{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 0,
  "global_step": 239,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0041841004184100415,
      "grad_norm": 0.6393154263496399,
      "learning_rate": 9.958158995815901e-06,
      "loss": 1.9217,
      "step": 1
    },
    {
      "epoch": 0.008368200836820083,
      "grad_norm": 0.63503098487854,
      "learning_rate": 9.9163179916318e-06,
      "loss": 1.9545,
      "step": 2
    },
    {
      "epoch": 0.012552301255230125,
      "grad_norm": 0.6524788737297058,
      "learning_rate": 9.874476987447699e-06,
      "loss": 2.0203,
      "step": 3
    },
    {
      "epoch": 0.016736401673640166,
      "grad_norm": 0.6311175227165222,
      "learning_rate": 9.8326359832636e-06,
      "loss": 1.9856,
      "step": 4
    },
    {
      "epoch": 0.02092050209205021,
      "grad_norm": 0.5862050652503967,
      "learning_rate": 9.790794979079498e-06,
      "loss": 1.921,
      "step": 5
    },
    {
      "epoch": 0.02510460251046025,
      "grad_norm": 0.5720086693763733,
      "learning_rate": 9.748953974895399e-06,
      "loss": 1.8198,
      "step": 6
    },
    {
      "epoch": 0.029288702928870293,
      "grad_norm": 0.6093158721923828,
      "learning_rate": 9.707112970711298e-06,
      "loss": 1.9426,
      "step": 7
    },
    {
      "epoch": 0.03347280334728033,
      "grad_norm": 0.5525884628295898,
      "learning_rate": 9.665271966527198e-06,
      "loss": 1.8915,
      "step": 8
    },
    {
      "epoch": 0.03765690376569038,
      "grad_norm": 0.5531294941902161,
      "learning_rate": 9.623430962343097e-06,
      "loss": 1.926,
      "step": 9
    },
    {
      "epoch": 0.04184100418410042,
      "grad_norm": 0.5190953612327576,
      "learning_rate": 9.581589958158996e-06,
      "loss": 1.8451,
      "step": 10
    },
    {
      "epoch": 0.04602510460251046,
      "grad_norm": 0.510351300239563,
      "learning_rate": 9.539748953974896e-06,
      "loss": 1.8092,
      "step": 11
    },
    {
      "epoch": 0.0502092050209205,
      "grad_norm": 0.4969535171985626,
      "learning_rate": 9.497907949790795e-06,
      "loss": 1.7935,
      "step": 12
    },
    {
      "epoch": 0.05439330543933055,
      "grad_norm": 0.5025241374969482,
      "learning_rate": 9.456066945606696e-06,
      "loss": 1.9057,
      "step": 13
    },
    {
      "epoch": 0.058577405857740586,
      "grad_norm": 0.5688336491584778,
      "learning_rate": 9.414225941422594e-06,
      "loss": 1.8393,
      "step": 14
    },
    {
      "epoch": 0.06276150627615062,
      "grad_norm": 0.47667714953422546,
      "learning_rate": 9.372384937238495e-06,
      "loss": 1.8035,
      "step": 15
    },
    {
      "epoch": 0.06694560669456066,
      "grad_norm": 0.5087830424308777,
      "learning_rate": 9.330543933054394e-06,
      "loss": 1.8427,
      "step": 16
    },
    {
      "epoch": 0.07112970711297072,
      "grad_norm": 0.4639821946620941,
      "learning_rate": 9.288702928870293e-06,
      "loss": 1.7071,
      "step": 17
    },
    {
      "epoch": 0.07531380753138076,
      "grad_norm": 0.4799270033836365,
      "learning_rate": 9.246861924686193e-06,
      "loss": 1.782,
      "step": 18
    },
    {
      "epoch": 0.0794979079497908,
      "grad_norm": 0.4342242181301117,
      "learning_rate": 9.205020920502092e-06,
      "loss": 1.7014,
      "step": 19
    },
    {
      "epoch": 0.08368200836820083,
      "grad_norm": 0.44706302881240845,
      "learning_rate": 9.163179916317992e-06,
      "loss": 1.7027,
      "step": 20
    },
    {
      "epoch": 0.08786610878661087,
      "grad_norm": 0.4259682893753052,
      "learning_rate": 9.121338912133893e-06,
      "loss": 1.745,
      "step": 21
    },
    {
      "epoch": 0.09205020920502092,
      "grad_norm": 0.40614959597587585,
      "learning_rate": 9.079497907949792e-06,
      "loss": 1.6348,
      "step": 22
    },
    {
      "epoch": 0.09623430962343096,
      "grad_norm": 0.41480347514152527,
      "learning_rate": 9.03765690376569e-06,
      "loss": 1.6336,
      "step": 23
    },
    {
      "epoch": 0.100418410041841,
      "grad_norm": 0.40425121784210205,
      "learning_rate": 8.995815899581591e-06,
      "loss": 1.6534,
      "step": 24
    },
    {
      "epoch": 0.10460251046025104,
      "grad_norm": 0.40777212381362915,
      "learning_rate": 8.95397489539749e-06,
      "loss": 1.7132,
      "step": 25
    },
    {
      "epoch": 0.1087866108786611,
      "grad_norm": 0.41130122542381287,
      "learning_rate": 8.91213389121339e-06,
      "loss": 1.6169,
      "step": 26
    },
    {
      "epoch": 0.11297071129707113,
      "grad_norm": 0.47633782029151917,
      "learning_rate": 8.87029288702929e-06,
      "loss": 1.6586,
      "step": 27
    },
    {
      "epoch": 0.11715481171548117,
      "grad_norm": 0.43167147040367126,
      "learning_rate": 8.82845188284519e-06,
      "loss": 1.6174,
      "step": 28
    },
    {
      "epoch": 0.12133891213389121,
      "grad_norm": 0.4019918739795685,
      "learning_rate": 8.786610878661089e-06,
      "loss": 1.6277,
      "step": 29
    },
    {
      "epoch": 0.12552301255230125,
      "grad_norm": 0.3712836802005768,
