{
  "best_global_step": null,
  "best_metric": 0.2180466502904892,
  "best_model_checkpoint": null,
  "epoch": 5.540723217303143,
  "eval_steps": 50,
  "global_step": 8200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0006759040216289286,
      "grad_norm": 2070.484130859375,
      "learning_rate": 0.0,
      "loss": 6.6982,
      "step": 1
    },
    {
      "epoch": 0.0013518080432578573,
      "grad_norm": 2034.6072998046875,
      "learning_rate": 4.504504504504505e-08,
      "loss": 7.1438,
      "step": 2
    },
    {
      "epoch": 0.002027712064886786,
      "grad_norm": 2172.909423828125,
      "learning_rate": 9.00900900900901e-08,
      "loss": 7.7546,
      "step": 3
    },
    {
      "epoch": 0.0027036160865157146,
      "grad_norm": 2017.2457275390625,
      "learning_rate": 1.3513513513513515e-07,
      "loss": 7.1674,
      "step": 4
    },
    {
      "epoch": 0.0033795201081446434,
      "grad_norm": 2135.654296875,
      "learning_rate": 1.801801801801802e-07,
      "loss": 6.9808,
      "step": 5
    },
    {
      "epoch": 0.004055424129773572,
      "grad_norm": 2085.8740234375,
      "learning_rate": 2.2522522522522524e-07,
      "loss": 7.0421,
      "step": 6
    },
    {
      "epoch": 0.004731328151402501,
      "grad_norm": 2102.819580078125,
      "learning_rate": 2.702702702702703e-07,
      "loss": 7.054,
      "step": 7
    },
    {
      "epoch": 0.005407232173031429,
      "grad_norm": 1950.758544921875,
      "learning_rate": 3.153153153153153e-07,
      "loss": 6.237,
      "step": 8
    },
    {
      "epoch": 0.006083136194660358,
      "grad_norm": 2137.362060546875,
      "learning_rate": 3.603603603603604e-07,
      "loss": 6.3447,
      "step": 9
    },
    {
      "epoch": 0.006759040216289287,
      "grad_norm": 2146.540283203125,
      "learning_rate": 4.0540540540540546e-07,
      "loss": 6.6189,
      "step": 10
    },
    {
      "epoch": 0.007434944237918215,
      "grad_norm": 1626.2615966796875,
      "learning_rate": 4.504504504504505e-07,
      "loss": 5.2975,
      "step": 11
    },
    {
      "epoch": 0.008110848259547145,
      "grad_norm": 1541.1419677734375,
      "learning_rate": 4.954954954954956e-07,
      "loss": 5.1497,
      "step": 12
    },
    {
      "epoch": 0.008786752281176074,
      "grad_norm": 1394.8128662109375,
      "learning_rate": 5.405405405405406e-07,
      "loss": 4.8215,
      "step": 13
    },
    {
      "epoch": 0.009462656302805001,
      "grad_norm": 1299.58056640625,
      "learning_rate": 5.855855855855856e-07,
      "loss": 4.7954,
      "step": 14
    },
    {
      "epoch": 0.01013856032443393,
      "grad_norm": 711.5867309570312,
      "learning_rate": 6.306306306306306e-07,
      "loss": 2.7501,
      "step": 15
    },
    {
      "epoch": 0.010814464346062858,
      "grad_norm": 741.6774291992188,
      "learning_rate": 6.756756756756758e-07,
      "loss": 2.8943,
      "step": 16
    },
    {
      "epoch": 0.011490368367691788,
      "grad_norm": 778.3511962890625,
      "learning_rate": 7.207207207207208e-07,
      "loss": 3.0308,
      "step": 17
    },
    {
      "epoch": 0.012166272389320717,
      "grad_norm": 767.487548828125,
      "learning_rate": 7.657657657657658e-07,
      "loss": 2.7969,
      "step": 18
    },
    {
      "epoch": 0.012842176410949644,
      "grad_norm": 742.028076171875,
      "learning_rate": 8.108108108108109e-07,
      "loss": 2.6171,
      "step": 19
    },
    {
      "epoch": 0.013518080432578574,
      "grad_norm": 544.3399658203125,
      "learning_rate": 8.55855855855856e-07,
      "loss": 1.4595,
      "step": 20
    },
    {
      "epoch": 0.014193984454207503,
      "grad_norm": 187.1193084716797,
      "learning_rate": 9.00900900900901e-07,
      "loss": 0.578,
      "step": 21
    },
    {
      "epoch": 0.01486988847583643,
      "grad_norm": 158.82821655273438,
      "learning_rate": 9.459459459459461e-07,
      "loss": 0.5568,
      "step": 22
    },
    {
      "epoch": 0.01554579249746536,
      "grad_norm": 101.81764221191406,
      "learning_rate": 9.909909909909911e-07,
      "loss": 0.4672,
      "step": 23
    },
    {
      "epoch": 0.01622169651909429,
      "grad_norm": 94.74485778808594,
      "learning_rate": 1.0360360360360361e-06,
      "loss": 0.4319,
      "step": 24
    },
    {
      "epoch": 0.01689760054072322,
      "grad_norm": 151.90870666503906,
      "learning_rate": 1.0810810810810812e-06,
      "loss": 0.505,
      "step": 25
    },
    {
      "epoch": 0.017573504562352148,
      "grad_norm": 105.22148895263672,
      "learning_rate": 1.1261261261261262e-06,
      "loss": 0.4546,
      "step": 26
    },
    {
      "epoch": 0.018249408583981074,
      "grad_norm": 139.1234130859375,
      "learning_rate": 1.1711711711711712e-06,
      "loss": 0.4765,
      "step": 27
    },
    {
      "epoch": 0.018925312605610003,
      "grad_norm": 40.476951599121094,
      "learning_rate": 1.2162162162162164e-06,
      "loss": 0.3601,
      "step": 28
    },
    {
      "epoch": 0.019601216627238932,
      "grad_norm": 23.708459854125977,
      "learning_rate": 1.2612612612612613e-06,
      "loss": 0.362,
      "step": 29
    },
    {
      "epoch": 0.02027712064886786,
      "grad_norm": 42.72507095336914,
      "learning_rate": 1.3063063063063065e-06,
      "loss": 0.3763,
      "step": 30
    },
    {
      "epoch": 0.02095302467049679,
      "grad_norm": 60.79398727416992,
      "learning_rate": 1.3513513513513515e-06,
      "loss": 0.348,
      "step": 31
    },
    {
      "epoch": 0.021628928692125717,
      "grad_norm": 12.653668403625488,
      "learning_rate": 1.3963963963963963e-06,
      "loss": 0.3467,
      "step": 32
    },
    {
      "epoch": 0.022304832713754646,
      "grad_norm": 19.537141799926758,
      "learning_rate": 1.4414414414414416e-06,
      "loss": 0.3351,
      "step": 33
    },
    {
      "epoch": 0.022980736735383575,
      "grad_norm": 11.202260971069336,
      "learning_rate": 1.4864864864864868e-06,
      "loss": 0.3369,
      "step": 34
    },
    {
      "epoch": 0.023656640757012504,
      "grad_norm": 35.051918029785156,
      "learning_rate": 1.5315315315315316e-06,
      "loss": 0.3386,
      "step": 35
    },
    {
      "epoch": 0.024332544778641434,
      "grad_norm": 21.535572052001953,
      "learning_rate": 1.5765765765765766e-06,
      "loss": 0.3222,
      "step": 36
    },
    {
      "epoch": 0.025008448800270363,
      "grad_norm": 23.429094314575195,
      "learning_rate": 1.6216216216216219e-06,
      "loss": 0.3453,
      "step": 37
    },
    {
      "epoch": 0.02568435282189929,
      "grad_norm": 19.155500411987305,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 0.3481,
      "step": 38
    },
    {
      "epoch": 0.026360256843528218,
      "grad_norm": 36.207061767578125,
      "learning_rate": 1.711711711711712e-06,
      "loss": 0.318,
      "step": 39
    },
    {
      "epoch": 0.027036160865157147,
      "grad_norm": 15.088589668273926,
      "learning_rate": 1.756756756756757e-06,
      "loss": 0.331,
      "step": 40
    },
    {
      "epoch": 0.027712064886786077,
      "grad_norm": 30.023080825805664,
      "learning_rate": 1.801801801801802e-06,
      "loss": 0.306,
      "step": 41
    },
    {
      "epoch": 0.028387968908415006,
      "grad_norm": 13.055476188659668,
      "learning_rate": 1.846846846846847e-06,
      "loss": 0.3331,
      "step": 42
    },
    {
      "epoch": 0.029063872930043935,
      "grad_norm": 25.66547203063965,
      "learning_rate": 1.8918918918918922e-06,
      "loss": 0.3052,
      "step": 43
    },
    {
      "epoch": 0.02973977695167286,
      "grad_norm": 14.33441162109375,
      "learning_rate": 1.9369369369369372e-06,
      "loss": 0.3165,
      "step": 44
    },
    {
      "epoch": 0.03041568097330179,
      "grad_norm": 25.875953674316406,
      "learning_rate": 1.9819819819819822e-06,
      "loss": 0.3234,
      "step": 45
    },
    {
      "epoch": 0.03109158499493072,
      "grad_norm": 30.295759201049805,
      "learning_rate": 2.0270270270270273e-06,
      "loss": 0.3328,
      "step": 46
    },
    {
      "epoch": 0.031767489016559645,
      "grad_norm": 7.533182621002197,
      "learning_rate": 2.0720720720720723e-06,
      "loss": 0.2972,
      "step": 47
    },
    {
      "epoch": 0.03244339303818858,
      "grad_norm": 56.362918853759766,
      "learning_rate": 2.1171171171171173e-06,
      "loss": 0.3488,
      "step": 48
    },
    {
      "epoch": 0.033119297059817504,
      "grad_norm": 10.13399887084961,
      "learning_rate": 2.1621621621621623e-06,
      "loss": 0.3154,
      "step": 49
    },
    {
      "epoch": 0.03379520108144644,
      "grad_norm": 11.681538581848145,
      "learning_rate": 2.2072072072072073e-06,
      "loss": 0.3122,
      "step": 50
    },
    {
      "epoch": 0.03379520108144644,
      "eval_accuracy": 0.26127533783783785,
      "eval_loss": 0.32419976592063904,
      "eval_runtime": 43.6361,
      "eval_samples_per_second": 542.67,
      "eval_steps_per_second": 16.958,
      "step": 50
    },
    {
      "epoch": 0.03447110510307536,
      "grad_norm": 11.827308654785156,
      "learning_rate": 2.2522522522522524e-06,
      "loss": 0.3353,
      "step": 51
    },
    {
      "epoch": 0.035147009124704295,
      "grad_norm": 32.8586311340332,
      "learning_rate": 2.297297297297298e-06,
      "loss": 0.3332,
      "step": 52
    },
    {
      "epoch": 0.03582291314633322,
      "grad_norm": 33.08034133911133,
      "learning_rate": 2.3423423423423424e-06,
      "loss": 0.3342,
      "step": 53
    },
    {
      "epoch": 0.03649881716796215,
      "grad_norm": 19.520170211791992,
      "learning_rate": 2.3873873873873874e-06,
      "loss": 0.3357,
      "step": 54
    },
    {
      "epoch": 0.03717472118959108,
      "grad_norm": 30.17961883544922,
      "learning_rate": 2.432432432432433e-06,
      "loss": 0.325,
      "step": 55
    },
    {
      "epoch": 0.037850625211220006,
      "grad_norm": 11.146134376525879,
      "learning_rate": 2.4774774774774775e-06,
      "loss": 0.3384,
      "step": 56
    },
    {
      "epoch": 0.03852652923284894,
      "grad_norm": 9.181519508361816,
      "learning_rate": 2.5225225225225225e-06,
      "loss": 0.33,
      "step": 57
    },
    {
      "epoch": 0.039202433254477864,
      "grad_norm": 6.26019287109375,
      "learning_rate": 2.5675675675675675e-06,
      "loss": 0.3201,
      "step": 58
    },
    {
      "epoch": 0.03987833727610679,
      "grad_norm": 32.51396942138672,
      "learning_rate": 2.612612612612613e-06,
      "loss": 0.3165,
      "step": 59
    },
    {
      "epoch": 0.04055424129773572,
      "grad_norm": 19.94057273864746,
      "learning_rate": 2.657657657657658e-06,
      "loss": 0.3147,
      "step": 60
    },
    {
      "epoch": 0.04123014531936465,
      "grad_norm": 20.111881256103516,
      "learning_rate": 2.702702702702703e-06,
      "loss": 0.3143,
      "step": 61
    },
    {
      "epoch": 0.04190604934099358,
      "grad_norm": 37.779022216796875,
      "learning_rate": 2.747747747747748e-06,
      "loss": 0.3274,
      "step": 62
    },
    {
      "epoch": 0.04258195336262251,
      "grad_norm": 38.85224533081055,
      "learning_rate": 2.7927927927927926e-06,
      "loss": 0.3426,
      "step": 63
    },
    {
      "epoch": 0.04325785738425143,
      "grad_norm": 27.557212829589844,
      "learning_rate": 2.837837837837838e-06,
      "loss": 0.3052,
      "step": 64
    },
    {
      "epoch": 0.043933761405880366,
      "grad_norm": 50.85004425048828,
      "learning_rate": 2.882882882882883e-06,
      "loss": 0.3093,
      "step": 65
    },
    {
      "epoch": 0.04460966542750929,
      "grad_norm": 14.456063270568848,
      "learning_rate": 2.927927927927928e-06,
      "loss": 0.3087,
      "step": 66
    },
    {
      "epoch": 0.045285569449138224,
      "grad_norm": 81.76181030273438,
      "learning_rate": 2.9729729729729736e-06,
      "loss": 0.3706,
      "step": 67
    },
    {
      "epoch": 0.04596147347076715,
      "grad_norm": 7.055322647094727,
      "learning_rate": 3.0180180180180186e-06,
      "loss": 0.2948,
      "step": 68
    },
    {
      "epoch": 0.04663737749239608,
      "grad_norm": 21.952960968017578,
      "learning_rate": 3.063063063063063e-06,
      "loss": 0.3255,
      "step": 69
    },
    {
      "epoch": 0.04731328151402501,
      "grad_norm": 37.055091857910156,
      "learning_rate": 3.1081081081081082e-06,
      "loss": 0.3137,
      "step": 70
    },
    {
      "epoch": 0.047989185535653935,
      "grad_norm": 9.407421112060547,
      "learning_rate": 3.1531531531531532e-06,
      "loss": 0.3326,
      "step": 71
    },
    {
      "epoch": 0.04866508955728287,
      "grad_norm": 39.629058837890625,
      "learning_rate": 3.1981981981981987e-06,
      "loss": 0.3188,
      "step": 72
    },
    {
      "epoch": 0.04934099357891179,
      "grad_norm": 47.593894958496094,
      "learning_rate": 3.2432432432432437e-06,
      "loss": 0.3382,
      "step": 73
    },
    {
      "epoch": 0.050016897600540726,
      "grad_norm": 11.171440124511719,
      "learning_rate": 3.2882882882882887e-06,
      "loss": 0.3234,
      "step": 74
    },
    {
      "epoch": 0.05069280162216965,
      "grad_norm": 7.109539031982422,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.3032,
      "step": 75
    },
    {
      "epoch": 0.05136870564379858,
      "grad_norm": 50.866188049316406,
      "learning_rate": 3.3783783783783788e-06,
      "loss": 0.3334,
      "step": 76
    },
    {
      "epoch": 0.05204460966542751,
      "grad_norm": 63.787696838378906,
      "learning_rate": 3.423423423423424e-06,
      "loss": 0.3516,
      "step": 77
    },
    {
      "epoch": 0.052720513687056436,
      "grad_norm": 24.337238311767578,
      "learning_rate": 3.468468468468469e-06,
      "loss": 0.3081,
      "step": 78
    },
    {
      "epoch": 0.05339641770868537,
      "grad_norm": 34.554840087890625,
      "learning_rate": 3.513513513513514e-06,
      "loss": 0.2975,
      "step": 79
    },
    {
      "epoch": 0.054072321730314295,
      "grad_norm": 36.74021911621094,
      "learning_rate": 3.5585585585585584e-06,
      "loss": 0.3272,
      "step": 80
    },
    {
      "epoch": 0.05474822575194323,
      "grad_norm": 26.778545379638672,
      "learning_rate": 3.603603603603604e-06,
      "loss": 0.3064,
      "step": 81
    },
    {
      "epoch": 0.05542412977357215,
      "grad_norm": 49.57334518432617,
      "learning_rate": 3.648648648648649e-06,
      "loss": 0.3442,
      "step": 82
    },
    {
      "epoch": 0.05610003379520108,
      "grad_norm": 9.214241981506348,
      "learning_rate": 3.693693693693694e-06,
      "loss": 0.32,
      "step": 83
    },
    {
      "epoch": 0.05677593781683001,
      "grad_norm": 18.205781936645508,
      "learning_rate": 3.7387387387387394e-06,
      "loss": 0.3353,
      "step": 84
    },
    {
      "epoch": 0.05745184183845894,
      "grad_norm": 37.84554672241211,
      "learning_rate": 3.7837837837837844e-06,
      "loss": 0.3369,
      "step": 85
    },
    {
      "epoch": 0.05812774586008787,
      "grad_norm": 54.41448211669922,
      "learning_rate": 3.828828828828829e-06,
      "loss": 0.2989,
      "step": 86
    },
    {
      "epoch": 0.058803649881716796,
      "grad_norm": 42.44352340698242,
      "learning_rate": 3.8738738738738744e-06,
      "loss": 0.3246,
      "step": 87
    },
    {
      "epoch": 0.05947955390334572,
      "grad_norm": 24.652870178222656,
      "learning_rate": 3.918918918918919e-06,
      "loss": 0.3268,
      "step": 88
    },
    {
      "epoch": 0.060155457924974655,
      "grad_norm": 43.99015808105469,
      "learning_rate": 3.9639639639639645e-06,
      "loss": 0.3097,
      "step": 89
    },
    {
      "epoch": 0.06083136194660358,
      "grad_norm": 23.306865692138672,
      "learning_rate": 4.009009009009009e-06,
      "loss": 0.325,
      "step": 90
    },
    {
      "epoch": 0.06150726596823251,
      "grad_norm": 18.154720306396484,
      "learning_rate": 4.0540540540540545e-06,
      "loss": 0.3124,
      "step": 91
    },
    {
      "epoch": 0.06218316998986144,
      "grad_norm": 6.390261650085449,
      "learning_rate": 4.099099099099099e-06,
      "loss": 0.3087,
      "step": 92
    },
    {
      "epoch": 0.06285907401149037,
      "grad_norm": 28.85924530029297,
      "learning_rate": 4.1441441441441446e-06,
      "loss": 0.3211,
      "step": 93
    },
    {
      "epoch": 0.06353497803311929,
      "grad_norm": 10.693065643310547,
      "learning_rate": 4.189189189189189e-06,
      "loss": 0.3031,
      "step": 94
    },
    {
      "epoch": 0.06421088205474823,
      "grad_norm": 43.638511657714844,
      "learning_rate": 4.234234234234235e-06,
      "loss": 0.2896,
      "step": 95
    },
    {
      "epoch": 0.06488678607637716,
      "grad_norm": 6.0606770515441895,
      "learning_rate": 4.27927927927928e-06,
      "loss": 0.3087,
      "step": 96
    },
    {
      "epoch": 0.06556269009800608,
      "grad_norm": 15.399177551269531,
      "learning_rate": 4.324324324324325e-06,
      "loss": 0.3051,
      "step": 97
    },
    {
      "epoch": 0.06623859411963501,
      "grad_norm": 69.56488800048828,
      "learning_rate": 4.369369369369369e-06,
      "loss": 0.3442,
      "step": 98
    },
    {
      "epoch": 0.06691449814126393,
      "grad_norm": 17.781269073486328,
      "learning_rate": 4.414414414414415e-06,
      "loss": 0.3165,
      "step": 99
    },
    {
      "epoch": 0.06759040216289287,
      "grad_norm": 50.67172622680664,
      "learning_rate": 4.45945945945946e-06,
      "loss": 0.305,
      "step": 100
    },
    {
      "epoch": 0.06759040216289287,
      "eval_accuracy": 0.22132601351351353,
      "eval_loss": 0.32813921570777893,
      "eval_runtime": 44.1801,
      "eval_samples_per_second": 535.988,
      "eval_steps_per_second": 16.75,
      "step": 100
    },
    {
      "epoch": 0.0682663061845218,
      "grad_norm": 40.05355453491211,
      "learning_rate": 4.504504504504505e-06,
      "loss": 0.3151,
      "step": 101
    },
    {
      "epoch": 0.06894221020615073,
      "grad_norm": 38.45130920410156,
      "learning_rate": 4.54954954954955e-06,
      "loss": 0.323,
      "step": 102
    },
    {
      "epoch": 0.06961811422777965,
      "grad_norm": 15.057827949523926,
      "learning_rate": 4.594594594594596e-06,
      "loss": 0.3015,
      "step": 103
    },
    {
      "epoch": 0.07029401824940859,
      "grad_norm": 18.441415786743164,
      "learning_rate": 4.63963963963964e-06,
      "loss": 0.2929,
      "step": 104
    },
    {
      "epoch": 0.07096992227103752,
      "grad_norm": 30.101444244384766,
      "learning_rate": 4.684684684684685e-06,
      "loss": 0.302,
      "step": 105
    },
    {
      "epoch": 0.07164582629266644,
      "grad_norm": 55.68412780761719,
      "learning_rate": 4.72972972972973e-06,
      "loss": 0.3334,
      "step": 106
    },
    {
      "epoch": 0.07232173031429537,
      "grad_norm": 15.319828033447266,
      "learning_rate": 4.774774774774775e-06,
      "loss": 0.3261,
      "step": 107
    },
    {
      "epoch": 0.0729976343359243,
      "grad_norm": 23.075336456298828,
      "learning_rate": 4.81981981981982e-06,
      "loss": 0.3075,
      "step": 108
    },
    {
      "epoch": 0.07367353835755323,
      "grad_norm": 88.35043334960938,
      "learning_rate": 4.864864864864866e-06,
      "loss": 0.3665,
      "step": 109
    },
    {
      "epoch": 0.07434944237918216,
      "grad_norm": 45.1597785949707,
      "learning_rate": 4.90990990990991e-06,
      "loss": 0.3397,
      "step": 110
    },
    {
      "epoch": 0.07502534640081109,
      "grad_norm": 10.711393356323242,
      "learning_rate": 4.954954954954955e-06,
      "loss": 0.3066,
      "step": 111
    },
    {
      "epoch": 0.07570125042244001,
      "grad_norm": 64.00765991210938,
      "learning_rate": 5e-06,
      "loss": 0.3303,
      "step": 112
    },
    {
      "epoch": 0.07637715444406894,
      "grad_norm": 144.59857177734375,
      "learning_rate": 5.045045045045045e-06,
      "loss": 0.4314,
      "step": 113
    },
    {
      "epoch": 0.07705305846569788,
      "grad_norm": 14.112588882446289,
      "learning_rate": 5.0900900900900905e-06,
      "loss": 0.3281,
      "step": 114
    },
    {
      "epoch": 0.0777289624873268,
      "grad_norm": 101.94636535644531,
      "learning_rate": 5.135135135135135e-06,
      "loss": 0.3959,
      "step": 115
    },
    {
      "epoch": 0.07840486650895573,
      "grad_norm": 98.03368377685547,
      "learning_rate": 5.180180180180181e-06,
      "loss": 0.4091,
      "step": 116
    },
    {
      "epoch": 0.07908077053058465,
      "grad_norm": 86.66287231445312,
      "learning_rate": 5.225225225225226e-06,
      "loss": 0.3586,
      "step": 117
    },
    {
      "epoch": 0.07975667455221358,
      "grad_norm": 16.383543014526367,
      "learning_rate": 5.2702702702702705e-06,
      "loss": 0.3069,
      "step": 118
    },
    {
      "epoch": 0.08043257857384252,
      "grad_norm": 104.17880249023438,
      "learning_rate": 5.315315315315316e-06,
      "loss": 0.4096,
      "step": 119
    },
    {
      "epoch": 0.08110848259547145,
      "grad_norm": 108.52730560302734,
      "learning_rate": 5.360360360360361e-06,
      "loss": 0.4403,
      "step": 120
    },
    {
      "epoch": 0.08178438661710037,
      "grad_norm": 49.37046432495117,
      "learning_rate": 5.405405405405406e-06,
      "loss": 0.3385,
      "step": 121
    },
    {
      "epoch": 0.0824602906387293,
      "grad_norm": 9.669893264770508,
      "learning_rate": 5.450450450450451e-06,
      "loss": 0.3109,
      "step": 122
    },
    {
      "epoch": 0.08313619466035822,
      "grad_norm": 36.525291442871094,
      "learning_rate": 5.495495495495496e-06,
      "loss": 0.336,
      "step": 123
    },
    {
      "epoch": 0.08381209868198716,
      "grad_norm": 42.838836669921875,
      "learning_rate": 5.540540540540541e-06,
      "loss": 0.3506,
      "step": 124
    },
    {
      "epoch": 0.08448800270361609,
      "grad_norm": 81.95569610595703,
      "learning_rate": 5.585585585585585e-06,
      "loss": 0.3962,
      "step": 125
    },
    {
      "epoch": 0.08516390672524501,
      "grad_norm": 80.83241271972656,
      "learning_rate": 5.6306306306306316e-06,
      "loss": 0.3666,
      "step": 126
    },
    {
      "epoch": 0.08583981074687394,
      "grad_norm": 32.22515869140625,
      "learning_rate": 5.675675675675676e-06,
      "loss": 0.3033,
      "step": 127
    },
    {
      "epoch": 0.08651571476850287,
      "grad_norm": 74.49701690673828,
      "learning_rate": 5.720720720720722e-06,
      "loss": 0.3927,
      "step": 128
    },
    {
      "epoch": 0.0871916187901318,
      "grad_norm": 52.743507385253906,
      "learning_rate": 5.765765765765766e-06,
      "loss": 0.3488,
      "step": 129
    },
    {
      "epoch": 0.08786752281176073,
      "grad_norm": 48.799320220947266,
      "learning_rate": 5.810810810810811e-06,
      "loss": 0.3415,
      "step": 130
    },
    {
      "epoch": 0.08854342683338966,
      "grad_norm": 33.620479583740234,
      "learning_rate": 5.855855855855856e-06,
      "loss": 0.3299,
      "step": 131
    },
    {
      "epoch": 0.08921933085501858,
      "grad_norm": 19.82341766357422,
      "learning_rate": 5.900900900900901e-06,
      "loss": 0.311,
      "step": 132
    },
    {
      "epoch": 0.08989523487664752,
      "grad_norm": 6.281918048858643,
      "learning_rate": 5.945945945945947e-06,
      "loss": 0.3102,
      "step": 133
    },
    {
      "epoch": 0.09057113889827645,
      "grad_norm": 10.104849815368652,
      "learning_rate": 5.990990990990992e-06,
      "loss": 0.319,
      "step": 134
    },
    {
      "epoch": 0.09124704291990537,
      "grad_norm": 32.904720306396484,
      "learning_rate": 6.036036036036037e-06,
      "loss": 0.3213,
      "step": 135
    },
    {
      "epoch": 0.0919229469415343,
      "grad_norm": 48.132911682128906,
      "learning_rate": 6.081081081081082e-06,
      "loss": 0.3415,
      "step": 136
    },
    {
      "epoch": 0.09259885096316323,
      "grad_norm": 19.54690933227539,
      "learning_rate": 6.126126126126126e-06,
      "loss": 0.3292,
      "step": 137
    },
    {
      "epoch": 0.09327475498479217,
      "grad_norm": 21.072050094604492,
      "learning_rate": 6.171171171171172e-06,
      "loss": 0.3289,
      "step": 138
    },
    {
      "epoch": 0.09395065900642109,
      "grad_norm": 10.111052513122559,
      "learning_rate": 6.2162162162162164e-06,
      "loss": 0.319,
      "step": 139
    },
    {
      "epoch": 0.09462656302805002,
      "grad_norm": 15.521512985229492,
      "learning_rate": 6.261261261261262e-06,
      "loss": 0.2997,
      "step": 140
    },
    {
      "epoch": 0.09530246704967894,
      "grad_norm": 32.24085998535156,
      "learning_rate": 6.3063063063063065e-06,
      "loss": 0.3398,
      "step": 141
    },
    {
      "epoch": 0.09597837107130787,
      "grad_norm": 5.101476669311523,
      "learning_rate": 6.351351351351351e-06,
      "loss": 0.3183,
      "step": 142
    },
    {
      "epoch": 0.09665427509293681,
      "grad_norm": 5.569261074066162,
      "learning_rate": 6.396396396396397e-06,
      "loss": 0.3159,
      "step": 143
    },
    {
      "epoch": 0.09733017911456573,
      "grad_norm": 20.40385627746582,
      "learning_rate": 6.441441441441442e-06,
      "loss": 0.3217,
      "step": 144
    },
    {
      "epoch": 0.09800608313619466,
      "grad_norm": 5.1705403327941895,
      "learning_rate": 6.486486486486487e-06,
      "loss": 0.3096,
      "step": 145
    },
    {
      "epoch": 0.09868198715782359,
      "grad_norm": 38.05413055419922,
      "learning_rate": 6.531531531531532e-06,
      "loss": 0.3164,
      "step": 146
    },
    {
      "epoch": 0.09935789117945251,
      "grad_norm": 11.002933502197266,
      "learning_rate": 6.5765765765765775e-06,
      "loss": 0.3164,
      "step": 147
    },
    {
      "epoch": 0.10003379520108145,
      "grad_norm": 3.074906826019287,
      "learning_rate": 6.621621621621622e-06,
      "loss": 0.3149,
      "step": 148
    },
    {
      "epoch": 0.10070969922271038,
      "grad_norm": 20.68535614013672,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.3191,
      "step": 149
    },
    {
      "epoch": 0.1013856032443393,
      "grad_norm": 13.093092918395996,
      "learning_rate": 6.711711711711713e-06,
      "loss": 0.3138,
      "step": 150
    },
    {
      "epoch": 0.1013856032443393,
      "eval_accuracy": 0.252027027027027,
      "eval_loss": 0.3118378818035126,
      "eval_runtime": 43.3039,
      "eval_samples_per_second": 546.833,
      "eval_steps_per_second": 17.089,
      "step": 150
    },
    {
      "epoch": 0.10206150726596823,
      "grad_norm": 12.758399963378906,
      "learning_rate": 6.7567567567567575e-06,
      "loss": 0.3269,
      "step": 151
    },
    {
      "epoch": 0.10273741128759716,
      "grad_norm": 8.120706558227539,
      "learning_rate": 6.801801801801803e-06,
      "loss": 0.3137,
      "step": 152
    },
    {
      "epoch": 0.1034133153092261,
      "grad_norm": 14.459858894348145,
      "learning_rate": 6.846846846846848e-06,
      "loss": 0.3183,
      "step": 153
    },
    {
      "epoch": 0.10408921933085502,
      "grad_norm": 33.33787536621094,
      "learning_rate": 6.891891891891892e-06,
      "loss": 0.3238,
      "step": 154
    },
    {
      "epoch": 0.10476512335248395,
      "grad_norm": 19.07410430908203,
      "learning_rate": 6.936936936936938e-06,
      "loss": 0.3118,
      "step": 155
    },
    {
      "epoch": 0.10544102737411287,
      "grad_norm": 16.95354652404785,
      "learning_rate": 6.981981981981982e-06,
      "loss": 0.3263,
      "step": 156
    },
    {
      "epoch": 0.1061169313957418,
      "grad_norm": 52.63601303100586,
      "learning_rate": 7.027027027027028e-06,
      "loss": 0.354,
      "step": 157
    },
    {
      "epoch": 0.10679283541737074,
      "grad_norm": 3.3446900844573975,
      "learning_rate": 7.072072072072072e-06,
      "loss": 0.3071,
      "step": 158
    },
    {
      "epoch": 0.10746873943899966,
      "grad_norm": 11.265155792236328,
      "learning_rate": 7.117117117117117e-06,
      "loss": 0.311,
      "step": 159
    },
    {
      "epoch": 0.10814464346062859,
      "grad_norm": 10.216545104980469,
      "learning_rate": 7.162162162162163e-06,
      "loss": 0.2964,
      "step": 160
    },
    {
      "epoch": 0.10882054748225752,
      "grad_norm": 7.049203872680664,
      "learning_rate": 7.207207207207208e-06,
      "loss": 0.3145,
      "step": 161
    },
    {
      "epoch": 0.10949645150388645,
      "grad_norm": 24.31463623046875,
      "learning_rate": 7.252252252252253e-06,
      "loss": 0.3133,
      "step": 162
    },
    {
      "epoch": 0.11017235552551538,
      "grad_norm": 23.22296905517578,
      "learning_rate": 7.297297297297298e-06,
      "loss": 0.3041,
      "step": 163
    },
    {
      "epoch": 0.1108482595471443,
      "grad_norm": 23.577558517456055,
      "learning_rate": 7.342342342342343e-06,
      "loss": 0.329,
      "step": 164
    },
    {
      "epoch": 0.11152416356877323,
      "grad_norm": 18.758020401000977,
      "learning_rate": 7.387387387387388e-06,
      "loss": 0.2977,
      "step": 165
    },
    {
      "epoch": 0.11220006759040216,
      "grad_norm": 38.65229797363281,
      "learning_rate": 7.4324324324324324e-06,
      "loss": 0.3309,
      "step": 166
    },
    {
      "epoch": 0.1128759716120311,
      "grad_norm": 3.814058542251587,
      "learning_rate": 7.477477477477479e-06,
      "loss": 0.2995,
      "step": 167
    },
    {
      "epoch": 0.11355187563366002,
      "grad_norm": 36.64772415161133,
      "learning_rate": 7.5225225225225225e-06,
      "loss": 0.3293,
      "step": 168
    },
    {
      "epoch": 0.11422777965528895,
      "grad_norm": 9.421634674072266,
      "learning_rate": 7.567567567567569e-06,
      "loss": 0.3325,
      "step": 169
    },
    {
      "epoch": 0.11490368367691788,
      "grad_norm": 60.44966506958008,
      "learning_rate": 7.612612612612613e-06,
      "loss": 0.3443,
      "step": 170
    },
    {
      "epoch": 0.1155795876985468,
      "grad_norm": 39.712276458740234,
      "learning_rate": 7.657657657657658e-06,
      "loss": 0.3115,
      "step": 171
    },
    {
      "epoch": 0.11625549172017574,
      "grad_norm": 5.191213130950928,
      "learning_rate": 7.702702702702704e-06,
      "loss": 0.2912,
      "step": 172
    },
    {
      "epoch": 0.11693139574180467,
      "grad_norm": 7.7331109046936035,
      "learning_rate": 7.747747747747749e-06,
      "loss": 0.3085,
      "step": 173
    },
    {
      "epoch": 0.11760729976343359,
      "grad_norm": 35.76954650878906,
      "learning_rate": 7.792792792792793e-06,
      "loss": 0.343,
      "step": 174
    },
    {
      "epoch": 0.11828320378506252,
      "grad_norm": 15.870879173278809,
      "learning_rate": 7.837837837837838e-06,
      "loss": 0.304,
      "step": 175
    },
    {
      "epoch": 0.11895910780669144,
      "grad_norm": 6.3920817375183105,
      "learning_rate": 7.882882882882884e-06,
      "loss": 0.3059,
      "step": 176
    },
    {
      "epoch": 0.11963501182832038,
      "grad_norm": 21.888822555541992,
      "learning_rate": 7.927927927927929e-06,
      "loss": 0.3073,
      "step": 177
    },
    {
      "epoch": 0.12031091584994931,
      "grad_norm": 4.58980131149292,
      "learning_rate": 7.972972972972974e-06,
      "loss": 0.3003,
      "step": 178
    },
    {
      "epoch": 0.12098681987157824,
      "grad_norm": 54.496238708496094,
      "learning_rate": 8.018018018018018e-06,
      "loss": 0.3448,
      "step": 179
    },
    {
      "epoch": 0.12166272389320716,
      "grad_norm": 33.6229362487793,
      "learning_rate": 8.063063063063063e-06,
      "loss": 0.3246,
      "step": 180
    },
    {
      "epoch": 0.12233862791483609,
      "grad_norm": 12.674449920654297,
      "learning_rate": 8.108108108108109e-06,
      "loss": 0.3208,
      "step": 181
    },
    {
      "epoch": 0.12301453193646503,
      "grad_norm": 17.364849090576172,
      "learning_rate": 8.153153153153154e-06,
      "loss": 0.2838,
      "step": 182
    },
    {
      "epoch": 0.12369043595809395,
      "grad_norm": 48.96987533569336,
      "learning_rate": 8.198198198198198e-06,
      "loss": 0.3271,
      "step": 183
    },
    {
      "epoch": 0.12436633997972288,
      "grad_norm": 30.79657745361328,
      "learning_rate": 8.243243243243245e-06,
      "loss": 0.3049,
      "step": 184
    },
    {
      "epoch": 0.1250422440013518,
      "grad_norm": 6.275866508483887,
      "learning_rate": 8.288288288288289e-06,
      "loss": 0.2769,
      "step": 185
    },
    {
      "epoch": 0.12571814802298073,
      "grad_norm": 12.865313529968262,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.3083,
      "step": 186
    },
    {
      "epoch": 0.12639405204460966,
      "grad_norm": 19.382774353027344,
      "learning_rate": 8.378378378378378e-06,
      "loss": 0.3046,
      "step": 187
    },
    {
      "epoch": 0.12706995606623858,
      "grad_norm": 12.652450561523438,
      "learning_rate": 8.423423423423423e-06,
      "loss": 0.3124,
      "step": 188
    },
    {
      "epoch": 0.12774586008786754,
      "grad_norm": 27.167505264282227,
      "learning_rate": 8.46846846846847e-06,
      "loss": 0.3082,
      "step": 189
    },
    {
      "epoch": 0.12842176410949646,
      "grad_norm": 8.244668006896973,
      "learning_rate": 8.513513513513514e-06,
      "loss": 0.2787,
      "step": 190
    },
    {
      "epoch": 0.1290976681311254,
      "grad_norm": 6.807435035705566,
      "learning_rate": 8.55855855855856e-06,
      "loss": 0.2629,
      "step": 191
    },
    {
      "epoch": 0.1297735721527543,
      "grad_norm": 19.980337142944336,
      "learning_rate": 8.603603603603605e-06,
      "loss": 0.2795,
      "step": 192
    },
    {
      "epoch": 0.13044947617438324,
      "grad_norm": 58.78095626831055,
      "learning_rate": 8.64864864864865e-06,
      "loss": 0.3531,
      "step": 193
    },
    {
      "epoch": 0.13112538019601216,
      "grad_norm": 47.949562072753906,
      "learning_rate": 8.693693693693694e-06,
      "loss": 0.3262,
      "step": 194
    },
    {
      "epoch": 0.1318012842176411,
      "grad_norm": 36.29952621459961,
      "learning_rate": 8.738738738738739e-06,
      "loss": 0.3128,
      "step": 195
    },
    {
      "epoch": 0.13247718823927002,
      "grad_norm": 26.755901336669922,
      "learning_rate": 8.783783783783785e-06,
      "loss": 0.3022,
      "step": 196
    },
    {
      "epoch": 0.13315309226089894,
      "grad_norm": 54.272193908691406,
      "learning_rate": 8.82882882882883e-06,
      "loss": 0.3433,
      "step": 197
    },
    {
      "epoch": 0.13382899628252787,
      "grad_norm": 7.653815269470215,
      "learning_rate": 8.873873873873876e-06,
      "loss": 0.3244,
      "step": 198
    },
    {
      "epoch": 0.13450490030415682,
      "grad_norm": 6.53266716003418,
      "learning_rate": 8.91891891891892e-06,
      "loss": 0.2703,
      "step": 199
    },
    {
      "epoch": 0.13518080432578575,
      "grad_norm": 41.0377311706543,
      "learning_rate": 8.963963963963965e-06,
      "loss": 0.3092,
      "step": 200
    },
    {
      "epoch": 0.13518080432578575,
      "eval_accuracy": 0.34256756756756757,
      "eval_loss": 0.30219197273254395,
      "eval_runtime": 42.5744,
      "eval_samples_per_second": 556.203,
      "eval_steps_per_second": 17.381,
      "step": 200
    },
    {
      "epoch": 0.13585670834741467,
      "grad_norm": 22.133607864379883,
      "learning_rate": 9.00900900900901e-06,
      "loss": 0.3071,
      "step": 201
    },
    {
      "epoch": 0.1365326123690436,
      "grad_norm": 28.495906829833984,
      "learning_rate": 9.054054054054054e-06,
      "loss": 0.332,
      "step": 202
    },
    {
      "epoch": 0.13720851639067252,
      "grad_norm": 34.705894470214844,
      "learning_rate": 9.0990990990991e-06,
      "loss": 0.2802,
      "step": 203
    },
    {
      "epoch": 0.13788442041230145,
      "grad_norm": 36.17578125,
      "learning_rate": 9.144144144144145e-06,
      "loss": 0.2968,
      "step": 204
    },
    {
      "epoch": 0.13856032443393038,
      "grad_norm": 9.468372344970703,
      "learning_rate": 9.189189189189191e-06,
      "loss": 0.3045,
      "step": 205
    },
    {
      "epoch": 0.1392362284555593,
      "grad_norm": 13.604094505310059,
      "learning_rate": 9.234234234234236e-06,
      "loss": 0.2739,
      "step": 206
    },
    {
      "epoch": 0.13991213247718823,
      "grad_norm": 7.345333099365234,
      "learning_rate": 9.27927927927928e-06,
      "loss": 0.2635,
      "step": 207
    },
    {
      "epoch": 0.14058803649881718,
      "grad_norm": 79.34138488769531,
      "learning_rate": 9.324324324324325e-06,
      "loss": 0.3631,
      "step": 208
    },
    {
      "epoch": 0.1412639405204461,
      "grad_norm": 13.533716201782227,
      "learning_rate": 9.36936936936937e-06,
      "loss": 0.2889,
      "step": 209
    },
    {
      "epoch": 0.14193984454207503,
      "grad_norm": 15.761404037475586,
      "learning_rate": 9.414414414414416e-06,
      "loss": 0.2912,
      "step": 210
    },
    {
      "epoch": 0.14261574856370396,
      "grad_norm": 38.67661666870117,
      "learning_rate": 9.45945945945946e-06,
      "loss": 0.3288,
      "step": 211
    },
    {
      "epoch": 0.14329165258533288,
      "grad_norm": 32.91361999511719,
      "learning_rate": 9.504504504504505e-06,
      "loss": 0.3151,
      "step": 212
    },
    {
      "epoch": 0.1439675566069618,
      "grad_norm": 10.84450912475586,
      "learning_rate": 9.54954954954955e-06,
      "loss": 0.2892,
      "step": 213
    },
    {
      "epoch": 0.14464346062859074,
      "grad_norm": 15.134542465209961,
      "learning_rate": 9.594594594594594e-06,
      "loss": 0.2969,
      "step": 214
    },
    {
      "epoch": 0.14531936465021966,
      "grad_norm": 19.319963455200195,
      "learning_rate": 9.63963963963964e-06,
      "loss": 0.2896,
      "step": 215
    },
    {
      "epoch": 0.1459952686718486,
      "grad_norm": 38.52631378173828,
      "learning_rate": 9.684684684684685e-06,
      "loss": 0.3058,
      "step": 216
    },
    {
      "epoch": 0.14667117269347751,
      "grad_norm": 6.934603214263916,
      "learning_rate": 9.729729729729732e-06,
      "loss": 0.2851,
      "step": 217
    },
    {
      "epoch": 0.14734707671510647,
      "grad_norm": 8.268049240112305,
      "learning_rate": 9.774774774774776e-06,
      "loss": 0.2611,
      "step": 218
    },
    {
      "epoch": 0.1480229807367354,
      "grad_norm": 32.98963165283203,
      "learning_rate": 9.81981981981982e-06,
      "loss": 0.2857,
      "step": 219
    },
    {
      "epoch": 0.14869888475836432,
      "grad_norm": 6.996422290802002,
      "learning_rate": 9.864864864864865e-06,
      "loss": 0.2868,
      "step": 220
    },
    {
      "epoch": 0.14937478877999325,
      "grad_norm": 28.030460357666016,
      "learning_rate": 9.90990990990991e-06,
      "loss": 0.3154,
      "step": 221
    },
    {
      "epoch": 0.15005069280162217,
      "grad_norm": 10.567412376403809,
      "learning_rate": 9.954954954954956e-06,
      "loss": 0.3289,
      "step": 222
    },
    {
      "epoch": 0.1507265968232511,
      "grad_norm": 18.604751586914062,
      "learning_rate": 1e-05,
      "loss": 0.2699,
      "step": 223
    },
    {
      "epoch": 0.15140250084488002,
      "grad_norm": 50.90874481201172,
      "learning_rate": 1.0045045045045045e-05,
      "loss": 0.3023,
      "step": 224
    },
    {
      "epoch": 0.15207840486650895,
      "grad_norm": 22.797761917114258,
      "learning_rate": 1.009009009009009e-05,
      "loss": 0.2877,
      "step": 225
    },
    {
      "epoch": 0.15275430888813787,
      "grad_norm": 19.7872371673584,
      "learning_rate": 1.0135135135135136e-05,
      "loss": 0.2764,
      "step": 226
    },
    {
      "epoch": 0.1534302129097668,
      "grad_norm": 33.477046966552734,
      "learning_rate": 1.0180180180180181e-05,
      "loss": 0.3086,
      "step": 227
    },
    {
      "epoch": 0.15410611693139575,
      "grad_norm": 31.19655418395996,
      "learning_rate": 1.0225225225225226e-05,
      "loss": 0.3136,
      "step": 228
    },
    {
      "epoch": 0.15478202095302468,
      "grad_norm": 10.91556167602539,
      "learning_rate": 1.027027027027027e-05,
      "loss": 0.2867,
      "step": 229
    },
    {
      "epoch": 0.1554579249746536,
      "grad_norm": 8.266558647155762,
      "learning_rate": 1.0315315315315315e-05,
      "loss": 0.2428,
      "step": 230
    },
    {
      "epoch": 0.15613382899628253,
      "grad_norm": 76.07386016845703,
      "learning_rate": 1.0360360360360363e-05,
      "loss": 0.3284,
      "step": 231
    },
    {
      "epoch": 0.15680973301791146,
      "grad_norm": 9.15540885925293,
      "learning_rate": 1.0405405405405407e-05,
      "loss": 0.2659,
      "step": 232
    },
    {
      "epoch": 0.15748563703954038,
      "grad_norm": 48.027278900146484,
      "learning_rate": 1.0450450450450452e-05,
      "loss": 0.3124,
      "step": 233
    },
    {
      "epoch": 0.1581615410611693,
      "grad_norm": 50.865779876708984,
      "learning_rate": 1.0495495495495496e-05,
      "loss": 0.3023,
      "step": 234
    },
    {
      "epoch": 0.15883744508279823,
      "grad_norm": 25.013639450073242,
      "learning_rate": 1.0540540540540541e-05,
      "loss": 0.3161,
      "step": 235
    },
    {
      "epoch": 0.15951334910442716,
      "grad_norm": 9.544523239135742,
      "learning_rate": 1.0585585585585587e-05,
      "loss": 0.29,
      "step": 236
    },
    {
      "epoch": 0.1601892531260561,
      "grad_norm": 8.434956550598145,
      "learning_rate": 1.0630630630630632e-05,
      "loss": 0.2898,
      "step": 237
    },
    {
      "epoch": 0.16086515714768504,
      "grad_norm": 24.728628158569336,
      "learning_rate": 1.0675675675675677e-05,
      "loss": 0.2734,
      "step": 238
    },
    {
      "epoch": 0.16154106116931397,
      "grad_norm": 68.85552978515625,
      "learning_rate": 1.0720720720720721e-05,
      "loss": 0.3177,
      "step": 239
    },
    {
      "epoch": 0.1622169651909429,
      "grad_norm": 45.969200134277344,
      "learning_rate": 1.0765765765765767e-05,
      "loss": 0.3064,
      "step": 240
    },
    {
      "epoch": 0.16289286921257182,
      "grad_norm": 67.59300994873047,
      "learning_rate": 1.0810810810810812e-05,
      "loss": 0.3167,
      "step": 241
    },
    {
      "epoch": 0.16356877323420074,
      "grad_norm": 43.392704010009766,
      "learning_rate": 1.0855855855855857e-05,
      "loss": 0.3176,
      "step": 242
    },
    {
      "epoch": 0.16424467725582967,
      "grad_norm": 27.959455490112305,
      "learning_rate": 1.0900900900900901e-05,
      "loss": 0.2948,
      "step": 243
    },
    {
      "epoch": 0.1649205812774586,
      "grad_norm": 44.7069206237793,
      "learning_rate": 1.0945945945945946e-05,
      "loss": 0.2994,
      "step": 244
    },
    {
      "epoch": 0.16559648529908752,
      "grad_norm": 65.22603607177734,
      "learning_rate": 1.0990990990990992e-05,
      "loss": 0.3741,
      "step": 245
    },
    {
      "epoch": 0.16627238932071645,
      "grad_norm": 10.029617309570312,
      "learning_rate": 1.1036036036036037e-05,
      "loss": 0.2608,
      "step": 246
    },
    {
      "epoch": 0.1669482933423454,
      "grad_norm": 37.45057678222656,
      "learning_rate": 1.1081081081081081e-05,
      "loss": 0.3048,
      "step": 247
    },
    {
      "epoch": 0.16762419736397433,
      "grad_norm": 58.8104362487793,
      "learning_rate": 1.1126126126126126e-05,
      "loss": 0.3341,
      "step": 248
    },
    {
      "epoch": 0.16830010138560325,
      "grad_norm": 45.97346878051758,
      "learning_rate": 1.117117117117117e-05,
      "loss": 0.317,
      "step": 249
    },
    {
      "epoch": 0.16897600540723218,
      "grad_norm": 25.392379760742188,
      "learning_rate": 1.1216216216216219e-05,
      "loss": 0.2935,
      "step": 250
    },
    {
      "epoch": 0.16897600540723218,
      "eval_accuracy": 0.3402449324324324,
      "eval_loss": 0.30567193031311035,
      "eval_runtime": 43.9669,
      "eval_samples_per_second": 538.587,
      "eval_steps_per_second": 16.831,
      "step": 250
    },
    {
      "epoch": 0.1696519094288611,
      "grad_norm": 46.97799301147461,
      "learning_rate": 1.1261261261261263e-05,
      "loss": 0.3187,
      "step": 251
    },
    {
      "epoch": 0.17032781345049003,
      "grad_norm": 63.25579833984375,
      "learning_rate": 1.1306306306306308e-05,
      "loss": 0.3674,
      "step": 252
    },
    {
      "epoch": 0.17100371747211895,
      "grad_norm": 14.791979789733887,
      "learning_rate": 1.1351351351351352e-05,
      "loss": 0.2545,
      "step": 253
    },
    {
      "epoch": 0.17167962149374788,
      "grad_norm": 21.663419723510742,
      "learning_rate": 1.1396396396396397e-05,
      "loss": 0.2757,
      "step": 254
    },
    {
      "epoch": 0.1723555255153768,
      "grad_norm": 33.15219497680664,
      "learning_rate": 1.1441441441441443e-05,
      "loss": 0.2627,
      "step": 255
    },
    {
      "epoch": 0.17303142953700573,
      "grad_norm": 38.973812103271484,
      "learning_rate": 1.1486486486486488e-05,
      "loss": 0.2787,
      "step": 256
    },
    {
      "epoch": 0.17370733355863469,
      "grad_norm": 10.700187683105469,
      "learning_rate": 1.1531531531531532e-05,
      "loss": 0.288,
      "step": 257
    },
    {
      "epoch": 0.1743832375802636,
      "grad_norm": 21.872146606445312,
      "learning_rate": 1.1576576576576577e-05,
      "loss": 0.2467,
      "step": 258
    },
    {
      "epoch": 0.17505914160189254,
      "grad_norm": 30.115171432495117,
      "learning_rate": 1.1621621621621622e-05,
      "loss": 0.2768,
      "step": 259
    },
    {
      "epoch": 0.17573504562352146,
      "grad_norm": 41.45762634277344,
      "learning_rate": 1.1666666666666668e-05,
      "loss": 0.2879,
      "step": 260
    },
    {
      "epoch": 0.1764109496451504,
      "grad_norm": 35.731529235839844,
      "learning_rate": 1.1711711711711713e-05,
      "loss": 0.3471,
      "step": 261
    },
    {
      "epoch": 0.17708685366677931,
      "grad_norm": 55.86954116821289,
      "learning_rate": 1.1756756756756757e-05,
      "loss": 0.3403,
      "step": 262
    },
    {
      "epoch": 0.17776275768840824,
      "grad_norm": 39.81838607788086,
      "learning_rate": 1.1801801801801802e-05,
      "loss": 0.3234,
      "step": 263
    },
    {
      "epoch": 0.17843866171003717,
      "grad_norm": 33.202850341796875,
      "learning_rate": 1.1846846846846846e-05,
      "loss": 0.3144,
      "step": 264
    },
    {
      "epoch": 0.1791145657316661,
      "grad_norm": 9.565064430236816,
      "learning_rate": 1.1891891891891894e-05,
      "loss": 0.2607,
      "step": 265
    },
    {
      "epoch": 0.17979046975329505,
      "grad_norm": 27.02252960205078,
      "learning_rate": 1.1936936936936939e-05,
      "loss": 0.2456,
      "step": 266
    },
    {
      "epoch": 0.18046637377492397,
      "grad_norm": 37.121246337890625,
      "learning_rate": 1.1981981981981983e-05,
      "loss": 0.3283,
      "step": 267
    },
    {
      "epoch": 0.1811422777965529,
      "grad_norm": 62.21095275878906,
      "learning_rate": 1.2027027027027028e-05,
      "loss": 0.3513,
      "step": 268
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 6.8497514724731445,
      "learning_rate": 1.2072072072072074e-05,
      "loss": 0.2619,
      "step": 269
    },
    {
      "epoch": 0.18249408583981075,
      "grad_norm": 27.067394256591797,
      "learning_rate": 1.2117117117117119e-05,
      "loss": 0.3066,
      "step": 270
    },
    {
      "epoch": 0.18316998986143967,
      "grad_norm": 22.04492950439453,
      "learning_rate": 1.2162162162162164e-05,
      "loss": 0.2784,
      "step": 271
    },
    {
      "epoch": 0.1838458938830686,
      "grad_norm": 56.1871337890625,
      "learning_rate": 1.2207207207207208e-05,
      "loss": 0.3327,
      "step": 272
    },
    {
      "epoch": 0.18452179790469753,
      "grad_norm": 27.577497482299805,
      "learning_rate": 1.2252252252252253e-05,
      "loss": 0.3241,
      "step": 273
    },
    {
      "epoch": 0.18519770192632645,
      "grad_norm": 9.122892379760742,
      "learning_rate": 1.2297297297297299e-05,
      "loss": 0.2667,
      "step": 274
    },
    {
      "epoch": 0.18587360594795538,
      "grad_norm": 28.702058792114258,
      "learning_rate": 1.2342342342342344e-05,
      "loss": 0.2849,
      "step": 275
    },
    {
      "epoch": 0.18654950996958433,
      "grad_norm": 29.983430862426758,
      "learning_rate": 1.2387387387387388e-05,
      "loss": 0.2672,
      "step": 276
    },
    {
      "epoch": 0.18722541399121326,
      "grad_norm": 67.43209075927734,
      "learning_rate": 1.2432432432432433e-05,
      "loss": 0.4042,
      "step": 277
    },
    {
      "epoch": 0.18790131801284218,
      "grad_norm": 8.171981811523438,
      "learning_rate": 1.2477477477477477e-05,
      "loss": 0.246,
      "step": 278
    },
    {
      "epoch": 0.1885772220344711,
      "grad_norm": 6.953465461730957,
      "learning_rate": 1.2522522522522524e-05,
      "loss": 0.3042,
      "step": 279
    },
    {
      "epoch": 0.18925312605610004,
      "grad_norm": 32.97325897216797,
      "learning_rate": 1.2567567567567568e-05,
      "loss": 0.2794,
      "step": 280
    },
    {
      "epoch": 0.18992903007772896,
      "grad_norm": 39.848594665527344,
      "learning_rate": 1.2612612612612613e-05,
      "loss": 0.3137,
      "step": 281
    },
    {
      "epoch": 0.1906049340993579,
      "grad_norm": 29.38252830505371,
      "learning_rate": 1.2657657657657658e-05,
      "loss": 0.2849,
      "step": 282
    },
    {
      "epoch": 0.1912808381209868,
      "grad_norm": 13.006146430969238,
      "learning_rate": 1.2702702702702702e-05,
      "loss": 0.2641,
      "step": 283
    },
    {
      "epoch": 0.19195674214261574,
      "grad_norm": 28.61220932006836,
      "learning_rate": 1.274774774774775e-05,
      "loss": 0.2934,
      "step": 284
    },
    {
      "epoch": 0.19263264616424466,
      "grad_norm": 35.80970764160156,
      "learning_rate": 1.2792792792792795e-05,
      "loss": 0.3241,
      "step": 285
    },
    {
      "epoch": 0.19330855018587362,
      "grad_norm": 51.582115173339844,
      "learning_rate": 1.283783783783784e-05,
      "loss": 0.3451,
      "step": 286
    },
    {
      "epoch": 0.19398445420750254,
      "grad_norm": 8.770044326782227,
      "learning_rate": 1.2882882882882884e-05,
      "loss": 0.3011,
      "step": 287
    },
    {
      "epoch": 0.19466035822913147,
      "grad_norm": 61.289039611816406,
      "learning_rate": 1.2927927927927929e-05,
      "loss": 0.3617,
      "step": 288
    },
    {
      "epoch": 0.1953362622507604,
      "grad_norm": 50.52556228637695,
      "learning_rate": 1.2972972972972975e-05,
      "loss": 0.3754,
      "step": 289
    },
    {
      "epoch": 0.19601216627238932,
      "grad_norm": 37.355918884277344,
      "learning_rate": 1.301801801801802e-05,
      "loss": 0.3107,
      "step": 290
    },
    {
      "epoch": 0.19668807029401825,
      "grad_norm": 16.883445739746094,
      "learning_rate": 1.3063063063063064e-05,
      "loss": 0.2745,
      "step": 291
    },
    {
      "epoch": 0.19736397431564717,
      "grad_norm": 28.419795989990234,
      "learning_rate": 1.3108108108108109e-05,
      "loss": 0.296,
      "step": 292
    },
    {
      "epoch": 0.1980398783372761,
      "grad_norm": 44.596282958984375,
      "learning_rate": 1.3153153153153155e-05,
      "loss": 0.3079,
      "step": 293
    },
    {
      "epoch": 0.19871578235890502,
      "grad_norm": 57.12053680419922,
      "learning_rate": 1.31981981981982e-05,
      "loss": 0.364,
      "step": 294
    },
    {
      "epoch": 0.19939168638053398,
      "grad_norm": 38.92661666870117,
      "learning_rate": 1.3243243243243244e-05,
      "loss": 0.335,
      "step": 295
    },
    {
      "epoch": 0.2000675904021629,
      "grad_norm": 6.355953216552734,
      "learning_rate": 1.3288288288288289e-05,
      "loss": 0.2588,
      "step": 296
    },
    {
      "epoch": 0.20074349442379183,
      "grad_norm": 39.481842041015625,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.3139,
      "step": 297
    },
    {
      "epoch": 0.20141939844542076,
      "grad_norm": 54.15153121948242,
      "learning_rate": 1.3378378378378381e-05,
      "loss": 0.3402,
      "step": 298
    },
    {
      "epoch": 0.20209530246704968,
      "grad_norm": 27.121000289916992,
      "learning_rate": 1.3423423423423426e-05,
      "loss": 0.2988,
      "step": 299
    },
    {
      "epoch": 0.2027712064886786,
      "grad_norm": 14.737486839294434,
      "learning_rate": 1.346846846846847e-05,
      "loss": 0.2765,
      "step": 300
    },
    {
      "epoch": 0.2027712064886786,
      "eval_accuracy": 0.3455658783783784,
      "eval_loss": 0.2748236358165741,
      "eval_runtime": 43.9212,
      "eval_samples_per_second": 539.147,
      "eval_steps_per_second": 16.848,
      "step": 300
    },
    {
      "epoch": 0.20344711051030753,
      "grad_norm": 6.383377552032471,
      "learning_rate": 1.3513513513513515e-05,
      "loss": 0.2815,
      "step": 301
    },
    {
      "epoch": 0.20412301453193646,
      "grad_norm": 55.12192153930664,
      "learning_rate": 1.355855855855856e-05,
      "loss": 0.3347,
      "step": 302
    },
    {
      "epoch": 0.20479891855356538,
      "grad_norm": 30.874704360961914,
      "learning_rate": 1.3603603603603606e-05,
      "loss": 0.3156,
      "step": 303
    },
    {
      "epoch": 0.2054748225751943,
      "grad_norm": 15.869712829589844,
      "learning_rate": 1.364864864864865e-05,
      "loss": 0.2242,
      "step": 304
    },
    {
      "epoch": 0.20615072659682326,
      "grad_norm": 21.23921775817871,
      "learning_rate": 1.3693693693693695e-05,
      "loss": 0.2698,
      "step": 305
    },
    {
      "epoch": 0.2068266306184522,
      "grad_norm": 21.48433494567871,
      "learning_rate": 1.373873873873874e-05,
      "loss": 0.2857,
      "step": 306
    },
    {
      "epoch": 0.20750253464008112,
      "grad_norm": 29.257539749145508,
      "learning_rate": 1.3783783783783784e-05,
      "loss": 0.2538,
      "step": 307
    },
    {
      "epoch": 0.20817843866171004,
      "grad_norm": 17.0058650970459,
      "learning_rate": 1.382882882882883e-05,
      "loss": 0.2585,
      "step": 308
    },
    {
      "epoch": 0.20885434268333897,
      "grad_norm": 35.79644012451172,
      "learning_rate": 1.3873873873873875e-05,
      "loss": 0.2691,
      "step": 309
    },
    {
      "epoch": 0.2095302467049679,
      "grad_norm": 7.479666709899902,
      "learning_rate": 1.391891891891892e-05,
      "loss": 0.266,
      "step": 310
    },
    {
      "epoch": 0.21020615072659682,
      "grad_norm": 9.142624855041504,
      "learning_rate": 1.3963963963963964e-05,
      "loss": 0.2419,
      "step": 311
    },
    {
      "epoch": 0.21088205474822574,
      "grad_norm": 21.355792999267578,
      "learning_rate": 1.4009009009009009e-05,
      "loss": 0.33,
      "step": 312
    },
    {
      "epoch": 0.21155795876985467,
      "grad_norm": 22.918132781982422,
      "learning_rate": 1.4054054054054055e-05,
      "loss": 0.2346,
      "step": 313
    },
    {
      "epoch": 0.2122338627914836,
      "grad_norm": 25.655885696411133,
      "learning_rate": 1.40990990990991e-05,
      "loss": 0.303,
      "step": 314
    },
    {
      "epoch": 0.21290976681311255,
      "grad_norm": 24.754261016845703,
      "learning_rate": 1.4144144144144145e-05,
      "loss": 0.2697,
      "step": 315
    },
    {
      "epoch": 0.21358567083474148,
      "grad_norm": 17.06878089904785,
      "learning_rate": 1.4189189189189189e-05,
      "loss": 0.2904,
      "step": 316
    },
    {
      "epoch": 0.2142615748563704,
      "grad_norm": 36.48949432373047,
      "learning_rate": 1.4234234234234234e-05,
      "loss": 0.2713,
      "step": 317
    },
    {
      "epoch": 0.21493747887799933,
      "grad_norm": 26.829757690429688,
      "learning_rate": 1.4279279279279282e-05,
      "loss": 0.3098,
      "step": 318
    },
    {
      "epoch": 0.21561338289962825,
      "grad_norm": 13.51832389831543,
      "learning_rate": 1.4324324324324326e-05,
      "loss": 0.2525,
      "step": 319
    },
    {
      "epoch": 0.21628928692125718,
      "grad_norm": 12.53862476348877,
      "learning_rate": 1.4369369369369371e-05,
      "loss": 0.2359,
      "step": 320
    },
    {
      "epoch": 0.2169651909428861,
      "grad_norm": 5.599822044372559,
      "learning_rate": 1.4414414414414416e-05,
      "loss": 0.2275,
      "step": 321
    },
    {
      "epoch": 0.21764109496451503,
      "grad_norm": 40.34337615966797,
      "learning_rate": 1.4459459459459462e-05,
      "loss": 0.3022,
      "step": 322
    },
    {
      "epoch": 0.21831699898614396,
      "grad_norm": 43.734012603759766,
      "learning_rate": 1.4504504504504506e-05,
      "loss": 0.3301,
      "step": 323
    },
    {
      "epoch": 0.2189929030077729,
      "grad_norm": 18.017704010009766,
      "learning_rate": 1.4549549549549551e-05,
      "loss": 0.2478,
      "step": 324
    },
    {
      "epoch": 0.21966880702940184,
      "grad_norm": 5.637330532073975,
      "learning_rate": 1.4594594594594596e-05,
      "loss": 0.2563,
      "step": 325
    },
    {
      "epoch": 0.22034471105103076,
      "grad_norm": 26.21967315673828,
      "learning_rate": 1.463963963963964e-05,
      "loss": 0.2606,
      "step": 326
    },
    {
      "epoch": 0.2210206150726597,
      "grad_norm": 52.56633377075195,
      "learning_rate": 1.4684684684684686e-05,
      "loss": 0.31,
      "step": 327
    },
    {
      "epoch": 0.2216965190942886,
      "grad_norm": 17.02704620361328,
      "learning_rate": 1.4729729729729731e-05,
      "loss": 0.2838,
      "step": 328
    },
    {
      "epoch": 0.22237242311591754,
      "grad_norm": 5.616494655609131,
      "learning_rate": 1.4774774774774776e-05,
      "loss": 0.2591,
      "step": 329
    },
    {
      "epoch": 0.22304832713754646,
      "grad_norm": 19.03318977355957,
      "learning_rate": 1.481981981981982e-05,
      "loss": 0.2753,
      "step": 330
    },
    {
      "epoch": 0.2237242311591754,
      "grad_norm": 7.4971208572387695,
      "learning_rate": 1.4864864864864865e-05,
      "loss": 0.2333,
      "step": 331
    },
    {
      "epoch": 0.22440013518080432,
      "grad_norm": 44.747642517089844,
      "learning_rate": 1.4909909909909913e-05,
      "loss": 0.2903,
      "step": 332
    },
    {
      "epoch": 0.22507603920243324,
      "grad_norm": 19.317218780517578,
      "learning_rate": 1.4954954954954957e-05,
      "loss": 0.2647,
      "step": 333
    },
    {
      "epoch": 0.2257519432240622,
      "grad_norm": 20.598121643066406,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.2588,
      "step": 334
    },
    {
      "epoch": 0.22642784724569112,
      "grad_norm": 6.493048191070557,
      "learning_rate": 1.5045045045045045e-05,
      "loss": 0.259,
      "step": 335
    },
    {
      "epoch": 0.22710375126732005,
      "grad_norm": 17.300582885742188,
      "learning_rate": 1.509009009009009e-05,
      "loss": 0.2402,
      "step": 336
    },
    {
      "epoch": 0.22777965528894897,
      "grad_norm": 7.773423671722412,
      "learning_rate": 1.5135135135135138e-05,
      "loss": 0.2454,
      "step": 337
    },
    {
      "epoch": 0.2284555593105779,
      "grad_norm": 30.24074363708496,
      "learning_rate": 1.5180180180180182e-05,
      "loss": 0.3272,
      "step": 338
    },
    {
      "epoch": 0.22913146333220683,
      "grad_norm": 10.738046646118164,
      "learning_rate": 1.5225225225225227e-05,
      "loss": 0.2739,
      "step": 339
    },
    {
      "epoch": 0.22980736735383575,
      "grad_norm": 9.040885925292969,
      "learning_rate": 1.527027027027027e-05,
      "loss": 0.2918,
      "step": 340
    },
    {
      "epoch": 0.23048327137546468,
      "grad_norm": 28.83804702758789,
      "learning_rate": 1.5315315315315316e-05,
      "loss": 0.2858,
      "step": 341
    },
    {
      "epoch": 0.2311591753970936,
      "grad_norm": 23.90815544128418,
      "learning_rate": 1.5360360360360362e-05,
      "loss": 0.2468,
      "step": 342
    },
    {
      "epoch": 0.23183507941872253,
      "grad_norm": 7.428371429443359,
      "learning_rate": 1.540540540540541e-05,
      "loss": 0.26,
      "step": 343
    },
    {
      "epoch": 0.23251098344035148,
      "grad_norm": 9.582365989685059,
      "learning_rate": 1.545045045045045e-05,
      "loss": 0.2364,
      "step": 344
    },
    {
      "epoch": 0.2331868874619804,
      "grad_norm": 24.658246994018555,
      "learning_rate": 1.5495495495495498e-05,
      "loss": 0.2834,
      "step": 345
    },
    {
      "epoch": 0.23386279148360933,
      "grad_norm": 10.297957420349121,
      "learning_rate": 1.554054054054054e-05,
      "loss": 0.2575,
      "step": 346
    },
    {
      "epoch": 0.23453869550523826,
      "grad_norm": 29.250194549560547,
      "learning_rate": 1.5585585585585587e-05,
      "loss": 0.2683,
      "step": 347
    },
    {
      "epoch": 0.23521459952686719,
      "grad_norm": 19.428937911987305,
      "learning_rate": 1.5630630630630633e-05,
      "loss": 0.2488,
      "step": 348
    },
    {
      "epoch": 0.2358905035484961,
      "grad_norm": 12.799714088439941,
      "learning_rate": 1.5675675675675676e-05,
      "loss": 0.2268,
      "step": 349
    },
    {
      "epoch": 0.23656640757012504,
      "grad_norm": 32.21827697753906,
      "learning_rate": 1.5720720720720722e-05,
      "loss": 0.2496,
      "step": 350
    },
    {
      "epoch": 0.23656640757012504,
      "eval_accuracy": 0.365625,
      "eval_loss": 0.2658956050872803,
      "eval_runtime": 45.0188,
      "eval_samples_per_second": 526.002,
      "eval_steps_per_second": 16.438,
      "step": 350
    },
    {
      "epoch": 0.23724231159175396,
      "grad_norm": 7.76975154876709,
      "learning_rate": 1.576576576576577e-05,
      "loss": 0.2745,
      "step": 351
    },
    {
      "epoch": 0.2379182156133829,
      "grad_norm": 9.59115219116211,
      "learning_rate": 1.581081081081081e-05,
      "loss": 0.2317,
      "step": 352
    },
    {
      "epoch": 0.23859411963501184,
      "grad_norm": 31.105690002441406,
      "learning_rate": 1.5855855855855858e-05,
      "loss": 0.2797,
      "step": 353
    },
    {
      "epoch": 0.23927002365664077,
      "grad_norm": 19.64571189880371,
      "learning_rate": 1.59009009009009e-05,
      "loss": 0.2724,
      "step": 354
    },
    {
      "epoch": 0.2399459276782697,
      "grad_norm": 17.50550651550293,
      "learning_rate": 1.5945945945945947e-05,
      "loss": 0.2522,
      "step": 355
    },
    {
      "epoch": 0.24062183169989862,
      "grad_norm": 13.670315742492676,
      "learning_rate": 1.5990990990990993e-05,
      "loss": 0.2731,
      "step": 356
    },
    {
      "epoch": 0.24129773572152755,
      "grad_norm": 29.79376220703125,
      "learning_rate": 1.6036036036036036e-05,
      "loss": 0.3266,
      "step": 357
    },
    {
      "epoch": 0.24197363974315647,
      "grad_norm": 14.058691024780273,
      "learning_rate": 1.6081081081081083e-05,
      "loss": 0.3472,
      "step": 358
    },
    {
      "epoch": 0.2426495437647854,
      "grad_norm": 27.8502254486084,
      "learning_rate": 1.6126126126126126e-05,
      "loss": 0.2877,
      "step": 359
    },
    {
      "epoch": 0.24332544778641432,
      "grad_norm": 26.8613224029541,
      "learning_rate": 1.6171171171171172e-05,
      "loss": 0.2827,
      "step": 360
    },
    {
      "epoch": 0.24400135180804325,
      "grad_norm": 37.26984786987305,
      "learning_rate": 1.6216216216216218e-05,
      "loss": 0.3178,
      "step": 361
    },
    {
      "epoch": 0.24467725582967217,
      "grad_norm": 17.115493774414062,
      "learning_rate": 1.6261261261261264e-05,
      "loss": 0.258,
      "step": 362
    },
    {
      "epoch": 0.24535315985130113,
      "grad_norm": 21.28546142578125,
      "learning_rate": 1.6306306306306307e-05,
      "loss": 0.2479,
      "step": 363
    },
    {
      "epoch": 0.24602906387293005,
      "grad_norm": 49.12176513671875,
      "learning_rate": 1.6351351351351354e-05,
      "loss": 0.3395,
      "step": 364
    },
    {
      "epoch": 0.24670496789455898,
      "grad_norm": 29.511077880859375,
      "learning_rate": 1.6396396396396396e-05,
      "loss": 0.2854,
      "step": 365
    },
    {
      "epoch": 0.2473808719161879,
      "grad_norm": 7.8312811851501465,
      "learning_rate": 1.6441441441441443e-05,
      "loss": 0.2634,
      "step": 366
    },
    {
      "epoch": 0.24805677593781683,
      "grad_norm": 24.444177627563477,
      "learning_rate": 1.648648648648649e-05,
      "loss": 0.2902,
      "step": 367
    },
    {
      "epoch": 0.24873267995944576,
      "grad_norm": 22.548418045043945,
      "learning_rate": 1.6531531531531532e-05,
      "loss": 0.27,
      "step": 368
    },
    {
      "epoch": 0.24940858398107468,
      "grad_norm": 22.425331115722656,
      "learning_rate": 1.6576576576576578e-05,
      "loss": 0.2709,
      "step": 369
    },
    {
      "epoch": 0.2500844880027036,
      "grad_norm": 20.650447845458984,
      "learning_rate": 1.662162162162162e-05,
      "loss": 0.2585,
      "step": 370
    },
    {
      "epoch": 0.25076039202433253,
      "grad_norm": 7.61582088470459,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.228,
      "step": 371
    },
    {
      "epoch": 0.25143629604596146,
      "grad_norm": 8.218512535095215,
      "learning_rate": 1.6711711711711714e-05,
      "loss": 0.2568,
      "step": 372
    },
    {
      "epoch": 0.2521122000675904,
      "grad_norm": 42.944435119628906,
      "learning_rate": 1.6756756756756757e-05,
      "loss": 0.2963,
      "step": 373
    },
    {
      "epoch": 0.2527881040892193,
      "grad_norm": 37.56005859375,
      "learning_rate": 1.6801801801801803e-05,
      "loss": 0.2857,
      "step": 374
    },
    {
      "epoch": 0.25346400811084824,
      "grad_norm": 6.39361047744751,
      "learning_rate": 1.6846846846846846e-05,
      "loss": 0.2583,
      "step": 375
    },
    {
      "epoch": 0.25413991213247716,
      "grad_norm": 8.943142890930176,
      "learning_rate": 1.6891891891891896e-05,
      "loss": 0.2495,
      "step": 376
    },
    {
      "epoch": 0.2548158161541061,
      "grad_norm": 7.621511459350586,
      "learning_rate": 1.693693693693694e-05,
      "loss": 0.2744,
      "step": 377
    },
    {
      "epoch": 0.25549172017573507,
      "grad_norm": 42.415279388427734,
      "learning_rate": 1.6981981981981985e-05,
      "loss": 0.3548,
      "step": 378
    },
    {
      "epoch": 0.256167624197364,
      "grad_norm": 53.698909759521484,
      "learning_rate": 1.7027027027027028e-05,
      "loss": 0.3523,
      "step": 379
    },
    {
      "epoch": 0.2568435282189929,
      "grad_norm": 11.290047645568848,
      "learning_rate": 1.7072072072072074e-05,
      "loss": 0.2489,
      "step": 380
    },
    {
      "epoch": 0.25751943224062185,
      "grad_norm": 15.774617195129395,
      "learning_rate": 1.711711711711712e-05,
      "loss": 0.2483,
      "step": 381
    },
    {
      "epoch": 0.2581953362622508,
      "grad_norm": 58.74515914916992,
      "learning_rate": 1.7162162162162163e-05,
      "loss": 0.3264,
      "step": 382
    },
    {
      "epoch": 0.2588712402838797,
      "grad_norm": 13.989523887634277,
      "learning_rate": 1.720720720720721e-05,
      "loss": 0.2496,
      "step": 383
    },
    {
      "epoch": 0.2595471443055086,
      "grad_norm": 15.255609512329102,
      "learning_rate": 1.7252252252252252e-05,
      "loss": 0.2417,
      "step": 384
    },
    {
      "epoch": 0.26022304832713755,
      "grad_norm": 15.812642097473145,
      "learning_rate": 1.72972972972973e-05,
      "loss": 0.2277,
      "step": 385
    },
    {
      "epoch": 0.2608989523487665,
      "grad_norm": 16.84299659729004,
      "learning_rate": 1.7342342342342345e-05,
      "loss": 0.2397,
      "step": 386
    },
    {
      "epoch": 0.2615748563703954,
      "grad_norm": 7.383591175079346,
      "learning_rate": 1.7387387387387388e-05,
      "loss": 0.243,
      "step": 387
    },
    {
      "epoch": 0.26225076039202433,
      "grad_norm": 7.706769943237305,
      "learning_rate": 1.7432432432432434e-05,
      "loss": 0.2272,
      "step": 388
    },
    {
      "epoch": 0.26292666441365325,
      "grad_norm": 37.56233596801758,
      "learning_rate": 1.7477477477477477e-05,
      "loss": 0.3438,
      "step": 389
    },
    {
      "epoch": 0.2636025684352822,
      "grad_norm": 12.62123966217041,
      "learning_rate": 1.7522522522522523e-05,
      "loss": 0.3444,
      "step": 390
    },
    {
      "epoch": 0.2642784724569111,
      "grad_norm": 23.13532829284668,
      "learning_rate": 1.756756756756757e-05,
      "loss": 0.2731,
      "step": 391
    },
    {
      "epoch": 0.26495437647854003,
      "grad_norm": 6.672341346740723,
      "learning_rate": 1.7612612612612613e-05,
      "loss": 0.2441,
      "step": 392
    },
    {
      "epoch": 0.26563028050016896,
      "grad_norm": 11.452704429626465,
      "learning_rate": 1.765765765765766e-05,
      "loss": 0.337,
      "step": 393
    },
    {
      "epoch": 0.2663061845217979,
      "grad_norm": 10.549775123596191,
      "learning_rate": 1.7702702702702702e-05,
      "loss": 0.259,
      "step": 394
    },
    {
      "epoch": 0.2669820885434268,
      "grad_norm": 16.266721725463867,
      "learning_rate": 1.774774774774775e-05,
      "loss": 0.2374,
      "step": 395
    },
    {
      "epoch": 0.26765799256505574,
      "grad_norm": 5.616227149963379,
      "learning_rate": 1.7792792792792794e-05,
      "loss": 0.291,
      "step": 396
    },
    {
      "epoch": 0.2683338965866847,
      "grad_norm": 7.635812282562256,
      "learning_rate": 1.783783783783784e-05,
      "loss": 0.2647,
      "step": 397
    },
    {
      "epoch": 0.26900980060831364,
      "grad_norm": 24.65031623840332,
      "learning_rate": 1.7882882882882883e-05,
      "loss": 0.2657,
      "step": 398
    },
    {
      "epoch": 0.26968570462994257,
      "grad_norm": 7.326444149017334,
      "learning_rate": 1.792792792792793e-05,
      "loss": 0.3077,
      "step": 399
    },
    {
      "epoch": 0.2703616086515715,
      "grad_norm": 18.758281707763672,
      "learning_rate": 1.7972972972972976e-05,
      "loss": 0.2451,
      "step": 400
    },
    {
      "epoch": 0.2703616086515715,
      "eval_accuracy": 0.40388513513513513,
      "eval_loss": 0.27211156487464905,
      "eval_runtime": 44.1599,
      "eval_samples_per_second": 536.233,
      "eval_steps_per_second": 16.757,
      "step": 400
    },
    {
      "epoch": 0.2710375126732004,
      "grad_norm": 23.268220901489258,
      "learning_rate": 1.801801801801802e-05,
      "loss": 0.2329,
      "step": 401
    },
    {
      "epoch": 0.27171341669482935,
      "grad_norm": 10.532570838928223,
      "learning_rate": 1.8063063063063065e-05,
      "loss": 0.2287,
      "step": 402
    },
    {
      "epoch": 0.27238932071645827,
      "grad_norm": 14.497779846191406,
      "learning_rate": 1.8108108108108108e-05,
      "loss": 0.2263,
      "step": 403
    },
    {
      "epoch": 0.2730652247380872,
      "grad_norm": 5.405073642730713,
      "learning_rate": 1.8153153153153154e-05,
      "loss": 0.2471,
      "step": 404
    },
    {
      "epoch": 0.2737411287597161,
      "grad_norm": 9.511168479919434,
      "learning_rate": 1.81981981981982e-05,
      "loss": 0.2489,
      "step": 405
    },
    {
      "epoch": 0.27441703278134505,
      "grad_norm": 19.796533584594727,
      "learning_rate": 1.8243243243243244e-05,
      "loss": 0.2113,
      "step": 406
    },
    {
      "epoch": 0.275092936802974,
      "grad_norm": 17.49905776977539,
      "learning_rate": 1.828828828828829e-05,
      "loss": 0.222,
      "step": 407
    },
    {
      "epoch": 0.2757688408246029,
      "grad_norm": 10.624908447265625,
      "learning_rate": 1.8333333333333333e-05,
      "loss": 0.282,
      "step": 408
    },
    {
      "epoch": 0.2764447448462318,
      "grad_norm": 8.585684776306152,
      "learning_rate": 1.8378378378378383e-05,
      "loss": 0.2635,
      "step": 409
    },
    {
      "epoch": 0.27712064886786075,
      "grad_norm": 33.59832763671875,
      "learning_rate": 1.8423423423423425e-05,
      "loss": 0.3366,
      "step": 410
    },
    {
      "epoch": 0.2777965528894897,
      "grad_norm": 21.95138931274414,
      "learning_rate": 1.8468468468468472e-05,
      "loss": 0.2568,
      "step": 411
    },
    {
      "epoch": 0.2784724569111186,
      "grad_norm": 9.23679256439209,
      "learning_rate": 1.8513513513513515e-05,
      "loss": 0.2536,
      "step": 412
    },
    {
      "epoch": 0.27914836093274753,
      "grad_norm": 11.463278770446777,
      "learning_rate": 1.855855855855856e-05,
      "loss": 0.2582,
      "step": 413
    },
    {
      "epoch": 0.27982426495437646,
      "grad_norm": 30.162878036499023,
      "learning_rate": 1.8603603603603607e-05,
      "loss": 0.2767,
      "step": 414
    },
    {
      "epoch": 0.2805001689760054,
      "grad_norm": 5.956443786621094,
      "learning_rate": 1.864864864864865e-05,
      "loss": 0.2357,
      "step": 415
    },
    {
      "epoch": 0.28117607299763436,
      "grad_norm": 7.451732158660889,
      "learning_rate": 1.8693693693693696e-05,
      "loss": 0.2838,
      "step": 416
    },
    {
      "epoch": 0.2818519770192633,
      "grad_norm": 14.508785247802734,
      "learning_rate": 1.873873873873874e-05,
      "loss": 0.2254,
      "step": 417
    },
    {
      "epoch": 0.2825278810408922,
      "grad_norm": 10.257606506347656,
      "learning_rate": 1.8783783783783786e-05,
      "loss": 0.2379,
      "step": 418
    },
    {
      "epoch": 0.28320378506252114,
      "grad_norm": 12.107852935791016,
      "learning_rate": 1.8828828828828832e-05,
      "loss": 0.2419,
      "step": 419
    },
    {
      "epoch": 0.28387968908415007,
      "grad_norm": 15.45522403717041,
      "learning_rate": 1.8873873873873875e-05,
      "loss": 0.2792,
      "step": 420
    },
    {
      "epoch": 0.284555593105779,
      "grad_norm": 8.541949272155762,
      "learning_rate": 1.891891891891892e-05,
      "loss": 0.308,
      "step": 421
    },
    {
      "epoch": 0.2852314971274079,
      "grad_norm": 12.07388687133789,
      "learning_rate": 1.8963963963963964e-05,
      "loss": 0.2655,
      "step": 422
    },
    {
      "epoch": 0.28590740114903684,
      "grad_norm": 5.895168304443359,
      "learning_rate": 1.900900900900901e-05,
      "loss": 0.2653,
      "step": 423
    },
    {
      "epoch": 0.28658330517066577,
      "grad_norm": 6.245179653167725,
      "learning_rate": 1.9054054054054057e-05,
      "loss": 0.264,
      "step": 424
    },
    {
      "epoch": 0.2872592091922947,
      "grad_norm": 4.858404636383057,
      "learning_rate": 1.90990990990991e-05,
      "loss": 0.2079,
      "step": 425
    },
    {
      "epoch": 0.2879351132139236,
      "grad_norm": 8.238146781921387,
      "learning_rate": 1.9144144144144146e-05,
      "loss": 0.2633,
      "step": 426
    },
    {
      "epoch": 0.28861101723555255,
      "grad_norm": 9.136231422424316,
      "learning_rate": 1.918918918918919e-05,
      "loss": 0.2306,
      "step": 427
    },
    {
      "epoch": 0.2892869212571815,
      "grad_norm": 12.827601432800293,
      "learning_rate": 1.9234234234234235e-05,
      "loss": 0.2667,
      "step": 428
    },
    {
      "epoch": 0.2899628252788104,
      "grad_norm": 17.73537826538086,
      "learning_rate": 1.927927927927928e-05,
      "loss": 0.2328,
      "step": 429
    },
    {
      "epoch": 0.2906387293004393,
      "grad_norm": 16.14219093322754,
      "learning_rate": 1.9324324324324328e-05,
      "loss": 0.2266,
      "step": 430
    },
    {
      "epoch": 0.29131463332206825,
      "grad_norm": 27.157175064086914,
      "learning_rate": 1.936936936936937e-05,
      "loss": 0.2854,
      "step": 431
    },
    {
      "epoch": 0.2919905373436972,
      "grad_norm": 22.319698333740234,
      "learning_rate": 1.9414414414414417e-05,
      "loss": 0.2544,
      "step": 432
    },
    {
      "epoch": 0.2926664413653261,
      "grad_norm": 7.77006196975708,
      "learning_rate": 1.9459459459459463e-05,
      "loss": 0.2533,
      "step": 433
    },
    {
      "epoch": 0.29334234538695503,
      "grad_norm": 23.4996395111084,
      "learning_rate": 1.9504504504504506e-05,
      "loss": 0.2633,
      "step": 434
    },
    {
      "epoch": 0.29401824940858395,
      "grad_norm": 7.7030463218688965,
      "learning_rate": 1.9549549549549552e-05,
      "loss": 0.2316,
      "step": 435
    },
    {
      "epoch": 0.29469415343021294,
      "grad_norm": 17.657413482666016,
      "learning_rate": 1.9594594594594595e-05,
      "loss": 0.2797,
      "step": 436
    },
    {
      "epoch": 0.29537005745184186,
      "grad_norm": 9.449223518371582,
      "learning_rate": 1.963963963963964e-05,
      "loss": 0.2587,
      "step": 437
    },
    {
      "epoch": 0.2960459614734708,
      "grad_norm": 23.927005767822266,
      "learning_rate": 1.9684684684684688e-05,
      "loss": 0.2658,
      "step": 438
    },
    {
      "epoch": 0.2967218654950997,
      "grad_norm": 14.248490333557129,
      "learning_rate": 1.972972972972973e-05,
      "loss": 0.2449,
      "step": 439
    },
    {
      "epoch": 0.29739776951672864,
      "grad_norm": 17.95062255859375,
      "learning_rate": 1.9774774774774777e-05,
      "loss": 0.2118,
      "step": 440
    },
    {
      "epoch": 0.29807367353835756,
      "grad_norm": 18.750036239624023,
      "learning_rate": 1.981981981981982e-05,
      "loss": 0.3016,
      "step": 441
    },
    {
      "epoch": 0.2987495775599865,
      "grad_norm": 19.29686164855957,
      "learning_rate": 1.9864864864864866e-05,
      "loss": 0.234,
      "step": 442
    },
    {
      "epoch": 0.2994254815816154,
      "grad_norm": 35.551090240478516,
      "learning_rate": 1.9909909909909912e-05,
      "loss": 0.2437,
      "step": 443
    },
    {
      "epoch": 0.30010138560324434,
      "grad_norm": 4.404517650604248,
      "learning_rate": 1.995495495495496e-05,
      "loss": 0.2588,
      "step": 444
    },
    {
      "epoch": 0.30077728962487327,
      "grad_norm": 7.432286262512207,
      "learning_rate": 2e-05,
      "loss": 0.2348,
      "step": 445
    },
    {
      "epoch": 0.3014531936465022,
      "grad_norm": 20.97759246826172,
      "learning_rate": 2e-05,
      "loss": 0.3221,
      "step": 446
    },
    {
      "epoch": 0.3021290976681311,
      "grad_norm": 55.454044342041016,
      "learning_rate": 2e-05,
      "loss": 0.3513,
      "step": 447
    },
    {
      "epoch": 0.30280500168976004,
      "grad_norm": 18.514883041381836,
      "learning_rate": 2e-05,
      "loss": 0.3177,
      "step": 448
    },
    {
      "epoch": 0.30348090571138897,
      "grad_norm": 25.97433853149414,
      "learning_rate": 2e-05,
      "loss": 0.2326,
      "step": 449
    },
    {
      "epoch": 0.3041568097330179,
      "grad_norm": 24.785057067871094,
      "learning_rate": 2e-05,
      "loss": 0.3131,
      "step": 450
    },
    {
      "epoch": 0.3041568097330179,
      "eval_accuracy": 0.47478885135135135,
      "eval_loss": 0.272676944732666,
      "eval_runtime": 43.394,
      "eval_samples_per_second": 545.697,
      "eval_steps_per_second": 17.053,
      "step": 450
    },
    {
      "epoch": 0.3048327137546468,
      "grad_norm": 28.057764053344727,
      "learning_rate": 2e-05,
      "loss": 0.2544,
      "step": 451
    },
    {
      "epoch": 0.30550861777627575,
      "grad_norm": 19.754398345947266,
      "learning_rate": 2e-05,
      "loss": 0.231,
      "step": 452
    },
    {
      "epoch": 0.3061845217979047,
      "grad_norm": 5.899191379547119,
      "learning_rate": 2e-05,
      "loss": 0.2326,
      "step": 453
    },
    {
      "epoch": 0.3068604258195336,
      "grad_norm": 6.860384941101074,
      "learning_rate": 2e-05,
      "loss": 0.2406,
      "step": 454
    },
    {
      "epoch": 0.3075363298411626,
      "grad_norm": 5.863921642303467,
      "learning_rate": 2e-05,
      "loss": 0.2683,
      "step": 455
    },
    {
      "epoch": 0.3082122338627915,
      "grad_norm": 11.171703338623047,
      "learning_rate": 2e-05,
      "loss": 0.2558,
      "step": 456
    },
    {
      "epoch": 0.30888813788442043,
      "grad_norm": 7.568742275238037,
      "learning_rate": 2e-05,
      "loss": 0.2405,
      "step": 457
    },
    {
      "epoch": 0.30956404190604936,
      "grad_norm": 8.397537231445312,
      "learning_rate": 2e-05,
      "loss": 0.2659,
      "step": 458
    },
    {
      "epoch": 0.3102399459276783,
      "grad_norm": 10.566232681274414,
      "learning_rate": 2e-05,
      "loss": 0.2918,
      "step": 459
    },
    {
      "epoch": 0.3109158499493072,
      "grad_norm": 32.572715759277344,
      "learning_rate": 2e-05,
      "loss": 0.2729,
      "step": 460
    },
    {
      "epoch": 0.31159175397093614,
      "grad_norm": 13.802708625793457,
      "learning_rate": 2e-05,
      "loss": 0.2899,
      "step": 461
    },
    {
      "epoch": 0.31226765799256506,
      "grad_norm": 14.609110832214355,
      "learning_rate": 2e-05,
      "loss": 0.2374,
      "step": 462
    },
    {
      "epoch": 0.312943562014194,
      "grad_norm": 12.26738452911377,
      "learning_rate": 2e-05,
      "loss": 0.2569,
      "step": 463
    },
    {
      "epoch": 0.3136194660358229,
      "grad_norm": 25.67567253112793,
      "learning_rate": 2e-05,
      "loss": 0.3,
      "step": 464
    },
    {
      "epoch": 0.31429537005745184,
      "grad_norm": 6.434136390686035,
      "learning_rate": 2e-05,
      "loss": 0.2516,
      "step": 465
    },
    {
      "epoch": 0.31497127407908077,
      "grad_norm": 29.788284301757812,
      "learning_rate": 2e-05,
      "loss": 0.3344,
      "step": 466
    },
    {
      "epoch": 0.3156471781007097,
      "grad_norm": 38.170654296875,
      "learning_rate": 2e-05,
      "loss": 0.3025,
      "step": 467
    },
    {
      "epoch": 0.3163230821223386,
      "grad_norm": 23.279903411865234,
