad_norm": 0.43095213174819946,
      "learning_rate": 9.48902269785371e-06,
      "loss": 0.379,
      "step": 99
    },
    {
      "epoch": 0.19147917663954045,
      "grad_norm": 0.4262320399284363,
      "learning_rate": 9.47498578065867e-06,
      "loss": 0.3666,
      "step": 100
    },
    {
      "epoch": 0.19339396840593587,
      "grad_norm": 0.4407063126564026,
      "learning_rate": 9.460769337775461e-06,
      "loss": 0.3857,
      "step": 101
    },
    {
      "epoch": 0.19530876017233126,
      "grad_norm": 0.39989781379699707,
      "learning_rate": 9.446373939533642e-06,
      "loss": 0.3568,
      "step": 102
    },
    {
      "epoch": 0.19722355193872668,
      "grad_norm": 0.43383434414863586,
      "learning_rate": 9.431800163442043e-06,
      "loss": 0.374,
      "step": 103
    },
    {
      "epoch": 0.19913834370512207,
      "grad_norm": 0.5074713230133057,
      "learning_rate": 9.417048594165572e-06,
      "loss": 0.4137,
      "step": 104
    },
    {
      "epoch": 0.20105313547151749,
      "grad_norm": 0.45141083002090454,
      "learning_rate": 9.402119823501787e-06,
      "loss": 0.3977,
      "step": 105
    },
    {
      "epoch": 0.20296792723791288,
      "grad_norm": 0.44136589765548706,
      "learning_rate": 9.387014450357128e-06,
      "loss": 0.3808,
      "step": 106
    },
    {
      "epoch": 0.2048827190043083,
      "grad_norm": 0.6126969456672668,
      "learning_rate": 9.371733080722911e-06,
      "loss": 0.4527,
      "step": 107
    },
    {
      "epoch": 0.20679751077070369,
      "grad_norm": 0.48484617471694946,
      "learning_rate": 9.356276327651006e-06,
      "loss": 0.4032,
      "step": 108
    },
    {
      "epoch": 0.2087123025370991,
      "grad_norm": 0.4131053388118744,
      "learning_rate": 9.340644811229243e-06,
      "loss": 0.3594,
      "step": 109
    },
    {
      "epoch": 0.2106270943034945,
      "grad_norm": 0.3629656434059143,
      "learning_rate": 9.324839158556542e-06,
      "loss": 0.3337,
      "step": 110
    },
    {
      "epoch": 0.2125418860698899,
      "grad_norm": 0.4247240424156189,
      "learning_rate": 9.308860003717748e-06,
      "loss": 0.3809,
      "step": 111
    },
    {
      "epoch": 0.2144566778362853,
      "grad_norm": 0.4432498812675476,
      "learning_rate": 9.292707987758202e-06,
      "loss": 0.398,
      "step": 112
    },
    {
      "epoch": 0.21637146960268072,
      "grad_norm": 0.45140188932418823,
      "learning_rate": 9.27638375865801e-06,
      "loss": 0.3938,
      "step": 113
    },
    {
      "epoch": 0.2182862613690761,
      "grad_norm": 0.5100187063217163,
      "learning_rate": 9.259887971306064e-06,
      "loss": 0.4427,
      "step": 114
    },
    {
      "epoch": 0.22020105313547153,
      "grad_norm": 0.5069835782051086,
      "learning_rate": 9.243221287473755e-06,
      "loss": 0.4131,
      "step": 115
    },
    {
      "epoch": 0.22211584490186692,
      "grad_norm": 0.4621468782424927,
      "learning_rate": 9.226384375788435e-06,
      "loss": 0.3788,
      "step": 116
    },
    {
      "epoch": 0.22403063666826234,
      "grad_norm": 0.4977872371673584,
      "learning_rate": 9.209377911706585e-06,
      "loss": 0.3719,
      "step": 117
    },
    {
      "epoch": 0.22594542843465773,
      "grad_norm": 0.4791455566883087,
      "learning_rate": 9.192202577486725e-06,
      "loss": 0.3562,
      "step": 118
    },
    {
      "epoch": 0.22786022020105315,
      "grad_norm": 0.4171106219291687,
      "learning_rate": 9.174859062162037e-06,
      "loss": 0.3568,
      "step": 119
    },
    {
      "epoch": 0.22977501196744854,
      "grad_norm": 0.4501950740814209,
      "learning_rate": 9.157348061512728e-06,
      "loss": 0.4015,
      "step": 120
    },
    {
      "epoch": 0.23168980373384396,
      "grad_norm": 0.4359966814517975,
      "learning_rate": 9.139670278038109e-06,
      "loss": 0.4014,