      "learning_rate": 8.744769874476987e-06,
      "loss": 1.6287,
      "step": 30
    },
    {
      "epoch": 0.1297071129707113,
      "grad_norm": 0.4354589283466339,
      "learning_rate": 8.702928870292888e-06,
      "loss": 1.6389,
      "step": 31
    },
    {
      "epoch": 0.13389121338912133,
      "grad_norm": 0.386348694562912,
      "learning_rate": 8.661087866108787e-06,
      "loss": 1.6217,
      "step": 32
    },
    {
      "epoch": 0.13807531380753138,
      "grad_norm": 0.3922019600868225,
      "learning_rate": 8.619246861924687e-06,
      "loss": 1.6063,
      "step": 33
    },
    {
      "epoch": 0.14225941422594143,
      "grad_norm": 0.34336555004119873,
      "learning_rate": 8.577405857740586e-06,
      "loss": 1.5818,
      "step": 34
    },
    {
      "epoch": 0.14644351464435146,
      "grad_norm": 0.39002758264541626,
      "learning_rate": 8.535564853556487e-06,
      "loss": 1.6544,
      "step": 35
    },
    {
      "epoch": 0.1506276150627615,
      "grad_norm": 0.32121986150741577,
      "learning_rate": 8.493723849372385e-06,
      "loss": 1.5688,
      "step": 36
    },
    {
      "epoch": 0.15481171548117154,
      "grad_norm": 0.33398061990737915,
      "learning_rate": 8.451882845188284e-06,
      "loss": 1.5353,
      "step": 37
    },
    {
      "epoch": 0.1589958158995816,
      "grad_norm": 0.32710814476013184,
      "learning_rate": 8.410041841004185e-06,
      "loss": 1.5489,
      "step": 38
    },
    {
      "epoch": 0.16317991631799164,
      "grad_norm": 0.3649294674396515,
      "learning_rate": 8.368200836820084e-06,
      "loss": 1.5735,
      "step": 39
    },
    {
      "epoch": 0.16736401673640167,
      "grad_norm": 0.3286899924278259,
      "learning_rate": 8.326359832635984e-06,
      "loss": 1.5176,
      "step": 40
    },
    {
      "epoch": 0.17154811715481172,
      "grad_norm": 0.4433610439300537,
      "learning_rate": 8.284518828451885e-06,
      "loss": 1.4534,
      "step": 41
    },
    {
      "epoch": 0.17573221757322174,
      "grad_norm": 0.2977966368198395,
      "learning_rate": 8.242677824267783e-06,
      "loss": 1.4857,
      "step": 42
    },
    {
      "epoch": 0.1799163179916318,
      "grad_norm": 0.356638103723526,
      "learning_rate": 8.200836820083682e-06,
      "loss": 1.5164,
      "step": 43
    },
    {
      "epoch": 0.18410041841004185,
      "grad_norm": 0.3443019390106201,
      "learning_rate": 8.158995815899581e-06,
      "loss": 1.6193,
      "step": 44
    },
    {
      "epoch": 0.18828451882845187,
      "grad_norm": 0.35785749554634094,
      "learning_rate": 8.117154811715482e-06,
      "loss": 1.5293,
      "step": 45
    },
    {
      "epoch": 0.19246861924686193,
      "grad_norm": 0.3255775272846222,
      "learning_rate": 8.075313807531382e-06,
      "loss": 1.5232,
      "step": 46
    },
    {
      "epoch": 0.19665271966527198,
      "grad_norm": 0.3624202311038971,
      "learning_rate": 8.033472803347281e-06,
      "loss": 1.4129,
      "step": 47
    },
    {
      "epoch": 0.200836820083682,
      "grad_norm": 0.3205145001411438,
      "learning_rate": 7.991631799163181e-06,
      "loss": 1.4812,
      "step": 48
    },
    {
      "epoch": 0.20502092050209206,
      "grad_norm": 0.43222349882125854,
      "learning_rate": 7.94979079497908e-06,
      "loss": 1.5202,
      "step": 49
    },
    {
      "epoch": 0.20920502092050208,
      "grad_norm": 0.31052061915397644,
      "learning_rate": 7.907949790794979e-06,
      "loss": 1.5129,
      "step": 50
    },
    {
      "epoch": 0.21338912133891214,
      "grad_norm": 0.33750268816947937,
      "learning_rate": 7.86610878661088e-06,
      "loss": 1.4717,
      "step": 51
    },
    {
      "epoch": 0.2175732217573222,
      "grad_norm": 0.3067304790019989,
      "learning_rate": 7.824267782426778e-06,
      "loss": 1.4624,
      "step": 52
    },
    {
      "epoch": 0.2217573221757322,
      "grad_norm": 0.3119943141937256,
      "learning_rate": 7.782426778242679e-06,
      "loss": 1.426,
      "step": 53
    },
    {
      "epoch": 0.22594142259414227,
      "grad_norm": 0.32612940669059753,
      "learning_rate": 7.740585774058578e-06,
      "loss": 1.4734,
      "step": 54
    },
    {
      "epoch": 0.2301255230125523,
      "grad_norm": 0.3576732277870178,
      "learning_rate": 7.698744769874478e-06,
      "loss": 1.4905,
      "step": 55
    },
    {
      "epoch": 0.23430962343096234,
      "grad_norm": 0.30397745966911316,
      "learning_rate": 7.656903765690377e-06,
      "loss": 1.4504,
      "step": 56
    },
    {
      "epoch": 0.2384937238493724,
      "grad_norm": 0.2985887825489044,
      "learning_rate": 7.615062761506277e-06,
      "loss": 1.4289,
      "step": 57
    },
    {
      "epoch": 0.24267782426778242,
      "grad_norm": 0.3366187512874603,
      "learning_rate": 7.573221757322176e-06,
      "loss": 1.4332,
      "step": 58
    },
    {
      "epoch": 0.24686192468619247,
      "grad_norm": 0.3567534387111664,
      "learning_rate": 7.531380753138075e-06,
      "loss": 1.3826,
      "step": 59
    },
    {
      "epoch": 0.2510460251046025,
      "grad_norm": 0.3008073568344116,
      "learning_rate": 7.489539748953976e-06,