      "learning_rate": 2e-05,
      "loss": 0.2888,
      "step": 468
    },
    {
      "epoch": 0.31699898614396754,
      "grad_norm": 6.785585880279541,
      "learning_rate": 2e-05,
      "loss": 0.2765,
      "step": 469
    },
    {
      "epoch": 0.31767489016559647,
      "grad_norm": 27.856338500976562,
      "learning_rate": 2e-05,
      "loss": 0.2623,
      "step": 470
    },
    {
      "epoch": 0.3183507941872254,
      "grad_norm": 51.53744125366211,
      "learning_rate": 2e-05,
      "loss": 0.3256,
      "step": 471
    },
    {
      "epoch": 0.3190266982088543,
      "grad_norm": 30.760831832885742,
      "learning_rate": 2e-05,
      "loss": 0.2833,
      "step": 472
    },
    {
      "epoch": 0.31970260223048325,
      "grad_norm": 24.788848876953125,
      "learning_rate": 2e-05,
      "loss": 0.2758,
      "step": 473
    },
    {
      "epoch": 0.3203785062521122,
      "grad_norm": 29.553756713867188,
      "learning_rate": 2e-05,
      "loss": 0.2542,
      "step": 474
    },
    {
      "epoch": 0.32105441027374115,
      "grad_norm": 44.220401763916016,
      "learning_rate": 2e-05,
      "loss": 0.3395,
      "step": 475
    },
    {
      "epoch": 0.3217303142953701,
      "grad_norm": 35.82133865356445,
      "learning_rate": 2e-05,
      "loss": 0.3345,
      "step": 476
    },
    {
      "epoch": 0.322406218316999,
      "grad_norm": 16.727977752685547,
      "learning_rate": 2e-05,
      "loss": 0.269,
      "step": 477
    },
    {
      "epoch": 0.32308212233862793,
      "grad_norm": 25.473691940307617,
      "learning_rate": 2e-05,
      "loss": 0.2413,
      "step": 478
    },
    {
      "epoch": 0.32375802636025686,
      "grad_norm": 24.619306564331055,
      "learning_rate": 2e-05,
      "loss": 0.2883,
      "step": 479
    },
    {
      "epoch": 0.3244339303818858,
      "grad_norm": 19.292226791381836,
      "learning_rate": 2e-05,
      "loss": 0.2488,
      "step": 480
    },
    {
      "epoch": 0.3251098344035147,
      "grad_norm": 20.396587371826172,
      "learning_rate": 2e-05,
      "loss": 0.303,
      "step": 481
    },
    {
      "epoch": 0.32578573842514363,
      "grad_norm": 11.841169357299805,
      "learning_rate": 2e-05,
      "loss": 0.2703,
      "step": 482
    },
    {
      "epoch": 0.32646164244677256,
      "grad_norm": 11.631719589233398,
      "learning_rate": 2e-05,
      "loss": 0.2367,
      "step": 483
    },
    {
      "epoch": 0.3271375464684015,
      "grad_norm": 23.28961753845215,
      "learning_rate": 2e-05,
      "loss": 0.2501,
      "step": 484
    },
    {
      "epoch": 0.3278134504900304,
      "grad_norm": 9.012608528137207,
      "learning_rate": 2e-05,
      "loss": 0.2984,
      "step": 485
    },
    {
      "epoch": 0.32848935451165934,
      "grad_norm": 9.729747772216797,
      "learning_rate": 2e-05,
      "loss": 0.2402,
      "step": 486
    },
    {
      "epoch": 0.32916525853328826,
      "grad_norm": 18.522336959838867,
      "learning_rate": 2e-05,
      "loss": 0.2629,
      "step": 487
    },
    {
      "epoch": 0.3298411625549172,
      "grad_norm": 19.287261962890625,
      "learning_rate": 2e-05,
      "loss": 0.2477,
      "step": 488
    },
    {
      "epoch": 0.3305170665765461,
      "grad_norm": 11.32188892364502,
      "learning_rate": 2e-05,
      "loss": 0.201,
      "step": 489
    },
    {
      "epoch": 0.33119297059817504,
      "grad_norm": 26.91994285583496,
      "learning_rate": 2e-05,
      "loss": 0.2669,
      "step": 490
    },
    {
      "epoch": 0.33186887461980397,
      "grad_norm": 13.397052764892578,
      "learning_rate": 2e-05,
      "loss": 0.2532,
      "step": 491
    },
    {
      "epoch": 0.3325447786414329,
      "grad_norm": 6.4740800857543945,
      "learning_rate": 2e-05,
      "loss": 0.168,
      "step": 492
    },
    {
      "epoch": 0.3332206826630618,
      "grad_norm": 28.9803409576416,
      "learning_rate": 2e-05,
      "loss": 0.3166,
      "step": 493
    },
    {
      "epoch": 0.3338965866846908,
      "grad_norm": 20.26554298400879,
      "learning_rate": 2e-05,
      "loss": 0.329,
      "step": 494
    },
    {
      "epoch": 0.3345724907063197,
      "grad_norm": 31.221900939941406,
      "learning_rate": 2e-05,
      "loss": 0.2801,
      "step": 495
    },
    {
      "epoch": 0.33524839472794865,
      "grad_norm": 3.933056354522705,
      "learning_rate": 2e-05,
      "loss": 0.2684,
      "step": 496
    },
    {
      "epoch": 0.3359242987495776,
      "grad_norm": 10.514286041259766,
      "learning_rate": 2e-05,
      "loss": 0.2368,
      "step": 497
    },
    {
      "epoch": 0.3366002027712065,
      "grad_norm": 22.384057998657227,
      "learning_rate": 2e-05,
      "loss": 0.2351,
      "step": 498
    },
    {
      "epoch": 0.33727610679283543,
      "grad_norm": 19.174257278442383,
      "learning_rate": 2e-05,
      "loss": 0.2441,
      "step": 499
    },
    {
      "epoch": 0.33795201081446435,
      "grad_norm": 21.364519119262695,
      "learning_rate": 2e-05,
      "loss": 0.2621,
      "step": 500
    },
    {
      "epoch": 0.33795201081446435,
      "eval_accuracy": 0.4654138513513513,
      "eval_loss": 0.2535924017429352,
      "eval_runtime": 43.2218,
      "eval_samples_per_second": 547.872,
      "eval_steps_per_second": 17.121,
      "step": 500
    },
    {
      "epoch": 0.3386279148360933,
      "grad_norm": 10.324834823608398,
      "learning_rate": 2e-05,
      "loss": 0.2332,
      "step": 501
    },
    {
      "epoch": 0.3393038188577222,
      "grad_norm": 4.294429302215576,
      "learning_rate": 2e-05,
      "loss": 0.2539,
      "step": 502
    },
    {
      "epoch": 0.33997972287935113,
      "grad_norm": 22.03772735595703,
      "learning_rate": 2e-05,
      "loss": 0.226,
      "step": 503
    },
    {
      "epoch": 0.34065562690098006,
      "grad_norm": 20.75609016418457,
      "learning_rate": 2e-05,
      "loss": 0.2355,
      "step": 504
    },
    {
      "epoch": 0.341331530922609,
      "grad_norm": 28.592885971069336,
      "learning_rate": 2e-05,
      "loss": 0.3042,
      "step": 505
    },
    {
      "epoch": 0.3420074349442379,
      "grad_norm": 15.966550827026367,
      "learning_rate": 2e-05,
      "loss": 0.2388,
      "step": 506
    },
    {
      "epoch": 0.34268333896586683,
      "grad_norm": 8.240887641906738,
      "learning_rate": 2e-05,
      "loss": 0.2442,
      "step": 507
    },
    {
      "epoch": 0.34335924298749576,
      "grad_norm": 46.294212341308594,
      "learning_rate": 2e-05,
      "loss": 0.363,
      "step": 508
    },
    {
      "epoch": 0.3440351470091247,
      "grad_norm": 53.40895080566406,
      "learning_rate": 2e-05,
      "loss": 0.3637,
      "step": 509
    },
    {
      "epoch": 0.3447110510307536,
      "grad_norm": 4.5886549949646,
      "learning_rate": 2e-05,
      "loss": 0.2533,
      "step": 510
    },
    {
      "epoch": 0.34538695505238254,
      "grad_norm": 18.716684341430664,
      "learning_rate": 2e-05,
      "loss": 0.2403,
      "step": 511
    },
    {
      "epoch": 0.34606285907401146,
      "grad_norm": 28.711111068725586,
      "learning_rate": 2e-05,
      "loss": 0.2773,
      "step": 512
    },
    {
      "epoch": 0.34673876309564045,
      "grad_norm": 38.96063995361328,
      "learning_rate": 2e-05,
      "loss": 0.3075,
      "step": 513
    },
    {
      "epoch": 0.34741466711726937,
      "grad_norm": 24.21227264404297,
      "learning_rate": 2e-05,
      "loss": 0.2834,
      "step": 514
    },
    {
      "epoch": 0.3480905711388983,
      "grad_norm": 9.264009475708008,
      "learning_rate": 2e-05,
      "loss": 0.2215,
      "step": 515
    },
    {
      "epoch": 0.3487664751605272,
      "grad_norm": 10.346829414367676,
      "learning_rate": 2e-05,
      "loss": 0.2889,
      "step": 516
    },
    {
      "epoch": 0.34944237918215615,
      "grad_norm": 40.69310760498047,
      "learning_rate": 2e-05,
      "loss": 0.3134,
      "step": 517
    },
    {
      "epoch": 0.3501182832037851,
      "grad_norm": 10.39887809753418,
      "learning_rate": 2e-05,
      "loss": 0.2537,
      "step": 518
    },
    {
      "epoch": 0.350794187225414,
      "grad_norm": 4.733892917633057,
      "learning_rate": 2e-05,
      "loss": 0.2403,
      "step": 519
    },
    {
      "epoch": 0.3514700912470429,
      "grad_norm": 4.56150484085083,
      "learning_rate": 2e-05,
      "loss": 0.2538,
      "step": 520
    },
    {
      "epoch": 0.35214599526867185,
      "grad_norm": 10.545814514160156,
      "learning_rate": 2e-05,
      "loss": 0.2319,
      "step": 521
    },
    {
      "epoch": 0.3528218992903008,
      "grad_norm": 16.684093475341797,
      "learning_rate": 2e-05,
      "loss": 0.2598,
      "step": 522
    },
    {
      "epoch": 0.3534978033119297,
      "grad_norm": 11.857154846191406,
      "learning_rate": 2e-05,
      "loss": 0.2257,
      "step": 523
    },
    {
      "epoch": 0.35417370733355863,
      "grad_norm": 16.423831939697266,
      "learning_rate": 2e-05,
      "loss": 0.2546,
      "step": 524
    },
    {
      "epoch": 0.35484961135518756,
      "grad_norm": 16.235641479492188,
      "learning_rate": 2e-05,
      "loss": 0.2806,
      "step": 525
    },
    {
      "epoch": 0.3555255153768165,
      "grad_norm": 4.872950553894043,
      "learning_rate": 2e-05,
      "loss": 0.2259,
      "step": 526
    },
    {
      "epoch": 0.3562014193984454,
      "grad_norm": 20.429086685180664,
      "learning_rate": 2e-05,
      "loss": 0.2459,
      "step": 527
    },
    {
      "epoch": 0.35687732342007433,
      "grad_norm": 10.833683967590332,
      "learning_rate": 2e-05,
      "loss": 0.2715,
      "step": 528
    },
    {
      "epoch": 0.35755322744170326,
      "grad_norm": 27.32050323486328,
      "learning_rate": 2e-05,
      "loss": 0.2504,
      "step": 529
    },
    {
      "epoch": 0.3582291314633322,
      "grad_norm": 10.801417350769043,
      "learning_rate": 2e-05,
      "loss": 0.2944,
      "step": 530
    },
    {
      "epoch": 0.3589050354849611,
      "grad_norm": 8.113250732421875,
      "learning_rate": 2e-05,
      "loss": 0.2356,
      "step": 531
    },
    {
      "epoch": 0.3595809395065901,
      "grad_norm": 7.069101333618164,
      "learning_rate": 2e-05,
      "loss": 0.2418,
      "step": 532
    },
    {
      "epoch": 0.360256843528219,
      "grad_norm": 12.299847602844238,
      "learning_rate": 2e-05,
      "loss": 0.2336,
      "step": 533
    },
    {
      "epoch": 0.36093274754984794,
      "grad_norm": 5.483318328857422,
      "learning_rate": 2e-05,
      "loss": 0.2068,
      "step": 534
    },
    {
      "epoch": 0.36160865157147687,
      "grad_norm": 17.276090621948242,
      "learning_rate": 2e-05,
      "loss": 0.2479,
      "step": 535
    },
    {
      "epoch": 0.3622845555931058,
      "grad_norm": 10.74336051940918,
      "learning_rate": 2e-05,
      "loss": 0.1883,
      "step": 536
    },
    {
      "epoch": 0.3629604596147347,
      "grad_norm": 4.883366584777832,
      "learning_rate": 2e-05,
      "loss": 0.1903,
      "step": 537
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 11.53194808959961,
      "learning_rate": 2e-05,
      "loss": 0.2452,
      "step": 538
    },
    {
      "epoch": 0.3643122676579926,
      "grad_norm": 22.010610580444336,
      "learning_rate": 2e-05,
      "loss": 0.2808,
      "step": 539
    },
    {
      "epoch": 0.3649881716796215,
      "grad_norm": 26.227867126464844,
      "learning_rate": 2e-05,
      "loss": 0.3323,
      "step": 540
    },
    {
      "epoch": 0.3656640757012504,
      "grad_norm": 9.882966995239258,
      "learning_rate": 2e-05,
      "loss": 0.2273,
      "step": 541
    },
    {
      "epoch": 0.36633997972287935,
      "grad_norm": 13.576104164123535,
      "learning_rate": 2e-05,
      "loss": 0.2463,
      "step": 542
    },
    {
      "epoch": 0.3670158837445083,
      "grad_norm": 10.270561218261719,
      "learning_rate": 2e-05,
      "loss": 0.2453,
      "step": 543
    },
    {
      "epoch": 0.3676917877661372,
      "grad_norm": 8.18601131439209,
      "learning_rate": 2e-05,
      "loss": 0.1781,
      "step": 544
    },
    {
      "epoch": 0.3683676917877661,
      "grad_norm": 24.564790725708008,
      "learning_rate": 2e-05,
      "loss": 0.2773,
      "step": 545
    },
    {
      "epoch": 0.36904359580939505,
      "grad_norm": 14.89745044708252,
      "learning_rate": 2e-05,
      "loss": 0.2435,
      "step": 546
    },
    {
      "epoch": 0.369719499831024,
      "grad_norm": 18.493202209472656,
      "learning_rate": 2e-05,
      "loss": 0.2691,
      "step": 547
    },
    {
      "epoch": 0.3703954038526529,
      "grad_norm": 27.84490966796875,
      "learning_rate": 2e-05,
      "loss": 0.2934,
      "step": 548
    },
    {
      "epoch": 0.37107130787428183,
      "grad_norm": 7.056121826171875,
      "learning_rate": 2e-05,
      "loss": 0.2422,
      "step": 549
    },
    {
      "epoch": 0.37174721189591076,
      "grad_norm": 6.907043933868408,
      "learning_rate": 2e-05,
      "loss": 0.2184,
      "step": 550
    },
    {
      "epoch": 0.37174721189591076,
      "eval_accuracy": 0.4627956081081081,
      "eval_loss": 0.2509574592113495,
      "eval_runtime": 51.4512,
      "eval_samples_per_second": 460.242,
      "eval_steps_per_second": 14.383,
      "step": 550
    },
    {
      "epoch": 0.3724231159175397,
      "grad_norm": 15.103052139282227,
      "learning_rate": 2e-05,
      "loss": 0.236,
      "step": 551
    },
    {
      "epoch": 0.37309901993916866,
      "grad_norm": 14.022075653076172,
      "learning_rate": 2e-05,
      "loss": 0.2299,
      "step": 552
    },
    {
      "epoch": 0.3737749239607976,
      "grad_norm": 14.494584083557129,
      "learning_rate": 2e-05,
      "loss": 0.2214,
      "step": 553
    },
    {
      "epoch": 0.3744508279824265,
      "grad_norm": 20.264755249023438,
      "learning_rate": 2e-05,
      "loss": 0.2524,
      "step": 554
    },
    {
      "epoch": 0.37512673200405544,
      "grad_norm": 14.280316352844238,
      "learning_rate": 2e-05,
      "loss": 0.2016,
      "step": 555
    },
    {
      "epoch": 0.37580263602568437,
      "grad_norm": 17.02768898010254,
      "learning_rate": 2e-05,
      "loss": 0.2624,
      "step": 556
    },
    {
      "epoch": 0.3764785400473133,
      "grad_norm": 17.18401527404785,
      "learning_rate": 2e-05,
      "loss": 0.2074,
      "step": 557
    },
    {
      "epoch": 0.3771544440689422,
      "grad_norm": 12.013110160827637,
      "learning_rate": 2e-05,
      "loss": 0.2172,
      "step": 558
    },
    {
      "epoch": 0.37783034809057114,
      "grad_norm": 45.43142318725586,
      "learning_rate": 2e-05,
      "loss": 0.3114,
      "step": 559
    },
    {
      "epoch": 0.37850625211220007,
      "grad_norm": 31.46410369873047,
      "learning_rate": 2e-05,
      "loss": 0.3562,
      "step": 560
    },
    {
      "epoch": 0.379182156133829,
      "grad_norm": 7.560055732727051,
      "learning_rate": 2e-05,
      "loss": 0.2009,
      "step": 561
    },
    {
      "epoch": 0.3798580601554579,
      "grad_norm": 34.85483932495117,
      "learning_rate": 2e-05,
      "loss": 0.3169,
      "step": 562
    },
    {
      "epoch": 0.38053396417708685,
      "grad_norm": 32.60721206665039,
      "learning_rate": 2e-05,
      "loss": 0.3027,
      "step": 563
    },
    {
      "epoch": 0.3812098681987158,
      "grad_norm": 34.28227233886719,
      "learning_rate": 2e-05,
      "loss": 0.3025,
      "step": 564
    },
    {
      "epoch": 0.3818857722203447,
      "grad_norm": 5.738337516784668,
      "learning_rate": 2e-05,
      "loss": 0.2175,
      "step": 565
    },
    {
      "epoch": 0.3825616762419736,
      "grad_norm": 10.929729461669922,
      "learning_rate": 2e-05,
      "loss": 0.2531,
      "step": 566
    },
    {
      "epoch": 0.38323758026360255,
      "grad_norm": 37.84080123901367,
      "learning_rate": 2e-05,
      "loss": 0.2864,
      "step": 567
    },
    {
      "epoch": 0.3839134842852315,
      "grad_norm": 14.876299858093262,
      "learning_rate": 2e-05,
      "loss": 0.2552,
      "step": 568
    },
    {
      "epoch": 0.3845893883068604,
      "grad_norm": 11.946091651916504,
      "learning_rate": 2e-05,
      "loss": 0.2564,
      "step": 569
    },
    {
      "epoch": 0.38526529232848933,
      "grad_norm": 13.906609535217285,
      "learning_rate": 2e-05,
      "loss": 0.2865,
      "step": 570
    },
    {
      "epoch": 0.3859411963501183,
      "grad_norm": 7.601094722747803,
      "learning_rate": 2e-05,
      "loss": 0.277,
      "step": 571
    },
    {
      "epoch": 0.38661710037174724,
      "grad_norm": 13.978620529174805,
      "learning_rate": 2e-05,
      "loss": 0.2629,
      "step": 572
    },
    {
      "epoch": 0.38729300439337616,
      "grad_norm": 17.96482276916504,
      "learning_rate": 2e-05,
      "loss": 0.2644,
      "step": 573
    },
    {
      "epoch": 0.3879689084150051,
      "grad_norm": 25.97359275817871,
      "learning_rate": 2e-05,
      "loss": 0.2462,
      "step": 574
    },
    {
      "epoch": 0.388644812436634,
      "grad_norm": 4.662927627563477,
      "learning_rate": 2e-05,
      "loss": 0.2608,
      "step": 575
    },
    {
      "epoch": 0.38932071645826294,
      "grad_norm": 15.310358047485352,
      "learning_rate": 2e-05,
      "loss": 0.2728,
      "step": 576
    },
    {
      "epoch": 0.38999662047989186,
      "grad_norm": 7.540192127227783,
      "learning_rate": 2e-05,
      "loss": 0.2595,
      "step": 577
    },
    {
      "epoch": 0.3906725245015208,
      "grad_norm": 8.407624244689941,
      "learning_rate": 2e-05,
      "loss": 0.1806,
      "step": 578
    },
    {
      "epoch": 0.3913484285231497,
      "grad_norm": 25.286590576171875,
      "learning_rate": 2e-05,
      "loss": 0.2563,
      "step": 579
    },
    {
      "epoch": 0.39202433254477864,
      "grad_norm": 10.83963394165039,
      "learning_rate": 2e-05,
      "loss": 0.2377,
      "step": 580
    },
    {
      "epoch": 0.39270023656640757,
      "grad_norm": 15.7402982711792,
      "learning_rate": 2e-05,
      "loss": 0.2643,
      "step": 581
    },
    {
      "epoch": 0.3933761405880365,
      "grad_norm": 10.31680679321289,
      "learning_rate": 2e-05,
      "loss": 0.2514,
      "step": 582
    },
    {
      "epoch": 0.3940520446096654,
      "grad_norm": 7.2291789054870605,
      "learning_rate": 2e-05,
      "loss": 0.2553,
      "step": 583
    },
    {
      "epoch": 0.39472794863129435,
      "grad_norm": 31.641695022583008,
      "learning_rate": 2e-05,
      "loss": 0.3109,
      "step": 584
    },
    {
      "epoch": 0.39540385265292327,
      "grad_norm": 7.274608612060547,
      "learning_rate": 2e-05,
      "loss": 0.2465,
      "step": 585
    },
    {
      "epoch": 0.3960797566745522,
      "grad_norm": 8.308979988098145,
      "learning_rate": 2e-05,
      "loss": 0.2004,
      "step": 586
    },
    {
      "epoch": 0.3967556606961811,
      "grad_norm": 13.090409278869629,
      "learning_rate": 2e-05,
      "loss": 0.2622,
      "step": 587
    },
    {
      "epoch": 0.39743156471781005,
      "grad_norm": 21.989341735839844,
      "learning_rate": 2e-05,
      "loss": 0.3026,
      "step": 588
    },
    {
      "epoch": 0.398107468739439,
      "grad_norm": 11.174470901489258,
      "learning_rate": 2e-05,
      "loss": 0.2487,
      "step": 589
    },
    {
      "epoch": 0.39878337276106796,
      "grad_norm": 12.85519027709961,
      "learning_rate": 2e-05,
      "loss": 0.2117,
      "step": 590
    },
    {
      "epoch": 0.3994592767826969,
      "grad_norm": 5.592171669006348,
      "learning_rate": 2e-05,
      "loss": 0.2059,
      "step": 591
    },
    {
      "epoch": 0.4001351808043258,
      "grad_norm": 19.291275024414062,
      "learning_rate": 2e-05,
      "loss": 0.3002,
      "step": 592
    },
    {
      "epoch": 0.40081108482595473,
      "grad_norm": 10.87536334991455,
      "learning_rate": 2e-05,
      "loss": 0.2177,
      "step": 593
    },
    {
      "epoch": 0.40148698884758366,
      "grad_norm": 13.620393753051758,
      "learning_rate": 2e-05,
      "loss": 0.2096,
      "step": 594
    },
    {
      "epoch": 0.4021628928692126,
      "grad_norm": 19.84151840209961,
      "learning_rate": 2e-05,
      "loss": 0.2124,
      "step": 595
    },
    {
      "epoch": 0.4028387968908415,
      "grad_norm": 28.7413330078125,
      "learning_rate": 2e-05,
      "loss": 0.2715,
      "step": 596
    },
    {
      "epoch": 0.40351470091247044,
      "grad_norm": 6.161742210388184,
      "learning_rate": 2e-05,
      "loss": 0.2378,
      "step": 597
    },
    {
      "epoch": 0.40419060493409936,
      "grad_norm": 23.92312240600586,
      "learning_rate": 2e-05,
      "loss": 0.2788,
      "step": 598
    },
    {
      "epoch": 0.4048665089557283,
      "grad_norm": 38.207054138183594,
      "learning_rate": 2e-05,
      "loss": 0.3485,
      "step": 599
    },
    {
      "epoch": 0.4055424129773572,
      "grad_norm": 11.446142196655273,
      "learning_rate": 2e-05,
      "loss": 0.2438,
      "step": 600
    },
    {
      "epoch": 0.4055424129773572,
      "eval_accuracy": 0.4564611486486487,
      "eval_loss": 0.25225114822387695,
      "eval_runtime": 53.6602,
      "eval_samples_per_second": 441.296,
      "eval_steps_per_second": 13.79,
      "step": 600
    },
    {
      "epoch": 0.40621831699898614,
      "grad_norm": 8.396985054016113,
      "learning_rate": 2e-05,
      "loss": 0.2007,
      "step": 601
    },
    {
      "epoch": 0.40689422102061507,
      "grad_norm": 33.24873733520508,
      "learning_rate": 2e-05,
      "loss": 0.3087,
      "step": 602
    },
    {
      "epoch": 0.407570125042244,
      "grad_norm": 8.559670448303223,
      "learning_rate": 2e-05,
      "loss": 0.2881,
      "step": 603
    },
    {
      "epoch": 0.4082460290638729,
      "grad_norm": 30.653242111206055,
      "learning_rate": 2e-05,
      "loss": 0.2822,
      "step": 604
    },
    {
      "epoch": 0.40892193308550184,
      "grad_norm": 20.757244110107422,
      "learning_rate": 2e-05,
      "loss": 0.2622,
      "step": 605
    },
    {
      "epoch": 0.40959783710713077,
      "grad_norm": 10.130572319030762,
      "learning_rate": 2e-05,
      "loss": 0.2831,
      "step": 606
    },
    {
      "epoch": 0.4102737411287597,
      "grad_norm": 31.593576431274414,
      "learning_rate": 2e-05,
      "loss": 0.2812,
      "step": 607
    },
    {
      "epoch": 0.4109496451503886,
      "grad_norm": 45.14284133911133,
      "learning_rate": 2e-05,
      "loss": 0.3113,
      "step": 608
    },
    {
      "epoch": 0.41162554917201755,
      "grad_norm": 6.411741256713867,
      "learning_rate": 2e-05,
      "loss": 0.2282,
      "step": 609
    },
    {
      "epoch": 0.4123014531936465,
      "grad_norm": 14.650525093078613,
      "learning_rate": 2e-05,
      "loss": 0.2224,
      "step": 610
    },
    {
      "epoch": 0.41297735721527545,
      "grad_norm": 8.40308666229248,
      "learning_rate": 2e-05,
      "loss": 0.2451,
      "step": 611
    },
    {
      "epoch": 0.4136532612369044,
      "grad_norm": 5.307476043701172,
      "learning_rate": 2e-05,
      "loss": 0.2452,
      "step": 612
    },
    {
      "epoch": 0.4143291652585333,
      "grad_norm": 5.4373321533203125,
      "learning_rate": 2e-05,
      "loss": 0.2246,
      "step": 613
    },
    {
      "epoch": 0.41500506928016223,
      "grad_norm": 23.124380111694336,
      "learning_rate": 2e-05,
      "loss": 0.2616,
      "step": 614
    },
    {
      "epoch": 0.41568097330179116,
      "grad_norm": 5.860323429107666,
      "learning_rate": 2e-05,
      "loss": 0.2058,
      "step": 615
    },
    {
      "epoch": 0.4163568773234201,
      "grad_norm": 7.813283443450928,
      "learning_rate": 2e-05,
      "loss": 0.2557,
      "step": 616
    },
    {
      "epoch": 0.417032781345049,
      "grad_norm": 5.725218772888184,
      "learning_rate": 2e-05,
      "loss": 0.2639,
      "step": 617
    },
    {
      "epoch": 0.41770868536667793,
      "grad_norm": 10.976724624633789,
      "learning_rate": 2e-05,
      "loss": 0.2711,
      "step": 618
    },
    {
      "epoch": 0.41838458938830686,
      "grad_norm": 7.5657639503479,
      "learning_rate": 2e-05,
      "loss": 0.1919,
      "step": 619
    },
    {
      "epoch": 0.4190604934099358,
      "grad_norm": 14.080645561218262,
      "learning_rate": 2e-05,
      "loss": 0.1778,
      "step": 620
    },
    {
      "epoch": 0.4197363974315647,
      "grad_norm": 17.534687042236328,
      "learning_rate": 2e-05,
      "loss": 0.2223,
      "step": 621
    },
    {
      "epoch": 0.42041230145319364,
      "grad_norm": 19.80284309387207,
      "learning_rate": 2e-05,
      "loss": 0.2677,
      "step": 622
    },
    {
      "epoch": 0.42108820547482256,
      "grad_norm": 7.196826934814453,
      "learning_rate": 2e-05,
      "loss": 0.1902,
      "step": 623
    },
    {
      "epoch": 0.4217641094964515,
      "grad_norm": 11.361509323120117,
      "learning_rate": 2e-05,
      "loss": 0.2673,
      "step": 624
    },
    {
      "epoch": 0.4224400135180804,
      "grad_norm": 10.464097023010254,
      "learning_rate": 2e-05,
      "loss": 0.2294,
      "step": 625
    },
    {
      "epoch": 0.42311591753970934,
      "grad_norm": 17.24912452697754,
      "learning_rate": 2e-05,
      "loss": 0.2189,
      "step": 626
    },
    {
      "epoch": 0.42379182156133827,
      "grad_norm": 15.688982009887695,
      "learning_rate": 2e-05,
      "loss": 0.2635,
      "step": 627
    },
    {
      "epoch": 0.4244677255829672,
      "grad_norm": 8.804266929626465,
      "learning_rate": 2e-05,
      "loss": 0.2508,
      "step": 628
    },
    {
      "epoch": 0.4251436296045962,
      "grad_norm": 13.522945404052734,
      "learning_rate": 2e-05,
      "loss": 0.2155,
      "step": 629
    },
    {
      "epoch": 0.4258195336262251,
      "grad_norm": 12.085273742675781,
      "learning_rate": 2e-05,
      "loss": 0.1931,
      "step": 630
    },
    {
      "epoch": 0.426495437647854,
      "grad_norm": 29.141868591308594,
      "learning_rate": 2e-05,
      "loss": 0.3074,
      "step": 631
    },
    {
      "epoch": 0.42717134166948295,
      "grad_norm": 9.974465370178223,
      "learning_rate": 2e-05,
      "loss": 0.2251,
      "step": 632
    },
    {
      "epoch": 0.4278472456911119,
      "grad_norm": 21.231918334960938,
      "learning_rate": 2e-05,
      "loss": 0.245,
      "step": 633
    },
    {
      "epoch": 0.4285231497127408,
      "grad_norm": 11.697405815124512,
      "learning_rate": 2e-05,
      "loss": 0.2609,
      "step": 634
    },
    {
      "epoch": 0.42919905373436973,
      "grad_norm": 32.082801818847656,
      "learning_rate": 2e-05,
      "loss": 0.2708,
      "step": 635
    },
    {
      "epoch": 0.42987495775599865,
      "grad_norm": 5.990884304046631,
      "learning_rate": 2e-05,
      "loss": 0.2166,
      "step": 636
    },
    {
      "epoch": 0.4305508617776276,
      "grad_norm": 10.21455192565918,
      "learning_rate": 2e-05,
      "loss": 0.2792,
      "step": 637
    },
    {
      "epoch": 0.4312267657992565,
      "grad_norm": 5.398487091064453,
      "learning_rate": 2e-05,
      "loss": 0.241,
      "step": 638
    },
    {
      "epoch": 0.43190266982088543,
      "grad_norm": 20.407052993774414,
      "learning_rate": 2e-05,
      "loss": 0.3049,
      "step": 639
    },
    {
      "epoch": 0.43257857384251436,
      "grad_norm": 33.5536003112793,
      "learning_rate": 2e-05,
      "loss": 0.2766,
      "step": 640
    },
    {
      "epoch": 0.4332544778641433,
      "grad_norm": 8.046171188354492,
      "learning_rate": 2e-05,
      "loss": 0.2454,
      "step": 641
    },
    {
      "epoch": 0.4339303818857722,
      "grad_norm": 6.703970432281494,
      "learning_rate": 2e-05,
      "loss": 0.2465,
      "step": 642
    },
    {
      "epoch": 0.43460628590740114,
      "grad_norm": 8.684975624084473,
      "learning_rate": 2e-05,
      "loss": 0.2406,
      "step": 643
    },
    {
      "epoch": 0.43528218992903006,
      "grad_norm": 29.707195281982422,
      "learning_rate": 2e-05,
      "loss": 0.3027,
      "step": 644
    },
    {
      "epoch": 0.435958093950659,
      "grad_norm": 27.96883201599121,
      "learning_rate": 2e-05,
      "loss": 0.2849,
      "step": 645
    },
    {
      "epoch": 0.4366339979722879,
      "grad_norm": 13.899518013000488,
      "learning_rate": 2e-05,
      "loss": 0.2583,
      "step": 646
    },
    {
      "epoch": 0.43730990199391684,
      "grad_norm": 4.008049964904785,
      "learning_rate": 2e-05,
      "loss": 0.2603,
      "step": 647
    },
    {
      "epoch": 0.4379858060155458,
      "grad_norm": 17.310728073120117,
      "learning_rate": 2e-05,
      "loss": 0.2267,
      "step": 648
    },
    {
      "epoch": 0.43866171003717475,
      "grad_norm": 24.714000701904297,
      "learning_rate": 2e-05,
      "loss": 0.2558,
      "step": 649
    },
    {
      "epoch": 0.43933761405880367,
      "grad_norm": 21.152280807495117,
      "learning_rate": 2e-05,
      "loss": 0.2463,
      "step": 650
    },
    {
      "epoch": 0.43933761405880367,
      "eval_accuracy": 0.48872466216216215,
      "eval_loss": 0.25620636343955994,
      "eval_runtime": 54.4323,
      "eval_samples_per_second": 435.036,
      "eval_steps_per_second": 13.595,
      "step": 650
    },
    {
      "epoch": 0.4400135180804326,
      "grad_norm": 13.977872848510742,
      "learning_rate": 2e-05,
      "loss": 0.222,
      "step": 651
    },
    {
      "epoch": 0.4406894221020615,
      "grad_norm": 4.368704319000244,
      "learning_rate": 2e-05,
      "loss": 0.196,
      "step": 652
    },
    {
      "epoch": 0.44136532612369045,
      "grad_norm": 13.556212425231934,
      "learning_rate": 2e-05,
      "loss": 0.2452,
      "step": 653
    },
    {
      "epoch": 0.4420412301453194,
      "grad_norm": 12.229365348815918,
      "learning_rate": 2e-05,
      "loss": 0.2967,
      "step": 654
    },
    {
      "epoch": 0.4427171341669483,
      "grad_norm": 13.790763854980469,
      "learning_rate": 2e-05,
      "loss": 0.256,
      "step": 655
    },
    {
      "epoch": 0.4433930381885772,
      "grad_norm": 32.792354583740234,
      "learning_rate": 2e-05,
      "loss": 0.2796,
      "step": 656
    },
    {
      "epoch": 0.44406894221020615,
      "grad_norm": 13.69776725769043,
      "learning_rate": 2e-05,
      "loss": 0.2735,
      "step": 657
    },
    {
      "epoch": 0.4447448462318351,
      "grad_norm": 22.52513885498047,
      "learning_rate": 2e-05,
      "loss": 0.2618,
      "step": 658
    },
    {
      "epoch": 0.445420750253464,
      "grad_norm": 16.65192222595215,
      "learning_rate": 2e-05,
      "loss": 0.2329,
      "step": 659
    },
    {
      "epoch": 0.44609665427509293,
      "grad_norm": 47.09639358520508,
      "learning_rate": 2e-05,
      "loss": 0.2942,
      "step": 660
    },
    {
      "epoch": 0.44677255829672186,
      "grad_norm": 5.504394054412842,
      "learning_rate": 2e-05,
      "loss": 0.2311,
      "step": 661
    },
    {
      "epoch": 0.4474484623183508,
      "grad_norm": 5.224880218505859,
      "learning_rate": 2e-05,
      "loss": 0.21,
      "step": 662
    },
    {
      "epoch": 0.4481243663399797,
      "grad_norm": 9.002572059631348,
      "learning_rate": 2e-05,
      "loss": 0.2463,
      "step": 663
    },
    {
      "epoch": 0.44880027036160863,
      "grad_norm": 14.988982200622559,
      "learning_rate": 2e-05,
      "loss": 0.2196,
      "step": 664
    },
    {
      "epoch": 0.44947617438323756,
      "grad_norm": 27.324880599975586,
      "learning_rate": 2e-05,
      "loss": 0.2742,
      "step": 665
    },
    {
      "epoch": 0.4501520784048665,
      "grad_norm": 28.757923126220703,
      "learning_rate": 2e-05,
      "loss": 0.2608,
      "step": 666
    },
    {
      "epoch": 0.4508279824264954,
      "grad_norm": 12.649532318115234,
      "learning_rate": 2e-05,
      "loss": 0.2507,
      "step": 667
    },
    {
      "epoch": 0.4515038864481244,
      "grad_norm": 14.712961196899414,
      "learning_rate": 2e-05,
      "loss": 0.2354,
      "step": 668
    },
    {
      "epoch": 0.4521797904697533,
      "grad_norm": 15.729104042053223,
      "learning_rate": 2e-05,
      "loss": 0.2355,
      "step": 669
    },
    {
      "epoch": 0.45285569449138224,
      "grad_norm": 28.63264274597168,
      "learning_rate": 2e-05,
      "loss": 0.2522,
      "step": 670
    },
    {
      "epoch": 0.45353159851301117,
      "grad_norm": 5.661179542541504,
      "learning_rate": 2e-05,
      "loss": 0.2101,
      "step": 671
    },
    {
      "epoch": 0.4542075025346401,
      "grad_norm": 5.4098992347717285,
      "learning_rate": 2e-05,
      "loss": 0.2486,
      "step": 672
    },
    {
      "epoch": 0.454883406556269,
      "grad_norm": 20.74265480041504,
      "learning_rate": 2e-05,
      "loss": 0.2453,
      "step": 673
    },
    {
      "epoch": 0.45555931057789795,
      "grad_norm": 11.112505912780762,
      "learning_rate": 2e-05,
      "loss": 0.2201,
      "step": 674
    },
    {
      "epoch": 0.4562352145995269,
      "grad_norm": 10.078001976013184,
      "learning_rate": 2e-05,
      "loss": 0.2188,
      "step": 675
    },
    {
      "epoch": 0.4569111186211558,
      "grad_norm": 10.964561462402344,
      "learning_rate": 2e-05,
      "loss": 0.2789,
      "step": 676
    },
    {
      "epoch": 0.4575870226427847,
      "grad_norm": 8.622685432434082,
      "learning_rate": 2e-05,
      "loss": 0.2013,
      "step": 677
    },
    {
      "epoch": 0.45826292666441365,
      "grad_norm": 4.949863910675049,
      "learning_rate": 2e-05,
      "loss": 0.186,
      "step": 678
    },
    {
      "epoch": 0.4589388306860426,
      "grad_norm": 7.447727680206299,
      "learning_rate": 2e-05,
      "loss": 0.2341,
      "step": 679
    },
    {
      "epoch": 0.4596147347076715,
      "grad_norm": 10.697237014770508,
      "learning_rate": 2e-05,
      "loss": 0.2345,
      "step": 680
    },
    {
      "epoch": 0.4602906387293004,
      "grad_norm": 14.21101188659668,
      "learning_rate": 2e-05,
      "loss": 0.259,
      "step": 681
    },
    {
      "epoch": 0.46096654275092935,
      "grad_norm": 26.095861434936523,
      "learning_rate": 2e-05,
      "loss": 0.2237,
      "step": 682
    },
    {
      "epoch": 0.4616424467725583,
      "grad_norm": 8.300591468811035,
      "learning_rate": 2e-05,
      "loss": 0.2365,
      "step": 683
    },
    {
      "epoch": 0.4623183507941872,
      "grad_norm": 27.740970611572266,
      "learning_rate": 2e-05,
      "loss": 0.24,
      "step": 684
    },
    {
      "epoch": 0.46299425481581613,
      "grad_norm": 26.450828552246094,
      "learning_rate": 2e-05,
      "loss": 0.2983,
      "step": 685
    },
    {
      "epoch": 0.46367015883744506,
      "grad_norm": 9.815147399902344,
      "learning_rate": 2e-05,
      "loss": 0.2708,
      "step": 686
    },
    {
      "epoch": 0.46434606285907404,
      "grad_norm": 7.831709861755371,
      "learning_rate": 2e-05,
      "loss": 0.2152,
      "step": 687
    },
    {
      "epoch": 0.46502196688070296,
      "grad_norm": 5.106151103973389,
      "learning_rate": 2e-05,
      "loss": 0.1677,
      "step": 688
    },
    {
      "epoch": 0.4656978709023319,
      "grad_norm": 33.37860870361328,
      "learning_rate": 2e-05,
      "loss": 0.2791,
      "step": 689
    },
    {
      "epoch": 0.4663737749239608,
      "grad_norm": 50.134952545166016,
      "learning_rate": 2e-05,
      "loss": 0.3468,
      "step": 690
    },
    {
      "epoch": 0.46704967894558974,
      "grad_norm": 9.629013061523438,
      "learning_rate": 2e-05,
      "loss": 0.2072,
      "step": 691
    },
    {
      "epoch": 0.46772558296721867,
      "grad_norm": 24.985658645629883,
      "learning_rate": 2e-05,
      "loss": 0.3175,
      "step": 692
    },
    {
      "epoch": 0.4684014869888476,
      "grad_norm": 22.907365798950195,
      "learning_rate": 2e-05,
      "loss": 0.2482,
      "step": 693
    },
    {
      "epoch": 0.4690773910104765,
      "grad_norm": 21.327999114990234,
      "learning_rate": 2e-05,
      "loss": 0.3142,
      "step": 694
    },
    {
      "epoch": 0.46975329503210544,
      "grad_norm": 5.798184871673584,
      "learning_rate": 2e-05,
      "loss": 0.2705,
      "step": 695
    },
    {
      "epoch": 0.47042919905373437,
      "grad_norm": 13.178022384643555,
      "learning_rate": 2e-05,
      "loss": 0.1967,
      "step": 696
    },
    {
      "epoch": 0.4711051030753633,
      "grad_norm": 5.837411403656006,
      "learning_rate": 2e-05,
      "loss": 0.213,
      "step": 697
    },
    {
      "epoch": 0.4717810070969922,
      "grad_norm": 10.437995910644531,
      "learning_rate": 2e-05,
      "loss": 0.2434,
      "step": 698
    },
    {
      "epoch": 0.47245691111862115,
      "grad_norm": 6.975067615509033,
      "learning_rate": 2e-05,
      "loss": 0.2368,
      "step": 699
    },
    {
      "epoch": 0.4731328151402501,
      "grad_norm": 4.262048244476318,
      "learning_rate": 2e-05,
      "loss": 0.2137,
      "step": 700
    },
    {
      "epoch": 0.4731328151402501,
      "eval_accuracy": 0.4838260135135135,
      "eval_loss": 0.2454683482646942,
      "eval_runtime": 43.542,
      "eval_samples_per_second": 543.843,
      "eval_steps_per_second": 16.995,
      "step": 700
    },
    {
      "epoch": 0.473808719161879,
      "grad_norm": 5.713321685791016,
      "learning_rate": 2e-05,
      "loss": 0.2182,
      "step": 701
    },
    {
      "epoch": 0.4744846231835079,
      "grad_norm": 5.373685836791992,
      "learning_rate": 2e-05,
      "loss": 0.207,
      "step": 702
    },
    {
      "epoch": 0.47516052720513685,
      "grad_norm": 5.068801403045654,
      "learning_rate": 2e-05,
      "loss": 0.22,
      "step": 703
    },
    {
      "epoch": 0.4758364312267658,
      "grad_norm": 6.384458065032959,
      "learning_rate": 2e-05,
      "loss": 0.2277,
      "step": 704
    },
    {
      "epoch": 0.4765123352483947,
      "grad_norm": 6.275250434875488,
      "learning_rate": 2e-05,
      "loss": 0.2237,
      "step": 705
    },
    {
      "epoch": 0.4771882392700237,
      "grad_norm": 7.105129718780518,
      "learning_rate": 2e-05,
      "loss": 0.2455,
      "step": 706
    },
    {
      "epoch": 0.4778641432916526,
      "grad_norm": 9.927888870239258,
      "learning_rate": 2e-05,
      "loss": 0.1986,
      "step": 707
    },
    {
      "epoch": 0.47854004731328154,
      "grad_norm": 6.860706329345703,
      "learning_rate": 2e-05,
      "loss": 0.2087,
      "step": 708
    },
    {
      "epoch": 0.47921595133491046,
      "grad_norm": 6.501542568206787,
      "learning_rate": 2e-05,
      "loss": 0.2377,
      "step": 709
    },
    {
      "epoch": 0.4798918553565394,
      "grad_norm": 13.474136352539062,
      "learning_rate": 2e-05,
      "loss": 0.2641,
      "step": 710
    },
    {
      "epoch": 0.4805677593781683,
      "grad_norm": 9.891058921813965,
      "learning_rate": 2e-05,
      "loss": 0.2051,
      "step": 711
    },
    {
      "epoch": 0.48124366339979724,
      "grad_norm": 8.466797828674316,
      "learning_rate": 2e-05,
      "loss": 0.1841,
      "step": 712
    },
    {
      "epoch": 0.48191956742142616,
      "grad_norm": 6.309836387634277,
      "learning_rate": 2e-05,
      "loss": 0.2024,
      "step": 713
    },
    {
      "epoch": 0.4825954714430551,
      "grad_norm": 15.04800796508789,
      "learning_rate": 2e-05,
      "loss": 0.2944,
      "step": 714
    },
    {
      "epoch": 0.483271375464684,
      "grad_norm": 16.803714752197266,
      "learning_rate": 2e-05,
      "loss": 0.2611,
      "step": 715
    },
    {
      "epoch": 0.48394727948631294,
      "grad_norm": 8.602987289428711,
      "learning_rate": 2e-05,
      "loss": 0.2577,
      "step": 716
    },
    {
      "epoch": 0.48462318350794187,
      "grad_norm": 18.947805404663086,
      "learning_rate": 2e-05,
      "loss": 0.2406,
      "step": 717
    },
    {
      "epoch": 0.4852990875295708,
      "grad_norm": 14.394124984741211,
      "learning_rate": 2e-05,
      "loss": 0.2616,
      "step": 718
    },
    {
      "epoch": 0.4859749915511997,
      "grad_norm": 24.078615188598633,
      "learning_rate": 2e-05,
      "loss": 0.2894,
      "step": 719
    },
    {
      "epoch": 0.48665089557282865,
      "grad_norm": 10.197669982910156,
      "learning_rate": 2e-05,
      "loss": 0.2753,
      "step": 720
    },
    {
      "epoch": 0.48732679959445757,
      "grad_norm": 12.005406379699707,
      "learning_rate": 2e-05,
      "loss": 0.2534,
      "step": 721
    },
    {
      "epoch": 0.4880027036160865,
      "grad_norm": 5.215416431427002,
      "learning_rate": 2e-05,
      "loss": 0.2396,
      "step": 722
    },
    {
      "epoch": 0.4886786076377154,
      "grad_norm": 20.210205078125,
      "learning_rate": 2e-05,
      "loss": 0.2932,
      "step": 723
    },
    {
      "epoch": 0.48935451165934435,
      "grad_norm": 5.5913004875183105,
      "learning_rate": 2e-05,
      "loss": 0.2403,
      "step": 724
    },
    {
      "epoch": 0.4900304156809733,
      "grad_norm": 7.585319995880127,
      "learning_rate": 2e-05,
      "loss": 0.2292,
      "step": 725
    },
    {
      "epoch": 0.49070631970260226,
      "grad_norm": 7.252737998962402,
      "learning_rate": 2e-05,
      "loss": 0.2031,
      "step": 726
    },
    {
      "epoch": 0.4913822237242312,
      "grad_norm": 5.767192840576172,
      "learning_rate": 2e-05,
      "loss": 0.2509,
      "step": 727
    },
    {
      "epoch": 0.4920581277458601,
      "grad_norm": 23.699129104614258,
      "learning_rate": 2e-05,
      "loss": 0.3178,
      "step": 728
    },
    {
      "epoch": 0.49273403176748903,
      "grad_norm": 25.347023010253906,
      "learning_rate": 2e-05,
      "loss": 0.2427,
      "step": 729
    },
    {
      "epoch": 0.49340993578911796,
      "grad_norm": 12.767412185668945,
      "learning_rate": 2e-05,
      "loss": 0.2637,
      "step": 730
    },
    {
      "epoch": 0.4940858398107469,
      "grad_norm": 12.110295295715332,
      "learning_rate": 2e-05,
      "loss": 0.2281,
      "step": 731
    },
    {
      "epoch": 0.4947617438323758,
      "grad_norm": 8.233778953552246,
      "learning_rate": 2e-05,
      "loss": 0.2684,
      "step": 732
    },
    {
      "epoch": 0.49543764785400474,
      "grad_norm": 9.911086082458496,
      "learning_rate": 2e-05,
      "loss": 0.257,
      "step": 733
    },
    {
      "epoch": 0.49611355187563366,
      "grad_norm": 15.240533828735352,
      "learning_rate": 2e-05,
      "loss": 0.2014,
      "step": 734
    },
    {
      "epoch": 0.4967894558972626,
      "grad_norm": 4.944741249084473,
      "learning_rate": 2e-05,
      "loss": 0.2356,
      "step": 735
    },
    {
      "epoch": 0.4974653599188915,
      "grad_norm": 6.706338882446289,
      "learning_rate": 2e-05,
      "loss": 0.2089,
      "step": 736
    },
    {
      "epoch": 0.49814126394052044,
      "grad_norm": 16.521263122558594,
      "learning_rate": 2e-05,
      "loss": 0.282,
      "step": 737
    },
    {
      "epoch": 0.49881716796214937,
      "grad_norm": 21.132631301879883,
      "learning_rate": 2e-05,
      "loss": 0.2713,
      "step": 738
    },
    {
      "epoch": 0.4994930719837783,
      "grad_norm": 25.87594985961914,
      "learning_rate": 2e-05,
      "loss": 0.3244,
      "step": 739
    },
    {
      "epoch": 0.5001689760054072,
      "grad_norm": 15.368523597717285,
      "learning_rate": 2e-05,
      "loss": 0.2526,
      "step": 740
    },
    {
      "epoch": 0.5008448800270362,
      "grad_norm": 24.8336238861084,
      "learning_rate": 2e-05,
      "loss": 0.2359,
      "step": 741
    },
    {
      "epoch": 0.5015207840486651,
      "grad_norm": 25.335954666137695,
      "learning_rate": 2e-05,
      "loss": 0.2718,
      "step": 742
    },
    {
      "epoch": 0.502196688070294,
      "grad_norm": 21.13730812072754,
      "learning_rate": 2e-05,
      "loss": 0.2623,
      "step": 743
    },
    {
      "epoch": 0.5028725920919229,
      "grad_norm": 14.057683944702148,
      "learning_rate": 2e-05,
      "loss": 0.293,
      "step": 744
    },
    {
      "epoch": 0.5035484961135519,
      "grad_norm": 6.086090564727783,
      "learning_rate": 2e-05,
      "loss": 0.2133,
      "step": 745
    },
    {
      "epoch": 0.5042244001351808,
      "grad_norm": 9.203995704650879,
      "learning_rate": 2e-05,
      "loss": 0.2475,
      "step": 746
    },
    {
      "epoch": 0.5049003041568098,
      "grad_norm": 22.84217643737793,
      "learning_rate": 2e-05,
      "loss": 0.2566,
      "step": 747
    },
    {
      "epoch": 0.5055762081784386,
      "grad_norm": 22.9800968170166,
      "learning_rate": 2e-05,
      "loss": 0.2511,
      "step": 748
    },
    {
      "epoch": 0.5062521122000676,
      "grad_norm": 33.33415985107422,
      "learning_rate": 2e-05,
      "loss": 0.2635,
      "step": 749
    },
    {
      "epoch": 0.5069280162216965,
      "grad_norm": 17.215524673461914,
      "learning_rate": 2e-05,
      "loss": 0.2031,
      "step": 750
    },
    {
      "epoch": 0.5069280162216965,
      "eval_accuracy": 0.4747466216216216,
      "eval_loss": 0.24398404359817505,
      "eval_runtime": 42.8602,
      "eval_samples_per_second": 552.494,
      "eval_steps_per_second": 17.265,
      "step": 750
    },
    {
      "epoch": 0.5076039202433255,
      "grad_norm": 8.766871452331543,
      "learning_rate": 2e-05,
      "loss": 0.1977,
      "step": 751
    },
    {
      "epoch": 0.5082798242649543,
      "grad_norm": 7.902562618255615,
      "learning_rate": 2e-05,
      "loss": 0.2236,
      "step": 752
    },
    {
      "epoch": 0.5089557282865833,
      "grad_norm": 19.342979431152344,
      "learning_rate": 2e-05,
      "loss": 0.2632,
      "step": 753
    },
    {
      "epoch": 0.5096316323082122,
      "grad_norm": 10.772290229797363,
      "learning_rate": 2e-05,
      "loss": 0.2715,
      "step": 754
    },
    {
      "epoch": 0.5103075363298412,
      "grad_norm": 24.518299102783203,
      "learning_rate": 2e-05,
      "loss": 0.2501,
      "step": 755
    },
    {
      "epoch": 0.5109834403514701,
      "grad_norm": 12.516921997070312,
      "learning_rate": 2e-05,
      "loss": 0.224,
      "step": 756
    },
    {
      "epoch": 0.511659344373099,
      "grad_norm": 32.63533401489258,
      "learning_rate": 2e-05,
      "loss": 0.2938,
      "step": 757
    },
    {
      "epoch": 0.512335248394728,
      "grad_norm": 6.152921676635742,
      "learning_rate": 2e-05,
      "loss": 0.2091,
      "step": 758
    },
    {
      "epoch": 0.5130111524163569,
      "grad_norm": 6.150788307189941,
      "learning_rate": 2e-05,
      "loss": 0.237,
      "step": 759
    },
    {
      "epoch": 0.5136870564379858,
      "grad_norm": 8.665495872497559,
      "learning_rate": 2e-05,
      "loss": 0.2019,
      "step": 760
    },
    {
      "epoch": 0.5143629604596147,
      "grad_norm": 12.776865005493164,
      "learning_rate": 2e-05,
      "loss": 0.2038,
      "step": 761
    },
    {
      "epoch": 0.5150388644812437,
      "grad_norm": 6.7819414138793945,
      "learning_rate": 2e-05,
      "loss": 0.2553,
      "step": 762
    },
    {
      "epoch": 0.5157147685028726,
      "grad_norm": 11.98867130279541,
      "learning_rate": 2e-05,
      "loss": 0.2428,
      "step": 763
    },
    {
      "epoch": 0.5163906725245015,
      "grad_norm": 17.58635711669922,
      "learning_rate": 2e-05,
      "loss": 0.2756,
      "step": 764
    },
    {
      "epoch": 0.5170665765461304,
      "grad_norm": 7.9190568923950195,
      "learning_rate": 2e-05,
      "loss": 0.2002,
      "step": 765
    },
    {
      "epoch": 0.5177424805677594,
      "grad_norm": 9.14595890045166,
      "learning_rate": 2e-05,
      "loss": 0.2431,
      "step": 766
    },
    {
      "epoch": 0.5184183845893883,
      "grad_norm": 14.508289337158203,
      "learning_rate": 2e-05,
      "loss": 0.293,
      "step": 767
    },
    {
      "epoch": 0.5190942886110173,
      "grad_norm": 5.583828926086426,
      "learning_rate": 2e-05,
      "loss": 0.2369,
      "step": 768
    },
    {
      "epoch": 0.5197701926326461,
      "grad_norm": 6.178142070770264,
      "learning_rate": 2e-05,
      "loss": 0.1963,
      "step": 769
    },
    {
      "epoch": 0.5204460966542751,
      "grad_norm": 7.729180335998535,
      "learning_rate": 2e-05,
      "loss": 0.2567,
      "step": 770
    },
    {
      "epoch": 0.521122000675904,
      "grad_norm": 7.023488998413086,
      "learning_rate": 2e-05,
      "loss": 0.258,
      "step": 771
    },
    {
      "epoch": 0.521797904697533,
      "grad_norm": 7.782037258148193,
      "learning_rate": 2e-05,
      "loss": 0.2904,
      "step": 772
    },
    {
      "epoch": 0.5224738087191618,
      "grad_norm": 16.021780014038086,
      "learning_rate": 2e-05,
      "loss": 0.2252,
      "step": 773
    },
    {
      "epoch": 0.5231497127407908,
      "grad_norm": 10.733901023864746,
      "learning_rate": 2e-05,
      "loss": 0.1876,
      "step": 774
    },
    {
      "epoch": 0.5238256167624198,
      "grad_norm": 4.662460803985596,
      "learning_rate": 2e-05,
      "loss": 0.188,
      "step": 775
    },
    {
      "epoch": 0.5245015207840487,
      "grad_norm": 14.435529708862305,
      "learning_rate": 2e-05,
      "loss": 0.2245,
      "step": 776
    },
    {
      "epoch": 0.5251774248056776,
      "grad_norm": 19.989212036132812,
      "learning_rate": 2e-05,
      "loss": 0.2335,
      "step": 777
    },
    {
      "epoch": 0.5258533288273065,
      "grad_norm": 17.345104217529297,
      "learning_rate": 2e-05,
      "loss": 0.2649,
      "step": 778
    },
    {
      "epoch": 0.5265292328489355,
      "grad_norm": 11.777247428894043,
      "learning_rate": 2e-05,
      "loss": 0.2452,
      "step": 779
    },
    {
      "epoch": 0.5272051368705644,
      "grad_norm": 13.34235668182373,
      "learning_rate": 2e-05,
      "loss": 0.2283,
      "step": 780
    },
    {
      "epoch": 0.5278810408921933,
      "grad_norm": 19.73564910888672,
      "learning_rate": 2e-05,
      "loss": 0.2056,
      "step": 781
    },
    {
      "epoch": 0.5285569449138222,
      "grad_norm": 3.989328622817993,
      "learning_rate": 2e-05,
      "loss": 0.2146,
      "step": 782
    },
    {
      "epoch": 0.5292328489354512,
      "grad_norm": 9.697778701782227,
      "learning_rate": 2e-05,
      "loss": 0.2264,
      "step": 783
    },
    {
      "epoch": 0.5299087529570801,
      "grad_norm": 8.201826095581055,
      "learning_rate": 2e-05,
      "loss": 0.2101,
      "step": 784
    },
    {
      "epoch": 0.530584656978709,
      "grad_norm": 12.118664741516113,
      "learning_rate": 2e-05,
      "loss": 0.2232,
      "step": 785
    },
    {
      "epoch": 0.5312605610003379,
      "grad_norm": 8.31401252746582,
      "learning_rate": 2e-05,
      "loss": 0.2275,
      "step": 786
    },
    {
      "epoch": 0.5319364650219669,
      "grad_norm": 5.233949184417725,
      "learning_rate": 2e-05,
      "loss": 0.1861,
      "step": 787
    },
    {
      "epoch": 0.5326123690435958,
      "grad_norm": 9.61313247680664,
      "learning_rate": 2e-05,
      "loss": 0.1906,
      "step": 788
    },
    {
      "epoch": 0.5332882730652247,
      "grad_norm": 9.397748947143555,
      "learning_rate": 2e-05,
      "loss": 0.1723,
      "step": 789
    },
    {
      "epoch": 0.5339641770868536,
      "grad_norm": 12.695356369018555,
      "learning_rate": 2e-05,
      "loss": 0.2714,
      "step": 790
    },
    {
      "epoch": 0.5346400811084826,
      "grad_norm": 15.562970161437988,
      "learning_rate": 2e-05,
      "loss": 0.2396,
      "step": 791
    },
    {
      "epoch": 0.5353159851301115,
      "grad_norm": 17.84036636352539,
      "learning_rate": 2e-05,
      "loss": 0.2519,
      "step": 792
    },
    {
      "epoch": 0.5359918891517405,
      "grad_norm": 12.072160720825195,
      "learning_rate": 2e-05,
      "loss": 0.1866,
      "step": 793
    },
    {
      "epoch": 0.5366677931733694,
      "grad_norm": 16.592008590698242,
      "learning_rate": 2e-05,
      "loss": 0.1867,
      "step": 794
    },
    {
      "epoch": 0.5373436971949983,
      "grad_norm": 15.631988525390625,
      "learning_rate": 2e-05,
      "loss": 0.2338,
      "step": 795
    },
    {
      "epoch": 0.5380196012166273,
      "grad_norm": 18.908798217773438,
      "learning_rate": 2e-05,
      "loss": 0.2454,
      "step": 796
    },
    {
      "epoch": 0.5386955052382562,
      "grad_norm": 10.594216346740723,
      "learning_rate": 2e-05,
      "loss": 0.2506,
      "step": 797
    },
    {
      "epoch": 0.5393714092598851,
      "grad_norm": 14.339956283569336,
      "learning_rate": 2e-05,
      "loss": 0.2207,
      "step": 798
    },
    {
      "epoch": 0.540047313281514,
      "grad_norm": 7.439448833465576,
      "learning_rate": 2e-05,
      "loss": 0.2415,
      "step": 799
    },
    {
      "epoch": 0.540723217303143,
      "grad_norm": 6.967495441436768,
      "learning_rate": 2e-05,
      "loss": 0.2206,
      "step": 800
    },
    {
      "epoch": 0.540723217303143,
      "eval_accuracy": 0.501097972972973,
      "eval_loss": 0.24171486496925354,
      "eval_runtime": 50.1506,
      "eval_samples_per_second": 472.177,
      "eval_steps_per_second": 14.756,
      "step": 800
    },
    {
      "epoch": 0.5413991213247719,
      "grad_norm": 5.49332332611084,
      "learning_rate": 2e-05,
      "loss": 0.2523,
      "step": 801
    },
    {
      "epoch": 0.5420750253464008,
      "grad_norm": 12.693426132202148,
      "learning_rate": 2e-05,
      "loss": 0.1893,
      "step": 802
    },
    {
      "epoch": 0.5427509293680297,
      "grad_norm": 8.018287658691406,
      "learning_rate": 2e-05,
      "loss": 0.2117,
      "step": 803
    },
    {
      "epoch": 0.5434268333896587,
      "grad_norm": 10.421152114868164,
      "learning_rate": 2e-05,
      "loss": 0.2192,
      "step": 804
    },
    {
      "epoch": 0.5441027374112876,
      "grad_norm": 6.275336742401123,
      "learning_rate": 2e-05,
      "loss": 0.2354,
      "step": 805
    },
    {
      "epoch": 0.5447786414329165,
      "grad_norm": 9.331214904785156,
      "learning_rate": 2e-05,
      "loss": 0.2104,
      "step": 806
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 25.183286666870117,
      "learning_rate": 2e-05,
      "loss": 0.2124,
      "step": 807
    },
    {
      "epoch": 0.5461304494761744,
      "grad_norm": 13.36805248260498,
      "learning_rate": 2e-05,
      "loss": 0.2167,
      "step": 808
    },
    {
      "epoch": 0.5468063534978033,
      "grad_norm": 18.761341094970703,
      "learning_rate": 2e-05,
      "loss": 0.2423,
      "step": 809
    },
    {
      "epoch": 0.5474822575194322,
      "grad_norm": 19.603567123413086,
      "learning_rate": 2e-05,
      "loss": 0.3018,
      "step": 810
    },
    {
      "epoch": 0.5481581615410611,
      "grad_norm": 10.518606185913086,
      "learning_rate": 2e-05,
      "loss": 0.2082,
      "step": 811
    },
    {
      "epoch": 0.5488340655626901,
      "grad_norm": 4.735916614532471,
      "learning_rate": 2e-05,
      "loss": 0.2277,
      "step": 812
    },
    {
      "epoch": 0.5495099695843191,
      "grad_norm": 7.163942813873291,
      "learning_rate": 2e-05,
      "loss": 0.2342,
      "step": 813
    },
    {
      "epoch": 0.550185873605948,
      "grad_norm": 11.552167892456055,
      "learning_rate": 2e-05,
      "loss": 0.182,
      "step": 814
    },
    {
      "epoch": 0.5508617776275769,
      "grad_norm": 5.470919132232666,
      "learning_rate": 2e-05,
      "loss": 0.1633,
      "step": 815
    },
    {
      "epoch": 0.5515376816492058,
      "grad_norm": 19.9710636138916,
      "learning_rate": 2e-05,
      "loss": 0.2379,
      "step": 816
    },
    {
      "epoch": 0.5522135856708348,
      "grad_norm": 16.759910583496094,
      "learning_rate": 2e-05,
      "loss": 0.2131,
      "step": 817
    },
    {
      "epoch": 0.5528894896924637,
      "grad_norm": 28.654191970825195,
      "learning_rate": 2e-05,
      "loss": 0.243,
      "step": 818
    },
    {
      "epoch": 0.5535653937140926,
      "grad_norm": 15.915573120117188,
      "learning_rate": 2e-05,
      "loss": 0.1759,
      "step": 819
    },
    {
      "epoch": 0.5542412977357215,
      "grad_norm": 20.26451301574707,
      "learning_rate": 2e-05,
      "loss": 0.25,
      "step": 820
    },
    {
      "epoch": 0.5549172017573505,
      "grad_norm": 6.014492988586426,
      "learning_rate": 2e-05,
      "loss": 0.1759,
      "step": 821
    },
    {
      "epoch": 0.5555931057789794,
      "grad_norm": 21.39944076538086,
      "learning_rate": 2e-05,
      "loss": 0.1907,
      "step": 822
    },
    {
      "epoch": 0.5562690098006083,
      "grad_norm": 25.57290267944336,
      "learning_rate": 2e-05,
      "loss": 0.2586,
      "step": 823
    },
    {
      "epoch": 0.5569449138222372,
      "grad_norm": 13.717406272888184,
      "learning_rate": 2e-05,
      "loss": 0.297,
      "step": 824
    },
    {
      "epoch": 0.5576208178438662,
      "grad_norm": 12.847921371459961,
      "learning_rate": 2e-05,
      "loss": 0.2157,
      "step": 825
    },
    {
      "epoch": 0.5582967218654951,
      "grad_norm": 41.37167739868164,
      "learning_rate": 2e-05,
      "loss": 0.3664,
      "step": 826
    },
    {
      "epoch": 0.558972625887124,
      "grad_norm": 25.807235717773438,
      "learning_rate": 2e-05,
      "loss": 0.2432,
      "step": 827
    },
    {
      "epoch": 0.5596485299087529,
      "grad_norm": 7.43070650100708,
      "learning_rate": 2e-05,
      "loss": 0.2589,
      "step": 828
    },
    {
      "epoch": 0.5603244339303819,
      "grad_norm": 30.882186889648438,
      "learning_rate": 2e-05,
      "loss": 0.2895,
      "step": 829
    },
    {
      "epoch": 0.5610003379520108,
      "grad_norm": 23.006864547729492,
      "learning_rate": 2e-05,
      "loss": 0.2891,
      "step": 830
    },
    {
      "epoch": 0.5616762419736397,
      "grad_norm": 7.514183521270752,
      "learning_rate": 2e-05,
      "loss": 0.2271,
      "step": 831
    },
    {
      "epoch": 0.5623521459952687,
      "grad_norm": 6.559591770172119,
      "learning_rate": 2e-05,
      "loss": 0.2581,
      "step": 832
    },
    {
      "epoch": 0.5630280500168976,
      "grad_norm": 18.156450271606445,
      "learning_rate": 2e-05,
      "loss": 0.2272,
      "step": 833
    },
    {
      "epoch": 0.5637039540385266,
      "grad_norm": 7.304296970367432,
      "learning_rate": 2e-05,
      "loss": 0.2164,
      "step": 834
    },
    {
      "epoch": 0.5643798580601554,
      "grad_norm": 19.68966293334961,
      "learning_rate": 2e-05,
      "loss": 0.2558,
      "step": 835
    },
    {
      "epoch": 0.5650557620817844,
      "grad_norm": 21.8721923828125,
      "learning_rate": 2e-05,
      "loss": 0.2412,
      "step": 836
    },
    {
      "epoch": 0.5657316661034133,
      "grad_norm": 4.59301233291626,
      "learning_rate": 2e-05,
      "loss": 0.2374,
      "step": 837
    },
    {
      "epoch": 0.5664075701250423,
      "grad_norm": 8.521605491638184,
      "learning_rate": 2e-05,
      "loss": 0.2307,
      "step": 838
    },
    {
      "epoch": 0.5670834741466712,
      "grad_norm": 6.292214393615723,
      "learning_rate": 2e-05,
      "loss": 0.2265,
      "step": 839
    },
    {
      "epoch": 0.5677593781683001,
      "grad_norm": 4.929276943206787,
      "learning_rate": 2e-05,
      "loss": 0.2193,
      "step": 840
    },
    {
      "epoch": 0.568435282189929,
      "grad_norm": 5.209563255310059,
      "learning_rate": 2e-05,
      "loss": 0.244,
      "step": 841
    },
    {
      "epoch": 0.569111186211558,
      "grad_norm": 8.945363998413086,
      "learning_rate": 2e-05,
      "loss": 0.2273,
      "step": 842
    },
    {
      "epoch": 0.5697870902331869,
      "grad_norm": 5.43878698348999,
      "learning_rate": 2e-05,
      "loss": 0.26,
      "step": 843
    },
    {
      "epoch": 0.5704629942548158,
      "grad_norm": 25.81899642944336,
      "learning_rate": 2e-05,
      "loss": 0.2928,
      "step": 844
    },
    {
      "epoch": 0.5711388982764447,
      "grad_norm": 27.98745346069336,
      "learning_rate": 2e-05,
      "loss": 0.2438,
      "step": 845
    },
    {
      "epoch": 0.5718148022980737,
      "grad_norm": 13.064859390258789,
      "learning_rate": 2e-05,
      "loss": 0.2429,
      "step": 846
    },
    {
      "epoch": 0.5724907063197026,
      "grad_norm": 19.52817153930664,
      "learning_rate": 2e-05,
      "loss": 0.2444,
      "step": 847
    },
    {
      "epoch": 0.5731666103413315,
      "grad_norm": 10.583756446838379,
      "learning_rate": 2e-05,
      "loss": 0.2279,
      "step": 848
    },
    {
      "epoch": 0.5738425143629604,
      "grad_norm": 31.996356964111328,
      "learning_rate": 2e-05,
      "loss": 0.219,
      "step": 849
    },
    {
      "epoch": 0.5745184183845894,
      "grad_norm": 14.50013542175293,
      "learning_rate": 2e-05,
      "loss": 0.2813,
      "step": 850
    },
    {
      "epoch": 0.5745184183845894,
      "eval_accuracy": 0.46841216216216214,
      "eval_loss": 0.2428000122308731,
      "eval_runtime": 51.0591,
      "eval_samples_per_second": 463.777,
      "eval_steps_per_second": 14.493,
      "step": 850
    },
    {
      "epoch": 0.5751943224062184,
      "grad_norm": 12.393126487731934,
      "learning_rate": 2e-05,
      "loss": 0.2003,
      "step": 851
    },
    {
      "epoch": 0.5758702264278472,
      "grad_norm": 22.28180503845215,
      "learning_rate": 2e-05,
      "loss": 0.249,
      "step": 852
    },
    {
      "epoch": 0.5765461304494762,
      "grad_norm": 18.110198974609375,
      "learning_rate": 2e-05,
      "loss": 0.2287,
      "step": 853
    },
    {
      "epoch": 0.5772220344711051,
      "grad_norm": 18.370519638061523,
      "learning_rate": 2e-05,
      "loss": 0.3031,
      "step": 854
    },
    {
      "epoch": 0.5778979384927341,
      "grad_norm": 6.29045295715332,
      "learning_rate": 2e-05,
      "loss": 0.2157,
      "step": 855
    },
    {
      "epoch": 0.578573842514363,
      "grad_norm": 4.097279071807861,
      "learning_rate": 2e-05,
      "loss": 0.2331,
      "step": 856
    },
    {
      "epoch": 0.5792497465359919,
      "grad_norm": 8.204437255859375,
      "learning_rate": 2e-05,
      "loss": 0.2215,
      "step": 857
    },
    {
      "epoch": 0.5799256505576208,
      "grad_norm": 14.319632530212402,
      "learning_rate": 2e-05,
      "loss": 0.248,
      "step": 858
    },
    {
      "epoch": 0.5806015545792498,
      "grad_norm": 12.233708381652832,
      "learning_rate": 2e-05,
      "loss": 0.229,
      "step": 859
    },
    {
      "epoch": 0.5812774586008786,
      "grad_norm": 9.040895462036133,
      "learning_rate": 2e-05,
      "loss": 0.2409,
      "step": 860
    },
    {
      "epoch": 0.5819533626225076,
      "grad_norm": 5.95002555847168,
      "learning_rate": 2e-05,
      "loss": 0.1733,
      "step": 861
    },
    {
      "epoch": 0.5826292666441365,
      "grad_norm": 4.910984516143799,
      "learning_rate": 2e-05,
      "loss": 0.2266,
      "step": 862
    },
    {
      "epoch": 0.5833051706657655,
      "grad_norm": 7.131229400634766,
      "learning_rate": 2e-05,
      "loss": 0.2111,
      "step": 863
    },
    {
      "epoch": 0.5839810746873944,
      "grad_norm": 5.1656904220581055,
      "learning_rate": 2e-05,
      "loss": 0.161,
      "step": 864
    },
    {
      "epoch": 0.5846569787090233,
      "grad_norm": 14.883503913879395,
      "learning_rate": 2e-05,
      "loss": 0.1825,
      "step": 865
    },
    {
      "epoch": 0.5853328827306522,
      "grad_norm": 6.652871131896973,
      "learning_rate": 2e-05,
      "loss": 0.1869,
      "step": 866
    },
    {
      "epoch": 0.5860087867522812,
      "grad_norm": 7.669157981872559,
      "learning_rate": 2e-05,
      "loss": 0.1587,
      "step": 867
    },
    {
      "epoch": 0.5866846907739101,
      "grad_norm": 25.5183048248291,
      "learning_rate": 2e-05,
      "loss": 0.3247,
      "step": 868
    },
    {
      "epoch": 0.587360594795539,
      "grad_norm": 8.981620788574219,
      "learning_rate": 2e-05,
      "loss": 0.2509,
      "step": 869
    },
    {
      "epoch": 0.5880364988171679,
      "grad_norm": 8.003873825073242,
      "learning_rate": 2e-05,
      "loss": 0.246,
      "step": 870
    },
    {
      "epoch": 0.5887124028387969,
      "grad_norm": 17.560935974121094,
      "learning_rate": 2e-05,
      "loss": 0.25,
      "step": 871
    },
    {
      "epoch": 0.5893883068604259,
      "grad_norm": 18.198057174682617,
      "learning_rate": 2e-05,
      "loss": 0.1985,
      "step": 872
    },
    {
      "epoch": 0.5900642108820547,
      "grad_norm": 14.000364303588867,
      "learning_rate": 2e-05,
      "loss": 0.2375,
      "step": 873
    },
    {
      "epoch": 0.5907401149036837,
      "grad_norm": 7.843807697296143,
      "learning_rate": 2e-05,
      "loss": 0.2458,
      "step": 874
    },
    {
      "epoch": 0.5914160189253126,
      "grad_norm": 12.056757926940918,
      "learning_rate": 2e-05,
      "loss": 0.2464,
      "step": 875
    },
    {
      "epoch": 0.5920919229469416,
      "grad_norm": 18.50706672668457,
      "learning_rate": 2e-05,
      "loss": 0.2122,
      "step": 876
    },
    {
      "epoch": 0.5927678269685704,
      "grad_norm": 20.386232376098633,
      "learning_rate": 2e-05,
      "loss": 0.2695,
      "step": 877
    },
    {
      "epoch": 0.5934437309901994,
      "grad_norm": 8.848958015441895,
      "learning_rate": 2e-05,
      "loss": 0.1784,
      "step": 878
    },
    {
      "epoch": 0.5941196350118283,
      "grad_norm": 7.481949329376221,
      "learning_rate": 2e-05,
      "loss": 0.1948,
      "step": 879
    },
    {
      "epoch": 0.5947955390334573,
      "grad_norm": 24.28002166748047,
      "learning_rate": 2e-05,
      "loss": 0.2597,
      "step": 880
    },
    {
      "epoch": 0.5954714430550861,
      "grad_norm": 7.155148506164551,
      "learning_rate": 2e-05,
      "loss": 0.1557,
      "step": 881
    },
    {
      "epoch": 0.5961473470767151,
      "grad_norm": 23.062162399291992,
      "learning_rate": 2e-05,
      "loss": 0.2688,
      "step": 882
    },
    {
      "epoch": 0.596823251098344,
      "grad_norm": 21.118284225463867,
      "learning_rate": 2e-05,
      "loss": 0.2529,
      "step": 883
    },
    {
      "epoch": 0.597499155119973,
      "grad_norm": 9.437238693237305,
      "learning_rate": 2e-05,
      "loss": 0.1995,
      "step": 884
    },
    {
      "epoch": 0.5981750591416019,
      "grad_norm": 22.119731903076172,
      "learning_rate": 2e-05,
      "loss": 0.333,
      "step": 885
    },
    {
      "epoch": 0.5988509631632308,
      "grad_norm": 14.369667053222656,
      "learning_rate": 2e-05,
      "loss": 0.2303,
      "step": 886
    },
    {
      "epoch": 0.5995268671848597,
      "grad_norm": 14.250020027160645,
      "learning_rate": 2e-05,
      "loss": 0.229,
      "step": 887
    },
    {
      "epoch": 0.6002027712064887,
      "grad_norm": 8.716592788696289,
      "learning_rate": 2e-05,
      "loss": 0.1873,
      "step": 888
    },
    {
      "epoch": 0.6008786752281176,
      "grad_norm": 29.83051300048828,
      "learning_rate": 2e-05,
      "loss": 0.2262,
      "step": 889
    },
    {
      "epoch": 0.6015545792497465,
      "grad_norm": 13.368353843688965,
      "learning_rate": 2e-05,
      "loss": 0.2534,
      "step": 890
    },
    {
      "epoch": 0.6022304832713755,
      "grad_norm": 27.32805061340332,
      "learning_rate": 2e-05,
      "loss": 0.2813,
      "step": 891
    },
    {
      "epoch": 0.6029063872930044,
      "grad_norm": 27.29034423828125,
      "learning_rate": 2e-05,
      "loss": 0.317,
      "step": 892
    },
    {
      "epoch": 0.6035822913146334,
      "grad_norm": 12.490097045898438,
      "learning_rate": 2e-05,
      "loss": 0.251,
      "step": 893
    },
    {
      "epoch": 0.6042581953362622,
      "grad_norm": 5.995603561401367,
      "learning_rate": 2e-05,
      "loss": 0.2615,
      "step": 894
    },
    {
      "epoch": 0.6049340993578912,
      "grad_norm": 5.359321117401123,
      "learning_rate": 2e-05,
      "loss": 0.2488,
      "step": 895
    },
    {
      "epoch": 0.6056100033795201,
      "grad_norm": 4.065511703491211,
      "learning_rate": 2e-05,
      "loss": 0.2529,
      "step": 896
    },
    {
      "epoch": 0.6062859074011491,
      "grad_norm": 4.584462642669678,
      "learning_rate": 2e-05,
      "loss": 0.2157,
      "step": 897
    },
    {
      "epoch": 0.6069618114227779,
      "grad_norm": 18.944622039794922,
      "learning_rate": 2e-05,
      "loss": 0.2506,
      "step": 898
    },
    {
      "epoch": 0.6076377154444069,
      "grad_norm": 12.426178932189941,
      "learning_rate": 2e-05,
      "loss": 0.2332,
      "step": 899
    },
    {
      "epoch": 0.6083136194660358,
      "grad_norm": 4.7093892097473145,
      "learning_rate": 2e-05,
      "loss": 0.1833,
      "step": 900
    },
    {
      "epoch": 0.6083136194660358,
      "eval_accuracy": 0.494214527027027,
      "eval_loss": 0.2376047968864441,
      "eval_runtime": 49.8226,
      "eval_samples_per_second": 475.286,
      "eval_steps_per_second": 14.853,
      "step": 900
    },
    {
      "epoch": 0.6089895234876648,
      "grad_norm": 10.227592468261719,
      "learning_rate": 2e-05,
      "loss": 0.2559,
      "step": 901
    },
    {
      "epoch": 0.6096654275092936,
      "grad_norm": 7.939753532409668,
      "learning_rate": 2e-05,
      "loss": 0.2291,
      "step": 902
    },
    {
      "epoch": 0.6103413315309226,
      "grad_norm": 8.873037338256836,
      "learning_rate": 2e-05,
      "loss": 0.2644,
      "step": 903
    },
    {
      "epoch": 0.6110172355525515,
      "grad_norm": 9.451284408569336,
      "learning_rate": 2e-05,
      "loss": 0.2355,
      "step": 904
    },
    {
      "epoch": 0.6116931395741805,
      "grad_norm": 14.779155731201172,
      "learning_rate": 2e-05,
      "loss": 0.2642,
      "step": 905
    },
    {
      "epoch": 0.6123690435958093,
      "grad_norm": 5.490085124969482,
      "learning_rate": 2e-05,
      "loss": 0.2411,
      "step": 906
    },
    {
      "epoch": 0.6130449476174383,
      "grad_norm": 10.876485824584961,
      "learning_rate": 2e-05,
      "loss": 0.2214,
      "step": 907
    },
    {
      "epoch": 0.6137208516390672,
      "grad_norm": 6.782824516296387,
      "learning_rate": 2e-05,
      "loss": 0.2607,
      "step": 908
    },
    {
      "epoch": 0.6143967556606962,
      "grad_norm": 7.840001106262207,
      "learning_rate": 2e-05,
      "loss": 0.2328,
      "step": 909
    },
    {
      "epoch": 0.6150726596823252,
      "grad_norm": 6.93187141418457,
      "learning_rate": 2e-05,
      "loss": 0.2355,
      "step": 910
    },
    {
      "epoch": 0.615748563703954,
      "grad_norm": 12.17858600616455,
      "learning_rate": 2e-05,
      "loss": 0.2604,
      "step": 911
    },
    {
      "epoch": 0.616424467725583,
      "grad_norm": 17.32192039489746,
      "learning_rate": 2e-05,
      "loss": 0.2367,
      "step": 912
    },
    {
      "epoch": 0.6171003717472119,
      "grad_norm": 5.254330158233643,
      "learning_rate": 2e-05,
      "loss": 0.2308,
      "step": 913
    },
    {
      "epoch": 0.6177762757688409,
      "grad_norm": 5.951586723327637,
      "learning_rate": 2e-05,
      "loss": 0.2216,
      "step": 914
    },
    {
      "epoch": 0.6184521797904697,
      "grad_norm": 5.777591705322266,
      "learning_rate": 2e-05,
      "loss": 0.2896,
      "step": 915
    },
    {
      "epoch": 0.6191280838120987,
      "grad_norm": 5.829374313354492,
      "learning_rate": 2e-05,
      "loss": 0.2194,
      "step": 916
    },
    {
      "epoch": 0.6198039878337276,
      "grad_norm": 13.255396842956543,
      "learning_rate": 2e-05,
      "loss": 0.2236,
      "step": 917
    },
    {
      "epoch": 0.6204798918553566,
      "grad_norm": 6.473336219787598,
      "learning_rate": 2e-05,
      "loss": 0.2362,
      "step": 918
    },
    {
      "epoch": 0.6211557958769854,
      "grad_norm": 13.834738731384277,
      "learning_rate": 2e-05,
      "loss": 0.2108,
      "step": 919
    },
    {
      "epoch": 0.6218316998986144,
      "grad_norm": 5.438971996307373,
      "learning_rate": 2e-05,
      "loss": 0.2253,
      "step": 920
    },
    {
      "epoch": 0.6225076039202433,
      "grad_norm": 8.005040168762207,
      "learning_rate": 2e-05,
      "loss": 0.2259,
      "step": 921
    },
    {
      "epoch": 0.6231835079418723,
      "grad_norm": 4.105720043182373,
      "learning_rate": 2e-05,
      "loss": 0.2262,
      "step": 922
    },
    {
      "epoch": 0.6238594119635011,
      "grad_norm": 6.449704170227051,
      "learning_rate": 2e-05,
      "loss": 0.1723,
      "step": 923
    },
    {
      "epoch": 0.6245353159851301,
      "grad_norm": 5.534755706787109,
      "learning_rate": 2e-05,
      "loss": 0.1789,
      "step": 924
    },
    {
      "epoch": 0.625211220006759,
      "grad_norm": 6.1030097007751465,
      "learning_rate": 2e-05,
      "loss": 0.201,
      "step": 925
    },
    {
      "epoch": 0.625887124028388,
      "grad_norm": 6.290901184082031,
      "learning_rate": 2e-05,
      "loss": 0.2578,
      "step": 926
    },
    {
      "epoch": 0.6265630280500168,
      "grad_norm": 7.832151889801025,
      "learning_rate": 2e-05,
      "loss": 0.2604,
      "step": 927
    },
    {
      "epoch": 0.6272389320716458,
      "grad_norm": 9.229602813720703,
      "learning_rate": 2e-05,
      "loss": 0.2456,
      "step": 928
    },
    {
      "epoch": 0.6279148360932748,
      "grad_norm": 14.984476089477539,
      "learning_rate": 2e-05,
      "loss": 0.2207,
      "step": 929
    },
    {
      "epoch": 0.6285907401149037,
      "grad_norm": 8.645849227905273,
      "learning_rate": 2e-05,
      "loss": 0.1696,
      "step": 930
    },
    {
      "epoch": 0.6292666441365327,
      "grad_norm": 9.641719818115234,
      "learning_rate": 2e-05,
      "loss": 0.2555,
      "step": 931
    },
    {
      "epoch": 0.6299425481581615,
      "grad_norm": 5.814601898193359,
      "learning_rate": 2e-05,
      "loss": 0.1962,
      "step": 932
    },
    {
      "epoch": 0.6306184521797905,
      "grad_norm": 21.282167434692383,
      "learning_rate": 2e-05,
      "loss": 0.2105,
      "step": 933
    },
    {
      "epoch": 0.6312943562014194,
      "grad_norm": 7.012668609619141,
      "learning_rate": 2e-05,
      "loss": 0.2,
      "step": 934
    },
    {
      "epoch": 0.6319702602230484,
      "grad_norm": 17.538867950439453,
      "learning_rate": 2e-05,
      "loss": 0.2361,
      "step": 935
    },
    {
      "epoch": 0.6326461642446772,
      "grad_norm": 8.175737380981445,
      "learning_rate": 2e-05,
      "loss": 0.2557,
      "step": 936
    },
    {
      "epoch": 0.6333220682663062,
      "grad_norm": 8.973394393920898,
      "learning_rate": 2e-05,
      "loss": 0.2215,
      "step": 937
    },
    {
      "epoch": 0.6339979722879351,
      "grad_norm": 7.609282970428467,
      "learning_rate": 2e-05,
      "loss": 0.2385,
      "step": 938
    },
    {
      "epoch": 0.6346738763095641,
      "grad_norm": 10.523405075073242,
      "learning_rate": 2e-05,
      "loss": 0.1831,
      "step": 939
    },
    {
      "epoch": 0.6353497803311929,
      "grad_norm": 8.435493469238281,
      "learning_rate": 2e-05,
      "loss": 0.2237,
      "step": 940
    },
    {
      "epoch": 0.6360256843528219,
      "grad_norm": 6.66005277633667,
      "learning_rate": 2e-05,
      "loss": 0.2159,
      "step": 941
    },
    {
      "epoch": 0.6367015883744508,
      "grad_norm": 8.253142356872559,
      "learning_rate": 2e-05,
      "loss": 0.2041,
      "step": 942
    },
    {
      "epoch": 0.6373774923960798,
      "grad_norm": 6.386720180511475,
      "learning_rate": 2e-05,
      "loss": 0.1979,
      "step": 943
    },
    {
      "epoch": 0.6380533964177086,
      "grad_norm": 14.370277404785156,
      "learning_rate": 2e-05,
      "loss": 0.2767,
      "step": 944
    },
    {
      "epoch": 0.6387293004393376,
      "grad_norm": 10.20983600616455,
      "learning_rate": 2e-05,
      "loss": 0.1982,
      "step": 945
    },
    {
      "epoch": 0.6394052044609665,
      "grad_norm": 6.028168201446533,
      "learning_rate": 2e-05,
      "loss": 0.1868,
      "step": 946
    },
    {
      "epoch": 0.6400811084825955,
      "grad_norm": 12.134034156799316,
      "learning_rate": 2e-05,
      "loss": 0.2321,
      "step": 947
    },
    {
      "epoch": 0.6407570125042245,
      "grad_norm": 10.148812294006348,
      "learning_rate": 2e-05,
      "loss": 0.1896,
      "step": 948
    },
    {
      "epoch": 0.6414329165258533,
      "grad_norm": 5.734457015991211,
      "learning_rate": 2e-05,
      "loss": 0.1837,
      "step": 949
    },
    {
      "epoch": 0.6421088205474823,
      "grad_norm": 6.308267593383789,
      "learning_rate": 2e-05,
      "loss": 0.2197,
      "step": 950
    },
    {
      "epoch": 0.6421088205474823,
      "eval_accuracy": 0.49197635135135137,
      "eval_loss": 0.24354594945907593,
      "eval_runtime": 48.6245,
      "eval_samples_per_second": 486.997,
      "eval_steps_per_second": 15.219,
      "step": 950
    },
    {
      "epoch": 0.6427847245691112,
      "grad_norm": 10.225892066955566,
      "learning_rate": 2e-05,
      "loss": 0.2524,
      "step": 951
    },
    {
      "epoch": 0.6434606285907402,
      "grad_norm": 8.87254810333252,
      "learning_rate": 2e-05,
      "loss": 0.1956,
      "step": 952
    },
    {
      "epoch": 0.644136532612369,
      "grad_norm": 6.702756881713867,
      "learning_rate": 2e-05,
      "loss": 0.2106,
      "step": 953
    },
    {
      "epoch": 0.644812436633998,
      "grad_norm": 22.123424530029297,
      "learning_rate": 2e-05,
      "loss": 0.1748,
      "step": 954
    },
    {
      "epoch": 0.6454883406556269,
      "grad_norm": 12.616996765136719,
      "learning_rate": 2e-05,
      "loss": 0.2231,
      "step": 955
    },
    {
      "epoch": 0.6461642446772559,
      "grad_norm": 21.087623596191406,
      "learning_rate": 2e-05,
      "loss": 0.2604,
      "step": 956
    },
    {
      "epoch": 0.6468401486988847,
      "grad_norm": 30.040668487548828,
      "learning_rate": 2e-05,
      "loss": 0.2283,
      "step": 957
    },
    {
      "epoch": 0.6475160527205137,
      "grad_norm": 7.995543003082275,
      "learning_rate": 2e-05,
      "loss": 0.2446,
      "step": 958
    },
    {
      "epoch": 0.6481919567421426,
      "grad_norm": 24.88127899169922,
      "learning_rate": 2e-05,
      "loss": 0.2065,
      "step": 959
    },
    {
      "epoch": 0.6488678607637716,
      "grad_norm": 24.258468627929688,
      "learning_rate": 2e-05,
      "loss": 0.2489,
      "step": 960
    },
    {
      "epoch": 0.6495437647854004,
      "grad_norm": 5.214273929595947,
      "learning_rate": 2e-05,
      "loss": 0.18,
      "step": 961
    },
    {
      "epoch": 0.6502196688070294,
      "grad_norm": 9.629908561706543,
      "learning_rate": 2e-05,
      "loss": 0.2388,
      "step": 962
    },
    {
      "epoch": 0.6508955728286583,
      "grad_norm": 6.560550689697266,
      "learning_rate": 2e-05,
      "loss": 0.2039,
      "step": 963
    },
    {
      "epoch": 0.6515714768502873,
      "grad_norm": 7.172302722930908,
      "learning_rate": 2e-05,
      "loss": 0.2029,
      "step": 964
    },
    {
      "epoch": 0.6522473808719161,
      "grad_norm": 9.361584663391113,
      "learning_rate": 2e-05,
      "loss": 0.2128,
      "step": 965
    },
    {
      "epoch": 0.6529232848935451,
      "grad_norm": 17.794818878173828,
      "learning_rate": 2e-05,
      "loss": 0.264,
      "step": 966
    },
    {
      "epoch": 0.6535991889151741,
      "grad_norm": 8.392476081848145,
      "learning_rate": 2e-05,
      "loss": 0.227,
      "step": 967
    },
    {
      "epoch": 0.654275092936803,
      "grad_norm": 15.336060523986816,
      "learning_rate": 2e-05,
      "loss": 0.2264,
      "step": 968
    },
    {
      "epoch": 0.654950996958432,
      "grad_norm": 5.842196941375732,
      "learning_rate": 2e-05,
      "loss": 0.1767,
      "step": 969
    },
    {
      "epoch": 0.6556269009800608,
      "grad_norm": 10.964766502380371,
      "learning_rate": 2e-05,
      "loss": 0.2578,
      "step": 970
    },
    {
      "epoch": 0.6563028050016898,
      "grad_norm": 7.23557186126709,
      "learning_rate": 2e-05,
      "loss": 0.2644,
      "step": 971
    },
    {
      "epoch": 0.6569787090233187,
      "grad_norm": 6.613933086395264,
      "learning_rate": 2e-05,
      "loss": 0.1778,
      "step": 972
    },
    {
      "epoch": 0.6576546130449477,
      "grad_norm": 11.548998832702637,
      "learning_rate": 2e-05,
      "loss": 0.192,
      "step": 973
    },
    {
      "epoch": 0.6583305170665765,
      "grad_norm": 21.71702766418457,
      "learning_rate": 2e-05,
      "loss": 0.256,
      "step": 974
    },
    {
      "epoch": 0.6590064210882055,
      "grad_norm": 6.417525291442871,
      "learning_rate": 2e-05,
      "loss": 0.2335,
      "step": 975
    },
    {
      "epoch": 0.6596823251098344,
      "grad_norm": 6.790191650390625,
      "learning_rate": 2e-05,
      "loss": 0.2032,
      "step": 976
    },
    {
      "epoch": 0.6603582291314634,
      "grad_norm": 19.544801712036133,
      "learning_rate": 2e-05,
      "loss": 0.2461,
      "step": 977
    },
    {
      "epoch": 0.6610341331530922,
      "grad_norm": 14.863484382629395,
      "learning_rate": 2e-05,
      "loss": 0.212,
      "step": 978
    },
    {
      "epoch": 0.6617100371747212,
      "grad_norm": 13.733709335327148,
      "learning_rate": 2e-05,
      "loss": 0.1827,
      "step": 979
    },
    {
      "epoch": 0.6623859411963501,
      "grad_norm": 13.68557357788086,
      "learning_rate": 2e-05,
      "loss": 0.2585,
      "step": 980
    },
    {
      "epoch": 0.6630618452179791,
      "grad_norm": 18.7807674407959,
      "learning_rate": 2e-05,
      "loss": 0.2719,
      "step": 981
    },
    {
      "epoch": 0.6637377492396079,
      "grad_norm": 11.724921226501465,
      "learning_rate": 2e-05,
      "loss": 0.2766,
      "step": 982
    },
    {