      "step": 121
    },
    {
      "epoch": 0.23360459550023935,
      "grad_norm": 0.3942509889602661,
      "learning_rate": 9.121826420928421e-06,
      "loss": 0.3328,
      "step": 122
    },
    {
      "epoch": 0.23551938726663477,
      "grad_norm": 0.4340513050556183,
      "learning_rate": 9.103817206036383e-06,
      "loss": 0.3628,
      "step": 123
    },
    {
      "epoch": 0.23743417903303016,
      "grad_norm": 0.44619980454444885,
      "learning_rate": 9.085643355848468e-06,
      "loss": 0.4003,
      "step": 124
    },
    {
      "epoch": 0.23934897079942558,
      "grad_norm": 0.43487605452537537,
      "learning_rate": 9.06730559945592e-06,
      "loss": 0.3558,
      "step": 125
    },
    {
      "epoch": 0.24126376256582097,
      "grad_norm": 0.4643803834915161,
      "learning_rate": 9.048804672525513e-06,
      "loss": 0.375,
      "step": 126
    },
    {
      "epoch": 0.24317855433221638,
      "grad_norm": 0.44975554943084717,
      "learning_rate": 9.030141317270026e-06,
      "loss": 0.3602,
      "step": 127
    },
    {
      "epoch": 0.24509334609861178,
      "grad_norm": 0.45412832498550415,
      "learning_rate": 9.011316282418474e-06,
      "loss": 0.3724,
      "step": 128
    },
    {
      "epoch": 0.2470081378650072,
      "grad_norm": 0.4402771592140198,
      "learning_rate": 8.992330323186069e-06,
      "loss": 0.3945,
      "step": 129
    },
    {
      "epoch": 0.24892292963140258,
      "grad_norm": 0.46597278118133545,
      "learning_rate": 8.973184201243922e-06,
      "loss": 0.4407,
      "step": 130
    },
    {
      "epoch": 0.250837721397798,
      "grad_norm": 0.44561654329299927,
      "learning_rate": 8.953878684688492e-06,
      "loss": 0.3814,
      "step": 131
    },
    {
      "epoch": 0.2527525131641934,
      "grad_norm": 0.4216248095035553,
      "learning_rate": 8.934414548010764e-06,
      "loss": 0.3647,
      "step": 132
    },
    {
      "epoch": 0.2546673049305888,
      "grad_norm": 0.4280121326446533,
      "learning_rate": 8.914792572065178e-06,
      "loss": 0.3733,
      "step": 133
    },
    {
      "epoch": 0.25658209669698423,
      "grad_norm": 0.5127595067024231,
      "learning_rate": 8.89501354403831e-06,
      "loss": 0.4195,
      "step": 134
    },
    {
      "epoch": 0.2584968884633796,
      "grad_norm": 0.5074846744537354,
      "learning_rate": 8.875078257417294e-06,
      "loss": 0.4204,
      "step": 135
    },
    {
      "epoch": 0.260411680229775,
      "grad_norm": 0.4372941553592682,
      "learning_rate": 8.854987511957974e-06,
      "loss": 0.3924,
      "step": 136
    },
    {
      "epoch": 0.26232647199617043,
      "grad_norm": 0.482693076133728,
      "learning_rate": 8.834742113652835e-06,
      "loss": 0.3989,
      "step": 137
    },
    {
      "epoch": 0.26424126376256585,
      "grad_norm": 0.40820303559303284,
      "learning_rate": 8.81434287469866e-06,
      "loss": 0.3339,
      "step": 138
    },
    {
      "epoch": 0.2661560555289612,
      "grad_norm": 0.5613487362861633,
      "learning_rate": 8.793790613463956e-06,
      "loss": 0.3689,
      "step": 139
    },
    {
      "epoch": 0.26807084729535663,
      "grad_norm": 0.4296424686908722,
      "learning_rate": 8.773086154456106e-06,
      "loss": 0.3781,
      "step": 140
    },
    {
      "epoch": 0.26998563906175205,
      "grad_norm": 0.4859068989753723,
      "learning_rate": 8.752230328288314e-06,
      "loss": 0.4299,
      "step": 141
    },
    {
      "epoch": 0.27190043082814747,
      "grad_norm": 0.4237840175628662,
      "learning_rate": 8.731223971646261e-06,
      "loss": 0.3507,
      "step": 142
    },
    {
      "epoch": 0.27381522259454283,
      "grad_norm": 0.40383824706077576,
      "learning_rate": 8.710067927254555e-06,
      "loss": 0.344,
      "step": 143
    },
    {
      "epoch": 0.27573001436093825,
      "grad_norm": 0.40671247243881226,