      "loss": 1.4754,
      "step": 60
    },
    {
      "epoch": 0.25523012552301255,
      "grad_norm": 0.29970571398735046,
      "learning_rate": 7.4476987447698746e-06,
      "loss": 1.459,
      "step": 61
    },
    {
      "epoch": 0.2594142259414226,
      "grad_norm": 0.4011840522289276,
      "learning_rate": 7.405857740585774e-06,
      "loss": 1.4442,
      "step": 62
    },
    {
      "epoch": 0.26359832635983266,
      "grad_norm": 0.3857061564922333,
      "learning_rate": 7.364016736401675e-06,
      "loss": 1.4669,
      "step": 63
    },
    {
      "epoch": 0.26778242677824265,
      "grad_norm": 0.34745919704437256,
      "learning_rate": 7.3221757322175736e-06,
      "loss": 1.386,
      "step": 64
    },
    {
      "epoch": 0.2719665271966527,
      "grad_norm": 0.37241384387016296,
      "learning_rate": 7.280334728033473e-06,
      "loss": 1.4307,
      "step": 65
    },
    {
      "epoch": 0.27615062761506276,
      "grad_norm": 0.3590523898601532,
      "learning_rate": 7.238493723849372e-06,
      "loss": 1.4181,
      "step": 66
    },
    {
      "epoch": 0.2803347280334728,
      "grad_norm": 0.3953969180583954,
      "learning_rate": 7.1966527196652726e-06,
      "loss": 1.3858,
      "step": 67
    },
    {
      "epoch": 0.28451882845188287,
      "grad_norm": 0.43755462765693665,
      "learning_rate": 7.154811715481172e-06,
      "loss": 1.4155,
      "step": 68
    },
    {
      "epoch": 0.28870292887029286,
      "grad_norm": 0.355756938457489,
      "learning_rate": 7.112970711297071e-06,
      "loss": 1.3971,
      "step": 69
    },
    {
      "epoch": 0.2928870292887029,
      "grad_norm": 0.4370933175086975,
      "learning_rate": 7.0711297071129716e-06,
      "loss": 1.3027,
      "step": 70
    },
    {
      "epoch": 0.29707112970711297,
      "grad_norm": 0.31537461280822754,
      "learning_rate": 7.02928870292887e-06,
      "loss": 1.4147,
      "step": 71
    },
    {
      "epoch": 0.301255230125523,
      "grad_norm": 0.32908475399017334,
      "learning_rate": 6.98744769874477e-06,
      "loss": 1.4637,
      "step": 72
    },
    {
      "epoch": 0.3054393305439331,
      "grad_norm": 0.2952563464641571,
      "learning_rate": 6.9456066945606706e-06,
      "loss": 1.3477,
      "step": 73
    },
    {
      "epoch": 0.30962343096234307,
      "grad_norm": 0.29872894287109375,
      "learning_rate": 6.903765690376569e-06,
      "loss": 1.4093,
      "step": 74
    },
    {
      "epoch": 0.3138075313807531,
      "grad_norm": 0.3141920566558838,
      "learning_rate": 6.861924686192469e-06,
      "loss": 1.3911,
      "step": 75
    },
    {
      "epoch": 0.3179916317991632,
      "grad_norm": 0.3398382067680359,
      "learning_rate": 6.820083682008368e-06,
      "loss": 1.437,
      "step": 76
    },
    {
      "epoch": 0.32217573221757323,
      "grad_norm": 0.40376701951026917,
      "learning_rate": 6.778242677824268e-06,
      "loss": 1.3147,
      "step": 77
    },
    {
      "epoch": 0.3263598326359833,
      "grad_norm": 0.30617496371269226,
      "learning_rate": 6.736401673640168e-06,
      "loss": 1.3865,
      "step": 78
    },
    {
      "epoch": 0.3305439330543933,
      "grad_norm": 0.338013619184494,
      "learning_rate": 6.694560669456067e-06,
      "loss": 1.3723,
      "step": 79
    },
    {
      "epoch": 0.33472803347280333,
      "grad_norm": 0.4524187445640564,
      "learning_rate": 6.652719665271967e-06,
      "loss": 1.3681,
      "step": 80
    },
    {
      "epoch": 0.3389121338912134,
      "grad_norm": 0.33039766550064087,
      "learning_rate": 6.610878661087866e-06,
      "loss": 1.4012,
      "step": 81
    },
    {
      "epoch": 0.34309623430962344,
      "grad_norm": 0.434511661529541,
      "learning_rate": 6.569037656903766e-06,
      "loss": 1.4246,
      "step": 82
    },
    {
      "epoch": 0.3472803347280335,
      "grad_norm": 0.3561013340950012,
      "learning_rate": 6.527196652719666e-06,
      "loss": 1.3597,
      "step": 83
    },
    {
      "epoch": 0.3514644351464435,
      "grad_norm": 0.4421449899673462,
      "learning_rate": 6.485355648535565e-06,
      "loss": 1.4445,
      "step": 84
    },
    {
      "epoch": 0.35564853556485354,
      "grad_norm": 0.332782119512558,
      "learning_rate": 6.443514644351465e-06,
      "loss": 1.4093,
      "step": 85
    },
    {
      "epoch": 0.3598326359832636,
      "grad_norm": 0.35183823108673096,
      "learning_rate": 6.401673640167364e-06,
      "loss": 1.3283,
      "step": 86
    },
    {
      "epoch": 0.36401673640167365,
      "grad_norm": 0.33525529503822327,
      "learning_rate": 6.359832635983264e-06,
      "loss": 1.4419,
      "step": 87
    },
    {
      "epoch": 0.3682008368200837,
      "grad_norm": 0.33096978068351746,
      "learning_rate": 6.317991631799164e-06,
      "loss": 1.3691,
      "step": 88
    },
    {
      "epoch": 0.3723849372384937,
      "grad_norm": 0.3147584795951843,
      "learning_rate": 6.276150627615063e-06,
      "loss": 1.4014,
      "step": 89
    },
    {
      "epoch": 0.37656903765690375,
      "grad_norm": 0.3018977642059326,
      "learning_rate": 6.234309623430963e-06,
      "loss": 1.3661,