      "epoch": 0.6644136532612369,
      "grad_norm": 13.09675121307373,
      "learning_rate": 2e-05,
      "loss": 0.2111,
      "step": 983
    },
    {
      "epoch": 0.6650895572828658,
      "grad_norm": 6.225228786468506,
      "learning_rate": 2e-05,
      "loss": 0.1957,
      "step": 984
    },
    {
      "epoch": 0.6657654613044948,
      "grad_norm": 5.069770336151123,
      "learning_rate": 2e-05,
      "loss": 0.1859,
      "step": 985
    },
    {
      "epoch": 0.6664413653261236,
      "grad_norm": 7.252865314483643,
      "learning_rate": 2e-05,
      "loss": 0.2167,
      "step": 986
    },
    {
      "epoch": 0.6671172693477526,
      "grad_norm": 6.324028015136719,
      "learning_rate": 2e-05,
      "loss": 0.2375,
      "step": 987
    },
    {
      "epoch": 0.6677931733693816,
      "grad_norm": 7.382453918457031,
      "learning_rate": 2e-05,
      "loss": 0.1932,
      "step": 988
    },
    {
      "epoch": 0.6684690773910105,
      "grad_norm": 18.762386322021484,
      "learning_rate": 2e-05,
      "loss": 0.2582,
      "step": 989
    },
    {
      "epoch": 0.6691449814126395,
      "grad_norm": 13.117562294006348,
      "learning_rate": 2e-05,
      "loss": 0.3303,
      "step": 990
    },
    {
      "epoch": 0.6698208854342683,
      "grad_norm": 6.673574924468994,
      "learning_rate": 2e-05,
      "loss": 0.2177,
      "step": 991
    },
    {
      "epoch": 0.6704967894558973,
      "grad_norm": 9.75029468536377,
      "learning_rate": 2e-05,
      "loss": 0.1988,
      "step": 992
    },
    {
      "epoch": 0.6711726934775262,
      "grad_norm": 13.865641593933105,
      "learning_rate": 2e-05,
      "loss": 0.2423,
      "step": 993
    },
    {
      "epoch": 0.6718485974991552,
      "grad_norm": 10.177987098693848,
      "learning_rate": 2e-05,
      "loss": 0.2392,
      "step": 994
    },
    {
      "epoch": 0.672524501520784,
      "grad_norm": 10.84850788116455,
      "learning_rate": 2e-05,
      "loss": 0.2429,
      "step": 995
    },
    {
      "epoch": 0.673200405542413,
      "grad_norm": 4.13468599319458,
      "learning_rate": 2e-05,
      "loss": 0.2086,
      "step": 996
    },
    {
      "epoch": 0.6738763095640419,
      "grad_norm": 14.262582778930664,
      "learning_rate": 2e-05,
      "loss": 0.259,
      "step": 997
    },
    {
      "epoch": 0.6745522135856709,
      "grad_norm": 17.153623580932617,
      "learning_rate": 2e-05,
      "loss": 0.2766,
      "step": 998
    },
    {
      "epoch": 0.6752281176072997,
      "grad_norm": 5.940430164337158,
      "learning_rate": 2e-05,
      "loss": 0.186,
      "step": 999
    },
    {
      "epoch": 0.6759040216289287,
      "grad_norm": 4.768993377685547,
      "learning_rate": 2e-05,
      "loss": 0.2499,
      "step": 1000
    },
    {
      "epoch": 0.6759040216289287,
      "eval_accuracy": 0.4924831081081081,
      "eval_loss": 0.23884893953800201,
      "eval_runtime": 43.237,
      "eval_samples_per_second": 547.679,
      "eval_steps_per_second": 17.115,
      "step": 1000
    },
    {
      "epoch": 0.6765799256505576,
      "grad_norm": 5.785939693450928,
      "learning_rate": 2e-05,
      "loss": 0.1994,
      "step": 1001
    },
    {
      "epoch": 0.6772558296721866,
      "grad_norm": 6.666167259216309,
      "learning_rate": 2e-05,
      "loss": 0.2548,
      "step": 1002
    },
    {
      "epoch": 0.6779317336938154,
      "grad_norm": 11.325234413146973,
      "learning_rate": 2e-05,
      "loss": 0.2323,
      "step": 1003
    },
    {
      "epoch": 0.6786076377154444,
      "grad_norm": 13.005061149597168,
      "learning_rate": 2e-05,
      "loss": 0.1953,
      "step": 1004
    },
    {
      "epoch": 0.6792835417370733,
      "grad_norm": 12.69682788848877,
      "learning_rate": 2e-05,
      "loss": 0.1959,
      "step": 1005
    },
    {
      "epoch": 0.6799594457587023,
      "grad_norm": 5.653406143188477,
      "learning_rate": 2e-05,
      "loss": 0.2171,
      "step": 1006
    },
    {
      "epoch": 0.6806353497803312,
      "grad_norm": 6.353088855743408,
      "learning_rate": 2e-05,
      "loss": 0.2064,
      "step": 1007
    },
    {
      "epoch": 0.6813112538019601,
      "grad_norm": 13.225577354431152,
      "learning_rate": 2e-05,
      "loss": 0.2164,
      "step": 1008
    },
    {
      "epoch": 0.6819871578235891,
      "grad_norm": 6.649354457855225,
      "learning_rate": 2e-05,
      "loss": 0.2121,
      "step": 1009
    },
    {
      "epoch": 0.682663061845218,
      "grad_norm": 5.866961479187012,
      "learning_rate": 2e-05,
      "loss": 0.2095,
      "step": 1010
    },
    {
      "epoch": 0.683338965866847,
      "grad_norm": 7.9010114669799805,
      "learning_rate": 2e-05,
      "loss": 0.187,
      "step": 1011
    },
    {
      "epoch": 0.6840148698884758,
      "grad_norm": 14.979111671447754,
      "learning_rate": 2e-05,
      "loss": 0.2172,
      "step": 1012
    },
    {
      "epoch": 0.6846907739101048,
      "grad_norm": 9.778497695922852,
      "learning_rate": 2e-05,
      "loss": 0.1863,
      "step": 1013
    },
    {
      "epoch": 0.6853666779317337,
      "grad_norm": 7.4816060066223145,
      "learning_rate": 2e-05,
      "loss": 0.238,
      "step": 1014
    },
    {
      "epoch": 0.6860425819533627,
      "grad_norm": 5.56958532333374,
      "learning_rate": 2e-05,
      "loss": 0.1762,
      "step": 1015
    },
    {
      "epoch": 0.6867184859749915,
      "grad_norm": 10.95412826538086,
      "learning_rate": 2e-05,
      "loss": 0.2151,
      "step": 1016
    },
    {
      "epoch": 0.6873943899966205,
      "grad_norm": 24.208843231201172,
      "learning_rate": 2e-05,
      "loss": 0.227,
      "step": 1017
    },
    {
      "epoch": 0.6880702940182494,
      "grad_norm": 11.7774019241333,
      "learning_rate": 2e-05,
      "loss": 0.2314,
      "step": 1018
    },
    {
      "epoch": 0.6887461980398784,
      "grad_norm": 7.816908836364746,
      "learning_rate": 2e-05,
      "loss": 0.2001,
      "step": 1019
    },
    {
      "epoch": 0.6894221020615072,
      "grad_norm": 31.480722427368164,
      "learning_rate": 2e-05,
      "loss": 0.2558,
      "step": 1020
    },
    {
      "epoch": 0.6900980060831362,
      "grad_norm": 34.0113525390625,
      "learning_rate": 2e-05,
      "loss": 0.2997,
      "step": 1021
    },
    {
      "epoch": 0.6907739101047651,
      "grad_norm": 6.395514488220215,
      "learning_rate": 2e-05,
      "loss": 0.1938,
      "step": 1022
    },
    {
      "epoch": 0.6914498141263941,
      "grad_norm": 7.920341491699219,
      "learning_rate": 2e-05,
      "loss": 0.1747,
      "step": 1023
    },
    {
      "epoch": 0.6921257181480229,
      "grad_norm": 27.69765853881836,
      "learning_rate": 2e-05,
      "loss": 0.265,
      "step": 1024
    },
    {
      "epoch": 0.6928016221696519,
      "grad_norm": 16.156330108642578,
      "learning_rate": 2e-05,
      "loss": 0.248,
      "step": 1025
    },
    {
      "epoch": 0.6934775261912809,
      "grad_norm": 15.444580078125,
      "learning_rate": 2e-05,
      "loss": 0.2257,
      "step": 1026
    },
    {
      "epoch": 0.6941534302129098,
      "grad_norm": 13.50002670288086,
      "learning_rate": 2e-05,
      "loss": 0.1777,
      "step": 1027
    },
    {
      "epoch": 0.6948293342345387,
      "grad_norm": 19.000240325927734,
      "learning_rate": 2e-05,
      "loss": 0.2208,
      "step": 1028
    },
    {
      "epoch": 0.6955052382561676,
      "grad_norm": 19.04424285888672,
      "learning_rate": 2e-05,
      "loss": 0.2492,
      "step": 1029
    },
    {
      "epoch": 0.6961811422777966,
      "grad_norm": 6.247470378875732,
      "learning_rate": 2e-05,
      "loss": 0.2076,
      "step": 1030
    },
    {
      "epoch": 0.6968570462994255,
      "grad_norm": 21.067174911499023,
      "learning_rate": 2e-05,
      "loss": 0.1716,
      "step": 1031
    },
    {
      "epoch": 0.6975329503210544,
      "grad_norm": 11.34173583984375,
      "learning_rate": 2e-05,
      "loss": 0.2247,
      "step": 1032
    },
    {
      "epoch": 0.6982088543426833,
      "grad_norm": 7.067166805267334,
      "learning_rate": 2e-05,
      "loss": 0.2179,
      "step": 1033
    },
    {
      "epoch": 0.6988847583643123,
      "grad_norm": 20.66969108581543,
      "learning_rate": 2e-05,
      "loss": 0.2146,
      "step": 1034
    },
    {
      "epoch": 0.6995606623859412,
      "grad_norm": 15.261811256408691,
      "learning_rate": 2e-05,
      "loss": 0.2096,
      "step": 1035
    },
    {
      "epoch": 0.7002365664075701,
      "grad_norm": 21.30474281311035,
      "learning_rate": 2e-05,
      "loss": 0.2373,
      "step": 1036
    },
    {
      "epoch": 0.700912470429199,
      "grad_norm": 5.884871482849121,
      "learning_rate": 2e-05,
      "loss": 0.2089,
      "step": 1037
    },
    {
      "epoch": 0.701588374450828,
      "grad_norm": 13.929277420043945,
      "learning_rate": 2e-05,
      "loss": 0.22,
      "step": 1038
    },
    {
      "epoch": 0.7022642784724569,
      "grad_norm": 5.325328826904297,
      "learning_rate": 2e-05,
      "loss": 0.2009,
      "step": 1039
    },
    {
      "epoch": 0.7029401824940859,
      "grad_norm": 8.752159118652344,
      "learning_rate": 2e-05,
      "loss": 0.2,
      "step": 1040
    },
    {
      "epoch": 0.7036160865157147,
      "grad_norm": 17.963991165161133,
      "learning_rate": 2e-05,
      "loss": 0.2571,
      "step": 1041
    },
    {
      "epoch": 0.7042919905373437,
      "grad_norm": 13.15282154083252,
      "learning_rate": 2e-05,
      "loss": 0.211,
      "step": 1042
    },
    {
      "epoch": 0.7049678945589726,
      "grad_norm": 10.036561012268066,
      "learning_rate": 2e-05,
      "loss": 0.2541,
      "step": 1043
    },
    {
      "epoch": 0.7056437985806016,
      "grad_norm": 24.137882232666016,
      "learning_rate": 2e-05,
      "loss": 0.2776,
      "step": 1044
    },
    {
      "epoch": 0.7063197026022305,
      "grad_norm": 6.232251167297363,
      "learning_rate": 2e-05,
      "loss": 0.1947,
      "step": 1045
    },
    {
      "epoch": 0.7069956066238594,
      "grad_norm": 8.344635963439941,
      "learning_rate": 2e-05,
      "loss": 0.2019,
      "step": 1046
    },
    {
      "epoch": 0.7076715106454884,
      "grad_norm": 10.985777854919434,
      "learning_rate": 2e-05,
      "loss": 0.2187,
      "step": 1047
    },
    {
      "epoch": 0.7083474146671173,
      "grad_norm": 8.773568153381348,
      "learning_rate": 2e-05,
      "loss": 0.1701,
      "step": 1048
    },
    {
      "epoch": 0.7090233186887462,
      "grad_norm": 11.976913452148438,
      "learning_rate": 2e-05,
      "loss": 0.187,
      "step": 1049
    },
    {
      "epoch": 0.7096992227103751,
      "grad_norm": 9.368999481201172,
      "learning_rate": 2e-05,
      "loss": 0.1948,
      "step": 1050
    },
    {
      "epoch": 0.7096992227103751,
      "eval_accuracy": 0.49408783783783783,
      "eval_loss": 0.2504537105560303,
      "eval_runtime": 42.981,
      "eval_samples_per_second": 550.942,
      "eval_steps_per_second": 17.217,
      "step": 1050
    },
    {
      "epoch": 0.7103751267320041,
      "grad_norm": 6.375654220581055,
      "learning_rate": 2e-05,
      "loss": 0.2161,
      "step": 1051
    },
    {
      "epoch": 0.711051030753633,
      "grad_norm": 10.110198974609375,
      "learning_rate": 2e-05,
      "loss": 0.2031,
      "step": 1052
    },
    {
      "epoch": 0.7117269347752619,
      "grad_norm": 7.251458168029785,
      "learning_rate": 2e-05,
      "loss": 0.1991,
      "step": 1053
    },
    {
      "epoch": 0.7124028387968908,
      "grad_norm": 7.68689489364624,
      "learning_rate": 2e-05,
      "loss": 0.194,
      "step": 1054
    },
    {
      "epoch": 0.7130787428185198,
      "grad_norm": 22.308534622192383,
      "learning_rate": 2e-05,
      "loss": 0.196,
      "step": 1055
    },
    {
      "epoch": 0.7137546468401487,
      "grad_norm": 8.831141471862793,
      "learning_rate": 2e-05,
      "loss": 0.1796,
      "step": 1056
    },
    {
      "epoch": 0.7144305508617776,
      "grad_norm": 6.209223747253418,
      "learning_rate": 2e-05,
      "loss": 0.1857,
      "step": 1057
    },
    {
      "epoch": 0.7151064548834065,
      "grad_norm": 8.361316680908203,
      "learning_rate": 2e-05,
      "loss": 0.227,
      "step": 1058
    },
    {
      "epoch": 0.7157823589050355,
      "grad_norm": 15.114814758300781,
      "learning_rate": 2e-05,
      "loss": 0.1693,
      "step": 1059
    },
    {
      "epoch": 0.7164582629266644,
      "grad_norm": 18.22860336303711,
      "learning_rate": 2e-05,
      "loss": 0.2919,
      "step": 1060
    },
    {
      "epoch": 0.7171341669482934,
      "grad_norm": 14.51125431060791,
      "learning_rate": 2e-05,
      "loss": 0.1835,
      "step": 1061
    },
    {
      "epoch": 0.7178100709699222,
      "grad_norm": 15.58604907989502,
      "learning_rate": 2e-05,
      "loss": 0.2127,
      "step": 1062
    },
    {
      "epoch": 0.7184859749915512,
      "grad_norm": 6.652542591094971,
      "learning_rate": 2e-05,
      "loss": 0.2283,
      "step": 1063
    },
    {
      "epoch": 0.7191618790131802,
      "grad_norm": 6.00473165512085,
      "learning_rate": 2e-05,
      "loss": 0.2145,
      "step": 1064
    },
    {
      "epoch": 0.719837783034809,
      "grad_norm": 23.68505096435547,
      "learning_rate": 2e-05,
      "loss": 0.2678,
      "step": 1065
    },
    {
      "epoch": 0.720513687056438,
      "grad_norm": 17.581308364868164,
      "learning_rate": 2e-05,
      "loss": 0.2366,
      "step": 1066
    },
    {
      "epoch": 0.7211895910780669,
      "grad_norm": 12.790616035461426,
      "learning_rate": 2e-05,
      "loss": 0.2199,
      "step": 1067
    },
    {
      "epoch": 0.7218654950996959,
      "grad_norm": 6.462890148162842,
      "learning_rate": 2e-05,
      "loss": 0.183,
      "step": 1068
    },
    {
      "epoch": 0.7225413991213248,
      "grad_norm": 15.078261375427246,
      "learning_rate": 2e-05,
      "loss": 0.3078,
      "step": 1069
    },
    {
      "epoch": 0.7232173031429537,
      "grad_norm": 5.97918176651001,
      "learning_rate": 2e-05,
      "loss": 0.2257,
      "step": 1070
    },
    {
      "epoch": 0.7238932071645826,
      "grad_norm": 5.661224365234375,
      "learning_rate": 2e-05,
      "loss": 0.1666,
      "step": 1071
    },
    {
      "epoch": 0.7245691111862116,
      "grad_norm": 5.955348014831543,
      "learning_rate": 2e-05,
      "loss": 0.1997,
      "step": 1072
    },
    {
      "epoch": 0.7252450152078405,
      "grad_norm": 14.627803802490234,
      "learning_rate": 2e-05,
      "loss": 0.1924,
      "step": 1073
    },
    {
      "epoch": 0.7259209192294694,
      "grad_norm": 9.531682968139648,
      "learning_rate": 2e-05,
      "loss": 0.1688,
      "step": 1074
    },
    {
      "epoch": 0.7265968232510983,
      "grad_norm": 5.604430675506592,
      "learning_rate": 2e-05,
      "loss": 0.183,
      "step": 1075
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 5.706563472747803,
      "learning_rate": 2e-05,
      "loss": 0.1445,
      "step": 1076
    },
    {
      "epoch": 0.7279486312943562,
      "grad_norm": 7.352299690246582,
      "learning_rate": 2e-05,
      "loss": 0.2202,
      "step": 1077
    },
    {
      "epoch": 0.7286245353159851,
      "grad_norm": 15.233855247497559,
      "learning_rate": 2e-05,
      "loss": 0.2796,
      "step": 1078
    },
    {
      "epoch": 0.729300439337614,
      "grad_norm": 18.072532653808594,
      "learning_rate": 2e-05,
      "loss": 0.2184,
      "step": 1079
    },
    {
      "epoch": 0.729976343359243,
      "grad_norm": 8.069234848022461,
      "learning_rate": 2e-05,
      "loss": 0.2624,
      "step": 1080
    },
    {
      "epoch": 0.7306522473808719,
      "grad_norm": 15.25510311126709,
      "learning_rate": 2e-05,
      "loss": 0.2669,
      "step": 1081
    },
    {
      "epoch": 0.7313281514025008,
      "grad_norm": 13.413167953491211,
      "learning_rate": 2e-05,
      "loss": 0.2271,
      "step": 1082
    },
    {
      "epoch": 0.7320040554241297,
      "grad_norm": 13.497303009033203,
      "learning_rate": 2e-05,
      "loss": 0.1992,
      "step": 1083
    },
    {
      "epoch": 0.7326799594457587,
      "grad_norm": 8.783838272094727,
      "learning_rate": 2e-05,
      "loss": 0.2155,
      "step": 1084
    },
    {
      "epoch": 0.7333558634673877,
      "grad_norm": 13.636452674865723,
      "learning_rate": 2e-05,
      "loss": 0.3004,
      "step": 1085
    },
    {
      "epoch": 0.7340317674890166,
      "grad_norm": 4.58281135559082,
      "learning_rate": 2e-05,
      "loss": 0.22,
      "step": 1086
    },
    {
      "epoch": 0.7347076715106455,
      "grad_norm": 9.679242134094238,
      "learning_rate": 2e-05,
      "loss": 0.2844,
      "step": 1087
    },
    {
      "epoch": 0.7353835755322744,
      "grad_norm": 9.497397422790527,
      "learning_rate": 2e-05,
      "loss": 0.1883,
      "step": 1088
    },
    {
      "epoch": 0.7360594795539034,
      "grad_norm": 12.795823097229004,
      "learning_rate": 2e-05,
      "loss": 0.1943,
      "step": 1089
    },
    {
      "epoch": 0.7367353835755323,
      "grad_norm": 15.079276084899902,
      "learning_rate": 2e-05,
      "loss": 0.2479,
      "step": 1090
    },
    {
      "epoch": 0.7374112875971612,
      "grad_norm": 19.834909439086914,
      "learning_rate": 2e-05,
      "loss": 0.2304,
      "step": 1091
    },
    {
      "epoch": 0.7380871916187901,
      "grad_norm": 13.948566436767578,
      "learning_rate": 2e-05,
      "loss": 0.2208,
      "step": 1092
    },
    {
      "epoch": 0.7387630956404191,
      "grad_norm": 5.014163017272949,
      "learning_rate": 2e-05,
      "loss": 0.2047,
      "step": 1093
    },
    {
      "epoch": 0.739438999662048,
      "grad_norm": 35.45634460449219,
      "learning_rate": 2e-05,
      "loss": 0.2913,
      "step": 1094
    },
    {
      "epoch": 0.7401149036836769,
      "grad_norm": 23.73724365234375,
      "learning_rate": 2e-05,
      "loss": 0.1858,
      "step": 1095
    },
    {
      "epoch": 0.7407908077053058,
      "grad_norm": 12.93753433227539,
      "learning_rate": 2e-05,
      "loss": 0.2357,
      "step": 1096
    },
    {
      "epoch": 0.7414667117269348,
      "grad_norm": 20.563678741455078,
      "learning_rate": 2e-05,
      "loss": 0.2563,
      "step": 1097
    },
    {
      "epoch": 0.7421426157485637,
      "grad_norm": 16.740474700927734,
      "learning_rate": 2e-05,
      "loss": 0.1917,
      "step": 1098
    },
    {
      "epoch": 0.7428185197701926,
      "grad_norm": 17.281414031982422,
      "learning_rate": 2e-05,
      "loss": 0.2206,
      "step": 1099
    },
    {
      "epoch": 0.7434944237918215,
      "grad_norm": 33.13484191894531,
      "learning_rate": 2e-05,
      "loss": 0.2477,
      "step": 1100
    },
    {
      "epoch": 0.7434944237918215,
      "eval_accuracy": 0.46368243243243246,
      "eval_loss": 0.2543576657772064,
      "eval_runtime": 43.667,
      "eval_samples_per_second": 542.287,
      "eval_steps_per_second": 16.946,
      "step": 1100
    },
    {
      "epoch": 0.7441703278134505,
      "grad_norm": 17.829689025878906,
      "learning_rate": 2e-05,
      "loss": 0.2114,
      "step": 1101
    },
    {
      "epoch": 0.7448462318350794,
      "grad_norm": 5.902535915374756,
      "learning_rate": 2e-05,
      "loss": 0.2295,
      "step": 1102
    },
    {
      "epoch": 0.7455221358567083,
      "grad_norm": 21.668373107910156,
      "learning_rate": 2e-05,
      "loss": 0.2584,
      "step": 1103
    },
    {
      "epoch": 0.7461980398783373,
      "grad_norm": 24.67908477783203,
      "learning_rate": 2e-05,
      "loss": 0.2007,
      "step": 1104
    },
    {
      "epoch": 0.7468739438999662,
      "grad_norm": 6.308023452758789,
      "learning_rate": 2e-05,
      "loss": 0.1407,
      "step": 1105
    },
    {
      "epoch": 0.7475498479215952,
      "grad_norm": 15.159717559814453,
      "learning_rate": 2e-05,
      "loss": 0.2093,
      "step": 1106
    },
    {
      "epoch": 0.748225751943224,
      "grad_norm": 7.329532146453857,
      "learning_rate": 2e-05,
      "loss": 0.202,
      "step": 1107
    },
    {
      "epoch": 0.748901655964853,
      "grad_norm": 10.736015319824219,
      "learning_rate": 2e-05,
      "loss": 0.2618,
      "step": 1108
    },
    {
      "epoch": 0.7495775599864819,
      "grad_norm": 10.156184196472168,
      "learning_rate": 2e-05,
      "loss": 0.1605,
      "step": 1109
    },
    {
      "epoch": 0.7502534640081109,
      "grad_norm": 9.366998672485352,
      "learning_rate": 2e-05,
      "loss": 0.1788,
      "step": 1110
    },
    {
      "epoch": 0.7509293680297398,
      "grad_norm": 10.315082550048828,
      "learning_rate": 2e-05,
      "loss": 0.2072,
      "step": 1111
    },
    {
      "epoch": 0.7516052720513687,
      "grad_norm": 6.533961772918701,
      "learning_rate": 2e-05,
      "loss": 0.1519,
      "step": 1112
    },
    {
      "epoch": 0.7522811760729976,
      "grad_norm": 15.932812690734863,
      "learning_rate": 2e-05,
      "loss": 0.2076,
      "step": 1113
    },
    {
      "epoch": 0.7529570800946266,
      "grad_norm": 10.178567886352539,
      "learning_rate": 2e-05,
      "loss": 0.2044,
      "step": 1114
    },
    {
      "epoch": 0.7536329841162555,
      "grad_norm": 7.442099094390869,
      "learning_rate": 2e-05,
      "loss": 0.1839,
      "step": 1115
    },
    {
      "epoch": 0.7543088881378844,
      "grad_norm": 17.365951538085938,
      "learning_rate": 2e-05,
      "loss": 0.1855,
      "step": 1116
    },
    {
      "epoch": 0.7549847921595133,
      "grad_norm": 7.542236804962158,
      "learning_rate": 2e-05,
      "loss": 0.2034,
      "step": 1117
    },
    {
      "epoch": 0.7556606961811423,
      "grad_norm": 6.920073509216309,
      "learning_rate": 2e-05,
      "loss": 0.1983,
      "step": 1118
    },
    {
      "epoch": 0.7563366002027712,
      "grad_norm": 7.537623405456543,
      "learning_rate": 2e-05,
      "loss": 0.1631,
      "step": 1119
    },
    {
      "epoch": 0.7570125042244001,
      "grad_norm": 7.0059990882873535,
      "learning_rate": 2e-05,
      "loss": 0.1866,
      "step": 1120
    },
    {
      "epoch": 0.757688408246029,
      "grad_norm": 8.363640785217285,
      "learning_rate": 2e-05,
      "loss": 0.1969,
      "step": 1121
    },
    {
      "epoch": 0.758364312267658,
      "grad_norm": 9.751455307006836,
      "learning_rate": 2e-05,
      "loss": 0.2458,
      "step": 1122
    },
    {
      "epoch": 0.759040216289287,
      "grad_norm": 9.328222274780273,
      "learning_rate": 2e-05,
      "loss": 0.1662,
      "step": 1123
    },
    {
      "epoch": 0.7597161203109158,
      "grad_norm": 7.672349452972412,
      "learning_rate": 2e-05,
      "loss": 0.1855,
      "step": 1124
    },
    {
      "epoch": 0.7603920243325448,
      "grad_norm": 13.264748573303223,
      "learning_rate": 2e-05,
      "loss": 0.2545,
      "step": 1125
    },
    {
      "epoch": 0.7610679283541737,
      "grad_norm": 6.912208080291748,
      "learning_rate": 2e-05,
      "loss": 0.176,
      "step": 1126
    },
    {
      "epoch": 0.7617438323758027,
      "grad_norm": 19.757699966430664,
      "learning_rate": 2e-05,
      "loss": 0.2382,
      "step": 1127
    },
    {
      "epoch": 0.7624197363974315,
      "grad_norm": 10.862689018249512,
      "learning_rate": 2e-05,
      "loss": 0.1743,
      "step": 1128
    },
    {
      "epoch": 0.7630956404190605,
      "grad_norm": 8.607519149780273,
      "learning_rate": 2e-05,
      "loss": 0.2056,
      "step": 1129
    },
    {
      "epoch": 0.7637715444406894,
      "grad_norm": 14.789392471313477,
      "learning_rate": 2e-05,
      "loss": 0.2437,
      "step": 1130
    },
    {
      "epoch": 0.7644474484623184,
      "grad_norm": 25.900230407714844,
      "learning_rate": 2e-05,
      "loss": 0.2692,
      "step": 1131
    },
    {
      "epoch": 0.7651233524839472,
      "grad_norm": 11.484848022460938,
      "learning_rate": 2e-05,
      "loss": 0.2368,
      "step": 1132
    },
    {
      "epoch": 0.7657992565055762,
      "grad_norm": 18.097991943359375,
      "learning_rate": 2e-05,
      "loss": 0.1971,
      "step": 1133
    },
    {
      "epoch": 0.7664751605272051,
      "grad_norm": 15.160357475280762,
      "learning_rate": 2e-05,
      "loss": 0.1999,
      "step": 1134
    },
    {
      "epoch": 0.7671510645488341,
      "grad_norm": 7.972028732299805,
      "learning_rate": 2e-05,
      "loss": 0.2045,
      "step": 1135
    },
    {
      "epoch": 0.767826968570463,
      "grad_norm": 5.024528980255127,
      "learning_rate": 2e-05,
      "loss": 0.1681,
      "step": 1136
    },
    {
      "epoch": 0.7685028725920919,
      "grad_norm": 7.397834300994873,
      "learning_rate": 2e-05,
      "loss": 0.2314,
      "step": 1137
    },
    {
      "epoch": 0.7691787766137208,
      "grad_norm": 8.272059440612793,
      "learning_rate": 2e-05,
      "loss": 0.2027,
      "step": 1138
    },
    {
      "epoch": 0.7698546806353498,
      "grad_norm": 16.07627296447754,
      "learning_rate": 2e-05,
      "loss": 0.2956,
      "step": 1139
    },
    {
      "epoch": 0.7705305846569787,
      "grad_norm": 5.9847822189331055,
      "learning_rate": 2e-05,
      "loss": 0.1899,
      "step": 1140
    },
    {
      "epoch": 0.7712064886786076,
      "grad_norm": 12.44365406036377,
      "learning_rate": 2e-05,
      "loss": 0.2288,
      "step": 1141
    },
    {
      "epoch": 0.7718823927002366,
      "grad_norm": 7.141899585723877,
      "learning_rate": 2e-05,
      "loss": 0.2367,
      "step": 1142
    },
    {
      "epoch": 0.7725582967218655,
      "grad_norm": 6.0400543212890625,
      "learning_rate": 2e-05,
      "loss": 0.1812,
      "step": 1143
    },
    {
      "epoch": 0.7732342007434945,
      "grad_norm": 9.75620174407959,
      "learning_rate": 2e-05,
      "loss": 0.2727,
      "step": 1144
    },
    {
      "epoch": 0.7739101047651233,
      "grad_norm": 5.1894354820251465,
      "learning_rate": 2e-05,
      "loss": 0.1922,
      "step": 1145
    },
    {
      "epoch": 0.7745860087867523,
      "grad_norm": 4.697371006011963,
      "learning_rate": 2e-05,
      "loss": 0.2221,
      "step": 1146
    },
    {
      "epoch": 0.7752619128083812,
      "grad_norm": 13.672688484191895,
      "learning_rate": 2e-05,
      "loss": 0.2294,
      "step": 1147
    },
    {
      "epoch": 0.7759378168300102,
      "grad_norm": 8.185792922973633,
      "learning_rate": 2e-05,
      "loss": 0.2436,
      "step": 1148
    },
    {
      "epoch": 0.776613720851639,
      "grad_norm": 12.582009315490723,
      "learning_rate": 2e-05,
      "loss": 0.2302,
      "step": 1149
    },
    {
      "epoch": 0.777289624873268,
      "grad_norm": 8.135284423828125,
      "learning_rate": 2e-05,
      "loss": 0.1806,
      "step": 1150
    },
    {
      "epoch": 0.777289624873268,
      "eval_accuracy": 0.506714527027027,
      "eval_loss": 0.2479238510131836,
      "eval_runtime": 42.9075,
      "eval_samples_per_second": 551.885,
      "eval_steps_per_second": 17.246,
      "step": 1150
    },
    {
      "epoch": 0.7779655288948969,
      "grad_norm": 8.771943092346191,
      "learning_rate": 2e-05,
      "loss": 0.1759,
      "step": 1151
    },
    {
      "epoch": 0.7786414329165259,
      "grad_norm": 6.68550968170166,
      "learning_rate": 2e-05,
      "loss": 0.1793,
      "step": 1152
    },
    {
      "epoch": 0.7793173369381547,
      "grad_norm": 20.542678833007812,
      "learning_rate": 2e-05,
      "loss": 0.3239,
      "step": 1153
    },
    {
      "epoch": 0.7799932409597837,
      "grad_norm": 15.024261474609375,
      "learning_rate": 2e-05,
      "loss": 0.2722,
      "step": 1154
    },
    {
      "epoch": 0.7806691449814126,
      "grad_norm": 4.131154537200928,
      "learning_rate": 2e-05,
      "loss": 0.1453,
      "step": 1155
    },
    {
      "epoch": 0.7813450490030416,
      "grad_norm": 9.97591781616211,
      "learning_rate": 2e-05,
      "loss": 0.1834,
      "step": 1156
    },
    {
      "epoch": 0.7820209530246705,
      "grad_norm": 7.113241195678711,
      "learning_rate": 2e-05,
      "loss": 0.2149,
      "step": 1157
    },
    {
      "epoch": 0.7826968570462994,
      "grad_norm": 24.356287002563477,
      "learning_rate": 2e-05,
      "loss": 0.2737,
      "step": 1158
    },
    {
      "epoch": 0.7833727610679283,
      "grad_norm": 20.656835556030273,
      "learning_rate": 2e-05,
      "loss": 0.2153,
      "step": 1159
    },
    {
      "epoch": 0.7840486650895573,
      "grad_norm": 5.064976215362549,
      "learning_rate": 2e-05,
      "loss": 0.2138,
      "step": 1160
    },
    {
      "epoch": 0.7847245691111863,
      "grad_norm": 7.328847885131836,
      "learning_rate": 2e-05,
      "loss": 0.2804,
      "step": 1161
    },
    {
      "epoch": 0.7854004731328151,
      "grad_norm": 9.451038360595703,
      "learning_rate": 2e-05,
      "loss": 0.1842,
      "step": 1162
    },
    {
      "epoch": 0.7860763771544441,
      "grad_norm": 10.150983810424805,
      "learning_rate": 2e-05,
      "loss": 0.2288,
      "step": 1163
    },
    {
      "epoch": 0.786752281176073,
      "grad_norm": 7.798166751861572,
      "learning_rate": 2e-05,
      "loss": 0.1585,
      "step": 1164
    },
    {
      "epoch": 0.787428185197702,
      "grad_norm": 7.807036399841309,
      "learning_rate": 2e-05,
      "loss": 0.2105,
      "step": 1165
    },
    {
      "epoch": 0.7881040892193308,
      "grad_norm": 11.663860321044922,
      "learning_rate": 2e-05,
      "loss": 0.1836,
      "step": 1166
    },
    {
      "epoch": 0.7887799932409598,
      "grad_norm": 6.4970550537109375,
      "learning_rate": 2e-05,
      "loss": 0.2168,
      "step": 1167
    },
    {
      "epoch": 0.7894558972625887,
      "grad_norm": 6.069588661193848,
      "learning_rate": 2e-05,
      "loss": 0.2362,
      "step": 1168
    },
    {
      "epoch": 0.7901318012842177,
      "grad_norm": 9.050861358642578,
      "learning_rate": 2e-05,
      "loss": 0.1991,
      "step": 1169
    },
    {
      "epoch": 0.7908077053058465,
      "grad_norm": 14.853404998779297,
      "learning_rate": 2e-05,
      "loss": 0.1889,
      "step": 1170
    },
    {
      "epoch": 0.7914836093274755,
      "grad_norm": 11.839611053466797,
      "learning_rate": 2e-05,
      "loss": 0.2273,
      "step": 1171
    },
    {
      "epoch": 0.7921595133491044,
      "grad_norm": 7.285501480102539,
      "learning_rate": 2e-05,
      "loss": 0.186,
      "step": 1172
    },
    {
      "epoch": 0.7928354173707334,
      "grad_norm": 6.155836582183838,
      "learning_rate": 2e-05,
      "loss": 0.224,
      "step": 1173
    },
    {
      "epoch": 0.7935113213923622,
      "grad_norm": 11.46066951751709,
      "learning_rate": 2e-05,
      "loss": 0.182,
      "step": 1174
    },
    {
      "epoch": 0.7941872254139912,
      "grad_norm": 7.4410719871521,
      "learning_rate": 2e-05,
      "loss": 0.2066,
      "step": 1175
    },
    {
      "epoch": 0.7948631294356201,
      "grad_norm": 6.505594730377197,
      "learning_rate": 2e-05,
      "loss": 0.2186,
      "step": 1176
    },
    {
      "epoch": 0.7955390334572491,
      "grad_norm": 6.885608673095703,
      "learning_rate": 2e-05,
      "loss": 0.1907,
      "step": 1177
    },
    {
      "epoch": 0.796214937478878,
      "grad_norm": 8.011312484741211,
      "learning_rate": 2e-05,
      "loss": 0.2374,
      "step": 1178
    },
    {
      "epoch": 0.7968908415005069,
      "grad_norm": 17.985292434692383,
      "learning_rate": 2e-05,
      "loss": 0.2565,
      "step": 1179
    },
    {
      "epoch": 0.7975667455221359,
      "grad_norm": 6.796809196472168,
      "learning_rate": 2e-05,
      "loss": 0.2096,
      "step": 1180
    },
    {
      "epoch": 0.7982426495437648,
      "grad_norm": 6.347496509552002,
      "learning_rate": 2e-05,
      "loss": 0.2109,
      "step": 1181
    },
    {
      "epoch": 0.7989185535653938,
      "grad_norm": 10.744550704956055,
      "learning_rate": 2e-05,
      "loss": 0.2198,
      "step": 1182
    },
    {
      "epoch": 0.7995944575870226,
      "grad_norm": 6.117054462432861,
      "learning_rate": 2e-05,
      "loss": 0.1659,
      "step": 1183
    },
    {
      "epoch": 0.8002703616086516,
      "grad_norm": 21.302350997924805,
      "learning_rate": 2e-05,
      "loss": 0.2265,
      "step": 1184
    },
    {
      "epoch": 0.8009462656302805,
      "grad_norm": 7.636870384216309,
      "learning_rate": 2e-05,
      "loss": 0.2694,
      "step": 1185
    },
    {
      "epoch": 0.8016221696519095,
      "grad_norm": 14.193074226379395,
      "learning_rate": 2e-05,
      "loss": 0.2435,
      "step": 1186
    },
    {
      "epoch": 0.8022980736735383,
      "grad_norm": 8.847366333007812,
      "learning_rate": 2e-05,
      "loss": 0.1845,
      "step": 1187
    },
    {
      "epoch": 0.8029739776951673,
      "grad_norm": 17.352949142456055,
      "learning_rate": 2e-05,
      "loss": 0.2038,
      "step": 1188
    },
    {
      "epoch": 0.8036498817167962,
      "grad_norm": 6.498169422149658,
      "learning_rate": 2e-05,
      "loss": 0.1916,
      "step": 1189
    },
    {
      "epoch": 0.8043257857384252,
      "grad_norm": 7.721607685089111,
      "learning_rate": 2e-05,
      "loss": 0.1686,
      "step": 1190
    },
    {
      "epoch": 0.805001689760054,
      "grad_norm": 11.49802017211914,
      "learning_rate": 2e-05,
      "loss": 0.2002,
      "step": 1191
    },
    {
      "epoch": 0.805677593781683,
      "grad_norm": 7.777790546417236,
      "learning_rate": 2e-05,
      "loss": 0.2222,
      "step": 1192
    },
    {
      "epoch": 0.8063534978033119,
      "grad_norm": 13.689692497253418,
      "learning_rate": 2e-05,
      "loss": 0.1983,
      "step": 1193
    },
    {
      "epoch": 0.8070294018249409,
      "grad_norm": 8.7461519241333,
      "learning_rate": 2e-05,
      "loss": 0.2147,
      "step": 1194
    },
    {
      "epoch": 0.8077053058465697,
      "grad_norm": 5.51593542098999,
      "learning_rate": 2e-05,
      "loss": 0.1525,
      "step": 1195
    },
    {
      "epoch": 0.8083812098681987,
      "grad_norm": 6.755614757537842,
      "learning_rate": 2e-05,
      "loss": 0.1995,
      "step": 1196
    },
    {
      "epoch": 0.8090571138898276,
      "grad_norm": 10.919930458068848,
      "learning_rate": 2e-05,
      "loss": 0.2008,
      "step": 1197
    },
    {
      "epoch": 0.8097330179114566,
      "grad_norm": 8.546280860900879,
      "learning_rate": 2e-05,
      "loss": 0.1794,
      "step": 1198
    },
    {
      "epoch": 0.8104089219330854,
      "grad_norm": 7.8644585609436035,
      "learning_rate": 2e-05,
      "loss": 0.2511,
      "step": 1199
    },
    {
      "epoch": 0.8110848259547144,
      "grad_norm": 6.196953773498535,
      "learning_rate": 2e-05,
      "loss": 0.2104,
      "step": 1200
    },
    {
      "epoch": 0.8110848259547144,
      "eval_accuracy": 0.5048141891891892,
      "eval_loss": 0.2510896325111389,
      "eval_runtime": 43.3431,
      "eval_samples_per_second": 546.339,
      "eval_steps_per_second": 17.073,
      "step": 1200
    },
    {
      "epoch": 0.8117607299763434,
      "grad_norm": 6.848859786987305,
      "learning_rate": 2e-05,
      "loss": 0.1585,
      "step": 1201
    },
    {
      "epoch": 0.8124366339979723,
      "grad_norm": 5.62040376663208,
      "learning_rate": 2e-05,
      "loss": 0.1511,
      "step": 1202
    },
    {
      "epoch": 0.8131125380196013,
      "grad_norm": 6.895204544067383,
      "learning_rate": 2e-05,
      "loss": 0.2003,
      "step": 1203
    },
    {
      "epoch": 0.8137884420412301,
      "grad_norm": 8.006102561950684,
      "learning_rate": 2e-05,
      "loss": 0.2143,
      "step": 1204
    },
    {
      "epoch": 0.8144643460628591,
      "grad_norm": 7.573379039764404,
      "learning_rate": 2e-05,
      "loss": 0.2732,
      "step": 1205
    },
    {
      "epoch": 0.815140250084488,
      "grad_norm": 4.748726844787598,
      "learning_rate": 2e-05,
      "loss": 0.1735,
      "step": 1206
    },
    {
      "epoch": 0.815816154106117,
      "grad_norm": 7.2698187828063965,
      "learning_rate": 2e-05,
      "loss": 0.1788,
      "step": 1207
    },
    {
      "epoch": 0.8164920581277458,
      "grad_norm": 19.041854858398438,
      "learning_rate": 2e-05,
      "loss": 0.2161,
      "step": 1208
    },
    {
      "epoch": 0.8171679621493748,
      "grad_norm": 5.435664176940918,
      "learning_rate": 2e-05,
      "loss": 0.1813,
      "step": 1209
    },
    {
      "epoch": 0.8178438661710037,
      "grad_norm": 11.679469108581543,
      "learning_rate": 2e-05,
      "loss": 0.2355,
      "step": 1210
    },
    {
      "epoch": 0.8185197701926327,
      "grad_norm": 8.989809036254883,
      "learning_rate": 2e-05,
      "loss": 0.1799,
      "step": 1211
    },
    {
      "epoch": 0.8191956742142615,
      "grad_norm": 18.823190689086914,
      "learning_rate": 2e-05,
      "loss": 0.1998,
      "step": 1212
    },
    {
      "epoch": 0.8198715782358905,
      "grad_norm": 18.91478157043457,
      "learning_rate": 2e-05,
      "loss": 0.2409,
      "step": 1213
    },
    {
      "epoch": 0.8205474822575194,
      "grad_norm": 8.507401466369629,
      "learning_rate": 2e-05,
      "loss": 0.1979,
      "step": 1214
    },
    {
      "epoch": 0.8212233862791484,
      "grad_norm": 12.792491912841797,
      "learning_rate": 2e-05,
      "loss": 0.2783,
      "step": 1215
    },
    {
      "epoch": 0.8218992903007772,
      "grad_norm": 8.762041091918945,
      "learning_rate": 2e-05,
      "loss": 0.2081,
      "step": 1216
    },
    {
      "epoch": 0.8225751943224062,
      "grad_norm": 23.453824996948242,
      "learning_rate": 2e-05,
      "loss": 0.2706,
      "step": 1217
    },
    {
      "epoch": 0.8232510983440351,
      "grad_norm": 10.428817749023438,
      "learning_rate": 2e-05,
      "loss": 0.2456,
      "step": 1218
    },
    {
      "epoch": 0.8239270023656641,
      "grad_norm": 9.321738243103027,
      "learning_rate": 2e-05,
      "loss": 0.2318,
      "step": 1219
    },
    {
      "epoch": 0.824602906387293,
      "grad_norm": 6.270871162414551,
      "learning_rate": 2e-05,
      "loss": 0.2064,
      "step": 1220
    },
    {
      "epoch": 0.8252788104089219,
      "grad_norm": 5.2087483406066895,
      "learning_rate": 2e-05,
      "loss": 0.1774,
      "step": 1221
    },
    {
      "epoch": 0.8259547144305509,
      "grad_norm": 13.255462646484375,
      "learning_rate": 2e-05,
      "loss": 0.2065,
      "step": 1222
    },
    {
      "epoch": 0.8266306184521798,
      "grad_norm": 16.872066497802734,
      "learning_rate": 2e-05,
      "loss": 0.223,
      "step": 1223
    },
    {
      "epoch": 0.8273065224738088,
      "grad_norm": 4.487804412841797,
      "learning_rate": 2e-05,
      "loss": 0.1463,
      "step": 1224
    },
    {
      "epoch": 0.8279824264954376,
      "grad_norm": 10.489372253417969,
      "learning_rate": 2e-05,
      "loss": 0.201,
      "step": 1225
    },
    {
      "epoch": 0.8286583305170666,
      "grad_norm": 11.024821281433105,
      "learning_rate": 2e-05,
      "loss": 0.1977,
      "step": 1226
    },
    {
      "epoch": 0.8293342345386955,
      "grad_norm": 20.890716552734375,
      "learning_rate": 2e-05,
      "loss": 0.2188,
      "step": 1227
    },
    {
      "epoch": 0.8300101385603245,
      "grad_norm": 8.517678260803223,
      "learning_rate": 2e-05,
      "loss": 0.1733,
      "step": 1228
    },
    {
      "epoch": 0.8306860425819533,
      "grad_norm": 6.923564434051514,
      "learning_rate": 2e-05,
      "loss": 0.1868,
      "step": 1229
    },
    {
      "epoch": 0.8313619466035823,
      "grad_norm": 6.372324466705322,
      "learning_rate": 2e-05,
      "loss": 0.1877,
      "step": 1230
    },
    {
      "epoch": 0.8320378506252112,
      "grad_norm": 11.669061660766602,
      "learning_rate": 2e-05,
      "loss": 0.2394,
      "step": 1231
    },
    {
      "epoch": 0.8327137546468402,
      "grad_norm": 9.385963439941406,
      "learning_rate": 2e-05,
      "loss": 0.2705,
      "step": 1232
    },
    {
      "epoch": 0.833389658668469,
      "grad_norm": 9.27115249633789,
      "learning_rate": 2e-05,
      "loss": 0.209,
      "step": 1233
    },
    {
      "epoch": 0.834065562690098,
      "grad_norm": 7.9884934425354,
      "learning_rate": 2e-05,
      "loss": 0.2414,
      "step": 1234
    },
    {
      "epoch": 0.8347414667117269,
      "grad_norm": 5.524908542633057,
      "learning_rate": 2e-05,
      "loss": 0.1687,
      "step": 1235
    },
    {
      "epoch": 0.8354173707333559,
      "grad_norm": 8.831918716430664,
      "learning_rate": 2e-05,
      "loss": 0.1889,
      "step": 1236
    },
    {
      "epoch": 0.8360932747549847,
      "grad_norm": 5.926281452178955,
      "learning_rate": 2e-05,
      "loss": 0.2121,
      "step": 1237
    },
    {
      "epoch": 0.8367691787766137,
      "grad_norm": 4.643338680267334,
      "learning_rate": 2e-05,
      "loss": 0.1495,
      "step": 1238
    },
    {
      "epoch": 0.8374450827982427,
      "grad_norm": 3.7261962890625,
      "learning_rate": 2e-05,
      "loss": 0.144,
      "step": 1239
    },
    {
      "epoch": 0.8381209868198716,
      "grad_norm": 9.820416450500488,
      "learning_rate": 2e-05,
      "loss": 0.2534,
      "step": 1240
    },
    {
      "epoch": 0.8387968908415006,
      "grad_norm": 11.715143203735352,
      "learning_rate": 2e-05,
      "loss": 0.2019,
      "step": 1241
    },
    {
      "epoch": 0.8394727948631294,
      "grad_norm": 7.56997013092041,
      "learning_rate": 2e-05,
      "loss": 0.1969,
      "step": 1242
    },
    {
      "epoch": 0.8401486988847584,
      "grad_norm": 5.484521865844727,
      "learning_rate": 2e-05,
      "loss": 0.2046,
      "step": 1243
    },
    {
      "epoch": 0.8408246029063873,
      "grad_norm": 7.299929618835449,
      "learning_rate": 2e-05,
      "loss": 0.2113,
      "step": 1244
    },
    {
      "epoch": 0.8415005069280163,
      "grad_norm": 10.518689155578613,
      "learning_rate": 2e-05,
      "loss": 0.2072,
      "step": 1245
    },
    {
      "epoch": 0.8421764109496451,
      "grad_norm": 7.255989074707031,
      "learning_rate": 2e-05,
      "loss": 0.2133,
      "step": 1246
    },
    {
      "epoch": 0.8428523149712741,
      "grad_norm": 9.57217025756836,
      "learning_rate": 2e-05,
      "loss": 0.2097,
      "step": 1247
    },
    {
      "epoch": 0.843528218992903,
      "grad_norm": 9.486115455627441,
      "learning_rate": 2e-05,
      "loss": 0.1786,
      "step": 1248
    },
    {
      "epoch": 0.844204123014532,
      "grad_norm": 6.4218244552612305,
      "learning_rate": 2e-05,
      "loss": 0.1817,
      "step": 1249
    },
    {
      "epoch": 0.8448800270361608,
      "grad_norm": 11.722517967224121,
      "learning_rate": 2e-05,
      "loss": 0.2211,
      "step": 1250
    },
    {
      "epoch": 0.8448800270361608,
      "eval_accuracy": 0.5355574324324325,
      "eval_loss": 0.24502938985824585,
      "eval_runtime": 42.7562,
      "eval_samples_per_second": 553.838,
      "eval_steps_per_second": 17.307,
      "step": 1250
    },
    {
      "epoch": 0.8455559310577898,
      "grad_norm": 6.453871726989746,
      "learning_rate": 2e-05,
      "loss": 0.2208,
      "step": 1251
    },
    {
      "epoch": 0.8462318350794187,
      "grad_norm": 11.477692604064941,
      "learning_rate": 2e-05,
      "loss": 0.2028,
      "step": 1252
    },
    {
      "epoch": 0.8469077391010477,
      "grad_norm": 14.508378028869629,
      "learning_rate": 2e-05,
      "loss": 0.2406,
      "step": 1253
    },
    {
      "epoch": 0.8475836431226765,
      "grad_norm": 18.37901496887207,
      "learning_rate": 2e-05,
      "loss": 0.2386,
      "step": 1254
    },
    {
      "epoch": 0.8482595471443055,
      "grad_norm": 4.863740921020508,
      "learning_rate": 2e-05,
      "loss": 0.1887,
      "step": 1255
    },
    {
      "epoch": 0.8489354511659344,
      "grad_norm": 10.315373420715332,
      "learning_rate": 2e-05,
      "loss": 0.2124,
      "step": 1256
    },
    {
      "epoch": 0.8496113551875634,
      "grad_norm": 18.155216217041016,
      "learning_rate": 2e-05,
      "loss": 0.1996,
      "step": 1257
    },
    {
      "epoch": 0.8502872592091923,
      "grad_norm": 8.713403701782227,
      "learning_rate": 2e-05,
      "loss": 0.2024,
      "step": 1258
    },
    {
      "epoch": 0.8509631632308212,
      "grad_norm": 4.6441826820373535,
      "learning_rate": 2e-05,
      "loss": 0.1962,
      "step": 1259
    },
    {
      "epoch": 0.8516390672524502,
      "grad_norm": 11.462042808532715,
      "learning_rate": 2e-05,
      "loss": 0.1695,
      "step": 1260
    },
    {
      "epoch": 0.8523149712740791,
      "grad_norm": 9.312410354614258,
      "learning_rate": 2e-05,
      "loss": 0.2146,
      "step": 1261
    },
    {
      "epoch": 0.852990875295708,
      "grad_norm": 5.628330707550049,
      "learning_rate": 2e-05,
      "loss": 0.2231,
      "step": 1262
    },
    {
      "epoch": 0.8536667793173369,
      "grad_norm": 10.98554801940918,
      "learning_rate": 2e-05,
      "loss": 0.1799,
      "step": 1263
    },
    {
      "epoch": 0.8543426833389659,
      "grad_norm": 6.092944145202637,
      "learning_rate": 2e-05,
      "loss": 0.1996,
      "step": 1264
    },
    {
      "epoch": 0.8550185873605948,
      "grad_norm": 9.386287689208984,
      "learning_rate": 2e-05,
      "loss": 0.164,
      "step": 1265
    },
    {
      "epoch": 0.8556944913822238,
      "grad_norm": 7.082285404205322,
      "learning_rate": 2e-05,
      "loss": 0.2265,
      "step": 1266
    },
    {
      "epoch": 0.8563703954038526,
      "grad_norm": 21.08319091796875,
      "learning_rate": 2e-05,
      "loss": 0.3169,
      "step": 1267
    },
    {
      "epoch": 0.8570462994254816,
      "grad_norm": 7.102109909057617,
      "learning_rate": 2e-05,
      "loss": 0.1856,
      "step": 1268
    },
    {
      "epoch": 0.8577222034471105,
      "grad_norm": 6.747745990753174,
      "learning_rate": 2e-05,
      "loss": 0.2658,
      "step": 1269
    },
    {
      "epoch": 0.8583981074687395,
      "grad_norm": 9.807302474975586,
      "learning_rate": 2e-05,
      "loss": 0.1822,
      "step": 1270
    },
    {
      "epoch": 0.8590740114903683,
      "grad_norm": 5.614322185516357,
      "learning_rate": 2e-05,
      "loss": 0.1696,
      "step": 1271
    },
    {
      "epoch": 0.8597499155119973,
      "grad_norm": 8.99000358581543,
      "learning_rate": 2e-05,
      "loss": 0.1817,
      "step": 1272
    },
    {
      "epoch": 0.8604258195336262,
      "grad_norm": 5.519763469696045,
      "learning_rate": 2e-05,
      "loss": 0.1996,
      "step": 1273
    },
    {
      "epoch": 0.8611017235552552,
      "grad_norm": 19.828474044799805,
      "learning_rate": 2e-05,
      "loss": 0.2093,
      "step": 1274
    },
    {
      "epoch": 0.861777627576884,
      "grad_norm": 8.07367992401123,
      "learning_rate": 2e-05,
      "loss": 0.1779,
      "step": 1275
    },
    {
      "epoch": 0.862453531598513,
      "grad_norm": 7.599564552307129,
      "learning_rate": 2e-05,
      "loss": 0.2078,
      "step": 1276
    },
    {
      "epoch": 0.863129435620142,
      "grad_norm": 13.987101554870605,
      "learning_rate": 2e-05,
      "loss": 0.2187,
      "step": 1277
    },
    {
      "epoch": 0.8638053396417709,
      "grad_norm": 22.092912673950195,
      "learning_rate": 2e-05,
      "loss": 0.2465,
      "step": 1278
    },
    {
      "epoch": 0.8644812436633998,
      "grad_norm": 9.018051147460938,
      "learning_rate": 2e-05,
      "loss": 0.1775,
      "step": 1279
    },
    {
      "epoch": 0.8651571476850287,
      "grad_norm": 9.70766830444336,
      "learning_rate": 2e-05,
      "loss": 0.2572,
      "step": 1280
    },
    {
      "epoch": 0.8658330517066577,
      "grad_norm": 8.481575965881348,
      "learning_rate": 2e-05,
      "loss": 0.2094,
      "step": 1281
    },
    {
      "epoch": 0.8665089557282866,
      "grad_norm": 11.429023742675781,
      "learning_rate": 2e-05,
      "loss": 0.1374,
      "step": 1282
    },
    {
      "epoch": 0.8671848597499155,
      "grad_norm": 13.500280380249023,
      "learning_rate": 2e-05,
      "loss": 0.1678,
      "step": 1283
    },
    {
      "epoch": 0.8678607637715444,
      "grad_norm": 12.098771095275879,
      "learning_rate": 2e-05,
      "loss": 0.1837,
      "step": 1284
    },
    {
      "epoch": 0.8685366677931734,
      "grad_norm": 4.430557727813721,
      "learning_rate": 2e-05,
      "loss": 0.1518,
      "step": 1285
    },
    {
      "epoch": 0.8692125718148023,
      "grad_norm": 29.035429000854492,
      "learning_rate": 2e-05,
      "loss": 0.3285,
      "step": 1286
    },
    {
      "epoch": 0.8698884758364313,
      "grad_norm": 14.897074699401855,
      "learning_rate": 2e-05,
      "loss": 0.1969,
      "step": 1287
    },
    {
      "epoch": 0.8705643798580601,
      "grad_norm": 5.606655120849609,
      "learning_rate": 2e-05,
      "loss": 0.19,
      "step": 1288
    },
    {
      "epoch": 0.8712402838796891,
      "grad_norm": 7.774875640869141,
      "learning_rate": 2e-05,
      "loss": 0.1896,
      "step": 1289
    },
    {
      "epoch": 0.871916187901318,
      "grad_norm": 10.693351745605469,
      "learning_rate": 2e-05,
      "loss": 0.1659,
      "step": 1290
    },
    {
      "epoch": 0.872592091922947,
      "grad_norm": 5.655900001525879,
      "learning_rate": 2e-05,
      "loss": 0.2405,
      "step": 1291
    },
    {
      "epoch": 0.8732679959445758,
      "grad_norm": 7.432564735412598,
      "learning_rate": 2e-05,
      "loss": 0.2322,
      "step": 1292
    },
    {
      "epoch": 0.8739438999662048,
      "grad_norm": 8.859232902526855,
      "learning_rate": 2e-05,
      "loss": 0.206,
      "step": 1293
    },
    {
      "epoch": 0.8746198039878337,
      "grad_norm": 9.045379638671875,
      "learning_rate": 2e-05,
      "loss": 0.2354,
      "step": 1294
    },
    {
      "epoch": 0.8752957080094627,
      "grad_norm": 11.812585830688477,
      "learning_rate": 2e-05,
      "loss": 0.1875,
      "step": 1295
    },
    {
      "epoch": 0.8759716120310916,
      "grad_norm": 4.715848922729492,
      "learning_rate": 2e-05,
      "loss": 0.1959,
      "step": 1296
    },
    {
      "epoch": 0.8766475160527205,
      "grad_norm": 5.5169997215271,
      "learning_rate": 2e-05,
      "loss": 0.1553,
      "step": 1297
    },
    {
      "epoch": 0.8773234200743495,
      "grad_norm": 5.727180480957031,
      "learning_rate": 2e-05,
      "loss": 0.1485,
      "step": 1298
    },
    {
      "epoch": 0.8779993240959784,
      "grad_norm": 16.564773559570312,
      "learning_rate": 2e-05,
      "loss": 0.1964,
      "step": 1299
    },
    {
      "epoch": 0.8786752281176073,
      "grad_norm": 15.917607307434082,
      "learning_rate": 2e-05,
      "loss": 0.2471,
      "step": 1300
    },
    {
      "epoch": 0.8786752281176073,
      "eval_accuracy": 0.5214949324324324,
      "eval_loss": 0.24719096720218658,
      "eval_runtime": 42.743,
      "eval_samples_per_second": 554.009,
      "eval_steps_per_second": 17.313,
      "step": 1300
    },
    {
      "epoch": 0.8793511321392362,
      "grad_norm": 7.029534339904785,
      "learning_rate": 2e-05,
      "loss": 0.194,
      "step": 1301
    },
    {
      "epoch": 0.8800270361608652,
      "grad_norm": 9.881065368652344,
      "learning_rate": 2e-05,
      "loss": 0.1831,
      "step": 1302
    },
    {
      "epoch": 0.8807029401824941,
      "grad_norm": 14.316802024841309,
      "learning_rate": 2e-05,
      "loss": 0.1906,
      "step": 1303
    },
    {
      "epoch": 0.881378844204123,
      "grad_norm": 13.665921211242676,
      "learning_rate": 2e-05,
      "loss": 0.1589,
      "step": 1304
    },
    {
      "epoch": 0.8820547482257519,
      "grad_norm": 9.587456703186035,
      "learning_rate": 2e-05,
      "loss": 0.2397,
      "step": 1305
    },
    {
      "epoch": 0.8827306522473809,
      "grad_norm": 6.790398120880127,
      "learning_rate": 2e-05,
      "loss": 0.1946,
      "step": 1306
    },
    {
      "epoch": 0.8834065562690098,
      "grad_norm": 9.336783409118652,
      "learning_rate": 2e-05,
      "loss": 0.2724,
      "step": 1307
    },
    {
      "epoch": 0.8840824602906387,
      "grad_norm": 8.844710350036621,
      "learning_rate": 2e-05,
      "loss": 0.173,
      "step": 1308
    },
    {
      "epoch": 0.8847583643122676,
      "grad_norm": 9.389384269714355,
      "learning_rate": 2e-05,
      "loss": 0.2489,
      "step": 1309
    },
    {
      "epoch": 0.8854342683338966,
      "grad_norm": 6.336228370666504,
      "learning_rate": 2e-05,
      "loss": 0.197,
      "step": 1310
    },
    {
      "epoch": 0.8861101723555255,
      "grad_norm": 6.35755729675293,
      "learning_rate": 2e-05,
      "loss": 0.1931,
      "step": 1311
    },
    {
      "epoch": 0.8867860763771545,
      "grad_norm": 7.834129333496094,
      "learning_rate": 2e-05,
      "loss": 0.1719,
      "step": 1312
    },
    {
      "epoch": 0.8874619803987833,
      "grad_norm": 6.420130729675293,
      "learning_rate": 2e-05,
      "loss": 0.1947,
      "step": 1313
    },
    {
      "epoch": 0.8881378844204123,
      "grad_norm": 9.024942398071289,
      "learning_rate": 2e-05,
      "loss": 0.1792,
      "step": 1314
    },
    {
      "epoch": 0.8888137884420412,
      "grad_norm": 7.959498405456543,
      "learning_rate": 2e-05,
      "loss": 0.2241,
      "step": 1315
    },
    {
      "epoch": 0.8894896924636702,
      "grad_norm": 4.692646503448486,
      "learning_rate": 2e-05,
      "loss": 0.1986,
      "step": 1316
    },
    {
      "epoch": 0.8901655964852991,
      "grad_norm": 10.953783988952637,
      "learning_rate": 2e-05,
      "loss": 0.2552,
      "step": 1317
    },
    {
      "epoch": 0.890841500506928,
      "grad_norm": 6.532024383544922,
      "learning_rate": 2e-05,
      "loss": 0.2412,
      "step": 1318
    },
    {
      "epoch": 0.891517404528557,
      "grad_norm": 7.065764427185059,
      "learning_rate": 2e-05,
      "loss": 0.207,
      "step": 1319
    },
    {
      "epoch": 0.8921933085501859,
      "grad_norm": 6.261386871337891,
      "learning_rate": 2e-05,
      "loss": 0.2192,
      "step": 1320
    },
    {
      "epoch": 0.8928692125718148,
      "grad_norm": 6.068378448486328,
      "learning_rate": 2e-05,
      "loss": 0.168,
      "step": 1321
    },
    {
      "epoch": 0.8935451165934437,
      "grad_norm": 14.019453048706055,
      "learning_rate": 2e-05,
      "loss": 0.2269,
      "step": 1322
    },
    {
      "epoch": 0.8942210206150727,
      "grad_norm": 8.851364135742188,
      "learning_rate": 2e-05,
      "loss": 0.2188,
      "step": 1323
    },
    {
      "epoch": 0.8948969246367016,
      "grad_norm": 5.782850742340088,
      "learning_rate": 2e-05,
      "loss": 0.1986,
      "step": 1324
    },
    {
      "epoch": 0.8955728286583305,
      "grad_norm": 9.825984001159668,
      "learning_rate": 2e-05,
      "loss": 0.2197,
      "step": 1325
    },
    {
      "epoch": 0.8962487326799594,
      "grad_norm": 5.604493618011475,
      "learning_rate": 2e-05,
      "loss": 0.2279,
      "step": 1326
    },
    {
      "epoch": 0.8969246367015884,
      "grad_norm": 8.774330139160156,
      "learning_rate": 2e-05,
      "loss": 0.1866,
      "step": 1327
    },
    {
      "epoch": 0.8976005407232173,
      "grad_norm": 11.542707443237305,
      "learning_rate": 2e-05,
      "loss": 0.1597,
      "step": 1328
    },
    {
      "epoch": 0.8982764447448462,
      "grad_norm": 9.064879417419434,
      "learning_rate": 2e-05,
      "loss": 0.2575,
      "step": 1329
    },
    {
      "epoch": 0.8989523487664751,
      "grad_norm": 4.677642345428467,
      "learning_rate": 2e-05,
      "loss": 0.1615,
      "step": 1330
    },
    {
      "epoch": 0.8996282527881041,
      "grad_norm": 9.845659255981445,
      "learning_rate": 2e-05,
      "loss": 0.2364,
      "step": 1331
    },
    {
      "epoch": 0.900304156809733,
      "grad_norm": 4.682253360748291,
      "learning_rate": 2e-05,
      "loss": 0.1837,
      "step": 1332
    },
    {
      "epoch": 0.900980060831362,
      "grad_norm": 6.681576728820801,
      "learning_rate": 2e-05,
      "loss": 0.1501,
      "step": 1333
    },
    {
      "epoch": 0.9016559648529908,
      "grad_norm": 14.403308868408203,
      "learning_rate": 2e-05,
      "loss": 0.2131,
      "step": 1334
    },
    {
      "epoch": 0.9023318688746198,
      "grad_norm": 13.329488754272461,
      "learning_rate": 2e-05,
      "loss": 0.2144,
      "step": 1335
    },
    {
      "epoch": 0.9030077728962488,
      "grad_norm": 4.763746738433838,
      "learning_rate": 2e-05,
      "loss": 0.1488,
      "step": 1336
    },
    {
      "epoch": 0.9036836769178777,
      "grad_norm": 8.368087768554688,
      "learning_rate": 2e-05,
      "loss": 0.2405,
      "step": 1337
    },
    {
      "epoch": 0.9043595809395066,
      "grad_norm": 11.699691772460938,
      "learning_rate": 2e-05,
      "loss": 0.2311,
      "step": 1338
    },
    {
      "epoch": 0.9050354849611355,
      "grad_norm": 9.941756248474121,
      "learning_rate": 2e-05,
      "loss": 0.1863,
      "step": 1339
    },
    {
      "epoch": 0.9057113889827645,
      "grad_norm": 13.111069679260254,
      "learning_rate": 2e-05,
      "loss": 0.2243,
      "step": 1340
    },
    {
      "epoch": 0.9063872930043934,
      "grad_norm": 5.702256202697754,
      "learning_rate": 2e-05,
      "loss": 0.164,
      "step": 1341
    },
    {
      "epoch": 0.9070631970260223,
      "grad_norm": 9.156183242797852,
      "learning_rate": 2e-05,
      "loss": 0.2295,
      "step": 1342
    },
    {
      "epoch": 0.9077391010476512,
      "grad_norm": 9.478164672851562,
      "learning_rate": 2e-05,
      "loss": 0.2006,
      "step": 1343
    },
    {
      "epoch": 0.9084150050692802,
      "grad_norm": 4.6168670654296875,
      "learning_rate": 2e-05,
      "loss": 0.167,
      "step": 1344
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 10.580148696899414,
      "learning_rate": 2e-05,
      "loss": 0.241,
      "step": 1345
    },
    {
      "epoch": 0.909766813112538,
      "grad_norm": 15.528704643249512,
      "learning_rate": 2e-05,
      "loss": 0.2605,
      "step": 1346
    },
    {
      "epoch": 0.9104427171341669,
      "grad_norm": 5.934053421020508,
      "learning_rate": 2e-05,
      "loss": 0.2334,
      "step": 1347
    },
    {
      "epoch": 0.9111186211557959,
      "grad_norm": 5.953293323516846,
      "learning_rate": 2e-05,
      "loss": 0.1881,
      "step": 1348
    },
    {
      "epoch": 0.9117945251774248,
      "grad_norm": 6.431759357452393,
      "learning_rate": 2e-05,
      "loss": 0.1904,
      "step": 1349
    },
    {
      "epoch": 0.9124704291990537,
      "grad_norm": 10.722702980041504,
      "learning_rate": 2e-05,
      "loss": 0.2205,
      "step": 1350
    },
    {
      "epoch": 0.9124704291990537,
      "eval_accuracy": 0.5083614864864865,
      "eval_loss": 0.2435036599636078,
      "eval_runtime": 43.2426,
      "eval_samples_per_second": 547.609,
      "eval_steps_per_second": 17.113,
      "step": 1350
    },
    {
      "epoch": 0.9131463332206826,
      "grad_norm": 23.397581100463867,
      "learning_rate": 2e-05,
      "loss": 0.2663,
      "step": 1351
    },
    {
      "epoch": 0.9138222372423116,
      "grad_norm": 7.792226791381836,
      "learning_rate": 2e-05,
      "loss": 0.2133,
      "step": 1352
    },
    {
      "epoch": 0.9144981412639405,
      "grad_norm": 13.499917030334473,
      "learning_rate": 2e-05,
      "loss": 0.2734,
      "step": 1353
    },
    {
      "epoch": 0.9151740452855694,
      "grad_norm": 7.270654201507568,
      "learning_rate": 2e-05,
      "loss": 0.1532,
      "step": 1354
    },
    {
      "epoch": 0.9158499493071984,
      "grad_norm": 12.761370658874512,
      "learning_rate": 2e-05,
      "loss": 0.2373,
      "step": 1355
    },
    {
      "epoch": 0.9165258533288273,
      "grad_norm": 6.501335144042969,
      "learning_rate": 2e-05,
      "loss": 0.2151,
      "step": 1356
    },
    {
      "epoch": 0.9172017573504563,
      "grad_norm": 6.510569095611572,
      "learning_rate": 2e-05,
      "loss": 0.1817,
      "step": 1357
    },
    {
      "epoch": 0.9178776613720852,
      "grad_norm": 4.450448513031006,
      "learning_rate": 2e-05,
      "loss": 0.2133,
      "step": 1358
    },
    {
      "epoch": 0.9185535653937141,
      "grad_norm": 4.948029041290283,
      "learning_rate": 2e-05,
      "loss": 0.2262,
      "step": 1359
    },
    {
      "epoch": 0.919229469415343,
      "grad_norm": 5.755936145782471,
      "learning_rate": 2e-05,
      "loss": 0.2301,
      "step": 1360
    },
    {
      "epoch": 0.919905373436972,
      "grad_norm": 14.457347869873047,
      "learning_rate": 2e-05,
      "loss": 0.1795,
      "step": 1361
    },
    {
      "epoch": 0.9205812774586009,
      "grad_norm": 5.785470485687256,
      "learning_rate": 2e-05,
      "loss": 0.2091,
      "step": 1362
    },
    {
      "epoch": 0.9212571814802298,
      "grad_norm": 5.467822551727295,
      "learning_rate": 2e-05,
      "loss": 0.203,
      "step": 1363
    },
    {
      "epoch": 0.9219330855018587,
      "grad_norm": 11.902308464050293,
      "learning_rate": 2e-05,
      "loss": 0.2388,
      "step": 1364
    },
    {
      "epoch": 0.9226089895234877,
      "grad_norm": 14.441679954528809,
      "learning_rate": 2e-05,
      "loss": 0.2357,
      "step": 1365
    },
    {
      "epoch": 0.9232848935451166,
      "grad_norm": 7.901401996612549,
      "learning_rate": 2e-05,
      "loss": 0.2014,
      "step": 1366
    },
    {
      "epoch": 0.9239607975667455,
      "grad_norm": 14.741616249084473,
      "learning_rate": 2e-05,
      "loss": 0.2004,
      "step": 1367
    },
    {
      "epoch": 0.9246367015883744,
      "grad_norm": 7.863753318786621,
      "learning_rate": 2e-05,
      "loss": 0.222,
      "step": 1368
    },
    {
      "epoch": 0.9253126056100034,
      "grad_norm": 6.023305892944336,
      "learning_rate": 2e-05,
      "loss": 0.2083,
      "step": 1369
    },
    {
      "epoch": 0.9259885096316323,
      "grad_norm": 8.30286693572998,
      "learning_rate": 2e-05,
      "loss": 0.2316,
      "step": 1370
    },
    {
      "epoch": 0.9266644136532612,
      "grad_norm": 11.35500431060791,
      "learning_rate": 2e-05,
      "loss": 0.1933,
      "step": 1371
    },
    {
      "epoch": 0.9273403176748901,
      "grad_norm": 22.3579158782959,
      "learning_rate": 2e-05,
      "loss": 0.2475,
      "step": 1372
    },
    {
      "epoch": 0.9280162216965191,
      "grad_norm": 6.859817028045654,
      "learning_rate": 2e-05,
      "loss": 0.1713,
      "step": 1373
    },
    {
      "epoch": 0.9286921257181481,
      "grad_norm": 5.11651611328125,
      "learning_rate": 2e-05,
      "loss": 0.1685,
      "step": 1374
    },
    {
      "epoch": 0.929368029739777,
      "grad_norm": 12.812508583068848,
      "learning_rate": 2e-05,
      "loss": 0.2355,
      "step": 1375
    },
    {
      "epoch": 0.9300439337614059,
      "grad_norm": 8.175957679748535,
      "learning_rate": 2e-05,
      "loss": 0.2182,
      "step": 1376
    },
    {
      "epoch": 0.9307198377830348,
      "grad_norm": 5.79949426651001,
      "learning_rate": 2e-05,
      "loss": 0.1386,
      "step": 1377
    },
    {
      "epoch": 0.9313957418046638,
      "grad_norm": 9.280027389526367,
      "learning_rate": 2e-05,
      "loss": 0.2123,
      "step": 1378
    },
    {
      "epoch": 0.9320716458262926,
      "grad_norm": 6.391855239868164,
      "learning_rate": 2e-05,
      "loss": 0.24,
      "step": 1379
    },
    {
      "epoch": 0.9327475498479216,
      "grad_norm": 5.140876770019531,
      "learning_rate": 2e-05,
      "loss": 0.237,
      "step": 1380
    },
    {
      "epoch": 0.9334234538695505,
      "grad_norm": 8.382298469543457,
      "learning_rate": 2e-05,
      "loss": 0.1459,
      "step": 1381
    },
    {
      "epoch": 0.9340993578911795,
      "grad_norm": 13.176122665405273,
      "learning_rate": 2e-05,
      "loss": 0.2264,
      "step": 1382
    },
    {
      "epoch": 0.9347752619128084,
      "grad_norm": 17.337602615356445,
      "learning_rate": 2e-05,
      "loss": 0.2523,
      "step": 1383
    },
    {
      "epoch": 0.9354511659344373,
      "grad_norm": 7.463795185089111,
      "learning_rate": 2e-05,
      "loss": 0.2319,
      "step": 1384
    },
    {
      "epoch": 0.9361270699560662,
      "grad_norm": 9.58946704864502,
      "learning_rate": 2e-05,
      "loss": 0.2407,
      "step": 1385
    },
    {
      "epoch": 0.9368029739776952,
      "grad_norm": 12.550409317016602,
      "learning_rate": 2e-05,
      "loss": 0.1849,
      "step": 1386
    },
    {
      "epoch": 0.9374788779993241,
      "grad_norm": 7.615727424621582,
      "learning_rate": 2e-05,
      "loss": 0.1934,
      "step": 1387
    },
    {
      "epoch": 0.938154782020953,
      "grad_norm": 10.833722114562988,
      "learning_rate": 2e-05,
      "loss": 0.2312,
      "step": 1388
    },
    {
      "epoch": 0.9388306860425819,
      "grad_norm": 9.494311332702637,
      "learning_rate": 2e-05,
      "loss": 0.1828,
      "step": 1389
    },
    {
      "epoch": 0.9395065900642109,
      "grad_norm": 4.821182727813721,
      "learning_rate": 2e-05,
      "loss": 0.1772,
      "step": 1390
    },
    {
      "epoch": 0.9401824940858398,
      "grad_norm": 6.338797092437744,
      "learning_rate": 2e-05,
      "loss": 0.1926,
      "step": 1391
    },
    {
      "epoch": 0.9408583981074687,
      "grad_norm": 17.986995697021484,
      "learning_rate": 2e-05,
      "loss": 0.2424,
      "step": 1392
    },
    {
      "epoch": 0.9415343021290977,
      "grad_norm": 13.755885124206543,
      "learning_rate": 2e-05,
      "loss": 0.1638,
      "step": 1393
    },
    {
      "epoch": 0.9422102061507266,
      "grad_norm": 8.161349296569824,
      "learning_rate": 2e-05,
      "loss": 0.1847,
      "step": 1394
    },
    {
      "epoch": 0.9428861101723556,
      "grad_norm": 6.8715128898620605,
      "learning_rate": 2e-05,
      "loss": 0.2622,
      "step": 1395
    },
    {
      "epoch": 0.9435620141939844,
      "grad_norm": 7.77981424331665,
      "learning_rate": 2e-05,
      "loss": 0.2185,
      "step": 1396
    },
    {
      "epoch": 0.9442379182156134,
      "grad_norm": 8.505199432373047,
      "learning_rate": 2e-05,
      "loss": 0.213,
      "step": 1397
    },
    {
      "epoch": 0.9449138222372423,
      "grad_norm": 14.53833293914795,
      "learning_rate": 2e-05,
      "loss": 0.2343,
      "step": 1398
    },
    {
      "epoch": 0.9455897262588713,
      "grad_norm": 8.017377853393555,
      "learning_rate": 2e-05,
      "loss": 0.198,
      "step": 1399
    },
    {
      "epoch": 0.9462656302805001,
      "grad_norm": 7.532063961029053,
      "learning_rate": 2e-05,
      "loss": 0.1879,
      "step": 1400
    },
    {
      "epoch": 0.9462656302805001,
      "eval_accuracy": 0.5655405405405406,
      "eval_loss": 0.24145914614200592,
      "eval_runtime": 43.2502,
      "eval_samples_per_second": 547.512,
      "eval_steps_per_second": 17.11,
      "step": 1400
    },
    {
      "epoch": 0.9469415343021291,
      "grad_norm": 11.871845245361328,
      "learning_rate": 2e-05,
      "loss": 0.2489,
      "step": 1401
    },
    {
      "epoch": 0.947617438323758,
      "grad_norm": 3.954641342163086,
      "learning_rate": 2e-05,
      "loss": 0.1555,
      "step": 1402
    },
    {
      "epoch": 0.948293342345387,
      "grad_norm": 12.058279037475586,
      "learning_rate": 2e-05,
      "loss": 0.2205,
      "step": 1403
    },
    {
      "epoch": 0.9489692463670159,
      "grad_norm": 10.926996231079102,
      "learning_rate": 2e-05,
      "loss": 0.2185,
      "step": 1404
    },
    {
      "epoch": 0.9496451503886448,
      "grad_norm": 17.732980728149414,
      "learning_rate": 2e-05,
      "loss": 0.2666,
      "step": 1405
    },
    {
      "epoch": 0.9503210544102737,
      "grad_norm": 8.76140308380127,
      "learning_rate": 2e-05,
      "loss": 0.1671,
      "step": 1406
    },
    {
      "epoch": 0.9509969584319027,
      "grad_norm": 10.106435775756836,
      "learning_rate": 2e-05,
      "loss": 0.1693,
      "step": 1407
    },
    {
      "epoch": 0.9516728624535316,
      "grad_norm": 7.838395118713379,
      "learning_rate": 2e-05,
      "loss": 0.2011,
      "step": 1408
    },
    {
      "epoch": 0.9523487664751605,
      "grad_norm": 13.766228675842285,
      "learning_rate": 2e-05,
      "loss": 0.254,
      "step": 1409
    },
    {
      "epoch": 0.9530246704967894,
      "grad_norm": 7.80083703994751,
      "learning_rate": 2e-05,
      "loss": 0.1959,
      "step": 1410
    },
    {
      "epoch": 0.9537005745184184,
      "grad_norm": 11.639693260192871,
      "learning_rate": 2e-05,
      "loss": 0.1756,
      "step": 1411
    },
    {
      "epoch": 0.9543764785400474,
      "grad_norm": 12.617870330810547,
      "learning_rate": 2e-05,
      "loss": 0.1922,
      "step": 1412
    },
    {
      "epoch": 0.9550523825616762,
      "grad_norm": 9.167438507080078,
      "learning_rate": 2e-05,
      "loss": 0.1608,
      "step": 1413
    },
    {
      "epoch": 0.9557282865833052,
      "grad_norm": 7.673298358917236,
      "learning_rate": 2e-05,
      "loss": 0.2614,
      "step": 1414
    },
    {
      "epoch": 0.9564041906049341,
      "grad_norm": 8.636027336120605,
      "learning_rate": 2e-05,
      "loss": 0.2004,
      "step": 1415
    },
    {
      "epoch": 0.9570800946265631,
      "grad_norm": 13.095100402832031,
      "learning_rate": 2e-05,
      "loss": 0.12,
      "step": 1416
    },
    {
      "epoch": 0.9577559986481919,
      "grad_norm": 6.782408714294434,
      "learning_rate": 2e-05,
      "loss": 0.1979,
      "step": 1417
    },
    {
      "epoch": 0.9584319026698209,
      "grad_norm": 11.162742614746094,
      "learning_rate": 2e-05,
      "loss": 0.2597,
      "step": 1418
    },
    {
      "epoch": 0.9591078066914498,
      "grad_norm": 8.159440040588379,
      "learning_rate": 2e-05,
      "loss": 0.2356,
      "step": 1419
    },
    {
      "epoch": 0.9597837107130788,
      "grad_norm": 17.93026351928711,
      "learning_rate": 2e-05,
      "loss": 0.2215,
      "step": 1420
    },
    {
      "epoch": 0.9604596147347076,
      "grad_norm": 9.422277450561523,
      "learning_rate": 2e-05,
      "loss": 0.2167,
      "step": 1421
    },
    {
      "epoch": 0.9611355187563366,
      "grad_norm": 7.1589508056640625,
      "learning_rate": 2e-05,
      "loss": 0.1818,
      "step": 1422
    },
    {
      "epoch": 0.9618114227779655,
      "grad_norm": 15.701791763305664,
      "learning_rate": 2e-05,
      "loss": 0.1833,
      "step": 1423
    },
    {
      "epoch": 0.9624873267995945,
      "grad_norm": 4.5370988845825195,
      "learning_rate": 2e-05,
      "loss": 0.1973,
      "step": 1424
    },
    {
      "epoch": 0.9631632308212233,
      "grad_norm": 11.787425994873047,
      "learning_rate": 2e-05,
      "loss": 0.1965,
      "step": 1425
    },
    {
      "epoch": 0.9638391348428523,
      "grad_norm": 4.357491493225098,
      "learning_rate": 2e-05,
      "loss": 0.1528,
      "step": 1426
    },
    {
      "epoch": 0.9645150388644812,
      "grad_norm": 19.76885414123535,
      "learning_rate": 2e-05,
      "loss": 0.2727,
      "step": 1427
    },
    {
      "epoch": 0.9651909428861102,
      "grad_norm": 6.536474227905273,
      "learning_rate": 2e-05,
      "loss": 0.2325,
      "step": 1428
    },
    {
      "epoch": 0.965866846907739,
      "grad_norm": 5.640315055847168,
      "learning_rate": 2e-05,
      "loss": 0.2021,
      "step": 1429
    },
    {
      "epoch": 0.966542750929368,
      "grad_norm": 4.722154140472412,
      "learning_rate": 2e-05,
      "loss": 0.1381,
      "step": 1430
    },
    {
      "epoch": 0.9672186549509969,
      "grad_norm": 5.78228235244751,
      "learning_rate": 2e-05,
      "loss": 0.193,
      "step": 1431
    },
    {
      "epoch": 0.9678945589726259,
      "grad_norm": 5.861937046051025,
      "learning_rate": 2e-05,
      "loss": 0.1989,
      "step": 1432
    },
    {
      "epoch": 0.9685704629942549,
      "grad_norm": 15.43178653717041,
      "learning_rate": 2e-05,
      "loss": 0.2236,
      "step": 1433
    },
    {
      "epoch": 0.9692463670158837,
      "grad_norm": 5.131720066070557,
      "learning_rate": 2e-05,
      "loss": 0.1965,
      "step": 1434
    },
    {
      "epoch": 0.9699222710375127,
      "grad_norm": 7.956301689147949,
      "learning_rate": 2e-05,
      "loss": 0.2489,
      "step": 1435
    },
    {
      "epoch": 0.9705981750591416,
      "grad_norm": 6.4857048988342285,
      "learning_rate": 2e-05,
      "loss": 0.1559,
      "step": 1436
    },
    {
      "epoch": 0.9712740790807706,
      "grad_norm": 9.459001541137695,
      "learning_rate": 2e-05,
      "loss": 0.1956,
      "step": 1437
    },
    {
      "epoch": 0.9719499831023994,
      "grad_norm": 9.905768394470215,
      "learning_rate": 2e-05,
      "loss": 0.2199,
      "step": 1438
    },
    {
      "epoch": 0.9726258871240284,
      "grad_norm": 11.162975311279297,
      "learning_rate": 2e-05,
      "loss": 0.2795,
      "step": 1439
    },
    {
      "epoch": 0.9733017911456573,
      "grad_norm": 8.349414825439453,
      "learning_rate": 2e-05,
      "loss": 0.2416,
      "step": 1440
    },
    {
      "epoch": 0.9739776951672863,
      "grad_norm": 5.573241233825684,
      "learning_rate": 2e-05,
      "loss": 0.2127,
      "step": 1441
    },
    {
      "epoch": 0.9746535991889151,
      "grad_norm": 10.376121520996094,
      "learning_rate": 2e-05,
      "loss": 0.1692,
      "step": 1442
    },
    {
      "epoch": 0.9753295032105441,
      "grad_norm": 11.532674789428711,
      "learning_rate": 2e-05,
      "loss": 0.211,
      "step": 1443
    },
    {
      "epoch": 0.976005407232173,
      "grad_norm": 14.336299896240234,
      "learning_rate": 2e-05,
      "loss": 0.2162,
      "step": 1444
    },
    {
      "epoch": 0.976681311253802,
      "grad_norm": 12.402562141418457,
      "learning_rate": 2e-05,
      "loss": 0.1968,
      "step": 1445
    },
    {
      "epoch": 0.9773572152754308,
      "grad_norm": 5.954377174377441,
      "learning_rate": 2e-05,
      "loss": 0.2145,
      "step": 1446
    },
    {
      "epoch": 0.9780331192970598,
      "grad_norm": 6.661704063415527,
      "learning_rate": 2e-05,
      "loss": 0.1855,
      "step": 1447
    },
    {
      "epoch": 0.9787090233186887,
      "grad_norm": 6.415578842163086,
      "learning_rate": 2e-05,
      "loss": 0.2079,
      "step": 1448
    },
    {
      "epoch": 0.9793849273403177,
      "grad_norm": 20.292686462402344,
      "learning_rate": 2e-05,
      "loss": 0.2563,
      "step": 1449
    },
    {
      "epoch": 0.9800608313619465,
      "grad_norm": 14.375391960144043,
      "learning_rate": 2e-05,
      "loss": 0.1963,
      "step": 1450
    },
    {
      "epoch": 0.9800608313619465,
      "eval_accuracy": 0.5608530405405405,
      "eval_loss": 0.24381186068058014,
      "eval_runtime": 43.1001,
      "eval_samples_per_second": 549.419,
      "eval_steps_per_second": 17.169,
      "step": 1450
    },
    {
      "epoch": 0.9807367353835755,
      "grad_norm": 6.668982028961182,
      "learning_rate": 2e-05,
      "loss": 0.1914,
      "step": 1451
    },
    {
      "epoch": 0.9814126394052045,
      "grad_norm": 8.152243614196777,
      "learning_rate": 2e-05,
      "loss": 0.2256,
      "step": 1452
    },
    {
      "epoch": 0.9820885434268334,
      "grad_norm": 15.24609661102295,
      "learning_rate": 2e-05,
      "loss": 0.2183,
      "step": 1453
    },
    {
      "epoch": 0.9827644474484624,
      "grad_norm": 12.639070510864258,
      "learning_rate": 2e-05,
      "loss": 0.2099,
      "step": 1454
    },
    {
      "epoch": 0.9834403514700912,
      "grad_norm": 11.708888053894043,
      "learning_rate": 2e-05,
      "loss": 0.1831,
      "step": 1455
    },
    {
      "epoch": 0.9841162554917202,
      "grad_norm": 9.001616477966309,
      "learning_rate": 2e-05,
      "loss": 0.2749,
      "step": 1456
    },
    {
      "epoch": 0.9847921595133491,
      "grad_norm": 10.976318359375,
      "learning_rate": 2e-05,
      "loss": 0.1678,
      "step": 1457
    },
    {
      "epoch": 0.9854680635349781,
      "grad_norm": 9.280449867248535,
      "learning_rate": 2e-05,
      "loss": 0.1747,
      "step": 1458
    },
    {
      "epoch": 0.9861439675566069,
      "grad_norm": 7.089550018310547,
      "learning_rate": 2e-05,
      "loss": 0.2179,
      "step": 1459
    },
    {
      "epoch": 0.9868198715782359,
      "grad_norm": 4.536575794219971,
      "learning_rate": 2e-05,
      "loss": 0.218,
      "step": 1460
    },
    {
      "epoch": 0.9874957755998648,
      "grad_norm": 7.5261335372924805,
      "learning_rate": 2e-05,
      "loss": 0.173,
      "step": 1461
    },
    {
      "epoch": 0.9881716796214938,
      "grad_norm": 6.0577311515808105,
      "learning_rate": 2e-05,
      "loss": 0.1405,
      "step": 1462
    },
    {
      "epoch": 0.9888475836431226,
      "grad_norm": 7.020982265472412,
      "learning_rate": 2e-05,
      "loss": 0.2108,
      "step": 1463
    },
    {
      "epoch": 0.9895234876647516,
      "grad_norm": 8.021031379699707,
      "learning_rate": 2e-05,
      "loss": 0.1574,
      "step": 1464
    },
    {
      "epoch": 0.9901993916863805,
      "grad_norm": 5.367835998535156,
      "learning_rate": 2e-05,
      "loss": 0.2128,
      "step": 1465
    },
    {
      "epoch": 0.9908752957080095,
      "grad_norm": 10.465010643005371,
      "learning_rate": 2e-05,
      "loss": 0.1311,
      "step": 1466
    },
    {
      "epoch": 0.9915511997296383,
      "grad_norm": 7.832886695861816,
      "learning_rate": 2e-05,
      "loss": 0.2304,
      "step": 1467
    },
    {
      "epoch": 0.9922271037512673,
      "grad_norm": 9.998841285705566,
      "learning_rate": 2e-05,
      "loss": 0.1466,
      "step": 1468
    },
    {
      "epoch": 0.9929030077728962,
      "grad_norm": 13.224456787109375,
      "learning_rate": 2e-05,
      "loss": 0.2396,
      "step": 1469
    },
    {
      "epoch": 0.9935789117945252,
      "grad_norm": 11.226777076721191,
      "learning_rate": 2e-05,
      "loss": 0.2676,
      "step": 1470
    },
    {
      "epoch": 0.9942548158161542,
      "grad_norm": 5.555713176727295,
      "learning_rate": 2e-05,
      "loss": 0.1673,
      "step": 1471
    },
    {
      "epoch": 0.994930719837783,
      "grad_norm": 9.338447570800781,
      "learning_rate": 2e-05,
      "loss": 0.1742,
      "step": 1472
    },
    {
      "epoch": 0.995606623859412,
      "grad_norm": 16.82993507385254,
      "learning_rate": 2e-05,
      "loss": 0.212,
      "step": 1473
    },
    {
      "epoch": 0.9962825278810409,
      "grad_norm": 18.847139358520508,
      "learning_rate": 2e-05,
      "loss": 0.2372,
      "step": 1474
    },
    {
      "epoch": 0.9969584319026699,
      "grad_norm": 5.872804164886475,
      "learning_rate": 2e-05,
      "loss": 0.2315,
      "step": 1475
    },
    {
      "epoch": 0.9976343359242987,
      "grad_norm": 4.786114692687988,
      "learning_rate": 2e-05,
      "loss": 0.1813,
      "step": 1476
    },
    {
      "epoch": 0.9983102399459277,
      "grad_norm": 28.35748291015625,
      "learning_rate": 2e-05,
      "loss": 0.3103,
      "step": 1477
    },
    {
      "epoch": 0.9989861439675566,
      "grad_norm": 14.680118560791016,
      "learning_rate": 2e-05,
      "loss": 0.1965,
      "step": 1478
    },
    {
      "epoch": 0.9996620479891856,
      "grad_norm": 4.458162784576416,
      "learning_rate": 2e-05,
      "loss": 0.1796,
      "step": 1479
    },
    {
      "epoch": 1.0,
      "grad_norm": 4.458162784576416,
      "learning_rate": 2e-05,
      "loss": 0.092,
      "step": 1480
    },
    {
      "epoch": 1.0006759040216289,
      "grad_norm": 21.5565242767334,
      "learning_rate": 2e-05,
      "loss": 0.1858,
      "step": 1481
    },
    {
      "epoch": 1.001351808043258,
      "grad_norm": 7.1103105545043945,
      "learning_rate": 2e-05,
      "loss": 0.1814,
      "step": 1482
    },
    {
      "epoch": 1.0020277120648868,
      "grad_norm": 7.880371570587158,
      "learning_rate": 2e-05,
      "loss": 0.1741,
      "step": 1483
    },
    {
      "epoch": 1.0027036160865157,
      "grad_norm": 7.954522132873535,
      "learning_rate": 2e-05,
      "loss": 0.132,
      "step": 1484
    },
    {
      "epoch": 1.0033795201081446,
      "grad_norm": 6.319475173950195,
      "learning_rate": 2e-05,
      "loss": 0.1909,
      "step": 1485
    },
    {
      "epoch": 1.0040554241297737,
      "grad_norm": 8.983370780944824,
      "learning_rate": 2e-05,
      "loss": 0.1896,
      "step": 1486
    },
    {
      "epoch": 1.0047313281514025,
      "grad_norm": 15.98290729522705,
      "learning_rate": 2e-05,
      "loss": 0.1985,
      "step": 1487
    },
    {
      "epoch": 1.0054072321730314,
      "grad_norm": 5.744195938110352,
      "learning_rate": 2e-05,
      "loss": 0.1705,
      "step": 1488
    },
    {
      "epoch": 1.0060831361946603,
      "grad_norm": 5.227535247802734,
      "learning_rate": 2e-05,
      "loss": 0.1524,
      "step": 1489
    },
    {
      "epoch": 1.0067590402162894,
      "grad_norm": 7.34737491607666,
      "learning_rate": 2e-05,
      "loss": 0.1653,
      "step": 1490
    },
    {
      "epoch": 1.0074349442379182,
      "grad_norm": 13.461752891540527,
      "learning_rate": 2e-05,
      "loss": 0.1659,
      "step": 1491
    },
    {
      "epoch": 1.008110848259547,
      "grad_norm": 8.869354248046875,
      "learning_rate": 2e-05,
      "loss": 0.1663,
      "step": 1492
    },
    {
      "epoch": 1.008786752281176,
      "grad_norm": 9.648326873779297,
      "learning_rate": 2e-05,
      "loss": 0.1885,
      "step": 1493
    },
    {
      "epoch": 1.009462656302805,
      "grad_norm": 8.934494972229004,
      "learning_rate": 2e-05,
      "loss": 0.1707,
      "step": 1494
    },
    {
      "epoch": 1.010138560324434,
      "grad_norm": 10.205559730529785,
      "learning_rate": 2e-05,