      "learning_rate": 8.688763043842916e-06,
      "loss": 0.361,
      "step": 144
    },
    {
      "epoch": 0.27764480612733367,
      "grad_norm": 0.4422585666179657,
      "learning_rate": 8.66731017611213e-06,
      "loss": 0.3884,
      "step": 145
    },
    {
      "epoch": 0.2795595978937291,
      "grad_norm": 0.4300481081008911,
      "learning_rate": 8.645710184699756e-06,
      "loss": 0.3804,
      "step": 146
    },
    {
      "epoch": 0.28147438966012445,
      "grad_norm": 0.43995732069015503,
      "learning_rate": 8.6239639361456e-06,
      "loss": 0.3715,
      "step": 147
    },
    {
      "epoch": 0.28338918142651986,
      "grad_norm": 0.39543983340263367,
      "learning_rate": 8.602072302856961e-06,
      "loss": 0.3664,
      "step": 148
    },
    {
      "epoch": 0.2853039731929153,
      "grad_norm": 0.4740181863307953,
      "learning_rate": 8.580036163073615e-06,
      "loss": 0.3998,
      "step": 149
    },
    {
      "epoch": 0.2872187649593107,
      "grad_norm": 0.4195634126663208,
      "learning_rate": 8.5578564008326e-06,
      "loss": 0.3514,
      "step": 150
    },
    {
      "epoch": 0.28913355672570606,
      "grad_norm": 0.4382091164588928,
      "learning_rate": 8.535533905932739e-06,
      "loss": 0.3645,
      "step": 151
    },
    {
      "epoch": 0.2910483484921015,
      "grad_norm": 0.47153332829475403,
      "learning_rate": 8.513069573898944e-06,
      "loss": 0.3709,
      "step": 152
    },
    {
      "epoch": 0.2929631402584969,
      "grad_norm": 0.428423672914505,
      "learning_rate": 8.490464305946296e-06,
      "loss": 0.377,
      "step": 153
    },
    {
      "epoch": 0.2948779320248923,
      "grad_norm": 0.4477589428424835,
      "learning_rate": 8.467719008943886e-06,
      "loss": 0.3696,
      "step": 154
    },
    {
      "epoch": 0.2967927237912877,
      "grad_norm": 0.44177618622779846,
      "learning_rate": 8.444834595378434e-06,
      "loss": 0.3793,
      "step": 155
    },
    {
      "epoch": 0.2987075155576831,
      "grad_norm": 0.40658193826675415,
      "learning_rate": 8.421811983317682e-06,
      "loss": 0.3564,
      "step": 156
    },
    {
      "epoch": 0.3006223073240785,
      "grad_norm": 0.40630021691322327,
      "learning_rate": 8.398652096373566e-06,
      "loss": 0.3743,
      "step": 157
    },
    {
      "epoch": 0.30253709909047394,
      "grad_norm": 0.419085294008255,
      "learning_rate": 8.375355863665155e-06,
      "loss": 0.3609,
      "step": 158
    },
    {
      "epoch": 0.3044518908568693,
      "grad_norm": 0.4741552174091339,
      "learning_rate": 8.351924219781393e-06,
      "loss": 0.4016,
      "step": 159
    },
    {
      "epoch": 0.3063666826232647,
      "grad_norm": 0.445846825838089,
      "learning_rate": 8.328358104743588e-06,
      "loss": 0.3714,
      "step": 160
    },
    {
      "epoch": 0.30828147438966014,
      "grad_norm": 0.4190356135368347,
      "learning_rate": 8.304658463967705e-06,
      "loss": 0.3718,
      "step": 161
    },
    {
      "epoch": 0.31019626615605556,
      "grad_norm": 0.4018896222114563,
      "learning_rate": 8.28082624822645e-06,
      "loss": 0.3352,
      "step": 162
    },
    {
      "epoch": 0.3121110579224509,
      "grad_norm": 0.41169753670692444,
      "learning_rate": 8.256862413611113e-06,
      "loss": 0.3439,
      "step": 163
    },
    {
      "epoch": 0.31402584968884634,
      "grad_norm": 0.4585376977920532,
      "learning_rate": 8.232767921493216e-06,
      "loss": 0.3901,
      "step": 164
    },
    {
      "epoch": 0.31594064145524176,
      "grad_norm": 0.4590586721897125,
      "learning_rate": 8.20854373848595e-06,
      "loss": 0.4042,
      "step": 165
    },
    {
      "epoch": 0.3178554332216372,
      "grad_norm": 0.42637982964515686,
      "learning_rate": 8.184190836405394e-06,
      "loss": 0.3295,
      "step": 166
    },
    {