      "step": 90
    },
    {
      "epoch": 0.3807531380753138,
      "grad_norm": 0.31138119101524353,
      "learning_rate": 6.192468619246862e-06,
      "loss": 1.3163,
      "step": 91
    },
    {
      "epoch": 0.38493723849372385,
      "grad_norm": 0.35807517170906067,
      "learning_rate": 6.150627615062762e-06,
      "loss": 1.3546,
      "step": 92
    },
    {
      "epoch": 0.3891213389121339,
      "grad_norm": 0.3687175512313843,
      "learning_rate": 6.108786610878662e-06,
      "loss": 1.3204,
      "step": 93
    },
    {
      "epoch": 0.39330543933054396,
      "grad_norm": 0.3205777704715729,
      "learning_rate": 6.066945606694561e-06,
      "loss": 1.3651,
      "step": 94
    },
    {
      "epoch": 0.39748953974895396,
      "grad_norm": 0.3424644470214844,
      "learning_rate": 6.025104602510461e-06,
      "loss": 1.3837,
      "step": 95
    },
    {
      "epoch": 0.401673640167364,
      "grad_norm": 0.334182471036911,
      "learning_rate": 5.9832635983263595e-06,
      "loss": 1.3737,
      "step": 96
    },
    {
      "epoch": 0.40585774058577406,
      "grad_norm": 0.32831668853759766,
      "learning_rate": 5.94142259414226e-06,
      "loss": 1.3455,
      "step": 97
    },
    {
      "epoch": 0.4100418410041841,
      "grad_norm": 0.2965930700302124,
      "learning_rate": 5.89958158995816e-06,
      "loss": 1.3143,
      "step": 98
    },
    {
      "epoch": 0.41422594142259417,
      "grad_norm": 0.38945430517196655,
      "learning_rate": 5.8577405857740585e-06,
      "loss": 1.3119,
      "step": 99
    },
    {
      "epoch": 0.41841004184100417,
      "grad_norm": 0.3289240896701813,
      "learning_rate": 5.815899581589959e-06,
      "loss": 1.3754,
      "step": 100
    },
    {
      "epoch": 0.4225941422594142,
      "grad_norm": 0.3681870996952057,
      "learning_rate": 5.774058577405858e-06,
      "loss": 1.3732,
      "step": 101
    },
    {
      "epoch": 0.42677824267782427,
      "grad_norm": 0.36226436495780945,
      "learning_rate": 5.7322175732217575e-06,
      "loss": 1.3542,
      "step": 102
    },
    {
      "epoch": 0.4309623430962343,
      "grad_norm": 0.5703559517860413,
      "learning_rate": 5.690376569037658e-06,
      "loss": 1.4442,
      "step": 103
    },
    {
      "epoch": 0.4351464435146444,
      "grad_norm": 0.3629617393016815,
      "learning_rate": 5.648535564853557e-06,
      "loss": 1.3407,
      "step": 104
    },
    {
      "epoch": 0.4393305439330544,
      "grad_norm": 0.31508779525756836,
      "learning_rate": 5.6066945606694565e-06,
      "loss": 1.369,
      "step": 105
    },
    {
      "epoch": 0.4435146443514644,
      "grad_norm": 0.4534814655780792,
      "learning_rate": 5.564853556485355e-06,
      "loss": 1.3886,
      "step": 106
    },
    {
      "epoch": 0.4476987447698745,
      "grad_norm": 0.3670834004878998,
      "learning_rate": 5.523012552301256e-06,
      "loss": 1.3609,
      "step": 107
    },
    {
      "epoch": 0.45188284518828453,
      "grad_norm": 0.3328559994697571,
      "learning_rate": 5.4811715481171555e-06,
      "loss": 1.2958,
      "step": 108
    },
    {
      "epoch": 0.4560669456066946,
      "grad_norm": 0.30234208703041077,
      "learning_rate": 5.439330543933054e-06,
      "loss": 1.3321,
      "step": 109
    },
    {
      "epoch": 0.4602510460251046,
      "grad_norm": 0.3368891179561615,
      "learning_rate": 5.397489539748955e-06,
      "loss": 1.4353,
      "step": 110
    },
    {
      "epoch": 0.46443514644351463,
      "grad_norm": 0.3192000389099121,
      "learning_rate": 5.355648535564854e-06,
      "loss": 1.3775,
      "step": 111
    },
    {
      "epoch": 0.4686192468619247,
      "grad_norm": 0.4967532753944397,
      "learning_rate": 5.313807531380753e-06,
      "loss": 1.3357,
      "step": 112
    },
    {
      "epoch": 0.47280334728033474,
      "grad_norm": 0.3520500063896179,
      "learning_rate": 5.271966527196654e-06,
      "loss": 1.3523,
      "step": 113
    },
    {
      "epoch": 0.4769874476987448,
      "grad_norm": 0.3255818486213684,
      "learning_rate": 5.230125523012553e-06,
      "loss": 1.3396,
      "step": 114
    },
    {
      "epoch": 0.4811715481171548,
      "grad_norm": 0.34796151518821716,
      "learning_rate": 5.188284518828452e-06,
      "loss": 1.3115,
      "step": 115
    },
    {
      "epoch": 0.48535564853556484,
      "grad_norm": 0.3277406692504883,
      "learning_rate": 5.146443514644351e-06,
      "loss": 1.3258,
      "step": 116
    },
    {
      "epoch": 0.4895397489539749,
      "grad_norm": 0.33442819118499756,
      "learning_rate": 5.104602510460252e-06,
      "loss": 1.348,
      "step": 117
    },
    {
      "epoch": 0.49372384937238495,
      "grad_norm": 0.32950884103775024,
      "learning_rate": 5.062761506276151e-06,
      "loss": 1.3909,
      "step": 118
    },
    {
      "epoch": 0.497907949790795,
      "grad_norm": 0.33458420634269714,
      "learning_rate": 5.02092050209205e-06,
      "loss": 1.2859,
      "step": 119
    },
    {
      "epoch": 0.502092050209205,
      "grad_norm": 0.396798312664032,
      "learning_rate": 4.979079497907951e-06,
      "loss": 1.3061,
      "step": 120