      "loss": 0.143,
      "step": 1495
    },
    {
      "epoch": 1.0108144643460628,
      "grad_norm": 12.012412071228027,
      "learning_rate": 2e-05,
      "loss": 0.1777,
      "step": 1496
    },
    {
      "epoch": 1.0114903683676917,
      "grad_norm": 8.470035552978516,
      "learning_rate": 2e-05,
      "loss": 0.1481,
      "step": 1497
    },
    {
      "epoch": 1.0121662723893208,
      "grad_norm": 13.590110778808594,
      "learning_rate": 2e-05,
      "loss": 0.2505,
      "step": 1498
    },
    {
      "epoch": 1.0128421764109496,
      "grad_norm": 10.510622024536133,
      "learning_rate": 2e-05,
      "loss": 0.1113,
      "step": 1499
    },
    {
      "epoch": 1.0135180804325785,
      "grad_norm": 5.009744167327881,
      "learning_rate": 2e-05,
      "loss": 0.1611,
      "step": 1500
    },
    {
      "epoch": 1.0135180804325785,
      "eval_accuracy": 0.5618243243243243,
      "eval_loss": 0.2593933045864105,
      "eval_runtime": 43.9381,
      "eval_samples_per_second": 538.94,
      "eval_steps_per_second": 16.842,
      "step": 1500
    },
    {
      "epoch": 1.0141939844542076,
      "grad_norm": 11.493053436279297,
      "learning_rate": 2e-05,
      "loss": 0.1934,
      "step": 1501
    },
    {
      "epoch": 1.0148698884758365,
      "grad_norm": 16.883634567260742,
      "learning_rate": 2e-05,
      "loss": 0.2285,
      "step": 1502
    },
    {
      "epoch": 1.0155457924974653,
      "grad_norm": 12.143837928771973,
      "learning_rate": 2e-05,
      "loss": 0.1549,
      "step": 1503
    },
    {
      "epoch": 1.0162216965190942,
      "grad_norm": 14.23536491394043,
      "learning_rate": 2e-05,
      "loss": 0.1952,
      "step": 1504
    },
    {
      "epoch": 1.0168976005407233,
      "grad_norm": 14.788068771362305,
      "learning_rate": 2e-05,
      "loss": 0.2332,
      "step": 1505
    },
    {
      "epoch": 1.0175735045623522,
      "grad_norm": 7.611464023590088,
      "learning_rate": 2e-05,
      "loss": 0.1805,
      "step": 1506
    },
    {
      "epoch": 1.018249408583981,
      "grad_norm": 6.606745719909668,
      "learning_rate": 2e-05,
      "loss": 0.2403,
      "step": 1507
    },
    {
      "epoch": 1.01892531260561,
      "grad_norm": 21.132848739624023,
      "learning_rate": 2e-05,
      "loss": 0.2214,
      "step": 1508
    },
    {
      "epoch": 1.019601216627239,
      "grad_norm": 5.072515487670898,
      "learning_rate": 2e-05,
      "loss": 0.2203,
      "step": 1509
    },
    {
      "epoch": 1.0202771206488679,
      "grad_norm": 16.314434051513672,
      "learning_rate": 2e-05,
      "loss": 0.1971,
      "step": 1510
    },
    {
      "epoch": 1.0209530246704968,
      "grad_norm": 16.486282348632812,
      "learning_rate": 2e-05,
      "loss": 0.1973,
      "step": 1511
    },
    {
      "epoch": 1.0216289286921256,
      "grad_norm": 12.184969902038574,
      "learning_rate": 2e-05,
      "loss": 0.1741,
      "step": 1512
    },
    {
      "epoch": 1.0223048327137547,
      "grad_norm": 13.917608261108398,
      "learning_rate": 2e-05,
      "loss": 0.2531,
      "step": 1513
    },
    {
      "epoch": 1.0229807367353836,
      "grad_norm": 20.18840980529785,
      "learning_rate": 2e-05,
      "loss": 0.2592,
      "step": 1514
    },
    {
      "epoch": 1.0236566407570125,
      "grad_norm": 13.323144912719727,
      "learning_rate": 2e-05,
      "loss": 0.1874,
      "step": 1515
    },
    {
      "epoch": 1.0243325447786413,
      "grad_norm": 8.44716739654541,
      "learning_rate": 2e-05,
      "loss": 0.1932,
      "step": 1516
    },
    {
      "epoch": 1.0250084488002704,
      "grad_norm": 5.5620269775390625,
      "learning_rate": 2e-05,
      "loss": 0.1979,
      "step": 1517
    },
    {
      "epoch": 1.0256843528218993,
      "grad_norm": 9.057921409606934,
      "learning_rate": 2e-05,
      "loss": 0.1989,
      "step": 1518
    },
    {
      "epoch": 1.0263602568435282,
      "grad_norm": 11.09757137298584,
      "learning_rate": 2e-05,
      "loss": 0.2024,
      "step": 1519
    },
    {
      "epoch": 1.0270361608651573,
      "grad_norm": 4.491654396057129,
      "learning_rate": 2e-05,
      "loss": 0.1204,
      "step": 1520
    },
    {
      "epoch": 1.0277120648867861,
      "grad_norm": 11.967937469482422,
      "learning_rate": 2e-05,
      "loss": 0.2882,
      "step": 1521
    },
    {
      "epoch": 1.028387968908415,
      "grad_norm": 18.121967315673828,
      "learning_rate": 2e-05,
      "loss": 0.1693,
      "step": 1522
    },
    {
      "epoch": 1.0290638729300439,
      "grad_norm": 5.297868251800537,
      "learning_rate": 2e-05,
      "loss": 0.1205,
      "step": 1523
    },
    {
      "epoch": 1.029739776951673,
      "grad_norm": 8.752513885498047,
      "learning_rate": 2e-05,
      "loss": 0.1492,
      "step": 1524
    },
    {
      "epoch": 1.0304156809733018,
      "grad_norm": 12.493378639221191,
      "learning_rate": 2e-05,
      "loss": 0.154,
      "step": 1525
    },
    {
      "epoch": 1.0310915849949307,
      "grad_norm": 10.127198219299316,
      "learning_rate": 2e-05,
      "loss": 0.2122,
      "step": 1526
    },
    {
      "epoch": 1.0317674890165596,
      "grad_norm": 19.531936645507812,
      "learning_rate": 2e-05,
      "loss": 0.2393,
      "step": 1527
    },
    {
      "epoch": 1.0324433930381887,
      "grad_norm": 14.250304222106934,
      "learning_rate": 2e-05,
      "loss": 0.2016,
      "step": 1528
    },
    {
      "epoch": 1.0331192970598175,
      "grad_norm": 7.805017948150635,
      "learning_rate": 2e-05,
      "loss": 0.2069,
      "step": 1529
    },
    {
      "epoch": 1.0337952010814464,
      "grad_norm": 17.818315505981445,
      "learning_rate": 2e-05,
      "loss": 0.2119,
      "step": 1530
    },
    {
      "epoch": 1.0344711051030753,
      "grad_norm": 16.684783935546875,
      "learning_rate": 2e-05,
      "loss": 0.1539,
      "step": 1531
    },
    {
      "epoch": 1.0351470091247044,
      "grad_norm": 9.802934646606445,
      "learning_rate": 2e-05,
      "loss": 0.1954,
      "step": 1532
    },
    {
      "epoch": 1.0358229131463332,
      "grad_norm": 20.43698501586914,
      "learning_rate": 2e-05,
      "loss": 0.1519,
      "step": 1533
    },
    {
      "epoch": 1.036498817167962,
      "grad_norm": 9.709138870239258,
      "learning_rate": 2e-05,
      "loss": 0.2494,
      "step": 1534
    },
    {
      "epoch": 1.037174721189591,
      "grad_norm": 7.344099044799805,
      "learning_rate": 2e-05,
      "loss": 0.1968,
      "step": 1535
    },
    {
      "epoch": 1.03785062521122,
      "grad_norm": 4.563737392425537,
      "learning_rate": 2e-05,
      "loss": 0.2262,
      "step": 1536
    },
    {
      "epoch": 1.038526529232849,
      "grad_norm": 25.876577377319336,
      "learning_rate": 2e-05,
      "loss": 0.2602,
      "step": 1537
    },
    {
      "epoch": 1.0392024332544778,
      "grad_norm": 5.60959529876709,
      "learning_rate": 2e-05,
      "loss": 0.1777,
      "step": 1538
    },
    {
      "epoch": 1.039878337276107,
      "grad_norm": 9.121501922607422,
      "learning_rate": 2e-05,
      "loss": 0.1814,
      "step": 1539
    },
    {
      "epoch": 1.0405542412977358,
      "grad_norm": 7.653229713439941,
      "learning_rate": 2e-05,
      "loss": 0.154,
      "step": 1540
    },
    {
      "epoch": 1.0412301453193646,
      "grad_norm": 7.714646816253662,
      "learning_rate": 2e-05,
      "loss": 0.1716,
      "step": 1541
    },
    {
      "epoch": 1.0419060493409935,
      "grad_norm": 10.15739917755127,
      "learning_rate": 2e-05,
      "loss": 0.2146,
      "step": 1542
    },
    {
      "epoch": 1.0425819533626226,
      "grad_norm": 10.382244110107422,
      "learning_rate": 2e-05,
      "loss": 0.1721,
      "step": 1543
    },
    {
      "epoch": 1.0432578573842515,
      "grad_norm": 8.84390640258789,
      "learning_rate": 2e-05,
      "loss": 0.2115,
      "step": 1544
    },
    {
      "epoch": 1.0439337614058803,
      "grad_norm": 10.150927543640137,
      "learning_rate": 2e-05,
      "loss": 0.1862,
      "step": 1545
    },
    {
      "epoch": 1.0446096654275092,
      "grad_norm": 5.540737628936768,
      "learning_rate": 2e-05,
      "loss": 0.1719,
      "step": 1546
    },
    {
      "epoch": 1.0452855694491383,
      "grad_norm": 6.719390869140625,
      "learning_rate": 2e-05,
      "loss": 0.2193,
      "step": 1547
    },
    {
      "epoch": 1.0459614734707672,
      "grad_norm": 5.99237060546875,
      "learning_rate": 2e-05,
      "loss": 0.2018,
      "step": 1548
    },
    {
      "epoch": 1.046637377492396,
      "grad_norm": 5.534433841705322,
      "learning_rate": 2e-05,
      "loss": 0.1483,
      "step": 1549
    },
    {
      "epoch": 1.047313281514025,
      "grad_norm": 7.99867582321167,
      "learning_rate": 2e-05,
      "loss": 0.1636,
      "step": 1550
    },
    {
      "epoch": 1.047313281514025,
      "eval_accuracy": 0.5410472972972973,
      "eval_loss": 0.24752496182918549,
      "eval_runtime": 44.1118,
      "eval_samples_per_second": 536.818,
      "eval_steps_per_second": 16.776,
      "step": 1550
    },
    {
      "epoch": 1.047989185535654,
      "grad_norm": 7.494906902313232,
      "learning_rate": 2e-05,
      "loss": 0.19,
      "step": 1551
    },
    {
      "epoch": 1.0486650895572829,
      "grad_norm": 6.5509138107299805,
      "learning_rate": 2e-05,
      "loss": 0.1867,
      "step": 1552
    },
    {
      "epoch": 1.0493409935789118,
      "grad_norm": 6.034501552581787,
      "learning_rate": 2e-05,
      "loss": 0.1952,
      "step": 1553
    },
    {
      "epoch": 1.0500168976005406,
      "grad_norm": 6.700221538543701,
      "learning_rate": 2e-05,
      "loss": 0.2099,
      "step": 1554
    },
    {
      "epoch": 1.0506928016221697,
      "grad_norm": 6.970874309539795,
      "learning_rate": 2e-05,
      "loss": 0.2154,
      "step": 1555
    },
    {
      "epoch": 1.0513687056437986,
      "grad_norm": 6.734160423278809,
      "learning_rate": 2e-05,
      "loss": 0.219,
      "step": 1556
    },
    {
      "epoch": 1.0520446096654275,
      "grad_norm": 15.16308307647705,
      "learning_rate": 2e-05,
      "loss": 0.1675,
      "step": 1557
    },
    {
      "epoch": 1.0527205136870563,
      "grad_norm": 5.907565593719482,
      "learning_rate": 2e-05,
      "loss": 0.1997,
      "step": 1558
    },
    {
      "epoch": 1.0533964177086854,
      "grad_norm": 5.556641101837158,
      "learning_rate": 2e-05,
      "loss": 0.142,
      "step": 1559
    },
    {
      "epoch": 1.0540723217303143,
      "grad_norm": 5.899667739868164,
      "learning_rate": 2e-05,
      "loss": 0.1531,
      "step": 1560
    },
    {
      "epoch": 1.0547482257519432,
      "grad_norm": 11.160567283630371,
      "learning_rate": 2e-05,
      "loss": 0.168,
      "step": 1561
    },
    {
      "epoch": 1.0554241297735723,
      "grad_norm": 5.03303337097168,
      "learning_rate": 2e-05,
      "loss": 0.1901,
      "step": 1562
    },
    {
      "epoch": 1.0561000337952011,
      "grad_norm": 5.753058433532715,
      "learning_rate": 2e-05,
      "loss": 0.1616,
      "step": 1563
    },
    {
      "epoch": 1.05677593781683,
      "grad_norm": 8.30715560913086,
      "learning_rate": 2e-05,
      "loss": 0.1894,
      "step": 1564
    },
    {
      "epoch": 1.0574518418384589,
      "grad_norm": 9.124565124511719,
      "learning_rate": 2e-05,
      "loss": 0.2021,
      "step": 1565
    },
    {
      "epoch": 1.058127745860088,
      "grad_norm": 12.789179801940918,
      "learning_rate": 2e-05,
      "loss": 0.1532,
      "step": 1566
    },
    {
      "epoch": 1.0588036498817168,
      "grad_norm": 6.772098541259766,
      "learning_rate": 2e-05,
      "loss": 0.2106,
      "step": 1567
    },
    {
      "epoch": 1.0594795539033457,
      "grad_norm": 11.740741729736328,
      "learning_rate": 2e-05,
      "loss": 0.1917,
      "step": 1568
    },
    {
      "epoch": 1.0601554579249746,
      "grad_norm": 4.266551971435547,
      "learning_rate": 2e-05,
      "loss": 0.2211,
      "step": 1569
    },
    {
      "epoch": 1.0608313619466037,
      "grad_norm": 18.40229606628418,
      "learning_rate": 2e-05,
      "loss": 0.179,
      "step": 1570
    },
    {
      "epoch": 1.0615072659682325,
      "grad_norm": 8.055603981018066,
      "learning_rate": 2e-05,
      "loss": 0.2429,
      "step": 1571
    },
    {
      "epoch": 1.0621831699898614,
      "grad_norm": 19.007583618164062,
      "learning_rate": 2e-05,
      "loss": 0.2536,
      "step": 1572
    },
    {
      "epoch": 1.0628590740114903,
      "grad_norm": 13.592592239379883,
      "learning_rate": 2e-05,
      "loss": 0.2061,
      "step": 1573
    },
    {
      "epoch": 1.0635349780331194,
      "grad_norm": 9.17104721069336,
      "learning_rate": 2e-05,
      "loss": 0.2253,
      "step": 1574
    },
    {
      "epoch": 1.0642108820547482,
      "grad_norm": 6.010147571563721,
      "learning_rate": 2e-05,
      "loss": 0.1816,
      "step": 1575
    },
    {
      "epoch": 1.064886786076377,
      "grad_norm": 7.839737415313721,
      "learning_rate": 2e-05,
      "loss": 0.2291,
      "step": 1576
    },
    {
      "epoch": 1.0655626900980062,
      "grad_norm": 21.790990829467773,
      "learning_rate": 2e-05,
      "loss": 0.1626,
      "step": 1577
    },
    {
      "epoch": 1.066238594119635,
      "grad_norm": 6.293765068054199,
      "learning_rate": 2e-05,
      "loss": 0.1969,
      "step": 1578
    },
    {
      "epoch": 1.066914498141264,
      "grad_norm": 6.565197467803955,
      "learning_rate": 2e-05,
      "loss": 0.2311,
      "step": 1579
    },
    {
      "epoch": 1.0675904021628928,
      "grad_norm": 9.486993789672852,
      "learning_rate": 2e-05,
      "loss": 0.1344,
      "step": 1580
    },
    {
      "epoch": 1.068266306184522,
      "grad_norm": 6.0743794441223145,
      "learning_rate": 2e-05,
      "loss": 0.1769,
      "step": 1581
    },
    {
      "epoch": 1.0689422102061508,
      "grad_norm": 8.510741233825684,
      "learning_rate": 2e-05,
      "loss": 0.1573,
      "step": 1582
    },
    {
      "epoch": 1.0696181142277796,
      "grad_norm": 8.71297550201416,
      "learning_rate": 2e-05,
      "loss": 0.2774,
      "step": 1583
    },
    {
      "epoch": 1.0702940182494085,
      "grad_norm": 14.98823070526123,
      "learning_rate": 2e-05,
      "loss": 0.1971,
      "step": 1584
    },
    {
      "epoch": 1.0709699222710376,
      "grad_norm": 9.315438270568848,
      "learning_rate": 2e-05,
      "loss": 0.1702,
      "step": 1585
    },
    {
      "epoch": 1.0716458262926665,
      "grad_norm": 9.593683242797852,
      "learning_rate": 2e-05,
      "loss": 0.2192,
      "step": 1586
    },
    {
      "epoch": 1.0723217303142953,
      "grad_norm": 13.582136154174805,
      "learning_rate": 2e-05,
      "loss": 0.2136,
      "step": 1587
    },
    {
      "epoch": 1.0729976343359242,
      "grad_norm": 7.657256126403809,
      "learning_rate": 2e-05,
      "loss": 0.182,
      "step": 1588
    },
    {
      "epoch": 1.0736735383575533,
      "grad_norm": 9.354362487792969,
      "learning_rate": 2e-05,
      "loss": 0.1652,
      "step": 1589
    },
    {
      "epoch": 1.0743494423791822,
      "grad_norm": 13.812020301818848,
      "learning_rate": 2e-05,
      "loss": 0.1791,
      "step": 1590
    },
    {
      "epoch": 1.075025346400811,
      "grad_norm": 11.258816719055176,
      "learning_rate": 2e-05,
      "loss": 0.1919,
      "step": 1591
    },
    {
      "epoch": 1.07570125042244,
      "grad_norm": 6.774445056915283,
      "learning_rate": 2e-05,
      "loss": 0.2464,
      "step": 1592
    },
    {
      "epoch": 1.076377154444069,
      "grad_norm": 10.06596851348877,
      "learning_rate": 2e-05,
      "loss": 0.226,
      "step": 1593
    },
    {
      "epoch": 1.0770530584656979,
      "grad_norm": 10.797426223754883,
      "learning_rate": 2e-05,
      "loss": 0.2485,
      "step": 1594
    },
    {
      "epoch": 1.0777289624873267,
      "grad_norm": 5.893416881561279,
      "learning_rate": 2e-05,
      "loss": 0.2072,
      "step": 1595
    },
    {
      "epoch": 1.0784048665089556,
      "grad_norm": 9.285983085632324,
      "learning_rate": 2e-05,
      "loss": 0.2086,
      "step": 1596
    },
    {
      "epoch": 1.0790807705305847,
      "grad_norm": 4.8580098152160645,
      "learning_rate": 2e-05,
      "loss": 0.1829,
      "step": 1597
    },
    {
      "epoch": 1.0797566745522136,
      "grad_norm": 4.218326568603516,
      "learning_rate": 2e-05,
      "loss": 0.1724,
      "step": 1598
    },
    {
      "epoch": 1.0804325785738425,
      "grad_norm": 4.816809177398682,
      "learning_rate": 2e-05,
      "loss": 0.203,
      "step": 1599
    },
    {
      "epoch": 1.0811084825954715,
      "grad_norm": 11.576028823852539,
      "learning_rate": 2e-05,
      "loss": 0.1907,
      "step": 1600
    },
    {
      "epoch": 1.0811084825954715,
      "eval_accuracy": 0.5567567567567567,
      "eval_loss": 0.2351687103509903,
      "eval_runtime": 43.1544,
      "eval_samples_per_second": 548.727,
      "eval_steps_per_second": 17.148,
      "step": 1600
    },
    {
      "epoch": 1.0817843866171004,
      "grad_norm": 4.173635482788086,
      "learning_rate": 2e-05,
      "loss": 0.1959,
      "step": 1601
    },
    {
      "epoch": 1.0824602906387293,
      "grad_norm": 6.389683246612549,
      "learning_rate": 2e-05,
      "loss": 0.2106,
      "step": 1602
    },
    {
      "epoch": 1.0831361946603582,
      "grad_norm": 8.429895401000977,
      "learning_rate": 2e-05,
      "loss": 0.2055,
      "step": 1603
    },
    {
      "epoch": 1.0838120986819872,
      "grad_norm": 14.725639343261719,
      "learning_rate": 2e-05,
      "loss": 0.1942,
      "step": 1604
    },
    {
      "epoch": 1.0844880027036161,
      "grad_norm": 14.07395076751709,
      "learning_rate": 2e-05,
      "loss": 0.2085,
      "step": 1605
    },
    {
      "epoch": 1.085163906725245,
      "grad_norm": 10.94120979309082,
      "learning_rate": 2e-05,
      "loss": 0.2017,
      "step": 1606
    },
    {
      "epoch": 1.0858398107468739,
      "grad_norm": 3.7622194290161133,
      "learning_rate": 2e-05,
      "loss": 0.1458,
      "step": 1607
    },
    {
      "epoch": 1.086515714768503,
      "grad_norm": 7.018836498260498,
      "learning_rate": 2e-05,
      "loss": 0.2083,
      "step": 1608
    },
    {
      "epoch": 1.0871916187901318,
      "grad_norm": 7.606606960296631,
      "learning_rate": 2e-05,
      "loss": 0.1441,
      "step": 1609
    },
    {
      "epoch": 1.0878675228117607,
      "grad_norm": 7.8314032554626465,
      "learning_rate": 2e-05,
      "loss": 0.2306,
      "step": 1610
    },
    {
      "epoch": 1.0885434268333896,
      "grad_norm": 14.65768051147461,
      "learning_rate": 2e-05,
      "loss": 0.2299,
      "step": 1611
    },
    {
      "epoch": 1.0892193308550187,
      "grad_norm": 7.190499782562256,
      "learning_rate": 2e-05,
      "loss": 0.2204,
      "step": 1612
    },
    {
      "epoch": 1.0898952348766475,
      "grad_norm": 10.105857849121094,
      "learning_rate": 2e-05,
      "loss": 0.1893,
      "step": 1613
    },
    {
      "epoch": 1.0905711388982764,
      "grad_norm": 5.7836222648620605,
      "learning_rate": 2e-05,
      "loss": 0.1535,
      "step": 1614
    },
    {
      "epoch": 1.0912470429199055,
      "grad_norm": 12.3967924118042,
      "learning_rate": 2e-05,
      "loss": 0.1768,
      "step": 1615
    },
    {
      "epoch": 1.0919229469415344,
      "grad_norm": 14.377320289611816,
      "learning_rate": 2e-05,
      "loss": 0.2219,
      "step": 1616
    },
    {
      "epoch": 1.0925988509631632,
      "grad_norm": 12.372457504272461,
      "learning_rate": 2e-05,
      "loss": 0.2046,
      "step": 1617
    },
    {
      "epoch": 1.093274754984792,
      "grad_norm": 10.94054126739502,
      "learning_rate": 2e-05,
      "loss": 0.2245,
      "step": 1618
    },
    {
      "epoch": 1.0939506590064212,
      "grad_norm": 5.4417877197265625,
      "learning_rate": 2e-05,
      "loss": 0.1928,
      "step": 1619
    },
    {
      "epoch": 1.09462656302805,
      "grad_norm": 4.946004390716553,
      "learning_rate": 2e-05,
      "loss": 0.157,
      "step": 1620
    },
    {
      "epoch": 1.095302467049679,
      "grad_norm": 7.771331310272217,
      "learning_rate": 2e-05,
      "loss": 0.1703,
      "step": 1621
    },
    {
      "epoch": 1.0959783710713078,
      "grad_norm": 24.649471282958984,
      "learning_rate": 2e-05,
      "loss": 0.2663,
      "step": 1622
    },
    {
      "epoch": 1.096654275092937,
      "grad_norm": 15.22679328918457,
      "learning_rate": 2e-05,
      "loss": 0.2025,
      "step": 1623
    },
    {
      "epoch": 1.0973301791145658,
      "grad_norm": 7.78530740737915,
      "learning_rate": 2e-05,
      "loss": 0.1277,
      "step": 1624
    },
    {
      "epoch": 1.0980060831361946,
      "grad_norm": 5.487343788146973,
      "learning_rate": 2e-05,
      "loss": 0.2171,
      "step": 1625
    },
    {
      "epoch": 1.0986819871578235,
      "grad_norm": 8.94001579284668,
      "learning_rate": 2e-05,
      "loss": 0.243,
      "step": 1626
    },
    {
      "epoch": 1.0993578911794526,
      "grad_norm": 24.005271911621094,
      "learning_rate": 2e-05,
      "loss": 0.2577,
      "step": 1627
    },
    {
      "epoch": 1.1000337952010815,
      "grad_norm": 22.163318634033203,
      "learning_rate": 2e-05,
      "loss": 0.242,
      "step": 1628
    },
    {
      "epoch": 1.1007096992227103,
      "grad_norm": 16.351396560668945,
      "learning_rate": 2e-05,
      "loss": 0.2216,
      "step": 1629
    },
    {
      "epoch": 1.1013856032443392,
      "grad_norm": 5.335818767547607,
      "learning_rate": 2e-05,
      "loss": 0.1763,
      "step": 1630
    },
    {
      "epoch": 1.1020615072659683,
      "grad_norm": 24.958620071411133,
      "learning_rate": 2e-05,
      "loss": 0.302,
      "step": 1631
    },
    {
      "epoch": 1.1027374112875972,
      "grad_norm": 21.881019592285156,
      "learning_rate": 2e-05,
      "loss": 0.2429,
      "step": 1632
    },
    {
      "epoch": 1.103413315309226,
      "grad_norm": 13.428984642028809,
      "learning_rate": 2e-05,
      "loss": 0.1844,
      "step": 1633
    },
    {
      "epoch": 1.104089219330855,
      "grad_norm": 9.528614044189453,
      "learning_rate": 2e-05,
      "loss": 0.1783,
      "step": 1634
    },
    {
      "epoch": 1.104765123352484,
      "grad_norm": 8.980578422546387,
      "learning_rate": 2e-05,
      "loss": 0.1631,
      "step": 1635
    },
    {
      "epoch": 1.1054410273741129,
      "grad_norm": 17.745315551757812,
      "learning_rate": 2e-05,
      "loss": 0.3105,
      "step": 1636
    },
    {
      "epoch": 1.1061169313957417,
      "grad_norm": 19.223962783813477,
      "learning_rate": 2e-05,
      "loss": 0.1582,
      "step": 1637
    },
    {
      "epoch": 1.1067928354173708,
      "grad_norm": 10.06949520111084,
      "learning_rate": 2e-05,
      "loss": 0.1995,
      "step": 1638
    },
    {
      "epoch": 1.1074687394389997,
      "grad_norm": 8.76278018951416,
      "learning_rate": 2e-05,
      "loss": 0.1898,
      "step": 1639
    },
    {
      "epoch": 1.1081446434606286,
      "grad_norm": 7.692691326141357,
      "learning_rate": 2e-05,
      "loss": 0.1217,
      "step": 1640
    },
    {
      "epoch": 1.1088205474822574,
      "grad_norm": 4.925232887268066,
      "learning_rate": 2e-05,
      "loss": 0.1506,
      "step": 1641
    },
    {
      "epoch": 1.1094964515038865,
      "grad_norm": 11.721088409423828,
      "learning_rate": 2e-05,
      "loss": 0.2145,
      "step": 1642
    },
    {
      "epoch": 1.1101723555255154,
      "grad_norm": 4.519599437713623,
      "learning_rate": 2e-05,
      "loss": 0.182,
      "step": 1643
    },
    {
      "epoch": 1.1108482595471443,
      "grad_norm": 7.33176326751709,
      "learning_rate": 2e-05,
      "loss": 0.184,
      "step": 1644
    },
    {
      "epoch": 1.1115241635687731,
      "grad_norm": 12.405784606933594,
      "learning_rate": 2e-05,
      "loss": 0.1537,
      "step": 1645
    },
    {
      "epoch": 1.1122000675904022,
      "grad_norm": 7.45192813873291,
      "learning_rate": 2e-05,
      "loss": 0.184,
      "step": 1646
    },
    {
      "epoch": 1.1128759716120311,
      "grad_norm": 5.690074443817139,
      "learning_rate": 2e-05,
      "loss": 0.1738,
      "step": 1647
    },
    {
      "epoch": 1.11355187563366,
      "grad_norm": 11.817813873291016,
      "learning_rate": 2e-05,
      "loss": 0.1827,
      "step": 1648
    },
    {
      "epoch": 1.1142277796552889,
      "grad_norm": 10.610444068908691,
      "learning_rate": 2e-05,
      "loss": 0.1561,
      "step": 1649
    },
    {
      "epoch": 1.114903683676918,
      "grad_norm": 10.52305793762207,
      "learning_rate": 2e-05,
      "loss": 0.2789,
      "step": 1650
    },
    {
      "epoch": 1.114903683676918,
      "eval_accuracy": 0.5798141891891891,
      "eval_loss": 0.27709394693374634,
      "eval_runtime": 43.2159,
      "eval_samples_per_second": 547.946,
      "eval_steps_per_second": 17.123,
      "step": 1650
    },
    {
      "epoch": 1.1155795876985468,
      "grad_norm": 14.026397705078125,
      "learning_rate": 2e-05,
      "loss": 0.1806,
      "step": 1651
    },
    {
      "epoch": 1.1162554917201757,
      "grad_norm": 6.374835014343262,
      "learning_rate": 2e-05,
      "loss": 0.1239,
      "step": 1652
    },
    {
      "epoch": 1.1169313957418048,
      "grad_norm": 9.793830871582031,
      "learning_rate": 2e-05,
      "loss": 0.1706,
      "step": 1653
    },
    {
      "epoch": 1.1176072997634336,
      "grad_norm": 10.387441635131836,
      "learning_rate": 2e-05,
      "loss": 0.2447,
      "step": 1654
    },
    {
      "epoch": 1.1182832037850625,
      "grad_norm": 9.949629783630371,
      "learning_rate": 2e-05,
      "loss": 0.2159,
      "step": 1655
    },
    {
      "epoch": 1.1189591078066914,
      "grad_norm": 5.654441833496094,
      "learning_rate": 2e-05,
      "loss": 0.2365,
      "step": 1656
    },
    {
      "epoch": 1.1196350118283205,
      "grad_norm": 5.238239765167236,
      "learning_rate": 2e-05,
      "loss": 0.187,
      "step": 1657
    },
    {
      "epoch": 1.1203109158499494,
      "grad_norm": 11.907963752746582,
      "learning_rate": 2e-05,
      "loss": 0.2148,
      "step": 1658
    },
    {
      "epoch": 1.1209868198715782,
      "grad_norm": 5.377718925476074,
      "learning_rate": 2e-05,
      "loss": 0.1749,
      "step": 1659
    },
    {
      "epoch": 1.121662723893207,
      "grad_norm": 9.349112510681152,
      "learning_rate": 2e-05,
      "loss": 0.2676,
      "step": 1660
    },
    {
      "epoch": 1.1223386279148362,
      "grad_norm": 8.235374450683594,
      "learning_rate": 2e-05,
      "loss": 0.1948,
      "step": 1661
    },
    {
      "epoch": 1.123014531936465,
      "grad_norm": 6.572847366333008,
      "learning_rate": 2e-05,
      "loss": 0.1312,
      "step": 1662
    },
    {
      "epoch": 1.123690435958094,
      "grad_norm": 6.189995288848877,
      "learning_rate": 2e-05,
      "loss": 0.1624,
      "step": 1663
    },
    {
      "epoch": 1.1243663399797228,
      "grad_norm": 5.19527530670166,
      "learning_rate": 2e-05,
      "loss": 0.2036,
      "step": 1664
    },
    {
      "epoch": 1.1250422440013519,
      "grad_norm": 4.6053266525268555,
      "learning_rate": 2e-05,
      "loss": 0.158,
      "step": 1665
    },
    {
      "epoch": 1.1257181480229808,
      "grad_norm": 7.997122287750244,
      "learning_rate": 2e-05,
      "loss": 0.1444,
      "step": 1666
    },
    {
      "epoch": 1.1263940520446096,
      "grad_norm": 5.123172760009766,
      "learning_rate": 2e-05,
      "loss": 0.1716,
      "step": 1667
    },
    {
      "epoch": 1.1270699560662385,
      "grad_norm": 5.152861595153809,
      "learning_rate": 2e-05,
      "loss": 0.1814,
      "step": 1668
    },
    {
      "epoch": 1.1277458600878676,
      "grad_norm": 8.465657234191895,
      "learning_rate": 2e-05,
      "loss": 0.2132,
      "step": 1669
    },
    {
      "epoch": 1.1284217641094965,
      "grad_norm": 8.66786003112793,
      "learning_rate": 2e-05,
      "loss": 0.1809,
      "step": 1670
    },
    {
      "epoch": 1.1290976681311253,
      "grad_norm": 8.214608192443848,
      "learning_rate": 2e-05,
      "loss": 0.2948,
      "step": 1671
    },
    {
      "epoch": 1.1297735721527542,
      "grad_norm": 20.78453254699707,
      "learning_rate": 2e-05,
      "loss": 0.2573,
      "step": 1672
    },
    {
      "epoch": 1.1304494761743833,
      "grad_norm": 6.199664115905762,
      "learning_rate": 2e-05,
      "loss": 0.2184,
      "step": 1673
    },
    {
      "epoch": 1.1311253801960122,
      "grad_norm": 8.72745418548584,
      "learning_rate": 2e-05,
      "loss": 0.158,
      "step": 1674
    },
    {
      "epoch": 1.131801284217641,
      "grad_norm": 5.726783752441406,
      "learning_rate": 2e-05,
      "loss": 0.1673,
      "step": 1675
    },
    {
      "epoch": 1.1324771882392701,
      "grad_norm": 11.06828498840332,
      "learning_rate": 2e-05,
      "loss": 0.2288,
      "step": 1676
    },
    {
      "epoch": 1.133153092260899,
      "grad_norm": 8.737191200256348,
      "learning_rate": 2e-05,
      "loss": 0.2439,
      "step": 1677
    },
    {
      "epoch": 1.1338289962825279,
      "grad_norm": 13.026644706726074,
      "learning_rate": 2e-05,
      "loss": 0.1561,
      "step": 1678
    },
    {
      "epoch": 1.1345049003041567,
      "grad_norm": 5.769721031188965,
      "learning_rate": 2e-05,
      "loss": 0.1714,
      "step": 1679
    },
    {
      "epoch": 1.1351808043257858,
      "grad_norm": 4.429483890533447,
      "learning_rate": 2e-05,
      "loss": 0.2313,
      "step": 1680
    },
    {
      "epoch": 1.1358567083474147,
      "grad_norm": 15.695980072021484,
      "learning_rate": 2e-05,
      "loss": 0.2342,
      "step": 1681
    },
    {
      "epoch": 1.1365326123690436,
      "grad_norm": 12.824810981750488,
      "learning_rate": 2e-05,
      "loss": 0.1667,
      "step": 1682
    },
    {
      "epoch": 1.1372085163906724,
      "grad_norm": 11.951675415039062,
      "learning_rate": 2e-05,
      "loss": 0.1978,
      "step": 1683
    },
    {
      "epoch": 1.1378844204123015,
      "grad_norm": 7.9456915855407715,
      "learning_rate": 2e-05,
      "loss": 0.2563,
      "step": 1684
    },
    {
      "epoch": 1.1385603244339304,
      "grad_norm": 5.929111480712891,
      "learning_rate": 2e-05,
      "loss": 0.1792,
      "step": 1685
    },
    {
      "epoch": 1.1392362284555593,
      "grad_norm": 11.506423950195312,
      "learning_rate": 2e-05,
      "loss": 0.2796,
      "step": 1686
    },
    {
      "epoch": 1.1399121324771881,
      "grad_norm": 10.89874267578125,
      "learning_rate": 2e-05,
      "loss": 0.2543,
      "step": 1687
    },
    {
      "epoch": 1.1405880364988172,
      "grad_norm": 12.746517181396484,
      "learning_rate": 2e-05,
      "loss": 0.2253,
      "step": 1688
    },
    {
      "epoch": 1.141263940520446,
      "grad_norm": 7.471184730529785,
      "learning_rate": 2e-05,
      "loss": 0.1533,
      "step": 1689
    },
    {
      "epoch": 1.141939844542075,
      "grad_norm": 8.1967191696167,
      "learning_rate": 2e-05,
      "loss": 0.1633,
      "step": 1690
    },
    {
      "epoch": 1.142615748563704,
      "grad_norm": 12.683608055114746,
      "learning_rate": 2e-05,
      "loss": 0.2197,
      "step": 1691
    },
    {
      "epoch": 1.143291652585333,
      "grad_norm": 6.299542427062988,
      "learning_rate": 2e-05,
      "loss": 0.2139,
      "step": 1692
    },
    {
      "epoch": 1.1439675566069618,
      "grad_norm": 4.293118476867676,
      "learning_rate": 2e-05,
      "loss": 0.1775,
      "step": 1693
    },
    {
      "epoch": 1.1446434606285907,
      "grad_norm": 7.741697311401367,
      "learning_rate": 2e-05,
      "loss": 0.2109,
      "step": 1694
    },
    {
      "epoch": 1.1453193646502196,
      "grad_norm": 8.122244834899902,
      "learning_rate": 2e-05,
      "loss": 0.19,
      "step": 1695
    },
    {
      "epoch": 1.1459952686718486,
      "grad_norm": 13.946004867553711,
      "learning_rate": 2e-05,
      "loss": 0.276,
      "step": 1696
    },
    {
      "epoch": 1.1466711726934775,
      "grad_norm": 8.122711181640625,
      "learning_rate": 2e-05,
      "loss": 0.2012,
      "step": 1697
    },
    {
      "epoch": 1.1473470767151064,
      "grad_norm": 8.3095064163208,
      "learning_rate": 2e-05,
      "loss": 0.1902,
      "step": 1698
    },
    {
      "epoch": 1.1480229807367355,
      "grad_norm": 12.27248477935791,
      "learning_rate": 2e-05,
      "loss": 0.2328,
      "step": 1699
    },
    {
      "epoch": 1.1486988847583643,
      "grad_norm": 5.6304755210876465,
      "learning_rate": 2e-05,
      "loss": 0.179,
      "step": 1700
    },
    {
      "epoch": 1.1486988847583643,
      "eval_accuracy": 0.5249155405405406,
      "eval_loss": 0.25242629647254944,
      "eval_runtime": 43.1843,
      "eval_samples_per_second": 548.348,
      "eval_steps_per_second": 17.136,
      "step": 1700
    },
    {
      "epoch": 1.1493747887799932,
      "grad_norm": 5.146942615509033,
      "learning_rate": 2e-05,
      "loss": 0.1834,
      "step": 1701
    },
    {
      "epoch": 1.150050692801622,
      "grad_norm": 9.923749923706055,
      "learning_rate": 2e-05,
      "loss": 0.2232,
      "step": 1702
    },
    {
      "epoch": 1.1507265968232512,
      "grad_norm": 6.141946315765381,
      "learning_rate": 2e-05,
      "loss": 0.1464,
      "step": 1703
    },
    {
      "epoch": 1.15140250084488,
      "grad_norm": 6.497961521148682,
      "learning_rate": 2e-05,
      "loss": 0.1948,
      "step": 1704
    },
    {
      "epoch": 1.152078404866509,
      "grad_norm": 8.933940887451172,
      "learning_rate": 2e-05,
      "loss": 0.1924,
      "step": 1705
    },
    {
      "epoch": 1.1527543088881378,
      "grad_norm": 7.441012382507324,
      "learning_rate": 2e-05,
      "loss": 0.1913,
      "step": 1706
    },
    {
      "epoch": 1.1534302129097669,
      "grad_norm": 9.877729415893555,
      "learning_rate": 2e-05,
      "loss": 0.2286,
      "step": 1707
    },
    {
      "epoch": 1.1541061169313958,
      "grad_norm": 20.113483428955078,
      "learning_rate": 2e-05,
      "loss": 0.2104,
      "step": 1708
    },
    {
      "epoch": 1.1547820209530246,
      "grad_norm": 5.081660270690918,
      "learning_rate": 2e-05,
      "loss": 0.1816,
      "step": 1709
    },
    {
      "epoch": 1.1554579249746535,
      "grad_norm": 10.323714256286621,
      "learning_rate": 2e-05,
      "loss": 0.1986,
      "step": 1710
    },
    {
      "epoch": 1.1561338289962826,
      "grad_norm": 12.184616088867188,
      "learning_rate": 2e-05,
      "loss": 0.1957,
      "step": 1711
    },
    {
      "epoch": 1.1568097330179115,
      "grad_norm": 7.0370330810546875,
      "learning_rate": 2e-05,
      "loss": 0.2237,
      "step": 1712
    },
    {
      "epoch": 1.1574856370395403,
      "grad_norm": 11.271484375,
      "learning_rate": 2e-05,
      "loss": 0.184,
      "step": 1713
    },
    {
      "epoch": 1.1581615410611694,
      "grad_norm": 12.165515899658203,
      "learning_rate": 2e-05,
      "loss": 0.1228,
      "step": 1714
    },
    {
      "epoch": 1.1588374450827983,
      "grad_norm": 6.417056083679199,
      "learning_rate": 2e-05,
      "loss": 0.1994,
      "step": 1715
    },
    {
      "epoch": 1.1595133491044272,
      "grad_norm": 9.35173511505127,
      "learning_rate": 2e-05,
      "loss": 0.2186,
      "step": 1716
    },
    {
      "epoch": 1.160189253126056,
      "grad_norm": 7.757330894470215,
      "learning_rate": 2e-05,
      "loss": 0.1557,
      "step": 1717
    },
    {
      "epoch": 1.1608651571476851,
      "grad_norm": 5.977793216705322,
      "learning_rate": 2e-05,
      "loss": 0.1901,
      "step": 1718
    },
    {
      "epoch": 1.161541061169314,
      "grad_norm": 7.802349090576172,
      "learning_rate": 2e-05,
      "loss": 0.1752,
      "step": 1719
    },
    {
      "epoch": 1.1622169651909429,
      "grad_norm": 5.732213020324707,
      "learning_rate": 2e-05,
      "loss": 0.1863,
      "step": 1720
    },
    {
      "epoch": 1.1628928692125717,
      "grad_norm": 12.12806224822998,
      "learning_rate": 2e-05,
      "loss": 0.2103,
      "step": 1721
    },
    {
      "epoch": 1.1635687732342008,
      "grad_norm": 5.558874607086182,
      "learning_rate": 2e-05,
      "loss": 0.1823,
      "step": 1722
    },
    {
      "epoch": 1.1642446772558297,
      "grad_norm": 9.363921165466309,
      "learning_rate": 2e-05,
      "loss": 0.2239,
      "step": 1723
    },
    {
      "epoch": 1.1649205812774586,
      "grad_norm": 13.906726837158203,
      "learning_rate": 2e-05,
      "loss": 0.2399,
      "step": 1724
    },
    {
      "epoch": 1.1655964852990874,
      "grad_norm": 14.188082695007324,
      "learning_rate": 2e-05,
      "loss": 0.1825,
      "step": 1725
    },
    {
      "epoch": 1.1662723893207165,
      "grad_norm": 6.657677173614502,
      "learning_rate": 2e-05,
      "loss": 0.1853,
      "step": 1726
    },
    {
      "epoch": 1.1669482933423454,
      "grad_norm": 4.3919172286987305,
      "learning_rate": 2e-05,
      "loss": 0.1577,
      "step": 1727
    },
    {
      "epoch": 1.1676241973639743,
      "grad_norm": 6.953029155731201,
      "learning_rate": 2e-05,
      "loss": 0.1639,
      "step": 1728
    },
    {
      "epoch": 1.1683001013856034,
      "grad_norm": 4.777359485626221,
      "learning_rate": 2e-05,
      "loss": 0.2182,
      "step": 1729
    },
    {
      "epoch": 1.1689760054072322,
      "grad_norm": 21.190176010131836,
      "learning_rate": 2e-05,
      "loss": 0.2563,
      "step": 1730
    },
    {
      "epoch": 1.169651909428861,
      "grad_norm": 24.972095489501953,
      "learning_rate": 2e-05,
      "loss": 0.2386,
      "step": 1731
    },
    {
      "epoch": 1.17032781345049,
      "grad_norm": 6.158979415893555,
      "learning_rate": 2e-05,
      "loss": 0.1813,
      "step": 1732
    },
    {
      "epoch": 1.1710037174721188,
      "grad_norm": 4.759649753570557,
      "learning_rate": 2e-05,
      "loss": 0.1703,
      "step": 1733
    },
    {
      "epoch": 1.171679621493748,
      "grad_norm": 12.787768363952637,
      "learning_rate": 2e-05,
      "loss": 0.1833,
      "step": 1734
    },
    {
      "epoch": 1.1723555255153768,
      "grad_norm": 8.353757858276367,
      "learning_rate": 2e-05,
      "loss": 0.1698,
      "step": 1735
    },
    {
      "epoch": 1.1730314295370057,
      "grad_norm": 6.793301105499268,
      "learning_rate": 2e-05,
      "loss": 0.2434,
      "step": 1736
    },
    {
      "epoch": 1.1737073335586348,
      "grad_norm": 10.916913986206055,
      "learning_rate": 2e-05,
      "loss": 0.2229,
      "step": 1737
    },
    {
      "epoch": 1.1743832375802636,
      "grad_norm": 13.575787544250488,
      "learning_rate": 2e-05,
      "loss": 0.1689,
      "step": 1738
    },
    {
      "epoch": 1.1750591416018925,
      "grad_norm": 12.402754783630371,
      "learning_rate": 2e-05,
      "loss": 0.1927,
      "step": 1739
    },
    {
      "epoch": 1.1757350456235214,
      "grad_norm": 10.311878204345703,
      "learning_rate": 2e-05,
      "loss": 0.1893,
      "step": 1740
    },
    {
      "epoch": 1.1764109496451505,
      "grad_norm": 8.042716979980469,
      "learning_rate": 2e-05,
      "loss": 0.1473,
      "step": 1741
    },
    {
      "epoch": 1.1770868536667793,
      "grad_norm": 16.54580307006836,
      "learning_rate": 2e-05,
      "loss": 0.2186,
      "step": 1742
    },
    {
      "epoch": 1.1777627576884082,
      "grad_norm": 6.49767541885376,
      "learning_rate": 2e-05,
      "loss": 0.1279,
      "step": 1743
    },
    {
      "epoch": 1.178438661710037,
      "grad_norm": 4.453315734863281,
      "learning_rate": 2e-05,
      "loss": 0.1385,
      "step": 1744
    },
    {
      "epoch": 1.1791145657316662,
      "grad_norm": 11.521588325500488,
      "learning_rate": 2e-05,
      "loss": 0.1927,
      "step": 1745
    },
    {
      "epoch": 1.179790469753295,
      "grad_norm": 6.927041053771973,
      "learning_rate": 2e-05,
      "loss": 0.1677,
      "step": 1746
    },
    {
      "epoch": 1.180466373774924,
      "grad_norm": 4.902695655822754,
      "learning_rate": 2e-05,
      "loss": 0.2328,
      "step": 1747
    },
    {
      "epoch": 1.1811422777965528,
      "grad_norm": 7.413721084594727,
      "learning_rate": 2e-05,
      "loss": 0.1489,
      "step": 1748
    },
    {
      "epoch": 1.1818181818181819,
      "grad_norm": 7.809430122375488,
      "learning_rate": 2e-05,
      "loss": 0.1703,
      "step": 1749
    },
    {
      "epoch": 1.1824940858398107,
      "grad_norm": 10.000692367553711,
      "learning_rate": 2e-05,
      "loss": 0.1553,
      "step": 1750
    },
    {
      "epoch": 1.1824940858398107,
      "eval_accuracy": 0.5375844594594594,
      "eval_loss": 0.2562502920627594,
      "eval_runtime": 43.8329,
      "eval_samples_per_second": 540.233,
      "eval_steps_per_second": 16.882,
      "step": 1750
    },
    {
      "epoch": 1.1831699898614396,
      "grad_norm": 9.801039695739746,
      "learning_rate": 2e-05,
      "loss": 0.1581,
      "step": 1751
    },
    {
      "epoch": 1.1838458938830687,
      "grad_norm": 12.205564498901367,
      "learning_rate": 2e-05,
      "loss": 0.1425,
      "step": 1752
    },
    {
      "epoch": 1.1845217979046976,
      "grad_norm": 12.849268913269043,
      "learning_rate": 2e-05,
      "loss": 0.2005,
      "step": 1753
    },
    {
      "epoch": 1.1851977019263265,
      "grad_norm": 17.78269386291504,
      "learning_rate": 2e-05,
      "loss": 0.1704,
      "step": 1754
    },
    {
      "epoch": 1.1858736059479553,
      "grad_norm": 5.69657039642334,
      "learning_rate": 2e-05,
      "loss": 0.1786,
      "step": 1755
    },
    {
      "epoch": 1.1865495099695844,
      "grad_norm": 6.391088008880615,
      "learning_rate": 2e-05,
      "loss": 0.1974,
      "step": 1756
    },
    {
      "epoch": 1.1872254139912133,
      "grad_norm": 9.732715606689453,
      "learning_rate": 2e-05,
      "loss": 0.1397,
      "step": 1757
    },
    {
      "epoch": 1.1879013180128422,
      "grad_norm": 5.983456611633301,
      "learning_rate": 2e-05,
      "loss": 0.1975,
      "step": 1758
    },
    {
      "epoch": 1.188577222034471,
      "grad_norm": 6.190369129180908,
      "learning_rate": 2e-05,
      "loss": 0.2352,
      "step": 1759
    },
    {
      "epoch": 1.1892531260561001,
      "grad_norm": 7.038388729095459,
      "learning_rate": 2e-05,
      "loss": 0.192,
      "step": 1760
    },
    {
      "epoch": 1.189929030077729,
      "grad_norm": 6.671899318695068,
      "learning_rate": 2e-05,
      "loss": 0.2465,
      "step": 1761
    },
    {
      "epoch": 1.1906049340993579,
      "grad_norm": 15.03160285949707,
      "learning_rate": 2e-05,
      "loss": 0.2115,
      "step": 1762
    },
    {
      "epoch": 1.1912808381209867,
      "grad_norm": 5.443784713745117,
      "learning_rate": 2e-05,
      "loss": 0.2006,
      "step": 1763
    },
    {
      "epoch": 1.1919567421426158,
      "grad_norm": 5.314581871032715,
      "learning_rate": 2e-05,
      "loss": 0.1496,
      "step": 1764
    },
    {
      "epoch": 1.1926326461642447,
      "grad_norm": 4.442335605621338,
      "learning_rate": 2e-05,
      "loss": 0.1859,
      "step": 1765
    },
    {
      "epoch": 1.1933085501858736,
      "grad_norm": 5.820644378662109,
      "learning_rate": 2e-05,
      "loss": 0.1876,
      "step": 1766
    },
    {
      "epoch": 1.1939844542075027,
      "grad_norm": 5.902360439300537,
      "learning_rate": 2e-05,
      "loss": 0.1921,
      "step": 1767
    },
    {
      "epoch": 1.1946603582291315,
      "grad_norm": 5.162236213684082,
      "learning_rate": 2e-05,
      "loss": 0.182,
      "step": 1768
    },
    {
      "epoch": 1.1953362622507604,
      "grad_norm": 6.795418739318848,
      "learning_rate": 2e-05,
      "loss": 0.1673,
      "step": 1769
    },
    {
      "epoch": 1.1960121662723893,
      "grad_norm": 4.9824748039245605,
      "learning_rate": 2e-05,
      "loss": 0.18,
      "step": 1770
    },
    {
      "epoch": 1.1966880702940181,
      "grad_norm": 8.24107551574707,
      "learning_rate": 2e-05,
      "loss": 0.2108,
      "step": 1771
    },
    {
      "epoch": 1.1973639743156472,
      "grad_norm": 9.094738006591797,
      "learning_rate": 2e-05,
      "loss": 0.2243,
      "step": 1772
    },
    {
      "epoch": 1.198039878337276,
      "grad_norm": 8.967723846435547,
      "learning_rate": 2e-05,
      "loss": 0.1953,
      "step": 1773
    },
    {
      "epoch": 1.198715782358905,
      "grad_norm": 5.480639934539795,
      "learning_rate": 2e-05,
      "loss": 0.2263,
      "step": 1774
    },
    {
      "epoch": 1.199391686380534,
      "grad_norm": 16.42664337158203,
      "learning_rate": 2e-05,
      "loss": 0.2207,
      "step": 1775
    },
    {
      "epoch": 1.200067590402163,
      "grad_norm": 17.21202278137207,
      "learning_rate": 2e-05,
      "loss": 0.22,
      "step": 1776
    },
    {
      "epoch": 1.2007434944237918,
      "grad_norm": 6.244174957275391,
      "learning_rate": 2e-05,
      "loss": 0.1911,
      "step": 1777
    },
    {
      "epoch": 1.2014193984454207,
      "grad_norm": 4.724922180175781,
      "learning_rate": 2e-05,
      "loss": 0.1465,
      "step": 1778
    },
    {
      "epoch": 1.2020953024670498,
      "grad_norm": 5.306416988372803,
      "learning_rate": 2e-05,
      "loss": 0.1323,
      "step": 1779
    },
    {
      "epoch": 1.2027712064886786,
      "grad_norm": 5.493519306182861,
      "learning_rate": 2e-05,
      "loss": 0.1483,
      "step": 1780
    },
    {
      "epoch": 1.2034471105103075,
      "grad_norm": 6.587663173675537,
      "learning_rate": 2e-05,
      "loss": 0.1478,
      "step": 1781
    },
    {
      "epoch": 1.2041230145319364,
      "grad_norm": 4.865078449249268,
      "learning_rate": 2e-05,
      "loss": 0.2297,
      "step": 1782
    },
    {
      "epoch": 1.2047989185535655,
      "grad_norm": 19.52175521850586,
      "learning_rate": 2e-05,
      "loss": 0.1843,
      "step": 1783
    },
    {
      "epoch": 1.2054748225751943,
      "grad_norm": 8.560564041137695,
      "learning_rate": 2e-05,
      "loss": 0.1847,
      "step": 1784
    },
    {
      "epoch": 1.2061507265968232,
      "grad_norm": 23.19159698486328,
      "learning_rate": 2e-05,
      "loss": 0.3087,
      "step": 1785
    },
    {
      "epoch": 1.206826630618452,
      "grad_norm": 12.84660530090332,
      "learning_rate": 2e-05,
      "loss": 0.1491,
      "step": 1786
    },
    {
      "epoch": 1.2075025346400812,
      "grad_norm": 6.975162506103516,
      "learning_rate": 2e-05,
      "loss": 0.2092,
      "step": 1787
    },
    {
      "epoch": 1.20817843866171,
      "grad_norm": 13.524344444274902,
      "learning_rate": 2e-05,
      "loss": 0.1576,
      "step": 1788
    },
    {
      "epoch": 1.208854342683339,
      "grad_norm": 15.739676475524902,
      "learning_rate": 2e-05,
      "loss": 0.1588,
      "step": 1789
    },
    {
      "epoch": 1.209530246704968,
      "grad_norm": 12.501124382019043,
      "learning_rate": 2e-05,
      "loss": 0.2087,
      "step": 1790
    },
    {
      "epoch": 1.2102061507265969,
      "grad_norm": 5.63452672958374,
      "learning_rate": 2e-05,
      "loss": 0.1822,
      "step": 1791
    },
    {
      "epoch": 1.2108820547482257,
      "grad_norm": 5.861694812774658,
      "learning_rate": 2e-05,
      "loss": 0.2114,
      "step": 1792
    },
    {
      "epoch": 1.2115579587698546,
      "grad_norm": 11.294877052307129,
      "learning_rate": 2e-05,
      "loss": 0.2252,
      "step": 1793
    },
    {
      "epoch": 1.2122338627914835,
      "grad_norm": 12.110435485839844,
      "learning_rate": 2e-05,
      "loss": 0.1818,
      "step": 1794
    },
    {
      "epoch": 1.2129097668131126,
      "grad_norm": 8.030058860778809,
      "learning_rate": 2e-05,
      "loss": 0.2461,
      "step": 1795
    },
    {
      "epoch": 1.2135856708347414,
      "grad_norm": 4.795174598693848,
      "learning_rate": 2e-05,
      "loss": 0.1724,
      "step": 1796
    },
    {
      "epoch": 1.2142615748563703,
      "grad_norm": 4.938699722290039,
      "learning_rate": 2e-05,
      "loss": 0.1818,
      "step": 1797
    },
    {
      "epoch": 1.2149374788779994,
      "grad_norm": 6.906745910644531,
      "learning_rate": 2e-05,
      "loss": 0.193,
      "step": 1798
    },
    {
      "epoch": 1.2156133828996283,
      "grad_norm": 5.3388214111328125,
      "learning_rate": 2e-05,
      "loss": 0.1625,
      "step": 1799
    },
    {
      "epoch": 1.2162892869212572,
      "grad_norm": 8.673335075378418,
      "learning_rate": 2e-05,
      "loss": 0.2284,
      "step": 1800
    },
    {
      "epoch": 1.2162892869212572,
      "eval_accuracy": 0.5300253378378378,
      "eval_loss": 0.25424784421920776,
      "eval_runtime": 49.1495,
      "eval_samples_per_second": 481.795,
      "eval_steps_per_second": 15.056,
      "step": 1800
    },
    {
      "epoch": 1.216965190942886,
      "grad_norm": 6.896942138671875,
      "learning_rate": 2e-05,
      "loss": 0.1862,
      "step": 1801
    },
    {
      "epoch": 1.2176410949645151,
      "grad_norm": 9.191743850708008,
      "learning_rate": 2e-05,
      "loss": 0.1872,
      "step": 1802
    },
    {
      "epoch": 1.218316998986144,
      "grad_norm": 9.363788604736328,
      "learning_rate": 2e-05,
      "loss": 0.1946,
      "step": 1803
    },
    {
      "epoch": 1.2189929030077729,
      "grad_norm": 6.419003486633301,
      "learning_rate": 2e-05,
      "loss": 0.1422,
      "step": 1804
    },
    {
      "epoch": 1.219668807029402,
      "grad_norm": 12.890032768249512,
      "learning_rate": 2e-05,
      "loss": 0.1982,
      "step": 1805
    },
    {
      "epoch": 1.2203447110510308,
      "grad_norm": 5.7455339431762695,
      "learning_rate": 2e-05,
      "loss": 0.2584,
      "step": 1806
    },
    {
      "epoch": 1.2210206150726597,
      "grad_norm": 14.446921348571777,
      "learning_rate": 2e-05,
      "loss": 0.2338,
      "step": 1807
    },
    {
      "epoch": 1.2216965190942886,
      "grad_norm": 16.134380340576172,
      "learning_rate": 2e-05,
      "loss": 0.1656,
      "step": 1808
    },
    {
      "epoch": 1.2223724231159174,
      "grad_norm": 7.192055702209473,
      "learning_rate": 2e-05,
      "loss": 0.1886,
      "step": 1809
    },
    {
      "epoch": 1.2230483271375465,
      "grad_norm": 18.008638381958008,
      "learning_rate": 2e-05,
      "loss": 0.2015,
      "step": 1810
    },
    {
      "epoch": 1.2237242311591754,
      "grad_norm": 4.611701488494873,
      "learning_rate": 2e-05,
      "loss": 0.2341,
      "step": 1811
    },
    {
      "epoch": 1.2244001351808043,
      "grad_norm": 15.221614837646484,
      "learning_rate": 2e-05,
      "loss": 0.1856,
      "step": 1812
    },
    {
      "epoch": 1.2250760392024334,
      "grad_norm": 15.420376777648926,
      "learning_rate": 2e-05,
      "loss": 0.136,
      "step": 1813
    },
    {
      "epoch": 1.2257519432240622,
      "grad_norm": 8.184415817260742,
      "learning_rate": 2e-05,
      "loss": 0.1741,
      "step": 1814
    },
    {
      "epoch": 1.226427847245691,
      "grad_norm": 4.147475719451904,
      "learning_rate": 2e-05,
      "loss": 0.1578,
      "step": 1815
    },
    {
      "epoch": 1.22710375126732,
      "grad_norm": 7.075599193572998,
      "learning_rate": 2e-05,
      "loss": 0.2083,
      "step": 1816
    },
    {
      "epoch": 1.227779655288949,
      "grad_norm": 4.614149570465088,
      "learning_rate": 2e-05,
      "loss": 0.1822,
      "step": 1817
    },
    {
      "epoch": 1.228455559310578,
      "grad_norm": 5.299441814422607,
      "learning_rate": 2e-05,
      "loss": 0.1694,
      "step": 1818
    },
    {
      "epoch": 1.2291314633322068,
      "grad_norm": 6.817848205566406,
      "learning_rate": 2e-05,
      "loss": 0.2067,
      "step": 1819
    },
    {
      "epoch": 1.2298073673538357,
      "grad_norm": 16.010360717773438,
      "learning_rate": 2e-05,
      "loss": 0.2598,
      "step": 1820
    },
    {
      "epoch": 1.2304832713754648,
      "grad_norm": 8.906267166137695,
      "learning_rate": 2e-05,
      "loss": 0.1654,
      "step": 1821
    },
    {
      "epoch": 1.2311591753970936,
      "grad_norm": 7.268871784210205,
      "learning_rate": 2e-05,
      "loss": 0.2197,
      "step": 1822
    },
    {
      "epoch": 1.2318350794187225,
      "grad_norm": 11.505237579345703,
      "learning_rate": 2e-05,
      "loss": 0.1781,
      "step": 1823
    },
    {
      "epoch": 1.2325109834403514,
      "grad_norm": 7.198616981506348,
      "learning_rate": 2e-05,
      "loss": 0.2033,
      "step": 1824
    },
    {
      "epoch": 1.2331868874619805,
      "grad_norm": 5.324962615966797,
      "learning_rate": 2e-05,
      "loss": 0.1373,
      "step": 1825
    },
    {
      "epoch": 1.2338627914836093,
      "grad_norm": 5.00131368637085,
      "learning_rate": 2e-05,
      "loss": 0.1941,
      "step": 1826
    },
    {
      "epoch": 1.2345386955052382,
      "grad_norm": 5.062750339508057,
      "learning_rate": 2e-05,
      "loss": 0.1503,
      "step": 1827
    },
    {
      "epoch": 1.2352145995268673,
      "grad_norm": 4.069732666015625,
      "learning_rate": 2e-05,
      "loss": 0.2009,
      "step": 1828
    },
    {
      "epoch": 1.2358905035484962,
      "grad_norm": 4.796449184417725,
      "learning_rate": 2e-05,
      "loss": 0.1686,
      "step": 1829
    },
    {
      "epoch": 1.236566407570125,
      "grad_norm": 8.326994895935059,
      "learning_rate": 2e-05,
      "loss": 0.1746,
      "step": 1830
    },
    {
      "epoch": 1.237242311591754,
      "grad_norm": 6.5724945068359375,
      "learning_rate": 2e-05,
      "loss": 0.1654,
      "step": 1831
    },
    {
      "epoch": 1.2379182156133828,
      "grad_norm": 5.32286262512207,
      "learning_rate": 2e-05,
      "loss": 0.2766,
      "step": 1832
    },
    {
      "epoch": 1.2385941196350119,
      "grad_norm": 5.573657035827637,
      "learning_rate": 2e-05,
      "loss": 0.1716,
      "step": 1833
    },
    {
      "epoch": 1.2392700236566407,
      "grad_norm": 9.84549617767334,
      "learning_rate": 2e-05,
      "loss": 0.1642,
      "step": 1834
    },
    {
      "epoch": 1.2399459276782696,
      "grad_norm": 4.09318208694458,
      "learning_rate": 2e-05,
      "loss": 0.1731,
      "step": 1835
    },
    {
      "epoch": 1.2406218316998987,
      "grad_norm": 6.5771989822387695,
      "learning_rate": 2e-05,
      "loss": 0.2423,
      "step": 1836
    },
    {
      "epoch": 1.2412977357215276,
      "grad_norm": 5.726827621459961,
      "learning_rate": 2e-05,
      "loss": 0.1668,
      "step": 1837
    },
    {
      "epoch": 1.2419736397431564,
      "grad_norm": 6.370915412902832,
      "learning_rate": 2e-05,
      "loss": 0.1222,
      "step": 1838
    },
    {
      "epoch": 1.2426495437647853,
      "grad_norm": 6.390203952789307,
      "learning_rate": 2e-05,
      "loss": 0.1632,
      "step": 1839
    },
    {
      "epoch": 1.2433254477864144,
      "grad_norm": 3.9887492656707764,
      "learning_rate": 2e-05,
      "loss": 0.1412,
      "step": 1840
    },
    {
      "epoch": 1.2440013518080433,
      "grad_norm": 4.565234184265137,
      "learning_rate": 2e-05,
      "loss": 0.152,
      "step": 1841
    },
    {
      "epoch": 1.2446772558296721,
      "grad_norm": 12.440657615661621,
      "learning_rate": 2e-05,
      "loss": 0.1708,
      "step": 1842
    },
    {
      "epoch": 1.2453531598513012,
      "grad_norm": 6.985007286071777,
      "learning_rate": 2e-05,
      "loss": 0.206,
      "step": 1843
    },
    {
      "epoch": 1.24602906387293,
      "grad_norm": 11.6002779006958,
      "learning_rate": 2e-05,
      "loss": 0.1869,
      "step": 1844
    },
    {
      "epoch": 1.246704967894559,
      "grad_norm": 17.70753288269043,
      "learning_rate": 2e-05,
      "loss": 0.1736,
      "step": 1845
    },
    {
      "epoch": 1.2473808719161878,
      "grad_norm": 7.624517917633057,
      "learning_rate": 2e-05,
      "loss": 0.1467,
      "step": 1846
    },
    {
      "epoch": 1.2480567759378167,
      "grad_norm": 10.224029541015625,
      "learning_rate": 2e-05,
      "loss": 0.1398,
      "step": 1847
    },
    {
      "epoch": 1.2487326799594458,
      "grad_norm": 7.823651313781738,
      "learning_rate": 2e-05,
      "loss": 0.1665,
      "step": 1848
    },
    {
      "epoch": 1.2494085839810747,
      "grad_norm": 13.50979232788086,
      "learning_rate": 2e-05,
      "loss": 0.1941,
      "step": 1849
    },
    {
      "epoch": 1.2500844880027036,
      "grad_norm": 7.744223117828369,
      "learning_rate": 2e-05,
      "loss": 0.2877,
      "step": 1850
    },
    {
      "epoch": 1.2500844880027036,
      "eval_accuracy": 0.5853040540540541,
      "eval_loss": 0.2664169371128082,
      "eval_runtime": 43.2424,
      "eval_samples_per_second": 547.611,
      "eval_steps_per_second": 17.113,
      "step": 1850
    },
    {
      "epoch": 1.2507603920243326,
      "grad_norm": 13.019590377807617,
      "learning_rate": 2e-05,
      "loss": 0.2284,
      "step": 1851
    },
    {
      "epoch": 1.2514362960459615,
      "grad_norm": 8.464447975158691,
      "learning_rate": 2e-05,
      "loss": 0.2434,
      "step": 1852
    },
    {
      "epoch": 1.2521122000675904,
      "grad_norm": 11.460489273071289,
      "learning_rate": 2e-05,
      "loss": 0.172,
      "step": 1853
    },
    {
      "epoch": 1.2527881040892193,
      "grad_norm": 7.235671043395996,
      "learning_rate": 2e-05,
      "loss": 0.139,
      "step": 1854
    },
    {
      "epoch": 1.2534640081108481,
      "grad_norm": 3.8853137493133545,
      "learning_rate": 2e-05,
      "loss": 0.1514,
      "step": 1855
    },
    {
      "epoch": 1.2541399121324772,
      "grad_norm": 5.043623447418213,
      "learning_rate": 2e-05,
      "loss": 0.139,
      "step": 1856
    },
    {
      "epoch": 1.254815816154106,
      "grad_norm": 6.943564414978027,
      "learning_rate": 2e-05,
      "loss": 0.2138,
      "step": 1857
    },
    {
      "epoch": 1.2554917201757352,
      "grad_norm": 5.591352939605713,
      "learning_rate": 2e-05,
      "loss": 0.1957,
      "step": 1858
    },
    {
      "epoch": 1.256167624197364,
      "grad_norm": 11.384085655212402,
      "learning_rate": 2e-05,
      "loss": 0.2266,
      "step": 1859
    },
    {
      "epoch": 1.256843528218993,
      "grad_norm": 11.875199317932129,
      "learning_rate": 2e-05,
      "loss": 0.224,
      "step": 1860
    },
    {
      "epoch": 1.2575194322406218,
      "grad_norm": 13.010139465332031,
      "learning_rate": 2e-05,
      "loss": 0.2005,
      "step": 1861
    },
    {
      "epoch": 1.2581953362622507,
      "grad_norm": 10.41720199584961,
      "learning_rate": 2e-05,
      "loss": 0.1912,
      "step": 1862
    },
    {
      "epoch": 1.2588712402838798,
      "grad_norm": 3.875239372253418,
      "learning_rate": 2e-05,
      "loss": 0.1773,
      "step": 1863
    },
    {
      "epoch": 1.2595471443055086,
      "grad_norm": 7.680438995361328,
      "learning_rate": 2e-05,
      "loss": 0.2225,
      "step": 1864
    },
    {
      "epoch": 1.2602230483271375,
      "grad_norm": 14.317904472351074,
      "learning_rate": 2e-05,
      "loss": 0.1758,
      "step": 1865
    },
    {
      "epoch": 1.2608989523487666,
      "grad_norm": 19.59193229675293,
      "learning_rate": 2e-05,
      "loss": 0.1646,
      "step": 1866
    },
    {
      "epoch": 1.2615748563703955,
      "grad_norm": 9.381613731384277,
      "learning_rate": 2e-05,
      "loss": 0.1483,
      "step": 1867
    },
    {
      "epoch": 1.2622507603920243,
      "grad_norm": 4.540507793426514,
      "learning_rate": 2e-05,
      "loss": 0.2291,
      "step": 1868
    },
    {
      "epoch": 1.2629266644136532,
      "grad_norm": 7.667973518371582,
      "learning_rate": 2e-05,
      "loss": 0.2307,
      "step": 1869
    },
    {
      "epoch": 1.263602568435282,
      "grad_norm": 8.528095245361328,
      "learning_rate": 2e-05,
      "loss": 0.2322,
      "step": 1870
    },
    {
      "epoch": 1.2642784724569112,
      "grad_norm": 3.984182596206665,
      "learning_rate": 2e-05,
      "loss": 0.1797,
      "step": 1871
    },
    {
      "epoch": 1.26495437647854,
      "grad_norm": 16.407001495361328,
      "learning_rate": 2e-05,
      "loss": 0.2027,
      "step": 1872
    },
    {
      "epoch": 1.265630280500169,
      "grad_norm": 19.284700393676758,
      "learning_rate": 2e-05,
      "loss": 0.2281,
      "step": 1873
    },
    {
      "epoch": 1.266306184521798,
      "grad_norm": 6.748196601867676,
      "learning_rate": 2e-05,
      "loss": 0.2568,
      "step": 1874
    },
    {
      "epoch": 1.2669820885434269,
      "grad_norm": 12.140364646911621,
      "learning_rate": 2e-05,
      "loss": 0.2031,
      "step": 1875
    },
    {
      "epoch": 1.2676579925650557,
      "grad_norm": 4.566565990447998,
      "learning_rate": 2e-05,
      "loss": 0.1789,
      "step": 1876
    },
    {
      "epoch": 1.2683338965866846,
      "grad_norm": 6.986762046813965,
      "learning_rate": 2e-05,
      "loss": 0.1514,
      "step": 1877
    },
    {
      "epoch": 1.2690098006083137,
      "grad_norm": 6.701267242431641,
      "learning_rate": 2e-05,
      "loss": 0.2491,
      "step": 1878
    },
    {
      "epoch": 1.2696857046299426,
      "grad_norm": 23.49757194519043,
      "learning_rate": 2e-05,
      "loss": 0.1727,
      "step": 1879
    },
    {
      "epoch": 1.2703616086515714,
      "grad_norm": 6.547677993774414,
      "learning_rate": 2e-05,
      "loss": 0.2402,
      "step": 1880
    },
    {
      "epoch": 1.2710375126732005,
      "grad_norm": 6.216747283935547,
      "learning_rate": 2e-05,
      "loss": 0.1779,
      "step": 1881
    },
    {
      "epoch": 1.2717134166948294,
      "grad_norm": 5.455733299255371,
      "learning_rate": 2e-05,
      "loss": 0.242,
      "step": 1882
    },
    {
      "epoch": 1.2723893207164583,
      "grad_norm": 15.995955467224121,
      "learning_rate": 2e-05,
      "loss": 0.2425,
      "step": 1883
    },
    {
      "epoch": 1.2730652247380871,
      "grad_norm": 13.078364372253418,
      "learning_rate": 2e-05,
      "loss": 0.3063,
      "step": 1884
    },
    {
      "epoch": 1.273741128759716,
      "grad_norm": 13.452291488647461,
      "learning_rate": 2e-05,
      "loss": 0.1823,
      "step": 1885
    },
    {
      "epoch": 1.274417032781345,
      "grad_norm": 6.481709957122803,
      "learning_rate": 2e-05,
      "loss": 0.167,
      "step": 1886
    },
    {
      "epoch": 1.275092936802974,
      "grad_norm": 3.967113971710205,
      "learning_rate": 2e-05,
      "loss": 0.1852,
      "step": 1887
    },
    {
      "epoch": 1.2757688408246028,
      "grad_norm": 4.7107834815979,
      "learning_rate": 2e-05,
      "loss": 0.2079,
      "step": 1888
    },
    {
      "epoch": 1.276444744846232,
      "grad_norm": 6.4967360496521,
      "learning_rate": 2e-05,
      "loss": 0.1702,
      "step": 1889
    },
    {
      "epoch": 1.2771206488678608,
      "grad_norm": 10.501167297363281,
      "learning_rate": 2e-05,
      "loss": 0.196,
      "step": 1890
    },
    {
      "epoch": 1.2777965528894897,
      "grad_norm": 8.474264144897461,
      "learning_rate": 2e-05,
      "loss": 0.186,
      "step": 1891
    },
    {
      "epoch": 1.2784724569111185,
      "grad_norm": 13.575934410095215,
      "learning_rate": 2e-05,
      "loss": 0.1988,
      "step": 1892
    },
    {
      "epoch": 1.2791483609327474,
      "grad_norm": 6.229480743408203,
      "learning_rate": 2e-05,
      "loss": 0.187,
      "step": 1893
    },
    {
      "epoch": 1.2798242649543765,
      "grad_norm": 11.305931091308594,
      "learning_rate": 2e-05,
      "loss": 0.1575,
      "step": 1894
    },
    {
      "epoch": 1.2805001689760054,
      "grad_norm": 9.88237476348877,
      "learning_rate": 2e-05,
      "loss": 0.1948,
      "step": 1895
    },
    {
      "epoch": 1.2811760729976345,
      "grad_norm": 7.044897079467773,
      "learning_rate": 2e-05,
      "loss": 0.2434,
      "step": 1896
    },
    {
      "epoch": 1.2818519770192633,
      "grad_norm": 10.417536735534668,
      "learning_rate": 2e-05,
      "loss": 0.1592,
      "step": 1897
    },
    {
      "epoch": 1.2825278810408922,
      "grad_norm": 5.4643449783325195,
      "learning_rate": 2e-05,
      "loss": 0.199,
      "step": 1898
    },
    {
      "epoch": 1.283203785062521,
      "grad_norm": 12.606752395629883,
      "learning_rate": 2e-05,
      "loss": 0.2653,
      "step": 1899
    },
    {
      "epoch": 1.28387968908415,
      "grad_norm": 21.580896377563477,
      "learning_rate": 2e-05,
      "loss": 0.1738,
      "step": 1900
    },
    {
      "epoch": 1.28387968908415,
      "eval_accuracy": 0.550464527027027,
      "eval_loss": 0.24923856556415558,
      "eval_runtime": 43.3397,
      "eval_samples_per_second": 546.381,
      "eval_steps_per_second": 17.074,
      "step": 1900
    },
    {
      "epoch": 1.284555593105779,
      "grad_norm": 9.045686721801758,
      "learning_rate": 2e-05,
      "loss": 0.2363,
      "step": 1901
    },
    {
      "epoch": 1.285231497127408,
      "grad_norm": 6.260179042816162,
      "learning_rate": 2e-05,
      "loss": 0.1491,
      "step": 1902
    },
    {
      "epoch": 1.2859074011490368,
      "grad_norm": 12.239228248596191,
      "learning_rate": 2e-05,
      "loss": 0.1544,
      "step": 1903
    },
    {
      "epoch": 1.2865833051706659,
      "grad_norm": 21.567676544189453,
      "learning_rate": 2e-05,
      "loss": 0.2484,
      "step": 1904
    },
    {
      "epoch": 1.2872592091922948,
      "grad_norm": 5.521478176116943,
      "learning_rate": 2e-05,
      "loss": 0.1444,
      "step": 1905
    },
    {
      "epoch": 1.2879351132139236,
      "grad_norm": 4.39363431930542,
      "learning_rate": 2e-05,
      "loss": 0.1974,
      "step": 1906
    },
    {
      "epoch": 1.2886110172355525,
      "grad_norm": 7.443763256072998,
      "learning_rate": 2e-05,
      "loss": 0.1734,
      "step": 1907
    },
    {
      "epoch": 1.2892869212571814,
      "grad_norm": 9.567221641540527,
      "learning_rate": 2e-05,
      "loss": 0.1747,
      "step": 1908
    },
    {
      "epoch": 1.2899628252788105,
      "grad_norm": 4.605379104614258,
      "learning_rate": 2e-05,
      "loss": 0.1881,
      "step": 1909
    },
    {
      "epoch": 1.2906387293004393,
      "grad_norm": 6.778324127197266,
      "learning_rate": 2e-05,
      "loss": 0.2072,
      "step": 1910
    },
    {
      "epoch": 1.2913146333220682,
      "grad_norm": 5.826206207275391,
      "learning_rate": 2e-05,
      "loss": 0.1453,
      "step": 1911
    },
    {
      "epoch": 1.2919905373436973,
      "grad_norm": 8.084344863891602,
      "learning_rate": 2e-05,
      "loss": 0.1873,
      "step": 1912
    },
    {
      "epoch": 1.2926664413653262,
      "grad_norm": 5.409360885620117,
      "learning_rate": 2e-05,
      "loss": 0.1763,
      "step": 1913
    },
    {
      "epoch": 1.293342345386955,
      "grad_norm": 4.929358005523682,
      "learning_rate": 2e-05,
      "loss": 0.1337,
      "step": 1914
    },
    {
      "epoch": 1.294018249408584,
      "grad_norm": 7.059220790863037,
      "learning_rate": 2e-05,
      "loss": 0.1903,
      "step": 1915
    },
    {
      "epoch": 1.294694153430213,
      "grad_norm": 13.921784400939941,
      "learning_rate": 2e-05,
      "loss": 0.1659,
      "step": 1916
    },
    {
      "epoch": 1.2953700574518419,
      "grad_norm": 5.5489277839660645,
      "learning_rate": 2e-05,
      "loss": 0.2381,
      "step": 1917
    },
    {
      "epoch": 1.2960459614734707,
      "grad_norm": 5.366847038269043,
      "learning_rate": 2e-05,
      "loss": 0.134,
      "step": 1918
    },
    {
      "epoch": 1.2967218654950998,
      "grad_norm": 5.960186958312988,
      "learning_rate": 2e-05,
      "loss": 0.1536,
      "step": 1919
    },
    {
      "epoch": 1.2973977695167287,
      "grad_norm": 4.907669544219971,
      "learning_rate": 2e-05,
      "loss": 0.1811,
      "step": 1920
    },
    {
      "epoch": 1.2980736735383576,
      "grad_norm": 8.940302848815918,
      "learning_rate": 2e-05,
      "loss": 0.1881,
      "step": 1921
    },
    {
      "epoch": 1.2987495775599864,
      "grad_norm": 6.0679473876953125,
      "learning_rate": 2e-05,
      "loss": 0.204,
      "step": 1922
    },
    {
      "epoch": 1.2994254815816153,
      "grad_norm": 5.588442325592041,
      "learning_rate": 2e-05,
      "loss": 0.1566,
      "step": 1923
    },
    {
      "epoch": 1.3001013856032444,
      "grad_norm": 6.246606826782227,
      "learning_rate": 2e-05,
      "loss": 0.1713,
      "step": 1924
    },
    {
      "epoch": 1.3007772896248733,
      "grad_norm": 7.371962547302246,
      "learning_rate": 2e-05,
      "loss": 0.1987,
      "step": 1925
    },
    {
      "epoch": 1.3014531936465021,
      "grad_norm": 8.71627426147461,
      "learning_rate": 2e-05,
      "loss": 0.179,
      "step": 1926
    },
    {
      "epoch": 1.3021290976681312,
      "grad_norm": 6.579427242279053,
      "learning_rate": 2e-05,
      "loss": 0.2073,
      "step": 1927
    },
    {
      "epoch": 1.30280500168976,
      "grad_norm": 4.729877471923828,
      "learning_rate": 2e-05,
      "loss": 0.1381,
      "step": 1928
    },
    {
      "epoch": 1.303480905711389,
      "grad_norm": 7.449542999267578,
      "learning_rate": 2e-05,
      "loss": 0.214,
      "step": 1929
    },
    {
      "epoch": 1.3041568097330178,
      "grad_norm": 8.971335411071777,
      "learning_rate": 2e-05,
      "loss": 0.2476,
      "step": 1930
    },
    {
      "epoch": 1.3048327137546467,
      "grad_norm": 4.638245582580566,
      "learning_rate": 2e-05,
      "loss": 0.195,
      "step": 1931
    },
    {
      "epoch": 1.3055086177762758,
      "grad_norm": 18.235916137695312,
      "learning_rate": 2e-05,
      "loss": 0.2292,
      "step": 1932
    },
    {
      "epoch": 1.3061845217979047,
      "grad_norm": 11.621353149414062,
      "learning_rate": 2e-05,
      "loss": 0.1509,
      "step": 1933
    },
    {
      "epoch": 1.3068604258195335,
      "grad_norm": 7.163653373718262,
      "learning_rate": 2e-05,
      "loss": 0.255,
      "step": 1934
    },
    {
      "epoch": 1.3075363298411626,
      "grad_norm": 11.058306694030762,
      "learning_rate": 2e-05,
      "loss": 0.1789,
      "step": 1935
    },
    {
      "epoch": 1.3082122338627915,
      "grad_norm": 4.688571453094482,
      "learning_rate": 2e-05,
      "loss": 0.1932,
      "step": 1936
    },
    {
      "epoch": 1.3088881378844204,
      "grad_norm": 9.821659088134766,
      "learning_rate": 2e-05,
      "loss": 0.1757,
      "step": 1937
    },
    {
      "epoch": 1.3095640419060492,
      "grad_norm": 13.18177318572998,
      "learning_rate": 2e-05,
      "loss": 0.1842,
      "step": 1938
    },
    {
      "epoch": 1.3102399459276783,
      "grad_norm": 9.797233581542969,
      "learning_rate": 2e-05,
      "loss": 0.1706,
      "step": 1939
    },
    {
      "epoch": 1.3109158499493072,
      "grad_norm": 4.413486480712891,
      "learning_rate": 2e-05,
      "loss": 0.2335,
      "step": 1940
    },
    {
      "epoch": 1.311591753970936,
      "grad_norm": 5.323004245758057,
      "learning_rate": 2e-05,
      "loss": 0.1862,
      "step": 1941
    },
    {
      "epoch": 1.3122676579925652,
      "grad_norm": 6.212184906005859,
      "learning_rate": 2e-05,
      "loss": 0.2289,
      "step": 1942
    },
    {
      "epoch": 1.312943562014194,
      "grad_norm": 5.104065418243408,
      "learning_rate": 2e-05,
      "loss": 0.1935,
      "step": 1943
    },
    {
      "epoch": 1.313619466035823,
      "grad_norm": 4.474874019622803,
      "learning_rate": 2e-05,
      "loss": 0.1468,
      "step": 1944
    },
    {
      "epoch": 1.3142953700574518,
      "grad_norm": 5.817380905151367,
      "learning_rate": 2e-05,
      "loss": 0.1936,
      "step": 1945
    },
    {
      "epoch": 1.3149712740790807,
      "grad_norm": 12.884904861450195,
      "learning_rate": 2e-05,
      "loss": 0.1934,
      "step": 1946
    },
    {
      "epoch": 1.3156471781007097,
      "grad_norm": 10.898008346557617,
      "learning_rate": 2e-05,
      "loss": 0.176,
      "step": 1947
    },
    {
      "epoch": 1.3163230821223386,
      "grad_norm": 5.706287384033203,
      "learning_rate": 2e-05,
      "loss": 0.1718,
      "step": 1948
    },
    {
      "epoch": 1.3169989861439675,
      "grad_norm": 4.690310955047607,
      "learning_rate": 2e-05,
      "loss": 0.1866,
      "step": 1949
    },
    {
      "epoch": 1.3176748901655966,
      "grad_norm": 6.478024005889893,
      "learning_rate": 2e-05,
      "loss": 0.2275,
      "step": 1950
    },
    {
      "epoch": 1.3176748901655966,
      "eval_accuracy": 0.5481418918918919,
      "eval_loss": 0.26147350668907166,
      "eval_runtime": 43.717,
      "eval_samples_per_second": 541.665,
      "eval_steps_per_second": 16.927,
      "step": 1950
    },
    {
      "epoch": 1.3183507941872254,
      "grad_norm": 13.045730590820312,
      "learning_rate": 2e-05,
      "loss": 0.1642,
      "step": 1951
    },
    {
      "epoch": 1.3190266982088543,
      "grad_norm": 9.647071838378906,
      "learning_rate": 2e-05,
      "loss": 0.119,
      "step": 1952
    },
    {
      "epoch": 1.3197026022304832,
      "grad_norm": 6.452217102050781,
      "learning_rate": 2e-05,
      "loss": 0.2778,
      "step": 1953
    },
    {
      "epoch": 1.3203785062521123,
      "grad_norm": 8.547514915466309,
      "learning_rate": 2e-05,
      "loss": 0.178,
      "step": 1954
    },
    {
      "epoch": 1.3210544102737412,
      "grad_norm": 7.549161434173584,
      "learning_rate": 2e-05,
      "loss": 0.1879,
      "step": 1955
    },
    {
      "epoch": 1.32173031429537,
      "grad_norm": 8.261528015136719,
      "learning_rate": 2e-05,
      "loss": 0.1907,
      "step": 1956
    },
    {
      "epoch": 1.3224062183169991,
      "grad_norm": 7.522820949554443,
      "learning_rate": 2e-05,
      "loss": 0.2374,
      "step": 1957
    },
    {
      "epoch": 1.323082122338628,
      "grad_norm": 7.375333309173584,
      "learning_rate": 2e-05,
      "loss": 0.1618,
      "step": 1958
    },
    {
      "epoch": 1.3237580263602569,
      "grad_norm": 7.6951093673706055,
      "learning_rate": 2e-05,
      "loss": 0.1666,
      "step": 1959
    },
    {
      "epoch": 1.3244339303818857,
      "grad_norm": 16.580968856811523,
      "learning_rate": 2e-05,
      "loss": 0.2118,
      "step": 1960
    },
    {
      "epoch": 1.3251098344035146,
      "grad_norm": 5.297538757324219,
      "learning_rate": 2e-05,
      "loss": 0.1507,
      "step": 1961
    },
    {
      "epoch": 1.3257857384251437,
      "grad_norm": 8.640284538269043,
      "learning_rate": 2e-05,
      "loss": 0.1506,
      "step": 1962
    },
    {
      "epoch": 1.3264616424467726,
      "grad_norm": 5.820493221282959,
      "learning_rate": 2e-05,
      "loss": 0.232,
      "step": 1963
    },
    {
      "epoch": 1.3271375464684014,
      "grad_norm": 5.430056095123291,
      "learning_rate": 2e-05,
      "loss": 0.1956,
      "step": 1964
    },
    {
      "epoch": 1.3278134504900305,
      "grad_norm": 5.763941287994385,
      "learning_rate": 2e-05,
      "loss": 0.1333,
      "step": 1965
    },
    {
      "epoch": 1.3284893545116594,
      "grad_norm": 10.531974792480469,
      "learning_rate": 2e-05,
      "loss": 0.183,
      "step": 1966
    },
    {
      "epoch": 1.3291652585332883,
      "grad_norm": 7.465350151062012,
      "learning_rate": 2e-05,
      "loss": 0.1786,
      "step": 1967
    },
    {
      "epoch": 1.3298411625549171,
      "grad_norm": 5.618406772613525,
      "learning_rate": 2e-05,
      "loss": 0.1346,
      "step": 1968
    },
    {
      "epoch": 1.330517066576546,
      "grad_norm": 6.605738162994385,
      "learning_rate": 2e-05,
      "loss": 0.1978,
      "step": 1969
    },
    {
      "epoch": 1.331192970598175,
      "grad_norm": 8.848418235778809,
      "learning_rate": 2e-05,
      "loss": 0.1758,
      "step": 1970
    },
    {
      "epoch": 1.331868874619804,
      "grad_norm": 5.835309028625488,
      "learning_rate": 2e-05,
      "loss": 0.1999,
      "step": 1971
    },
    {
      "epoch": 1.3325447786414328,
      "grad_norm": 10.541238784790039,
      "learning_rate": 2e-05,
      "loss": 0.2403,
      "step": 1972
    },
    {
      "epoch": 1.333220682663062,
      "grad_norm": 14.38760757446289,
      "learning_rate": 2e-05,
      "loss": 0.1688,
      "step": 1973
    },
    {
      "epoch": 1.3338965866846908,
      "grad_norm": 10.2054443359375,
      "learning_rate": 2e-05,
      "loss": 0.1451,
      "step": 1974
    },
    {
      "epoch": 1.3345724907063197,
      "grad_norm": 7.7367167472839355,
      "learning_rate": 2e-05,
      "loss": 0.2498,
      "step": 1975
    },
    {
      "epoch": 1.3352483947279485,
      "grad_norm": 9.478334426879883,
      "learning_rate": 2e-05,
      "loss": 0.1537,
      "step": 1976
    },
    {
      "epoch": 1.3359242987495776,
      "grad_norm": 3.9714813232421875,
      "learning_rate": 2e-05,
      "loss": 0.1239,
      "step": 1977
    },
    {
      "epoch": 1.3366002027712065,
      "grad_norm": 8.5205659866333,
      "learning_rate": 2e-05,
      "loss": 0.196,
      "step": 1978
    },
    {
      "epoch": 1.3372761067928354,
      "grad_norm": 10.215191841125488,
      "learning_rate": 2e-05,
      "loss": 0.1799,
      "step": 1979
    },
    {
      "epoch": 1.3379520108144645,
      "grad_norm": 6.6132330894470215,
      "learning_rate": 2e-05,
      "loss": 0.1592,
      "step": 1980
    },
    {
      "epoch": 1.3386279148360933,
      "grad_norm": 5.72751522064209,
      "learning_rate": 2e-05,
      "loss": 0.2062,
      "step": 1981
    },
    {
      "epoch": 1.3393038188577222,
      "grad_norm": 11.426005363464355,
      "learning_rate": 2e-05,
      "loss": 0.2442,
      "step": 1982
    },
    {
      "epoch": 1.339979722879351,
      "grad_norm": 10.031704902648926,
      "learning_rate": 2e-05,
      "loss": 0.2608,
      "step": 1983
    },
    {
      "epoch": 1.34065562690098,
      "grad_norm": 11.50489330291748,
      "learning_rate": 2e-05,
      "loss": 0.1698,
      "step": 1984
    },
    {
      "epoch": 1.341331530922609,
      "grad_norm": 5.350318431854248,
      "learning_rate": 2e-05,
      "loss": 0.1675,
      "step": 1985
    },
    {
      "epoch": 1.342007434944238,
      "grad_norm": 11.4668550491333,
      "learning_rate": 2e-05,
      "loss": 0.1727,
      "step": 1986
    },
    {
      "epoch": 1.3426833389658668,
      "grad_norm": 5.068203449249268,
      "learning_rate": 2e-05,
      "loss": 0.2493,
      "step": 1987
    },
    {
      "epoch": 1.3433592429874959,
      "grad_norm": 6.206216335296631,
      "learning_rate": 2e-05,
      "loss": 0.2206,
      "step": 1988
    },
    {
      "epoch": 1.3440351470091247,
      "grad_norm": 4.705418586730957,
      "learning_rate": 2e-05,
      "loss": 0.1689,
      "step": 1989
    },
    {
      "epoch": 1.3447110510307536,
      "grad_norm": 7.866369247436523,
      "learning_rate": 2e-05,
      "loss": 0.219,
      "step": 1990
    },
    {
      "epoch": 1.3453869550523825,
      "grad_norm": 10.211504936218262,
      "learning_rate": 2e-05,
      "loss": 0.1766,
      "step": 1991
    },
    {
      "epoch": 1.3460628590740114,
      "grad_norm": 5.499320983886719,
      "learning_rate": 2e-05,
      "loss": 0.2346,
      "step": 1992
    },
    {
      "epoch": 1.3467387630956404,
      "grad_norm": 9.570453643798828,
      "learning_rate": 2e-05,
      "loss": 0.1948,
      "step": 1993
    },
    {
      "epoch": 1.3474146671172693,
      "grad_norm": 4.686243534088135,
      "learning_rate": 2e-05,
      "loss": 0.147,
      "step": 1994
    },
    {
      "epoch": 1.3480905711388984,
      "grad_norm": 6.592576026916504,
      "learning_rate": 2e-05,
      "loss": 0.2175,
      "step": 1995
    },
    {
      "epoch": 1.3487664751605273,
      "grad_norm": 6.003763198852539,
      "learning_rate": 2e-05,
      "loss": 0.2426,
      "step": 1996
    },
    {
      "epoch": 1.3494423791821561,
      "grad_norm": 17.966516494750977,
      "learning_rate": 2e-05,
      "loss": 0.2187,
      "step": 1997
    },
    {
      "epoch": 1.350118283203785,
      "grad_norm": 6.132434368133545,
      "learning_rate": 2e-05,
      "loss": 0.1831,
      "step": 1998
    },
    {
      "epoch": 1.350794187225414,
      "grad_norm": 12.548261642456055,
      "learning_rate": 2e-05,
      "loss": 0.1687,
      "step": 1999
    },
    {
      "epoch": 1.351470091247043,
      "grad_norm": 6.947999954223633,
      "learning_rate": 2e-05,
      "loss": 0.1671,
      "step": 2000
    },
    {
      "epoch": 1.351470091247043,
      "eval_accuracy": 0.5547297297297298,
      "eval_loss": 0.2423630654811859,
      "eval_runtime": 44.5408,
      "eval_samples_per_second": 531.647,
      "eval_steps_per_second": 16.614,
      "step": 2000
    },
    {
      "epoch": 1.3521459952686719,
      "grad_norm": 4.164329528808594,
      "learning_rate": 2e-05,
      "loss": 0.232,
      "step": 2001
    },
    {
      "epoch": 1.3528218992903007,
      "grad_norm": 14.433164596557617,
      "learning_rate": 2e-05,
      "loss": 0.2002,
      "step": 2002
    },
    {
      "epoch": 1.3534978033119298,
      "grad_norm": 5.5497331619262695,
      "learning_rate": 2e-05,
      "loss": 0.1611,
      "step": 2003
    },
    {
      "epoch": 1.3541737073335587,
      "grad_norm": 4.189729690551758,
      "learning_rate": 2e-05,
      "loss": 0.1876,
      "step": 2004
    },
    {
      "epoch": 1.3548496113551876,
      "grad_norm": 5.174462795257568,
      "learning_rate": 2e-05,
      "loss": 0.2273,
      "step": 2005
    },
    {
      "epoch": 1.3555255153768164,
      "grad_norm": 13.285398483276367,
      "learning_rate": 2e-05,
      "loss": 0.1748,
      "step": 2006
    },