      "epoch": 0.31977022498803254,
      "grad_norm": 0.45910605788230896,
      "learning_rate": 8.15971019223152e-06,
      "loss": 0.3759,
      "step": 167
    },
    {
      "epoch": 0.32168501675442795,
      "grad_norm": 0.3774631917476654,
      "learning_rate": 8.135102788069015e-06,
      "loss": 0.3142,
      "step": 168
    },
    {
      "epoch": 0.3235998085208234,
      "grad_norm": 0.3829216957092285,
      "learning_rate": 8.110369611107869e-06,
      "loss": 0.3288,
      "step": 169
    },
    {
      "epoch": 0.3255146002872188,
      "grad_norm": 0.42439737915992737,
      "learning_rate": 8.085511653583772e-06,
      "loss": 0.3452,
      "step": 170
    },
    {
      "epoch": 0.32742939205361415,
      "grad_norm": 0.4537997543811798,
      "learning_rate": 8.060529912738316e-06,
      "loss": 0.4024,
      "step": 171
    },
    {
      "epoch": 0.3293441838200096,
      "grad_norm": 0.5059738755226135,
      "learning_rate": 8.035425390778975e-06,
      "loss": 0.4093,
      "step": 172
    },
    {
      "epoch": 0.331258975586405,
      "grad_norm": 0.41058847308158875,
      "learning_rate": 8.010199094838915e-06,
      "loss": 0.3704,
      "step": 173
    },
    {
      "epoch": 0.3331737673528004,
      "grad_norm": 0.43019795417785645,
      "learning_rate": 7.984852036936578e-06,
      "loss": 0.3463,
      "step": 174
    },
    {
      "epoch": 0.3350885591191958,
      "grad_norm": 0.4632018506526947,
      "learning_rate": 7.959385233935087e-06,
      "loss": 0.4076,
      "step": 175
    },
    {
      "epoch": 0.3370033508855912,
      "grad_norm": 0.363118976354599,
      "learning_rate": 7.933799707501448e-06,
      "loss": 0.3089,
      "step": 176
    },
    {
      "epoch": 0.3389181426519866,
      "grad_norm": 0.414247065782547,
      "learning_rate": 7.908096484065569e-06,
      "loss": 0.3691,
      "step": 177
    },
    {
      "epoch": 0.340832934418382,
      "grad_norm": 0.4680274724960327,
      "learning_rate": 7.88227659477908e-06,
      "loss": 0.3742,
      "step": 178
    },
    {
      "epoch": 0.3427477261847774,
      "grad_norm": 0.4838680624961853,
      "learning_rate": 7.856341075473963e-06,
      "loss": 0.3509,
      "step": 179
    },
    {
      "epoch": 0.3446625179511728,
      "grad_norm": 0.39406099915504456,
      "learning_rate": 7.830290966620997e-06,
      "loss": 0.327,
      "step": 180
    },
    {
      "epoch": 0.3465773097175682,
      "grad_norm": 0.4498644471168518,
      "learning_rate": 7.804127313288023e-06,
      "loss": 0.3665,
      "step": 181
    },
    {
      "epoch": 0.34849210148396365,
      "grad_norm": 0.42900529503822327,
      "learning_rate": 7.777851165098012e-06,
      "loss": 0.3779,
      "step": 182
    },
    {
      "epoch": 0.350406893250359,
      "grad_norm": 0.3877162039279938,
      "learning_rate": 7.751463576186957e-06,
      "loss": 0.3394,
      "step": 183
    },
    {
      "epoch": 0.3523216850167544,
      "grad_norm": 0.6071410179138184,
      "learning_rate": 7.72496560516159e-06,
      "loss": 0.3813,
      "step": 184
    },
    {
      "epoch": 0.35423647678314985,
      "grad_norm": 0.5272730588912964,
      "learning_rate": 7.6983583150569e-06,
      "loss": 0.3817,
      "step": 185
    },
    {
      "epoch": 0.35615126854954526,
      "grad_norm": 0.4268261194229126,
      "learning_rate": 7.671642773293506e-06,
      "loss": 0.3415,
      "step": 186
    },
    {
      "epoch": 0.3580660603159406,
      "grad_norm": 0.4886492192745209,
      "learning_rate": 7.644820051634813e-06,
      "loss": 0.3846,
      "step": 187
    },
    {
      "epoch": 0.35998085208233604,
      "grad_norm": 0.4869603216648102,
      "learning_rate": 7.617891226144034e-06,
      "loss": 0.4274,
      "step": 188
    },
    {
      "epoch": 0.36189564384873146,
      "grad_norm": 0.42415544390678406,
      "learning_rate": 7.59085737714101e-06,