    },
    {
      "epoch": 0.5062761506276151,
      "grad_norm": 0.3303851783275604,
      "learning_rate": 4.9372384937238495e-06,
      "loss": 1.3011,
      "step": 121
    },
    {
      "epoch": 0.5104602510460251,
      "grad_norm": 0.3201311528682709,
      "learning_rate": 4.895397489539749e-06,
      "loss": 1.274,
      "step": 122
    },
    {
      "epoch": 0.5146443514644351,
      "grad_norm": 0.48614901304244995,
      "learning_rate": 4.853556485355649e-06,
      "loss": 1.3494,
      "step": 123
    },
    {
      "epoch": 0.5188284518828452,
      "grad_norm": 0.31757262349128723,
      "learning_rate": 4.8117154811715485e-06,
      "loss": 1.3184,
      "step": 124
    },
    {
      "epoch": 0.5230125523012552,
      "grad_norm": 0.329109251499176,
      "learning_rate": 4.769874476987448e-06,
      "loss": 1.3895,
      "step": 125
    },
    {
      "epoch": 0.5271966527196653,
      "grad_norm": 0.5232523679733276,
      "learning_rate": 4.728033472803348e-06,
      "loss": 1.3904,
      "step": 126
    },
    {
      "epoch": 0.5313807531380753,
      "grad_norm": 0.332783043384552,
      "learning_rate": 4.6861924686192475e-06,
      "loss": 1.3277,
      "step": 127
    },
    {
      "epoch": 0.5355648535564853,
      "grad_norm": 0.31772381067276,
      "learning_rate": 4.644351464435146e-06,
      "loss": 1.3137,
      "step": 128
    },
    {
      "epoch": 0.5397489539748954,
      "grad_norm": 0.4796607792377472,
      "learning_rate": 4.602510460251046e-06,
      "loss": 1.3967,
      "step": 129
    },
    {
      "epoch": 0.5439330543933054,
      "grad_norm": 0.4687912166118622,
      "learning_rate": 4.5606694560669465e-06,
      "loss": 1.3456,
      "step": 130
    },
    {
      "epoch": 0.5481171548117155,
      "grad_norm": 0.33422547578811646,
      "learning_rate": 4.518828451882845e-06,
      "loss": 1.2768,
      "step": 131
    },
    {
      "epoch": 0.5523012552301255,
      "grad_norm": 0.3974786698818207,
      "learning_rate": 4.476987447698745e-06,
      "loss": 1.2981,
      "step": 132
    },
    {
      "epoch": 0.5564853556485355,
      "grad_norm": 0.34788477420806885,
      "learning_rate": 4.435146443514645e-06,
      "loss": 1.2914,
      "step": 133
    },
    {
      "epoch": 0.5606694560669456,
      "grad_norm": 0.3345721364021301,
      "learning_rate": 4.393305439330544e-06,
      "loss": 1.3216,
      "step": 134
    },
    {
      "epoch": 0.5648535564853556,
      "grad_norm": 0.3529014587402344,
      "learning_rate": 4.351464435146444e-06,
      "loss": 1.2673,
      "step": 135
    },
    {
      "epoch": 0.5690376569037657,
      "grad_norm": 0.32846778631210327,
      "learning_rate": 4.309623430962344e-06,
      "loss": 1.3022,
      "step": 136
    },
    {
      "epoch": 0.5732217573221757,
      "grad_norm": 0.5123084187507629,
      "learning_rate": 4.267782426778243e-06,
      "loss": 1.3141,
      "step": 137
    },
    {
      "epoch": 0.5774058577405857,
      "grad_norm": 0.3312026560306549,
      "learning_rate": 4.225941422594142e-06,
      "loss": 1.2946,
      "step": 138
    },
    {
      "epoch": 0.5815899581589958,
      "grad_norm": 0.4771478772163391,
      "learning_rate": 4.184100418410042e-06,
      "loss": 1.2487,
      "step": 139
    },
    {
      "epoch": 0.5857740585774058,
      "grad_norm": 0.38952118158340454,
      "learning_rate": 4.142259414225942e-06,
      "loss": 1.2993,
      "step": 140
    },
    {
      "epoch": 0.5899581589958159,
      "grad_norm": 0.43234747648239136,
      "learning_rate": 4.100418410041841e-06,
      "loss": 1.2749,
      "step": 141
    },
    {
      "epoch": 0.5941422594142259,
      "grad_norm": 0.5283378958702087,
      "learning_rate": 4.058577405857741e-06,
      "loss": 1.3033,
      "step": 142
    },
    {
      "epoch": 0.5983263598326359,
      "grad_norm": 0.3307763934135437,
      "learning_rate": 4.0167364016736405e-06,
      "loss": 1.289,
      "step": 143
    },
    {
      "epoch": 0.602510460251046,
      "grad_norm": 0.3551783263683319,
      "learning_rate": 3.97489539748954e-06,
      "loss": 1.3522,
      "step": 144
    },
    {
      "epoch": 0.606694560669456,
      "grad_norm": 0.35077103972435,
      "learning_rate": 3.93305439330544e-06,
      "loss": 1.3701,
      "step": 145
    },
    {
      "epoch": 0.6108786610878661,
      "grad_norm": 0.32242652773857117,
      "learning_rate": 3.8912133891213395e-06,
      "loss": 1.3106,
      "step": 146
    },
    {
      "epoch": 0.6150627615062761,
      "grad_norm": 0.33282139897346497,
      "learning_rate": 3.849372384937239e-06,
      "loss": 1.2679,
      "step": 147
    },
    {
      "epoch": 0.6192468619246861,
      "grad_norm": 0.35546955466270447,
      "learning_rate": 3.8075313807531384e-06,
      "loss": 1.256,
      "step": 148
    },
    {
      "epoch": 0.6234309623430963,
      "grad_norm": 0.3453517556190491,
      "learning_rate": 3.7656903765690376e-06,
      "loss": 1.32,
      "step": 149
    },
    {
      "epoch": 0.6276150627615062,
      "grad_norm": 0.36277034878730774,
      "learning_rate": 3.7238493723849373e-06,
      "loss": 1.3104,
      "step": 150
    },
    {