    {
      "epoch": 1.3562014193984453,
      "grad_norm": 6.065054416656494,
      "learning_rate": 2e-05,
      "loss": 0.1745,
      "step": 2007
    },
    {
      "epoch": 1.3568773234200744,
      "grad_norm": 5.3203606605529785,
      "learning_rate": 2e-05,
      "loss": 0.1373,
      "step": 2008
    },
    {
      "epoch": 1.3575532274417033,
      "grad_norm": 10.601472854614258,
      "learning_rate": 2e-05,
      "loss": 0.2139,
      "step": 2009
    },
    {
      "epoch": 1.3582291314633321,
      "grad_norm": 6.4249773025512695,
      "learning_rate": 2e-05,
      "loss": 0.2054,
      "step": 2010
    },
    {
      "epoch": 1.3589050354849612,
      "grad_norm": 9.552762985229492,
      "learning_rate": 2e-05,
      "loss": 0.1741,
      "step": 2011
    },
    {
      "epoch": 1.35958093950659,
      "grad_norm": 5.87086296081543,
      "learning_rate": 2e-05,
      "loss": 0.1678,
      "step": 2012
    },
    {
      "epoch": 1.360256843528219,
      "grad_norm": 11.474069595336914,
      "learning_rate": 2e-05,
      "loss": 0.2592,
      "step": 2013
    },
    {
      "epoch": 1.3609327475498478,
      "grad_norm": 14.914764404296875,
      "learning_rate": 2e-05,
      "loss": 0.1189,
      "step": 2014
    },
    {
      "epoch": 1.361608651571477,
      "grad_norm": 4.724549770355225,
      "learning_rate": 2e-05,
      "loss": 0.2166,
      "step": 2015
    },
    {
      "epoch": 1.3622845555931058,
      "grad_norm": 7.682941436767578,
      "learning_rate": 2e-05,
      "loss": 0.1899,
      "step": 2016
    },
    {
      "epoch": 1.3629604596147347,
      "grad_norm": 5.496298789978027,
      "learning_rate": 2e-05,
      "loss": 0.1776,
      "step": 2017
    },
    {
      "epoch": 1.3636363636363638,
      "grad_norm": 11.909602165222168,
      "learning_rate": 2e-05,
      "loss": 0.1799,
      "step": 2018
    },
    {
      "epoch": 1.3643122676579926,
      "grad_norm": 14.16934585571289,
      "learning_rate": 2e-05,
      "loss": 0.1907,
      "step": 2019
    },
    {
      "epoch": 1.3649881716796215,
      "grad_norm": 4.205503940582275,
      "learning_rate": 2e-05,
      "loss": 0.2632,
      "step": 2020
    },
    {
      "epoch": 1.3656640757012504,
      "grad_norm": 5.961310386657715,
      "learning_rate": 2e-05,
      "loss": 0.1372,
      "step": 2021
    },
    {
      "epoch": 1.3663399797228792,
      "grad_norm": 7.347164154052734,
      "learning_rate": 2e-05,
      "loss": 0.1907,
      "step": 2022
    },
    {
      "epoch": 1.3670158837445083,
      "grad_norm": 4.92029333114624,
      "learning_rate": 2e-05,
      "loss": 0.2069,
      "step": 2023
    },
    {
      "epoch": 1.3676917877661372,
      "grad_norm": 5.059731483459473,
      "learning_rate": 2e-05,
      "loss": 0.1598,
      "step": 2024
    },
    {
      "epoch": 1.368367691787766,
      "grad_norm": 7.631248474121094,
      "learning_rate": 2e-05,
      "loss": 0.2067,
      "step": 2025
    },
    {
      "epoch": 1.3690435958093952,
      "grad_norm": 9.32215690612793,
      "learning_rate": 2e-05,
      "loss": 0.1767,
      "step": 2026
    },
    {
      "epoch": 1.369719499831024,
      "grad_norm": 4.218494892120361,
      "learning_rate": 2e-05,
      "loss": 0.1712,
      "step": 2027
    },
    {
      "epoch": 1.370395403852653,
      "grad_norm": 7.304398536682129,
      "learning_rate": 2e-05,
      "loss": 0.219,
      "step": 2028
    },
    {
      "epoch": 1.3710713078742818,
      "grad_norm": 3.607694387435913,
      "learning_rate": 2e-05,
      "loss": 0.1908,
      "step": 2029
    },
    {
      "epoch": 1.3717472118959106,
      "grad_norm": 12.167262077331543,
      "learning_rate": 2e-05,
      "loss": 0.2368,
      "step": 2030
    },
    {
      "epoch": 1.3724231159175397,
      "grad_norm": 5.9942097663879395,
      "learning_rate": 2e-05,
      "loss": 0.1805,
      "step": 2031
    },
    {
      "epoch": 1.3730990199391686,
      "grad_norm": 4.767143249511719,
      "learning_rate": 2e-05,
      "loss": 0.1605,
      "step": 2032
    },
    {
      "epoch": 1.3737749239607977,
      "grad_norm": 4.129868030548096,
      "learning_rate": 2e-05,
      "loss": 0.1975,
      "step": 2033
    },
    {
      "epoch": 1.3744508279824266,
      "grad_norm": 5.324492454528809,
      "learning_rate": 2e-05,
      "loss": 0.2099,
      "step": 2034
    },
    {
      "epoch": 1.3751267320040554,
      "grad_norm": 5.252218723297119,
      "learning_rate": 2e-05,
      "loss": 0.1894,
      "step": 2035
    },
    {
      "epoch": 1.3758026360256843,
      "grad_norm": 7.192310333251953,
      "learning_rate": 2e-05,
      "loss": 0.1841,
      "step": 2036
    },
    {
      "epoch": 1.3764785400473132,
      "grad_norm": 8.263927459716797,
      "learning_rate": 2e-05,
      "loss": 0.1891,
      "step": 2037
    },
    {
      "epoch": 1.3771544440689423,
      "grad_norm": 3.9769811630249023,
      "learning_rate": 2e-05,
      "loss": 0.1893,
      "step": 2038
    },
    {
      "epoch": 1.3778303480905711,
      "grad_norm": 6.595765590667725,
      "learning_rate": 2e-05,
      "loss": 0.2076,
      "step": 2039
    },
    {
      "epoch": 1.3785062521122,
      "grad_norm": 12.31153678894043,
      "learning_rate": 2e-05,
      "loss": 0.1589,
      "step": 2040
    },
    {
      "epoch": 1.379182156133829,
      "grad_norm": 9.353832244873047,
      "learning_rate": 2e-05,
      "loss": 0.1743,
      "step": 2041
    },
    {
      "epoch": 1.379858060155458,
      "grad_norm": 4.911788463592529,
      "learning_rate": 2e-05,
      "loss": 0.198,
      "step": 2042
    },
    {
      "epoch": 1.3805339641770868,
      "grad_norm": 4.23148250579834,
      "learning_rate": 2e-05,
      "loss": 0.1969,
      "step": 2043
    },
    {
      "epoch": 1.3812098681987157,
      "grad_norm": 13.838946342468262,
      "learning_rate": 2e-05,
      "loss": 0.2086,
      "step": 2044
    },
    {
      "epoch": 1.3818857722203446,
      "grad_norm": 5.690454006195068,
      "learning_rate": 2e-05,
      "loss": 0.1838,
      "step": 2045
    },
    {
      "epoch": 1.3825616762419737,
      "grad_norm": 9.441311836242676,
      "learning_rate": 2e-05,
      "loss": 0.1563,
      "step": 2046
    },
    {
      "epoch": 1.3832375802636026,
      "grad_norm": 5.681439399719238,
      "learning_rate": 2e-05,
      "loss": 0.211,
      "step": 2047
    },
    {
      "epoch": 1.3839134842852314,
      "grad_norm": 5.774424076080322,
      "learning_rate": 2e-05,
      "loss": 0.1916,
      "step": 2048
    },
    {
      "epoch": 1.3845893883068605,
      "grad_norm": 8.960269927978516,
      "learning_rate": 2e-05,
      "loss": 0.2235,
      "step": 2049
    },
    {
      "epoch": 1.3852652923284894,
      "grad_norm": 8.997650146484375,
      "learning_rate": 2e-05,
      "loss": 0.2495,
      "step": 2050
    },
    {
      "epoch": 1.3852652923284894,
      "eval_accuracy": 0.5834881756756757,
      "eval_loss": 0.2470819056034088,
      "eval_runtime": 42.907,
      "eval_samples_per_second": 551.892,
      "eval_steps_per_second": 17.247,
      "step": 2050
    },
    {
      "epoch": 1.3859411963501183,
      "grad_norm": 12.562325477600098,
      "learning_rate": 2e-05,
      "loss": 0.1485,
      "step": 2051
    },
    {
      "epoch": 1.3866171003717471,
      "grad_norm": 5.126410961151123,
      "learning_rate": 2e-05,
      "loss": 0.1514,
      "step": 2052
    },
    {
      "epoch": 1.3872930043933762,
      "grad_norm": 4.096684455871582,
      "learning_rate": 2e-05,
      "loss": 0.1555,
      "step": 2053
    },
    {
      "epoch": 1.387968908415005,
      "grad_norm": 6.777998447418213,
      "learning_rate": 2e-05,
      "loss": 0.1918,
      "step": 2054
    },
    {
      "epoch": 1.388644812436634,
      "grad_norm": 12.364429473876953,
      "learning_rate": 2e-05,
      "loss": 0.1863,
      "step": 2055
    },
    {
      "epoch": 1.389320716458263,
      "grad_norm": 13.178617477416992,
      "learning_rate": 2e-05,
      "loss": 0.1607,
      "step": 2056
    },
    {
      "epoch": 1.389996620479892,
      "grad_norm": 5.161850929260254,
      "learning_rate": 2e-05,
      "loss": 0.2065,
      "step": 2057
    },
    {
      "epoch": 1.3906725245015208,
      "grad_norm": 5.752294540405273,
      "learning_rate": 2e-05,
      "loss": 0.1438,
      "step": 2058
    },
    {
      "epoch": 1.3913484285231497,
      "grad_norm": 8.796151161193848,
      "learning_rate": 2e-05,
      "loss": 0.1769,
      "step": 2059
    },
    {
      "epoch": 1.3920243325447785,
      "grad_norm": 6.317341327667236,
      "learning_rate": 2e-05,
      "loss": 0.2133,
      "step": 2060
    },
    {
      "epoch": 1.3927002365664076,
      "grad_norm": 13.05630874633789,
      "learning_rate": 2e-05,
      "loss": 0.1579,
      "step": 2061
    },
    {
      "epoch": 1.3933761405880365,
      "grad_norm": 4.74344539642334,
      "learning_rate": 2e-05,
      "loss": 0.155,
      "step": 2062
    },
    {
      "epoch": 1.3940520446096654,
      "grad_norm": 6.421231269836426,
      "learning_rate": 2e-05,
      "loss": 0.1661,
      "step": 2063
    },
    {
      "epoch": 1.3947279486312945,
      "grad_norm": 4.61155366897583,
      "learning_rate": 2e-05,
      "loss": 0.1749,
      "step": 2064
    },
    {
      "epoch": 1.3954038526529233,
      "grad_norm": 7.245513916015625,
      "learning_rate": 2e-05,
      "loss": 0.1697,
      "step": 2065
    },
    {
      "epoch": 1.3960797566745522,
      "grad_norm": 4.338123798370361,
      "learning_rate": 2e-05,
      "loss": 0.1947,
      "step": 2066
    },
    {
      "epoch": 1.396755660696181,
      "grad_norm": 4.983460426330566,
      "learning_rate": 2e-05,
      "loss": 0.2466,
      "step": 2067
    },
    {
      "epoch": 1.39743156471781,
      "grad_norm": 11.079565048217773,
      "learning_rate": 2e-05,
      "loss": 0.2204,
      "step": 2068
    },
    {
      "epoch": 1.398107468739439,
      "grad_norm": 8.722928047180176,
      "learning_rate": 2e-05,
      "loss": 0.1607,
      "step": 2069
    },
    {
      "epoch": 1.398783372761068,
      "grad_norm": 9.074298858642578,
      "learning_rate": 2e-05,
      "loss": 0.2293,
      "step": 2070
    },
    {
      "epoch": 1.399459276782697,
      "grad_norm": 5.493626594543457,
      "learning_rate": 2e-05,
      "loss": 0.2198,
      "step": 2071
    },
    {
      "epoch": 1.4001351808043259,
      "grad_norm": 4.922308921813965,
      "learning_rate": 2e-05,
      "loss": 0.1702,
      "step": 2072
    },
    {
      "epoch": 1.4008110848259547,
      "grad_norm": 5.051478862762451,
      "learning_rate": 2e-05,
      "loss": 0.2179,
      "step": 2073
    },
    {
      "epoch": 1.4014869888475836,
      "grad_norm": 12.076773643493652,
      "learning_rate": 2e-05,
      "loss": 0.1925,
      "step": 2074
    },
    {
      "epoch": 1.4021628928692125,
      "grad_norm": 5.308622360229492,
      "learning_rate": 2e-05,
      "loss": 0.1567,
      "step": 2075
    },
    {
      "epoch": 1.4028387968908416,
      "grad_norm": 7.618307113647461,
      "learning_rate": 2e-05,
      "loss": 0.1469,
      "step": 2076
    },
    {
      "epoch": 1.4035147009124704,
      "grad_norm": 6.2768683433532715,
      "learning_rate": 2e-05,
      "loss": 0.1804,
      "step": 2077
    },
    {
      "epoch": 1.4041906049340993,
      "grad_norm": 6.09349250793457,
      "learning_rate": 2e-05,
      "loss": 0.1763,
      "step": 2078
    },
    {
      "epoch": 1.4048665089557284,
      "grad_norm": 7.016401767730713,
      "learning_rate": 2e-05,
      "loss": 0.1746,
      "step": 2079
    },
    {
      "epoch": 1.4055424129773573,
      "grad_norm": 5.4194254875183105,
      "learning_rate": 2e-05,
      "loss": 0.1651,
      "step": 2080
    },
    {
      "epoch": 1.4062183169989861,
      "grad_norm": 7.77448844909668,
      "learning_rate": 2e-05,
      "loss": 0.1741,
      "step": 2081
    },
    {
      "epoch": 1.406894221020615,
      "grad_norm": 6.203484535217285,
      "learning_rate": 2e-05,
      "loss": 0.2272,
      "step": 2082
    },
    {
      "epoch": 1.4075701250422439,
      "grad_norm": 10.899011611938477,
      "learning_rate": 2e-05,
      "loss": 0.1708,
      "step": 2083
    },
    {
      "epoch": 1.408246029063873,
      "grad_norm": 6.4422736167907715,
      "learning_rate": 2e-05,
      "loss": 0.1291,
      "step": 2084
    },
    {
      "epoch": 1.4089219330855018,
      "grad_norm": 8.354059219360352,
      "learning_rate": 2e-05,
      "loss": 0.1432,
      "step": 2085
    },
    {
      "epoch": 1.4095978371071307,
      "grad_norm": 7.324949741363525,
      "learning_rate": 2e-05,
      "loss": 0.1912,
      "step": 2086
    },
    {
      "epoch": 1.4102737411287598,
      "grad_norm": 4.837865352630615,
      "learning_rate": 2e-05,
      "loss": 0.186,
      "step": 2087
    },
    {
      "epoch": 1.4109496451503887,
      "grad_norm": 9.170174598693848,
      "learning_rate": 2e-05,
      "loss": 0.1891,
      "step": 2088
    },
    {
      "epoch": 1.4116255491720175,
      "grad_norm": 5.323495864868164,
      "learning_rate": 2e-05,
      "loss": 0.2213,
      "step": 2089
    },
    {
      "epoch": 1.4123014531936464,
      "grad_norm": 15.268009185791016,
      "learning_rate": 2e-05,
      "loss": 0.2285,
      "step": 2090
    },
    {
      "epoch": 1.4129773572152755,
      "grad_norm": 5.491397857666016,
      "learning_rate": 2e-05,
      "loss": 0.1727,
      "step": 2091
    },
    {
      "epoch": 1.4136532612369044,
      "grad_norm": 14.244745254516602,
      "learning_rate": 2e-05,
      "loss": 0.1784,
      "step": 2092
    },
    {
      "epoch": 1.4143291652585332,
      "grad_norm": 5.385805606842041,
      "learning_rate": 2e-05,
      "loss": 0.1842,
      "step": 2093
    },
    {
      "epoch": 1.4150050692801623,
      "grad_norm": 6.143791675567627,
      "learning_rate": 2e-05,
      "loss": 0.1993,
      "step": 2094
    },
    {
      "epoch": 1.4156809733017912,
      "grad_norm": 5.258811950683594,
      "learning_rate": 2e-05,
      "loss": 0.1701,
      "step": 2095
    },
    {
      "epoch": 1.41635687732342,
      "grad_norm": 4.348021984100342,
      "learning_rate": 2e-05,
      "loss": 0.1801,
      "step": 2096
    },
    {
      "epoch": 1.417032781345049,
      "grad_norm": 9.983406066894531,
      "learning_rate": 2e-05,
      "loss": 0.2136,
      "step": 2097
    },
    {
      "epoch": 1.4177086853666778,
      "grad_norm": 6.516223907470703,
      "learning_rate": 2e-05,
      "loss": 0.2622,
      "step": 2098
    },
    {
      "epoch": 1.418384589388307,
      "grad_norm": 14.846181869506836,
      "learning_rate": 2e-05,
      "loss": 0.1356,
      "step": 2099
    },
    {
      "epoch": 1.4190604934099358,
      "grad_norm": 7.3362836837768555,
      "learning_rate": 2e-05,
      "loss": 0.1815,
      "step": 2100
    },
    {
      "epoch": 1.4190604934099358,
      "eval_accuracy": 0.5858108108108108,
      "eval_loss": 0.2511114180088043,
      "eval_runtime": 48.1618,
      "eval_samples_per_second": 491.676,
      "eval_steps_per_second": 15.365,
      "step": 2100
    },
    {
      "epoch": 1.4197363974315647,
      "grad_norm": 5.6508965492248535,
      "learning_rate": 2e-05,
      "loss": 0.2142,
      "step": 2101
    },
    {
      "epoch": 1.4204123014531937,
      "grad_norm": 4.842804431915283,
      "learning_rate": 2e-05,
      "loss": 0.163,
      "step": 2102
    },
    {
      "epoch": 1.4210882054748226,
      "grad_norm": 4.581643581390381,
      "learning_rate": 2e-05,
      "loss": 0.189,
      "step": 2103
    },
    {
      "epoch": 1.4217641094964515,
      "grad_norm": 5.937347412109375,
      "learning_rate": 2e-05,
      "loss": 0.1702,
      "step": 2104
    },
    {
      "epoch": 1.4224400135180804,
      "grad_norm": 10.337443351745605,
      "learning_rate": 2e-05,
      "loss": 0.2348,
      "step": 2105
    },
    {
      "epoch": 1.4231159175397092,
      "grad_norm": 7.004786491394043,
      "learning_rate": 2e-05,
      "loss": 0.2119,
      "step": 2106
    },
    {
      "epoch": 1.4237918215613383,
      "grad_norm": 11.350846290588379,
      "learning_rate": 2e-05,
      "loss": 0.2143,
      "step": 2107
    },
    {
      "epoch": 1.4244677255829672,
      "grad_norm": 4.834009647369385,
      "learning_rate": 2e-05,
      "loss": 0.1721,
      "step": 2108
    },
    {
      "epoch": 1.4251436296045963,
      "grad_norm": 4.095510482788086,
      "learning_rate": 2e-05,
      "loss": 0.1542,
      "step": 2109
    },
    {
      "epoch": 1.4258195336262252,
      "grad_norm": 6.5880045890808105,
      "learning_rate": 2e-05,
      "loss": 0.1689,
      "step": 2110
    },
    {
      "epoch": 1.426495437647854,
      "grad_norm": 9.641397476196289,
      "learning_rate": 2e-05,
      "loss": 0.1811,
      "step": 2111
    },
    {
      "epoch": 1.427171341669483,
      "grad_norm": 10.076664924621582,
      "learning_rate": 2e-05,
      "loss": 0.125,
      "step": 2112
    },
    {
      "epoch": 1.4278472456911118,
      "grad_norm": 6.1512274742126465,
      "learning_rate": 2e-05,
      "loss": 0.244,
      "step": 2113
    },
    {
      "epoch": 1.4285231497127409,
      "grad_norm": 7.681443691253662,
      "learning_rate": 2e-05,
      "loss": 0.1664,
      "step": 2114
    },
    {
      "epoch": 1.4291990537343697,
      "grad_norm": 5.373749732971191,
      "learning_rate": 2e-05,
      "loss": 0.1342,
      "step": 2115
    },
    {
      "epoch": 1.4298749577559986,
      "grad_norm": 7.861267566680908,
      "learning_rate": 2e-05,
      "loss": 0.1498,
      "step": 2116
    },
    {
      "epoch": 1.4305508617776277,
      "grad_norm": 8.720152854919434,
      "learning_rate": 2e-05,
      "loss": 0.2401,
      "step": 2117
    },
    {
      "epoch": 1.4312267657992566,
      "grad_norm": 10.948539733886719,
      "learning_rate": 2e-05,
      "loss": 0.2375,
      "step": 2118
    },
    {
      "epoch": 1.4319026698208854,
      "grad_norm": 9.004132270812988,
      "learning_rate": 2e-05,
      "loss": 0.1829,
      "step": 2119
    },
    {
      "epoch": 1.4325785738425143,
      "grad_norm": 15.756948471069336,
      "learning_rate": 2e-05,
      "loss": 0.1352,
      "step": 2120
    },
    {
      "epoch": 1.4332544778641432,
      "grad_norm": 3.862898588180542,
      "learning_rate": 2e-05,
      "loss": 0.1461,
      "step": 2121
    },
    {
      "epoch": 1.4339303818857723,
      "grad_norm": 3.8253321647644043,
      "learning_rate": 2e-05,
      "loss": 0.1314,
      "step": 2122
    },
    {
      "epoch": 1.4346062859074011,
      "grad_norm": 5.72938871383667,
      "learning_rate": 2e-05,
      "loss": 0.1218,
      "step": 2123
    },
    {
      "epoch": 1.43528218992903,
      "grad_norm": 8.463970184326172,
      "learning_rate": 2e-05,
      "loss": 0.1588,
      "step": 2124
    },
    {
      "epoch": 1.435958093950659,
      "grad_norm": 9.170289039611816,
      "learning_rate": 2e-05,
      "loss": 0.1772,
      "step": 2125
    },
    {
      "epoch": 1.436633997972288,
      "grad_norm": 5.7200541496276855,
      "learning_rate": 2e-05,
      "loss": 0.2232,
      "step": 2126
    },
    {
      "epoch": 1.4373099019939168,
      "grad_norm": 6.515347480773926,
      "learning_rate": 2e-05,
      "loss": 0.2139,
      "step": 2127
    },
    {
      "epoch": 1.4379858060155457,
      "grad_norm": 7.929168224334717,
      "learning_rate": 2e-05,
      "loss": 0.2501,
      "step": 2128
    },
    {
      "epoch": 1.4386617100371748,
      "grad_norm": 5.898550510406494,
      "learning_rate": 2e-05,
      "loss": 0.1739,
      "step": 2129
    },
    {
      "epoch": 1.4393376140588037,
      "grad_norm": 5.371425628662109,
      "learning_rate": 2e-05,
      "loss": 0.1707,
      "step": 2130
    },
    {
      "epoch": 1.4400135180804325,
      "grad_norm": 4.931488037109375,
      "learning_rate": 2e-05,
      "loss": 0.1821,
      "step": 2131
    },
    {
      "epoch": 1.4406894221020616,
      "grad_norm": 9.548002243041992,
      "learning_rate": 2e-05,
      "loss": 0.2033,
      "step": 2132
    },
    {
      "epoch": 1.4413653261236905,
      "grad_norm": 4.5445685386657715,
      "learning_rate": 2e-05,
      "loss": 0.1625,
      "step": 2133
    },
    {
      "epoch": 1.4420412301453194,
      "grad_norm": 7.636917591094971,
      "learning_rate": 2e-05,
      "loss": 0.2427,
      "step": 2134
    },
    {
      "epoch": 1.4427171341669482,
      "grad_norm": 17.72125244140625,
      "learning_rate": 2e-05,
      "loss": 0.225,
      "step": 2135
    },
    {
      "epoch": 1.4433930381885771,
      "grad_norm": 7.98087215423584,
      "learning_rate": 2e-05,
      "loss": 0.2229,
      "step": 2136
    },
    {
      "epoch": 1.4440689422102062,
      "grad_norm": 9.231107711791992,
      "learning_rate": 2e-05,
      "loss": 0.1598,
      "step": 2137
    },
    {
      "epoch": 1.444744846231835,
      "grad_norm": 8.550161361694336,
      "learning_rate": 2e-05,
      "loss": 0.1865,
      "step": 2138
    },
    {
      "epoch": 1.445420750253464,
      "grad_norm": 9.468100547790527,
      "learning_rate": 2e-05,
      "loss": 0.1837,
      "step": 2139
    },
    {
      "epoch": 1.446096654275093,
      "grad_norm": 12.03797435760498,
      "learning_rate": 2e-05,
      "loss": 0.1759,
      "step": 2140
    },
    {
      "epoch": 1.446772558296722,
      "grad_norm": 4.305181980133057,
      "learning_rate": 2e-05,
      "loss": 0.1511,
      "step": 2141
    },
    {
      "epoch": 1.4474484623183508,
      "grad_norm": 7.499657154083252,
      "learning_rate": 2e-05,
      "loss": 0.1835,
      "step": 2142
    },
    {
      "epoch": 1.4481243663399797,
      "grad_norm": 4.144172668457031,
      "learning_rate": 2e-05,
      "loss": 0.1859,
      "step": 2143
    },
    {
      "epoch": 1.4488002703616085,
      "grad_norm": 8.463600158691406,
      "learning_rate": 2e-05,
      "loss": 0.2042,
      "step": 2144
    },
    {
      "epoch": 1.4494761743832376,
      "grad_norm": 7.159388542175293,
      "learning_rate": 2e-05,
      "loss": 0.2354,
      "step": 2145
    },
    {
      "epoch": 1.4501520784048665,
      "grad_norm": 6.080113887786865,
      "learning_rate": 2e-05,
      "loss": 0.1768,
      "step": 2146
    },
    {
      "epoch": 1.4508279824264954,
      "grad_norm": 6.5162553787231445,
      "learning_rate": 2e-05,
      "loss": 0.1795,
      "step": 2147
    },
    {
      "epoch": 1.4515038864481244,
      "grad_norm": 4.523468494415283,
      "learning_rate": 2e-05,
      "loss": 0.1989,
      "step": 2148
    },
    {
      "epoch": 1.4521797904697533,
      "grad_norm": 10.39310359954834,
      "learning_rate": 2e-05,
      "loss": 0.2114,
      "step": 2149
    },
    {
      "epoch": 1.4528556944913822,
      "grad_norm": 8.092377662658691,
      "learning_rate": 2e-05,
      "loss": 0.1591,
      "step": 2150
    },
    {
      "epoch": 1.4528556944913822,
      "eval_accuracy": 0.5788429054054054,
      "eval_loss": 0.24172343313694,
      "eval_runtime": 43.1331,
      "eval_samples_per_second": 548.998,
      "eval_steps_per_second": 17.156,
      "step": 2150
    },
    {
      "epoch": 1.453531598513011,
      "grad_norm": 7.654262065887451,
      "learning_rate": 2e-05,
      "loss": 0.1625,
      "step": 2151
    },
    {
      "epoch": 1.4542075025346402,
      "grad_norm": 4.882418632507324,
      "learning_rate": 2e-05,
      "loss": 0.2151,
      "step": 2152
    },
    {
      "epoch": 1.454883406556269,
      "grad_norm": 9.050207138061523,
      "learning_rate": 2e-05,
      "loss": 0.207,
      "step": 2153
    },
    {
      "epoch": 1.455559310577898,
      "grad_norm": 5.264075756072998,
      "learning_rate": 2e-05,
      "loss": 0.1813,
      "step": 2154
    },
    {
      "epoch": 1.456235214599527,
      "grad_norm": 9.392800331115723,
      "learning_rate": 2e-05,
      "loss": 0.1451,
      "step": 2155
    },
    {
      "epoch": 1.4569111186211559,
      "grad_norm": 7.446670055389404,
      "learning_rate": 2e-05,
      "loss": 0.248,
      "step": 2156
    },
    {
      "epoch": 1.4575870226427847,
      "grad_norm": 11.775524139404297,
      "learning_rate": 2e-05,
      "loss": 0.1794,
      "step": 2157
    },
    {
      "epoch": 1.4582629266644136,
      "grad_norm": 8.781785011291504,
      "learning_rate": 2e-05,
      "loss": 0.2014,
      "step": 2158
    },
    {
      "epoch": 1.4589388306860425,
      "grad_norm": 3.9164316654205322,
      "learning_rate": 2e-05,
      "loss": 0.1559,
      "step": 2159
    },
    {
      "epoch": 1.4596147347076716,
      "grad_norm": 10.21069049835205,
      "learning_rate": 2e-05,
      "loss": 0.1734,
      "step": 2160
    },
    {
      "epoch": 1.4602906387293004,
      "grad_norm": 13.495207786560059,
      "learning_rate": 2e-05,
      "loss": 0.2255,
      "step": 2161
    },
    {
      "epoch": 1.4609665427509293,
      "grad_norm": 10.61190128326416,
      "learning_rate": 2e-05,
      "loss": 0.2066,
      "step": 2162
    },
    {
      "epoch": 1.4616424467725584,
      "grad_norm": 10.702459335327148,
      "learning_rate": 2e-05,
      "loss": 0.2086,
      "step": 2163
    },
    {
      "epoch": 1.4623183507941873,
      "grad_norm": 4.48975133895874,
      "learning_rate": 2e-05,
      "loss": 0.17,
      "step": 2164
    },
    {
      "epoch": 1.4629942548158161,
      "grad_norm": 8.763418197631836,
      "learning_rate": 2e-05,
      "loss": 0.1783,
      "step": 2165
    },
    {
      "epoch": 1.463670158837445,
      "grad_norm": 15.992945671081543,
      "learning_rate": 2e-05,
      "loss": 0.3171,
      "step": 2166
    },
    {
      "epoch": 1.464346062859074,
      "grad_norm": 5.126511096954346,
      "learning_rate": 2e-05,
      "loss": 0.1979,
      "step": 2167
    },
    {
      "epoch": 1.465021966880703,
      "grad_norm": 5.7425994873046875,
      "learning_rate": 2e-05,
      "loss": 0.1665,
      "step": 2168
    },
    {
      "epoch": 1.4656978709023318,
      "grad_norm": 5.492970943450928,
      "learning_rate": 2e-05,
      "loss": 0.1396,
      "step": 2169
    },
    {
      "epoch": 1.466373774923961,
      "grad_norm": 9.688923835754395,
      "learning_rate": 2e-05,
      "loss": 0.1678,
      "step": 2170
    },
    {
      "epoch": 1.4670496789455898,
      "grad_norm": 14.705467224121094,
      "learning_rate": 2e-05,
      "loss": 0.2895,
      "step": 2171
    },
    {
      "epoch": 1.4677255829672187,
      "grad_norm": 8.234030723571777,
      "learning_rate": 2e-05,
      "loss": 0.1646,
      "step": 2172
    },
    {
      "epoch": 1.4684014869888475,
      "grad_norm": 9.24417495727539,
      "learning_rate": 2e-05,
      "loss": 0.1578,
      "step": 2173
    },
    {
      "epoch": 1.4690773910104764,
      "grad_norm": 10.435982704162598,
      "learning_rate": 2e-05,
      "loss": 0.1455,
      "step": 2174
    },
    {
      "epoch": 1.4697532950321055,
      "grad_norm": 4.505556106567383,
      "learning_rate": 2e-05,
      "loss": 0.1917,
      "step": 2175
    },
    {
      "epoch": 1.4704291990537344,
      "grad_norm": 5.995444297790527,
      "learning_rate": 2e-05,
      "loss": 0.1707,
      "step": 2176
    },
    {
      "epoch": 1.4711051030753632,
      "grad_norm": 4.8840436935424805,
      "learning_rate": 2e-05,
      "loss": 0.1673,
      "step": 2177
    },
    {
      "epoch": 1.4717810070969923,
      "grad_norm": 7.204172134399414,
      "learning_rate": 2e-05,
      "loss": 0.2533,
      "step": 2178
    },
    {
      "epoch": 1.4724569111186212,
      "grad_norm": 12.251921653747559,
      "learning_rate": 2e-05,
      "loss": 0.146,
      "step": 2179
    },
    {
      "epoch": 1.47313281514025,
      "grad_norm": 4.832569122314453,
      "learning_rate": 2e-05,
      "loss": 0.2039,
      "step": 2180
    },
    {
      "epoch": 1.473808719161879,
      "grad_norm": 5.905416011810303,
      "learning_rate": 2e-05,
      "loss": 0.1528,
      "step": 2181
    },
    {
      "epoch": 1.4744846231835078,
      "grad_norm": 7.101807594299316,
      "learning_rate": 2e-05,
      "loss": 0.1426,
      "step": 2182
    },
    {
      "epoch": 1.475160527205137,
      "grad_norm": 5.10390043258667,
      "learning_rate": 2e-05,
      "loss": 0.181,
      "step": 2183
    },
    {
      "epoch": 1.4758364312267658,
      "grad_norm": 5.634151935577393,
      "learning_rate": 2e-05,
      "loss": 0.1595,
      "step": 2184
    },
    {
      "epoch": 1.4765123352483946,
      "grad_norm": 8.057448387145996,
      "learning_rate": 2e-05,
      "loss": 0.1519,
      "step": 2185
    },
    {
      "epoch": 1.4771882392700237,
      "grad_norm": 9.430185317993164,
      "learning_rate": 2e-05,
      "loss": 0.1313,
      "step": 2186
    },
    {
      "epoch": 1.4778641432916526,
      "grad_norm": 4.729147434234619,
      "learning_rate": 2e-05,
      "loss": 0.1753,
      "step": 2187
    },
    {
      "epoch": 1.4785400473132815,
      "grad_norm": 10.875343322753906,
      "learning_rate": 2e-05,
      "loss": 0.16,
      "step": 2188
    },
    {
      "epoch": 1.4792159513349104,
      "grad_norm": 4.842147350311279,
      "learning_rate": 2e-05,
      "loss": 0.1868,
      "step": 2189
    },
    {
      "epoch": 1.4798918553565394,
      "grad_norm": 9.174854278564453,
      "learning_rate": 2e-05,
      "loss": 0.1433,
      "step": 2190
    },
    {
      "epoch": 1.4805677593781683,
      "grad_norm": 5.437632083892822,
      "learning_rate": 2e-05,
      "loss": 0.1885,
      "step": 2191
    },
    {
      "epoch": 1.4812436633997972,
      "grad_norm": 8.09791088104248,
      "learning_rate": 2e-05,
      "loss": 0.1384,
      "step": 2192
    },
    {
      "epoch": 1.4819195674214263,
      "grad_norm": 4.850093841552734,
      "learning_rate": 2e-05,
      "loss": 0.2339,
      "step": 2193
    },
    {
      "epoch": 1.4825954714430551,
      "grad_norm": 11.142571449279785,
      "learning_rate": 2e-05,
      "loss": 0.1929,
      "step": 2194
    },
    {
      "epoch": 1.483271375464684,
      "grad_norm": 7.109097480773926,
      "learning_rate": 2e-05,
      "loss": 0.214,
      "step": 2195
    },
    {
      "epoch": 1.4839472794863129,
      "grad_norm": 4.472867012023926,
      "learning_rate": 2e-05,
      "loss": 0.1569,
      "step": 2196
    },
    {
      "epoch": 1.4846231835079418,
      "grad_norm": 6.484102725982666,
      "learning_rate": 2e-05,
      "loss": 0.1952,
      "step": 2197
    },
    {
      "epoch": 1.4852990875295708,
      "grad_norm": 8.640192985534668,
      "learning_rate": 2e-05,
      "loss": 0.1998,
      "step": 2198
    },
    {
      "epoch": 1.4859749915511997,
      "grad_norm": 3.552938222885132,
      "learning_rate": 2e-05,
      "loss": 0.2118,
      "step": 2199
    },
    {
      "epoch": 1.4866508955728286,
      "grad_norm": 15.382543563842773,
      "learning_rate": 2e-05,
      "loss": 0.231,
      "step": 2200
    },
    {
      "epoch": 1.4866508955728286,
      "eval_accuracy": 0.5773648648648648,
      "eval_loss": 0.2474917620420456,
      "eval_runtime": 43.4176,
      "eval_samples_per_second": 545.402,
      "eval_steps_per_second": 17.044,
      "step": 2200
    },
    {
      "epoch": 1.4873267995944577,
      "grad_norm": 4.840421676635742,
      "learning_rate": 2e-05,
      "loss": 0.2202,
      "step": 2201
    },
    {
      "epoch": 1.4880027036160866,
      "grad_norm": 5.706277847290039,
      "learning_rate": 2e-05,
      "loss": 0.2183,
      "step": 2202
    },
    {
      "epoch": 1.4886786076377154,
      "grad_norm": 5.44762659072876,
      "learning_rate": 2e-05,
      "loss": 0.1614,
      "step": 2203
    },
    {
      "epoch": 1.4893545116593443,
      "grad_norm": 6.158466815948486,
      "learning_rate": 2e-05,
      "loss": 0.1987,
      "step": 2204
    },
    {
      "epoch": 1.4900304156809732,
      "grad_norm": 3.2267990112304688,
      "learning_rate": 2e-05,
      "loss": 0.179,
      "step": 2205
    },
    {
      "epoch": 1.4907063197026023,
      "grad_norm": 7.055068016052246,
      "learning_rate": 2e-05,
      "loss": 0.1813,
      "step": 2206
    },
    {
      "epoch": 1.4913822237242311,
      "grad_norm": 10.071561813354492,
      "learning_rate": 2e-05,
      "loss": 0.2329,
      "step": 2207
    },
    {
      "epoch": 1.4920581277458602,
      "grad_norm": 17.695072174072266,
      "learning_rate": 2e-05,
      "loss": 0.1754,
      "step": 2208
    },
    {
      "epoch": 1.492734031767489,
      "grad_norm": 3.7514796257019043,
      "learning_rate": 2e-05,
      "loss": 0.1787,
      "step": 2209
    },
    {
      "epoch": 1.493409935789118,
      "grad_norm": 4.486561298370361,
      "learning_rate": 2e-05,
      "loss": 0.1856,
      "step": 2210
    },
    {
      "epoch": 1.4940858398107468,
      "grad_norm": 5.0584330558776855,
      "learning_rate": 2e-05,
      "loss": 0.1518,
      "step": 2211
    },
    {
      "epoch": 1.4947617438323757,
      "grad_norm": 4.961783409118652,
      "learning_rate": 2e-05,
      "loss": 0.1477,
      "step": 2212
    },
    {
      "epoch": 1.4954376478540048,
      "grad_norm": 4.054569244384766,
      "learning_rate": 2e-05,
      "loss": 0.2561,
      "step": 2213
    },
    {
      "epoch": 1.4961135518756337,
      "grad_norm": 17.047414779663086,
      "learning_rate": 2e-05,
      "loss": 0.2804,
      "step": 2214
    },
    {
      "epoch": 1.4967894558972625,
      "grad_norm": 13.178735733032227,
      "learning_rate": 2e-05,
      "loss": 0.2416,
      "step": 2215
    },
    {
      "epoch": 1.4974653599188916,
      "grad_norm": 3.969373941421509,
      "learning_rate": 2e-05,
      "loss": 0.1361,
      "step": 2216
    },
    {
      "epoch": 1.4981412639405205,
      "grad_norm": 11.148656845092773,
      "learning_rate": 2e-05,
      "loss": 0.154,
      "step": 2217
    },
    {
      "epoch": 1.4988171679621494,
      "grad_norm": 7.069233417510986,
      "learning_rate": 2e-05,
      "loss": 0.1965,
      "step": 2218
    },
    {
      "epoch": 1.4994930719837782,
      "grad_norm": 8.285849571228027,
      "learning_rate": 2e-05,
      "loss": 0.2002,
      "step": 2219
    },
    {
      "epoch": 1.500168976005407,
      "grad_norm": 4.704885482788086,
      "learning_rate": 2e-05,
      "loss": 0.1125,
      "step": 2220
    },
    {
      "epoch": 1.5008448800270362,
      "grad_norm": 4.855522155761719,
      "learning_rate": 2e-05,
      "loss": 0.1487,
      "step": 2221
    },
    {
      "epoch": 1.501520784048665,
      "grad_norm": 8.186415672302246,
      "learning_rate": 2e-05,
      "loss": 0.2169,
      "step": 2222
    },
    {
      "epoch": 1.5021966880702942,
      "grad_norm": 13.282955169677734,
      "learning_rate": 2e-05,
      "loss": 0.2917,
      "step": 2223
    },
    {
      "epoch": 1.502872592091923,
      "grad_norm": 5.817259311676025,
      "learning_rate": 2e-05,
      "loss": 0.1665,
      "step": 2224
    },
    {
      "epoch": 1.503548496113552,
      "grad_norm": 4.731293678283691,
      "learning_rate": 2e-05,
      "loss": 0.1346,
      "step": 2225
    },
    {
      "epoch": 1.5042244001351808,
      "grad_norm": 5.403949737548828,
      "learning_rate": 2e-05,
      "loss": 0.1739,
      "step": 2226
    },
    {
      "epoch": 1.5049003041568096,
      "grad_norm": 10.30024528503418,
      "learning_rate": 2e-05,
      "loss": 0.1859,
      "step": 2227
    },
    {
      "epoch": 1.5055762081784385,
      "grad_norm": 9.857585906982422,
      "learning_rate": 2e-05,
      "loss": 0.2024,
      "step": 2228
    },
    {
      "epoch": 1.5062521122000676,
      "grad_norm": 7.376889228820801,
      "learning_rate": 2e-05,
      "loss": 0.1722,
      "step": 2229
    },
    {
      "epoch": 1.5069280162216965,
      "grad_norm": 7.947818756103516,
      "learning_rate": 2e-05,
      "loss": 0.1603,
      "step": 2230
    },
    {
      "epoch": 1.5076039202433256,
      "grad_norm": 4.684057712554932,
      "learning_rate": 2e-05,
      "loss": 0.1305,
      "step": 2231
    },
    {
      "epoch": 1.5082798242649544,
      "grad_norm": 5.461733341217041,
      "learning_rate": 2e-05,
      "loss": 0.2036,
      "step": 2232
    },
    {
      "epoch": 1.5089557282865833,
      "grad_norm": 17.389698028564453,
      "learning_rate": 2e-05,
      "loss": 0.2366,
      "step": 2233
    },
    {
      "epoch": 1.5096316323082122,
      "grad_norm": 18.442707061767578,
      "learning_rate": 2e-05,
      "loss": 0.2018,
      "step": 2234
    },
    {
      "epoch": 1.510307536329841,
      "grad_norm": 4.597135543823242,
      "learning_rate": 2e-05,
      "loss": 0.1788,
      "step": 2235
    },
    {
      "epoch": 1.5109834403514701,
      "grad_norm": 5.084779262542725,
      "learning_rate": 2e-05,
      "loss": 0.2524,
      "step": 2236
    },
    {
      "epoch": 1.511659344373099,
      "grad_norm": 8.046063423156738,
      "learning_rate": 2e-05,
      "loss": 0.1938,
      "step": 2237
    },
    {
      "epoch": 1.512335248394728,
      "grad_norm": 5.279938697814941,
      "learning_rate": 2e-05,
      "loss": 0.2101,
      "step": 2238
    },
    {
      "epoch": 1.513011152416357,
      "grad_norm": 10.87525749206543,
      "learning_rate": 2e-05,
      "loss": 0.2224,
      "step": 2239
    },
    {
      "epoch": 1.5136870564379858,
      "grad_norm": 6.03635835647583,
      "learning_rate": 2e-05,
      "loss": 0.1896,
      "step": 2240
    },
    {
      "epoch": 1.5143629604596147,
      "grad_norm": 9.059898376464844,
      "learning_rate": 2e-05,
      "loss": 0.1382,
      "step": 2241
    },
    {
      "epoch": 1.5150388644812436,
      "grad_norm": 6.340293884277344,
      "learning_rate": 2e-05,
      "loss": 0.1724,
      "step": 2242
    },
    {
      "epoch": 1.5157147685028725,
      "grad_norm": 4.903107643127441,
      "learning_rate": 2e-05,
      "loss": 0.2003,
      "step": 2243
    },
    {
      "epoch": 1.5163906725245015,
      "grad_norm": 6.612017631530762,
      "learning_rate": 2e-05,
      "loss": 0.1236,
      "step": 2244
    },
    {
      "epoch": 1.5170665765461304,
      "grad_norm": 10.272838592529297,
      "learning_rate": 2e-05,
      "loss": 0.227,
      "step": 2245
    },
    {
      "epoch": 1.5177424805677595,
      "grad_norm": 7.0355143547058105,
      "learning_rate": 2e-05,
      "loss": 0.2345,
      "step": 2246
    },
    {
      "epoch": 1.5184183845893884,
      "grad_norm": 10.808367729187012,
      "learning_rate": 2e-05,
      "loss": 0.2106,
      "step": 2247
    },
    {
      "epoch": 1.5190942886110173,
      "grad_norm": 13.721657752990723,
      "learning_rate": 2e-05,
      "loss": 0.2074,
      "step": 2248
    },
    {
      "epoch": 1.5197701926326461,
      "grad_norm": 10.856085777282715,
      "learning_rate": 2e-05,
      "loss": 0.2516,
      "step": 2249
    },
    {
      "epoch": 1.520446096654275,
      "grad_norm": 10.965245246887207,
      "learning_rate": 2e-05,
      "loss": 0.1927,
      "step": 2250
    },
    {
      "epoch": 1.520446096654275,
      "eval_accuracy": 0.5722128378378378,
      "eval_loss": 0.23945848643779755,
      "eval_runtime": 43.4346,
      "eval_samples_per_second": 545.188,
      "eval_steps_per_second": 17.037,
      "step": 2250
    },
    {
      "epoch": 1.5211220006759039,
      "grad_norm": 4.381246566772461,
      "learning_rate": 2e-05,
      "loss": 0.1512,
      "step": 2251
    },
    {
      "epoch": 1.521797904697533,
      "grad_norm": 6.765244007110596,
      "learning_rate": 2e-05,
      "loss": 0.1705,
      "step": 2252
    },
    {
      "epoch": 1.5224738087191618,
      "grad_norm": 5.378074645996094,
      "learning_rate": 2e-05,
      "loss": 0.2264,
      "step": 2253
    },
    {
      "epoch": 1.523149712740791,
      "grad_norm": 5.012163162231445,
      "learning_rate": 2e-05,
      "loss": 0.1671,
      "step": 2254
    },
    {
      "epoch": 1.5238256167624198,
      "grad_norm": 4.966264247894287,
      "learning_rate": 2e-05,
      "loss": 0.1915,
      "step": 2255
    },
    {
      "epoch": 1.5245015207840487,
      "grad_norm": 4.353409290313721,
      "learning_rate": 2e-05,
      "loss": 0.1704,
      "step": 2256
    },
    {
      "epoch": 1.5251774248056775,
      "grad_norm": 4.247950077056885,
      "learning_rate": 2e-05,
      "loss": 0.1875,
      "step": 2257
    },
    {
      "epoch": 1.5258533288273064,
      "grad_norm": 5.50316047668457,
      "learning_rate": 2e-05,
      "loss": 0.2073,
      "step": 2258
    },
    {
      "epoch": 1.5265292328489355,
      "grad_norm": 7.349883079528809,
      "learning_rate": 2e-05,
      "loss": 0.2133,
      "step": 2259
    },
    {
      "epoch": 1.5272051368705644,
      "grad_norm": 6.416101932525635,
      "learning_rate": 2e-05,
      "loss": 0.1992,
      "step": 2260
    },
    {
      "epoch": 1.5278810408921935,
      "grad_norm": 14.082527160644531,
      "learning_rate": 2e-05,
      "loss": 0.3069,
      "step": 2261
    },
    {
      "epoch": 1.5285569449138223,
      "grad_norm": 15.03607177734375,
      "learning_rate": 2e-05,
      "loss": 0.1798,
      "step": 2262
    },
    {
      "epoch": 1.5292328489354512,
      "grad_norm": 15.619485855102539,
      "learning_rate": 2e-05,
      "loss": 0.1971,
      "step": 2263
    },
    {
      "epoch": 1.52990875295708,
      "grad_norm": 15.546542167663574,
      "learning_rate": 2e-05,
      "loss": 0.2124,
      "step": 2264
    },
    {
      "epoch": 1.530584656978709,
      "grad_norm": 9.49348258972168,
      "learning_rate": 2e-05,
      "loss": 0.1608,
      "step": 2265
    },
    {
      "epoch": 1.5312605610003378,
      "grad_norm": 3.584841012954712,
      "learning_rate": 2e-05,
      "loss": 0.188,
      "step": 2266
    },
    {
      "epoch": 1.531936465021967,
      "grad_norm": 7.976410388946533,
      "learning_rate": 2e-05,
      "loss": 0.2567,
      "step": 2267
    },
    {
      "epoch": 1.5326123690435958,
      "grad_norm": 12.009162902832031,
      "learning_rate": 2e-05,
      "loss": 0.1988,
      "step": 2268
    },
    {
      "epoch": 1.5332882730652249,
      "grad_norm": 12.632593154907227,
      "learning_rate": 2e-05,
      "loss": 0.1663,
      "step": 2269
    },
    {
      "epoch": 1.5339641770868537,
      "grad_norm": 4.518133163452148,
      "learning_rate": 2e-05,
      "loss": 0.1953,
      "step": 2270
    },
    {
      "epoch": 1.5346400811084826,
      "grad_norm": 4.8873162269592285,
      "learning_rate": 2e-05,
      "loss": 0.1291,
      "step": 2271
    },
    {
      "epoch": 1.5353159851301115,
      "grad_norm": 10.1647310256958,
      "learning_rate": 2e-05,
      "loss": 0.1413,
      "step": 2272
    },
    {
      "epoch": 1.5359918891517403,
      "grad_norm": 8.487892150878906,
      "learning_rate": 2e-05,
      "loss": 0.1474,
      "step": 2273
    },
    {
      "epoch": 1.5366677931733694,
      "grad_norm": 5.174804210662842,
      "learning_rate": 2e-05,
      "loss": 0.1239,
      "step": 2274
    },
    {
      "epoch": 1.5373436971949983,
      "grad_norm": 3.9673819541931152,
      "learning_rate": 2e-05,
      "loss": 0.1946,
      "step": 2275
    },
    {
      "epoch": 1.5380196012166274,
      "grad_norm": 14.35671615600586,
      "learning_rate": 2e-05,
      "loss": 0.1838,
      "step": 2276
    },
    {
      "epoch": 1.5386955052382563,
      "grad_norm": 4.901583194732666,
      "learning_rate": 2e-05,
      "loss": 0.1943,
      "step": 2277
    },
    {
      "epoch": 1.5393714092598851,
      "grad_norm": 13.62162971496582,
      "learning_rate": 2e-05,
      "loss": 0.1437,
      "step": 2278
    },
    {
      "epoch": 1.540047313281514,
      "grad_norm": 6.3414130210876465,
      "learning_rate": 2e-05,
      "loss": 0.1486,
      "step": 2279
    },
    {
      "epoch": 1.5407232173031429,
      "grad_norm": 5.703877925872803,
      "learning_rate": 2e-05,
      "loss": 0.1743,
      "step": 2280
    },
    {
      "epoch": 1.5413991213247717,
      "grad_norm": 9.699417114257812,
      "learning_rate": 2e-05,
      "loss": 0.1662,
      "step": 2281
    },
    {
      "epoch": 1.5420750253464008,
      "grad_norm": 12.090535163879395,
      "learning_rate": 2e-05,
      "loss": 0.1575,
      "step": 2282
    },
    {
      "epoch": 1.5427509293680297,
      "grad_norm": 5.752380847930908,
      "learning_rate": 2e-05,
      "loss": 0.1856,
      "step": 2283
    },
    {
      "epoch": 1.5434268333896588,
      "grad_norm": 10.353852272033691,
      "learning_rate": 2e-05,
      "loss": 0.1625,
      "step": 2284
    },
    {
      "epoch": 1.5441027374112877,
      "grad_norm": 8.841320037841797,
      "learning_rate": 2e-05,
      "loss": 0.1986,
      "step": 2285
    },
    {
      "epoch": 1.5447786414329165,
      "grad_norm": 5.844013690948486,
      "learning_rate": 2e-05,
      "loss": 0.1663,
      "step": 2286
    },
    {
      "epoch": 1.5454545454545454,
      "grad_norm": 6.027374267578125,
      "learning_rate": 2e-05,
      "loss": 0.1583,
      "step": 2287
    },
    {
      "epoch": 1.5461304494761743,
      "grad_norm": 6.576290607452393,
      "learning_rate": 2e-05,
      "loss": 0.1655,
      "step": 2288
    },
    {
      "epoch": 1.5468063534978032,
      "grad_norm": 9.782381057739258,
      "learning_rate": 2e-05,
      "loss": 0.144,
      "step": 2289
    },
    {
      "epoch": 1.5474822575194322,
      "grad_norm": 8.616189002990723,
      "learning_rate": 2e-05,
      "loss": 0.2029,
      "step": 2290
    },
    {
      "epoch": 1.5481581615410611,
      "grad_norm": 4.321144104003906,
      "learning_rate": 2e-05,
      "loss": 0.1799,
      "step": 2291
    },
    {
      "epoch": 1.5488340655626902,
      "grad_norm": 11.155842781066895,
      "learning_rate": 2e-05,
      "loss": 0.2316,
      "step": 2292
    },
    {
      "epoch": 1.549509969584319,
      "grad_norm": 16.357318878173828,
      "learning_rate": 2e-05,
      "loss": 0.1548,
      "step": 2293
    },
    {
      "epoch": 1.550185873605948,
      "grad_norm": 12.613271713256836,
      "learning_rate": 2e-05,
      "loss": 0.1789,
      "step": 2294
    },
    {
      "epoch": 1.5508617776275768,
      "grad_norm": 5.494408130645752,
      "learning_rate": 2e-05,
      "loss": 0.16,
      "step": 2295
    },
    {
      "epoch": 1.5515376816492057,
      "grad_norm": 4.317010402679443,
      "learning_rate": 2e-05,
      "loss": 0.205,
      "step": 2296
    },
    {
      "epoch": 1.5522135856708348,
      "grad_norm": 11.385794639587402,
      "learning_rate": 2e-05,
      "loss": 0.14,
      "step": 2297
    },
    {
      "epoch": 1.5528894896924637,
      "grad_norm": 10.275989532470703,
      "learning_rate": 2e-05,
      "loss": 0.1732,
      "step": 2298
    },
    {
      "epoch": 1.5535653937140927,
      "grad_norm": 5.948975563049316,
      "learning_rate": 2e-05,
      "loss": 0.1699,
      "step": 2299
    },
    {
      "epoch": 1.5542412977357216,
      "grad_norm": 4.115506649017334,
      "learning_rate": 2e-05,
      "loss": 0.1324,
      "step": 2300
    },
    {
      "epoch": 1.5542412977357216,
      "eval_accuracy": 0.5864864864864865,
      "eval_loss": 0.25849655270576477,
      "eval_runtime": 43.9186,
      "eval_samples_per_second": 539.18,
      "eval_steps_per_second": 16.849,
      "step": 2300
    },
    {
      "epoch": 1.5549172017573505,
      "grad_norm": 8.776203155517578,
      "learning_rate": 2e-05,
      "loss": 0.2016,
      "step": 2301
    },
    {
      "epoch": 1.5555931057789794,
      "grad_norm": 16.72568130493164,
      "learning_rate": 2e-05,
      "loss": 0.2085,
      "step": 2302
    },
    {
      "epoch": 1.5562690098006082,
      "grad_norm": 5.601991653442383,
      "learning_rate": 2e-05,
      "loss": 0.1929,
      "step": 2303
    },
    {
      "epoch": 1.556944913822237,
      "grad_norm": 8.078157424926758,
      "learning_rate": 2e-05,
      "loss": 0.1867,
      "step": 2304
    },
    {
      "epoch": 1.5576208178438662,
      "grad_norm": 8.878862380981445,
      "learning_rate": 2e-05,
      "loss": 0.2114,
      "step": 2305
    },
    {
      "epoch": 1.558296721865495,
      "grad_norm": 9.608660697937012,
      "learning_rate": 2e-05,
      "loss": 0.2183,
      "step": 2306
    },
    {
      "epoch": 1.5589726258871242,
      "grad_norm": 12.102940559387207,
      "learning_rate": 2e-05,
      "loss": 0.2201,
      "step": 2307
    },
    {
      "epoch": 1.559648529908753,
      "grad_norm": 19.34791374206543,
      "learning_rate": 2e-05,
      "loss": 0.2482,
      "step": 2308
    },
    {
      "epoch": 1.560324433930382,
      "grad_norm": 4.467925548553467,
      "learning_rate": 2e-05,
      "loss": 0.162,
      "step": 2309
    },
    {
      "epoch": 1.5610003379520108,
      "grad_norm": 4.155332088470459,
      "learning_rate": 2e-05,
      "loss": 0.2179,
      "step": 2310
    },
    {
      "epoch": 1.5616762419736396,
      "grad_norm": 11.80615234375,
      "learning_rate": 2e-05,
      "loss": 0.2155,
      "step": 2311
    },
    {
      "epoch": 1.5623521459952687,
      "grad_norm": 4.693686485290527,
      "learning_rate": 2e-05,
      "loss": 0.1424,
      "step": 2312
    },
    {
      "epoch": 1.5630280500168976,
      "grad_norm": 3.609858751296997,
      "learning_rate": 2e-05,
      "loss": 0.1872,
      "step": 2313
    },
    {
      "epoch": 1.5637039540385267,
      "grad_norm": 4.907233715057373,
      "learning_rate": 2e-05,
      "loss": 0.2253,
      "step": 2314
    },
    {
      "epoch": 1.5643798580601556,
      "grad_norm": 9.265901565551758,
      "learning_rate": 2e-05,
      "loss": 0.1585,
      "step": 2315
    },
    {
      "epoch": 1.5650557620817844,
      "grad_norm": 5.0789313316345215,
      "learning_rate": 2e-05,
      "loss": 0.1713,
      "step": 2316
    },
    {
      "epoch": 1.5657316661034133,
      "grad_norm": 4.391917705535889,
      "learning_rate": 2e-05,
      "loss": 0.2029,
      "step": 2317
    },
    {
      "epoch": 1.5664075701250422,
      "grad_norm": 5.53513765335083,
      "learning_rate": 2e-05,
      "loss": 0.1772,
      "step": 2318
    },
    {
      "epoch": 1.567083474146671,
      "grad_norm": 13.731374740600586,
      "learning_rate": 2e-05,
      "loss": 0.2498,
      "step": 2319
    },
    {
      "epoch": 1.5677593781683001,
      "grad_norm": 4.941620826721191,
      "learning_rate": 2e-05,
      "loss": 0.1369,
      "step": 2320
    },
    {
      "epoch": 1.568435282189929,
      "grad_norm": 11.691679000854492,
      "learning_rate": 2e-05,
      "loss": 0.1686,
      "step": 2321
    },
    {
      "epoch": 1.569111186211558,
      "grad_norm": 4.242433547973633,
      "learning_rate": 2e-05,
      "loss": 0.1816,
      "step": 2322
    },
    {
      "epoch": 1.569787090233187,
      "grad_norm": 8.025243759155273,
      "learning_rate": 2e-05,
      "loss": 0.2183,
      "step": 2323
    },
    {
      "epoch": 1.5704629942548158,
      "grad_norm": 5.103522300720215,
      "learning_rate": 2e-05,
      "loss": 0.1465,
      "step": 2324
    },
    {
      "epoch": 1.5711388982764447,
      "grad_norm": 5.336676120758057,
      "learning_rate": 2e-05,
      "loss": 0.1768,
      "step": 2325
    },
    {
      "epoch": 1.5718148022980736,
      "grad_norm": 18.870641708374023,
      "learning_rate": 2e-05,
      "loss": 0.248,
      "step": 2326
    },
    {
      "epoch": 1.5724907063197024,
      "grad_norm": 7.780706405639648,
      "learning_rate": 2e-05,
      "loss": 0.1874,
      "step": 2327
    },
    {
      "epoch": 1.5731666103413315,
      "grad_norm": 6.818233013153076,
      "learning_rate": 2e-05,
      "loss": 0.1943,
      "step": 2328
    },
    {
      "epoch": 1.5738425143629604,
      "grad_norm": 4.998536586761475,
      "learning_rate": 2e-05,
      "loss": 0.2069,
      "step": 2329
    },
    {
      "epoch": 1.5745184183845895,
      "grad_norm": 4.884218692779541,
      "learning_rate": 2e-05,
      "loss": 0.1581,
      "step": 2330
    },
    {
      "epoch": 1.5751943224062184,
      "grad_norm": 10.900787353515625,
      "learning_rate": 2e-05,
      "loss": 0.1949,
      "step": 2331
    },
    {
      "epoch": 1.5758702264278472,
      "grad_norm": 11.378925323486328,
      "learning_rate": 2e-05,
      "loss": 0.1715,
      "step": 2332
    },
    {
      "epoch": 1.5765461304494761,
      "grad_norm": 17.85470199584961,
      "learning_rate": 2e-05,
      "loss": 0.259,
      "step": 2333
    },
    {
      "epoch": 1.577222034471105,
      "grad_norm": 13.534132957458496,
      "learning_rate": 2e-05,
      "loss": 0.1838,
      "step": 2334
    },
    {
      "epoch": 1.577897938492734,
      "grad_norm": 10.618051528930664,
      "learning_rate": 2e-05,
      "loss": 0.213,
      "step": 2335
    },
    {
      "epoch": 1.578573842514363,
      "grad_norm": 4.630187034606934,
      "learning_rate": 2e-05,
      "loss": 0.1315,
      "step": 2336
    },
    {
      "epoch": 1.579249746535992,
      "grad_norm": 15.9749174118042,
      "learning_rate": 2e-05,
      "loss": 0.3333,
      "step": 2337
    },
    {
      "epoch": 1.579925650557621,
      "grad_norm": 25.845983505249023,
      "learning_rate": 2e-05,
      "loss": 0.1721,
      "step": 2338
    },
    {
      "epoch": 1.5806015545792498,
      "grad_norm": 5.581686496734619,
      "learning_rate": 2e-05,
      "loss": 0.1873,
      "step": 2339
    },
    {
      "epoch": 1.5812774586008786,
      "grad_norm": 4.403299808502197,
      "learning_rate": 2e-05,
      "loss": 0.1995,
      "step": 2340
    },
    {
      "epoch": 1.5819533626225075,
      "grad_norm": 6.205742835998535,
      "learning_rate": 2e-05,
      "loss": 0.1943,
      "step": 2341
    },
    {
      "epoch": 1.5826292666441364,
      "grad_norm": 18.68754768371582,
      "learning_rate": 2e-05,
      "loss": 0.2124,
      "step": 2342
    },
    {
      "epoch": 1.5833051706657655,
      "grad_norm": 14.0961332321167,
      "learning_rate": 2e-05,
      "loss": 0.2086,
      "step": 2343
    },
    {
      "epoch": 1.5839810746873944,
      "grad_norm": 18.117244720458984,
      "learning_rate": 2e-05,
      "loss": 0.1887,
      "step": 2344
    },
    {
      "epoch": 1.5846569787090234,
      "grad_norm": 6.50591516494751,
      "learning_rate": 2e-05,
      "loss": 0.2179,
      "step": 2345
    },
    {
      "epoch": 1.5853328827306523,
      "grad_norm": 4.6561384201049805,
      "learning_rate": 2e-05,
      "loss": 0.1435,
      "step": 2346
    },
    {
      "epoch": 1.5860087867522812,
      "grad_norm": 4.542756080627441,
      "learning_rate": 2e-05,
      "loss": 0.2414,
      "step": 2347
    },
    {
      "epoch": 1.58668469077391,
      "grad_norm": 20.08609962463379,
      "learning_rate": 2e-05,
      "loss": 0.229,
      "step": 2348
    },
    {
      "epoch": 1.587360594795539,
      "grad_norm": 14.879796028137207,
      "learning_rate": 2e-05,
      "loss": 0.1725,
      "step": 2349
    },
    {
      "epoch": 1.5880364988171678,
      "grad_norm": 10.37997817993164,
      "learning_rate": 2e-05,
      "loss": 0.1965,
      "step": 2350
    },
    {
      "epoch": 1.5880364988171678,
      "eval_accuracy": 0.5790962837837837,
      "eval_loss": 0.2376258820295334,
      "eval_runtime": 44.7723,
      "eval_samples_per_second": 528.898,
      "eval_steps_per_second": 16.528,
      "step": 2350
    },
    {
      "epoch": 1.5887124028387969,
      "grad_norm": 4.8163933753967285,
      "learning_rate": 2e-05,
      "loss": 0.1787,
      "step": 2351
    },
    {
      "epoch": 1.589388306860426,
      "grad_norm": 3.6620702743530273,
      "learning_rate": 2e-05,
      "loss": 0.1129,
      "step": 2352
    },
    {
      "epoch": 1.5900642108820549,
      "grad_norm": 8.62623119354248,
      "learning_rate": 2e-05,
      "loss": 0.189,
      "step": 2353
    },
    {
      "epoch": 1.5907401149036837,
      "grad_norm": 9.799690246582031,
      "learning_rate": 2e-05,
      "loss": 0.2285,
      "step": 2354
    },
    {
      "epoch": 1.5914160189253126,
      "grad_norm": 11.166980743408203,
      "learning_rate": 2e-05,
      "loss": 0.1675,
      "step": 2355
    },
    {
      "epoch": 1.5920919229469415,
      "grad_norm": 14.605269432067871,
      "learning_rate": 2e-05,
      "loss": 0.2312,
      "step": 2356
    },
    {
      "epoch": 1.5927678269685703,
      "grad_norm": 9.851414680480957,
      "learning_rate": 2e-05,
      "loss": 0.1432,
      "step": 2357
    },
    {
      "epoch": 1.5934437309901994,
      "grad_norm": 12.165542602539062,
      "learning_rate": 2e-05,
      "loss": 0.2035,
      "step": 2358
    },
    {
      "epoch": 1.5941196350118283,
      "grad_norm": 12.985828399658203,
      "learning_rate": 2e-05,
      "loss": 0.1645,
      "step": 2359
    },
    {
      "epoch": 1.5947955390334574,
      "grad_norm": 16.755958557128906,
      "learning_rate": 2e-05,
      "loss": 0.281,
      "step": 2360
    },
    {
      "epoch": 1.5954714430550863,
      "grad_norm": 26.619184494018555,
      "learning_rate": 2e-05,
      "loss": 0.2543,
      "step": 2361
    },
    {
      "epoch": 1.5961473470767151,
      "grad_norm": 10.029675483703613,
      "learning_rate": 2e-05,
      "loss": 0.1949,
      "step": 2362
    },
    {
      "epoch": 1.596823251098344,
      "grad_norm": 9.09751033782959,
      "learning_rate": 2e-05,
      "loss": 0.1612,
      "step": 2363
    },
    {
      "epoch": 1.5974991551199729,
      "grad_norm": 6.31437349319458,
      "learning_rate": 2e-05,
      "loss": 0.1824,
      "step": 2364
    },
    {
      "epoch": 1.5981750591416017,
      "grad_norm": 8.929709434509277,
      "learning_rate": 2e-05,
      "loss": 0.2107,
      "step": 2365
    },
    {
      "epoch": 1.5988509631632308,
      "grad_norm": 20.294939041137695,
      "learning_rate": 2e-05,
      "loss": 0.2028,
      "step": 2366
    },
    {
      "epoch": 1.5995268671848597,
      "grad_norm": 17.63393211364746,
      "learning_rate": 2e-05,
      "loss": 0.2209,
      "step": 2367
    },
    {
      "epoch": 1.6002027712064888,
      "grad_norm": 19.241361618041992,
      "learning_rate": 2e-05,
      "loss": 0.2092,
      "step": 2368
    },
    {
      "epoch": 1.6008786752281177,
      "grad_norm": 3.7149479389190674,
      "learning_rate": 2e-05,
      "loss": 0.1917,
      "step": 2369
    },
    {
      "epoch": 1.6015545792497465,
      "grad_norm": 5.900968074798584,
      "learning_rate": 2e-05,
      "loss": 0.2818,
      "step": 2370
    },
    {
      "epoch": 1.6022304832713754,
      "grad_norm": 26.1469669342041,
      "learning_rate": 2e-05,
      "loss": 0.1822,
      "step": 2371
    },
    {
      "epoch": 1.6029063872930043,
      "grad_norm": 21.845806121826172,
      "learning_rate": 2e-05,
      "loss": 0.2432,
      "step": 2372
    },
    {
      "epoch": 1.6035822913146334,
      "grad_norm": 7.149702548980713,
      "learning_rate": 2e-05,
      "loss": 0.2075,
      "step": 2373
    },
    {
      "epoch": 1.6042581953362622,
      "grad_norm": 8.07692813873291,
      "learning_rate": 2e-05,
      "loss": 0.1673,
      "step": 2374
    },
    {
      "epoch": 1.6049340993578913,
      "grad_norm": 4.245861530303955,
      "learning_rate": 2e-05,
      "loss": 0.1606,
      "step": 2375
    },
    {
      "epoch": 1.6056100033795202,
      "grad_norm": 3.7724454402923584,
      "learning_rate": 2e-05,
      "loss": 0.1329,
      "step": 2376
    },
    {
      "epoch": 1.606285907401149,
      "grad_norm": 6.395638465881348,
      "learning_rate": 2e-05,
      "loss": 0.1634,
      "step": 2377
    },
    {
      "epoch": 1.606961811422778,
      "grad_norm": 7.544825077056885,
      "learning_rate": 2e-05,
      "loss": 0.2056,
      "step": 2378
    },
    {
      "epoch": 1.6076377154444068,
      "grad_norm": 4.146425724029541,
      "learning_rate": 2e-05,
      "loss": 0.1902,
      "step": 2379
    },
    {
      "epoch": 1.6083136194660357,
      "grad_norm": 7.0751166343688965,
      "learning_rate": 2e-05,
      "loss": 0.1565,
      "step": 2380
    },
    {
      "epoch": 1.6089895234876648,
      "grad_norm": 4.2245988845825195,
      "learning_rate": 2e-05,
      "loss": 0.1732,
      "step": 2381
    },
    {
      "epoch": 1.6096654275092936,
      "grad_norm": 7.011125564575195,
      "learning_rate": 2e-05,
      "loss": 0.2293,
      "step": 2382
    },
    {
      "epoch": 1.6103413315309227,
      "grad_norm": 4.2283244132995605,
      "learning_rate": 2e-05,
      "loss": 0.2676,
      "step": 2383
    },
    {
      "epoch": 1.6110172355525516,
      "grad_norm": 11.525164604187012,
      "learning_rate": 2e-05,
      "loss": 0.2033,
      "step": 2384
    },
    {
      "epoch": 1.6116931395741805,
      "grad_norm": 6.459043502807617,
      "learning_rate": 2e-05,
      "loss": 0.1689,
      "step": 2385
    },
    {
      "epoch": 1.6123690435958093,
      "grad_norm": 9.029086112976074,
      "learning_rate": 2e-05,
      "loss": 0.1844,
      "step": 2386
    },
    {
      "epoch": 1.6130449476174382,
      "grad_norm": 5.073009967803955,
      "learning_rate": 2e-05,
      "loss": 0.1735,
      "step": 2387
    },
    {
      "epoch": 1.613720851639067,
      "grad_norm": 6.158257484436035,
      "learning_rate": 2e-05,
      "loss": 0.1777,
      "step": 2388
    },
    {
      "epoch": 1.6143967556606962,
      "grad_norm": 4.28267765045166,
      "learning_rate": 2e-05,
      "loss": 0.2343,
      "step": 2389
    },
    {
      "epoch": 1.6150726596823253,
      "grad_norm": 6.429845809936523,
      "learning_rate": 2e-05,
      "loss": 0.1752,
      "step": 2390
    },
    {
      "epoch": 1.6157485637039541,
      "grad_norm": 5.980356693267822,
      "learning_rate": 2e-05,
      "loss": 0.1604,
      "step": 2391
    },
    {
      "epoch": 1.616424467725583,
      "grad_norm": 7.064152240753174,
      "learning_rate": 2e-05,
      "loss": 0.1803,
      "step": 2392
    },
    {
      "epoch": 1.6171003717472119,
      "grad_norm": 5.489639759063721,
      "learning_rate": 2e-05,
      "loss": 0.1953,
      "step": 2393
    },
    {
      "epoch": 1.6177762757688408,
      "grad_norm": 7.328545570373535,
      "learning_rate": 2e-05,
      "loss": 0.1894,
      "step": 2394
    },
    {
      "epoch": 1.6184521797904696,
      "grad_norm": 8.030374526977539,
      "learning_rate": 2e-05,
      "loss": 0.1887,
      "step": 2395
    },
    {
      "epoch": 1.6191280838120987,
      "grad_norm": 8.570148468017578,
      "learning_rate": 2e-05,
      "loss": 0.1407,
      "step": 2396
    },
    {
      "epoch": 1.6198039878337276,
      "grad_norm": 7.231975555419922,
      "learning_rate": 2e-05,
      "loss": 0.1572,
      "step": 2397
    },
    {
      "epoch": 1.6204798918553567,
      "grad_norm": 15.542350769042969,
      "learning_rate": 2e-05,
      "loss": 0.2034,
      "step": 2398
    },
    {
      "epoch": 1.6211557958769856,
      "grad_norm": 3.5745925903320312,
      "learning_rate": 2e-05,
      "loss": 0.1637,
      "step": 2399
    },
    {
      "epoch": 1.6218316998986144,
      "grad_norm": 5.485713958740234,
      "learning_rate": 2e-05,
      "loss": 0.1366,
      "step": 2400
    },
    {
      "epoch": 1.6218316998986144,
      "eval_accuracy": 0.5775760135135135,
      "eval_loss": 0.25444385409355164,
      "eval_runtime": 43.0225,
      "eval_samples_per_second": 550.41,
      "eval_steps_per_second": 17.2,
      "step": 2400
    },
    {
      "epoch": 1.6225076039202433,
      "grad_norm": 8.351386070251465,
      "learning_rate": 2e-05,
      "loss": 0.1532,
      "step": 2401
    },
    {
      "epoch": 1.6231835079418722,
      "grad_norm": 6.4900126457214355,
      "learning_rate": 2e-05,
      "loss": 0.2029,
      "step": 2402
    },
    {
      "epoch": 1.623859411963501,
      "grad_norm": 5.751043319702148,
      "learning_rate": 2e-05,
      "loss": 0.1974,
      "step": 2403
    },
    {
      "epoch": 1.6245353159851301,
      "grad_norm": 6.958613395690918,
      "learning_rate": 2e-05,
      "loss": 0.1602,
      "step": 2404
    },
    {
      "epoch": 1.625211220006759,
      "grad_norm": 4.356684684753418,
      "learning_rate": 2e-05,
      "loss": 0.2185,
      "step": 2405
    },
    {
      "epoch": 1.625887124028388,
      "grad_norm": 4.376003742218018,
      "learning_rate": 2e-05,
      "loss": 0.1564,
      "step": 2406
    },
    {
      "epoch": 1.626563028050017,
      "grad_norm": 5.476803302764893,
      "learning_rate": 2e-05,
      "loss": 0.1809,
      "step": 2407
    },
    {
      "epoch": 1.6272389320716458,
      "grad_norm": 5.447032928466797,
      "learning_rate": 2e-05,
      "loss": 0.2167,
      "step": 2408
    },
    {
      "epoch": 1.6279148360932747,
      "grad_norm": 11.803844451904297,
      "learning_rate": 2e-05,
      "loss": 0.1908,
      "step": 2409
    },
    {
      "epoch": 1.6285907401149036,
      "grad_norm": 8.88444709777832,
      "learning_rate": 2e-05,
      "loss": 0.2308,
      "step": 2410
    },
    {
      "epoch": 1.6292666441365327,
      "grad_norm": 5.55924129486084,
      "learning_rate": 2e-05,
      "loss": 0.2141,
      "step": 2411
    },
    {
      "epoch": 1.6299425481581615,
      "grad_norm": 5.141617774963379,
      "learning_rate": 2e-05,
      "loss": 0.1886,
      "step": 2412
    },
    {
      "epoch": 1.6306184521797906,
      "grad_norm": 5.5594682693481445,
      "learning_rate": 2e-05,
      "loss": 0.1623,
      "step": 2413
    },
    {
      "epoch": 1.6312943562014195,
      "grad_norm": 4.058574199676514,
      "learning_rate": 2e-05,
      "loss": 0.2053,
      "step": 2414
    },
    {
      "epoch": 1.6319702602230484,
      "grad_norm": 8.939290046691895,
      "learning_rate": 2e-05,
      "loss": 0.1545,
      "step": 2415
    },
    {
      "epoch": 1.6326461642446772,
      "grad_norm": 7.376232624053955,
      "learning_rate": 2e-05,
      "loss": 0.1922,
      "step": 2416
    },
    {
      "epoch": 1.633322068266306,
      "grad_norm": 3.845019817352295,
      "learning_rate": 2e-05,
      "loss": 0.1259,
      "step": 2417
    },
    {
      "epoch": 1.633997972287935,
      "grad_norm": 6.508897304534912,
      "learning_rate": 2e-05,
      "loss": 0.1534,
      "step": 2418
    },
    {
      "epoch": 1.634673876309564,
      "grad_norm": 5.805614948272705,
      "learning_rate": 2e-05,
      "loss": 0.1372,
      "step": 2419
    },
    {
      "epoch": 1.635349780331193,
      "grad_norm": 5.080332279205322,
      "learning_rate": 2e-05,
      "loss": 0.1612,
      "step": 2420
    },
    {
      "epoch": 1.636025684352822,
      "grad_norm": 5.265460968017578,
      "learning_rate": 2e-05,
      "loss": 0.1386,
      "step": 2421
    },
    {
      "epoch": 1.636701588374451,
      "grad_norm": 8.879932403564453,
      "learning_rate": 2e-05,
      "loss": 0.2919,
      "step": 2422
    },
    {
      "epoch": 1.6373774923960798,
      "grad_norm": 15.19747257232666,
      "learning_rate": 2e-05,
      "loss": 0.2168,
      "step": 2423
    },
    {
      "epoch": 1.6380533964177086,
      "grad_norm": 5.382063865661621,
      "learning_rate": 2e-05,
      "loss": 0.2269,
      "step": 2424
    },
    {
      "epoch": 1.6387293004393375,
      "grad_norm": 6.144845008850098,
      "learning_rate": 2e-05,
      "loss": 0.142,
      "step": 2425
    },
    {
      "epoch": 1.6394052044609664,
      "grad_norm": 7.953306198120117,
      "learning_rate": 2e-05,
      "loss": 0.2067,
      "step": 2426
    },
    {
      "epoch": 1.6400811084825955,
      "grad_norm": 5.251932144165039,
      "learning_rate": 2e-05,
      "loss": 0.1661,
      "step": 2427
    },
    {
      "epoch": 1.6407570125042246,
      "grad_norm": 3.259294033050537,
      "learning_rate": 2e-05,
      "loss": 0.1357,
      "step": 2428
    },
    {
      "epoch": 1.6414329165258534,
      "grad_norm": 5.3510422706604,
      "learning_rate": 2e-05,
      "loss": 0.171,
      "step": 2429
    },
    {
      "epoch": 1.6421088205474823,
      "grad_norm": 13.025080680847168,
      "learning_rate": 2e-05,
      "loss": 0.2044,
      "step": 2430
    },
    {
      "epoch": 1.6427847245691112,
      "grad_norm": 4.298712730407715,
      "learning_rate": 2e-05,
      "loss": 0.212,
      "step": 2431
    },
    {
      "epoch": 1.64346062859074,
      "grad_norm": 10.400077819824219,
      "learning_rate": 2e-05,
      "loss": 0.1599,
      "step": 2432
    },
    {
      "epoch": 1.644136532612369,
      "grad_norm": 4.480478763580322,
      "learning_rate": 2e-05,
      "loss": 0.1552,
      "step": 2433
    },
    {
      "epoch": 1.644812436633998,
      "grad_norm": 8.666173934936523,
      "learning_rate": 2e-05,
      "loss": 0.1842,
      "step": 2434
    },
    {
      "epoch": 1.6454883406556269,
      "grad_norm": 7.5277791023254395,
      "learning_rate": 2e-05,
      "loss": 0.1499,
      "step": 2435
    },
    {
      "epoch": 1.646164244677256,
      "grad_norm": 10.284749984741211,
      "learning_rate": 2e-05,
      "loss": 0.1451,
      "step": 2436
    },
    {
      "epoch": 1.6468401486988848,
      "grad_norm": 3.6647725105285645,
      "learning_rate": 2e-05,
      "loss": 0.1293,
      "step": 2437
    },
    {
      "epoch": 1.6475160527205137,
      "grad_norm": 5.029379844665527,
      "learning_rate": 2e-05,
      "loss": 0.1737,
      "step": 2438
    },
    {
      "epoch": 1.6481919567421426,
      "grad_norm": 8.175917625427246,
      "learning_rate": 2e-05,
      "loss": 0.2293,
      "step": 2439
    },
    {
      "epoch": 1.6488678607637715,
      "grad_norm": 12.828500747680664,
      "learning_rate": 2e-05,
      "loss": 0.169,
      "step": 2440
    },
    {
      "epoch": 1.6495437647854003,
      "grad_norm": 6.131080627441406,
      "learning_rate": 2e-05,
      "loss": 0.1486,
      "step": 2441
    },
    {
      "epoch": 1.6502196688070294,
      "grad_norm": 5.607979774475098,
      "learning_rate": 2e-05,
      "loss": 0.2068,
      "step": 2442
    },
    {
      "epoch": 1.6508955728286583,
      "grad_norm": 6.850038051605225,
      "learning_rate": 2e-05,
      "loss": 0.3038,
      "step": 2443
    },
    {
      "epoch": 1.6515714768502874,
      "grad_norm": 7.675298690795898,
      "learning_rate": 2e-05,
      "loss": 0.1558,
      "step": 2444
    },
    {
      "epoch": 1.6522473808719162,
      "grad_norm": 5.793602466583252,
      "learning_rate": 2e-05,
      "loss": 0.2255,
      "step": 2445
    },
    {
      "epoch": 1.6529232848935451,
      "grad_norm": 10.947349548339844,
      "learning_rate": 2e-05,
      "loss": 0.1694,
      "step": 2446
    },
    {
      "epoch": 1.653599188915174,
      "grad_norm": 10.88366413116455,
      "learning_rate": 2e-05,
      "loss": 0.1854,
      "step": 2447
    },
    {
      "epoch": 1.6542750929368029,
      "grad_norm": 4.187343597412109,
      "learning_rate": 2e-05,
      "loss": 0.2017,
      "step": 2448
    },
    {
      "epoch": 1.654950996958432,
      "grad_norm": 4.29161262512207,
      "learning_rate": 2e-05,
      "loss": 0.2116,
      "step": 2449
    },
    {
      "epoch": 1.6556269009800608,
      "grad_norm": 4.166978359222412,
      "learning_rate": 2e-05,
      "loss": 0.16,
      "step": 2450
    },
    {
      "epoch": 1.6556269009800608,
      "eval_accuracy": 0.5794763513513513,
      "eval_loss": 0.24034349620342255,
      "eval_runtime": 43.2893,
      "eval_samples_per_second": 547.017,
      "eval_steps_per_second": 17.094,
      "step": 2450
    },
    {
      "epoch": 1.65630280500169,
      "grad_norm": 4.183224678039551,
      "learning_rate": 2e-05,
      "loss": 0.1767,
      "step": 2451
    },
    {
      "epoch": 1.6569787090233188,
      "grad_norm": 5.305939674377441,
      "learning_rate": 2e-05,
      "loss": 0.1423,
      "step": 2452
    },
    {
      "epoch": 1.6576546130449477,
      "grad_norm": 4.5353851318359375,
      "learning_rate": 2e-05,
      "loss": 0.1959,
      "step": 2453
    },
    {
      "epoch": 1.6583305170665765,
      "grad_norm": 6.4164958000183105,
      "learning_rate": 2e-05,
      "loss": 0.22,
      "step": 2454
    },
    {
      "epoch": 1.6590064210882054,
      "grad_norm": 9.458227157592773,
      "learning_rate": 2e-05,
      "loss": 0.1712,
      "step": 2455
    },
    {
      "epoch": 1.6596823251098343,
      "grad_norm": 8.647174835205078,
      "learning_rate": 2e-05,
      "loss": 0.1736,
      "step": 2456
    },
    {
      "epoch": 1.6603582291314634,
      "grad_norm": 12.20089054107666,
      "learning_rate": 2e-05,
      "loss": 0.2062,
      "step": 2457
    },
    {
      "epoch": 1.6610341331530922,
      "grad_norm": 11.11092472076416,
      "learning_rate": 2e-05,
      "loss": 0.2691,
      "step": 2458
    },
    {
      "epoch": 1.6617100371747213,
      "grad_norm": 6.341793537139893,
      "learning_rate": 2e-05,
      "loss": 0.2195,
      "step": 2459
    },
    {
      "epoch": 1.6623859411963502,
      "grad_norm": 5.966394901275635,
      "learning_rate": 2e-05,
      "loss": 0.2116,
      "step": 2460
    },
    {
      "epoch": 1.663061845217979,
      "grad_norm": 12.409872055053711,
      "learning_rate": 2e-05,
      "loss": 0.2366,
      "step": 2461
    },
    {
      "epoch": 1.663737749239608,
      "grad_norm": 7.669351577758789,
      "learning_rate": 2e-05,
      "loss": 0.1175,
      "step": 2462
    },
    {
      "epoch": 1.6644136532612368,
      "grad_norm": 10.877948760986328,
      "learning_rate": 2e-05,
      "loss": 0.2118,
      "step": 2463
    },
    {
      "epoch": 1.6650895572828657,
      "grad_norm": 11.667335510253906,
      "learning_rate": 2e-05,
      "loss": 0.1507,
      "step": 2464
    },
    {
      "epoch": 1.6657654613044948,
      "grad_norm": 5.127872467041016,
      "learning_rate": 2e-05,
      "loss": 0.2057,
      "step": 2465
    },
    {
      "epoch": 1.6664413653261236,
      "grad_norm": 9.405195236206055,
      "learning_rate": 2e-05,
      "loss": 0.212,
      "step": 2466
    },
    {
      "epoch": 1.6671172693477527,
      "grad_norm": 13.1704740524292,
      "learning_rate": 2e-05,
      "loss": 0.2302,
      "step": 2467
    },
    {
      "epoch": 1.6677931733693816,
      "grad_norm": 11.675824165344238,
      "learning_rate": 2e-05,
      "loss": 0.1725,
      "step": 2468
    },
    {
      "epoch": 1.6684690773910105,
      "grad_norm": 22.56087303161621,
      "learning_rate": 2e-05,
      "loss": 0.261,
      "step": 2469
    },
    {
      "epoch": 1.6691449814126393,
      "grad_norm": 9.754384994506836,
      "learning_rate": 2e-05,
      "loss": 0.1612,
      "step": 2470
    },
    {
      "epoch": 1.6698208854342682,
      "grad_norm": 5.7620062828063965,
      "learning_rate": 2e-05,
      "loss": 0.1963,
      "step": 2471
    },
    {
      "epoch": 1.6704967894558973,
      "grad_norm": 5.260942459106445,
      "learning_rate": 2e-05,
      "loss": 0.1953,
      "step": 2472
    },
    {
      "epoch": 1.6711726934775262,
      "grad_norm": 8.865323066711426,
      "learning_rate": 2e-05,
      "loss": 0.1895,
      "step": 2473
    },
    {
      "epoch": 1.6718485974991553,
      "grad_norm": 8.427356719970703,
      "learning_rate": 2e-05,
      "loss": 0.2182,
      "step": 2474
    },
    {
      "epoch": 1.6725245015207841,
      "grad_norm": 4.676563262939453,
      "learning_rate": 2e-05,
      "loss": 0.2097,
      "step": 2475
    },
    {
      "epoch": 1.673200405542413,
      "grad_norm": 5.45790433883667,
      "learning_rate": 2e-05,
      "loss": 0.1667,
      "step": 2476
    },
    {
      "epoch": 1.6738763095640419,
      "grad_norm": 11.678303718566895,
      "learning_rate": 2e-05,
      "loss": 0.1786,
      "step": 2477
    },
    {
      "epoch": 1.6745522135856707,
      "grad_norm": 4.3555121421813965,
      "learning_rate": 2e-05,
      "loss": 0.231,
      "step": 2478
    },
    {
      "epoch": 1.6752281176072996,
      "grad_norm": 6.58701229095459,
      "learning_rate": 2e-05,
      "loss": 0.2034,
      "step": 2479
    },
    {
      "epoch": 1.6759040216289287,
      "grad_norm": 5.317763805389404,
      "learning_rate": 2e-05,
      "loss": 0.1734,
      "step": 2480
    },
    {
      "epoch": 1.6765799256505576,
      "grad_norm": 14.798017501831055,
      "learning_rate": 2e-05,
      "loss": 0.1628,
      "step": 2481
    },
    {
      "epoch": 1.6772558296721867,
      "grad_norm": 6.995798110961914,
      "learning_rate": 2e-05,
      "loss": 0.122,
      "step": 2482
    },
    {
      "epoch": 1.6779317336938155,
      "grad_norm": 6.18882417678833,
      "learning_rate": 2e-05,
      "loss": 0.189,
      "step": 2483
    },
    {
      "epoch": 1.6786076377154444,
      "grad_norm": 10.648630142211914,
      "learning_rate": 2e-05,
      "loss": 0.1772,
      "step": 2484
    },
    {
      "epoch": 1.6792835417370733,
      "grad_norm": 5.636312961578369,
      "learning_rate": 2e-05,
      "loss": 0.2209,
      "step": 2485
    },
    {
      "epoch": 1.6799594457587022,
      "grad_norm": 6.773881435394287,
      "learning_rate": 2e-05,
      "loss": 0.2698,
      "step": 2486
    },
    {
      "epoch": 1.6806353497803312,
      "grad_norm": 15.515212059020996,
      "learning_rate": 2e-05,
      "loss": 0.1928,
      "step": 2487
    },
    {
      "epoch": 1.6813112538019601,
      "grad_norm": 6.560619354248047,
      "learning_rate": 2e-05,
      "loss": 0.1405,
      "step": 2488
    },
    {
      "epoch": 1.6819871578235892,
      "grad_norm": 4.492429256439209,
      "learning_rate": 2e-05,
      "loss": 0.2389,
      "step": 2489
    },
    {
      "epoch": 1.682663061845218,
      "grad_norm": 5.49231481552124,
      "learning_rate": 2e-05,
      "loss": 0.1946,
      "step": 2490
    },
    {
      "epoch": 1.683338965866847,
      "grad_norm": 6.298854351043701,
      "learning_rate": 2e-05,
      "loss": 0.1982,
      "step": 2491
    },
    {
      "epoch": 1.6840148698884758,
      "grad_norm": 6.166478157043457,
      "learning_rate": 2e-05,
      "loss": 0.2091,
      "step": 2492
    },
    {
      "epoch": 1.6846907739101047,
      "grad_norm": 4.0110764503479,
      "learning_rate": 2e-05,
      "loss": 0.2167,
      "step": 2493
    },
    {
      "epoch": 1.6853666779317336,
      "grad_norm": 4.374052047729492,
      "learning_rate": 2e-05,
      "loss": 0.2075,
      "step": 2494
    },
    {
      "epoch": 1.6860425819533627,
      "grad_norm": 4.271756172180176,
      "learning_rate": 2e-05,
      "loss": 0.165,
      "step": 2495
    },
    {
      "epoch": 1.6867184859749915,
      "grad_norm": 4.718948841094971,
      "learning_rate": 2e-05,
      "loss": 0.2013,
      "step": 2496
    },
    {
      "epoch": 1.6873943899966206,
      "grad_norm": 5.281561374664307,
      "learning_rate": 2e-05,
      "loss": 0.1644,
      "step": 2497
    },
    {
      "epoch": 1.6880702940182495,
      "grad_norm": 8.68126392364502,
      "learning_rate": 2e-05,
      "loss": 0.1488,
      "step": 2498
    },
    {
      "epoch": 1.6887461980398784,
      "grad_norm": 4.877569675445557,
      "learning_rate": 2e-05,
      "loss": 0.1274,
      "step": 2499
    },
    {
      "epoch": 1.6894221020615072,
      "grad_norm": 6.515623092651367,
      "learning_rate": 2e-05,
      "loss": 0.1924,
      "step": 2500
    },
    {
      "epoch": 1.6894221020615072,
      "eval_accuracy": 0.5816300675675675,
      "eval_loss": 0.23791509866714478,
      "eval_runtime": 43.1598,
      "eval_samples_per_second": 548.658,
      "eval_steps_per_second": 17.146,
      "step": 2500
    },
    {
      "epoch": 1.690098006083136,
      "grad_norm": 19.544435501098633,
      "learning_rate": 2e-05,
      "loss": 0.2173,
      "step": 2501
    },
    {
      "epoch": 1.690773910104765,
      "grad_norm": 6.0716776847839355,
      "learning_rate": 2e-05,
      "loss": 0.1879,
      "step": 2502
    },
    {
      "epoch": 1.691449814126394,
      "grad_norm": 7.121850967407227,
      "learning_rate": 2e-05,
      "loss": 0.1744,
      "step": 2503
    },
    {
      "epoch": 1.692125718148023,
      "grad_norm": 6.930884838104248,
      "learning_rate": 2e-05,
      "loss": 0.2054,
      "step": 2504
    },
    {
      "epoch": 1.692801622169652,
      "grad_norm": 8.176604270935059,
      "learning_rate": 2e-05,
      "loss": 0.2127,
      "step": 2505
    },
    {
      "epoch": 1.693477526191281,
      "grad_norm": 8.76045036315918,
      "learning_rate": 2e-05,
      "loss": 0.2107,
      "step": 2506
    },
    {
      "epoch": 1.6941534302129098,
      "grad_norm": 3.854976177215576,
      "learning_rate": 2e-05,
      "loss": 0.1335,
      "step": 2507
    },
    {
      "epoch": 1.6948293342345386,
      "grad_norm": 4.059969902038574,
      "learning_rate": 2e-05,
      "loss": 0.1813,
      "step": 2508
    },
    {
      "epoch": 1.6955052382561675,
      "grad_norm": 4.398321628570557,
      "learning_rate": 2e-05,
      "loss": 0.1584,
      "step": 2509
    },
    {
      "epoch": 1.6961811422777966,
      "grad_norm": 5.424098014831543,
      "learning_rate": 2e-05,
      "loss": 0.1947,
      "step": 2510
    },
    {
      "epoch": 1.6968570462994255,
      "grad_norm": 8.433823585510254,
      "learning_rate": 2e-05,
      "loss": 0.191,
      "step": 2511
    },
    {
      "epoch": 1.6975329503210546,
      "grad_norm": 4.464845657348633,
      "learning_rate": 2e-05,
      "loss": 0.1577,
      "step": 2512
    },
    {
      "epoch": 1.6982088543426834,
      "grad_norm": 14.834308624267578,
      "learning_rate": 2e-05,
      "loss": 0.2291,
      "step": 2513
    },
    {
      "epoch": 1.6988847583643123,
      "grad_norm": 7.633622646331787,
      "learning_rate": 2e-05,
      "loss": 0.1991,
      "step": 2514
    },
    {
      "epoch": 1.6995606623859412,
      "grad_norm": 7.648365020751953,
      "learning_rate": 2e-05,
      "loss": 0.1844,
      "step": 2515
    },
    {
      "epoch": 1.70023656640757,
      "grad_norm": 7.556852340698242,
      "learning_rate": 2e-05,
      "loss": 0.179,
      "step": 2516
    },
    {
      "epoch": 1.700912470429199,
      "grad_norm": 8.534584045410156,
      "learning_rate": 2e-05,
      "loss": 0.1339,
      "step": 2517
    },
    {
      "epoch": 1.701588374450828,
      "grad_norm": 5.528745174407959,
      "learning_rate": 2e-05,
      "loss": 0.175,
      "step": 2518
    },
    {
      "epoch": 1.7022642784724569,
      "grad_norm": 4.877151012420654,
      "learning_rate": 2e-05,
      "loss": 0.1928,
      "step": 2519
    },