      "loss": 0.3438,
      "step": 189
    },
    {
      "epoch": 0.3638104356151269,
      "grad_norm": 0.5172827243804932,
      "learning_rate": 7.563719589158874e-06,
      "loss": 0.3836,
      "step": 190
    },
    {
      "epoch": 0.36572522738152224,
      "grad_norm": 0.4153439998626709,
      "learning_rate": 7.536478950900537e-06,
      "loss": 0.3636,
      "step": 191
    },
    {
      "epoch": 0.36764001914791766,
      "grad_norm": 0.42014452815055847,
      "learning_rate": 7.509136555195025e-06,
      "loss": 0.3501,
      "step": 192
    },
    {
      "epoch": 0.3695548109143131,
      "grad_norm": 0.4194999933242798,
      "learning_rate": 7.481693498953621e-06,
      "loss": 0.3745,
      "step": 193
    },
    {
      "epoch": 0.3714696026807085,
      "grad_norm": 0.4361554980278015,
      "learning_rate": 7.4541508831258695e-06,
      "loss": 0.3469,
      "step": 194
    },
    {
      "epoch": 0.37338439444710386,
      "grad_norm": 0.3900616765022278,
      "learning_rate": 7.4265098126554065e-06,
      "loss": 0.3354,
      "step": 195
    },
    {
      "epoch": 0.3752991862134993,
      "grad_norm": 0.47491058707237244,
      "learning_rate": 7.3987713964356335e-06,
      "loss": 0.4077,
      "step": 196
    },
    {
      "epoch": 0.3772139779798947,
      "grad_norm": 0.5020147562026978,
      "learning_rate": 7.370936747265226e-06,
      "loss": 0.4198,
      "step": 197
    },
    {
      "epoch": 0.3791287697462901,
      "grad_norm": 0.4094443619251251,
      "learning_rate": 7.3430069818035e-06,
      "loss": 0.3482,
      "step": 198
    },
    {
      "epoch": 0.3810435615126855,
      "grad_norm": 0.4021066427230835,
      "learning_rate": 7.314983220525604e-06,
      "loss": 0.3561,
      "step": 199
    },
    {
      "epoch": 0.3829583532790809,
      "grad_norm": 0.3932279348373413,
      "learning_rate": 7.286866587677576e-06,
      "loss": 0.3414,
      "step": 200
    },
    {
      "epoch": 0.3848731450454763,
      "grad_norm": 0.466545045375824,
      "learning_rate": 7.2586582112312355e-06,
      "loss": 0.3868,
      "step": 201
    },
    {
      "epoch": 0.38678793681187174,
      "grad_norm": 0.4097374677658081,
      "learning_rate": 7.230359222838939e-06,
      "loss": 0.3468,
      "step": 202
    },
    {
      "epoch": 0.3887027285782671,
      "grad_norm": 0.4792751967906952,
      "learning_rate": 7.201970757788172e-06,
      "loss": 0.4131,
      "step": 203
    },
    {
      "epoch": 0.3906175203446625,
      "grad_norm": 0.40435218811035156,
      "learning_rate": 7.173493954956012e-06,
      "loss": 0.3456,
      "step": 204
    },
    {
      "epoch": 0.39253231211105793,
      "grad_norm": 0.45885634422302246,
      "learning_rate": 7.144929956763438e-06,
      "loss": 0.3672,
      "step": 205
    },
    {
      "epoch": 0.39444710387745335,
      "grad_norm": 0.5055181384086609,
      "learning_rate": 7.116279909129492e-06,
      "loss": 0.4122,
      "step": 206
    },
    {
      "epoch": 0.3963618956438487,
      "grad_norm": 0.42308080196380615,
      "learning_rate": 7.087544961425317e-06,
      "loss": 0.3621,
      "step": 207
    },
    {
      "epoch": 0.39827668741024413,
      "grad_norm": 0.41654130816459656,
      "learning_rate": 7.058726266428042e-06,
      "loss": 0.3623,
      "step": 208
    },
    {
      "epoch": 0.40019147917663955,
      "grad_norm": 0.3884025812149048,
      "learning_rate": 7.029824980274536e-06,
      "loss": 0.3302,
      "step": 209
    },
    {
      "epoch": 0.40210627094303497,
      "grad_norm": 0.43903034925460815,
      "learning_rate": 7.0008422624150285e-06,
      "loss": 0.3728,
      "step": 210
    }
  ],
  "logging_steps": 1,
  "max_steps": 522,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 105,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.1029906388628275e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}