      "epoch": 0.6317991631799164,
      "grad_norm": 0.3291323482990265,
      "learning_rate": 3.6820083682008374e-06,
      "loss": 1.2807,
      "step": 151
    },
    {
      "epoch": 0.6359832635983264,
      "grad_norm": 0.5009307265281677,
      "learning_rate": 3.6401673640167366e-06,
      "loss": 1.2196,
      "step": 152
    },
    {
      "epoch": 0.6401673640167364,
      "grad_norm": 0.3647055923938751,
      "learning_rate": 3.5983263598326363e-06,
      "loss": 1.2719,
      "step": 153
    },
    {
      "epoch": 0.6443514644351465,
      "grad_norm": 0.3235703110694885,
      "learning_rate": 3.5564853556485355e-06,
      "loss": 1.2871,
      "step": 154
    },
    {
      "epoch": 0.6485355648535565,
      "grad_norm": 0.4133061170578003,
      "learning_rate": 3.514644351464435e-06,
      "loss": 1.3943,
      "step": 155
    },
    {
      "epoch": 0.6527196652719666,
      "grad_norm": 0.3607695400714874,
      "learning_rate": 3.4728033472803353e-06,
      "loss": 1.3791,
      "step": 156
    },
    {
      "epoch": 0.6569037656903766,
      "grad_norm": 0.5460565686225891,
      "learning_rate": 3.4309623430962345e-06,
      "loss": 1.2695,
      "step": 157
    },
    {
      "epoch": 0.6610878661087866,
      "grad_norm": 0.3771589994430542,
      "learning_rate": 3.389121338912134e-06,
      "loss": 1.2643,
      "step": 158
    },
    {
      "epoch": 0.6652719665271967,
      "grad_norm": 0.3366248309612274,
      "learning_rate": 3.3472803347280334e-06,
      "loss": 1.2972,
      "step": 159
    },
    {
      "epoch": 0.6694560669456067,
      "grad_norm": 0.3988908529281616,
      "learning_rate": 3.305439330543933e-06,
      "loss": 1.2929,
      "step": 160
    },
    {
      "epoch": 0.6736401673640168,
      "grad_norm": 0.339682400226593,
      "learning_rate": 3.263598326359833e-06,
      "loss": 1.3184,
      "step": 161
    },
    {
      "epoch": 0.6778242677824268,
      "grad_norm": 0.3286002576351166,
      "learning_rate": 3.2217573221757324e-06,
      "loss": 1.2801,
      "step": 162
    },
    {
      "epoch": 0.6820083682008368,
      "grad_norm": 0.33581218123435974,
      "learning_rate": 3.179916317991632e-06,
      "loss": 1.3073,
      "step": 163
    },
    {
      "epoch": 0.6861924686192469,
      "grad_norm": 0.3373326361179352,
      "learning_rate": 3.1380753138075313e-06,
      "loss": 1.2853,
      "step": 164
    },
    {
      "epoch": 0.6903765690376569,
      "grad_norm": 0.3551255464553833,
      "learning_rate": 3.096234309623431e-06,
      "loss": 1.2795,
      "step": 165
    },
    {
      "epoch": 0.694560669456067,
      "grad_norm": 0.33370694518089294,
      "learning_rate": 3.054393305439331e-06,
      "loss": 1.3062,
      "step": 166
    },
    {
      "epoch": 0.698744769874477,
      "grad_norm": 0.3263309895992279,
      "learning_rate": 3.0125523012552303e-06,
      "loss": 1.3348,
      "step": 167
    },
    {
      "epoch": 0.702928870292887,
      "grad_norm": 0.3350699245929718,
      "learning_rate": 2.97071129707113e-06,
      "loss": 1.3154,
      "step": 168
    },
    {
      "epoch": 0.7071129707112971,
      "grad_norm": 0.33460667729377747,
      "learning_rate": 2.9288702928870293e-06,
      "loss": 1.2403,
      "step": 169
    },
    {
      "epoch": 0.7112970711297071,
      "grad_norm": 0.42305517196655273,
      "learning_rate": 2.887029288702929e-06,
      "loss": 1.2679,
      "step": 170
    },
    {
      "epoch": 0.7154811715481172,
      "grad_norm": 0.5009521245956421,
      "learning_rate": 2.845188284518829e-06,
      "loss": 1.2979,
      "step": 171
    },
    {
      "epoch": 0.7196652719665272,
      "grad_norm": 0.35044097900390625,
      "learning_rate": 2.8033472803347283e-06,
      "loss": 1.2417,
      "step": 172
    },
    {
      "epoch": 0.7238493723849372,
      "grad_norm": 0.3329603672027588,
      "learning_rate": 2.761506276150628e-06,
      "loss": 1.2727,
      "step": 173
    },
    {
      "epoch": 0.7280334728033473,
      "grad_norm": 0.553231954574585,
      "learning_rate": 2.719665271966527e-06,
      "loss": 1.2906,
      "step": 174
    },
    {
      "epoch": 0.7322175732217573,
      "grad_norm": 0.3310469686985016,
      "learning_rate": 2.677824267782427e-06,
      "loss": 1.2466,
      "step": 175
    },
    {
      "epoch": 0.7364016736401674,
      "grad_norm": 0.3337252736091614,
      "learning_rate": 2.635983263598327e-06,
      "loss": 1.2753,
      "step": 176
    },
    {
      "epoch": 0.7405857740585774,
      "grad_norm": 0.5241883993148804,
      "learning_rate": 2.594142259414226e-06,
      "loss": 1.3364,
      "step": 177
    },
    {
      "epoch": 0.7447698744769874,
      "grad_norm": 0.3621697723865509,
      "learning_rate": 2.552301255230126e-06,
      "loss": 1.2091,
      "step": 178
    },
    {
      "epoch": 0.7489539748953975,
      "grad_norm": 0.40887123346328735,
      "learning_rate": 2.510460251046025e-06,
      "loss": 1.3031,
      "step": 179
    },
    {
      "epoch": 0.7531380753138075,
      "grad_norm": 0.34628698229789734,
      "learning_rate": 2.4686192468619247e-06,
      "loss": 1.2093,
      "step": 180
    },
    {