    {
      "epoch": 1.702940182494086,
      "grad_norm": 5.051514625549316,
      "learning_rate": 2e-05,
      "loss": 0.1091,
      "step": 2520
    },
    {
      "epoch": 1.7036160865157148,
      "grad_norm": 4.397736549377441,
      "learning_rate": 2e-05,
      "loss": 0.2328,
      "step": 2521
    },
    {
      "epoch": 1.7042919905373437,
      "grad_norm": 10.45970344543457,
      "learning_rate": 2e-05,
      "loss": 0.1733,
      "step": 2522
    },
    {
      "epoch": 1.7049678945589726,
      "grad_norm": 6.983922481536865,
      "learning_rate": 2e-05,
      "loss": 0.2309,
      "step": 2523
    },
    {
      "epoch": 1.7056437985806014,
      "grad_norm": 6.789629936218262,
      "learning_rate": 2e-05,
      "loss": 0.1801,
      "step": 2524
    },
    {
      "epoch": 1.7063197026022305,
      "grad_norm": 4.981929302215576,
      "learning_rate": 2e-05,
      "loss": 0.184,
      "step": 2525
    },
    {
      "epoch": 1.7069956066238594,
      "grad_norm": 5.496044635772705,
      "learning_rate": 2e-05,
      "loss": 0.1973,
      "step": 2526
    },
    {
      "epoch": 1.7076715106454885,
      "grad_norm": 6.198970317840576,
      "learning_rate": 2e-05,
      "loss": 0.2099,
      "step": 2527
    },
    {
      "epoch": 1.7083474146671174,
      "grad_norm": 3.071101427078247,
      "learning_rate": 2e-05,
      "loss": 0.1632,
      "step": 2528
    },
    {
      "epoch": 1.7090233186887462,
      "grad_norm": 9.70447063446045,
      "learning_rate": 2e-05,
      "loss": 0.199,
      "step": 2529
    },
    {
      "epoch": 1.709699222710375,
      "grad_norm": 7.225060939788818,
      "learning_rate": 2e-05,
      "loss": 0.2071,
      "step": 2530
    },
    {
      "epoch": 1.710375126732004,
      "grad_norm": 4.797160625457764,
      "learning_rate": 2e-05,
      "loss": 0.1516,
      "step": 2531
    },
    {
      "epoch": 1.7110510307536329,
      "grad_norm": 3.6961772441864014,
      "learning_rate": 2e-05,
      "loss": 0.1771,
      "step": 2532
    },
    {
      "epoch": 1.711726934775262,
      "grad_norm": 6.2434892654418945,
      "learning_rate": 2e-05,
      "loss": 0.2195,
      "step": 2533
    },
    {
      "epoch": 1.7124028387968908,
      "grad_norm": 5.427141189575195,
      "learning_rate": 2e-05,
      "loss": 0.242,
      "step": 2534
    },
    {
      "epoch": 1.71307874281852,
      "grad_norm": 3.419060468673706,
      "learning_rate": 2e-05,
      "loss": 0.2017,
      "step": 2535
    },
    {
      "epoch": 1.7137546468401488,
      "grad_norm": 7.359076976776123,
      "learning_rate": 2e-05,
      "loss": 0.182,
      "step": 2536
    },
    {
      "epoch": 1.7144305508617776,
      "grad_norm": 7.003704071044922,
      "learning_rate": 2e-05,
      "loss": 0.2078,
      "step": 2537
    },
    {
      "epoch": 1.7151064548834065,
      "grad_norm": 8.134764671325684,
      "learning_rate": 2e-05,
      "loss": 0.1192,
      "step": 2538
    },
    {
      "epoch": 1.7157823589050354,
      "grad_norm": 7.914211273193359,
      "learning_rate": 2e-05,
      "loss": 0.1982,
      "step": 2539
    },
    {
      "epoch": 1.7164582629266643,
      "grad_norm": 7.769547462463379,
      "learning_rate": 2e-05,
      "loss": 0.1889,
      "step": 2540
    },
    {
      "epoch": 1.7171341669482934,
      "grad_norm": 6.228601455688477,
      "learning_rate": 2e-05,
      "loss": 0.1589,
      "step": 2541
    },
    {
      "epoch": 1.7178100709699222,
      "grad_norm": 7.32451868057251,
      "learning_rate": 2e-05,
      "loss": 0.2502,
      "step": 2542
    },
    {
      "epoch": 1.7184859749915513,
      "grad_norm": 8.49647045135498,
      "learning_rate": 2e-05,
      "loss": 0.2158,
      "step": 2543
    },
    {
      "epoch": 1.7191618790131802,
      "grad_norm": 14.597912788391113,
      "learning_rate": 2e-05,
      "loss": 0.1785,
      "step": 2544
    },
    {
      "epoch": 1.719837783034809,
      "grad_norm": 4.766815662384033,
      "learning_rate": 2e-05,
      "loss": 0.2365,
      "step": 2545
    },
    {
      "epoch": 1.720513687056438,
      "grad_norm": 5.222998142242432,
      "learning_rate": 2e-05,
      "loss": 0.2133,
      "step": 2546
    },
    {
      "epoch": 1.7211895910780668,
      "grad_norm": 11.843677520751953,
      "learning_rate": 2e-05,
      "loss": 0.1971,
      "step": 2547
    },
    {
      "epoch": 1.7218654950996959,
      "grad_norm": 14.084156036376953,
      "learning_rate": 2e-05,
      "loss": 0.1978,
      "step": 2548
    },
    {
      "epoch": 1.7225413991213248,
      "grad_norm": 11.864409446716309,
      "learning_rate": 2e-05,
      "loss": 0.1962,
      "step": 2549
    },
    {
      "epoch": 1.7232173031429538,
      "grad_norm": 9.647034645080566,
      "learning_rate": 2e-05,
      "loss": 0.2342,
      "step": 2550
    },
    {
      "epoch": 1.7232173031429538,
      "eval_accuracy": 0.5771537162162163,
      "eval_loss": 0.24552273750305176,
      "eval_runtime": 43.9551,
      "eval_samples_per_second": 538.731,
      "eval_steps_per_second": 16.835,
      "step": 2550
    },
    {
      "epoch": 1.7238932071645827,
      "grad_norm": 11.45572280883789,
      "learning_rate": 2e-05,
      "loss": 0.1919,
      "step": 2551
    },
    {
      "epoch": 1.7245691111862116,
      "grad_norm": 7.440885066986084,
      "learning_rate": 2e-05,
      "loss": 0.25,
      "step": 2552
    },
    {
      "epoch": 1.7252450152078405,
      "grad_norm": 14.133171081542969,
      "learning_rate": 2e-05,
      "loss": 0.2205,
      "step": 2553
    },
    {
      "epoch": 1.7259209192294693,
      "grad_norm": 6.328035831451416,
      "learning_rate": 2e-05,
      "loss": 0.1957,
      "step": 2554
    },
    {
      "epoch": 1.7265968232510982,
      "grad_norm": 3.8738412857055664,
      "learning_rate": 2e-05,
      "loss": 0.2262,
      "step": 2555
    },
    {
      "epoch": 1.7272727272727273,
      "grad_norm": 11.071885108947754,
      "learning_rate": 2e-05,
      "loss": 0.1985,
      "step": 2556
    },
    {
      "epoch": 1.7279486312943562,
      "grad_norm": 5.160275459289551,
      "learning_rate": 2e-05,
      "loss": 0.1762,
      "step": 2557
    },
    {
      "epoch": 1.7286245353159853,
      "grad_norm": 8.434906005859375,
      "learning_rate": 2e-05,
      "loss": 0.1473,
      "step": 2558
    },
    {
      "epoch": 1.7293004393376141,
      "grad_norm": 4.939137935638428,
      "learning_rate": 2e-05,
      "loss": 0.2053,
      "step": 2559
    },
    {
      "epoch": 1.729976343359243,
      "grad_norm": 3.0579586029052734,
      "learning_rate": 2e-05,
      "loss": 0.1995,
      "step": 2560
    },
    {
      "epoch": 1.7306522473808719,
      "grad_norm": 4.308902740478516,
      "learning_rate": 2e-05,
      "loss": 0.2232,
      "step": 2561
    },
    {
      "epoch": 1.7313281514025007,
      "grad_norm": 3.50070858001709,
      "learning_rate": 2e-05,
      "loss": 0.1729,
      "step": 2562
    },
    {
      "epoch": 1.7320040554241296,
      "grad_norm": 11.545621871948242,
      "learning_rate": 2e-05,
      "loss": 0.1812,
      "step": 2563
    },
    {
      "epoch": 1.7326799594457587,
      "grad_norm": 10.04864501953125,
      "learning_rate": 2e-05,
      "loss": 0.1618,
      "step": 2564
    },
    {
      "epoch": 1.7333558634673878,
      "grad_norm": 5.869135856628418,
      "learning_rate": 2e-05,
      "loss": 0.2285,
      "step": 2565
    },
    {
      "epoch": 1.7340317674890167,
      "grad_norm": 12.521825790405273,
      "learning_rate": 2e-05,
      "loss": 0.1964,
      "step": 2566
    },
    {
      "epoch": 1.7347076715106455,
      "grad_norm": 4.224028587341309,
      "learning_rate": 2e-05,
      "loss": 0.2252,
      "step": 2567
    },
    {
      "epoch": 1.7353835755322744,
      "grad_norm": 9.328328132629395,
      "learning_rate": 2e-05,
      "loss": 0.2283,
      "step": 2568
    },
    {
      "epoch": 1.7360594795539033,
      "grad_norm": 4.190976619720459,
      "learning_rate": 2e-05,
      "loss": 0.1795,
      "step": 2569
    },
    {
      "epoch": 1.7367353835755321,
      "grad_norm": 16.466846466064453,
      "learning_rate": 2e-05,
      "loss": 0.1965,
      "step": 2570
    },
    {
      "epoch": 1.7374112875971612,
      "grad_norm": 3.198859930038452,
      "learning_rate": 2e-05,
      "loss": 0.1366,
      "step": 2571
    },
    {
      "epoch": 1.73808719161879,
      "grad_norm": 3.350399971008301,
      "learning_rate": 2e-05,
      "loss": 0.1767,
      "step": 2572
    },
    {
      "epoch": 1.7387630956404192,
      "grad_norm": 5.573922634124756,
      "learning_rate": 2e-05,
      "loss": 0.1751,
      "step": 2573
    },
    {
      "epoch": 1.739438999662048,
      "grad_norm": 8.478436470031738,
      "learning_rate": 2e-05,
      "loss": 0.235,
      "step": 2574
    },
    {
      "epoch": 1.740114903683677,
      "grad_norm": 5.708214282989502,
      "learning_rate": 2e-05,
      "loss": 0.2061,
      "step": 2575
    },
    {
      "epoch": 1.7407908077053058,
      "grad_norm": 10.476367950439453,
      "learning_rate": 2e-05,
      "loss": 0.1837,
      "step": 2576
    },
    {
      "epoch": 1.7414667117269347,
      "grad_norm": 5.433126926422119,
      "learning_rate": 2e-05,
      "loss": 0.1568,
      "step": 2577
    },
    {
      "epoch": 1.7421426157485635,
      "grad_norm": 5.976624011993408,
      "learning_rate": 2e-05,
      "loss": 0.2144,
      "step": 2578
    },
    {
      "epoch": 1.7428185197701926,
      "grad_norm": 6.455981731414795,
      "learning_rate": 2e-05,
      "loss": 0.1662,
      "step": 2579
    },
    {
      "epoch": 1.7434944237918215,
      "grad_norm": 3.8413801193237305,
      "learning_rate": 2e-05,
      "loss": 0.1332,
      "step": 2580
    },
    {
      "epoch": 1.7441703278134506,
      "grad_norm": 4.399928092956543,
      "learning_rate": 2e-05,
      "loss": 0.2017,
      "step": 2581
    },
    {
      "epoch": 1.7448462318350795,
      "grad_norm": 5.0574750900268555,
      "learning_rate": 2e-05,
      "loss": 0.1538,
      "step": 2582
    },
    {
      "epoch": 1.7455221358567083,
      "grad_norm": 7.296647548675537,
      "learning_rate": 2e-05,
      "loss": 0.1913,
      "step": 2583
    },
    {
      "epoch": 1.7461980398783372,
      "grad_norm": 6.979553699493408,
      "learning_rate": 2e-05,
      "loss": 0.2033,
      "step": 2584
    },
    {
      "epoch": 1.746873943899966,
      "grad_norm": 4.836071491241455,
      "learning_rate": 2e-05,
      "loss": 0.1178,
      "step": 2585
    },
    {
      "epoch": 1.7475498479215952,
      "grad_norm": 6.946887016296387,
      "learning_rate": 2e-05,
      "loss": 0.2256,
      "step": 2586
    },
    {
      "epoch": 1.748225751943224,
      "grad_norm": 8.422430992126465,
      "learning_rate": 2e-05,
      "loss": 0.1867,
      "step": 2587
    },
    {
      "epoch": 1.7489016559648531,
      "grad_norm": 4.7318949699401855,
      "learning_rate": 2e-05,
      "loss": 0.1295,
      "step": 2588
    },
    {
      "epoch": 1.749577559986482,
      "grad_norm": 7.3790202140808105,
      "learning_rate": 2e-05,
      "loss": 0.2021,
      "step": 2589
    },
    {
      "epoch": 1.7502534640081109,
      "grad_norm": 3.1768226623535156,
      "learning_rate": 2e-05,
      "loss": 0.1499,
      "step": 2590
    },
    {
      "epoch": 1.7509293680297398,
      "grad_norm": 11.089823722839355,
      "learning_rate": 2e-05,
      "loss": 0.2196,
      "step": 2591
    },
    {
      "epoch": 1.7516052720513686,
      "grad_norm": 9.016037940979004,
      "learning_rate": 2e-05,
      "loss": 0.21,
      "step": 2592
    },
    {
      "epoch": 1.7522811760729975,
      "grad_norm": 7.374248027801514,
      "learning_rate": 2e-05,
      "loss": 0.1713,
      "step": 2593
    },
    {
      "epoch": 1.7529570800946266,
      "grad_norm": 9.101129531860352,
      "learning_rate": 2e-05,
      "loss": 0.1887,
      "step": 2594
    },
    {
      "epoch": 1.7536329841162555,
      "grad_norm": 3.322144031524658,
      "learning_rate": 2e-05,
      "loss": 0.2014,
      "step": 2595
    },
    {
      "epoch": 1.7543088881378845,
      "grad_norm": 10.312385559082031,
      "learning_rate": 2e-05,
      "loss": 0.2007,
      "step": 2596
    },
    {
      "epoch": 1.7549847921595134,
      "grad_norm": 13.90938949584961,
      "learning_rate": 2e-05,
      "loss": 0.1971,
      "step": 2597
    },
    {
      "epoch": 1.7556606961811423,
      "grad_norm": 7.390103816986084,
      "learning_rate": 2e-05,
      "loss": 0.2246,
      "step": 2598
    },
    {
      "epoch": 1.7563366002027712,
      "grad_norm": 9.669831275939941,
      "learning_rate": 2e-05,
      "loss": 0.277,
      "step": 2599
    },
    {
      "epoch": 1.7570125042244,
      "grad_norm": 9.848329544067383,
      "learning_rate": 2e-05,
      "loss": 0.1077,
      "step": 2600
    },
    {
      "epoch": 1.7570125042244,
      "eval_accuracy": 0.5880912162162162,
      "eval_loss": 0.23444630205631256,
      "eval_runtime": 43.553,
      "eval_samples_per_second": 543.706,
      "eval_steps_per_second": 16.991,
      "step": 2600
    },
    {
      "epoch": 1.757688408246029,
      "grad_norm": 4.029735088348389,
      "learning_rate": 2e-05,
      "loss": 0.2017,
      "step": 2601
    },
    {
      "epoch": 1.758364312267658,
      "grad_norm": 7.763247489929199,
      "learning_rate": 2e-05,
      "loss": 0.1698,
      "step": 2602
    },
    {
      "epoch": 1.759040216289287,
      "grad_norm": 12.335209846496582,
      "learning_rate": 2e-05,
      "loss": 0.1824,
      "step": 2603
    },
    {
      "epoch": 1.759716120310916,
      "grad_norm": 13.87010383605957,
      "learning_rate": 2e-05,
      "loss": 0.2335,
      "step": 2604
    },
    {
      "epoch": 1.7603920243325448,
      "grad_norm": 3.8047091960906982,
      "learning_rate": 2e-05,
      "loss": 0.1944,
      "step": 2605
    },
    {
      "epoch": 1.7610679283541737,
      "grad_norm": 7.6428117752075195,
      "learning_rate": 2e-05,
      "loss": 0.0983,
      "step": 2606
    },
    {
      "epoch": 1.7617438323758026,
      "grad_norm": 5.194156169891357,
      "learning_rate": 2e-05,
      "loss": 0.2027,
      "step": 2607
    },
    {
      "epoch": 1.7624197363974314,
      "grad_norm": 6.164571285247803,
      "learning_rate": 2e-05,
      "loss": 0.1655,
      "step": 2608
    },
    {
      "epoch": 1.7630956404190605,
      "grad_norm": 5.961998462677002,
      "learning_rate": 2e-05,
      "loss": 0.1631,
      "step": 2609
    },
    {
      "epoch": 1.7637715444406894,
      "grad_norm": 4.3097004890441895,
      "learning_rate": 2e-05,
      "loss": 0.149,
      "step": 2610
    },
    {
      "epoch": 1.7644474484623185,
      "grad_norm": 5.784277439117432,
      "learning_rate": 2e-05,
      "loss": 0.1863,
      "step": 2611
    },
    {
      "epoch": 1.7651233524839474,
      "grad_norm": 3.3257055282592773,
      "learning_rate": 2e-05,
      "loss": 0.1536,
      "step": 2612
    },
    {
      "epoch": 1.7657992565055762,
      "grad_norm": 7.754985332489014,
      "learning_rate": 2e-05,
      "loss": 0.1514,
      "step": 2613
    },
    {
      "epoch": 1.766475160527205,
      "grad_norm": 7.481513023376465,
      "learning_rate": 2e-05,
      "loss": 0.2389,
      "step": 2614
    },
    {
      "epoch": 1.767151064548834,
      "grad_norm": 21.713146209716797,
      "learning_rate": 2e-05,
      "loss": 0.2382,
      "step": 2615
    },
    {
      "epoch": 1.7678269685704628,
      "grad_norm": 6.970318794250488,
      "learning_rate": 2e-05,
      "loss": 0.1837,
      "step": 2616
    },
    {
      "epoch": 1.768502872592092,
      "grad_norm": 4.511751651763916,
      "learning_rate": 2e-05,
      "loss": 0.2093,
      "step": 2617
    },
    {
      "epoch": 1.7691787766137208,
      "grad_norm": 6.783756256103516,
      "learning_rate": 2e-05,
      "loss": 0.1517,
      "step": 2618
    },
    {
      "epoch": 1.76985468063535,
      "grad_norm": 4.149144649505615,
      "learning_rate": 2e-05,
      "loss": 0.1731,
      "step": 2619
    },
    {
      "epoch": 1.7705305846569788,
      "grad_norm": 4.369739532470703,
      "learning_rate": 2e-05,
      "loss": 0.1943,
      "step": 2620
    },
    {
      "epoch": 1.7712064886786076,
      "grad_norm": 8.759808540344238,
      "learning_rate": 2e-05,
      "loss": 0.2076,
      "step": 2621
    },
    {
      "epoch": 1.7718823927002365,
      "grad_norm": 11.403414726257324,
      "learning_rate": 2e-05,
      "loss": 0.1741,
      "step": 2622
    },
    {
      "epoch": 1.7725582967218654,
      "grad_norm": 15.274518966674805,
      "learning_rate": 2e-05,
      "loss": 0.1795,
      "step": 2623
    },
    {
      "epoch": 1.7732342007434945,
      "grad_norm": 9.883648872375488,
      "learning_rate": 2e-05,
      "loss": 0.2174,
      "step": 2624
    },
    {
      "epoch": 1.7739101047651233,
      "grad_norm": 9.530634880065918,
      "learning_rate": 2e-05,
      "loss": 0.1599,
      "step": 2625
    },
    {
      "epoch": 1.7745860087867524,
      "grad_norm": 17.3511962890625,
      "learning_rate": 2e-05,
      "loss": 0.2101,
      "step": 2626
    },
    {
      "epoch": 1.7752619128083813,
      "grad_norm": 5.638111114501953,
      "learning_rate": 2e-05,
      "loss": 0.2374,
      "step": 2627
    },
    {
      "epoch": 1.7759378168300102,
      "grad_norm": 22.273069381713867,
      "learning_rate": 2e-05,
      "loss": 0.2467,
      "step": 2628
    },
    {
      "epoch": 1.776613720851639,
      "grad_norm": 17.633724212646484,
      "learning_rate": 2e-05,
      "loss": 0.1924,
      "step": 2629
    },
    {
      "epoch": 1.777289624873268,
      "grad_norm": 5.024192810058594,
      "learning_rate": 2e-05,
      "loss": 0.2158,
      "step": 2630
    },
    {
      "epoch": 1.7779655288948968,
      "grad_norm": 6.539625644683838,
      "learning_rate": 2e-05,
      "loss": 0.1579,
      "step": 2631
    },
    {
      "epoch": 1.7786414329165259,
      "grad_norm": 10.131094932556152,
      "learning_rate": 2e-05,
      "loss": 0.179,
      "step": 2632
    },
    {
      "epoch": 1.7793173369381547,
      "grad_norm": 15.439739227294922,
      "learning_rate": 2e-05,
      "loss": 0.2452,
      "step": 2633
    },
    {
      "epoch": 1.7799932409597838,
      "grad_norm": 17.14769744873047,
      "learning_rate": 2e-05,
      "loss": 0.1795,
      "step": 2634
    },
    {
      "epoch": 1.7806691449814127,
      "grad_norm": 8.714799880981445,
      "learning_rate": 2e-05,
      "loss": 0.1201,
      "step": 2635
    },
    {
      "epoch": 1.7813450490030416,
      "grad_norm": 6.651059150695801,
      "learning_rate": 2e-05,
      "loss": 0.1783,
      "step": 2636
    },
    {
      "epoch": 1.7820209530246705,
      "grad_norm": 4.382108211517334,
      "learning_rate": 2e-05,
      "loss": 0.1672,
      "step": 2637
    },
    {
      "epoch": 1.7826968570462993,
      "grad_norm": 8.549869537353516,
      "learning_rate": 2e-05,
      "loss": 0.163,
      "step": 2638
    },
    {
      "epoch": 1.7833727610679282,
      "grad_norm": 4.106595993041992,
      "learning_rate": 2e-05,
      "loss": 0.1791,
      "step": 2639
    },
    {
      "epoch": 1.7840486650895573,
      "grad_norm": 7.812234401702881,
      "learning_rate": 2e-05,
      "loss": 0.176,
      "step": 2640
    },
    {
      "epoch": 1.7847245691111864,
      "grad_norm": 9.703205108642578,
      "learning_rate": 2e-05,
      "loss": 0.1683,
      "step": 2641
    },
    {
      "epoch": 1.7854004731328152,
      "grad_norm": 11.245566368103027,
      "learning_rate": 2e-05,
      "loss": 0.2019,
      "step": 2642
    },
    {
      "epoch": 1.7860763771544441,
      "grad_norm": 6.093090534210205,
      "learning_rate": 2e-05,
      "loss": 0.1559,
      "step": 2643
    },
    {
      "epoch": 1.786752281176073,
      "grad_norm": 7.6890411376953125,
      "learning_rate": 2e-05,
      "loss": 0.0913,
      "step": 2644
    },
    {
      "epoch": 1.7874281851977019,
      "grad_norm": 4.711739540100098,
      "learning_rate": 2e-05,
      "loss": 0.2222,
      "step": 2645
    },
    {
      "epoch": 1.7881040892193307,
      "grad_norm": 5.371128559112549,
      "learning_rate": 2e-05,
      "loss": 0.2222,
      "step": 2646
    },
    {
      "epoch": 1.7887799932409598,
      "grad_norm": 15.231616020202637,
      "learning_rate": 2e-05,
      "loss": 0.161,
      "step": 2647
    },
    {
      "epoch": 1.7894558972625887,
      "grad_norm": 6.385985851287842,
      "learning_rate": 2e-05,
      "loss": 0.1038,
      "step": 2648
    },
    {
      "epoch": 1.7901318012842178,
      "grad_norm": 5.480828285217285,
      "learning_rate": 2e-05,
      "loss": 0.1592,
      "step": 2649
    },
    {
      "epoch": 1.7908077053058467,
      "grad_norm": 9.38731861114502,
      "learning_rate": 2e-05,
      "loss": 0.2621,
      "step": 2650
    },
    {
      "epoch": 1.7908077053058467,
      "eval_accuracy": 0.5980996621621621,
      "eval_loss": 0.24836421012878418,
      "eval_runtime": 43.5141,
      "eval_samples_per_second": 544.192,
      "eval_steps_per_second": 17.006,
      "step": 2650
    },
    {
      "epoch": 1.7914836093274755,
      "grad_norm": 11.399660110473633,
      "learning_rate": 2e-05,
      "loss": 0.1404,
      "step": 2651
    },
    {
      "epoch": 1.7921595133491044,
      "grad_norm": 3.3008532524108887,
      "learning_rate": 2e-05,
      "loss": 0.1804,
      "step": 2652
    },
    {
      "epoch": 1.7928354173707333,
      "grad_norm": 9.726880073547363,
      "learning_rate": 2e-05,
      "loss": 0.1574,
      "step": 2653
    },
    {
      "epoch": 1.7935113213923621,
      "grad_norm": 5.608015537261963,
      "learning_rate": 2e-05,
      "loss": 0.1957,
      "step": 2654
    },
    {
      "epoch": 1.7941872254139912,
      "grad_norm": 10.68464469909668,
      "learning_rate": 2e-05,
      "loss": 0.2086,
      "step": 2655
    },
    {
      "epoch": 1.79486312943562,
      "grad_norm": 11.644718170166016,
      "learning_rate": 2e-05,
      "loss": 0.171,
      "step": 2656
    },
    {
      "epoch": 1.7955390334572492,
      "grad_norm": 6.091114044189453,
      "learning_rate": 2e-05,
      "loss": 0.186,
      "step": 2657
    },
    {
      "epoch": 1.796214937478878,
      "grad_norm": 5.484583377838135,
      "learning_rate": 2e-05,
      "loss": 0.1786,
      "step": 2658
    },
    {
      "epoch": 1.796890841500507,
      "grad_norm": 6.156371116638184,
      "learning_rate": 2e-05,
      "loss": 0.1619,
      "step": 2659
    },
    {
      "epoch": 1.7975667455221358,
      "grad_norm": 7.2929158210754395,
      "learning_rate": 2e-05,
      "loss": 0.1548,
      "step": 2660
    },
    {
      "epoch": 1.7982426495437647,
      "grad_norm": 8.586711883544922,
      "learning_rate": 2e-05,
      "loss": 0.2011,
      "step": 2661
    },
    {
      "epoch": 1.7989185535653938,
      "grad_norm": 3.7596936225891113,
      "learning_rate": 2e-05,
      "loss": 0.1752,
      "step": 2662
    },
    {
      "epoch": 1.7995944575870226,
      "grad_norm": 9.687914848327637,
      "learning_rate": 2e-05,
      "loss": 0.1644,
      "step": 2663
    },
    {
      "epoch": 1.8002703616086517,
      "grad_norm": 5.594640254974365,
      "learning_rate": 2e-05,
      "loss": 0.2223,
      "step": 2664
    },
    {
      "epoch": 1.8009462656302806,
      "grad_norm": 11.177318572998047,
      "learning_rate": 2e-05,
      "loss": 0.212,
      "step": 2665
    },
    {
      "epoch": 1.8016221696519095,
      "grad_norm": 5.25739860534668,
      "learning_rate": 2e-05,
      "loss": 0.1696,
      "step": 2666
    },
    {
      "epoch": 1.8022980736735383,
      "grad_norm": 7.418591022491455,
      "learning_rate": 2e-05,
      "loss": 0.1724,
      "step": 2667
    },
    {
      "epoch": 1.8029739776951672,
      "grad_norm": 5.855809211730957,
      "learning_rate": 2e-05,
      "loss": 0.1862,
      "step": 2668
    },
    {
      "epoch": 1.803649881716796,
      "grad_norm": 5.494976997375488,
      "learning_rate": 2e-05,
      "loss": 0.1681,
      "step": 2669
    },
    {
      "epoch": 1.8043257857384252,
      "grad_norm": 11.779149055480957,
      "learning_rate": 2e-05,
      "loss": 0.223,
      "step": 2670
    },
    {
      "epoch": 1.805001689760054,
      "grad_norm": 12.60892391204834,
      "learning_rate": 2e-05,
      "loss": 0.1852,
      "step": 2671
    },
    {
      "epoch": 1.8056775937816831,
      "grad_norm": 5.007839202880859,
      "learning_rate": 2e-05,
      "loss": 0.2275,
      "step": 2672
    },
    {
      "epoch": 1.806353497803312,
      "grad_norm": 6.869328022003174,
      "learning_rate": 2e-05,
      "loss": 0.1362,
      "step": 2673
    },
    {
      "epoch": 1.8070294018249409,
      "grad_norm": 3.29062819480896,
      "learning_rate": 2e-05,
      "loss": 0.154,
      "step": 2674
    },
    {
      "epoch": 1.8077053058465697,
      "grad_norm": 4.839600563049316,
      "learning_rate": 2e-05,
      "loss": 0.1658,
      "step": 2675
    },
    {
      "epoch": 1.8083812098681986,
      "grad_norm": 3.7732577323913574,
      "learning_rate": 2e-05,
      "loss": 0.1708,
      "step": 2676
    },
    {
      "epoch": 1.8090571138898275,
      "grad_norm": 3.999107599258423,
      "learning_rate": 2e-05,
      "loss": 0.1018,
      "step": 2677
    },
    {
      "epoch": 1.8097330179114566,
      "grad_norm": 12.22292423248291,
      "learning_rate": 2e-05,
      "loss": 0.2094,
      "step": 2678
    },
    {
      "epoch": 1.8104089219330854,
      "grad_norm": 3.9914746284484863,
      "learning_rate": 2e-05,
      "loss": 0.1211,
      "step": 2679
    },
    {
      "epoch": 1.8110848259547145,
      "grad_norm": 4.68793249130249,
      "learning_rate": 2e-05,
      "loss": 0.1749,
      "step": 2680
    },
    {
      "epoch": 1.8117607299763434,
      "grad_norm": 6.0852813720703125,
      "learning_rate": 2e-05,
      "loss": 0.1469,
      "step": 2681
    },
    {
      "epoch": 1.8124366339979723,
      "grad_norm": 6.017002105712891,
      "learning_rate": 2e-05,
      "loss": 0.1756,
      "step": 2682
    },
    {
      "epoch": 1.8131125380196011,
      "grad_norm": 8.435280799865723,
      "learning_rate": 2e-05,
      "loss": 0.148,
      "step": 2683
    },
    {
      "epoch": 1.81378844204123,
      "grad_norm": 7.500074863433838,
      "learning_rate": 2e-05,
      "loss": 0.1975,
      "step": 2684
    },
    {
      "epoch": 1.8144643460628591,
      "grad_norm": 5.351524353027344,
      "learning_rate": 2e-05,
      "loss": 0.1947,
      "step": 2685
    },
    {
      "epoch": 1.815140250084488,
      "grad_norm": 15.804051399230957,
      "learning_rate": 2e-05,
      "loss": 0.1431,
      "step": 2686
    },
    {
      "epoch": 1.815816154106117,
      "grad_norm": 5.6042351722717285,
      "learning_rate": 2e-05,
      "loss": 0.1092,
      "step": 2687
    },
    {
      "epoch": 1.816492058127746,
      "grad_norm": 6.975046634674072,
      "learning_rate": 2e-05,
      "loss": 0.1423,
      "step": 2688
    },
    {
      "epoch": 1.8171679621493748,
      "grad_norm": 6.7935895919799805,
      "learning_rate": 2e-05,
      "loss": 0.1965,
      "step": 2689
    },
    {
      "epoch": 1.8178438661710037,
      "grad_norm": 7.772512912750244,
      "learning_rate": 2e-05,
      "loss": 0.222,
      "step": 2690
    },
    {
      "epoch": 1.8185197701926326,
      "grad_norm": 8.189723014831543,
      "learning_rate": 2e-05,
      "loss": 0.2791,
      "step": 2691
    },
    {
      "epoch": 1.8191956742142614,
      "grad_norm": 9.871321678161621,
      "learning_rate": 2e-05,
      "loss": 0.1742,
      "step": 2692
    },
    {
      "epoch": 1.8198715782358905,
      "grad_norm": 7.352211952209473,
      "learning_rate": 2e-05,
      "loss": 0.2306,
      "step": 2693
    },
    {
      "epoch": 1.8205474822575194,
      "grad_norm": 5.238424777984619,
      "learning_rate": 2e-05,
      "loss": 0.1435,
      "step": 2694
    },
    {
      "epoch": 1.8212233862791485,
      "grad_norm": 7.480871200561523,
      "learning_rate": 2e-05,
      "loss": 0.1517,
      "step": 2695
    },
    {
      "epoch": 1.8218992903007774,
      "grad_norm": 8.36291217803955,
      "learning_rate": 2e-05,
      "loss": 0.1701,
      "step": 2696
    },
    {
      "epoch": 1.8225751943224062,
      "grad_norm": 5.798957824707031,
      "learning_rate": 2e-05,
      "loss": 0.1409,
      "step": 2697
    },
    {
      "epoch": 1.823251098344035,
      "grad_norm": 6.841068744659424,
      "learning_rate": 2e-05,
      "loss": 0.1815,
      "step": 2698
    },
    {
      "epoch": 1.823927002365664,
      "grad_norm": 4.006419658660889,
      "learning_rate": 2e-05,
      "loss": 0.1675,
      "step": 2699
    },
    {
      "epoch": 1.824602906387293,
      "grad_norm": 5.738894939422607,
      "learning_rate": 2e-05,
      "loss": 0.2347,
      "step": 2700
    },
    {
      "epoch": 1.824602906387293,
      "eval_accuracy": 0.5678631756756757,
      "eval_loss": 0.24507129192352295,
      "eval_runtime": 42.9337,
      "eval_samples_per_second": 551.548,
      "eval_steps_per_second": 17.236,
      "step": 2700
    },
    {
      "epoch": 1.825278810408922,
      "grad_norm": 5.894851207733154,
      "learning_rate": 2e-05,
      "loss": 0.1825,
      "step": 2701
    },
    {
      "epoch": 1.825954714430551,
      "grad_norm": 3.277672290802002,
      "learning_rate": 2e-05,
      "loss": 0.1892,
      "step": 2702
    },
    {
      "epoch": 1.8266306184521799,
      "grad_norm": 5.616495609283447,
      "learning_rate": 2e-05,
      "loss": 0.2145,
      "step": 2703
    },
    {
      "epoch": 1.8273065224738088,
      "grad_norm": 5.564042091369629,
      "learning_rate": 2e-05,
      "loss": 0.2128,
      "step": 2704
    },
    {
      "epoch": 1.8279824264954376,
      "grad_norm": 6.311061382293701,
      "learning_rate": 2e-05,
      "loss": 0.1284,
      "step": 2705
    },
    {
      "epoch": 1.8286583305170665,
      "grad_norm": 3.4651951789855957,
      "learning_rate": 2e-05,
      "loss": 0.2279,
      "step": 2706
    },
    {
      "epoch": 1.8293342345386954,
      "grad_norm": 6.329956531524658,
      "learning_rate": 2e-05,
      "loss": 0.205,
      "step": 2707
    },
    {
      "epoch": 1.8300101385603245,
      "grad_norm": 8.398343086242676,
      "learning_rate": 2e-05,
      "loss": 0.1759,
      "step": 2708
    },
    {
      "epoch": 1.8306860425819533,
      "grad_norm": 4.207131862640381,
      "learning_rate": 2e-05,
      "loss": 0.1835,
      "step": 2709
    },
    {
      "epoch": 1.8313619466035824,
      "grad_norm": 17.322328567504883,
      "learning_rate": 2e-05,
      "loss": 0.2283,
      "step": 2710
    },
    {
      "epoch": 1.8320378506252113,
      "grad_norm": 11.650321960449219,
      "learning_rate": 2e-05,
      "loss": 0.1918,
      "step": 2711
    },
    {
      "epoch": 1.8327137546468402,
      "grad_norm": 5.175717830657959,
      "learning_rate": 2e-05,
      "loss": 0.2317,
      "step": 2712
    },
    {
      "epoch": 1.833389658668469,
      "grad_norm": 7.719907283782959,
      "learning_rate": 2e-05,
      "loss": 0.1824,
      "step": 2713
    },
    {
      "epoch": 1.834065562690098,
      "grad_norm": 4.088478088378906,
      "learning_rate": 2e-05,
      "loss": 0.1285,
      "step": 2714
    },
    {
      "epoch": 1.8347414667117268,
      "grad_norm": 4.4212775230407715,
      "learning_rate": 2e-05,
      "loss": 0.1764,
      "step": 2715
    },
    {
      "epoch": 1.8354173707333559,
      "grad_norm": 4.863746166229248,
      "learning_rate": 2e-05,
      "loss": 0.2006,
      "step": 2716
    },
    {
      "epoch": 1.8360932747549847,
      "grad_norm": 15.684491157531738,
      "learning_rate": 2e-05,
      "loss": 0.2007,
      "step": 2717
    },
    {
      "epoch": 1.8367691787766138,
      "grad_norm": 12.90668773651123,
      "learning_rate": 2e-05,
      "loss": 0.1954,
      "step": 2718
    },
    {
      "epoch": 1.8374450827982427,
      "grad_norm": 16.494592666625977,
      "learning_rate": 2e-05,
      "loss": 0.173,
      "step": 2719
    },
    {
      "epoch": 1.8381209868198716,
      "grad_norm": 4.371163845062256,
      "learning_rate": 2e-05,
      "loss": 0.1582,
      "step": 2720
    },
    {
      "epoch": 1.8387968908415004,
      "grad_norm": 5.406315803527832,
      "learning_rate": 2e-05,
      "loss": 0.165,
      "step": 2721
    },
    {
      "epoch": 1.8394727948631293,
      "grad_norm": 4.790228843688965,
      "learning_rate": 2e-05,
      "loss": 0.1576,
      "step": 2722
    },
    {
      "epoch": 1.8401486988847584,
      "grad_norm": 12.593172073364258,
      "learning_rate": 2e-05,
      "loss": 0.1851,
      "step": 2723
    },
    {
      "epoch": 1.8408246029063873,
      "grad_norm": 10.618154525756836,
      "learning_rate": 2e-05,
      "loss": 0.2513,
      "step": 2724
    },
    {
      "epoch": 1.8415005069280164,
      "grad_norm": 13.46878433227539,
      "learning_rate": 2e-05,
      "loss": 0.1849,
      "step": 2725
    },
    {
      "epoch": 1.8421764109496452,
      "grad_norm": 5.81688928604126,
      "learning_rate": 2e-05,
      "loss": 0.1761,
      "step": 2726
    },
    {
      "epoch": 1.842852314971274,
      "grad_norm": 6.340455055236816,
      "learning_rate": 2e-05,
      "loss": 0.3092,
      "step": 2727
    },
    {
      "epoch": 1.843528218992903,
      "grad_norm": 14.016443252563477,
      "learning_rate": 2e-05,
      "loss": 0.2011,
      "step": 2728
    },
    {
      "epoch": 1.8442041230145318,
      "grad_norm": 4.903017520904541,
      "learning_rate": 2e-05,
      "loss": 0.1945,
      "step": 2729
    },
    {
      "epoch": 1.8448800270361607,
      "grad_norm": 12.618764877319336,
      "learning_rate": 2e-05,
      "loss": 0.1964,
      "step": 2730
    },
    {
      "epoch": 1.8455559310577898,
      "grad_norm": 10.985270500183105,
      "learning_rate": 2e-05,
      "loss": 0.2342,
      "step": 2731
    },
    {
      "epoch": 1.8462318350794187,
      "grad_norm": 4.410651206970215,
      "learning_rate": 2e-05,
      "loss": 0.1153,
      "step": 2732
    },
    {
      "epoch": 1.8469077391010478,
      "grad_norm": 3.5916385650634766,
      "learning_rate": 2e-05,
      "loss": 0.2104,
      "step": 2733
    },
    {
      "epoch": 1.8475836431226766,
      "grad_norm": 5.499990940093994,
      "learning_rate": 2e-05,
      "loss": 0.2373,
      "step": 2734
    },
    {
      "epoch": 1.8482595471443055,
      "grad_norm": 11.053502082824707,
      "learning_rate": 2e-05,
      "loss": 0.2033,
      "step": 2735
    },
    {
      "epoch": 1.8489354511659344,
      "grad_norm": 7.8896565437316895,
      "learning_rate": 2e-05,
      "loss": 0.1719,
      "step": 2736
    },
    {
      "epoch": 1.8496113551875633,
      "grad_norm": 11.607213020324707,
      "learning_rate": 2e-05,
      "loss": 0.1888,
      "step": 2737
    },
    {
      "epoch": 1.8502872592091923,
      "grad_norm": 8.946487426757812,
      "learning_rate": 2e-05,
      "loss": 0.1745,
      "step": 2738
    },
    {
      "epoch": 1.8509631632308212,
      "grad_norm": 4.203725337982178,
      "learning_rate": 2e-05,
      "loss": 0.2985,
      "step": 2739
    },
    {
      "epoch": 1.8516390672524503,
      "grad_norm": 3.9979352951049805,
      "learning_rate": 2e-05,
      "loss": 0.1559,
      "step": 2740
    },
    {
      "epoch": 1.8523149712740792,
      "grad_norm": 5.383571147918701,
      "learning_rate": 2e-05,
      "loss": 0.1586,
      "step": 2741
    },
    {
      "epoch": 1.852990875295708,
      "grad_norm": 11.859116554260254,
      "learning_rate": 2e-05,
      "loss": 0.2211,
      "step": 2742
    },
    {
      "epoch": 1.853666779317337,
      "grad_norm": 5.734777927398682,
      "learning_rate": 2e-05,
      "loss": 0.1876,
      "step": 2743
    },
    {
      "epoch": 1.8543426833389658,
      "grad_norm": 7.260355472564697,
      "learning_rate": 2e-05,
      "loss": 0.1836,
      "step": 2744
    },
    {
      "epoch": 1.8550185873605947,
      "grad_norm": 3.752039670944214,
      "learning_rate": 2e-05,
      "loss": 0.1473,
      "step": 2745
    },
    {
      "epoch": 1.8556944913822238,
      "grad_norm": 7.485644817352295,
      "learning_rate": 2e-05,
      "loss": 0.1788,
      "step": 2746
    },
    {
      "epoch": 1.8563703954038526,
      "grad_norm": 6.429305553436279,
      "learning_rate": 2e-05,
      "loss": 0.1601,
      "step": 2747
    },
    {
      "epoch": 1.8570462994254817,
      "grad_norm": 3.2717812061309814,
      "learning_rate": 2e-05,
      "loss": 0.1449,
      "step": 2748
    },
    {
      "epoch": 1.8577222034471106,
      "grad_norm": 8.036136627197266,
      "learning_rate": 2e-05,
      "loss": 0.2086,
      "step": 2749
    },
    {
      "epoch": 1.8583981074687395,
      "grad_norm": 6.824588298797607,
      "learning_rate": 2e-05,
      "loss": 0.2036,
      "step": 2750
    },
    {
      "epoch": 1.8583981074687395,
      "eval_accuracy": 0.5482685810810811,
      "eval_loss": 0.24389486014842987,
      "eval_runtime": 43.4386,
      "eval_samples_per_second": 545.137,
      "eval_steps_per_second": 17.036,
      "step": 2750
    },
    {
      "epoch": 1.8590740114903683,
      "grad_norm": 12.68576431274414,
      "learning_rate": 2e-05,
      "loss": 0.2031,
      "step": 2751
    },
    {
      "epoch": 1.8597499155119972,
      "grad_norm": 10.93404483795166,
      "learning_rate": 2e-05,
      "loss": 0.1856,
      "step": 2752
    },
    {
      "epoch": 1.860425819533626,
      "grad_norm": 3.636402130126953,
      "learning_rate": 2e-05,
      "loss": 0.1678,
      "step": 2753
    },
    {
      "epoch": 1.8611017235552552,
      "grad_norm": 5.8929643630981445,
      "learning_rate": 2e-05,
      "loss": 0.2303,
      "step": 2754
    },
    {
      "epoch": 1.861777627576884,
      "grad_norm": 5.686343193054199,
      "learning_rate": 2e-05,
      "loss": 0.1317,
      "step": 2755
    },
    {
      "epoch": 1.8624535315985131,
      "grad_norm": 5.60164737701416,
      "learning_rate": 2e-05,
      "loss": 0.1755,
      "step": 2756
    },
    {
      "epoch": 1.863129435620142,
      "grad_norm": 3.9800546169281006,
      "learning_rate": 2e-05,
      "loss": 0.1859,
      "step": 2757
    },
    {
      "epoch": 1.8638053396417709,
      "grad_norm": 6.842212200164795,
      "learning_rate": 2e-05,
      "loss": 0.1759,
      "step": 2758
    },
    {
      "epoch": 1.8644812436633997,
      "grad_norm": 8.406274795532227,
      "learning_rate": 2e-05,
      "loss": 0.2012,
      "step": 2759
    },
    {
      "epoch": 1.8651571476850286,
      "grad_norm": 14.377900123596191,
      "learning_rate": 2e-05,
      "loss": 0.2222,
      "step": 2760
    },
    {
      "epoch": 1.8658330517066577,
      "grad_norm": 12.718677520751953,
      "learning_rate": 2e-05,
      "loss": 0.1863,
      "step": 2761
    },
    {
      "epoch": 1.8665089557282866,
      "grad_norm": 5.260409832000732,
      "learning_rate": 2e-05,
      "loss": 0.2003,
      "step": 2762
    },
    {
      "epoch": 1.8671848597499157,
      "grad_norm": 12.591012954711914,
      "learning_rate": 2e-05,
      "loss": 0.2407,
      "step": 2763
    },
    {
      "epoch": 1.8678607637715445,
      "grad_norm": 7.632625579833984,
      "learning_rate": 2e-05,
      "loss": 0.1766,
      "step": 2764
    },
    {
      "epoch": 1.8685366677931734,
      "grad_norm": 6.503602027893066,
      "learning_rate": 2e-05,
      "loss": 0.2157,
      "step": 2765
    },
    {
      "epoch": 1.8692125718148023,
      "grad_norm": 6.701712608337402,
      "learning_rate": 2e-05,
      "loss": 0.2127,
      "step": 2766
    },
    {
      "epoch": 1.8698884758364311,
      "grad_norm": 6.23762845993042,
      "learning_rate": 2e-05,
      "loss": 0.1377,
      "step": 2767
    },
    {
      "epoch": 1.87056437985806,
      "grad_norm": 4.502397060394287,
      "learning_rate": 2e-05,
      "loss": 0.2298,
      "step": 2768
    },
    {
      "epoch": 1.871240283879689,
      "grad_norm": 6.32965087890625,
      "learning_rate": 2e-05,
      "loss": 0.1791,
      "step": 2769
    },
    {
      "epoch": 1.871916187901318,
      "grad_norm": 2.9488518238067627,
      "learning_rate": 2e-05,
      "loss": 0.1532,
      "step": 2770
    },
    {
      "epoch": 1.872592091922947,
      "grad_norm": 4.1780290603637695,
      "learning_rate": 2e-05,
      "loss": 0.1792,
      "step": 2771
    },
    {
      "epoch": 1.873267995944576,
      "grad_norm": 4.919784069061279,
      "learning_rate": 2e-05,
      "loss": 0.2147,
      "step": 2772
    },
    {
      "epoch": 1.8739438999662048,
      "grad_norm": 7.426290035247803,
      "learning_rate": 2e-05,
      "loss": 0.204,
      "step": 2773
    },
    {
      "epoch": 1.8746198039878337,
      "grad_norm": 9.903227806091309,
      "learning_rate": 2e-05,
      "loss": 0.1586,
      "step": 2774
    },
    {
      "epoch": 1.8752957080094625,
      "grad_norm": 3.4525110721588135,
      "learning_rate": 2e-05,
      "loss": 0.1777,
      "step": 2775
    },
    {
      "epoch": 1.8759716120310916,
      "grad_norm": 4.358468532562256,
      "learning_rate": 2e-05,
      "loss": 0.2093,
      "step": 2776
    },
    {
      "epoch": 1.8766475160527205,
      "grad_norm": 9.612951278686523,
      "learning_rate": 2e-05,
      "loss": 0.2087,
      "step": 2777
    },
    {
      "epoch": 1.8773234200743496,
      "grad_norm": 3.321638584136963,
      "learning_rate": 2e-05,
      "loss": 0.1767,
      "step": 2778
    },
    {
      "epoch": 1.8779993240959785,
      "grad_norm": 5.449265956878662,
      "learning_rate": 2e-05,
      "loss": 0.1965,
      "step": 2779
    },
    {
      "epoch": 1.8786752281176073,
      "grad_norm": 6.67354679107666,
      "learning_rate": 2e-05,
      "loss": 0.2142,
      "step": 2780
    },
    {
      "epoch": 1.8793511321392362,
      "grad_norm": 18.4562931060791,
      "learning_rate": 2e-05,
      "loss": 0.3054,
      "step": 2781
    },
    {
      "epoch": 1.880027036160865,
      "grad_norm": 14.796256065368652,
      "learning_rate": 2e-05,
      "loss": 0.1623,
      "step": 2782
    },
    {
      "epoch": 1.880702940182494,
      "grad_norm": 3.050717830657959,
      "learning_rate": 2e-05,
      "loss": 0.1464,
      "step": 2783
    },
    {
      "epoch": 1.881378844204123,
      "grad_norm": 6.183348178863525,
      "learning_rate": 2e-05,
      "loss": 0.1491,
      "step": 2784
    },
    {
      "epoch": 1.882054748225752,
      "grad_norm": 8.259084701538086,
      "learning_rate": 2e-05,
      "loss": 0.131,
      "step": 2785
    },
    {
      "epoch": 1.882730652247381,
      "grad_norm": 3.8852925300598145,
      "learning_rate": 2e-05,
      "loss": 0.2408,
      "step": 2786
    },
    {
      "epoch": 1.8834065562690099,
      "grad_norm": 13.149214744567871,
      "learning_rate": 2e-05,
      "loss": 0.1659,
      "step": 2787
    },
    {
      "epoch": 1.8840824602906387,
      "grad_norm": 7.442408561706543,
      "learning_rate": 2e-05,
      "loss": 0.2941,
      "step": 2788
    },
    {
      "epoch": 1.8847583643122676,
      "grad_norm": 3.9350929260253906,
      "learning_rate": 2e-05,
      "loss": 0.1194,
      "step": 2789
    },
    {
      "epoch": 1.8854342683338965,
      "grad_norm": 2.4489426612854004,
      "learning_rate": 2e-05,
      "loss": 0.1641,
      "step": 2790
    },
    {
      "epoch": 1.8861101723555254,
      "grad_norm": 8.698687553405762,
      "learning_rate": 2e-05,
      "loss": 0.1436,
      "step": 2791
    },
    {
      "epoch": 1.8867860763771545,
      "grad_norm": 8.73154354095459,
      "learning_rate": 2e-05,
      "loss": 0.1805,
      "step": 2792
    },
    {
      "epoch": 1.8874619803987833,
      "grad_norm": 4.743855953216553,
      "learning_rate": 2e-05,
      "loss": 0.1715,
      "step": 2793
    },
    {
      "epoch": 1.8881378844204124,
      "grad_norm": 10.911948204040527,
      "learning_rate": 2e-05,
      "loss": 0.1769,
      "step": 2794
    },
    {
      "epoch": 1.8888137884420413,
      "grad_norm": 10.998044967651367,
      "learning_rate": 2e-05,
      "loss": 0.1644,
      "step": 2795
    },
    {
      "epoch": 1.8894896924636702,
      "grad_norm": 6.860535621643066,
      "learning_rate": 2e-05,
      "loss": 0.2261,
      "step": 2796
    },
    {
      "epoch": 1.890165596485299,
      "grad_norm": 10.268416404724121,
      "learning_rate": 2e-05,
      "loss": 0.2615,
      "step": 2797
    },
    {
      "epoch": 1.890841500506928,
      "grad_norm": 6.328078746795654,
      "learning_rate": 2e-05,
      "loss": 0.1539,
      "step": 2798
    },
    {
      "epoch": 1.891517404528557,
      "grad_norm": 4.592935562133789,
      "learning_rate": 2e-05,
      "loss": 0.1899,
      "step": 2799
    },
    {
      "epoch": 1.8921933085501859,
      "grad_norm": 5.793074607849121,
      "learning_rate": 2e-05,
      "loss": 0.1826,
      "step": 2800
    },
    {
      "epoch": 1.8921933085501859,
      "eval_accuracy": 0.5891047297297297,
      "eval_loss": 0.2733939588069916,
      "eval_runtime": 43.6539,
      "eval_samples_per_second": 542.448,
      "eval_steps_per_second": 16.952,
      "step": 2800
    },
    {
      "epoch": 1.892869212571815,
      "grad_norm": 12.105826377868652,
      "learning_rate": 2e-05,
      "loss": 0.1823,
      "step": 2801
    },
    {
      "epoch": 1.8935451165934438,
      "grad_norm": 12.35008430480957,
      "learning_rate": 2e-05,
      "loss": 0.2428,
      "step": 2802
    },
    {
      "epoch": 1.8942210206150727,
      "grad_norm": 8.897770881652832,
      "learning_rate": 2e-05,
      "loss": 0.2298,
      "step": 2803
    },
    {
      "epoch": 1.8948969246367016,
      "grad_norm": 3.8769006729125977,
      "learning_rate": 2e-05,
      "loss": 0.1825,
      "step": 2804
    },
    {
      "epoch": 1.8955728286583304,
      "grad_norm": 7.741106033325195,
      "learning_rate": 2e-05,
      "loss": 0.2374,
      "step": 2805
    },
    {
      "epoch": 1.8962487326799593,
      "grad_norm": 5.4894843101501465,
      "learning_rate": 2e-05,
      "loss": 0.1991,
      "step": 2806
    },
    {
      "epoch": 1.8969246367015884,
      "grad_norm": 3.5829946994781494,
      "learning_rate": 2e-05,
      "loss": 0.2102,
      "step": 2807
    },
    {
      "epoch": 1.8976005407232173,
      "grad_norm": 5.0757670402526855,
      "learning_rate": 2e-05,
      "loss": 0.1608,
      "step": 2808
    },
    {
      "epoch": 1.8982764447448464,
      "grad_norm": 4.403623580932617,
      "learning_rate": 2e-05,
      "loss": 0.2436,
      "step": 2809
    },
    {
      "epoch": 1.8989523487664752,
      "grad_norm": 13.676843643188477,
      "learning_rate": 2e-05,
      "loss": 0.1943,
      "step": 2810
    },
    {
      "epoch": 1.899628252788104,
      "grad_norm": 8.326689720153809,
      "learning_rate": 2e-05,
      "loss": 0.1737,
      "step": 2811
    },
    {
      "epoch": 1.900304156809733,
      "grad_norm": 3.8745570182800293,
      "learning_rate": 2e-05,
      "loss": 0.1876,
      "step": 2812
    },
    {
      "epoch": 1.9009800608313618,
      "grad_norm": 11.273282051086426,
      "learning_rate": 2e-05,
      "loss": 0.1992,
      "step": 2813
    },
    {
      "epoch": 1.9016559648529907,
      "grad_norm": 7.559131622314453,
      "learning_rate": 2e-05,
      "loss": 0.183,
      "step": 2814
    },
    {
      "epoch": 1.9023318688746198,
      "grad_norm": 8.616376876831055,
      "learning_rate": 2e-05,
      "loss": 0.2065,
      "step": 2815
    },
    {
      "epoch": 1.903007772896249,
      "grad_norm": 3.000570297241211,
      "learning_rate": 2e-05,
      "loss": 0.1306,
      "step": 2816
    },
    {
      "epoch": 1.9036836769178778,
      "grad_norm": 3.7790160179138184,
      "learning_rate": 2e-05,
      "loss": 0.2062,
      "step": 2817
    },
    {
      "epoch": 1.9043595809395066,
      "grad_norm": 7.347280979156494,
      "learning_rate": 2e-05,
      "loss": 0.1589,
      "step": 2818
    },
    {
      "epoch": 1.9050354849611355,
      "grad_norm": 7.530834197998047,
      "learning_rate": 2e-05,
      "loss": 0.2039,
      "step": 2819
    },
    {
      "epoch": 1.9057113889827644,
      "grad_norm": 8.861859321594238,
      "learning_rate": 2e-05,
      "loss": 0.2167,
      "step": 2820
    },
    {
      "epoch": 1.9063872930043932,
      "grad_norm": 4.127299785614014,
      "learning_rate": 2e-05,
      "loss": 0.1987,
      "step": 2821
    },
    {
      "epoch": 1.9070631970260223,
      "grad_norm": 6.076162815093994,
      "learning_rate": 2e-05,
      "loss": 0.1851,
      "step": 2822
    },
    {
      "epoch": 1.9077391010476512,
      "grad_norm": 4.075569152832031,
      "learning_rate": 2e-05,
      "loss": 0.1813,
      "step": 2823
    },
    {
      "epoch": 1.9084150050692803,
      "grad_norm": 3.855742931365967,
      "learning_rate": 2e-05,
      "loss": 0.1938,
      "step": 2824
    },
    {
      "epoch": 1.9090909090909092,
      "grad_norm": 8.623655319213867,
      "learning_rate": 2e-05,
      "loss": 0.1554,
      "step": 2825
    },
    {
      "epoch": 1.909766813112538,
      "grad_norm": 6.362133502960205,
      "learning_rate": 2e-05,
      "loss": 0.193,
      "step": 2826
    },
    {
      "epoch": 1.910442717134167,
      "grad_norm": 3.6771328449249268,
      "learning_rate": 2e-05,
      "loss": 0.1559,
      "step": 2827
    },
    {
      "epoch": 1.9111186211557958,
      "grad_norm": 3.73022723197937,
      "learning_rate": 2e-05,
      "loss": 0.1599,
      "step": 2828
    },
    {
      "epoch": 1.9117945251774247,
      "grad_norm": 13.279247283935547,
      "learning_rate": 2e-05,
      "loss": 0.36,
      "step": 2829
    },
    {
      "epoch": 1.9124704291990537,
      "grad_norm": 17.65468978881836,
      "learning_rate": 2e-05,
      "loss": 0.2167,
      "step": 2830
    },
    {
      "epoch": 1.9131463332206826,
      "grad_norm": 8.829855918884277,
      "learning_rate": 2e-05,
      "loss": 0.2223,
      "step": 2831
    },
    {
      "epoch": 1.9138222372423117,
      "grad_norm": 6.301202774047852,
      "learning_rate": 2e-05,
      "loss": 0.2354,
      "step": 2832
    },
    {
      "epoch": 1.9144981412639406,
      "grad_norm": 18.643455505371094,
      "learning_rate": 2e-05,
      "loss": 0.2284,
      "step": 2833
    },
    {
      "epoch": 1.9151740452855694,
      "grad_norm": 12.795949935913086,
      "learning_rate": 2e-05,
      "loss": 0.178,
      "step": 2834
    },
    {
      "epoch": 1.9158499493071983,
      "grad_norm": 10.732818603515625,
      "learning_rate": 2e-05,
      "loss": 0.2069,
      "step": 2835
    },
    {
      "epoch": 1.9165258533288272,
      "grad_norm": 3.3544466495513916,
      "learning_rate": 2e-05,
      "loss": 0.1959,
      "step": 2836
    },
    {
      "epoch": 1.9172017573504563,
      "grad_norm": 4.970009803771973,
      "learning_rate": 2e-05,
      "loss": 0.1433,
      "step": 2837
    },
    {
      "epoch": 1.9178776613720852,
      "grad_norm": 4.551453113555908,
      "learning_rate": 2e-05,
      "loss": 0.1659,
      "step": 2838
    },
    {
      "epoch": 1.9185535653937142,
      "grad_norm": 7.480842113494873,
      "learning_rate": 2e-05,
      "loss": 0.1367,
      "step": 2839
    },
    {
      "epoch": 1.9192294694153431,
      "grad_norm": 3.3002588748931885,
      "learning_rate": 2e-05,
      "loss": 0.149,
      "step": 2840
    },
    {
      "epoch": 1.919905373436972,
      "grad_norm": 8.559991836547852,
      "learning_rate": 2e-05,
      "loss": 0.1825,
      "step": 2841
    },
    {
      "epoch": 1.9205812774586009,
      "grad_norm": 11.25338077545166,
      "learning_rate": 2e-05,
      "loss": 0.1673,
      "step": 2842
    },
    {
      "epoch": 1.9212571814802297,
      "grad_norm": 6.881085395812988,
      "learning_rate": 2e-05,
      "loss": 0.1355,
      "step": 2843
    },
    {
      "epoch": 1.9219330855018586,
      "grad_norm": 4.199821472167969,
      "learning_rate": 2e-05,
      "loss": 0.2266,
      "step": 2844
    },
    {
      "epoch": 1.9226089895234877,
      "grad_norm": 8.923846244812012,
      "learning_rate": 2e-05,
      "loss": 0.1533,
      "step": 2845
    },
    {
      "epoch": 1.9232848935451166,
      "grad_norm": 6.2072834968566895,
      "learning_rate": 2e-05,
      "loss": 0.2072,
      "step": 2846
    },
    {
      "epoch": 1.9239607975667457,
      "grad_norm": 7.811628818511963,
      "learning_rate": 2e-05,
      "loss": 0.2453,
      "step": 2847
    },
    {
      "epoch": 1.9246367015883745,
      "grad_norm": 10.847253799438477,
      "learning_rate": 2e-05,
      "loss": 0.1689,
      "step": 2848
    },
    {
      "epoch": 1.9253126056100034,
      "grad_norm": 7.073367118835449,
      "learning_rate": 2e-05,
      "loss": 0.1487,
      "step": 2849
    },
    {
      "epoch": 1.9259885096316323,
      "grad_norm": 4.746853828430176,
      "learning_rate": 2e-05,
      "loss": 0.2061,
      "step": 2850
    },
    {
      "epoch": 1.9259885096316323,
      "eval_accuracy": 0.5817989864864865,
      "eval_loss": 0.24908575415611267,
      "eval_runtime": 43.7922,
      "eval_samples_per_second": 540.736,
      "eval_steps_per_second": 16.898,
      "step": 2850
    },
    {
      "epoch": 1.9266644136532611,
      "grad_norm": 7.5206379890441895,
      "learning_rate": 2e-05,
      "loss": 0.2068,
      "step": 2851
    },
    {
      "epoch": 1.92734031767489,
      "grad_norm": 10.845320701599121,
      "learning_rate": 2e-05,
      "loss": 0.1433,
      "step": 2852
    },
    {
      "epoch": 1.928016221696519,
      "grad_norm": 5.754593372344971,
      "learning_rate": 2e-05,
      "loss": 0.2046,
      "step": 2853
    },
    {
      "epoch": 1.9286921257181482,
      "grad_norm": 4.3944220542907715,
      "learning_rate": 2e-05,
      "loss": 0.2306,
      "step": 2854
    },
    {
      "epoch": 1.929368029739777,
      "grad_norm": 8.105534553527832,
      "learning_rate": 2e-05,
      "loss": 0.1872,
      "step": 2855
    },
    {
      "epoch": 1.930043933761406,
      "grad_norm": 4.804765701293945,
      "learning_rate": 2e-05,
      "loss": 0.171,
      "step": 2856
    },
    {
      "epoch": 1.9307198377830348,
      "grad_norm": 7.661672592163086,
      "learning_rate": 2e-05,
      "loss": 0.1805,
      "step": 2857
    },
    {
      "epoch": 1.9313957418046637,
      "grad_norm": 3.499898672103882,
      "learning_rate": 2e-05,
      "loss": 0.171,
      "step": 2858
    },
    {
      "epoch": 1.9320716458262925,
      "grad_norm": 3.5822105407714844,
      "learning_rate": 2e-05,
      "loss": 0.1343,
      "step": 2859
    },
    {
      "epoch": 1.9327475498479216,
      "grad_norm": 5.272571563720703,
      "learning_rate": 2e-05,
      "loss": 0.2237,
      "step": 2860
    },
    {
      "epoch": 1.9334234538695505,
      "grad_norm": 11.466962814331055,
      "learning_rate": 2e-05,
      "loss": 0.1808,
      "step": 2861
    },
    {
      "epoch": 1.9340993578911796,
      "grad_norm": 8.41195011138916,
      "learning_rate": 2e-05,
      "loss": 0.2356,
      "step": 2862
    },
    {
      "epoch": 1.9347752619128085,
      "grad_norm": 4.572606086730957,
      "learning_rate": 2e-05,
      "loss": 0.1892,
      "step": 2863
    },
    {
      "epoch": 1.9354511659344373,
      "grad_norm": 3.017068862915039,
      "learning_rate": 2e-05,
      "loss": 0.1394,
      "step": 2864
    },
    {
      "epoch": 1.9361270699560662,
      "grad_norm": 5.866125106811523,
      "learning_rate": 2e-05,
      "loss": 0.1524,
      "step": 2865
    },
    {
      "epoch": 1.936802973977695,
      "grad_norm": 7.522221565246582,
      "learning_rate": 2e-05,
      "loss": 0.1569,
      "step": 2866
    },
    {
      "epoch": 1.937478877999324,
      "grad_norm": 4.005246162414551,
      "learning_rate": 2e-05,
      "loss": 0.1581,
      "step": 2867
    },
    {
      "epoch": 1.938154782020953,
      "grad_norm": 8.334013938903809,
      "learning_rate": 2e-05,
      "loss": 0.1679,
      "step": 2868
    },
    {
      "epoch": 1.938830686042582,
      "grad_norm": 5.114103317260742,
      "learning_rate": 2e-05,
      "loss": 0.184,
      "step": 2869
    },
    {
      "epoch": 1.939506590064211,
      "grad_norm": 3.466862440109253,
      "learning_rate": 2e-05,
      "loss": 0.1896,
      "step": 2870
    },
    {
      "epoch": 1.9401824940858399,
      "grad_norm": 3.6039469242095947,
      "learning_rate": 2e-05,
      "loss": 0.1562,
      "step": 2871
    },
    {
      "epoch": 1.9408583981074687,
      "grad_norm": 5.491974830627441,
      "learning_rate": 2e-05,
      "loss": 0.2296,
      "step": 2872
    },
    {
      "epoch": 1.9415343021290976,
      "grad_norm": 9.485787391662598,
      "learning_rate": 2e-05,
      "loss": 0.1385,
      "step": 2873
    },
    {
      "epoch": 1.9422102061507265,
      "grad_norm": 5.98376989364624,
      "learning_rate": 2e-05,
      "loss": 0.1667,
      "step": 2874
    },
    {
      "epoch": 1.9428861101723556,
      "grad_norm": 5.454115867614746,
      "learning_rate": 2e-05,
      "loss": 0.1766,
      "step": 2875
    },
    {
      "epoch": 1.9435620141939844,
      "grad_norm": 5.069123268127441,
      "learning_rate": 2e-05,
      "loss": 0.1773,
      "step": 2876
    },
    {
      "epoch": 1.9442379182156135,
      "grad_norm": 11.89149284362793,
      "learning_rate": 2e-05,
      "loss": 0.2047,
      "step": 2877
    },
    {
      "epoch": 1.9449138222372424,
      "grad_norm": 4.64158821105957,
      "learning_rate": 2e-05,
      "loss": 0.2519,
      "step": 2878
    },
    {
      "epoch": 1.9455897262588713,
      "grad_norm": 5.843343734741211,
      "learning_rate": 2e-05,
      "loss": 0.1527,
      "step": 2879
    },
    {
      "epoch": 1.9462656302805001,
      "grad_norm": 5.756446361541748,
      "learning_rate": 2e-05,
      "loss": 0.1611,
      "step": 2880
    },
    {
      "epoch": 1.946941534302129,
      "grad_norm": 7.921277046203613,
      "learning_rate": 2e-05,
      "loss": 0.2168,
      "step": 2881
    },
    {
      "epoch": 1.9476174383237579,
      "grad_norm": 9.443964958190918,
      "learning_rate": 2e-05,
      "loss": 0.2125,
      "step": 2882
    },
    {
      "epoch": 1.948293342345387,
      "grad_norm": 6.596774578094482,
      "learning_rate": 2e-05,
      "loss": 0.2172,
      "step": 2883
    },
    {
      "epoch": 1.9489692463670159,
      "grad_norm": 9.665492057800293,
      "learning_rate": 2e-05,
      "loss": 0.1841,
      "step": 2884
    },
    {
      "epoch": 1.949645150388645,
      "grad_norm": 8.871766090393066,
      "learning_rate": 2e-05,
      "loss": 0.1646,
      "step": 2885
    },
    {
      "epoch": 1.9503210544102738,
      "grad_norm": 5.621859073638916,
      "learning_rate": 2e-05,
      "loss": 0.1724,
      "step": 2886
    },
    {
      "epoch": 1.9509969584319027,
      "grad_norm": 7.3798136711120605,
      "learning_rate": 2e-05,
      "loss": 0.1697,
      "step": 2887
    },
    {
      "epoch": 1.9516728624535316,
      "grad_norm": 14.888413429260254,
      "learning_rate": 2e-05,
      "loss": 0.2135,
      "step": 2888
    },
    {
      "epoch": 1.9523487664751604,
      "grad_norm": 12.269875526428223,
      "learning_rate": 2e-05,
      "loss": 0.2124,
      "step": 2889
    },
    {
      "epoch": 1.9530246704967893,
      "grad_norm": 3.257439136505127,
      "learning_rate": 2e-05,
      "loss": 0.1702,
      "step": 2890
    },
    {
      "epoch": 1.9537005745184184,
      "grad_norm": 12.326751708984375,
      "learning_rate": 2e-05,
      "loss": 0.2533,
      "step": 2891
    },
    {
      "epoch": 1.9543764785400475,
      "grad_norm": 17.621788024902344,
      "learning_rate": 2e-05,
      "loss": 0.1752,
      "step": 2892
    },
    {
      "epoch": 1.9550523825616763,
      "grad_norm": 4.84025239944458,
      "learning_rate": 2e-05,
      "loss": 0.2343,
      "step": 2893
    },
    {
      "epoch": 1.9557282865833052,
      "grad_norm": 15.434209823608398,
      "learning_rate": 2e-05,
      "loss": 0.2302,
      "step": 2894
    },
    {
      "epoch": 1.956404190604934,
      "grad_norm": 4.919122695922852,
      "learning_rate": 2e-05,
      "loss": 0.1691,
      "step": 2895
    },
    {
      "epoch": 1.957080094626563,
      "grad_norm": 3.271238327026367,
      "learning_rate": 2e-05,
      "loss": 0.176,
      "step": 2896
    },
    {
      "epoch": 1.9577559986481918,
      "grad_norm": 4.000260829925537,
      "learning_rate": 2e-05,
      "loss": 0.1767,
      "step": 2897
    },
    {
      "epoch": 1.958431902669821,
      "grad_norm": 7.548071384429932,
      "learning_rate": 2e-05,
      "loss": 0.2061,
      "step": 2898
    },
    {
      "epoch": 1.9591078066914498,
      "grad_norm": 8.457720756530762,
      "learning_rate": 2e-05,
      "loss": 0.1634,
      "step": 2899
    },
    {
      "epoch": 1.9597837107130789,
      "grad_norm": 11.326650619506836,
      "learning_rate": 2e-05,
      "loss": 0.1984,
      "step": 2900
    },
    {
      "epoch": 1.9597837107130789,
      "eval_accuracy": 0.5546875,
      "eval_loss": 0.24482764303684235,
      "eval_runtime": 43.719,
      "eval_samples_per_second": 541.641,
      "eval_steps_per_second": 16.926,
      "step": 2900
    },
    {
      "epoch": 1.9604596147347078,
      "grad_norm": 8.271347999572754,
      "learning_rate": 2e-05,
      "loss": 0.1647,
      "step": 2901
    },
    {
      "epoch": 1.9611355187563366,
      "grad_norm": 8.007658004760742,
      "learning_rate": 2e-05,
      "loss": 0.1678,
      "step": 2902
    },
    {
      "epoch": 1.9618114227779655,
      "grad_norm": 4.80509090423584,
      "learning_rate": 2e-05,
      "loss": 0.1426,
      "step": 2903
    },
    {
      "epoch": 1.9624873267995944,
      "grad_norm": 4.784300804138184,
      "learning_rate": 2e-05,
      "loss": 0.2,
      "step": 2904
    },
    {
      "epoch": 1.9631632308212232,
      "grad_norm": 5.337709426879883,
      "learning_rate": 2e-05,
      "loss": 0.1418,
      "step": 2905
    },
    {
      "epoch": 1.9638391348428523,
      "grad_norm": 13.378520011901855,
      "learning_rate": 2e-05,
      "loss": 0.1913,
      "step": 2906
    },
    {
      "epoch": 1.9645150388644812,
      "grad_norm": 5.909094333648682,
      "learning_rate": 2e-05,
      "loss": 0.2524,
      "step": 2907
    },
    {
      "epoch": 1.9651909428861103,
      "grad_norm": 9.7367525100708,
      "learning_rate": 2e-05,
      "loss": 0.1631,
      "step": 2908
    },
    {
      "epoch": 1.9658668469077392,
      "grad_norm": 4.675846576690674,
      "learning_rate": 2e-05,
      "loss": 0.1845,
      "step": 2909
    },
    {
      "epoch": 1.966542750929368,
      "grad_norm": 10.139694213867188,
      "learning_rate": 2e-05,
      "loss": 0.2108,
      "step": 2910
    },
    {
      "epoch": 1.967218654950997,
      "grad_norm": 6.667794704437256,
      "learning_rate": 2e-05,
      "loss": 0.172,
      "step": 2911
    },
    {
      "epoch": 1.9678945589726258,
      "grad_norm": 8.518877983093262,
      "learning_rate": 2e-05,
      "loss": 0.165,
      "step": 2912
    },
    {
      "epoch": 1.9685704629942549,
      "grad_norm": 15.649295806884766,
      "learning_rate": 2e-05,
      "loss": 0.2453,
      "step": 2913
    },
    {
      "epoch": 1.9692463670158837,
      "grad_norm": 8.8789701461792,
      "learning_rate": 2e-05,
      "loss": 0.1814,
      "step": 2914
    },
    {
      "epoch": 1.9699222710375128,
      "grad_norm": 4.221238136291504,
      "learning_rate": 2e-05,
      "loss": 0.1632,
      "step": 2915
    },
    {
      "epoch": 1.9705981750591417,
      "grad_norm": 5.267458438873291,
      "learning_rate": 2e-05,
      "loss": 0.1244,
      "step": 2916
    },
    {
      "epoch": 1.9712740790807706,
      "grad_norm": 5.69966983795166,
      "learning_rate": 2e-05,
      "loss": 0.1349,
      "step": 2917
    },
    {
      "epoch": 1.9719499831023994,
      "grad_norm": 3.8558523654937744,
      "learning_rate": 2e-05,
      "loss": 0.1108,
      "step": 2918
    },
    {
      "epoch": 1.9726258871240283,
      "grad_norm": 3.4546284675598145,
      "learning_rate": 2e-05,
      "loss": 0.1655,
      "step": 2919
    },
    {
      "epoch": 1.9733017911456572,
      "grad_norm": 7.102457046508789,
      "learning_rate": 2e-05,
      "loss": 0.1588,
      "step": 2920
    },
    {
      "epoch": 1.9739776951672863,
      "grad_norm": 10.218047142028809,
      "learning_rate": 2e-05,
      "loss": 0.2331,
      "step": 2921
    },
    {
      "epoch": 1.9746535991889151,
      "grad_norm": 5.357170581817627,
      "learning_rate": 2e-05,
      "loss": 0.1752,
      "step": 2922
    },
    {
      "epoch": 1.9753295032105442,
      "grad_norm": 6.4291276931762695,
      "learning_rate": 2e-05,
      "loss": 0.1694,
      "step": 2923
    },
    {
      "epoch": 1.976005407232173,
      "grad_norm": 3.9666686058044434,
      "learning_rate": 2e-05,
      "loss": 0.2129,
      "step": 2924
    },
    {
      "epoch": 1.976681311253802,
      "grad_norm": 4.652652740478516,
      "learning_rate": 2e-05,
      "loss": 0.1588,
      "step": 2925
    },
    {
      "epoch": 1.9773572152754308,
      "grad_norm": 9.393472671508789,
      "learning_rate": 2e-05,
      "loss": 0.1846,
      "step": 2926
    },
    {
      "epoch": 1.9780331192970597,
      "grad_norm": 4.401183128356934,
      "learning_rate": 2e-05,
      "loss": 0.1381,
      "step": 2927
    },
    {
      "epoch": 1.9787090233186886,
      "grad_norm": 6.686988830566406,
      "learning_rate": 2e-05,
      "loss": 0.1926,
      "step": 2928
    },
    {
      "epoch": 1.9793849273403177,
      "grad_norm": 4.928613662719727,
      "learning_rate": 2e-05,
      "loss": 0.1675,
      "step": 2929
    },
    {
      "epoch": 1.9800608313619465,
      "grad_norm": 4.313234806060791,
      "learning_rate": 2e-05,
      "loss": 0.1577,
      "step": 2930
    },
    {
      "epoch": 1.9807367353835756,
      "grad_norm": 8.284016609191895,
      "learning_rate": 2e-05,
      "loss": 0.1498,
      "step": 2931
    },
    {
      "epoch": 1.9814126394052045,
      "grad_norm": 3.5303051471710205,
      "learning_rate": 2e-05,
      "loss": 0.2164,
      "step": 2932
    },
    {
      "epoch": 1.9820885434268334,
      "grad_norm": 5.7605061531066895,
      "learning_rate": 2e-05,
      "loss": 0.1746,
      "step": 2933
    },
    {
      "epoch": 1.9827644474484623,
      "grad_norm": 3.0147488117218018,
      "learning_rate": 2e-05,
      "loss": 0.1653,
      "step": 2934
    },
    {
      "epoch": 1.9834403514700911,
      "grad_norm": 5.692258358001709,
      "learning_rate": 2e-05,
      "loss": 0.1311,
      "step": 2935
    },
    {
      "epoch": 1.9841162554917202,
      "grad_norm": 6.569450855255127,
      "learning_rate": 2e-05,
      "loss": 0.1237,
      "step": 2936
    },
    {
      "epoch": 1.984792159513349,
      "grad_norm": 4.27687406539917,
      "learning_rate": 2e-05,
      "loss": 0.1954,
      "step": 2937
    },
    {
      "epoch": 1.9854680635349782,
      "grad_norm": 6.633908748626709,
      "learning_rate": 2e-05,
      "loss": 0.1502,
      "step": 2938
    },
    {
      "epoch": 1.986143967556607,
      "grad_norm": 4.053294658660889,
      "learning_rate": 2e-05,
      "loss": 0.1925,
      "step": 2939
    },
    {
      "epoch": 1.986819871578236,
      "grad_norm": 4.516960620880127,
      "learning_rate": 2e-05,
      "loss": 0.1202,
      "step": 2940
    },
    {
      "epoch": 1.9874957755998648,
      "grad_norm": 7.344974994659424,
      "learning_rate": 2e-05,
      "loss": 0.1833,
      "step": 2941
    },
    {
      "epoch": 1.9881716796214937,
      "grad_norm": 11.731520652770996,
      "learning_rate": 2e-05,
      "loss": 0.2379,
      "step": 2942
    },
    {
      "epoch": 1.9888475836431225,
      "grad_norm": 11.45028018951416,
      "learning_rate": 2e-05,
      "loss": 0.153,
      "step": 2943
    },
    {
      "epoch": 1.9895234876647516,
      "grad_norm": 5.908964157104492,
      "learning_rate": 2e-05,
      "loss": 0.1891,
      "step": 2944
    },
    {
      "epoch": 1.9901993916863805,
      "grad_norm": 4.992810249328613,
      "learning_rate": 2e-05,
      "loss": 0.1585,
      "step": 2945
    },
    {
      "epoch": 1.9908752957080096,
      "grad_norm": 10.803153991699219,
      "learning_rate": 2e-05,
      "loss": 0.1663,
      "step": 2946
    },
    {
      "epoch": 1.9915511997296385,
      "grad_norm": 6.042548179626465,
      "learning_rate": 2e-05,
      "loss": 0.2718,
      "step": 2947
    },
    {
      "epoch": 1.9922271037512673,
      "grad_norm": 6.572315216064453,
      "learning_rate": 2e-05,
      "loss": 0.1325,
      "step": 2948
    },
    {
      "epoch": 1.9929030077728962,
      "grad_norm": 6.5174665451049805,
      "learning_rate": 2e-05,
      "loss": 0.1952,
      "step": 2949
    },
    {
      "epoch": 1.993578911794525,
      "grad_norm": 6.284380912780762,
      "learning_rate": 2e-05,
      "loss": 0.1692,
      "step": 2950
    },
    {
      "epoch": 1.993578911794525,
      "eval_accuracy": 0.6034628378378378,
      "eval_loss": 0.24298271536827087,
      "eval_runtime": 43.1678,
      "eval_samples_per_second": 548.557,
      "eval_steps_per_second": 17.142,
      "step": 2950
    },
    {
      "epoch": 1.9942548158161542,
      "grad_norm": 3.2937819957733154,
      "learning_rate": 2e-05,
      "loss": 0.1658,
      "step": 2951
    },
    {
      "epoch": 1.994930719837783,
      "grad_norm": 5.049665927886963,
      "learning_rate": 2e-05,
      "loss": 0.1844,
      "step": 2952
    },
    {
      "epoch": 1.9956066238594121,
      "grad_norm": 11.591720581054688,
      "learning_rate": 2e-05,
      "loss": 0.1798,
      "step": 2953
    },
    {
      "epoch": 1.996282527881041,
      "grad_norm": 14.118362426757812,
      "learning_rate": 2e-05,
      "loss": 0.2104,
      "step": 2954
    },
    {
      "epoch": 1.9969584319026699,
      "grad_norm": 2.9852046966552734,
      "learning_rate": 2e-05,
      "loss": 0.1558,
      "step": 2955
    },
    {
      "epoch": 1.9976343359242987,
      "grad_norm": 4.10506010055542,
      "learning_rate": 2e-05,
      "loss": 0.1519,
      "step": 2956
    },
    {
      "epoch": 1.9983102399459276,
      "grad_norm": 4.632903575897217,
      "learning_rate": 2e-05,
      "loss": 0.119,
      "step": 2957
    },
    {
      "epoch": 1.9989861439675565,
      "grad_norm": 4.232982635498047,
      "learning_rate": 2e-05,
      "loss": 0.2211,
      "step": 2958
    },
    {
      "epoch": 1.9996620479891856,
      "grad_norm": 5.2129034996032715,
      "learning_rate": 2e-05,
      "loss": 0.147,
      "step": 2959
    },
    {
      "epoch": 2.0,
      "grad_norm": 3.6078925132751465,
      "learning_rate": 2e-05,
      "loss": 0.0742,
      "step": 2960
    },
    {
      "epoch": 2.000675904021629,
      "grad_norm": 4.6735334396362305,
      "learning_rate": 2e-05,
      "loss": 0.1317,
      "step": 2961
    },
    {
      "epoch": 2.0013518080432577,
      "grad_norm": 4.6005473136901855,
      "learning_rate": 2e-05,
      "loss": 0.2249,
      "step": 2962
    },
    {
      "epoch": 2.0020277120648866,
      "grad_norm": 4.877197265625,
      "learning_rate": 2e-05,
      "loss": 0.1337,
      "step": 2963
    },
    {
      "epoch": 2.002703616086516,
      "grad_norm": 3.1036055088043213,
      "learning_rate": 2e-05,
      "loss": 0.1478,
      "step": 2964
    },
    {
      "epoch": 2.003379520108145,
      "grad_norm": 5.319400787353516,
      "learning_rate": 2e-05,
      "loss": 0.2309,
      "step": 2965
    },
    {
      "epoch": 2.0040554241297737,
      "grad_norm": 3.7249183654785156,
      "learning_rate": 2e-05,
      "loss": 0.1835,
      "step": 2966
    },
    {
      "epoch": 2.0047313281514025,
      "grad_norm": 4.209641456604004,
      "learning_rate": 2e-05,
      "loss": 0.1504,
      "step": 2967
    },
    {
      "epoch": 2.0054072321730314,
      "grad_norm": 4.227386474609375,
      "learning_rate": 2e-05,
      "loss": 0.1664,
      "step": 2968
    },
    {
      "epoch": 2.0060831361946603,
      "grad_norm": 3.8710951805114746,
      "learning_rate": 2e-05,
      "loss": 0.1791,
      "step": 2969
    },
    {
      "epoch": 2.006759040216289,
      "grad_norm": 5.418036937713623,
      "learning_rate": 2e-05,
      "loss": 0.1537,
      "step": 2970
    },
    {
      "epoch": 2.007434944237918,
      "grad_norm": 4.285545825958252,
      "learning_rate": 2e-05,
      "loss": 0.1516,
      "step": 2971
    },
    {
      "epoch": 2.0081108482595473,
      "grad_norm": 7.1214518547058105,
      "learning_rate": 2e-05,
      "loss": 0.1347,
      "step": 2972
    },
    {
      "epoch": 2.008786752281176,
      "grad_norm": 6.34596061706543,
      "learning_rate": 2e-05,
      "loss": 0.17,
      "step": 2973
    },
    {
      "epoch": 2.009462656302805,
      "grad_norm": 4.461135387420654,
      "learning_rate": 2e-05,
      "loss": 0.1347,
      "step": 2974
    },
    {
      "epoch": 2.010138560324434,
      "grad_norm": 5.788484573364258,
      "learning_rate": 2e-05,
      "loss": 0.202,
      "step": 2975
    },
    {
      "epoch": 2.010814464346063,
      "grad_norm": 8.429254531860352,
      "learning_rate": 2e-05,
      "loss": 0.2396,
      "step": 2976
    },
    {
      "epoch": 2.0114903683676917,
      "grad_norm": 8.989744186401367,
      "learning_rate": 2e-05,
      "loss": 0.1601,
      "step": 2977
    },
    {
      "epoch": 2.0121662723893206,
      "grad_norm": 6.173910140991211,
      "learning_rate": 2e-05,
      "loss": 0.1576,
      "step": 2978
    },
    {
      "epoch": 2.0128421764109494,
      "grad_norm": 8.21849250793457,
      "learning_rate": 2e-05,
      "loss": 0.1596,
      "step": 2979
    },
    {
      "epoch": 2.0135180804325787,
      "grad_norm": 4.760103225708008,
      "learning_rate": 2e-05,
      "loss": 0.2017,
      "step": 2980
    },
    {
      "epoch": 2.0141939844542076,
      "grad_norm": 4.564212799072266,
      "learning_rate": 2e-05,
      "loss": 0.1801,
      "step": 2981
    },
    {
      "epoch": 2.0148698884758365,
      "grad_norm": 7.960360050201416,
      "learning_rate": 2e-05,
      "loss": 0.1184,
      "step": 2982
    },
    {
      "epoch": 2.0155457924974653,
      "grad_norm": 6.082637786865234,
      "learning_rate": 2e-05,
      "loss": 0.1915,
      "step": 2983
    },
    {
      "epoch": 2.016221696519094,
      "grad_norm": 5.003920555114746,
      "learning_rate": 2e-05,
      "loss": 0.159,
      "step": 2984
    },
    {
      "epoch": 2.016897600540723,
      "grad_norm": 5.937395095825195,
      "learning_rate": 2e-05,
      "loss": 0.1432,
      "step": 2985
    },
    {
      "epoch": 2.017573504562352,
      "grad_norm": 11.74141788482666,
      "learning_rate": 2e-05,
      "loss": 0.1668,
      "step": 2986
    },
    {
      "epoch": 2.0182494085839813,
      "grad_norm": 9.998269081115723,
      "learning_rate": 2e-05,
      "loss": 0.19,
      "step": 2987
    },
    {
      "epoch": 2.01892531260561,
      "grad_norm": 4.348331928253174,
      "learning_rate": 2e-05,
      "loss": 0.1457,
      "step": 2988
    },
    {
      "epoch": 2.019601216627239,
      "grad_norm": 11.877017974853516,
      "learning_rate": 2e-05,
      "loss": 0.1801,
      "step": 2989
    },
    {
      "epoch": 2.020277120648868,
      "grad_norm": 9.183175086975098,
      "learning_rate": 2e-05,
      "loss": 0.1285,
      "step": 2990
    },
    {
      "epoch": 2.0209530246704968,
      "grad_norm": 4.95518159866333,
      "learning_rate": 2e-05,
      "loss": 0.1421,
      "step": 2991
    },
    {
      "epoch": 2.0216289286921256,
      "grad_norm": 4.082427024841309,
      "learning_rate": 2e-05,
      "loss": 0.1517,
      "step": 2992
    },
    {
      "epoch": 2.0223048327137545,
      "grad_norm": 5.052013397216797,
      "learning_rate": 2e-05,
      "loss": 0.2054,
      "step": 2993
    },
    {
      "epoch": 2.0229807367353834,
      "grad_norm": 12.371784210205078,
      "learning_rate": 2e-05,
      "loss": 0.2041,
      "step": 2994
    },
    {
      "epoch": 2.0236566407570127,
      "grad_norm": 4.37654447555542,
      "learning_rate": 2e-05,
      "loss": 0.1223,
      "step": 2995
    },
    {
      "epoch": 2.0243325447786416,
      "grad_norm": 5.615379333496094,
      "learning_rate": 2e-05,
      "loss": 0.2228,
      "step": 2996
    },
    {
      "epoch": 2.0250084488002704,
      "grad_norm": 4.81010627746582,
      "learning_rate": 2e-05,
      "loss": 0.1415,
      "step": 2997
    },
    {
      "epoch": 2.0256843528218993,
      "grad_norm": 4.751645565032959,
      "learning_rate": 2e-05,
      "loss": 0.1788,
      "step": 2998
    },
    {
      "epoch": 2.026360256843528,
      "grad_norm": 9.062533378601074,
      "learning_rate": 2e-05,
      "loss": 0.2487,
      "step": 2999
    },
    {
      "epoch": 2.027036160865157,
      "grad_norm": 11.601116180419922,
      "learning_rate": 2e-05,
      "loss": 0.1894,
      "step": 3000
    },
    {
      "epoch": 2.027036160865157,
      "eval_accuracy": 0.6025760135135135,
      "eval_loss": 0.24444490671157837,
      "eval_runtime": 42.9618,
      "eval_samples_per_second": 551.188,
      "eval_steps_per_second": 17.225,
      "step": 3000
    },
    {
      "epoch": 2.027712064886786,
      "grad_norm": 10.005276679992676,
      "learning_rate": 2e-05,
      "loss": 0.2532,
      "step": 3001
    },
    {
      "epoch": 2.028387968908415,
      "grad_norm": 5.847565174102783,
      "learning_rate": 2e-05,
      "loss": 0.1533,
      "step": 3002
    },
    {
      "epoch": 2.029063872930044,
      "grad_norm": 11.00701904296875,
      "learning_rate": 2e-05,
      "loss": 0.1697,
      "step": 3003
    },
    {
      "epoch": 2.029739776951673,
      "grad_norm": 13.192605018615723,
      "learning_rate": 2e-05,
      "loss": 0.1913,
      "step": 3004
    },
    {
      "epoch": 2.030415680973302,
      "grad_norm": 5.137044906616211,
      "learning_rate": 2e-05,
      "loss": 0.169,
      "step": 3005
    },
    {
      "epoch": 2.0310915849949307,
      "grad_norm": 3.6105823516845703,
      "learning_rate": 2e-05,
      "loss": 0.2038,
      "step": 3006
    },
    {
      "epoch": 2.0317674890165596,
      "grad_norm": 5.426303863525391,
      "learning_rate": 2e-05,
      "loss": 0.2377,
      "step": 3007
    },
    {
      "epoch": 2.0324433930381884,
      "grad_norm": 4.012668132781982,
      "learning_rate": 2e-05,
      "loss": 0.1836,
      "step": 3008
    },
    {
      "epoch": 2.0331192970598173,
      "grad_norm": 2.8176541328430176,
      "learning_rate": 2e-05,
      "loss": 0.126,
      "step": 3009
    },
    {
      "epoch": 2.0337952010814466,
      "grad_norm": 9.950310707092285,
      "learning_rate": 2e-05,
      "loss": 0.1923,
      "step": 3010
    },
    {
      "epoch": 2.0344711051030755,
      "grad_norm": 10.31648063659668,
      "learning_rate": 2e-05,
      "loss": 0.1517,
      "step": 3011
    },
    {
      "epoch": 2.0351470091247044,
      "grad_norm": 8.717665672302246,
      "learning_rate": 2e-05,
      "loss": 0.2197,
      "step": 3012
    },
    {
      "epoch": 2.0358229131463332,
      "grad_norm": 3.6964938640594482,
      "learning_rate": 2e-05,
      "loss": 0.1432,
      "step": 3013
    },
    {
      "epoch": 2.036498817167962,
      "grad_norm": 5.616652488708496,
      "learning_rate": 2e-05,
      "loss": 0.1395,
      "step": 3014
    },
    {
      "epoch": 2.037174721189591,
      "grad_norm": 3.12808895111084,
      "learning_rate": 2e-05,
      "loss": 0.1426,
      "step": 3015
    },
    {
      "epoch": 2.03785062521122,
      "grad_norm": 6.358743667602539,
      "learning_rate": 2e-05,
      "loss": 0.1628,
      "step": 3016
    },
    {
      "epoch": 2.0385265292328487,
      "grad_norm": 5.668044567108154,
      "learning_rate": 2e-05,
      "loss": 0.1982,
      "step": 3017
    },
    {
      "epoch": 2.039202433254478,
      "grad_norm": 7.454906463623047,
      "learning_rate": 2e-05,
      "loss": 0.1683,
      "step": 3018
    },
    {
      "epoch": 2.039878337276107,
      "grad_norm": 10.45825481414795,
      "learning_rate": 2e-05,
      "loss": 0.1837,
      "step": 3019
    },
    {
      "epoch": 2.0405542412977358,
      "grad_norm": 7.6037750244140625,
      "learning_rate": 2e-05,
      "loss": 0.2191,
      "step": 3020
    },
    {
      "epoch": 2.0412301453193646,
      "grad_norm": 9.708761215209961,
      "learning_rate": 2e-05,
      "loss": 0.1484,
      "step": 3021
    },
    {
      "epoch": 2.0419060493409935,
      "grad_norm": 6.406172275543213,
      "learning_rate": 2e-05,
      "loss": 0.1728,
      "step": 3022
    },
    {
      "epoch": 2.0425819533626224,
      "grad_norm": 8.277690887451172,
      "learning_rate": 2e-05,
      "loss": 0.1731,
      "step": 3023
    },
    {
      "epoch": 2.0432578573842513,
      "grad_norm": 8.816667556762695,
      "learning_rate": 2e-05,
      "loss": 0.1629,
      "step": 3024
    },
    {
      "epoch": 2.0439337614058806,
      "grad_norm": 5.795119762420654,
      "learning_rate": 2e-05,
      "loss": 0.2238,
      "step": 3025
    },
    {
      "epoch": 2.0446096654275094,
      "grad_norm": 4.676114082336426,
      "learning_rate": 2e-05,
      "loss": 0.1591,
      "step": 3026
    },
    {
      "epoch": 2.0452855694491383,
      "grad_norm": 6.482534408569336,
      "learning_rate": 2e-05,
      "loss": 0.2361,
      "step": 3027
    },
    {
      "epoch": 2.045961473470767,
      "grad_norm": 3.884361743927002,
      "learning_rate": 2e-05,
      "loss": 0.1803,
      "step": 3028
    },
    {
      "epoch": 2.046637377492396,
      "grad_norm": 9.405523300170898,
      "learning_rate": 2e-05,
      "loss": 0.1624,
      "step": 3029
    },
    {
      "epoch": 2.047313281514025,
      "grad_norm": 11.672199249267578,
      "learning_rate": 2e-05,
      "loss": 0.2371,
      "step": 3030
    },
    {
      "epoch": 2.047989185535654,
      "grad_norm": 9.58065414428711,
      "learning_rate": 2e-05,
      "loss": 0.1934,
      "step": 3031
    },
    {
      "epoch": 2.0486650895572827,
      "grad_norm": 10.119881629943848,
      "learning_rate": 2e-05,
      "loss": 0.2377,
      "step": 3032
    },
    {