      "epoch": 0.7573221757322176,
      "grad_norm": 0.3536050617694855,
      "learning_rate": 2.4267782426778244e-06,
      "loss": 1.3339,
      "step": 181
    },
    {
      "epoch": 0.7615062761506276,
      "grad_norm": 0.44155943393707275,
      "learning_rate": 2.384937238493724e-06,
      "loss": 1.3256,
      "step": 182
    },
    {
      "epoch": 0.7656903765690377,
      "grad_norm": 0.3510695695877075,
      "learning_rate": 2.3430962343096237e-06,
      "loss": 1.246,
      "step": 183
    },
    {
      "epoch": 0.7698744769874477,
      "grad_norm": 0.39234158396720886,
      "learning_rate": 2.301255230125523e-06,
      "loss": 1.1959,
      "step": 184
    },
    {
      "epoch": 0.7740585774058577,
      "grad_norm": 0.38711047172546387,
      "learning_rate": 2.2594142259414227e-06,
      "loss": 1.2806,
      "step": 185
    },
    {
      "epoch": 0.7782426778242678,
      "grad_norm": 0.34728190302848816,
      "learning_rate": 2.2175732217573223e-06,
      "loss": 1.2284,
      "step": 186
    },
    {
      "epoch": 0.7824267782426778,
      "grad_norm": 0.4292190372943878,
      "learning_rate": 2.175732217573222e-06,
      "loss": 1.317,
      "step": 187
    },
    {
      "epoch": 0.7866108786610879,
      "grad_norm": 0.3527109920978546,
      "learning_rate": 2.1338912133891217e-06,
      "loss": 1.2481,
      "step": 188
    },
    {
      "epoch": 0.7907949790794979,
      "grad_norm": 0.41887742280960083,
      "learning_rate": 2.092050209205021e-06,
      "loss": 1.2332,
      "step": 189
    },
    {
      "epoch": 0.7949790794979079,
      "grad_norm": 0.40693023800849915,
      "learning_rate": 2.0502092050209206e-06,
      "loss": 1.2143,
      "step": 190
    },
    {
      "epoch": 0.799163179916318,
      "grad_norm": 0.33459046483039856,
      "learning_rate": 2.0083682008368202e-06,
      "loss": 1.2407,
      "step": 191
    },
    {
      "epoch": 0.803347280334728,
      "grad_norm": 0.3630668520927429,
      "learning_rate": 1.96652719665272e-06,
      "loss": 1.2559,
      "step": 192
    },
    {
      "epoch": 0.8075313807531381,
      "grad_norm": 0.4772534668445587,
      "learning_rate": 1.9246861924686196e-06,
      "loss": 1.2456,
      "step": 193
    },
    {
      "epoch": 0.8117154811715481,
      "grad_norm": 0.36172667145729065,
      "learning_rate": 1.8828451882845188e-06,
      "loss": 1.2749,
      "step": 194
    },
    {
      "epoch": 0.8158995815899581,
      "grad_norm": 0.37000811100006104,
      "learning_rate": 1.8410041841004187e-06,
      "loss": 1.2395,
      "step": 195
    },
    {
      "epoch": 0.8200836820083682,
      "grad_norm": 0.49782806634902954,
      "learning_rate": 1.7991631799163181e-06,
      "loss": 1.2788,
      "step": 196
    },
    {
      "epoch": 0.8242677824267782,
      "grad_norm": 0.3391241133213043,
      "learning_rate": 1.7573221757322176e-06,
      "loss": 1.2772,
      "step": 197
    },
    {
      "epoch": 0.8284518828451883,
      "grad_norm": 0.3950854539871216,
      "learning_rate": 1.7154811715481173e-06,
      "loss": 1.231,
      "step": 198
    },
    {
      "epoch": 0.8326359832635983,
      "grad_norm": 0.4793711304664612,
      "learning_rate": 1.6736401673640167e-06,
      "loss": 1.248,
      "step": 199
    },
    {
      "epoch": 0.8368200836820083,
      "grad_norm": 0.51435387134552,
      "learning_rate": 1.6317991631799166e-06,
      "loss": 1.2203,
      "step": 200
    },
    {
      "epoch": 0.8410041841004184,
      "grad_norm": 0.6365242600440979,
      "learning_rate": 1.589958158995816e-06,
      "loss": 1.3182,
      "step": 201
    },
    {
      "epoch": 0.8451882845188284,
      "grad_norm": 0.3626181483268738,
      "learning_rate": 1.5481171548117155e-06,
      "loss": 1.286,
      "step": 202
    },
    {
      "epoch": 0.8493723849372385,
      "grad_norm": 0.5900272130966187,
      "learning_rate": 1.5062761506276152e-06,
      "loss": 1.2312,
      "step": 203
    },
    {
      "epoch": 0.8535564853556485,
      "grad_norm": 0.3442046642303467,
      "learning_rate": 1.4644351464435146e-06,
      "loss": 1.2708,
      "step": 204
    },
    {
      "epoch": 0.8577405857740585,
      "grad_norm": 0.3256811797618866,
      "learning_rate": 1.4225941422594145e-06,
      "loss": 1.2402,
      "step": 205
    },
    {
      "epoch": 0.8619246861924686,
      "grad_norm": 0.3342282474040985,
      "learning_rate": 1.380753138075314e-06,
      "loss": 1.2461,
      "step": 206
    },
    {
      "epoch": 0.8661087866108786,
      "grad_norm": 0.33777275681495667,
      "learning_rate": 1.3389121338912134e-06,
      "loss": 1.2926,
      "step": 207
    },
    {
      "epoch": 0.8702928870292888,
      "grad_norm": 0.3379373848438263,
      "learning_rate": 1.297071129707113e-06,
      "loss": 1.2772,
      "step": 208
    },
    {
      "epoch": 0.8744769874476988,
      "grad_norm": 0.6433287262916565,
      "learning_rate": 1.2552301255230125e-06,
      "loss": 1.2669,
      "step": 209
    },
    {
      "epoch": 0.8786610878661087,
      "grad_norm": 0.6349135041236877,
      "learning_rate": 1.2133891213389122e-06,
      "loss": 1.305,
      "step": 210
    },
    {