      "epoch": 2.049340993578912,
      "grad_norm": 3.700838565826416,
      "learning_rate": 2e-05,
      "loss": 0.1689,
      "step": 3033
    },
    {
      "epoch": 2.050016897600541,
      "grad_norm": 4.781131267547607,
      "learning_rate": 2e-05,
      "loss": 0.1975,
      "step": 3034
    },
    {
      "epoch": 2.0506928016221697,
      "grad_norm": 5.485964298248291,
      "learning_rate": 2e-05,
      "loss": 0.1936,
      "step": 3035
    },
    {
      "epoch": 2.0513687056437986,
      "grad_norm": 3.096673011779785,
      "learning_rate": 2e-05,
      "loss": 0.1578,
      "step": 3036
    },
    {
      "epoch": 2.0520446096654275,
      "grad_norm": 9.307065963745117,
      "learning_rate": 2e-05,
      "loss": 0.1616,
      "step": 3037
    },
    {
      "epoch": 2.0527205136870563,
      "grad_norm": 6.640214920043945,
      "learning_rate": 2e-05,
      "loss": 0.1845,
      "step": 3038
    },
    {
      "epoch": 2.053396417708685,
      "grad_norm": 3.7676360607147217,
      "learning_rate": 2e-05,
      "loss": 0.1745,
      "step": 3039
    },
    {
      "epoch": 2.0540723217303145,
      "grad_norm": 10.616251945495605,
      "learning_rate": 2e-05,
      "loss": 0.1994,
      "step": 3040
    },
    {
      "epoch": 2.0547482257519434,
      "grad_norm": 7.360347747802734,
      "learning_rate": 2e-05,
      "loss": 0.1809,
      "step": 3041
    },
    {
      "epoch": 2.0554241297735723,
      "grad_norm": 3.8854875564575195,
      "learning_rate": 2e-05,
      "loss": 0.1449,
      "step": 3042
    },
    {
      "epoch": 2.056100033795201,
      "grad_norm": 3.352506399154663,
      "learning_rate": 2e-05,
      "loss": 0.1446,
      "step": 3043
    },
    {
      "epoch": 2.05677593781683,
      "grad_norm": 8.17944049835205,
      "learning_rate": 2e-05,
      "loss": 0.1658,
      "step": 3044
    },
    {
      "epoch": 2.057451841838459,
      "grad_norm": 3.659284830093384,
      "learning_rate": 2e-05,
      "loss": 0.1643,
      "step": 3045
    },
    {
      "epoch": 2.0581277458600877,
      "grad_norm": 6.787537574768066,
      "learning_rate": 2e-05,
      "loss": 0.1701,
      "step": 3046
    },
    {
      "epoch": 2.0588036498817166,
      "grad_norm": 8.294686317443848,
      "learning_rate": 2e-05,
      "loss": 0.0979,
      "step": 3047
    },
    {
      "epoch": 2.059479553903346,
      "grad_norm": 3.785752773284912,
      "learning_rate": 2e-05,
      "loss": 0.1748,
      "step": 3048
    },
    {
      "epoch": 2.060155457924975,
      "grad_norm": 5.009303092956543,
      "learning_rate": 2e-05,
      "loss": 0.1544,
      "step": 3049
    },
    {
      "epoch": 2.0608313619466037,
      "grad_norm": 3.68192458152771,
      "learning_rate": 2e-05,
      "loss": 0.1731,
      "step": 3050
    },
    {
      "epoch": 2.0608313619466037,
      "eval_accuracy": 0.5962837837837838,
      "eval_loss": 0.24677318334579468,
      "eval_runtime": 43.6505,
      "eval_samples_per_second": 542.491,
      "eval_steps_per_second": 16.953,
      "step": 3050
    },
    {
      "epoch": 2.0615072659682325,
      "grad_norm": 12.031888008117676,
      "learning_rate": 2e-05,
      "loss": 0.1759,
      "step": 3051
    },
    {
      "epoch": 2.0621831699898614,
      "grad_norm": 6.794201850891113,
      "learning_rate": 2e-05,
      "loss": 0.1086,
      "step": 3052
    },
    {
      "epoch": 2.0628590740114903,
      "grad_norm": 5.413728713989258,
      "learning_rate": 2e-05,
      "loss": 0.1783,
      "step": 3053
    },
    {
      "epoch": 2.063534978033119,
      "grad_norm": 10.313142776489258,
      "learning_rate": 2e-05,
      "loss": 0.1761,
      "step": 3054
    },
    {
      "epoch": 2.0642108820547485,
      "grad_norm": 3.073026418685913,
      "learning_rate": 2e-05,
      "loss": 0.1096,
      "step": 3055
    },
    {
      "epoch": 2.0648867860763773,
      "grad_norm": 5.179638862609863,
      "learning_rate": 2e-05,
      "loss": 0.1681,
      "step": 3056
    },
    {
      "epoch": 2.065562690098006,
      "grad_norm": 4.794300079345703,
      "learning_rate": 2e-05,
      "loss": 0.1455,
      "step": 3057
    },
    {
      "epoch": 2.066238594119635,
      "grad_norm": 7.410665035247803,
      "learning_rate": 2e-05,
      "loss": 0.2369,
      "step": 3058
    },
    {
      "epoch": 2.066914498141264,
      "grad_norm": 5.090938568115234,
      "learning_rate": 2e-05,
      "loss": 0.1971,
      "step": 3059
    },
    {
      "epoch": 2.067590402162893,
      "grad_norm": 15.89980697631836,
      "learning_rate": 2e-05,
      "loss": 0.2726,
      "step": 3060
    },
    {
      "epoch": 2.0682663061845217,
      "grad_norm": 8.74927806854248,
      "learning_rate": 2e-05,
      "loss": 0.1931,
      "step": 3061
    },
    {
      "epoch": 2.0689422102061505,
      "grad_norm": 2.999441146850586,
      "learning_rate": 2e-05,
      "loss": 0.1065,
      "step": 3062
    },
    {
      "epoch": 2.06961811422778,
      "grad_norm": 12.450057983398438,
      "learning_rate": 2e-05,
      "loss": 0.1836,
      "step": 3063
    },
    {
      "epoch": 2.0702940182494087,
      "grad_norm": 7.204087257385254,
      "learning_rate": 2e-05,
      "loss": 0.1712,
      "step": 3064
    },
    {
      "epoch": 2.0709699222710376,
      "grad_norm": 5.514335632324219,
      "learning_rate": 2e-05,
      "loss": 0.203,
      "step": 3065
    },
    {
      "epoch": 2.0716458262926665,
      "grad_norm": 8.17913818359375,
      "learning_rate": 2e-05,
      "loss": 0.1646,
      "step": 3066
    },
    {
      "epoch": 2.0723217303142953,
      "grad_norm": 3.133633852005005,
      "learning_rate": 2e-05,
      "loss": 0.1277,
      "step": 3067
    },
    {
      "epoch": 2.072997634335924,
      "grad_norm": 3.362186908721924,
      "learning_rate": 2e-05,
      "loss": 0.1406,
      "step": 3068
    },
    {
      "epoch": 2.073673538357553,
      "grad_norm": 7.835089683532715,
      "learning_rate": 2e-05,
      "loss": 0.2226,
      "step": 3069
    },
    {
      "epoch": 2.074349442379182,
      "grad_norm": 16.854398727416992,
      "learning_rate": 2e-05,
      "loss": 0.2136,
      "step": 3070
    },
    {
      "epoch": 2.0750253464008113,
      "grad_norm": 3.6954283714294434,
      "learning_rate": 2e-05,
      "loss": 0.1435,
      "step": 3071
    },
    {
      "epoch": 2.07570125042244,
      "grad_norm": 3.755856513977051,
      "learning_rate": 2e-05,
      "loss": 0.2163,
      "step": 3072
    },
    {
      "epoch": 2.076377154444069,
      "grad_norm": 5.692213535308838,
      "learning_rate": 2e-05,
      "loss": 0.1709,
      "step": 3073
    },
    {
      "epoch": 2.077053058465698,
      "grad_norm": 5.823268890380859,
      "learning_rate": 2e-05,
      "loss": 0.2183,
      "step": 3074
    },
    {
      "epoch": 2.0777289624873267,
      "grad_norm": 3.8899178504943848,
      "learning_rate": 2e-05,
      "loss": 0.1582,
      "step": 3075
    },
    {
      "epoch": 2.0784048665089556,
      "grad_norm": 3.8967831134796143,
      "learning_rate": 2e-05,
      "loss": 0.1927,
      "step": 3076
    },
    {
      "epoch": 2.0790807705305845,
      "grad_norm": 9.340778350830078,
      "learning_rate": 2e-05,
      "loss": 0.223,
      "step": 3077
    },
    {
      "epoch": 2.079756674552214,
      "grad_norm": 7.446197032928467,
      "learning_rate": 2e-05,
      "loss": 0.1921,
      "step": 3078
    },
    {
      "epoch": 2.0804325785738427,
      "grad_norm": 6.011845111846924,
      "learning_rate": 2e-05,
      "loss": 0.2064,
      "step": 3079
    },
    {
      "epoch": 2.0811084825954715,
      "grad_norm": 4.458156108856201,
      "learning_rate": 2e-05,
      "loss": 0.1582,
      "step": 3080
    },
    {
      "epoch": 2.0817843866171004,
      "grad_norm": 13.981649398803711,
      "learning_rate": 2e-05,
      "loss": 0.1841,
      "step": 3081
    },
    {
      "epoch": 2.0824602906387293,
      "grad_norm": 6.567946910858154,
      "learning_rate": 2e-05,
      "loss": 0.2044,
      "step": 3082
    },
    {
      "epoch": 2.083136194660358,
      "grad_norm": 6.279582500457764,
      "learning_rate": 2e-05,
      "loss": 0.1292,
      "step": 3083
    },
    {
      "epoch": 2.083812098681987,
      "grad_norm": 4.783864974975586,
      "learning_rate": 2e-05,
      "loss": 0.1939,
      "step": 3084
    },
    {
      "epoch": 2.084488002703616,
      "grad_norm": 5.035356521606445,
      "learning_rate": 2e-05,
      "loss": 0.1664,
      "step": 3085
    },
    {
      "epoch": 2.085163906725245,
      "grad_norm": 2.933652639389038,
      "learning_rate": 2e-05,
      "loss": 0.1151,
      "step": 3086
    },
    {
      "epoch": 2.085839810746874,
      "grad_norm": 5.895377159118652,
      "learning_rate": 2e-05,
      "loss": 0.1591,
      "step": 3087
    },
    {
      "epoch": 2.086515714768503,
      "grad_norm": 8.77101993560791,
      "learning_rate": 2e-05,
      "loss": 0.2006,
      "step": 3088
    },
    {
      "epoch": 2.087191618790132,
      "grad_norm": 9.538093566894531,
      "learning_rate": 2e-05,
      "loss": 0.1645,
      "step": 3089
    },
    {
      "epoch": 2.0878675228117607,
      "grad_norm": 9.071789741516113,
      "learning_rate": 2e-05,
      "loss": 0.1723,
      "step": 3090
    },
    {
      "epoch": 2.0885434268333896,
      "grad_norm": 8.510079383850098,
      "learning_rate": 2e-05,
      "loss": 0.2287,
      "step": 3091
    },
    {
      "epoch": 2.0892193308550184,
      "grad_norm": 6.541302680969238,
      "learning_rate": 2e-05,
      "loss": 0.1407,
      "step": 3092
    },
    {
      "epoch": 2.0898952348766473,
      "grad_norm": 9.028719902038574,
      "learning_rate": 2e-05,
      "loss": 0.2816,
      "step": 3093
    },
    {
      "epoch": 2.0905711388982766,
      "grad_norm": 6.769493103027344,
      "learning_rate": 2e-05,
      "loss": 0.1794,
      "step": 3094
    },
    {
      "epoch": 2.0912470429199055,
      "grad_norm": 4.790868759155273,
      "learning_rate": 2e-05,
      "loss": 0.1182,
      "step": 3095
    },
    {
      "epoch": 2.0919229469415344,
      "grad_norm": 8.728163719177246,
      "learning_rate": 2e-05,
      "loss": 0.1793,
      "step": 3096
    },
    {
      "epoch": 2.0925988509631632,
      "grad_norm": 4.219030857086182,
      "learning_rate": 2e-05,
      "loss": 0.1379,
      "step": 3097
    },
    {
      "epoch": 2.093274754984792,
      "grad_norm": 9.693530082702637,
      "learning_rate": 2e-05,
      "loss": 0.1829,
      "step": 3098
    },
    {
      "epoch": 2.093950659006421,
      "grad_norm": 9.99069881439209,
      "learning_rate": 2e-05,
      "loss": 0.207,
      "step": 3099
    },
    {
      "epoch": 2.09462656302805,
      "grad_norm": 7.041345596313477,
      "learning_rate": 2e-05,
      "loss": 0.1977,
      "step": 3100
    },
    {
      "epoch": 2.09462656302805,
      "eval_accuracy": 0.6035472972972973,
      "eval_loss": 0.24456217885017395,
      "eval_runtime": 43.621,
      "eval_samples_per_second": 542.858,
      "eval_steps_per_second": 16.964,
      "step": 3100
    },
    {
      "epoch": 2.095302467049679,
      "grad_norm": 4.213728427886963,
      "learning_rate": 2e-05,
      "loss": 0.129,
      "step": 3101
    },
    {
      "epoch": 2.095978371071308,
      "grad_norm": 3.445549488067627,
      "learning_rate": 2e-05,
      "loss": 0.1574,
      "step": 3102
    },
    {
      "epoch": 2.096654275092937,
      "grad_norm": 3.7089526653289795,
      "learning_rate": 2e-05,
      "loss": 0.1627,
      "step": 3103
    },
    {
      "epoch": 2.0973301791145658,
      "grad_norm": 13.804893493652344,
      "learning_rate": 2e-05,
      "loss": 0.2211,
      "step": 3104
    },
    {
      "epoch": 2.0980060831361946,
      "grad_norm": 4.414833068847656,
      "learning_rate": 2e-05,
      "loss": 0.1905,
      "step": 3105
    },
    {
      "epoch": 2.0986819871578235,
      "grad_norm": 3.63696026802063,
      "learning_rate": 2e-05,
      "loss": 0.1485,
      "step": 3106
    },
    {
      "epoch": 2.0993578911794524,
      "grad_norm": 3.6523048877716064,
      "learning_rate": 2e-05,
      "loss": 0.1687,
      "step": 3107
    },
    {
      "epoch": 2.1000337952010812,
      "grad_norm": 3.3543989658355713,
      "learning_rate": 2e-05,
      "loss": 0.1439,
      "step": 3108
    },
    {
      "epoch": 2.1007096992227106,
      "grad_norm": 6.494397163391113,
      "learning_rate": 2e-05,
      "loss": 0.1967,
      "step": 3109
    },
    {
      "epoch": 2.1013856032443394,
      "grad_norm": 6.415686130523682,
      "learning_rate": 2e-05,
      "loss": 0.1548,
      "step": 3110
    },
    {
      "epoch": 2.1020615072659683,
      "grad_norm": 8.738008499145508,
      "learning_rate": 2e-05,
      "loss": 0.1547,
      "step": 3111
    },
    {
      "epoch": 2.102737411287597,
      "grad_norm": 5.148190498352051,
      "learning_rate": 2e-05,
      "loss": 0.1708,
      "step": 3112
    },
    {
      "epoch": 2.103413315309226,
      "grad_norm": 7.366543769836426,
      "learning_rate": 2e-05,
      "loss": 0.1437,
      "step": 3113
    },
    {
      "epoch": 2.104089219330855,
      "grad_norm": 6.5242600440979,
      "learning_rate": 2e-05,
      "loss": 0.1945,
      "step": 3114
    },
    {
      "epoch": 2.104765123352484,
      "grad_norm": 5.333303451538086,
      "learning_rate": 2e-05,
      "loss": 0.116,
      "step": 3115
    },
    {
      "epoch": 2.1054410273741127,
      "grad_norm": 11.689277648925781,
      "learning_rate": 2e-05,
      "loss": 0.2454,
      "step": 3116
    },
    {
      "epoch": 2.106116931395742,
      "grad_norm": 4.443140029907227,
      "learning_rate": 2e-05,
      "loss": 0.1195,
      "step": 3117
    },
    {
      "epoch": 2.106792835417371,
      "grad_norm": 7.86335563659668,
      "learning_rate": 2e-05,
      "loss": 0.21,
      "step": 3118
    },
    {
      "epoch": 2.1074687394389997,
      "grad_norm": 3.4494547843933105,
      "learning_rate": 2e-05,
      "loss": 0.1404,
      "step": 3119
    },
    {
      "epoch": 2.1081446434606286,
      "grad_norm": 11.829693794250488,
      "learning_rate": 2e-05,
      "loss": 0.2254,
      "step": 3120
    },
    {
      "epoch": 2.1088205474822574,
      "grad_norm": 11.516547203063965,
      "learning_rate": 2e-05,
      "loss": 0.1857,
      "step": 3121
    },
    {
      "epoch": 2.1094964515038863,
      "grad_norm": 3.490159749984741,
      "learning_rate": 2e-05,
      "loss": 0.1272,
      "step": 3122
    },
    {
      "epoch": 2.110172355525515,
      "grad_norm": 8.317041397094727,
      "learning_rate": 2e-05,
      "loss": 0.1767,
      "step": 3123
    },
    {
      "epoch": 2.1108482595471445,
      "grad_norm": 6.226221084594727,
      "learning_rate": 2e-05,
      "loss": 0.1668,
      "step": 3124
    },
    {
      "epoch": 2.1115241635687734,
      "grad_norm": 4.412379741668701,
      "learning_rate": 2e-05,
      "loss": 0.233,
      "step": 3125
    },
    {
      "epoch": 2.1122000675904022,
      "grad_norm": 8.910784721374512,
      "learning_rate": 2e-05,
      "loss": 0.2058,
      "step": 3126
    },
    {
      "epoch": 2.112875971612031,
      "grad_norm": 3.878466844558716,
      "learning_rate": 2e-05,
      "loss": 0.1313,
      "step": 3127
    },
    {
      "epoch": 2.11355187563366,
      "grad_norm": 4.4734206199646,
      "learning_rate": 2e-05,
      "loss": 0.1571,
      "step": 3128
    },
    {
      "epoch": 2.114227779655289,
      "grad_norm": 4.8730998039245605,
      "learning_rate": 2e-05,
      "loss": 0.1996,
      "step": 3129
    },
    {
      "epoch": 2.1149036836769177,
      "grad_norm": 3.1972901821136475,
      "learning_rate": 2e-05,
      "loss": 0.1002,
      "step": 3130
    },
    {
      "epoch": 2.1155795876985466,
      "grad_norm": 5.084463596343994,
      "learning_rate": 2e-05,
      "loss": 0.1446,
      "step": 3131
    },
    {
      "epoch": 2.116255491720176,
      "grad_norm": 3.6921608448028564,
      "learning_rate": 2e-05,
      "loss": 0.1539,
      "step": 3132
    },
    {
      "epoch": 2.1169313957418048,
      "grad_norm": 3.3622002601623535,
      "learning_rate": 2e-05,
      "loss": 0.1108,
      "step": 3133
    },
    {
      "epoch": 2.1176072997634336,
      "grad_norm": 3.606389045715332,
      "learning_rate": 2e-05,
      "loss": 0.1461,
      "step": 3134
    },
    {
      "epoch": 2.1182832037850625,
      "grad_norm": 7.90379524230957,
      "learning_rate": 2e-05,
      "loss": 0.1753,
      "step": 3135
    },
    {
      "epoch": 2.1189591078066914,
      "grad_norm": 6.426236152648926,
      "learning_rate": 2e-05,
      "loss": 0.2063,
      "step": 3136
    },
    {
      "epoch": 2.1196350118283203,
      "grad_norm": 4.41853141784668,
      "learning_rate": 2e-05,
      "loss": 0.1501,
      "step": 3137
    },
    {
      "epoch": 2.120310915849949,
      "grad_norm": 5.13426399230957,
      "learning_rate": 2e-05,
      "loss": 0.1292,
      "step": 3138
    },
    {
      "epoch": 2.1209868198715784,
      "grad_norm": 4.725967884063721,
      "learning_rate": 2e-05,
      "loss": 0.2022,
      "step": 3139
    },
    {
      "epoch": 2.1216627238932073,
      "grad_norm": 4.623298168182373,
      "learning_rate": 2e-05,
      "loss": 0.1799,
      "step": 3140
    },
    {
      "epoch": 2.122338627914836,
      "grad_norm": 7.722148418426514,
      "learning_rate": 2e-05,
      "loss": 0.1545,
      "step": 3141
    },
    {
      "epoch": 2.123014531936465,
      "grad_norm": 4.646947860717773,
      "learning_rate": 2e-05,
      "loss": 0.1408,
      "step": 3142
    },
    {
      "epoch": 2.123690435958094,
      "grad_norm": 3.8990731239318848,
      "learning_rate": 2e-05,
      "loss": 0.1275,
      "step": 3143
    },
    {
      "epoch": 2.124366339979723,
      "grad_norm": 11.815875053405762,
      "learning_rate": 2e-05,
      "loss": 0.2186,
      "step": 3144
    },
    {
      "epoch": 2.1250422440013517,
      "grad_norm": 4.157047271728516,
      "learning_rate": 2e-05,
      "loss": 0.1886,
      "step": 3145
    },
    {
      "epoch": 2.1257181480229805,
      "grad_norm": 7.639450550079346,
      "learning_rate": 2e-05,
      "loss": 0.163,
      "step": 3146
    },
    {
      "epoch": 2.12639405204461,
      "grad_norm": 7.498966217041016,
      "learning_rate": 2e-05,
      "loss": 0.2004,
      "step": 3147
    },
    {
      "epoch": 2.1270699560662387,
      "grad_norm": 6.309972763061523,
      "learning_rate": 2e-05,
      "loss": 0.1744,
      "step": 3148
    },
    {
      "epoch": 2.1277458600878676,
      "grad_norm": 4.759450435638428,
      "learning_rate": 2e-05,
      "loss": 0.1543,
      "step": 3149
    },
    {
      "epoch": 2.1284217641094965,
      "grad_norm": 11.800680160522461,
      "learning_rate": 2e-05,
      "loss": 0.2536,
      "step": 3150
    },
    {
      "epoch": 2.1284217641094965,
      "eval_accuracy": 0.5943834459459459,
      "eval_loss": 0.24277140200138092,
      "eval_runtime": 43.284,
      "eval_samples_per_second": 547.084,
      "eval_steps_per_second": 17.096,
      "step": 3150
    },
    {
      "epoch": 2.1290976681311253,
      "grad_norm": 9.763533592224121,
      "learning_rate": 2e-05,
      "loss": 0.2343,
      "step": 3151
    },
    {
      "epoch": 2.129773572152754,
      "grad_norm": 4.771275043487549,
      "learning_rate": 2e-05,
      "loss": 0.2521,
      "step": 3152
    },
    {
      "epoch": 2.130449476174383,
      "grad_norm": 6.450836658477783,
      "learning_rate": 2e-05,
      "loss": 0.2261,
      "step": 3153
    },
    {
      "epoch": 2.1311253801960124,
      "grad_norm": 8.70740795135498,
      "learning_rate": 2e-05,
      "loss": 0.1607,
      "step": 3154
    },
    {
      "epoch": 2.1318012842176413,
      "grad_norm": 11.853043556213379,
      "learning_rate": 2e-05,
      "loss": 0.1894,
      "step": 3155
    },
    {
      "epoch": 2.13247718823927,
      "grad_norm": 6.056429862976074,
      "learning_rate": 2e-05,
      "loss": 0.2092,
      "step": 3156
    },
    {
      "epoch": 2.133153092260899,
      "grad_norm": 4.074942588806152,
      "learning_rate": 2e-05,
      "loss": 0.1995,
      "step": 3157
    },
    {
      "epoch": 2.133828996282528,
      "grad_norm": 5.064111232757568,
      "learning_rate": 2e-05,
      "loss": 0.1855,
      "step": 3158
    },
    {
      "epoch": 2.1345049003041567,
      "grad_norm": 7.881313323974609,
      "learning_rate": 2e-05,
      "loss": 0.1767,
      "step": 3159
    },
    {
      "epoch": 2.1351808043257856,
      "grad_norm": 5.936881065368652,
      "learning_rate": 2e-05,
      "loss": 0.1343,
      "step": 3160
    },
    {
      "epoch": 2.1358567083474145,
      "grad_norm": 4.2075982093811035,
      "learning_rate": 2e-05,
      "loss": 0.1525,
      "step": 3161
    },
    {
      "epoch": 2.136532612369044,
      "grad_norm": 4.53906774520874,
      "learning_rate": 2e-05,
      "loss": 0.1891,
      "step": 3162
    },
    {
      "epoch": 2.1372085163906727,
      "grad_norm": 5.103899955749512,
      "learning_rate": 2e-05,
      "loss": 0.1975,
      "step": 3163
    },
    {
      "epoch": 2.1378844204123015,
      "grad_norm": 5.558284282684326,
      "learning_rate": 2e-05,
      "loss": 0.1898,
      "step": 3164
    },
    {
      "epoch": 2.1385603244339304,
      "grad_norm": 3.8485207557678223,
      "learning_rate": 2e-05,
      "loss": 0.1708,
      "step": 3165
    },
    {
      "epoch": 2.1392362284555593,
      "grad_norm": 3.328108787536621,
      "learning_rate": 2e-05,
      "loss": 0.1024,
      "step": 3166
    },
    {
      "epoch": 2.139912132477188,
      "grad_norm": 5.200891017913818,
      "learning_rate": 2e-05,
      "loss": 0.1501,
      "step": 3167
    },
    {
      "epoch": 2.140588036498817,
      "grad_norm": 3.764679431915283,
      "learning_rate": 2e-05,
      "loss": 0.1654,
      "step": 3168
    },
    {
      "epoch": 2.141263940520446,
      "grad_norm": 3.8609917163848877,
      "learning_rate": 2e-05,
      "loss": 0.1572,
      "step": 3169
    },
    {
      "epoch": 2.141939844542075,
      "grad_norm": 7.967406272888184,
      "learning_rate": 2e-05,
      "loss": 0.1891,
      "step": 3170
    },
    {
      "epoch": 2.142615748563704,
      "grad_norm": 5.429915904998779,
      "learning_rate": 2e-05,
      "loss": 0.1265,
      "step": 3171
    },
    {
      "epoch": 2.143291652585333,
      "grad_norm": 6.476261615753174,
      "learning_rate": 2e-05,
      "loss": 0.2458,
      "step": 3172
    },
    {
      "epoch": 2.143967556606962,
      "grad_norm": 4.117125034332275,
      "learning_rate": 2e-05,
      "loss": 0.1982,
      "step": 3173
    },
    {
      "epoch": 2.1446434606285907,
      "grad_norm": 5.634766101837158,
      "learning_rate": 2e-05,
      "loss": 0.0929,
      "step": 3174
    },
    {
      "epoch": 2.1453193646502196,
      "grad_norm": 5.313941478729248,
      "learning_rate": 2e-05,
      "loss": 0.1815,
      "step": 3175
    },
    {
      "epoch": 2.1459952686718484,
      "grad_norm": 6.390480995178223,
      "learning_rate": 2e-05,
      "loss": 0.1991,
      "step": 3176
    },
    {
      "epoch": 2.1466711726934777,
      "grad_norm": 8.011385917663574,
      "learning_rate": 2e-05,
      "loss": 0.2242,
      "step": 3177
    },
    {
      "epoch": 2.1473470767151066,
      "grad_norm": 9.767807006835938,
      "learning_rate": 2e-05,
      "loss": 0.1668,
      "step": 3178
    },
    {
      "epoch": 2.1480229807367355,
      "grad_norm": 6.0176873207092285,
      "learning_rate": 2e-05,
      "loss": 0.2156,
      "step": 3179
    },
    {
      "epoch": 2.1486988847583643,
      "grad_norm": 4.412847518920898,
      "learning_rate": 2e-05,
      "loss": 0.1615,
      "step": 3180
    },
    {
      "epoch": 2.149374788779993,
      "grad_norm": 4.831414222717285,
      "learning_rate": 2e-05,
      "loss": 0.1489,
      "step": 3181
    },
    {
      "epoch": 2.150050692801622,
      "grad_norm": 3.231546640396118,
      "learning_rate": 2e-05,
      "loss": 0.1549,
      "step": 3182
    },
    {
      "epoch": 2.150726596823251,
      "grad_norm": 3.8682236671447754,
      "learning_rate": 2e-05,
      "loss": 0.1509,
      "step": 3183
    },
    {
      "epoch": 2.15140250084488,
      "grad_norm": 6.37618350982666,
      "learning_rate": 2e-05,
      "loss": 0.1948,
      "step": 3184
    },
    {
      "epoch": 2.152078404866509,
      "grad_norm": 3.7866528034210205,
      "learning_rate": 2e-05,
      "loss": 0.1786,
      "step": 3185
    },
    {
      "epoch": 2.152754308888138,
      "grad_norm": 11.494038581848145,
      "learning_rate": 2e-05,
      "loss": 0.1536,
      "step": 3186
    },
    {
      "epoch": 2.153430212909767,
      "grad_norm": 9.503454208374023,
      "learning_rate": 2e-05,
      "loss": 0.2133,
      "step": 3187
    },
    {
      "epoch": 2.1541061169313958,
      "grad_norm": 6.074669361114502,
      "learning_rate": 2e-05,
      "loss": 0.1963,
      "step": 3188
    },
    {
      "epoch": 2.1547820209530246,
      "grad_norm": 3.4947216510772705,
      "learning_rate": 2e-05,
      "loss": 0.1442,
      "step": 3189
    },
    {
      "epoch": 2.1554579249746535,
      "grad_norm": 5.0107340812683105,
      "learning_rate": 2e-05,
      "loss": 0.1576,
      "step": 3190
    },
    {
      "epoch": 2.1561338289962824,
      "grad_norm": 5.114109992980957,
      "learning_rate": 2e-05,
      "loss": 0.1837,
      "step": 3191
    },
    {
      "epoch": 2.1568097330179112,
      "grad_norm": 4.087992191314697,
      "learning_rate": 2e-05,
      "loss": 0.1943,
      "step": 3192
    },
    {
      "epoch": 2.1574856370395405,
      "grad_norm": 8.127744674682617,
      "learning_rate": 2e-05,
      "loss": 0.1847,
      "step": 3193
    },
    {
      "epoch": 2.1581615410611694,
      "grad_norm": 12.328630447387695,
      "learning_rate": 2e-05,
      "loss": 0.2503,
      "step": 3194
    },
    {
      "epoch": 2.1588374450827983,
      "grad_norm": 4.856424331665039,
      "learning_rate": 2e-05,
      "loss": 0.1926,
      "step": 3195
    },
    {
      "epoch": 2.159513349104427,
      "grad_norm": 3.2503879070281982,
      "learning_rate": 2e-05,
      "loss": 0.1262,
      "step": 3196
    },
    {
      "epoch": 2.160189253126056,
      "grad_norm": 9.877945899963379,
      "learning_rate": 2e-05,
      "loss": 0.1657,
      "step": 3197
    },
    {
      "epoch": 2.160865157147685,
      "grad_norm": 2.5658576488494873,
      "learning_rate": 2e-05,
      "loss": 0.1177,
      "step": 3198
    },
    {
      "epoch": 2.1615410611693138,
      "grad_norm": 4.326022148132324,
      "learning_rate": 2e-05,
      "loss": 0.2439,
      "step": 3199
    },
    {
      "epoch": 2.162216965190943,
      "grad_norm": 7.753981113433838,
      "learning_rate": 2e-05,
      "loss": 0.186,
      "step": 3200
    },
    {
      "epoch": 2.162216965190943,
      "eval_accuracy": 0.6052364864864865,
      "eval_loss": 0.24144019186496735,
      "eval_runtime": 43.6366,
      "eval_samples_per_second": 542.664,
      "eval_steps_per_second": 16.958,
      "step": 3200
    },
    {
      "epoch": 2.162892869212572,
      "grad_norm": 4.461979389190674,
      "learning_rate": 2e-05,
      "loss": 0.1516,
      "step": 3201
    },
    {
      "epoch": 2.163568773234201,
      "grad_norm": 10.518665313720703,
      "learning_rate": 2e-05,
      "loss": 0.1801,
      "step": 3202
    },
    {
      "epoch": 2.1642446772558297,
      "grad_norm": 3.310713052749634,
      "learning_rate": 2e-05,
      "loss": 0.1455,
      "step": 3203
    },
    {
      "epoch": 2.1649205812774586,
      "grad_norm": 4.8109049797058105,
      "learning_rate": 2e-05,
      "loss": 0.1841,
      "step": 3204
    },
    {
      "epoch": 2.1655964852990874,
      "grad_norm": 3.48355770111084,
      "learning_rate": 2e-05,
      "loss": 0.1259,
      "step": 3205
    },
    {
      "epoch": 2.1662723893207163,
      "grad_norm": 3.916614294052124,
      "learning_rate": 2e-05,
      "loss": 0.1656,
      "step": 3206
    },
    {
      "epoch": 2.1669482933423456,
      "grad_norm": 3.758842706680298,
      "learning_rate": 2e-05,
      "loss": 0.1428,
      "step": 3207
    },
    {
      "epoch": 2.1676241973639745,
      "grad_norm": 3.7408957481384277,
      "learning_rate": 2e-05,
      "loss": 0.2024,
      "step": 3208
    },
    {
      "epoch": 2.1683001013856034,
      "grad_norm": 3.7503747940063477,
      "learning_rate": 2e-05,
      "loss": 0.1944,
      "step": 3209
    },
    {
      "epoch": 2.1689760054072322,
      "grad_norm": 8.04417610168457,
      "learning_rate": 2e-05,
      "loss": 0.1935,
      "step": 3210
    },
    {
      "epoch": 2.169651909428861,
      "grad_norm": 4.7889018058776855,
      "learning_rate": 2e-05,
      "loss": 0.1795,
      "step": 3211
    },
    {
      "epoch": 2.17032781345049,
      "grad_norm": 4.8595099449157715,
      "learning_rate": 2e-05,
      "loss": 0.1732,
      "step": 3212
    },
    {
      "epoch": 2.171003717472119,
      "grad_norm": 4.803548812866211,
      "learning_rate": 2e-05,
      "loss": 0.1511,
      "step": 3213
    },
    {
      "epoch": 2.1716796214937477,
      "grad_norm": 3.50727915763855,
      "learning_rate": 2e-05,
      "loss": 0.1702,
      "step": 3214
    },
    {
      "epoch": 2.1723555255153766,
      "grad_norm": 2.9792022705078125,
      "learning_rate": 2e-05,
      "loss": 0.1319,
      "step": 3215
    },
    {
      "epoch": 2.173031429537006,
      "grad_norm": 6.154310703277588,
      "learning_rate": 2e-05,
      "loss": 0.1333,
      "step": 3216
    },
    {
      "epoch": 2.1737073335586348,
      "grad_norm": 2.9292361736297607,
      "learning_rate": 2e-05,
      "loss": 0.1321,
      "step": 3217
    },
    {
      "epoch": 2.1743832375802636,
      "grad_norm": 6.368291854858398,
      "learning_rate": 2e-05,
      "loss": 0.089,
      "step": 3218
    },
    {
      "epoch": 2.1750591416018925,
      "grad_norm": 6.803452014923096,
      "learning_rate": 2e-05,
      "loss": 0.2053,
      "step": 3219
    },
    {
      "epoch": 2.1757350456235214,
      "grad_norm": 3.5454823970794678,
      "learning_rate": 2e-05,
      "loss": 0.1474,
      "step": 3220
    },
    {
      "epoch": 2.1764109496451503,
      "grad_norm": 4.612344741821289,
      "learning_rate": 2e-05,
      "loss": 0.1454,
      "step": 3221
    },
    {
      "epoch": 2.177086853666779,
      "grad_norm": 8.076301574707031,
      "learning_rate": 2e-05,
      "loss": 0.2675,
      "step": 3222
    },
    {
      "epoch": 2.1777627576884084,
      "grad_norm": 4.85161018371582,
      "learning_rate": 2e-05,
      "loss": 0.2002,
      "step": 3223
    },
    {
      "epoch": 2.1784386617100373,
      "grad_norm": 4.625650405883789,
      "learning_rate": 2e-05,
      "loss": 0.1414,
      "step": 3224
    },
    {
      "epoch": 2.179114565731666,
      "grad_norm": 11.34421443939209,
      "learning_rate": 2e-05,
      "loss": 0.1872,
      "step": 3225
    },
    {
      "epoch": 2.179790469753295,
      "grad_norm": 6.218505382537842,
      "learning_rate": 2e-05,
      "loss": 0.149,
      "step": 3226
    },
    {
      "epoch": 2.180466373774924,
      "grad_norm": 5.426534652709961,
      "learning_rate": 2e-05,
      "loss": 0.2604,
      "step": 3227
    },
    {
      "epoch": 2.181142277796553,
      "grad_norm": 8.568424224853516,
      "learning_rate": 2e-05,
      "loss": 0.1875,
      "step": 3228
    },
    {
      "epoch": 2.1818181818181817,
      "grad_norm": 2.770448684692383,
      "learning_rate": 2e-05,
      "loss": 0.158,
      "step": 3229
    },
    {
      "epoch": 2.182494085839811,
      "grad_norm": 6.3830952644348145,
      "learning_rate": 2e-05,
      "loss": 0.1676,
      "step": 3230
    },
    {
      "epoch": 2.18316998986144,
      "grad_norm": 5.813825607299805,
      "learning_rate": 2e-05,
      "loss": 0.1492,
      "step": 3231
    },
    {
      "epoch": 2.1838458938830687,
      "grad_norm": 3.0028014183044434,
      "learning_rate": 2e-05,
      "loss": 0.1531,
      "step": 3232
    },
    {
      "epoch": 2.1845217979046976,
      "grad_norm": 2.5705974102020264,
      "learning_rate": 2e-05,
      "loss": 0.172,
      "step": 3233
    },
    {
      "epoch": 2.1851977019263265,
      "grad_norm": 3.52238392829895,
      "learning_rate": 2e-05,
      "loss": 0.1033,
      "step": 3234
    },
    {
      "epoch": 2.1858736059479553,
      "grad_norm": 4.3263139724731445,
      "learning_rate": 2e-05,
      "loss": 0.1468,
      "step": 3235
    },
    {
      "epoch": 2.186549509969584,
      "grad_norm": 2.668415069580078,
      "learning_rate": 2e-05,
      "loss": 0.1044,
      "step": 3236
    },
    {
      "epoch": 2.187225413991213,
      "grad_norm": 11.69015121459961,
      "learning_rate": 2e-05,
      "loss": 0.1528,
      "step": 3237
    },
    {
      "epoch": 2.1879013180128424,
      "grad_norm": 10.755455017089844,
      "learning_rate": 2e-05,
      "loss": 0.1698,
      "step": 3238
    },
    {
      "epoch": 2.1885772220344712,
      "grad_norm": 5.8964762687683105,
      "learning_rate": 2e-05,
      "loss": 0.1682,
      "step": 3239
    },
    {
      "epoch": 2.1892531260561,
      "grad_norm": 5.413486957550049,
      "learning_rate": 2e-05,
      "loss": 0.1977,
      "step": 3240
    },
    {
      "epoch": 2.189929030077729,
      "grad_norm": 14.622255325317383,
      "learning_rate": 2e-05,
      "loss": 0.2097,
      "step": 3241
    },
    {
      "epoch": 2.190604934099358,
      "grad_norm": 6.706539154052734,
      "learning_rate": 2e-05,
      "loss": 0.1642,
      "step": 3242
    },
    {
      "epoch": 2.1912808381209867,
      "grad_norm": 7.367602348327637,
      "learning_rate": 2e-05,
      "loss": 0.2098,
      "step": 3243
    },
    {
      "epoch": 2.1919567421426156,
      "grad_norm": 5.999882221221924,
      "learning_rate": 2e-05,
      "loss": 0.2151,
      "step": 3244
    },
    {
      "epoch": 2.1926326461642445,
      "grad_norm": 5.779503345489502,
      "learning_rate": 2e-05,
      "loss": 0.2261,
      "step": 3245
    },
    {
      "epoch": 2.193308550185874,
      "grad_norm": 5.511745452880859,
      "learning_rate": 2e-05,
      "loss": 0.1956,
      "step": 3246
    },
    {
      "epoch": 2.1939844542075027,
      "grad_norm": 6.898604869842529,
      "learning_rate": 2e-05,
      "loss": 0.2097,
      "step": 3247
    },
    {
      "epoch": 2.1946603582291315,
      "grad_norm": 11.452569961547852,
      "learning_rate": 2e-05,
      "loss": 0.1568,
      "step": 3248
    },
    {
      "epoch": 2.1953362622507604,
      "grad_norm": 11.427278518676758,
      "learning_rate": 2e-05,
      "loss": 0.2205,
      "step": 3249
    },
    {
      "epoch": 2.1960121662723893,
      "grad_norm": 4.644765853881836,
      "learning_rate": 2e-05,
      "loss": 0.1255,
      "step": 3250
    },
    {
      "epoch": 2.1960121662723893,
      "eval_accuracy": 0.5858952702702702,
      "eval_loss": 0.23227770626544952,
      "eval_runtime": 42.9892,
      "eval_samples_per_second": 550.835,
      "eval_steps_per_second": 17.214,
      "step": 3250
    },
    {
      "epoch": 2.196688070294018,
      "grad_norm": 7.356512069702148,
      "learning_rate": 2e-05,
      "loss": 0.1882,
      "step": 3251
    },
    {
      "epoch": 2.197363974315647,
      "grad_norm": 3.2795941829681396,
      "learning_rate": 2e-05,
      "loss": 0.1848,
      "step": 3252
    },
    {
      "epoch": 2.1980398783372763,
      "grad_norm": 4.867314338684082,
      "learning_rate": 2e-05,
      "loss": 0.1652,
      "step": 3253
    },
    {
      "epoch": 2.198715782358905,
      "grad_norm": 9.142512321472168,
      "learning_rate": 2e-05,
      "loss": 0.196,
      "step": 3254
    },
    {
      "epoch": 2.199391686380534,
      "grad_norm": 5.0288166999816895,
      "learning_rate": 2e-05,
      "loss": 0.1694,
      "step": 3255
    },
    {
      "epoch": 2.200067590402163,
      "grad_norm": 4.122348308563232,
      "learning_rate": 2e-05,
      "loss": 0.1322,
      "step": 3256
    },
    {
      "epoch": 2.200743494423792,
      "grad_norm": 2.946875810623169,
      "learning_rate": 2e-05,
      "loss": 0.159,
      "step": 3257
    },
    {
      "epoch": 2.2014193984454207,
      "grad_norm": 7.801653861999512,
      "learning_rate": 2e-05,
      "loss": 0.1336,
      "step": 3258
    },
    {
      "epoch": 2.2020953024670495,
      "grad_norm": 3.8833436965942383,
      "learning_rate": 2e-05,
      "loss": 0.1965,
      "step": 3259
    },
    {
      "epoch": 2.2027712064886784,
      "grad_norm": 4.403292655944824,
      "learning_rate": 2e-05,
      "loss": 0.1651,
      "step": 3260
    },
    {
      "epoch": 2.2034471105103077,
      "grad_norm": 4.996804714202881,
      "learning_rate": 2e-05,
      "loss": 0.207,
      "step": 3261
    },
    {
      "epoch": 2.2041230145319366,
      "grad_norm": 11.509166717529297,
      "learning_rate": 2e-05,
      "loss": 0.1609,
      "step": 3262
    },
    {
      "epoch": 2.2047989185535655,
      "grad_norm": 6.8609395027160645,
      "learning_rate": 2e-05,
      "loss": 0.2487,
      "step": 3263
    },
    {
      "epoch": 2.2054748225751943,
      "grad_norm": 6.081023693084717,
      "learning_rate": 2e-05,
      "loss": 0.1401,
      "step": 3264
    },
    {
      "epoch": 2.206150726596823,
      "grad_norm": 9.604809761047363,
      "learning_rate": 2e-05,
      "loss": 0.1662,
      "step": 3265
    },
    {
      "epoch": 2.206826630618452,
      "grad_norm": 4.024880886077881,
      "learning_rate": 2e-05,
      "loss": 0.2199,
      "step": 3266
    },
    {
      "epoch": 2.207502534640081,
      "grad_norm": 10.118607521057129,
      "learning_rate": 2e-05,
      "loss": 0.1739,
      "step": 3267
    },
    {
      "epoch": 2.20817843866171,
      "grad_norm": 8.389554977416992,
      "learning_rate": 2e-05,
      "loss": 0.1595,
      "step": 3268
    },
    {
      "epoch": 2.208854342683339,
      "grad_norm": 6.468325614929199,
      "learning_rate": 2e-05,
      "loss": 0.153,
      "step": 3269
    },
    {
      "epoch": 2.209530246704968,
      "grad_norm": 12.115082740783691,
      "learning_rate": 2e-05,
      "loss": 0.1715,
      "step": 3270
    },
    {
      "epoch": 2.210206150726597,
      "grad_norm": 3.5346126556396484,
      "learning_rate": 2e-05,
      "loss": 0.1752,
      "step": 3271
    },
    {
      "epoch": 2.2108820547482257,
      "grad_norm": 10.470577239990234,
      "learning_rate": 2e-05,
      "loss": 0.2437,
      "step": 3272
    },
    {
      "epoch": 2.2115579587698546,
      "grad_norm": 3.4891722202301025,
      "learning_rate": 2e-05,
      "loss": 0.1455,
      "step": 3273
    },
    {
      "epoch": 2.2122338627914835,
      "grad_norm": 18.01983070373535,
      "learning_rate": 2e-05,
      "loss": 0.246,
      "step": 3274
    },
    {
      "epoch": 2.2129097668131124,
      "grad_norm": 7.538489818572998,
      "learning_rate": 2e-05,
      "loss": 0.1226,
      "step": 3275
    },
    {
      "epoch": 2.2135856708347417,
      "grad_norm": 8.32938289642334,
      "learning_rate": 2e-05,
      "loss": 0.1877,
      "step": 3276
    },
    {
      "epoch": 2.2142615748563705,
      "grad_norm": 8.28532600402832,
      "learning_rate": 2e-05,
      "loss": 0.1953,
      "step": 3277
    },
    {
      "epoch": 2.2149374788779994,
      "grad_norm": 4.876653671264648,
      "learning_rate": 2e-05,
      "loss": 0.1802,
      "step": 3278
    },
    {
      "epoch": 2.2156133828996283,
      "grad_norm": 4.218432426452637,
      "learning_rate": 2e-05,
      "loss": 0.1299,
      "step": 3279
    },
    {
      "epoch": 2.216289286921257,
      "grad_norm": 5.065981864929199,
      "learning_rate": 2e-05,
      "loss": 0.1411,
      "step": 3280
    },
    {
      "epoch": 2.216965190942886,
      "grad_norm": 2.746223211288452,
      "learning_rate": 2e-05,
      "loss": 0.1447,
      "step": 3281
    },
    {
      "epoch": 2.217641094964515,
      "grad_norm": 3.691584348678589,
      "learning_rate": 2e-05,
      "loss": 0.1296,
      "step": 3282
    },
    {
      "epoch": 2.2183169989861438,
      "grad_norm": 3.909642457962036,
      "learning_rate": 2e-05,
      "loss": 0.1704,
      "step": 3283
    },
    {
      "epoch": 2.218992903007773,
      "grad_norm": 6.441038608551025,
      "learning_rate": 2e-05,
      "loss": 0.2862,
      "step": 3284
    },
    {
      "epoch": 2.219668807029402,
      "grad_norm": 9.059642791748047,
      "learning_rate": 2e-05,
      "loss": 0.1751,
      "step": 3285
    },
    {
      "epoch": 2.220344711051031,
      "grad_norm": 11.763044357299805,
      "learning_rate": 2e-05,
      "loss": 0.2343,
      "step": 3286
    },
    {
      "epoch": 2.2210206150726597,
      "grad_norm": 3.464024066925049,
      "learning_rate": 2e-05,
      "loss": 0.1554,
      "step": 3287
    },
    {
      "epoch": 2.2216965190942886,
      "grad_norm": 7.648268699645996,
      "learning_rate": 2e-05,
      "loss": 0.1983,
      "step": 3288
    },
    {
      "epoch": 2.2223724231159174,
      "grad_norm": 4.339693069458008,
      "learning_rate": 2e-05,
      "loss": 0.1891,
      "step": 3289
    },
    {
      "epoch": 2.2230483271375463,
      "grad_norm": 8.732832908630371,
      "learning_rate": 2e-05,
      "loss": 0.192,
      "step": 3290
    },
    {
      "epoch": 2.223724231159175,
      "grad_norm": 3.6575446128845215,
      "learning_rate": 2e-05,
      "loss": 0.1831,
      "step": 3291
    },
    {
      "epoch": 2.2244001351808045,
      "grad_norm": 4.372243404388428,
      "learning_rate": 2e-05,
      "loss": 0.185,
      "step": 3292
    },
    {
      "epoch": 2.2250760392024334,
      "grad_norm": 7.240643501281738,
      "learning_rate": 2e-05,
      "loss": 0.1624,
      "step": 3293
    },
    {
      "epoch": 2.2257519432240622,
      "grad_norm": 3.7088863849639893,
      "learning_rate": 2e-05,
      "loss": 0.1991,
      "step": 3294
    },
    {
      "epoch": 2.226427847245691,
      "grad_norm": 6.232091426849365,
      "learning_rate": 2e-05,
      "loss": 0.1823,
      "step": 3295
    },
    {
      "epoch": 2.22710375126732,
      "grad_norm": 2.8064465522766113,
      "learning_rate": 2e-05,
      "loss": 0.1463,
      "step": 3296
    },
    {
      "epoch": 2.227779655288949,
      "grad_norm": 4.561887264251709,
      "learning_rate": 2e-05,
      "loss": 0.1531,
      "step": 3297
    },
    {
      "epoch": 2.2284555593105777,
      "grad_norm": 4.1704630851745605,
      "learning_rate": 2e-05,
      "loss": 0.1639,
      "step": 3298
    },
    {
      "epoch": 2.229131463332207,
      "grad_norm": 3.756300449371338,
      "learning_rate": 2e-05,
      "loss": 0.1948,
      "step": 3299
    },
    {
      "epoch": 2.229807367353836,
      "grad_norm": 9.37271785736084,
      "learning_rate": 2e-05,
      "loss": 0.1682,
      "step": 3300
    },
    {
      "epoch": 2.229807367353836,
      "eval_accuracy": 0.6205236486486486,
      "eval_loss": 0.23990772664546967,
      "eval_runtime": 43.2066,
      "eval_samples_per_second": 548.064,
      "eval_steps_per_second": 17.127,
      "step": 3300
    },
    {
      "epoch": 2.2304832713754648,
      "grad_norm": 8.793599128723145,
      "learning_rate": 2e-05,
      "loss": 0.2141,
      "step": 3301
    },
    {
      "epoch": 2.2311591753970936,
      "grad_norm": 7.260626316070557,
      "learning_rate": 2e-05,
      "loss": 0.1553,
      "step": 3302
    },
    {
      "epoch": 2.2318350794187225,
      "grad_norm": 4.608921051025391,
      "learning_rate": 2e-05,
      "loss": 0.1913,
      "step": 3303
    },
    {
      "epoch": 2.2325109834403514,
      "grad_norm": 5.585414409637451,
      "learning_rate": 2e-05,
      "loss": 0.1813,
      "step": 3304
    },
    {
      "epoch": 2.2331868874619802,
      "grad_norm": 10.24202823638916,
      "learning_rate": 2e-05,
      "loss": 0.1791,
      "step": 3305
    },
    {
      "epoch": 2.2338627914836096,
      "grad_norm": 3.997720241546631,
      "learning_rate": 2e-05,
      "loss": 0.1415,
      "step": 3306
    },
    {
      "epoch": 2.2345386955052384,
      "grad_norm": 4.925719738006592,
      "learning_rate": 2e-05,
      "loss": 0.1955,
      "step": 3307
    },
    {
      "epoch": 2.2352145995268673,
      "grad_norm": 8.441012382507324,
      "learning_rate": 2e-05,
      "loss": 0.1809,
      "step": 3308
    },
    {
      "epoch": 2.235890503548496,
      "grad_norm": 5.84602165222168,
      "learning_rate": 2e-05,
      "loss": 0.1933,
      "step": 3309
    },
    {
      "epoch": 2.236566407570125,
      "grad_norm": 13.316950798034668,
      "learning_rate": 2e-05,
      "loss": 0.1849,
      "step": 3310
    },
    {
      "epoch": 2.237242311591754,
      "grad_norm": 5.194865703582764,
      "learning_rate": 2e-05,
      "loss": 0.2196,
      "step": 3311
    },
    {
      "epoch": 2.2379182156133828,
      "grad_norm": 6.090098857879639,
      "learning_rate": 2e-05,
      "loss": 0.1693,
      "step": 3312
    },
    {
      "epoch": 2.2385941196350116,
      "grad_norm": 2.5871477127075195,
      "learning_rate": 2e-05,
      "loss": 0.1281,
      "step": 3313
    },
    {
      "epoch": 2.239270023656641,
      "grad_norm": 9.752044677734375,
      "learning_rate": 2e-05,
      "loss": 0.2253,
      "step": 3314
    },
    {
      "epoch": 2.23994592767827,
      "grad_norm": 3.833843231201172,
      "learning_rate": 2e-05,
      "loss": 0.1294,
      "step": 3315
    },
    {
      "epoch": 2.2406218316998987,
      "grad_norm": 4.260839462280273,
      "learning_rate": 2e-05,
      "loss": 0.1593,
      "step": 3316
    },
    {
      "epoch": 2.2412977357215276,
      "grad_norm": 4.1634345054626465,
      "learning_rate": 2e-05,
      "loss": 0.1973,
      "step": 3317
    },
    {
      "epoch": 2.2419736397431564,
      "grad_norm": 6.317977428436279,
      "learning_rate": 2e-05,
      "loss": 0.2015,
      "step": 3318
    },
    {
      "epoch": 2.2426495437647853,
      "grad_norm": 6.547689914703369,
      "learning_rate": 2e-05,
      "loss": 0.1482,
      "step": 3319
    },
    {
      "epoch": 2.243325447786414,
      "grad_norm": 4.845648288726807,
      "learning_rate": 2e-05,
      "loss": 0.1858,
      "step": 3320
    },
    {
      "epoch": 2.244001351808043,
      "grad_norm": 5.605300426483154,
      "learning_rate": 2e-05,
      "loss": 0.1833,
      "step": 3321
    },
    {
      "epoch": 2.2446772558296724,
      "grad_norm": 9.666301727294922,
      "learning_rate": 2e-05,
      "loss": 0.175,
      "step": 3322
    },
    {
      "epoch": 2.2453531598513012,
      "grad_norm": 6.612582206726074,
      "learning_rate": 2e-05,
      "loss": 0.1702,
      "step": 3323
    },
    {
      "epoch": 2.24602906387293,
      "grad_norm": 7.716861724853516,
      "learning_rate": 2e-05,
      "loss": 0.1716,
      "step": 3324
    },
    {
      "epoch": 2.246704967894559,
      "grad_norm": 6.116706848144531,
      "learning_rate": 2e-05,
      "loss": 0.1135,
      "step": 3325
    },
    {
      "epoch": 2.247380871916188,
      "grad_norm": 10.202956199645996,
      "learning_rate": 2e-05,
      "loss": 0.1408,
      "step": 3326
    },
    {
      "epoch": 2.2480567759378167,
      "grad_norm": 3.7535979747772217,
      "learning_rate": 2e-05,
      "loss": 0.1512,
      "step": 3327
    },
    {
      "epoch": 2.2487326799594456,
      "grad_norm": 3.9959065914154053,
      "learning_rate": 2e-05,
      "loss": 0.1773,
      "step": 3328
    },
    {
      "epoch": 2.249408583981075,
      "grad_norm": 9.778607368469238,
      "learning_rate": 2e-05,
      "loss": 0.1678,
      "step": 3329
    },
    {
      "epoch": 2.2500844880027038,
      "grad_norm": 6.176625728607178,
      "learning_rate": 2e-05,
      "loss": 0.1823,
      "step": 3330
    },
    {
      "epoch": 2.2507603920243326,
      "grad_norm": 6.047693729400635,
      "learning_rate": 2e-05,
      "loss": 0.1704,
      "step": 3331
    },
    {
      "epoch": 2.2514362960459615,
      "grad_norm": 6.231898784637451,
      "learning_rate": 2e-05,
      "loss": 0.1674,
      "step": 3332
    },
    {
      "epoch": 2.2521122000675904,
      "grad_norm": 10.183764457702637,
      "learning_rate": 2e-05,
      "loss": 0.2128,
      "step": 3333
    },
    {
      "epoch": 2.2527881040892193,
      "grad_norm": 5.37260627746582,
      "learning_rate": 2e-05,
      "loss": 0.1355,
      "step": 3334
    },
    {
      "epoch": 2.253464008110848,
      "grad_norm": 11.589009284973145,
      "learning_rate": 2e-05,
      "loss": 0.202,
      "step": 3335
    },
    {
      "epoch": 2.254139912132477,
      "grad_norm": 2.9599897861480713,
      "learning_rate": 2e-05,
      "loss": 0.1382,
      "step": 3336
    },
    {
      "epoch": 2.254815816154106,
      "grad_norm": 5.797420978546143,
      "learning_rate": 2e-05,
      "loss": 0.1299,
      "step": 3337
    },
    {
      "epoch": 2.255491720175735,
      "grad_norm": 8.565637588500977,
      "learning_rate": 2e-05,
      "loss": 0.1356,
      "step": 3338
    },
    {
      "epoch": 2.256167624197364,
      "grad_norm": 12.31452465057373,
      "learning_rate": 2e-05,
      "loss": 0.2311,
      "step": 3339
    },
    {
      "epoch": 2.256843528218993,
      "grad_norm": 8.965505599975586,
      "learning_rate": 2e-05,
      "loss": 0.146,
      "step": 3340
    },
    {
      "epoch": 2.257519432240622,
      "grad_norm": 8.910396575927734,
      "learning_rate": 2e-05,
      "loss": 0.2221,
      "step": 3341
    },
    {
      "epoch": 2.2581953362622507,
      "grad_norm": 3.5007598400115967,
      "learning_rate": 2e-05,
      "loss": 0.1451,
      "step": 3342
    },
    {
      "epoch": 2.2588712402838795,
      "grad_norm": 10.561697959899902,
      "learning_rate": 2e-05,
      "loss": 0.1741,
      "step": 3343
    },
    {
      "epoch": 2.2595471443055084,
      "grad_norm": 3.4376463890075684,
      "learning_rate": 2e-05,
      "loss": 0.1468,
      "step": 3344
    },
    {
      "epoch": 2.2602230483271377,
      "grad_norm": 9.397671699523926,
      "learning_rate": 2e-05,
      "loss": 0.1947,
      "step": 3345
    },
    {
      "epoch": 2.2608989523487666,
      "grad_norm": 4.235092639923096,
      "learning_rate": 2e-05,
      "loss": 0.1855,
      "step": 3346
    },
    {
      "epoch": 2.2615748563703955,
      "grad_norm": 5.999369144439697,
      "learning_rate": 2e-05,
      "loss": 0.1709,
      "step": 3347
    },
    {
      "epoch": 2.2622507603920243,
      "grad_norm": 6.663065433502197,
      "learning_rate": 2e-05,
      "loss": 0.2849,
      "step": 3348
    },
    {
      "epoch": 2.262926664413653,
      "grad_norm": 5.947935581207275,
      "learning_rate": 2e-05,
      "loss": 0.1352,
      "step": 3349
    },
    {
      "epoch": 2.263602568435282,
      "grad_norm": 3.333280086517334,
      "learning_rate": 2e-05,
      "loss": 0.1651,
      "step": 3350
    },
    {
      "epoch": 2.263602568435282,
      "eval_accuracy": 0.5948057432432432,
      "eval_loss": 0.23637646436691284,
      "eval_runtime": 43.88,
      "eval_samples_per_second": 539.654,
      "eval_steps_per_second": 16.864,
      "step": 3350
    },
    {
      "epoch": 2.264278472456911,
      "grad_norm": 3.2116878032684326,
      "learning_rate": 2e-05,
      "loss": 0.1557,
      "step": 3351
    },
    {
      "epoch": 2.2649543764785403,
      "grad_norm": 7.409358501434326,
      "learning_rate": 2e-05,
      "loss": 0.1573,
      "step": 3352
    },
    {
      "epoch": 2.265630280500169,
      "grad_norm": 3.891572952270508,
      "learning_rate": 2e-05,
      "loss": 0.15,
      "step": 3353
    },
    {
      "epoch": 2.266306184521798,
      "grad_norm": 3.689561605453491,
      "learning_rate": 2e-05,
      "loss": 0.1635,
      "step": 3354
    },
    {
      "epoch": 2.266982088543427,
      "grad_norm": 4.749916076660156,
      "learning_rate": 2e-05,
      "loss": 0.172,
      "step": 3355
    },
    {
      "epoch": 2.2676579925650557,
      "grad_norm": 5.277047157287598,
      "learning_rate": 2e-05,
      "loss": 0.1422,
      "step": 3356
    },
    {
      "epoch": 2.2683338965866846,
      "grad_norm": 10.052003860473633,
      "learning_rate": 2e-05,
      "loss": 0.1609,
      "step": 3357
    },
    {
      "epoch": 2.2690098006083135,
      "grad_norm": 4.5497965812683105,
      "learning_rate": 2e-05,
      "loss": 0.1458,
      "step": 3358
    },
    {
      "epoch": 2.269685704629943,
      "grad_norm": 4.279760837554932,
      "learning_rate": 2e-05,
      "loss": 0.1198,
      "step": 3359
    },
    {
      "epoch": 2.2703616086515717,
      "grad_norm": 5.208939552307129,
      "learning_rate": 2e-05,
      "loss": 0.1971,
      "step": 3360
    },
    {
      "epoch": 2.2710375126732005,
      "grad_norm": 4.2755560874938965,
      "learning_rate": 2e-05,
      "loss": 0.1135,
      "step": 3361
    },
    {
      "epoch": 2.2717134166948294,
      "grad_norm": 4.007674694061279,
      "learning_rate": 2e-05,
      "loss": 0.1255,
      "step": 3362
    },
    {
      "epoch": 2.2723893207164583,
      "grad_norm": 8.393821716308594,
      "learning_rate": 2e-05,
      "loss": 0.15,
      "step": 3363
    },
    {
      "epoch": 2.273065224738087,
      "grad_norm": 8.059743881225586,
      "learning_rate": 2e-05,
      "loss": 0.2117,
      "step": 3364
    },
    {
      "epoch": 2.273741128759716,
      "grad_norm": 5.673038005828857,
      "learning_rate": 2e-05,
      "loss": 0.1942,
      "step": 3365
    },
    {
      "epoch": 2.274417032781345,
      "grad_norm": 5.784035682678223,
      "learning_rate": 2e-05,
      "loss": 0.1522,
      "step": 3366
    },
    {
      "epoch": 2.2750929368029738,
      "grad_norm": 9.026654243469238,
      "learning_rate": 2e-05,
      "loss": 0.1642,
      "step": 3367
    },
    {
      "epoch": 2.275768840824603,
      "grad_norm": 7.638371467590332,
      "learning_rate": 2e-05,
      "loss": 0.1231,
      "step": 3368
    },
    {
      "epoch": 2.276444744846232,
      "grad_norm": 4.356266021728516,
      "learning_rate": 2e-05,
      "loss": 0.166,
      "step": 3369
    },
    {
      "epoch": 2.277120648867861,
      "grad_norm": 3.244919776916504,
      "learning_rate": 2e-05,
      "loss": 0.1247,
      "step": 3370
    },
    {
      "epoch": 2.2777965528894897,
      "grad_norm": 3.3583271503448486,
      "learning_rate": 2e-05,
      "loss": 0.1068,
      "step": 3371
    },
    {
      "epoch": 2.2784724569111185,
      "grad_norm": 6.828993797302246,
      "learning_rate": 2e-05,
      "loss": 0.2435,
      "step": 3372
    },
    {
      "epoch": 2.2791483609327474,
      "grad_norm": 11.702268600463867,
      "learning_rate": 2e-05,
      "loss": 0.2973,
      "step": 3373
    },
    {
      "epoch": 2.2798242649543763,
      "grad_norm": 4.286538600921631,
      "learning_rate": 2e-05,
      "loss": 0.1135,
      "step": 3374
    },
    {
      "epoch": 2.2805001689760056,
      "grad_norm": 6.029041290283203,
      "learning_rate": 2e-05,
      "loss": 0.1724,
      "step": 3375
    },
    {
      "epoch": 2.2811760729976345,
      "grad_norm": 7.209542751312256,
      "learning_rate": 2e-05,
      "loss": 0.2132,
      "step": 3376
    },
    {
      "epoch": 2.2818519770192633,
      "grad_norm": 5.124207973480225,
      "learning_rate": 2e-05,
      "loss": 0.1983,
      "step": 3377
    },
    {
      "epoch": 2.282527881040892,
      "grad_norm": 2.7804529666900635,
      "learning_rate": 2e-05,
      "loss": 0.1757,
      "step": 3378
    },
    {
      "epoch": 2.283203785062521,
      "grad_norm": 4.957258701324463,
      "learning_rate": 2e-05,
      "loss": 0.1418,
      "step": 3379
    },
    {
      "epoch": 2.28387968908415,
      "grad_norm": 3.830329656600952,
      "learning_rate": 2e-05,
      "loss": 0.2451,
      "step": 3380
    },
    {
      "epoch": 2.284555593105779,
      "grad_norm": 3.6943233013153076,
      "learning_rate": 2e-05,
      "loss": 0.1204,
      "step": 3381
    },
    {
      "epoch": 2.285231497127408,
      "grad_norm": 6.72092866897583,
      "learning_rate": 2e-05,
      "loss": 0.1609,
      "step": 3382
    },
    {
      "epoch": 2.285907401149037,
      "grad_norm": 4.70988655090332,
      "learning_rate": 2e-05,
      "loss": 0.2221,
      "step": 3383
    },
    {
      "epoch": 2.286583305170666,
      "grad_norm": 7.916769027709961,
      "learning_rate": 2e-05,
      "loss": 0.2141,
      "step": 3384
    },
    {
      "epoch": 2.2872592091922948,
      "grad_norm": 5.768481731414795,
      "learning_rate": 2e-05,
      "loss": 0.1768,
      "step": 3385
    },
    {
      "epoch": 2.2879351132139236,
      "grad_norm": 3.3432281017303467,
      "learning_rate": 2e-05,
      "loss": 0.1299,
      "step": 3386
    },
    {
      "epoch": 2.2886110172355525,
      "grad_norm": 4.621403694152832,
      "learning_rate": 2e-05,
      "loss": 0.1635,
      "step": 3387
    },
    {
      "epoch": 2.2892869212571814,
      "grad_norm": 2.9021315574645996,
      "learning_rate": 2e-05,
      "loss": 0.1591,
      "step": 3388
    },
    {
      "epoch": 2.2899628252788102,
      "grad_norm": 5.355762481689453,
      "learning_rate": 2e-05,
      "loss": 0.1343,
      "step": 3389
    },
    {
      "epoch": 2.290638729300439,
      "grad_norm": 5.268439292907715,
      "learning_rate": 2e-05,
      "loss": 0.1733,
      "step": 3390
    },
    {
      "epoch": 2.2913146333220684,
      "grad_norm": 9.169044494628906,
      "learning_rate": 2e-05,
      "loss": 0.1588,
      "step": 3391
    },
    {
      "epoch": 2.2919905373436973,
      "grad_norm": 3.226876974105835,
      "learning_rate": 2e-05,
      "loss": 0.1624,
      "step": 3392
    },
    {
      "epoch": 2.292666441365326,
      "grad_norm": 10.897741317749023,
      "learning_rate": 2e-05,
      "loss": 0.171,
      "step": 3393
    },
    {
      "epoch": 2.293342345386955,
      "grad_norm": 11.215130805969238,
      "learning_rate": 2e-05,
      "loss": 0.1701,
      "step": 3394
    },
    {
      "epoch": 2.294018249408584,
      "grad_norm": 3.7406058311462402,
      "learning_rate": 2e-05,
      "loss": 0.1698,
      "step": 3395
    },
    {
      "epoch": 2.2946941534302128,
      "grad_norm": 3.8354482650756836,
      "learning_rate": 2e-05,
      "loss": 0.1372,
      "step": 3396
    },
    {
      "epoch": 2.2953700574518416,
      "grad_norm": 5.13549280166626,
      "learning_rate": 2e-05,
      "loss": 0.1687,
      "step": 3397
    },
    {
      "epoch": 2.296045961473471,
      "grad_norm": 4.400850296020508,
      "learning_rate": 2e-05,
      "loss": 0.1467,
      "step": 3398
    },
    {
      "epoch": 2.2967218654951,
      "grad_norm": 6.608180046081543,
      "learning_rate": 2e-05,
      "loss": 0.163,
      "step": 3399
    },
    {
      "epoch": 2.2973977695167287,
      "grad_norm": 5.134616851806641,
      "learning_rate": 2e-05,
      "loss": 0.1942,
      "step": 3400
    },
    {
      "epoch": 2.2973977695167287,
      "eval_accuracy": 0.6225506756756757,
      "eval_loss": 0.2588506042957306,
      "eval_runtime": 42.9137,
      "eval_samples_per_second": 551.805,
      "eval_steps_per_second": 17.244,
      "step": 3400
    },
    {
      "epoch": 2.2980736735383576,
      "grad_norm": 5.8486456871032715,
      "learning_rate": 2e-05,
      "loss": 0.2013,
      "step": 3401
    },
    {
      "epoch": 2.2987495775599864,
      "grad_norm": 5.3029584884643555,
      "learning_rate": 2e-05,
      "loss": 0.2428,
      "step": 3402
    },
    {
      "epoch": 2.2994254815816153,
      "grad_norm": 6.474770545959473,
      "learning_rate": 2e-05,
      "loss": 0.2097,
      "step": 3403
    },
    {
      "epoch": 2.300101385603244,
      "grad_norm": 4.848434925079346,
      "learning_rate": 2e-05,
      "loss": 0.1867,
      "step": 3404
    },
    {
      "epoch": 2.3007772896248735,
      "grad_norm": 5.522232532501221,
      "learning_rate": 2e-05,
      "loss": 0.2155,
      "step": 3405
    },
    {
      "epoch": 2.3014531936465024,
      "grad_norm": 3.732842206954956,
      "learning_rate": 2e-05,
      "loss": 0.1375,
      "step": 3406
    },
    {
      "epoch": 2.3021290976681312,
      "grad_norm": 3.17150616645813,
      "learning_rate": 2e-05,
      "loss": 0.1667,
      "step": 3407
    },
    {
      "epoch": 2.30280500168976,
      "grad_norm": 6.7350664138793945,
      "learning_rate": 2e-05,
      "loss": 0.1999,
      "step": 3408
    },
    {
      "epoch": 2.303480905711389,
      "grad_norm": 10.877307891845703,
      "learning_rate": 2e-05,
      "loss": 0.1638,
      "step": 3409
    },
    {
      "epoch": 2.304156809733018,
      "grad_norm": 3.8318591117858887,
      "learning_rate": 2e-05,
      "loss": 0.1998,
      "step": 3410
    },
    {
      "epoch": 2.3048327137546467,
      "grad_norm": 3.490509271621704,
      "learning_rate": 2e-05,
      "loss": 0.1951,
      "step": 3411
    },
    {
      "epoch": 2.3055086177762756,
      "grad_norm": 8.993873596191406,
      "learning_rate": 2e-05,
      "loss": 0.129,
      "step": 3412
    },
    {
      "epoch": 2.3061845217979045,
      "grad_norm": 11.646345138549805,
      "learning_rate": 2e-05,
      "loss": 0.2374,
      "step": 3413
    },
    {
      "epoch": 2.3068604258195338,
      "grad_norm": 13.090377807617188,
      "learning_rate": 2e-05,
      "loss": 0.2393,
      "step": 3414
    },
    {
      "epoch": 2.3075363298411626,
      "grad_norm": 8.895794868469238,
      "learning_rate": 2e-05,
      "loss": 0.1641,
      "step": 3415
    },
    {
      "epoch": 2.3082122338627915,
      "grad_norm": 3.5558063983917236,
      "learning_rate": 2e-05,
      "loss": 0.2094,
      "step": 3416
    },
    {
      "epoch": 2.3088881378844204,
      "grad_norm": 3.1957976818084717,
      "learning_rate": 2e-05,
      "loss": 0.2046,
      "step": 3417
    },
    {
      "epoch": 2.3095640419060492,
      "grad_norm": 7.18776273727417,
      "learning_rate": 2e-05,
      "loss": 0.1639,
      "step": 3418
    },
    {
      "epoch": 2.310239945927678,
      "grad_norm": 10.970337867736816,
      "learning_rate": 2e-05,
      "loss": 0.1875,
      "step": 3419
    },
    {
      "epoch": 2.310915849949307,
      "grad_norm": 6.61180305480957,
      "learning_rate": 2e-05,
      "loss": 0.1789,
      "step": 3420
    },
    {
      "epoch": 2.3115917539709363,
      "grad_norm": 5.863350868225098,
      "learning_rate": 2e-05,
      "loss": 0.1574,
      "step": 3421
    },
    {
      "epoch": 2.312267657992565,
      "grad_norm": 5.791599750518799,
      "learning_rate": 2e-05,
      "loss": 0.1566,
      "step": 3422
    },
    {
      "epoch": 2.312943562014194,
      "grad_norm": 7.2386932373046875,
      "learning_rate": 2e-05,
      "loss": 0.2325,
      "step": 3423
    },
    {
      "epoch": 2.313619466035823,
      "grad_norm": 8.387593269348145,
      "learning_rate": 2e-05,
      "loss": 0.1875,
      "step": 3424
    },
    {
      "epoch": 2.314295370057452,
      "grad_norm": 9.802776336669922,
      "learning_rate": 2e-05,
      "loss": 0.1621,
      "step": 3425
    },
    {
      "epoch": 2.3149712740790807,
      "grad_norm": 3.5909860134124756,
      "learning_rate": 2e-05,
      "loss": 0.1525,
      "step": 3426
    },
    {
      "epoch": 2.3156471781007095,
      "grad_norm": 4.424538612365723,
      "learning_rate": 2e-05,
      "loss": 0.134,
      "step": 3427
    },
    {
      "epoch": 2.316323082122339,
      "grad_norm": 8.8836088180542,
      "learning_rate": 2e-05,
      "loss": 0.2047,
      "step": 3428
    },
    {
      "epoch": 2.3169989861439677,
      "grad_norm": 4.748858451843262,
      "learning_rate": 2e-05,
      "loss": 0.1806,
      "step": 3429
    },
    {
      "epoch": 2.3176748901655966,
      "grad_norm": 4.790547847747803,
      "learning_rate": 2e-05,
      "loss": 0.2645,
      "step": 3430
    },
    {
      "epoch": 2.3183507941872254,
      "grad_norm": 8.153573036193848,
      "learning_rate": 2e-05,
      "loss": 0.2136,
      "step": 3431
    },
    {
      "epoch": 2.3190266982088543,
      "grad_norm": 3.68000864982605,
      "learning_rate": 2e-05,
      "loss": 0.1213,
      "step": 3432
    },
    {
      "epoch": 2.319702602230483,
      "grad_norm": 4.958583831787109,
      "learning_rate": 2e-05,
      "loss": 0.1646,
      "step": 3433
    },
    {
      "epoch": 2.320378506252112,
      "grad_norm": 12.112847328186035,
      "learning_rate": 2e-05,
      "loss": 0.1944,
      "step": 3434
    },
    {
      "epoch": 2.3210544102737414,
      "grad_norm": 7.059319972991943,
      "learning_rate": 2e-05,
      "loss": 0.1841,
      "step": 3435
    },
    {
      "epoch": 2.3217303142953702,
      "grad_norm": 7.60038948059082,
      "learning_rate": 2e-05,
      "loss": 0.1587,
      "step": 3436
    },
    {
      "epoch": 2.322406218316999,
      "grad_norm": 3.8598945140838623,
      "learning_rate": 2e-05,
      "loss": 0.1846,
      "step": 3437
    },
    {
      "epoch": 2.323082122338628,
      "grad_norm": 3.7289273738861084,
      "learning_rate": 2e-05,
      "loss": 0.1681,
      "step": 3438
    },
    {
      "epoch": 2.323758026360257,
      "grad_norm": 5.635745048522949,
      "learning_rate": 2e-05,
      "loss": 0.2375,
      "step": 3439
    },
    {
      "epoch": 2.3244339303818857,
      "grad_norm": 3.194148063659668,
      "learning_rate": 2e-05,
      "loss": 0.1194,
      "step": 3440
    },
    {
      "epoch": 2.3251098344035146,
      "grad_norm": 3.841714859008789,
      "learning_rate": 2e-05,
      "loss": 0.1239,
      "step": 3441
    },
    {
      "epoch": 2.3257857384251435,
      "grad_norm": 6.641201496124268,
      "learning_rate": 2e-05,
      "loss": 0.123,
      "step": 3442
    },
    {
      "epoch": 2.3264616424467723,
      "grad_norm": 6.15051794052124,
      "learning_rate": 2e-05,
      "loss": 0.1418,
      "step": 3443
    },
    {
      "epoch": 2.3271375464684017,
      "grad_norm": 4.397256851196289,
      "learning_rate": 2e-05,
      "loss": 0.1446,
      "step": 3444
    },
    {
      "epoch": 2.3278134504900305,
      "grad_norm": 5.51765251159668,
      "learning_rate": 2e-05,
      "loss": 0.2432,
      "step": 3445
    },
    {
      "epoch": 2.3284893545116594,
      "grad_norm": 4.066997051239014,
      "learning_rate": 2e-05,
      "loss": 0.2065,
      "step": 3446
    },
    {
      "epoch": 2.3291652585332883,
      "grad_norm": 9.741137504577637,
      "learning_rate": 2e-05,
      "loss": 0.2241,
      "step": 3447
    },
    {
      "epoch": 2.329841162554917,
      "grad_norm": 5.532139301300049,
      "learning_rate": 2e-05,
      "loss": 0.1882,
      "step": 3448
    },
    {
      "epoch": 2.330517066576546,
      "grad_norm": 4.564309120178223,
      "learning_rate": 2e-05,
      "loss": 0.1607,
      "step": 3449
    },
    {
      "epoch": 2.331192970598175,
      "grad_norm": 3.845898151397705,
      "learning_rate": 2e-05,
      "loss": 0.1626,
      "step": 3450
    },
    {
      "epoch": 2.331192970598175,
      "eval_accuracy": 0.5911739864864864,
      "eval_loss": 0.23568399250507355,
      "eval_runtime": 43.0765,
      "eval_samples_per_second": 549.719,
      "eval_steps_per_second": 17.179,
      "step": 3450
    },
    {
      "epoch": 2.331868874619804,
      "grad_norm": 3.921767234802246,
      "learning_rate": 2e-05,
      "loss": 0.168,
      "step": 3451
    },
    {
      "epoch": 2.332544778641433,
      "grad_norm": 3.653740882873535,
      "learning_rate": 2e-05,
      "loss": 0.185,
      "step": 3452
    },
    {
      "epoch": 2.333220682663062,
      "grad_norm": 8.064520835876465,
      "learning_rate": 2e-05,
      "loss": 0.2527,
      "step": 3453
    },
    {
      "epoch": 2.333896586684691,
      "grad_norm": 7.814143180847168,
      "learning_rate": 2e-05,
      "loss": 0.2237,
      "step": 3454
    },
    {
      "epoch": 2.3345724907063197,
      "grad_norm": 6.045163631439209,
      "learning_rate": 2e-05,
      "loss": 0.2006,
      "step": 3455
    },
    {
      "epoch": 2.3352483947279485,
      "grad_norm": 4.963010787963867,
      "learning_rate": 2e-05,
      "loss": 0.1317,
      "step": 3456
    },
    {
      "epoch": 2.3359242987495774,
      "grad_norm": 4.678638458251953,
      "learning_rate": 2e-05,
      "loss": 0.1644,
      "step": 3457
    },
    {
      "epoch": 2.3366002027712067,
      "grad_norm": 2.415482521057129,
      "learning_rate": 2e-05,
      "loss": 0.1582,
      "step": 3458
    },
    {
      "epoch": 2.3372761067928356,
      "grad_norm": 10.953478813171387,
      "learning_rate": 2e-05,
      "loss": 0.2072,
      "step": 3459
    },
    {
      "epoch": 2.3379520108144645,
      "grad_norm": 5.980770111083984,
      "learning_rate": 2e-05,
      "loss": 0.1265,
      "step": 3460
    },
    {
      "epoch": 2.3386279148360933,
      "grad_norm": 5.874148368835449,
      "learning_rate": 2e-05,
      "loss": 0.176,
      "step": 3461
    },
    {
      "epoch": 2.339303818857722,
      "grad_norm": 3.020826578140259,
      "learning_rate": 2e-05,
      "loss": 0.1685,
      "step": 3462
    },
    {
      "epoch": 2.339979722879351,
      "grad_norm": 4.983199119567871,
      "learning_rate": 2e-05,
      "loss": 0.1844,
      "step": 3463
    },
    {
      "epoch": 2.34065562690098,
      "grad_norm": 10.844513893127441,
      "learning_rate": 2e-05,
      "loss": 0.1828,
      "step": 3464
    },
    {
      "epoch": 2.341331530922609,
      "grad_norm": 9.137381553649902,
      "learning_rate": 2e-05,
      "loss": 0.1769,
      "step": 3465
    },
    {
      "epoch": 2.3420074349442377,
      "grad_norm": 4.012994289398193,
      "learning_rate": 2e-05,
      "loss": 0.1872,
      "step": 3466
    },
    {
      "epoch": 2.342683338965867,
      "grad_norm": 4.98275899887085,
      "learning_rate": 2e-05,
      "loss": 0.2084,
      "step": 3467
    },
    {
      "epoch": 2.343359242987496,
      "grad_norm": 4.677314281463623,
      "learning_rate": 2e-05,
      "loss": 0.15,
      "step": 3468
    },
    {
      "epoch": 2.3440351470091247,
      "grad_norm": 9.890164375305176,
      "learning_rate": 2e-05,
      "loss": 0.1732,
      "step": 3469
    },
    {
      "epoch": 2.3447110510307536,
      "grad_norm": 7.913033485412598,
      "learning_rate": 2e-05,
      "loss": 0.2014,
      "step": 3470
    },
    {
      "epoch": 2.3453869550523825,
      "grad_norm": 7.337737560272217,
      "learning_rate": 2e-05,
      "loss": 0.2283,
      "step": 3471
    },
    {
      "epoch": 2.3460628590740114,
      "grad_norm": 5.568354606628418,
      "learning_rate": 2e-05,
      "loss": 0.1989,
      "step": 3472
    },
    {
      "epoch": 2.3467387630956402,
      "grad_norm": 10.048598289489746,
      "learning_rate": 2e-05,
      "loss": 0.2817,
      "step": 3473
    },
    {
      "epoch": 2.3474146671172695,
      "grad_norm": 7.9533491134643555,
      "learning_rate": 2e-05,
      "loss": 0.2254,
      "step": 3474
    },
    {
      "epoch": 2.3480905711388984,
      "grad_norm": 11.134841918945312,
      "learning_rate": 2e-05,
      "loss": 0.2299,
      "step": 3475
    },
    {
      "epoch": 2.3487664751605273,
      "grad_norm": 14.39871597290039,
      "learning_rate": 2e-05,
      "loss": 0.1581,
      "step": 3476
    },
    {
      "epoch": 2.349442379182156,
      "grad_norm": 3.314241886138916,
      "learning_rate": 2e-05,
      "loss": 0.2024,
      "step": 3477
    },
    {
      "epoch": 2.350118283203785,
      "grad_norm": 4.7493438720703125,
      "learning_rate": 2e-05,
      "loss": 0.1362,
      "step": 3478
    },
    {
      "epoch": 2.350794187225414,
      "grad_norm": 3.884521484375,
      "learning_rate": 2e-05,
      "loss": 0.1502,
      "step": 3479
    },
    {
      "epoch": 2.3514700912470428,
      "grad_norm": 3.233128786087036,
      "learning_rate": 2e-05,
      "loss": 0.1726,
      "step": 3480
    },
    {
      "epoch": 2.352145995268672,
      "grad_norm": 2.957909107208252,
      "learning_rate": 2e-05,
      "loss": 0.1576,
      "step": 3481
    },
    {
      "epoch": 2.352821899290301,
      "grad_norm": 5.721560955047607,
      "learning_rate": 2e-05,
      "loss": 0.1966,
      "step": 3482
    },
    {
      "epoch": 2.35349780331193,
      "grad_norm": 6.479107856750488,
      "learning_rate": 2e-05,
      "loss": 0.1456,
      "step": 3483
    },
    {
      "epoch": 2.3541737073335587,
      "grad_norm": 13.436686515808105,
      "learning_rate": 2e-05,
      "loss": 0.2506,
      "step": 3484
    },
    {
      "epoch": 2.3548496113551876,
      "grad_norm": 6.670453071594238,
      "learning_rate": 2e-05,
      "loss": 0.1552,
      "step": 3485
    },
    {
      "epoch": 2.3555255153768164,
      "grad_norm": 9.524511337280273,
      "learning_rate": 2e-05,
      "loss": 0.2071,
      "step": 3486
    },
    {
      "epoch": 2.3562014193984453,
      "grad_norm": 3.1257176399230957,
      "learning_rate": 2e-05,
      "loss": 0.1591,
      "step": 3487
    },
    {
      "epoch": 2.356877323420074,
      "grad_norm": 6.041022300720215,
      "learning_rate": 2e-05,
      "loss": 0.1842,
      "step": 3488
    },
    {
      "epoch": 2.357553227441703,
      "grad_norm": 2.9090487957000732,
      "learning_rate": 2e-05,
      "loss": 0.1539,
      "step": 3489
    },
    {
      "epoch": 2.3582291314633324,
      "grad_norm": 6.214077472686768,
      "learning_rate": 2e-05,
      "loss": 0.1979,
      "step": 3490
    },
    {
      "epoch": 2.358905035484961,
      "grad_norm": 8.772430419921875,
      "learning_rate": 2e-05,
      "loss": 0.1684,
      "step": 3491
    },
    {
      "epoch": 2.35958093950659,
      "grad_norm": 2.4072837829589844,
      "learning_rate": 2e-05,
      "loss": 0.1043,
      "step": 3492
    },
    {
      "epoch": 2.360256843528219,
      "grad_norm": 4.493781089782715,
      "learning_rate": 2e-05,
      "loss": 0.2086,
      "step": 3493
    },
    {
      "epoch": 2.360932747549848,
      "grad_norm": 4.729082107543945,
      "learning_rate": 2e-05,
      "loss": 0.1819,
      "step": 3494
    },
    {
      "epoch": 2.3616086515714767,
      "grad_norm": 3.530740261077881,
      "learning_rate": 2e-05,
      "loss": 0.1319,
      "step": 3495
    },
    {
      "epoch": 2.3622845555931056,
      "grad_norm": 12.37961483001709,
      "learning_rate": 2e-05,
      "loss": 0.1979,
      "step": 3496
    },
    {
      "epoch": 2.362960459614735,
      "grad_norm": 6.164299488067627,
      "learning_rate": 2e-05,
      "loss": 0.1502,
      "step": 3497
    },
    {
      "epoch": 2.3636363636363638,
      "grad_norm": 3.168825626373291,
      "learning_rate": 2e-05,
      "loss": 0.1343,
      "step": 3498
    },
    {
      "epoch": 2.3643122676579926,
      "grad_norm": 8.395450592041016,
      "learning_rate": 2e-05,
      "loss": 0.1263,
      "step": 3499
    },
    {
      "epoch": 2.3649881716796215,
      "grad_norm": 7.3723835945129395,
      "learning_rate": 2e-05,
      "loss": 0.1801,
      "step": 3500
    },
    {
      "epoch": 2.3649881716796215,
      "eval_accuracy": 0.6009712837837838,
      "eval_loss": 0.2416248470544815,
      "eval_runtime": 43.1558,
      "eval_samples_per_second": 548.71,
      "eval_steps_per_second": 17.147,
      "step": 3500
    },
    {
      "epoch": 2.3656640757012504,
      "grad_norm": 4.798622131347656,
      "learning_rate": 2e-05,
      "loss": 0.1754,
      "step": 3501
    },
    {
      "epoch": 2.3663399797228792,
      "grad_norm": 5.091726303100586,
      "learning_rate": 2e-05,
      "loss": 0.1836,
      "step": 3502
    },
    {
      "epoch": 2.367015883744508,
      "grad_norm": 7.071547508239746,
      "learning_rate": 2e-05,
      "loss": 0.1373,
      "step": 3503
    },
    {
      "epoch": 2.3676917877661374,
      "grad_norm": 9.170117378234863,
      "learning_rate": 2e-05,
      "loss": 0.2472,
      "step": 3504
    },
    {
      "epoch": 2.3683676917877663,
      "grad_norm": 5.1770405769348145,
      "learning_rate": 2e-05,
      "loss": 0.1581,
      "step": 3505
    },
    {
      "epoch": 2.369043595809395,
      "grad_norm": 7.819122791290283,
      "learning_rate": 2e-05,
      "loss": 0.1211,
      "step": 3506
    },
    {
      "epoch": 2.369719499831024,
      "grad_norm": 5.031421661376953,
      "learning_rate": 2e-05,
      "loss": 0.1837,
      "step": 3507
    },
    {
      "epoch": 2.370395403852653,
      "grad_norm": 6.376677989959717,
      "learning_rate": 2e-05,
      "loss": 0.1933,
      "step": 3508
    },
    {
      "epoch": 2.3710713078742818,
      "grad_norm": 4.0316081047058105,
      "learning_rate": 2e-05,
      "loss": 0.1395,
      "step": 3509
    },
    {
      "epoch": 2.3717472118959106,
      "grad_norm": 5.593104839324951,
      "learning_rate": 2e-05,
      "loss": 0.1489,
      "step": 3510
    },
    {
      "epoch": 2.3724231159175395,
      "grad_norm": 3.939556121826172,
      "learning_rate": 2e-05,
      "loss": 0.1679,
      "step": 3511
    },
    {
      "epoch": 2.373099019939169,
      "grad_norm": 5.373107433319092,
      "learning_rate": 2e-05,
      "loss": 0.1772,
      "step": 3512
    },
    {
      "epoch": 2.3737749239607977,
      "grad_norm": 5.134762763977051,
      "learning_rate": 2e-05,
      "loss": 0.145,
      "step": 3513
    },
    {
      "epoch": 2.3744508279824266,
      "grad_norm": 4.812692165374756,
      "learning_rate": 2e-05,
      "loss": 0.1642,
      "step": 3514
    },
    {
      "epoch": 2.3751267320040554,
      "grad_norm": 3.660113573074341,
      "learning_rate": 2e-05,
      "loss": 0.1805,
      "step": 3515
    },
    {
      "epoch": 2.3758026360256843,
      "grad_norm": 4.317796230316162,
      "learning_rate": 2e-05,
      "loss": 0.1757,
      "step": 3516
    },
    {
      "epoch": 2.376478540047313,
      "grad_norm": 4.481330871582031,
      "learning_rate": 2e-05,
      "loss": 0.1262,
      "step": 3517
    },
    {
      "epoch": 2.377154444068942,
      "grad_norm": 4.748600959777832,
      "learning_rate": 2e-05,
      "loss": 0.1268,
      "step": 3518
    },
    {
      "epoch": 2.377830348090571,
      "grad_norm": 2.9484903812408447,
      "learning_rate": 2e-05,
      "loss": 0.1503,
      "step": 3519
    },
    {
      "epoch": 2.3785062521122002,
      "grad_norm": 4.309611797332764,
      "learning_rate": 2e-05,
      "loss": 0.1434,
      "step": 3520
    },
    {
      "epoch": 2.379182156133829,
      "grad_norm": 4.735438823699951,
      "learning_rate": 2e-05,
      "loss": 0.1398,
      "step": 3521
    },
    {
      "epoch": 2.379858060155458,
      "grad_norm": 4.737332820892334,
      "learning_rate": 2e-05,
      "loss": 0.1705,
      "step": 3522
    },
    {
      "epoch": 2.380533964177087,
      "grad_norm": 4.88022518157959,
      "learning_rate": 2e-05,
      "loss": 0.134,
      "step": 3523
    },
    {
      "epoch": 2.3812098681987157,
      "grad_norm": 3.9362502098083496,
      "learning_rate": 2e-05,
      "loss": 0.1676,
      "step": 3524
    },
    {
      "epoch": 2.3818857722203446,
      "grad_norm": 8.2464599609375,
      "learning_rate": 2e-05,
      "loss": 0.2238,
      "step": 3525
    },
    {
      "epoch": 2.3825616762419735,
      "grad_norm": 5.775880336761475,
      "learning_rate": 2e-05,
      "loss": 0.1281,
      "step": 3526
    },
    {
      "epoch": 2.3832375802636028,
      "grad_norm": 3.683084011077881,
      "learning_rate": 2e-05,
      "loss": 0.1417,
      "step": 3527
    },
    {
      "epoch": 2.3839134842852316,
      "grad_norm": 5.029477119445801,
      "learning_rate": 2e-05,
      "loss": 0.1767,
      "step": 3528
    },
    {
      "epoch": 2.3845893883068605,
      "grad_norm": 4.4477667808532715,
      "learning_rate": 2e-05,
      "loss": 0.1965,
      "step": 3529
    },
    {
      "epoch": 2.3852652923284894,
      "grad_norm": 6.685290336608887,
      "learning_rate": 2e-05,
      "loss": 0.1481,
      "step": 3530
    },
    {
      "epoch": 2.3859411963501183,
      "grad_norm": 6.064494609832764,
      "learning_rate": 2e-05,
      "loss": 0.1285,
      "step": 3531
    },
    {
      "epoch": 2.386617100371747,
      "grad_norm": 6.255228042602539,
      "learning_rate": 2e-05,
      "loss": 0.2299,
      "step": 3532
    },
    {
      "epoch": 2.387293004393376,
      "grad_norm": 4.975205898284912,
      "learning_rate": 2e-05,
      "loss": 0.1638,
      "step": 3533
    },
    {
      "epoch": 2.3879689084150053,
      "grad_norm": 4.931118488311768,
      "learning_rate": 2e-05,
      "loss": 0.1342,
      "step": 3534
    },
    {
      "epoch": 2.388644812436634,
      "grad_norm": 3.0937533378601074,
      "learning_rate": 2e-05,
      "loss": 0.1123,
      "step": 3535
    },
    {
      "epoch": 2.389320716458263,
      "grad_norm": 8.796273231506348,
      "learning_rate": 2e-05,
      "loss": 0.1746,
      "step": 3536
    },
    {
      "epoch": 2.389996620479892,
      "grad_norm": 4.368141174316406,
      "learning_rate": 2e-05,
      "loss": 0.1389,
      "step": 3537
    },
    {
      "epoch": 2.390672524501521,
      "grad_norm": 5.365710258483887,
      "learning_rate": 2e-05,
      "loss": 0.1938,
      "step": 3538
    },
    {
      "epoch": 2.3913484285231497,
      "grad_norm": 6.443450450897217,
      "learning_rate": 2e-05,
      "loss": 0.2609,
      "step": 3539
    },
    {
      "epoch": 2.3920243325447785,
      "grad_norm": 4.4272847175598145,
      "learning_rate": 2e-05,
      "loss": 0.1623,
      "step": 3540
    },
    {
      "epoch": 2.3927002365664074,
      "grad_norm": 4.234251976013184,
      "learning_rate": 2e-05,
      "loss": 0.2126,
      "step": 3541
    },
    {
      "epoch": 2.3933761405880363,
      "grad_norm": 3.1130051612854004,
      "learning_rate": 2e-05,
      "loss": 0.1482,
      "step": 3542
    },
    {
      "epoch": 2.3940520446096656,
      "grad_norm": 3.8227174282073975,
      "learning_rate": 2e-05,
      "loss": 0.132,
      "step": 3543
    },
    {
      "epoch": 2.3947279486312945,
      "grad_norm": 3.1941630840301514,
      "learning_rate": 2e-05,
      "loss": 0.1641,
      "step": 3544
    },
    {
      "epoch": 2.3954038526529233,
      "grad_norm": 4.530529022216797,
      "learning_rate": 2e-05,
      "loss": 0.2324,
      "step": 3545
    },
    {
      "epoch": 2.396079756674552,
      "grad_norm": 2.946449041366577,
      "learning_rate": 2e-05,