      "epoch": 0.8828451882845189,
      "grad_norm": 0.33414891362190247,
      "learning_rate": 1.1715481171548119e-06,
      "loss": 1.2529,
      "step": 211
    },
    {
      "epoch": 0.8870292887029289,
      "grad_norm": 0.3989071547985077,
      "learning_rate": 1.1297071129707113e-06,
      "loss": 1.2506,
      "step": 212
    },
    {
      "epoch": 0.891213389121339,
      "grad_norm": 0.3862692415714264,
      "learning_rate": 1.087866108786611e-06,
      "loss": 1.3046,
      "step": 213
    },
    {
      "epoch": 0.895397489539749,
      "grad_norm": 0.34327417612075806,
      "learning_rate": 1.0460251046025104e-06,
      "loss": 1.227,
      "step": 214
    },
    {
      "epoch": 0.899581589958159,
      "grad_norm": 0.3878503441810608,
      "learning_rate": 1.0041841004184101e-06,
      "loss": 1.2668,
      "step": 215
    },
    {
      "epoch": 0.9037656903765691,
      "grad_norm": 0.32297104597091675,
      "learning_rate": 9.623430962343098e-07,
      "loss": 1.2597,
      "step": 216
    },
    {
      "epoch": 0.9079497907949791,
      "grad_norm": 0.44788116216659546,
      "learning_rate": 9.205020920502093e-07,
      "loss": 1.3145,
      "step": 217
    },
    {
      "epoch": 0.9121338912133892,
      "grad_norm": 0.35166341066360474,
      "learning_rate": 8.786610878661088e-07,
      "loss": 1.2948,
      "step": 218
    },
    {
      "epoch": 0.9163179916317992,
      "grad_norm": 0.3332233428955078,
      "learning_rate": 8.368200836820084e-07,
      "loss": 1.2554,
      "step": 219
    },
    {
      "epoch": 0.9205020920502092,
      "grad_norm": 0.3346772789955139,
      "learning_rate": 7.94979079497908e-07,
      "loss": 1.2279,
      "step": 220
    },
    {
      "epoch": 0.9246861924686193,
      "grad_norm": 0.3380833864212036,
      "learning_rate": 7.531380753138076e-07,
      "loss": 1.2176,
      "step": 221
    },
    {
      "epoch": 0.9288702928870293,
      "grad_norm": 0.5502627491950989,
      "learning_rate": 7.112970711297073e-07,
      "loss": 1.2271,
      "step": 222
    },
    {
      "epoch": 0.9330543933054394,
      "grad_norm": 0.3738647401332855,
      "learning_rate": 6.694560669456067e-07,
      "loss": 1.3101,
      "step": 223
    },
    {
      "epoch": 0.9372384937238494,
      "grad_norm": 0.4273400902748108,
      "learning_rate": 6.276150627615063e-07,
      "loss": 1.2648,
      "step": 224
    },
    {
      "epoch": 0.9414225941422594,
      "grad_norm": 0.5694935917854309,
      "learning_rate": 5.857740585774059e-07,
      "loss": 1.301,
      "step": 225
    },
    {
      "epoch": 0.9456066945606695,
      "grad_norm": 0.37318360805511475,
      "learning_rate": 5.439330543933055e-07,
      "loss": 1.2353,
      "step": 226
    },
    {
      "epoch": 0.9497907949790795,
      "grad_norm": 0.37486758828163147,
      "learning_rate": 5.020920502092051e-07,
      "loss": 1.2324,
      "step": 227
    },
    {
      "epoch": 0.9539748953974896,
      "grad_norm": 0.342326283454895,
      "learning_rate": 4.6025104602510467e-07,
      "loss": 1.265,
      "step": 228
    },
    {
      "epoch": 0.9581589958158996,
      "grad_norm": 0.3409275710582733,
      "learning_rate": 4.184100418410042e-07,
      "loss": 1.2498,
      "step": 229
    },
    {
      "epoch": 0.9623430962343096,
      "grad_norm": 0.41361624002456665,
      "learning_rate": 3.765690376569038e-07,
      "loss": 1.243,
      "step": 230
    },
    {
      "epoch": 0.9665271966527197,
      "grad_norm": 0.3393700122833252,
      "learning_rate": 3.3472803347280335e-07,
      "loss": 1.2796,
      "step": 231
    },
    {
      "epoch": 0.9707112970711297,
      "grad_norm": 0.3456774055957794,
      "learning_rate": 2.9288702928870297e-07,
      "loss": 1.3252,
      "step": 232
    },
    {
      "epoch": 0.9748953974895398,
      "grad_norm": 0.5081372261047363,
      "learning_rate": 2.5104602510460253e-07,
      "loss": 1.2795,
      "step": 233
    },
    {
      "epoch": 0.9790794979079498,
      "grad_norm": 0.35145166516304016,
      "learning_rate": 2.092050209205021e-07,
      "loss": 1.2994,
      "step": 234
    },
    {
      "epoch": 0.9832635983263598,
      "grad_norm": 0.40834537148475647,
      "learning_rate": 1.6736401673640168e-07,
      "loss": 1.2504,
      "step": 235
    },
    {
      "epoch": 0.9874476987447699,
      "grad_norm": 0.338552862405777,
      "learning_rate": 1.2552301255230126e-07,
      "loss": 1.2258,
      "step": 236
    },
    {
      "epoch": 0.9916317991631799,
      "grad_norm": 0.35948339104652405,
      "learning_rate": 8.368200836820084e-08,
      "loss": 1.2088,
      "step": 237
    },
    {
      "epoch": 0.99581589958159,
      "grad_norm": 0.36100009083747864,
      "learning_rate": 4.184100418410042e-08,
      "loss": 1.1964,
      "step": 238
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.3361973166465759,
      "learning_rate": 0.0,
      "loss": 1.2125,
      "step": 239
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 239,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 0,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8.011575614609818e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}