      "loss": 0.1926,
      "step": 3546
    },
    {
      "epoch": 2.396755660696181,
      "grad_norm": 8.104991912841797,
      "learning_rate": 2e-05,
      "loss": 0.1712,
      "step": 3547
    },
    {
      "epoch": 2.39743156471781,
      "grad_norm": 4.209113597869873,
      "learning_rate": 2e-05,
      "loss": 0.1736,
      "step": 3548
    },
    {
      "epoch": 2.398107468739439,
      "grad_norm": 3.7515151500701904,
      "learning_rate": 2e-05,
      "loss": 0.138,
      "step": 3549
    },
    {
      "epoch": 2.398783372761068,
      "grad_norm": 3.824816942214966,
      "learning_rate": 2e-05,
      "loss": 0.1929,
      "step": 3550
    },
    {
      "epoch": 2.398783372761068,
      "eval_accuracy": 0.6175675675675676,
      "eval_loss": 0.2377575784921646,
      "eval_runtime": 43.9697,
      "eval_samples_per_second": 538.552,
      "eval_steps_per_second": 16.83,
      "step": 3550
    },
    {
      "epoch": 2.399459276782697,
      "grad_norm": 5.810705184936523,
      "learning_rate": 2e-05,
      "loss": 0.1606,
      "step": 3551
    },
    {
      "epoch": 2.400135180804326,
      "grad_norm": 5.512827396392822,
      "learning_rate": 2e-05,
      "loss": 0.1957,
      "step": 3552
    },
    {
      "epoch": 2.4008110848259547,
      "grad_norm": 8.96096420288086,
      "learning_rate": 2e-05,
      "loss": 0.2551,
      "step": 3553
    },
    {
      "epoch": 2.4014869888475836,
      "grad_norm": 8.88686466217041,
      "learning_rate": 2e-05,
      "loss": 0.2276,
      "step": 3554
    },
    {
      "epoch": 2.4021628928692125,
      "grad_norm": 5.826314449310303,
      "learning_rate": 2e-05,
      "loss": 0.1254,
      "step": 3555
    },
    {
      "epoch": 2.4028387968908413,
      "grad_norm": 7.353268623352051,
      "learning_rate": 2e-05,
      "loss": 0.1569,
      "step": 3556
    },
    {
      "epoch": 2.4035147009124707,
      "grad_norm": 10.254307746887207,
      "learning_rate": 2e-05,
      "loss": 0.1793,
      "step": 3557
    },
    {
      "epoch": 2.4041906049340995,
      "grad_norm": 8.428903579711914,
      "learning_rate": 2e-05,
      "loss": 0.1913,
      "step": 3558
    },
    {
      "epoch": 2.4048665089557284,
      "grad_norm": 7.618976593017578,
      "learning_rate": 2e-05,
      "loss": 0.2034,
      "step": 3559
    },
    {
      "epoch": 2.4055424129773573,
      "grad_norm": 11.462730407714844,
      "learning_rate": 2e-05,
      "loss": 0.2434,
      "step": 3560
    },
    {
      "epoch": 2.406218316998986,
      "grad_norm": 11.912050247192383,
      "learning_rate": 2e-05,
      "loss": 0.1937,
      "step": 3561
    },
    {
      "epoch": 2.406894221020615,
      "grad_norm": 5.461316108703613,
      "learning_rate": 2e-05,
      "loss": 0.2018,
      "step": 3562
    },
    {
      "epoch": 2.407570125042244,
      "grad_norm": 9.773530006408691,
      "learning_rate": 2e-05,
      "loss": 0.234,
      "step": 3563
    },
    {
      "epoch": 2.4082460290638728,
      "grad_norm": 3.2932982444763184,
      "learning_rate": 2e-05,
      "loss": 0.1063,
      "step": 3564
    },
    {
      "epoch": 2.4089219330855016,
      "grad_norm": 11.696737289428711,
      "learning_rate": 2e-05,
      "loss": 0.2578,
      "step": 3565
    },
    {
      "epoch": 2.409597837107131,
      "grad_norm": 10.361413955688477,
      "learning_rate": 2e-05,
      "loss": 0.2096,
      "step": 3566
    },
    {
      "epoch": 2.41027374112876,
      "grad_norm": 5.4052934646606445,
      "learning_rate": 2e-05,
      "loss": 0.2043,
      "step": 3567
    },
    {
      "epoch": 2.4109496451503887,
      "grad_norm": 6.412773132324219,
      "learning_rate": 2e-05,
      "loss": 0.1872,
      "step": 3568
    },
    {
      "epoch": 2.4116255491720175,
      "grad_norm": 3.888967514038086,
      "learning_rate": 2e-05,
      "loss": 0.1972,
      "step": 3569
    },
    {
      "epoch": 2.4123014531936464,
      "grad_norm": 4.711049556732178,
      "learning_rate": 2e-05,
      "loss": 0.1821,
      "step": 3570
    },
    {
      "epoch": 2.4129773572152753,
      "grad_norm": 10.273355484008789,
      "learning_rate": 2e-05,
      "loss": 0.229,
      "step": 3571
    },
    {
      "epoch": 2.413653261236904,
      "grad_norm": 10.52583122253418,
      "learning_rate": 2e-05,
      "loss": 0.2148,
      "step": 3572
    },
    {
      "epoch": 2.4143291652585335,
      "grad_norm": 11.134725570678711,
      "learning_rate": 2e-05,
      "loss": 0.2108,
      "step": 3573
    },
    {
      "epoch": 2.4150050692801623,
      "grad_norm": 4.809671878814697,
      "learning_rate": 2e-05,
      "loss": 0.1644,
      "step": 3574
    },
    {
      "epoch": 2.415680973301791,
      "grad_norm": 7.8823933601379395,
      "learning_rate": 2e-05,
      "loss": 0.1868,
      "step": 3575
    },
    {
      "epoch": 2.41635687732342,
      "grad_norm": 2.6260170936584473,
      "learning_rate": 2e-05,
      "loss": 0.1608,
      "step": 3576
    },
    {
      "epoch": 2.417032781345049,
      "grad_norm": 6.996401786804199,
      "learning_rate": 2e-05,
      "loss": 0.2523,
      "step": 3577
    },
    {
      "epoch": 2.417708685366678,
      "grad_norm": 16.429948806762695,
      "learning_rate": 2e-05,
      "loss": 0.2419,
      "step": 3578
    },
    {
      "epoch": 2.4183845893883067,
      "grad_norm": 5.822106838226318,
      "learning_rate": 2e-05,
      "loss": 0.1217,
      "step": 3579
    },
    {
      "epoch": 2.419060493409936,
      "grad_norm": 3.5314886569976807,
      "learning_rate": 2e-05,
      "loss": 0.1732,
      "step": 3580
    },
    {
      "epoch": 2.419736397431565,
      "grad_norm": 2.928501605987549,
      "learning_rate": 2e-05,
      "loss": 0.1701,
      "step": 3581
    },
    {
      "epoch": 2.4204123014531937,
      "grad_norm": 5.849956512451172,
      "learning_rate": 2e-05,
      "loss": 0.233,
      "step": 3582
    },
    {
      "epoch": 2.4210882054748226,
      "grad_norm": 5.050788879394531,
      "learning_rate": 2e-05,
      "loss": 0.1678,
      "step": 3583
    },
    {
      "epoch": 2.4217641094964515,
      "grad_norm": 8.021020889282227,
      "learning_rate": 2e-05,
      "loss": 0.1587,
      "step": 3584
    },
    {
      "epoch": 2.4224400135180804,
      "grad_norm": 3.2651925086975098,
      "learning_rate": 2e-05,
      "loss": 0.1484,
      "step": 3585
    },
    {
      "epoch": 2.4231159175397092,
      "grad_norm": 13.680055618286133,
      "learning_rate": 2e-05,
      "loss": 0.1895,
      "step": 3586
    },
    {
      "epoch": 2.423791821561338,
      "grad_norm": 5.245504856109619,
      "learning_rate": 2e-05,
      "loss": 0.1933,
      "step": 3587
    },
    {
      "epoch": 2.424467725582967,
      "grad_norm": 6.4251322746276855,
      "learning_rate": 2e-05,
      "loss": 0.2157,
      "step": 3588
    },
    {
      "epoch": 2.4251436296045963,
      "grad_norm": 3.6663687229156494,
      "learning_rate": 2e-05,
      "loss": 0.1653,
      "step": 3589
    },
    {
      "epoch": 2.425819533626225,
      "grad_norm": 5.5530524253845215,
      "learning_rate": 2e-05,
      "loss": 0.2208,
      "step": 3590
    },
    {
      "epoch": 2.426495437647854,
      "grad_norm": 6.398958683013916,
      "learning_rate": 2e-05,
      "loss": 0.1391,
      "step": 3591
    },
    {
      "epoch": 2.427171341669483,
      "grad_norm": 4.2970757484436035,
      "learning_rate": 2e-05,
      "loss": 0.1401,
      "step": 3592
    },
    {
      "epoch": 2.4278472456911118,
      "grad_norm": 5.411555290222168,
      "learning_rate": 2e-05,
      "loss": 0.219,
      "step": 3593
    },
    {
      "epoch": 2.4285231497127406,
      "grad_norm": 2.916627883911133,
      "learning_rate": 2e-05,
      "loss": 0.1177,
      "step": 3594
    },
    {
      "epoch": 2.4291990537343695,
      "grad_norm": 3.5563061237335205,
      "learning_rate": 2e-05,
      "loss": 0.1727,
      "step": 3595
    },
    {
      "epoch": 2.429874957755999,
      "grad_norm": 3.129744529724121,
      "learning_rate": 2e-05,
      "loss": 0.1571,
      "step": 3596
    },
    {
      "epoch": 2.4305508617776277,
      "grad_norm": 3.3798346519470215,
      "learning_rate": 2e-05,
      "loss": 0.1642,
      "step": 3597
    },
    {
      "epoch": 2.4312267657992566,
      "grad_norm": 16.574304580688477,
      "learning_rate": 2e-05,
      "loss": 0.2643,
      "step": 3598
    },
    {
      "epoch": 2.4319026698208854,
      "grad_norm": 4.745944976806641,
      "learning_rate": 2e-05,
      "loss": 0.1752,
      "step": 3599
    },
    {
      "epoch": 2.4325785738425143,
      "grad_norm": 4.139151096343994,
      "learning_rate": 2e-05,
      "loss": 0.202,
      "step": 3600
    },
    {
      "epoch": 2.4325785738425143,
      "eval_accuracy": 0.5850506756756757,
      "eval_loss": 0.23639054596424103,
      "eval_runtime": 43.1046,
      "eval_samples_per_second": 549.361,
      "eval_steps_per_second": 17.168,
      "step": 3600
    },
    {
      "epoch": 2.433254477864143,
      "grad_norm": 6.096058368682861,
      "learning_rate": 2e-05,
      "loss": 0.1197,
      "step": 3601
    },
    {
      "epoch": 2.433930381885772,
      "grad_norm": 4.986451148986816,
      "learning_rate": 2e-05,
      "loss": 0.1682,
      "step": 3602
    },
    {
      "epoch": 2.4346062859074014,
      "grad_norm": 8.427064895629883,
      "learning_rate": 2e-05,
      "loss": 0.1798,
      "step": 3603
    },
    {
      "epoch": 2.4352821899290302,
      "grad_norm": 9.04959487915039,
      "learning_rate": 2e-05,
      "loss": 0.2396,
      "step": 3604
    },
    {
      "epoch": 2.435958093950659,
      "grad_norm": 5.080387115478516,
      "learning_rate": 2e-05,
      "loss": 0.1972,
      "step": 3605
    },
    {
      "epoch": 2.436633997972288,
      "grad_norm": 8.318197250366211,
      "learning_rate": 2e-05,
      "loss": 0.1798,
      "step": 3606
    },
    {
      "epoch": 2.437309901993917,
      "grad_norm": 3.472079277038574,
      "learning_rate": 2e-05,
      "loss": 0.1358,
      "step": 3607
    },
    {
      "epoch": 2.4379858060155457,
      "grad_norm": 7.786823272705078,
      "learning_rate": 2e-05,
      "loss": 0.1433,
      "step": 3608
    },
    {
      "epoch": 2.4386617100371746,
      "grad_norm": 11.232202529907227,
      "learning_rate": 2e-05,
      "loss": 0.1862,
      "step": 3609
    },
    {
      "epoch": 2.439337614058804,
      "grad_norm": 7.665849685668945,
      "learning_rate": 2e-05,
      "loss": 0.1651,
      "step": 3610
    },
    {
      "epoch": 2.4400135180804328,
      "grad_norm": 7.294016361236572,
      "learning_rate": 2e-05,
      "loss": 0.1929,
      "step": 3611
    },
    {
      "epoch": 2.4406894221020616,
      "grad_norm": 3.1783769130706787,
      "learning_rate": 2e-05,
      "loss": 0.1539,
      "step": 3612
    },
    {
      "epoch": 2.4413653261236905,
      "grad_norm": 3.5696709156036377,
      "learning_rate": 2e-05,
      "loss": 0.141,
      "step": 3613
    },
    {
      "epoch": 2.4420412301453194,
      "grad_norm": 12.37879753112793,
      "learning_rate": 2e-05,
      "loss": 0.2106,
      "step": 3614
    },
    {
      "epoch": 2.4427171341669482,
      "grad_norm": 3.3170015811920166,
      "learning_rate": 2e-05,
      "loss": 0.1478,
      "step": 3615
    },
    {
      "epoch": 2.443393038188577,
      "grad_norm": 3.431633710861206,
      "learning_rate": 2e-05,
      "loss": 0.1804,
      "step": 3616
    },
    {
      "epoch": 2.444068942210206,
      "grad_norm": 3.2003285884857178,
      "learning_rate": 2e-05,
      "loss": 0.1394,
      "step": 3617
    },
    {
      "epoch": 2.444744846231835,
      "grad_norm": 8.573968887329102,
      "learning_rate": 2e-05,
      "loss": 0.1661,
      "step": 3618
    },
    {
      "epoch": 2.445420750253464,
      "grad_norm": 3.2574329376220703,
      "learning_rate": 2e-05,
      "loss": 0.1257,
      "step": 3619
    },
    {
      "epoch": 2.446096654275093,
      "grad_norm": 8.928265571594238,
      "learning_rate": 2e-05,
      "loss": 0.2388,
      "step": 3620
    },
    {
      "epoch": 2.446772558296722,
      "grad_norm": 3.72212290763855,
      "learning_rate": 2e-05,
      "loss": 0.142,
      "step": 3621
    },
    {
      "epoch": 2.447448462318351,
      "grad_norm": 6.520442008972168,
      "learning_rate": 2e-05,
      "loss": 0.1625,
      "step": 3622
    },
    {
      "epoch": 2.4481243663399797,
      "grad_norm": 4.637546062469482,
      "learning_rate": 2e-05,
      "loss": 0.2096,
      "step": 3623
    },
    {
      "epoch": 2.4488002703616085,
      "grad_norm": 3.585965394973755,
      "learning_rate": 2e-05,
      "loss": 0.1449,
      "step": 3624
    },
    {
      "epoch": 2.4494761743832374,
      "grad_norm": 3.0943186283111572,
      "learning_rate": 2e-05,
      "loss": 0.1595,
      "step": 3625
    },
    {
      "epoch": 2.4501520784048667,
      "grad_norm": 3.5424232482910156,
      "learning_rate": 2e-05,
      "loss": 0.1184,
      "step": 3626
    },
    {
      "epoch": 2.4508279824264956,
      "grad_norm": 2.987668037414551,
      "learning_rate": 2e-05,
      "loss": 0.1548,
      "step": 3627
    },
    {
      "epoch": 2.4515038864481244,
      "grad_norm": 2.7403626441955566,
      "learning_rate": 2e-05,
      "loss": 0.1129,
      "step": 3628
    },
    {
      "epoch": 2.4521797904697533,
      "grad_norm": 6.073212146759033,
      "learning_rate": 2e-05,
      "loss": 0.2344,
      "step": 3629
    },
    {
      "epoch": 2.452855694491382,
      "grad_norm": 9.38095760345459,
      "learning_rate": 2e-05,
      "loss": 0.161,
      "step": 3630
    },
    {
      "epoch": 2.453531598513011,
      "grad_norm": 3.772554636001587,
      "learning_rate": 2e-05,
      "loss": 0.1524,
      "step": 3631
    },
    {
      "epoch": 2.45420750253464,
      "grad_norm": 8.261722564697266,
      "learning_rate": 2e-05,
      "loss": 0.1971,
      "step": 3632
    },
    {
      "epoch": 2.4548834065562692,
      "grad_norm": 5.569125175476074,
      "learning_rate": 2e-05,
      "loss": 0.1782,
      "step": 3633
    },
    {
      "epoch": 2.455559310577898,
      "grad_norm": 3.35858154296875,
      "learning_rate": 2e-05,
      "loss": 0.1406,
      "step": 3634
    },
    {
      "epoch": 2.456235214599527,
      "grad_norm": 6.787368297576904,
      "learning_rate": 2e-05,
      "loss": 0.2029,
      "step": 3635
    },
    {
      "epoch": 2.456911118621156,
      "grad_norm": 6.944825649261475,
      "learning_rate": 2e-05,
      "loss": 0.1579,
      "step": 3636
    },
    {
      "epoch": 2.4575870226427847,
      "grad_norm": 9.272619247436523,
      "learning_rate": 2e-05,
      "loss": 0.1647,
      "step": 3637
    },
    {
      "epoch": 2.4582629266644136,
      "grad_norm": 9.28803825378418,
      "learning_rate": 2e-05,
      "loss": 0.1938,
      "step": 3638
    },
    {
      "epoch": 2.4589388306860425,
      "grad_norm": 6.169096946716309,
      "learning_rate": 2e-05,
      "loss": 0.1508,
      "step": 3639
    },
    {
      "epoch": 2.4596147347076713,
      "grad_norm": 3.4150195121765137,
      "learning_rate": 2e-05,
      "loss": 0.1125,
      "step": 3640
    },
    {
      "epoch": 2.4602906387293,
      "grad_norm": 5.129098892211914,
      "learning_rate": 2e-05,
      "loss": 0.2303,
      "step": 3641
    },
    {
      "epoch": 2.4609665427509295,
      "grad_norm": 3.5400962829589844,
      "learning_rate": 2e-05,
      "loss": 0.1146,
      "step": 3642
    },
    {
      "epoch": 2.4616424467725584,
      "grad_norm": 4.7925944328308105,
      "learning_rate": 2e-05,
      "loss": 0.1742,
      "step": 3643
    },
    {
      "epoch": 2.4623183507941873,
      "grad_norm": 5.4755964279174805,
      "learning_rate": 2e-05,
      "loss": 0.1828,
      "step": 3644
    },
    {
      "epoch": 2.462994254815816,
      "grad_norm": 4.846229553222656,
      "learning_rate": 2e-05,
      "loss": 0.177,
      "step": 3645
    },
    {
      "epoch": 2.463670158837445,
      "grad_norm": 4.284670352935791,
      "learning_rate": 2e-05,
      "loss": 0.1949,
      "step": 3646
    },
    {
      "epoch": 2.464346062859074,
      "grad_norm": 3.0707144737243652,
      "learning_rate": 2e-05,
      "loss": 0.1743,
      "step": 3647
    },
    {
      "epoch": 2.4650219668807027,
      "grad_norm": 7.0638933181762695,
      "learning_rate": 2e-05,
      "loss": 0.1596,
      "step": 3648
    },
    {
      "epoch": 2.465697870902332,
      "grad_norm": 5.090981483459473,
      "learning_rate": 2e-05,
      "loss": 0.2081,
      "step": 3649
    },
    {
      "epoch": 2.466373774923961,
      "grad_norm": 6.48937463760376,
      "learning_rate": 2e-05,
      "loss": 0.1333,
      "step": 3650
    },
    {
      "epoch": 2.466373774923961,
      "eval_accuracy": 0.6196790540540541,
      "eval_loss": 0.25149840116500854,
      "eval_runtime": 43.6842,
      "eval_samples_per_second": 542.072,
      "eval_steps_per_second": 16.94,
      "step": 3650
    },
    {
      "epoch": 2.46704967894559,
      "grad_norm": 6.577058792114258,
      "learning_rate": 2e-05,
      "loss": 0.1712,
      "step": 3651
    },
    {
      "epoch": 2.4677255829672187,
      "grad_norm": 2.850867509841919,
      "learning_rate": 2e-05,
      "loss": 0.1678,
      "step": 3652
    },
    {
      "epoch": 2.4684014869888475,
      "grad_norm": 5.821547508239746,
      "learning_rate": 2e-05,
      "loss": 0.1495,
      "step": 3653
    },
    {
      "epoch": 2.4690773910104764,
      "grad_norm": 7.215985298156738,
      "learning_rate": 2e-05,
      "loss": 0.206,
      "step": 3654
    },
    {
      "epoch": 2.4697532950321053,
      "grad_norm": 7.36564826965332,
      "learning_rate": 2e-05,
      "loss": 0.1928,
      "step": 3655
    },
    {
      "epoch": 2.4704291990537346,
      "grad_norm": 7.695664405822754,
      "learning_rate": 2e-05,
      "loss": 0.1591,
      "step": 3656
    },
    {
      "epoch": 2.4711051030753635,
      "grad_norm": 8.782844543457031,
      "learning_rate": 2e-05,
      "loss": 0.21,
      "step": 3657
    },
    {
      "epoch": 2.4717810070969923,
      "grad_norm": 5.658388614654541,
      "learning_rate": 2e-05,
      "loss": 0.1338,
      "step": 3658
    },
    {
      "epoch": 2.472456911118621,
      "grad_norm": 3.6739861965179443,
      "learning_rate": 2e-05,
      "loss": 0.2071,
      "step": 3659
    },
    {
      "epoch": 2.47313281514025,
      "grad_norm": 11.852744102478027,
      "learning_rate": 2e-05,
      "loss": 0.2431,
      "step": 3660
    },
    {
      "epoch": 2.473808719161879,
      "grad_norm": 3.1877598762512207,
      "learning_rate": 2e-05,
      "loss": 0.1741,
      "step": 3661
    },
    {
      "epoch": 2.474484623183508,
      "grad_norm": 4.018684387207031,
      "learning_rate": 2e-05,
      "loss": 0.2191,
      "step": 3662
    },
    {
      "epoch": 2.4751605272051367,
      "grad_norm": 3.733031988143921,
      "learning_rate": 2e-05,
      "loss": 0.1544,
      "step": 3663
    },
    {
      "epoch": 2.4758364312267656,
      "grad_norm": 2.6399409770965576,
      "learning_rate": 2e-05,
      "loss": 0.1513,
      "step": 3664
    },
    {
      "epoch": 2.476512335248395,
      "grad_norm": 7.235827922821045,
      "learning_rate": 2e-05,
      "loss": 0.1794,
      "step": 3665
    },
    {
      "epoch": 2.4771882392700237,
      "grad_norm": 6.347423553466797,
      "learning_rate": 2e-05,
      "loss": 0.1651,
      "step": 3666
    },
    {
      "epoch": 2.4778641432916526,
      "grad_norm": 4.702339172363281,
      "learning_rate": 2e-05,
      "loss": 0.1591,
      "step": 3667
    },
    {
      "epoch": 2.4785400473132815,
      "grad_norm": 9.33214282989502,
      "learning_rate": 2e-05,
      "loss": 0.2421,
      "step": 3668
    },
    {
      "epoch": 2.4792159513349104,
      "grad_norm": 3.916281223297119,
      "learning_rate": 2e-05,
      "loss": 0.1929,
      "step": 3669
    },
    {
      "epoch": 2.479891855356539,
      "grad_norm": 3.7465481758117676,
      "learning_rate": 2e-05,
      "loss": 0.2157,
      "step": 3670
    },
    {
      "epoch": 2.480567759378168,
      "grad_norm": 9.174432754516602,
      "learning_rate": 2e-05,
      "loss": 0.1585,
      "step": 3671
    },
    {
      "epoch": 2.4812436633997974,
      "grad_norm": 5.460710048675537,
      "learning_rate": 2e-05,
      "loss": 0.1821,
      "step": 3672
    },
    {
      "epoch": 2.4819195674214263,
      "grad_norm": 5.598506450653076,
      "learning_rate": 2e-05,
      "loss": 0.2368,
      "step": 3673
    },
    {
      "epoch": 2.482595471443055,
      "grad_norm": 2.6404807567596436,
      "learning_rate": 2e-05,
      "loss": 0.1578,
      "step": 3674
    },
    {
      "epoch": 2.483271375464684,
      "grad_norm": 4.975667476654053,
      "learning_rate": 2e-05,
      "loss": 0.1766,
      "step": 3675
    },
    {
      "epoch": 2.483947279486313,
      "grad_norm": 3.8979263305664062,
      "learning_rate": 2e-05,
      "loss": 0.1702,
      "step": 3676
    },
    {
      "epoch": 2.4846231835079418,
      "grad_norm": 6.578461170196533,
      "learning_rate": 2e-05,
      "loss": 0.1758,
      "step": 3677
    },
    {
      "epoch": 2.4852990875295706,
      "grad_norm": 2.561985731124878,
      "learning_rate": 2e-05,
      "loss": 0.1243,
      "step": 3678
    },
    {
      "epoch": 2.4859749915512,
      "grad_norm": 4.182596683502197,
      "learning_rate": 2e-05,
      "loss": 0.147,
      "step": 3679
    },
    {
      "epoch": 2.486650895572829,
      "grad_norm": 5.816317081451416,
      "learning_rate": 2e-05,
      "loss": 0.2083,
      "step": 3680
    },
    {
      "epoch": 2.4873267995944577,
      "grad_norm": 4.848917007446289,
      "learning_rate": 2e-05,
      "loss": 0.1358,
      "step": 3681
    },
    {
      "epoch": 2.4880027036160866,
      "grad_norm": 3.7754077911376953,
      "learning_rate": 2e-05,
      "loss": 0.154,
      "step": 3682
    },
    {
      "epoch": 2.4886786076377154,
      "grad_norm": 3.3090763092041016,
      "learning_rate": 2e-05,
      "loss": 0.1513,
      "step": 3683
    },
    {
      "epoch": 2.4893545116593443,
      "grad_norm": 3.9185028076171875,
      "learning_rate": 2e-05,
      "loss": 0.2121,
      "step": 3684
    },
    {
      "epoch": 2.490030415680973,
      "grad_norm": 4.008417129516602,
      "learning_rate": 2e-05,
      "loss": 0.1398,
      "step": 3685
    },
    {
      "epoch": 2.4907063197026025,
      "grad_norm": 5.263805389404297,
      "learning_rate": 2e-05,
      "loss": 0.169,
      "step": 3686
    },
    {
      "epoch": 2.4913822237242313,
      "grad_norm": 5.403930187225342,
      "learning_rate": 2e-05,
      "loss": 0.2072,
      "step": 3687
    },
    {
      "epoch": 2.49205812774586,
      "grad_norm": 3.8072431087493896,
      "learning_rate": 2e-05,
      "loss": 0.1092,
      "step": 3688
    },
    {
      "epoch": 2.492734031767489,
      "grad_norm": 4.375130653381348,
      "learning_rate": 2e-05,
      "loss": 0.1489,
      "step": 3689
    },
    {
      "epoch": 2.493409935789118,
      "grad_norm": 5.004925727844238,
      "learning_rate": 2e-05,
      "loss": 0.1898,
      "step": 3690
    },
    {
      "epoch": 2.494085839810747,
      "grad_norm": 2.9622061252593994,
      "learning_rate": 2e-05,
      "loss": 0.1447,
      "step": 3691
    },
    {
      "epoch": 2.4947617438323757,
      "grad_norm": 7.797007083892822,
      "learning_rate": 2e-05,
      "loss": 0.1642,
      "step": 3692
    },
    {
      "epoch": 2.4954376478540046,
      "grad_norm": 3.575624465942383,
      "learning_rate": 2e-05,
      "loss": 0.1756,
      "step": 3693
    },
    {
      "epoch": 2.4961135518756334,
      "grad_norm": 5.310842990875244,
      "learning_rate": 2e-05,
      "loss": 0.2452,
      "step": 3694
    },
    {
      "epoch": 2.4967894558972628,
      "grad_norm": 4.662956714630127,
      "learning_rate": 2e-05,
      "loss": 0.159,
      "step": 3695
    },
    {
      "epoch": 2.4974653599188916,
      "grad_norm": 3.832106590270996,
      "learning_rate": 2e-05,
      "loss": 0.1686,
      "step": 3696
    },
    {
      "epoch": 2.4981412639405205,
      "grad_norm": 6.430568695068359,
      "learning_rate": 2e-05,
      "loss": 0.1393,
      "step": 3697
    },
    {
      "epoch": 2.4988171679621494,
      "grad_norm": 4.170271873474121,
      "learning_rate": 2e-05,
      "loss": 0.1567,
      "step": 3698
    },
    {
      "epoch": 2.4994930719837782,
      "grad_norm": 6.4853620529174805,
      "learning_rate": 2e-05,
      "loss": 0.1398,
      "step": 3699
    },
    {
      "epoch": 2.500168976005407,
      "grad_norm": 8.266644477844238,
      "learning_rate": 2e-05,
      "loss": 0.1788,
      "step": 3700
    },
    {
      "epoch": 2.500168976005407,
      "eval_accuracy": 0.6128378378378379,
      "eval_loss": 0.24195338785648346,
      "eval_runtime": 44.7262,
      "eval_samples_per_second": 529.443,
      "eval_steps_per_second": 16.545,
      "step": 3700
    },
    {
      "epoch": 2.500844880027036,
      "grad_norm": 6.516618251800537,
      "learning_rate": 2e-05,
      "loss": 0.1328,
      "step": 3701
    },
    {
      "epoch": 2.5015207840486653,
      "grad_norm": 3.39359450340271,
      "learning_rate": 2e-05,
      "loss": 0.1394,
      "step": 3702
    },
    {
      "epoch": 2.502196688070294,
      "grad_norm": 5.915496349334717,
      "learning_rate": 2e-05,
      "loss": 0.1323,
      "step": 3703
    },
    {
      "epoch": 2.502872592091923,
      "grad_norm": 3.6432924270629883,
      "learning_rate": 2e-05,
      "loss": 0.1584,
      "step": 3704
    },
    {
      "epoch": 2.503548496113552,
      "grad_norm": 12.44284439086914,
      "learning_rate": 2e-05,
      "loss": 0.2496,
      "step": 3705
    },
    {
      "epoch": 2.5042244001351808,
      "grad_norm": 9.720717430114746,
      "learning_rate": 2e-05,
      "loss": 0.2417,
      "step": 3706
    },
    {
      "epoch": 2.5049003041568096,
      "grad_norm": 6.071999549865723,
      "learning_rate": 2e-05,
      "loss": 0.1451,
      "step": 3707
    },
    {
      "epoch": 2.5055762081784385,
      "grad_norm": 4.051117897033691,
      "learning_rate": 2e-05,
      "loss": 0.1504,
      "step": 3708
    },
    {
      "epoch": 2.506252112200068,
      "grad_norm": 11.606890678405762,
      "learning_rate": 2e-05,
      "loss": 0.2615,
      "step": 3709
    },
    {
      "epoch": 2.5069280162216963,
      "grad_norm": 7.94819450378418,
      "learning_rate": 2e-05,
      "loss": 0.2291,
      "step": 3710
    },
    {
      "epoch": 2.5076039202433256,
      "grad_norm": 7.01585578918457,
      "learning_rate": 2e-05,
      "loss": 0.199,
      "step": 3711
    },
    {
      "epoch": 2.5082798242649544,
      "grad_norm": 7.199394226074219,
      "learning_rate": 2e-05,
      "loss": 0.1979,
      "step": 3712
    },
    {
      "epoch": 2.5089557282865833,
      "grad_norm": 9.821786880493164,
      "learning_rate": 2e-05,
      "loss": 0.163,
      "step": 3713
    },
    {
      "epoch": 2.509631632308212,
      "grad_norm": 2.824984550476074,
      "learning_rate": 2e-05,
      "loss": 0.1526,
      "step": 3714
    },
    {
      "epoch": 2.510307536329841,
      "grad_norm": 3.4457921981811523,
      "learning_rate": 2e-05,
      "loss": 0.1653,
      "step": 3715
    },
    {
      "epoch": 2.5109834403514704,
      "grad_norm": 11.498390197753906,
      "learning_rate": 2e-05,
      "loss": 0.1978,
      "step": 3716
    },
    {
      "epoch": 2.511659344373099,
      "grad_norm": 8.387451171875,
      "learning_rate": 2e-05,
      "loss": 0.1812,
      "step": 3717
    },
    {
      "epoch": 2.512335248394728,
      "grad_norm": 2.4313137531280518,
      "learning_rate": 2e-05,
      "loss": 0.132,
      "step": 3718
    },
    {
      "epoch": 2.513011152416357,
      "grad_norm": 3.8302266597747803,
      "learning_rate": 2e-05,
      "loss": 0.1682,
      "step": 3719
    },
    {
      "epoch": 2.513687056437986,
      "grad_norm": 4.0799407958984375,
      "learning_rate": 2e-05,
      "loss": 0.1937,
      "step": 3720
    },
    {
      "epoch": 2.5143629604596147,
      "grad_norm": 4.345461368560791,
      "learning_rate": 2e-05,
      "loss": 0.1993,
      "step": 3721
    },
    {
      "epoch": 2.5150388644812436,
      "grad_norm": 14.312705993652344,
      "learning_rate": 2e-05,
      "loss": 0.201,
      "step": 3722
    },
    {
      "epoch": 2.5157147685028725,
      "grad_norm": 11.570003509521484,
      "learning_rate": 2e-05,
      "loss": 0.1864,
      "step": 3723
    },
    {
      "epoch": 2.5163906725245013,
      "grad_norm": 5.256471633911133,
      "learning_rate": 2e-05,
      "loss": 0.1533,
      "step": 3724
    },
    {
      "epoch": 2.5170665765461306,
      "grad_norm": 2.896127700805664,
      "learning_rate": 2e-05,
      "loss": 0.1444,
      "step": 3725
    },
    {
      "epoch": 2.5177424805677595,
      "grad_norm": 5.360356330871582,
      "learning_rate": 2e-05,
      "loss": 0.2099,
      "step": 3726
    },
    {
      "epoch": 2.5184183845893884,
      "grad_norm": 4.019027233123779,
      "learning_rate": 2e-05,
      "loss": 0.1304,
      "step": 3727
    },
    {
      "epoch": 2.5190942886110173,
      "grad_norm": 11.412945747375488,
      "learning_rate": 2e-05,
      "loss": 0.225,
      "step": 3728
    },
    {
      "epoch": 2.519770192632646,
      "grad_norm": 4.437242031097412,
      "learning_rate": 2e-05,
      "loss": 0.1237,
      "step": 3729
    },
    {
      "epoch": 2.520446096654275,
      "grad_norm": 9.523957252502441,
      "learning_rate": 2e-05,
      "loss": 0.2241,
      "step": 3730
    },
    {
      "epoch": 2.521122000675904,
      "grad_norm": 8.656919479370117,
      "learning_rate": 2e-05,
      "loss": 0.1657,
      "step": 3731
    },
    {
      "epoch": 2.521797904697533,
      "grad_norm": 10.647966384887695,
      "learning_rate": 2e-05,
      "loss": 0.1936,
      "step": 3732
    },
    {
      "epoch": 2.5224738087191616,
      "grad_norm": 4.285972595214844,
      "learning_rate": 2e-05,
      "loss": 0.1352,
      "step": 3733
    },
    {
      "epoch": 2.523149712740791,
      "grad_norm": 12.56174373626709,
      "learning_rate": 2e-05,
      "loss": 0.2581,
      "step": 3734
    },
    {
      "epoch": 2.52382561676242,
      "grad_norm": 10.07620906829834,
      "learning_rate": 2e-05,
      "loss": 0.2952,
      "step": 3735
    },
    {
      "epoch": 2.5245015207840487,
      "grad_norm": 13.584002494812012,
      "learning_rate": 2e-05,
      "loss": 0.2152,
      "step": 3736
    },
    {
      "epoch": 2.5251774248056775,
      "grad_norm": 15.87375259399414,
      "learning_rate": 2e-05,
      "loss": 0.1878,
      "step": 3737
    },
    {
      "epoch": 2.5258533288273064,
      "grad_norm": 9.007230758666992,
      "learning_rate": 2e-05,
      "loss": 0.181,
      "step": 3738
    },
    {
      "epoch": 2.5265292328489357,
      "grad_norm": 3.555891752243042,
      "learning_rate": 2e-05,
      "loss": 0.1647,
      "step": 3739
    },
    {
      "epoch": 2.527205136870564,
      "grad_norm": 7.595727920532227,
      "learning_rate": 2e-05,
      "loss": 0.1812,
      "step": 3740
    },
    {
      "epoch": 2.5278810408921935,
      "grad_norm": 19.01483726501465,
      "learning_rate": 2e-05,
      "loss": 0.2636,
      "step": 3741
    },
    {
      "epoch": 2.5285569449138223,
      "grad_norm": 16.427669525146484,
      "learning_rate": 2e-05,
      "loss": 0.2579,
      "step": 3742
    },
    {
      "epoch": 2.529232848935451,
      "grad_norm": 7.527764797210693,
      "learning_rate": 2e-05,
      "loss": 0.2173,
      "step": 3743
    },
    {
      "epoch": 2.52990875295708,
      "grad_norm": 2.5841336250305176,
      "learning_rate": 2e-05,
      "loss": 0.1939,
      "step": 3744
    },
    {
      "epoch": 2.530584656978709,
      "grad_norm": 8.149871826171875,
      "learning_rate": 2e-05,
      "loss": 0.2089,
      "step": 3745
    },
    {
      "epoch": 2.531260561000338,
      "grad_norm": 2.133481025695801,
      "learning_rate": 2e-05,
      "loss": 0.1692,
      "step": 3746
    },
    {
      "epoch": 2.5319364650219667,
      "grad_norm": 4.135637283325195,
      "learning_rate": 2e-05,
      "loss": 0.1623,
      "step": 3747
    },
    {
      "epoch": 2.532612369043596,
      "grad_norm": 9.291674613952637,
      "learning_rate": 2e-05,
      "loss": 0.1819,
      "step": 3748
    },
    {
      "epoch": 2.533288273065225,
      "grad_norm": 10.112967491149902,
      "learning_rate": 2e-05,
      "loss": 0.1891,
      "step": 3749
    },
    {
      "epoch": 2.5339641770868537,
      "grad_norm": 9.783638954162598,
      "learning_rate": 2e-05,
      "loss": 0.1775,
      "step": 3750
    },
    {
      "epoch": 2.5339641770868537,
      "eval_accuracy": 0.5306587837837838,
      "eval_loss": 0.254186749458313,
      "eval_runtime": 43.2118,
      "eval_samples_per_second": 547.999,
      "eval_steps_per_second": 17.125,
      "step": 3750
    },
    {
      "epoch": 2.5346400811084826,
      "grad_norm": 16.536514282226562,
      "learning_rate": 2e-05,
      "loss": 0.2328,
      "step": 3751
    },
    {
      "epoch": 2.5353159851301115,
      "grad_norm": 4.992948055267334,
      "learning_rate": 2e-05,
      "loss": 0.1422,
      "step": 3752
    },
    {
      "epoch": 2.5359918891517403,
      "grad_norm": 3.177072286605835,
      "learning_rate": 2e-05,
      "loss": 0.1921,
      "step": 3753
    },
    {
      "epoch": 2.536667793173369,
      "grad_norm": 2.9893949031829834,
      "learning_rate": 2e-05,
      "loss": 0.1897,
      "step": 3754
    },
    {
      "epoch": 2.5373436971949985,
      "grad_norm": 2.5667643547058105,
      "learning_rate": 2e-05,
      "loss": 0.1498,
      "step": 3755
    },
    {
      "epoch": 2.5380196012166274,
      "grad_norm": 18.502717971801758,
      "learning_rate": 2e-05,
      "loss": 0.3394,
      "step": 3756
    },
    {
      "epoch": 2.5386955052382563,
      "grad_norm": 6.377475738525391,
      "learning_rate": 2e-05,
      "loss": 0.1614,
      "step": 3757
    },
    {
      "epoch": 2.539371409259885,
      "grad_norm": 12.274182319641113,
      "learning_rate": 2e-05,
      "loss": 0.2474,
      "step": 3758
    },
    {
      "epoch": 2.540047313281514,
      "grad_norm": 2.0814223289489746,
      "learning_rate": 2e-05,
      "loss": 0.129,
      "step": 3759
    },
    {
      "epoch": 2.540723217303143,
      "grad_norm": 2.9459316730499268,
      "learning_rate": 2e-05,
      "loss": 0.172,
      "step": 3760
    },
    {
      "epoch": 2.5413991213247717,
      "grad_norm": 2.1656854152679443,
      "learning_rate": 2e-05,
      "loss": 0.148,
      "step": 3761
    },
    {
      "epoch": 2.542075025346401,
      "grad_norm": 3.552518129348755,
      "learning_rate": 2e-05,
      "loss": 0.1995,
      "step": 3762
    },
    {
      "epoch": 2.5427509293680295,
      "grad_norm": 3.223020553588867,
      "learning_rate": 2e-05,
      "loss": 0.1781,
      "step": 3763
    },
    {
      "epoch": 2.543426833389659,
      "grad_norm": 6.991052627563477,
      "learning_rate": 2e-05,
      "loss": 0.1911,
      "step": 3764
    },
    {
      "epoch": 2.5441027374112877,
      "grad_norm": 8.696425437927246,
      "learning_rate": 2e-05,
      "loss": 0.1956,
      "step": 3765
    },
    {
      "epoch": 2.5447786414329165,
      "grad_norm": 6.304619312286377,
      "learning_rate": 2e-05,
      "loss": 0.1634,
      "step": 3766
    },
    {
      "epoch": 2.5454545454545454,
      "grad_norm": 6.773791313171387,
      "learning_rate": 2e-05,
      "loss": 0.1259,
      "step": 3767
    },
    {
      "epoch": 2.5461304494761743,
      "grad_norm": 3.234262704849243,
      "learning_rate": 2e-05,
      "loss": 0.1534,
      "step": 3768
    },
    {
      "epoch": 2.546806353497803,
      "grad_norm": 4.205082893371582,
      "learning_rate": 2e-05,
      "loss": 0.2247,
      "step": 3769
    },
    {
      "epoch": 2.547482257519432,
      "grad_norm": 7.469764709472656,
      "learning_rate": 2e-05,
      "loss": 0.2171,
      "step": 3770
    },
    {
      "epoch": 2.5481581615410613,
      "grad_norm": 5.174887180328369,
      "learning_rate": 2e-05,
      "loss": 0.26,
      "step": 3771
    },
    {
      "epoch": 2.54883406556269,
      "grad_norm": 6.856235980987549,
      "learning_rate": 2e-05,
      "loss": 0.2053,
      "step": 3772
    },
    {
      "epoch": 2.549509969584319,
      "grad_norm": 3.6185362339019775,
      "learning_rate": 2e-05,
      "loss": 0.1588,
      "step": 3773
    },
    {
      "epoch": 2.550185873605948,
      "grad_norm": 6.973910331726074,
      "learning_rate": 2e-05,
      "loss": 0.2154,
      "step": 3774
    },
    {
      "epoch": 2.550861777627577,
      "grad_norm": 5.229891300201416,
      "learning_rate": 2e-05,
      "loss": 0.1981,
      "step": 3775
    },
    {
      "epoch": 2.5515376816492057,
      "grad_norm": 6.490280628204346,
      "learning_rate": 2e-05,
      "loss": 0.2251,
      "step": 3776
    },
    {
      "epoch": 2.5522135856708346,
      "grad_norm": 2.6070618629455566,
      "learning_rate": 2e-05,
      "loss": 0.1938,
      "step": 3777
    },
    {
      "epoch": 2.552889489692464,
      "grad_norm": 4.160554885864258,
      "learning_rate": 2e-05,
      "loss": 0.1487,
      "step": 3778
    },
    {
      "epoch": 2.5535653937140927,
      "grad_norm": 4.159438133239746,
      "learning_rate": 2e-05,
      "loss": 0.1698,
      "step": 3779
    },
    {
      "epoch": 2.5542412977357216,
      "grad_norm": 2.687220573425293,
      "learning_rate": 2e-05,
      "loss": 0.1622,
      "step": 3780
    },
    {
      "epoch": 2.5549172017573505,
      "grad_norm": 2.8703272342681885,
      "learning_rate": 2e-05,
      "loss": 0.1733,
      "step": 3781
    },
    {
      "epoch": 2.5555931057789794,
      "grad_norm": 4.249571323394775,
      "learning_rate": 2e-05,
      "loss": 0.242,
      "step": 3782
    },
    {
      "epoch": 2.5562690098006082,
      "grad_norm": 2.494717836380005,
      "learning_rate": 2e-05,
      "loss": 0.1034,
      "step": 3783
    },
    {
      "epoch": 2.556944913822237,
      "grad_norm": 6.239696979522705,
      "learning_rate": 2e-05,
      "loss": 0.1663,
      "step": 3784
    },
    {
      "epoch": 2.5576208178438664,
      "grad_norm": 3.6900289058685303,
      "learning_rate": 2e-05,
      "loss": 0.1984,
      "step": 3785
    },
    {
      "epoch": 2.558296721865495,
      "grad_norm": 4.887582302093506,
      "learning_rate": 2e-05,
      "loss": 0.1699,
      "step": 3786
    },
    {
      "epoch": 2.558972625887124,
      "grad_norm": 4.697784423828125,
      "learning_rate": 2e-05,
      "loss": 0.1811,
      "step": 3787
    },
    {
      "epoch": 2.559648529908753,
      "grad_norm": 3.5335211753845215,
      "learning_rate": 2e-05,
      "loss": 0.1578,
      "step": 3788
    },
    {
      "epoch": 2.560324433930382,
      "grad_norm": 4.038930416107178,
      "learning_rate": 2e-05,
      "loss": 0.251,
      "step": 3789
    },
    {
      "epoch": 2.5610003379520108,
      "grad_norm": 2.5993902683258057,
      "learning_rate": 2e-05,
      "loss": 0.163,
      "step": 3790
    },
    {
      "epoch": 2.5616762419736396,
      "grad_norm": 6.5578227043151855,
      "learning_rate": 2e-05,
      "loss": 0.1547,
      "step": 3791
    },
    {
      "epoch": 2.562352145995269,
      "grad_norm": 5.210549831390381,
      "learning_rate": 2e-05,
      "loss": 0.162,
      "step": 3792
    },
    {
      "epoch": 2.5630280500168974,
      "grad_norm": 3.660827398300171,
      "learning_rate": 2e-05,
      "loss": 0.2032,
      "step": 3793
    },
    {
      "epoch": 2.5637039540385267,
      "grad_norm": 3.6849365234375,
      "learning_rate": 2e-05,
      "loss": 0.1879,
      "step": 3794
    },
    {
      "epoch": 2.5643798580601556,
      "grad_norm": 8.583539962768555,
      "learning_rate": 2e-05,
      "loss": 0.1756,
      "step": 3795
    },
    {
      "epoch": 2.5650557620817844,
      "grad_norm": 5.107412815093994,
      "learning_rate": 2e-05,
      "loss": 0.1417,
      "step": 3796
    },
    {
      "epoch": 2.5657316661034133,
      "grad_norm": 10.308847427368164,
      "learning_rate": 2e-05,
      "loss": 0.1735,
      "step": 3797
    },
    {
      "epoch": 2.566407570125042,
      "grad_norm": 6.4034624099731445,
      "learning_rate": 2e-05,
      "loss": 0.1746,
      "step": 3798
    },
    {
      "epoch": 2.567083474146671,
      "grad_norm": 6.073577880859375,
      "learning_rate": 2e-05,
      "loss": 0.142,
      "step": 3799
    },
    {
      "epoch": 2.5677593781683,
      "grad_norm": 5.266944885253906,
      "learning_rate": 2e-05,
      "loss": 0.1565,
      "step": 3800
    },
    {
      "epoch": 2.5677593781683,
      "eval_accuracy": 0.6453547297297297,
      "eval_loss": 0.2386253923177719,
      "eval_runtime": 43.9713,
      "eval_samples_per_second": 538.533,
      "eval_steps_per_second": 16.829,
      "step": 3800
    },
    {
      "epoch": 2.5684352821899292,
      "grad_norm": 6.357748985290527,
      "learning_rate": 2e-05,
      "loss": 0.1567,
      "step": 3801
    },
    {
      "epoch": 2.569111186211558,
      "grad_norm": 4.024009704589844,
      "learning_rate": 2e-05,
      "loss": 0.1825,
      "step": 3802
    },
    {
      "epoch": 2.569787090233187,
      "grad_norm": 2.684163808822632,
      "learning_rate": 2e-05,
      "loss": 0.1405,
      "step": 3803
    },
    {
      "epoch": 2.570462994254816,
      "grad_norm": 4.6390485763549805,
      "learning_rate": 2e-05,
      "loss": 0.2229,
      "step": 3804
    },
    {
      "epoch": 2.5711388982764447,
      "grad_norm": 3.9853525161743164,
      "learning_rate": 2e-05,
      "loss": 0.1895,
      "step": 3805
    },
    {
      "epoch": 2.5718148022980736,
      "grad_norm": 3.3414101600646973,
      "learning_rate": 2e-05,
      "loss": 0.157,
      "step": 3806
    },
    {
      "epoch": 2.5724907063197024,
      "grad_norm": 5.381470680236816,
      "learning_rate": 2e-05,
      "loss": 0.231,
      "step": 3807
    },
    {
      "epoch": 2.5731666103413318,
      "grad_norm": 4.216530799865723,
      "learning_rate": 2e-05,
      "loss": 0.1393,
      "step": 3808
    },
    {
      "epoch": 2.57384251436296,
      "grad_norm": 5.454718589782715,
      "learning_rate": 2e-05,
      "loss": 0.1638,
      "step": 3809
    },
    {
      "epoch": 2.5745184183845895,
      "grad_norm": 7.9302778244018555,
      "learning_rate": 2e-05,
      "loss": 0.2021,
      "step": 3810
    },
    {
      "epoch": 2.5751943224062184,
      "grad_norm": 7.141448974609375,
      "learning_rate": 2e-05,
      "loss": 0.1462,
      "step": 3811
    },
    {
      "epoch": 2.5758702264278472,
      "grad_norm": 2.8769116401672363,
      "learning_rate": 2e-05,
      "loss": 0.1425,
      "step": 3812
    },
    {
      "epoch": 2.576546130449476,
      "grad_norm": 3.3876357078552246,
      "learning_rate": 2e-05,
      "loss": 0.1324,
      "step": 3813
    },
    {
      "epoch": 2.577222034471105,
      "grad_norm": 6.2871246337890625,
      "learning_rate": 2e-05,
      "loss": 0.1356,
      "step": 3814
    },
    {
      "epoch": 2.5778979384927343,
      "grad_norm": 3.1090171337127686,
      "learning_rate": 2e-05,
      "loss": 0.1656,
      "step": 3815
    },
    {
      "epoch": 2.5785738425143627,
      "grad_norm": 3.903395891189575,
      "learning_rate": 2e-05,
      "loss": 0.1314,
      "step": 3816
    },
    {
      "epoch": 2.579249746535992,
      "grad_norm": 5.498440742492676,
      "learning_rate": 2e-05,
      "loss": 0.1397,
      "step": 3817
    },
    {
      "epoch": 2.579925650557621,
      "grad_norm": 3.4385592937469482,
      "learning_rate": 2e-05,
      "loss": 0.1739,
      "step": 3818
    },
    {
      "epoch": 2.58060155457925,
      "grad_norm": 9.958258628845215,
      "learning_rate": 2e-05,
      "loss": 0.2268,
      "step": 3819
    },
    {
      "epoch": 2.5812774586008786,
      "grad_norm": 5.978479385375977,
      "learning_rate": 2e-05,
      "loss": 0.1596,
      "step": 3820
    },
    {
      "epoch": 2.5819533626225075,
      "grad_norm": 5.947302341461182,
      "learning_rate": 2e-05,
      "loss": 0.1707,
      "step": 3821
    },
    {
      "epoch": 2.5826292666441364,
      "grad_norm": 3.232821464538574,
      "learning_rate": 2e-05,
      "loss": 0.1405,
      "step": 3822
    },
    {
      "epoch": 2.5833051706657653,
      "grad_norm": 2.679095506668091,
      "learning_rate": 2e-05,
      "loss": 0.1378,
      "step": 3823
    },
    {
      "epoch": 2.5839810746873946,
      "grad_norm": 10.79228401184082,
      "learning_rate": 2e-05,
      "loss": 0.1628,
      "step": 3824
    },
    {
      "epoch": 2.5846569787090234,
      "grad_norm": 7.686194896697998,
      "learning_rate": 2e-05,
      "loss": 0.157,
      "step": 3825
    },
    {
      "epoch": 2.5853328827306523,
      "grad_norm": 2.7443084716796875,
      "learning_rate": 2e-05,
      "loss": 0.1537,
      "step": 3826
    },
    {
      "epoch": 2.586008786752281,
      "grad_norm": 5.666914939880371,
      "learning_rate": 2e-05,
      "loss": 0.1875,
      "step": 3827
    },
    {
      "epoch": 2.58668469077391,
      "grad_norm": 3.9466044902801514,
      "learning_rate": 2e-05,
      "loss": 0.1971,
      "step": 3828
    },
    {
      "epoch": 2.587360594795539,
      "grad_norm": 2.81189227104187,
      "learning_rate": 2e-05,
      "loss": 0.1302,
      "step": 3829
    },
    {
      "epoch": 2.588036498817168,
      "grad_norm": 6.807549476623535,
      "learning_rate": 2e-05,
      "loss": 0.194,
      "step": 3830
    },
    {
      "epoch": 2.588712402838797,
      "grad_norm": 7.393130302429199,
      "learning_rate": 2e-05,
      "loss": 0.1489,
      "step": 3831
    },
    {
      "epoch": 2.589388306860426,
      "grad_norm": 9.677915573120117,
      "learning_rate": 2e-05,
      "loss": 0.2499,
      "step": 3832
    },
    {
      "epoch": 2.590064210882055,
      "grad_norm": 2.7042629718780518,
      "learning_rate": 2e-05,
      "loss": 0.1255,
      "step": 3833
    },
    {
      "epoch": 2.5907401149036837,
      "grad_norm": 9.093623161315918,
      "learning_rate": 2e-05,
      "loss": 0.2577,
      "step": 3834
    },
    {
      "epoch": 2.5914160189253126,
      "grad_norm": 6.729109287261963,
      "learning_rate": 2e-05,
      "loss": 0.2167,
      "step": 3835
    },
    {
      "epoch": 2.5920919229469415,
      "grad_norm": 4.306965351104736,
      "learning_rate": 2e-05,
      "loss": 0.2284,
      "step": 3836
    },
    {
      "epoch": 2.5927678269685703,
      "grad_norm": 7.064921855926514,
      "learning_rate": 2e-05,
      "loss": 0.1748,
      "step": 3837
    },
    {
      "epoch": 2.5934437309901996,
      "grad_norm": 9.39675235748291,
      "learning_rate": 2e-05,
      "loss": 0.1777,
      "step": 3838
    },
    {
      "epoch": 2.594119635011828,
      "grad_norm": 7.53825044631958,
      "learning_rate": 2e-05,
      "loss": 0.2181,
      "step": 3839
    },
    {
      "epoch": 2.5947955390334574,
      "grad_norm": 10.145267486572266,
      "learning_rate": 2e-05,
      "loss": 0.1818,
      "step": 3840
    },
    {
      "epoch": 2.5954714430550863,
      "grad_norm": 7.497081279754639,
      "learning_rate": 2e-05,
      "loss": 0.1654,
      "step": 3841
    },
    {
      "epoch": 2.596147347076715,
      "grad_norm": 3.0035948753356934,
      "learning_rate": 2e-05,
      "loss": 0.1437,
      "step": 3842
    },
    {
      "epoch": 2.596823251098344,
      "grad_norm": 2.34785795211792,
      "learning_rate": 2e-05,
      "loss": 0.1266,
      "step": 3843
    },
    {
      "epoch": 2.597499155119973,
      "grad_norm": 3.3371164798736572,
      "learning_rate": 2e-05,
      "loss": 0.1336,
      "step": 3844
    },
    {
      "epoch": 2.5981750591416017,
      "grad_norm": 6.341569900512695,
      "learning_rate": 2e-05,
      "loss": 0.2061,
      "step": 3845
    },
    {
      "epoch": 2.5988509631632306,
      "grad_norm": 3.11114764213562,
      "learning_rate": 2e-05,
      "loss": 0.1742,
      "step": 3846
    },
    {
      "epoch": 2.59952686718486,
      "grad_norm": 3.9601855278015137,
      "learning_rate": 2e-05,
      "loss": 0.1353,
      "step": 3847
    },
    {
      "epoch": 2.600202771206489,
      "grad_norm": 4.781423091888428,
      "learning_rate": 2e-05,
      "loss": 0.1632,
      "step": 3848
    },
    {
      "epoch": 2.6008786752281177,
      "grad_norm": 5.938412189483643,
      "learning_rate": 2e-05,
      "loss": 0.2098,
      "step": 3849
    },
    {
      "epoch": 2.6015545792497465,
      "grad_norm": 4.020205020904541,
      "learning_rate": 2e-05,
      "loss": 0.1698,
      "step": 3850
    },
    {
      "epoch": 2.6015545792497465,
      "eval_accuracy": 0.6154138513513514,
      "eval_loss": 0.2306618094444275,
      "eval_runtime": 43.6036,
      "eval_samples_per_second": 543.075,
      "eval_steps_per_second": 16.971,
      "step": 3850
    },
    {
      "epoch": 2.6022304832713754,
      "grad_norm": 2.4604928493499756,
      "learning_rate": 2e-05,
      "loss": 0.1715,
      "step": 3851
    },
    {
      "epoch": 2.6029063872930043,
      "grad_norm": 2.357781171798706,
      "learning_rate": 2e-05,
      "loss": 0.1629,
      "step": 3852
    },
    {
      "epoch": 2.603582291314633,
      "grad_norm": 4.680060386657715,
      "learning_rate": 2e-05,
      "loss": 0.1411,
      "step": 3853
    },
    {
      "epoch": 2.6042581953362625,
      "grad_norm": 5.841970920562744,
      "learning_rate": 2e-05,
      "loss": 0.1234,
      "step": 3854
    },
    {
      "epoch": 2.6049340993578913,
      "grad_norm": 4.186134338378906,
      "learning_rate": 2e-05,
      "loss": 0.17,
      "step": 3855
    },
    {
      "epoch": 2.60561000337952,
      "grad_norm": 5.284934997558594,
      "learning_rate": 2e-05,
      "loss": 0.1617,
      "step": 3856
    },
    {
      "epoch": 2.606285907401149,
      "grad_norm": 11.94904899597168,
      "learning_rate": 2e-05,
      "loss": 0.2653,
      "step": 3857
    },
    {
      "epoch": 2.606961811422778,
      "grad_norm": 2.8571274280548096,
      "learning_rate": 2e-05,
      "loss": 0.1083,
      "step": 3858
    },
    {
      "epoch": 2.607637715444407,
      "grad_norm": 8.25114631652832,
      "learning_rate": 2e-05,
      "loss": 0.1599,
      "step": 3859
    },
    {
      "epoch": 2.6083136194660357,
      "grad_norm": 2.9940707683563232,
      "learning_rate": 2e-05,
      "loss": 0.149,
      "step": 3860
    },
    {
      "epoch": 2.608989523487665,
      "grad_norm": 5.7448930740356445,
      "learning_rate": 2e-05,
      "loss": 0.1614,
      "step": 3861
    },
    {
      "epoch": 2.6096654275092934,
      "grad_norm": 4.815892696380615,
      "learning_rate": 2e-05,
      "loss": 0.1658,
      "step": 3862
    },
    {
      "epoch": 2.6103413315309227,
      "grad_norm": 6.1313629150390625,
      "learning_rate": 2e-05,
      "loss": 0.2039,
      "step": 3863
    },
    {
      "epoch": 2.6110172355525516,
      "grad_norm": 2.952054977416992,
      "learning_rate": 2e-05,
      "loss": 0.1364,
      "step": 3864
    },
    {
      "epoch": 2.6116931395741805,
      "grad_norm": 4.2586774826049805,
      "learning_rate": 2e-05,
      "loss": 0.1448,
      "step": 3865
    },
    {
      "epoch": 2.6123690435958093,
      "grad_norm": 5.018320083618164,
      "learning_rate": 2e-05,
      "loss": 0.1499,
      "step": 3866
    },
    {
      "epoch": 2.613044947617438,
      "grad_norm": 9.956719398498535,
      "learning_rate": 2e-05,
      "loss": 0.1392,
      "step": 3867
    },
    {
      "epoch": 2.613720851639067,
      "grad_norm": 4.287433624267578,
      "learning_rate": 2e-05,
      "loss": 0.1585,
      "step": 3868
    },
    {
      "epoch": 2.614396755660696,
      "grad_norm": 4.067876815795898,
      "learning_rate": 2e-05,
      "loss": 0.1943,
      "step": 3869
    },
    {
      "epoch": 2.6150726596823253,
      "grad_norm": 5.251091003417969,
      "learning_rate": 2e-05,
      "loss": 0.1627,
      "step": 3870
    },
    {
      "epoch": 2.615748563703954,
      "grad_norm": 2.7000410556793213,
      "learning_rate": 2e-05,
      "loss": 0.1517,
      "step": 3871
    },
    {
      "epoch": 2.616424467725583,
      "grad_norm": 5.239898204803467,
      "learning_rate": 2e-05,
      "loss": 0.2171,
      "step": 3872
    },
    {
      "epoch": 2.617100371747212,
      "grad_norm": 5.27008056640625,
      "learning_rate": 2e-05,
      "loss": 0.2073,
      "step": 3873
    },
    {
      "epoch": 2.6177762757688408,
      "grad_norm": 7.250558853149414,
      "learning_rate": 2e-05,
      "loss": 0.163,
      "step": 3874
    },
    {
      "epoch": 2.6184521797904696,
      "grad_norm": 8.200263977050781,
      "learning_rate": 2e-05,
      "loss": 0.1965,
      "step": 3875
    },
    {
      "epoch": 2.6191280838120985,
      "grad_norm": 4.7708539962768555,
      "learning_rate": 2e-05,
      "loss": 0.157,
      "step": 3876
    },
    {
      "epoch": 2.619803987833728,
      "grad_norm": 6.144678592681885,
      "learning_rate": 2e-05,
      "loss": 0.1214,
      "step": 3877
    },
    {
      "epoch": 2.6204798918553567,
      "grad_norm": 3.6367435455322266,
      "learning_rate": 2e-05,
      "loss": 0.1631,
      "step": 3878
    },
    {
      "epoch": 2.6211557958769856,
      "grad_norm": 3.080364465713501,
      "learning_rate": 2e-05,
      "loss": 0.1424,
      "step": 3879
    },
    {
      "epoch": 2.6218316998986144,
      "grad_norm": 6.035419940948486,
      "learning_rate": 2e-05,
      "loss": 0.1749,
      "step": 3880
    },
    {
      "epoch": 2.6225076039202433,
      "grad_norm": 2.81796932220459,
      "learning_rate": 2e-05,
      "loss": 0.1778,
      "step": 3881
    },
    {
      "epoch": 2.623183507941872,
      "grad_norm": 4.599078178405762,
      "learning_rate": 2e-05,
      "loss": 0.1837,
      "step": 3882
    },
    {
      "epoch": 2.623859411963501,
      "grad_norm": 3.844712257385254,
      "learning_rate": 2e-05,
      "loss": 0.1624,
      "step": 3883
    },
    {
      "epoch": 2.6245353159851303,
      "grad_norm": 6.131062030792236,
      "learning_rate": 2e-05,
      "loss": 0.2163,
      "step": 3884
    },
    {
      "epoch": 2.6252112200067588,
      "grad_norm": 6.639335632324219,
      "learning_rate": 2e-05,
      "loss": 0.1794,
      "step": 3885
    },
    {
      "epoch": 2.625887124028388,
      "grad_norm": 4.206928253173828,
      "learning_rate": 2e-05,
      "loss": 0.1764,
      "step": 3886
    },
    {
      "epoch": 2.626563028050017,
      "grad_norm": 7.392221927642822,
      "learning_rate": 2e-05,
      "loss": 0.1831,
      "step": 3887
    },
    {
      "epoch": 2.627238932071646,
      "grad_norm": 6.696908473968506,
      "learning_rate": 2e-05,
      "loss": 0.188,
      "step": 3888
    },
    {
      "epoch": 2.6279148360932747,
      "grad_norm": 7.68948221206665,
      "learning_rate": 2e-05,
      "loss": 0.2098,
      "step": 3889
    },
    {
      "epoch": 2.6285907401149036,
      "grad_norm": 3.593106508255005,
      "learning_rate": 2e-05,
      "loss": 0.1346,
      "step": 3890
    },
    {
      "epoch": 2.629266644136533,
      "grad_norm": 3.639265537261963,
      "learning_rate": 2e-05,
      "loss": 0.1369,
      "step": 3891
    },
    {
      "epoch": 2.6299425481581613,
      "grad_norm": 4.698686599731445,
      "learning_rate": 2e-05,
      "loss": 0.1849,
      "step": 3892
    },
    {
      "epoch": 2.6306184521797906,
      "grad_norm": 7.516557693481445,
      "learning_rate": 2e-05,
      "loss": 0.1398,
      "step": 3893
    },
    {
      "epoch": 2.6312943562014195,
      "grad_norm": 9.85696029663086,
      "learning_rate": 2e-05,
      "loss": 0.2339,
      "step": 3894
    },
    {
      "epoch": 2.6319702602230484,
      "grad_norm": 8.781182289123535,
      "learning_rate": 2e-05,
      "loss": 0.186,
      "step": 3895
    },
    {
      "epoch": 2.6326461642446772,
      "grad_norm": 4.395594596862793,
      "learning_rate": 2e-05,
      "loss": 0.1195,
      "step": 3896
    },
    {
      "epoch": 2.633322068266306,
      "grad_norm": 4.57072639465332,
      "learning_rate": 2e-05,
      "loss": 0.1642,
      "step": 3897
    },
    {
      "epoch": 2.633997972287935,
      "grad_norm": 7.972568988800049,
      "learning_rate": 2e-05,
      "loss": 0.1682,
      "step": 3898
    },
    {
      "epoch": 2.634673876309564,
      "grad_norm": 3.5821585655212402,
      "learning_rate": 2e-05,
      "loss": 0.1176,
      "step": 3899
    },
    {
      "epoch": 2.635349780331193,
      "grad_norm": 14.33480167388916,
      "learning_rate": 2e-05,
      "loss": 0.3035,
      "step": 3900
    },
    {
      "epoch": 2.635349780331193,
      "eval_accuracy": 0.6195101351351351,
      "eval_loss": 0.2542659044265747,
      "eval_runtime": 43.6731,
      "eval_samples_per_second": 542.21,
      "eval_steps_per_second": 16.944,
      "step": 3900
    },
    {
      "epoch": 2.636025684352822,
      "grad_norm": 4.07547664642334,
      "learning_rate": 2e-05,
      "loss": 0.1571,
      "step": 3901
    },
    {
      "epoch": 2.636701588374451,
      "grad_norm": 5.537904262542725,
      "learning_rate": 2e-05,
      "loss": 0.1642,
      "step": 3902
    },
    {
      "epoch": 2.6373774923960798,
      "grad_norm": 9.093957901000977,
      "learning_rate": 2e-05,
      "loss": 0.1717,
      "step": 3903
    },
    {
      "epoch": 2.6380533964177086,
      "grad_norm": 2.440197229385376,
      "learning_rate": 2e-05,
      "loss": 0.1632,
      "step": 3904
    },
    {
      "epoch": 2.6387293004393375,
      "grad_norm": 4.437382221221924,
      "learning_rate": 2e-05,
      "loss": 0.2057,
      "step": 3905
    },
    {
      "epoch": 2.6394052044609664,
      "grad_norm": 6.223577499389648,
      "learning_rate": 2e-05,
      "loss": 0.152,
      "step": 3906
    },
    {
      "epoch": 2.6400811084825957,
      "grad_norm": 7.173955917358398,
      "learning_rate": 2e-05,
      "loss": 0.1963,
      "step": 3907
    },
    {
      "epoch": 2.6407570125042246,
      "grad_norm": 6.309013843536377,
      "learning_rate": 2e-05,
      "loss": 0.1487,
      "step": 3908
    },
    {
      "epoch": 2.6414329165258534,
      "grad_norm": 5.464279651641846,
      "learning_rate": 2e-05,
      "loss": 0.1621,
      "step": 3909
    },
    {
      "epoch": 2.6421088205474823,
      "grad_norm": 6.278139114379883,
      "learning_rate": 2e-05,
      "loss": 0.1898,
      "step": 3910
    },
    {
      "epoch": 2.642784724569111,
      "grad_norm": 2.949488878250122,
      "learning_rate": 2e-05,
      "loss": 0.1415,
      "step": 3911
    },
    {
      "epoch": 2.64346062859074,
      "grad_norm": 3.6515345573425293,
      "learning_rate": 2e-05,
      "loss": 0.1897,
      "step": 3912
    },
    {
      "epoch": 2.644136532612369,
      "grad_norm": 4.102426052093506,
      "learning_rate": 2e-05,
      "loss": 0.1602,
      "step": 3913
    },
    {
      "epoch": 2.6448124366339982,
      "grad_norm": 6.245779037475586,
      "learning_rate": 2e-05,
      "loss": 0.1237,
      "step": 3914
    },
    {
      "epoch": 2.6454883406556267,
      "grad_norm": 7.387156009674072,
      "learning_rate": 2e-05,
      "loss": 0.1767,
      "step": 3915
    },
    {
      "epoch": 2.646164244677256,
      "grad_norm": 6.259372234344482,
      "learning_rate": 2e-05,
      "loss": 0.1477,
      "step": 3916
    },
    {
      "epoch": 2.646840148698885,
      "grad_norm": 3.381678581237793,
      "learning_rate": 2e-05,
      "loss": 0.1147,
      "step": 3917
    },
    {
      "epoch": 2.6475160527205137,
      "grad_norm": 6.4828643798828125,
      "learning_rate": 2e-05,
      "loss": 0.177,
      "step": 3918
    },
    {
      "epoch": 2.6481919567421426,
      "grad_norm": 7.484105110168457,
      "learning_rate": 2e-05,
      "loss": 0.2058,
      "step": 3919
    },
    {
      "epoch": 2.6488678607637715,
      "grad_norm": 4.925253868103027,
      "learning_rate": 2e-05,
      "loss": 0.187,
      "step": 3920
    },
    {
      "epoch": 2.6495437647854003,
      "grad_norm": 4.529942512512207,
      "learning_rate": 2e-05,
      "loss": 0.1102,
      "step": 3921
    },
    {
      "epoch": 2.650219668807029,
      "grad_norm": 6.846439838409424,
      "learning_rate": 2e-05,
      "loss": 0.1785,
      "step": 3922
    },
    {
      "epoch": 2.6508955728286585,
      "grad_norm": 6.945457935333252,
      "learning_rate": 2e-05,
      "loss": 0.1336,
      "step": 3923
    },
    {
      "epoch": 2.6515714768502874,
      "grad_norm": 5.201254844665527,
      "learning_rate": 2e-05,
      "loss": 0.1592,
      "step": 3924
    },
    {
      "epoch": 2.6522473808719162,
      "grad_norm": 7.897097110748291,
      "learning_rate": 2e-05,
      "loss": 0.233,
      "step": 3925
    },
    {
      "epoch": 2.652923284893545,
      "grad_norm": 3.1241564750671387,
      "learning_rate": 2e-05,
      "loss": 0.1247,
      "step": 3926
    },
    {
      "epoch": 2.653599188915174,
      "grad_norm": 2.940185070037842,
      "learning_rate": 2e-05,
      "loss": 0.1658,
      "step": 3927
    },
    {
      "epoch": 2.654275092936803,
      "grad_norm": 10.478843688964844,
      "learning_rate": 2e-05,
      "loss": 0.1773,
      "step": 3928
    },
    {
      "epoch": 2.6549509969584317,
      "grad_norm": 3.286783456802368,
      "learning_rate": 2e-05,
      "loss": 0.1331,
      "step": 3929
    },
    {
      "epoch": 2.655626900980061,
      "grad_norm": 3.791581153869629,
      "learning_rate": 2e-05,
      "loss": 0.1828,
      "step": 3930
    },
    {
      "epoch": 2.65630280500169,
      "grad_norm": 5.032402515411377,
      "learning_rate": 2e-05,
      "loss": 0.1914,
      "step": 3931
    },
    {
      "epoch": 2.656978709023319,
      "grad_norm": 5.713887691497803,
      "learning_rate": 2e-05,
      "loss": 0.1552,
      "step": 3932
    },
    {
      "epoch": 2.6576546130449477,
      "grad_norm": 3.039876699447632,
      "learning_rate": 2e-05,
      "loss": 0.1552,
      "step": 3933
    },
    {
      "epoch": 2.6583305170665765,
      "grad_norm": 4.744739532470703,
      "learning_rate": 2e-05,
      "loss": 0.1398,
      "step": 3934
    },
    {
      "epoch": 2.6590064210882054,
      "grad_norm": 4.150801658630371,
      "learning_rate": 2e-05,
      "loss": 0.181,
      "step": 3935
    },
    {
      "epoch": 2.6596823251098343,
      "grad_norm": 3.6169631481170654,
      "learning_rate": 2e-05,
      "loss": 0.1738,
      "step": 3936
    },
    {
      "epoch": 2.6603582291314636,
      "grad_norm": 6.696247100830078,
      "learning_rate": 2e-05,
      "loss": 0.2467,
      "step": 3937
    },
    {
      "epoch": 2.661034133153092,
      "grad_norm": 6.386106491088867,
      "learning_rate": 2e-05,
      "loss": 0.1769,
      "step": 3938
    },
    {
      "epoch": 2.6617100371747213,
      "grad_norm": 9.772092819213867,
      "learning_rate": 2e-05,
      "loss": 0.1932,
      "step": 3939
    },
    {
      "epoch": 2.66238594119635,
      "grad_norm": 9.638934135437012,
      "learning_rate": 2e-05,
      "loss": 0.2047,
      "step": 3940
    },
    {
      "epoch": 2.663061845217979,
      "grad_norm": 3.6192626953125,
      "learning_rate": 2e-05,
      "loss": 0.1348,
      "step": 3941
    },
    {
      "epoch": 2.663737749239608,
      "grad_norm": 7.832583904266357,
      "learning_rate": 2e-05,
      "loss": 0.196,
      "step": 3942
    },
    {
      "epoch": 2.664413653261237,
      "grad_norm": 4.077947616577148,
      "learning_rate": 2e-05,
      "loss": 0.1322,
      "step": 3943
    },
    {
      "epoch": 2.6650895572828657,
      "grad_norm": 12.221175193786621,
      "learning_rate": 2e-05,
      "loss": 0.2064,
      "step": 3944
    },
    {
      "epoch": 2.6657654613044945,
      "grad_norm": 6.203710556030273,
      "learning_rate": 2e-05,
      "loss": 0.1535,
      "step": 3945
    },
    {
      "epoch": 2.666441365326124,
      "grad_norm": 3.2879369258880615,
      "learning_rate": 2e-05,
      "loss": 0.1416,
      "step": 3946
    },
    {
      "epoch": 2.6671172693477527,
      "grad_norm": 3.888216733932495,
      "learning_rate": 2e-05,
      "loss": 0.1928,
      "step": 3947
    },
    {
      "epoch": 2.6677931733693816,
      "grad_norm": 3.2732605934143066,
      "learning_rate": 2e-05,
      "loss": 0.103,
      "step": 3948
    },
    {
      "epoch": 2.6684690773910105,
      "grad_norm": 2.9895548820495605,
      "learning_rate": 2e-05,
      "loss": 0.1588,
      "step": 3949
    },
    {
      "epoch": 2.6691449814126393,
      "grad_norm": 3.2193915843963623,
      "learning_rate": 2e-05,
      "loss": 0.1704,
      "step": 3950
    },
    {
      "epoch": 2.6691449814126393,
      "eval_accuracy": 0.6204391891891892,
      "eval_loss": 0.2545826733112335,
      "eval_runtime": 43.2462,
      "eval_samples_per_second": 547.563,
      "eval_steps_per_second": 17.111,
      "step": 3950
    },
    {
      "epoch": 2.669820885434268,
      "grad_norm": 9.332496643066406,
      "learning_rate": 2e-05,
      "loss": 0.149,
      "step": 3951
    },
    {
      "epoch": 2.670496789455897,
      "grad_norm": 6.144619941711426,
      "learning_rate": 2e-05,
      "loss": 0.2463,
      "step": 3952
    },
    {
      "epoch": 2.6711726934775264,
      "grad_norm": 7.907084941864014,
      "learning_rate": 2e-05,
      "loss": 0.1363,
      "step": 3953
    },
    {
      "epoch": 2.6718485974991553,
      "grad_norm": 4.921655178070068,
      "learning_rate": 2e-05,
      "loss": 0.201,
      "step": 3954
    },
    {
      "epoch": 2.672524501520784,
      "grad_norm": 6.950207233428955,
      "learning_rate": 2e-05,
      "loss": 0.1515,
      "step": 3955
    },
    {
      "epoch": 2.673200405542413,
      "grad_norm": 7.517873287200928,
      "learning_rate": 2e-05,
      "loss": 0.1976,
      "step": 3956
    },
    {
      "epoch": 2.673876309564042,
      "grad_norm": 3.7530570030212402,
      "learning_rate": 2e-05,
      "loss": 0.1015,
      "step": 3957
    },
    {
      "epoch": 2.6745522135856707,
      "grad_norm": 4.11577033996582,
      "learning_rate": 2e-05,
      "loss": 0.1202,
      "step": 3958
    },
    {
      "epoch": 2.6752281176072996,
      "grad_norm": 11.02484130859375,
      "learning_rate": 2e-05,
      "loss": 0.2837,
      "step": 3959
    },
    {
      "epoch": 2.675904021628929,
      "grad_norm": 8.650595664978027,
      "learning_rate": 2e-05,
      "loss": 0.2411,
      "step": 3960
    },
    {
      "epoch": 2.6765799256505574,
      "grad_norm": 3.75227689743042,
      "learning_rate": 2e-05,
      "loss": 0.1148,
      "step": 3961
    },
    {
      "epoch": 2.6772558296721867,
      "grad_norm": 3.346264362335205,
      "learning_rate": 2e-05,
      "loss": 0.1507,
      "step": 3962
    },
    {
      "epoch": 2.6779317336938155,
      "grad_norm": 5.367396831512451,
      "learning_rate": 2e-05,
      "loss": 0.1094,
      "step": 3963
    },
    {
      "epoch": 2.6786076377154444,
      "grad_norm": 8.28704833984375,
      "learning_rate": 2e-05,
      "loss": 0.2199,
      "step": 3964
    },
    {
      "epoch": 2.6792835417370733,
      "grad_norm": 8.293229103088379,
      "learning_rate": 2e-05,
      "loss": 0.1702,
      "step": 3965
    },
    {
      "epoch": 2.679959445758702,
      "grad_norm": 4.515837669372559,
      "learning_rate": 2e-05,
      "loss": 0.168,
      "step": 3966
    },
    {
      "epoch": 2.6806353497803315,
      "grad_norm": 4.416254997253418,
      "learning_rate": 2e-05,
      "loss": 0.1782,
      "step": 3967
    },
    {
      "epoch": 2.68131125380196,
      "grad_norm": 4.482529163360596,
      "learning_rate": 2e-05,
      "loss": 0.1409,
      "step": 3968
    },
    {
      "epoch": 2.681987157823589,
      "grad_norm": 4.320587158203125,
      "learning_rate": 2e-05,
      "loss": 0.2002,
      "step": 3969
    },
    {
      "epoch": 2.682663061845218,
      "grad_norm": 4.150219917297363,
      "learning_rate": 2e-05,
      "loss": 0.1919,
      "step": 3970
    },
    {
      "epoch": 2.683338965866847,
      "grad_norm": 2.829458713531494,
      "learning_rate": 2e-05,
      "loss": 0.1314,
      "step": 3971
    },
    {
      "epoch": 2.684014869888476,
      "grad_norm": 5.278309345245361,
      "learning_rate": 2e-05,
      "loss": 0.1506,
      "step": 3972
    },
    {
      "epoch": 2.6846907739101047,
      "grad_norm": 9.28503131866455,
      "learning_rate": 2e-05,
      "loss": 0.18,
      "step": 3973
    },
    {
      "epoch": 2.6853666779317336,
      "grad_norm": 8.615150451660156,
      "learning_rate": 2e-05,
      "loss": 0.231,
      "step": 3974
    },
    {
      "epoch": 2.6860425819533624,
      "grad_norm": 2.371910333633423,
      "learning_rate": 2e-05,
      "loss": 0.1127,
      "step": 3975
    },
    {
      "epoch": 2.6867184859749917,
      "grad_norm": 3.8825602531433105,
      "learning_rate": 2e-05,
      "loss": 0.1687,
      "step": 3976
    },
    {
      "epoch": 2.6873943899966206,
      "grad_norm": 4.020174026489258,
      "learning_rate": 2e-05,
      "loss": 0.1234,
      "step": 3977
    },
    {
      "epoch": 2.6880702940182495,
      "grad_norm": 6.041571140289307,
      "learning_rate": 2e-05,
      "loss": 0.2137,
      "step": 3978
    },
    {
      "epoch": 2.6887461980398784,
      "grad_norm": 10.062735557556152,
      "learning_rate": 2e-05,
      "loss": 0.2508,
      "step": 3979
    },
    {
      "epoch": 2.6894221020615072,
      "grad_norm": 8.508631706237793,
      "learning_rate": 2e-05,
      "loss": 0.1556,
      "step": 3980
    },
    {
      "epoch": 2.690098006083136,
      "grad_norm": 6.412632465362549,
      "learning_rate": 2e-05,
      "loss": 0.2171,
      "step": 3981
    },
    {
      "epoch": 2.690773910104765,
      "grad_norm": 3.643195390701294,
      "learning_rate": 2e-05,
      "loss": 0.1591,
      "step": 3982
    },
    {
      "epoch": 2.6914498141263943,
      "grad_norm": 7.452207088470459,
      "learning_rate": 2e-05,
      "loss": 0.1514,
      "step": 3983
    },
    {
      "epoch": 2.6921257181480227,
      "grad_norm": 10.446961402893066,
      "learning_rate": 2e-05,
      "loss": 0.1805,
      "step": 3984
    },
    {
      "epoch": 2.692801622169652,
      "grad_norm": 3.9950571060180664,
      "learning_rate": 2e-05,
      "loss": 0.1696,
      "step": 3985
    },
    {
      "epoch": 2.693477526191281,
      "grad_norm": 5.098580360412598,
      "learning_rate": 2e-05,
      "loss": 0.1391,
      "step": 3986
    },
    {
      "epoch": 2.6941534302129098,
      "grad_norm": 7.451459884643555,
      "learning_rate": 2e-05,
      "loss": 0.1748,
      "step": 3987
    },
    {
      "epoch": 2.6948293342345386,
      "grad_norm": 8.18530559539795,
      "learning_rate": 2e-05,
      "loss": 0.1578,
      "step": 3988
    },
    {
      "epoch": 2.6955052382561675,
      "grad_norm": 2.920577049255371,
      "learning_rate": 2e-05,
      "loss": 0.1144,
      "step": 3989
    },
    {
      "epoch": 2.696181142277797,
      "grad_norm": 3.644012689590454,
      "learning_rate": 2e-05,
      "loss": 0.1935,
      "step": 3990
    },
    {
      "epoch": 2.6968570462994252,
      "grad_norm": 5.739291667938232,
      "learning_rate": 2e-05,
      "loss": 0.1833,
      "step": 3991
    },
    {
      "epoch": 2.6975329503210546,
      "grad_norm": 9.068730354309082,
      "learning_rate": 2e-05,
      "loss": 0.2109,
      "step": 3992
    },
    {
      "epoch": 2.6982088543426834,
      "grad_norm": 2.780134439468384,
      "learning_rate": 2e-05,
      "loss": 0.1326,
      "step": 3993
    },
    {
      "epoch": 2.6988847583643123,
      "grad_norm": 3.4869091510772705,
      "learning_rate": 2e-05,
      "loss": 0.1591,
      "step": 3994
    },
    {
      "epoch": 2.699560662385941,
      "grad_norm": 5.980189800262451,
      "learning_rate": 2e-05,
      "loss": 0.184,
      "step": 3995
    },
    {
      "epoch": 2.70023656640757,
      "grad_norm": 5.103518486022949,
      "learning_rate": 2e-05,
      "loss": 0.2016,
      "step": 3996
    },
    {
      "epoch": 2.700912470429199,
      "grad_norm": 6.285627365112305,
      "learning_rate": 2e-05,
      "loss": 0.1976,
      "step": 3997
    },
    {
      "epoch": 2.701588374450828,
      "grad_norm": 6.427217960357666,
      "learning_rate": 2e-05,
      "loss": 0.139,
      "step": 3998
    },
    {
      "epoch": 2.702264278472457,
      "grad_norm": 4.921926498413086,
      "learning_rate": 2e-05,
      "loss": 0.1554,
      "step": 3999
    },
    {
      "epoch": 2.702940182494086,
      "grad_norm": 4.923445224761963,
      "learning_rate": 2e-05,
      "loss": 0.1816,
      "step": 4000
    },
    {
      "epoch": 2.702940182494086,
      "eval_accuracy": 0.6177787162162162,
      "eval_loss": 0.2440662682056427,
      "eval_runtime": 43.7858,
      "eval_samples_per_second": 540.815,
      "eval_steps_per_second": 16.9,
      "step": 4000
    },
    {
      "epoch": 2.703616086515715,
      "grad_norm": 4.585531234741211,
      "learning_rate": 2e-05,
      "loss": 0.1034,
      "step": 4001
    },
    {
      "epoch": 2.7042919905373437,
      "grad_norm": 3.816891670227051,
      "learning_rate": 2e-05,
      "loss": 0.1769,
      "step": 4002
    },
    {
      "epoch": 2.7049678945589726,
      "grad_norm": 3.676326036453247,
      "learning_rate": 2e-05,
      "loss": 0.1669,
      "step": 4003
    },
    {
      "epoch": 2.7056437985806014,
      "grad_norm": 3.0309276580810547,
      "learning_rate": 2e-05,
      "loss": 0.1575,
      "step": 4004
    },
    {
      "epoch": 2.7063197026022303,
      "grad_norm": 7.131407737731934,
      "learning_rate": 2e-05,
      "loss": 0.2022,
      "step": 4005
    },
    {
      "epoch": 2.7069956066238596,
      "grad_norm": 8.2808256149292,
      "learning_rate": 2e-05,
      "loss": 0.1865,
      "step": 4006
    },
    {
      "epoch": 2.7076715106454885,
      "grad_norm": 11.504780769348145,
      "learning_rate": 2e-05,
      "loss": 0.2393,
      "step": 4007
    },
    {
      "epoch": 2.7083474146671174,
      "grad_norm": 4.6348066329956055,
      "learning_rate": 2e-05,
      "loss": 0.1062,
      "step": 4008
    },
    {
      "epoch": 2.7090233186887462,
      "grad_norm": 2.8618414402008057,
      "learning_rate": 2e-05,
      "loss": 0.1111,
      "step": 4009
    },
    {
      "epoch": 2.709699222710375,
      "grad_norm": 6.647571563720703,
      "learning_rate": 2e-05,
      "loss": 0.1727,
      "step": 4010
    },
    {
      "epoch": 2.710375126732004,
      "grad_norm": 7.566481590270996,
      "learning_rate": 2e-05,
      "loss": 0.2025,
      "step": 4011
    },
    {
      "epoch": 2.711051030753633,
      "grad_norm": 7.776313304901123,
      "learning_rate": 2e-05,
      "loss": 0.1757,
      "step": 4012
    },
    {
      "epoch": 2.711726934775262,
      "grad_norm": 13.726325988769531,
      "learning_rate": 2e-05,
      "loss": 0.1974,
      "step": 4013
    },
    {
      "epoch": 2.7124028387968906,
      "grad_norm": 2.8205463886260986,
      "learning_rate": 2e-05,
      "loss": 0.2066,
      "step": 4014
    },
    {
      "epoch": 2.71307874281852,
      "grad_norm": 2.772888422012329,
      "learning_rate": 2e-05,
      "loss": 0.1642,
      "step": 4015
    },
    {
      "epoch": 2.7137546468401488,
      "grad_norm": 3.922152519226074,
      "learning_rate": 2e-05,
      "loss": 0.1623,
      "step": 4016
    },
    {
      "epoch": 2.7144305508617776,
      "grad_norm": 3.8021433353424072,
      "learning_rate": 2e-05,
      "loss": 0.1213,
      "step": 4017
    },
    {
      "epoch": 2.7151064548834065,
      "grad_norm": 5.7329254150390625,
      "learning_rate": 2e-05,
      "loss": 0.1797,
      "step": 4018
    },
    {
      "epoch": 2.7157823589050354,
      "grad_norm": 5.181338310241699,
      "learning_rate": 2e-05,
      "loss": 0.1532,
      "step": 4019
    },
    {
      "epoch": 2.7164582629266643,
      "grad_norm": 7.237488746643066,
      "learning_rate": 2e-05,
      "loss": 0.1884,
      "step": 4020
    },
    {
      "epoch": 2.717134166948293,
      "grad_norm": 6.219926357269287,
      "learning_rate": 2e-05,
      "loss": 0.2059,
      "step": 4021
    },
    {
      "epoch": 2.7178100709699224,
      "grad_norm": 4.6019721031188965,
      "learning_rate": 2e-05,
      "loss": 0.1201,
      "step": 4022
    },
    {
      "epoch": 2.7184859749915513,
      "grad_norm": 6.685354232788086,
      "learning_rate": 2e-05,
      "loss": 0.2678,
      "step": 4023
    },
    {
      "epoch": 2.71916187901318,
      "grad_norm": 4.047056198120117,
      "learning_rate": 2e-05,
      "loss": 0.1484,
      "step": 4024
    },
    {
      "epoch": 2.719837783034809,
      "grad_norm": 3.0660908222198486,
      "learning_rate": 2e-05,
      "loss": 0.1139,
      "step": 4025
    },
    {
      "epoch": 2.720513687056438,
      "grad_norm": 4.235180854797363,
      "learning_rate": 2e-05,
      "loss": 0.1061,
      "step": 4026
    },
    {
      "epoch": 2.721189591078067,
      "grad_norm": 7.116016864776611,
      "learning_rate": 2e-05,
      "loss": 0.2129,
      "step": 4027
    },
    {
      "epoch": 2.7218654950996957,
      "grad_norm": 3.46651554107666,
      "learning_rate": 2e-05,
      "loss": 0.1595,
      "step": 4028
    },
    {
      "epoch": 2.722541399121325,
      "grad_norm": 3.357184648513794,
      "learning_rate": 2e-05,
      "loss": 0.1452,
      "step": 4029
    },
    {
      "epoch": 2.723217303142954,
      "grad_norm": 4.003505706787109,
      "learning_rate": 2e-05,
      "loss": 0.1433,
      "step": 4030
    },
    {
      "epoch": 2.7238932071645827,
      "grad_norm": 11.556512832641602,
      "learning_rate": 2e-05,
      "loss": 0.1827,
      "step": 4031
    },
    {
      "epoch": 2.7245691111862116,
      "grad_norm": 3.570859670639038,
      "learning_rate": 2e-05,
      "loss": 0.1512,
      "step": 4032
    },
    {
      "epoch": 2.7252450152078405,
      "grad_norm": 4.287653923034668,
      "learning_rate": 2e-05,
      "loss": 0.1324,
      "step": 4033
    },
    {
      "epoch": 2.7259209192294693,
      "grad_norm": 3.2347168922424316,
      "learning_rate": 2e-05,
      "loss": 0.1576,
      "step": 4034
    },
    {
      "epoch": 2.726596823251098,
      "grad_norm": 4.570337772369385,
      "learning_rate": 2e-05,
      "loss": 0.1286,
      "step": 4035
    },
    {
      "epoch": 2.7272727272727275,
      "grad_norm": 2.769334554672241,
      "learning_rate": 2e-05,
      "loss": 0.1335,
      "step": 4036
    },
    {
      "epoch": 2.727948631294356,
      "grad_norm": 12.837052345275879,
      "learning_rate": 2e-05,
      "loss": 0.1763,
      "step": 4037
    },
    {
      "epoch": 2.7286245353159853,
      "grad_norm": 6.957407474517822,
      "learning_rate": 2e-05,
      "loss": 0.1646,
      "step": 4038
    },
    {
      "epoch": 2.729300439337614,
      "grad_norm": 10.262145042419434,
      "learning_rate": 2e-05,
      "loss": 0.2555,
      "step": 4039
    },
    {
      "epoch": 2.729976343359243,
      "grad_norm": 9.020309448242188,
      "learning_rate": 2e-05,
      "loss": 0.2657,
      "step": 4040
    },
    {
      "epoch": 2.730652247380872,
      "grad_norm": 4.5144944190979,
      "learning_rate": 2e-05,
      "loss": 0.1571,
      "step": 4041
    },
    {
      "epoch": 2.7313281514025007,
      "grad_norm": 9.912672996520996,
      "learning_rate": 2e-05,
      "loss": 0.2057,
      "step": 4042
    },
    {
      "epoch": 2.7320040554241296,
      "grad_norm": 7.483288288116455,
      "learning_rate": 2e-05,
      "loss": 0.1757,
      "step": 4043
    },
    {
      "epoch": 2.7326799594457585,
      "grad_norm": 4.738915920257568,
      "learning_rate": 2e-05,
      "loss": 0.1569,
      "step": 4044
    },
    {
      "epoch": 2.733355863467388,
      "grad_norm": 5.940584182739258,
      "learning_rate": 2e-05,
      "loss": 0.1544,
      "step": 4045
    },
    {
      "epoch": 2.7340317674890167,
      "grad_norm": 2.5112557411193848,
      "learning_rate": 2e-05,
      "loss": 0.169,
      "step": 4046
    },
    {
      "epoch": 2.7347076715106455,
      "grad_norm": 3.11336088180542,
      "learning_rate": 2e-05,
      "loss": 0.1674,
      "step": 4047
    },
    {
      "epoch": 2.7353835755322744,
      "grad_norm": 7.079949855804443,
      "learning_rate": 2e-05,
      "loss": 0.2095,
      "step": 4048
    },
    {
      "epoch": 2.7360594795539033,
      "grad_norm": 4.270845890045166,
      "learning_rate": 2e-05,
      "loss": 0.1906,
      "step": 4049
    },
    {
      "epoch": 2.736735383575532,
      "grad_norm": 3.725804328918457,
      "learning_rate": 2e-05,
      "loss": 0.1376,
      "step": 4050
    },
    {
      "epoch": 2.736735383575532,
      "eval_accuracy": 0.6127956081081081,
      "eval_loss": 0.2327306866645813,
      "eval_runtime": 43.6102,
      "eval_samples_per_second": 542.992,
      "eval_steps_per_second": 16.969,
      "step": 4050
    },
    {
      "epoch": 2.737411287597161,
      "grad_norm": 2.590064525604248,
      "learning_rate": 2e-05,
      "loss": 0.1628,
      "step": 4051
    },
    {
      "epoch": 2.7380871916187903,
      "grad_norm": 6.1630659103393555,
      "learning_rate": 2e-05,
      "loss": 0.1764,
      "step": 4052
    },
    {
      "epoch": 2.738763095640419,
      "grad_norm": 5.174619197845459,
      "learning_rate": 2e-05,
      "loss": 0.1414,
      "step": 4053
    },
    {
      "epoch": 2.739438999662048,
      "grad_norm": 5.92640495300293,
      "learning_rate": 2e-05,
      "loss": 0.2127,
      "step": 4054
    },
    {
      "epoch": 2.740114903683677,
      "grad_norm": 4.364621639251709,
      "learning_rate": 2e-05,
      "loss": 0.1796,
      "step": 4055
    },
    {
      "epoch": 2.740790807705306,
      "grad_norm": 4.164730548858643,
      "learning_rate": 2e-05,
      "loss": 0.1404,
      "step": 4056
    },
    {
      "epoch": 2.7414667117269347,
      "grad_norm": 2.70204496383667,
      "learning_rate": 2e-05,
      "loss": 0.1365,
      "step": 4057
    },
    {
      "epoch": 2.7421426157485635,
      "grad_norm": 5.356203556060791,
      "learning_rate": 2e-05,
      "loss": 0.1894,
      "step": 4058
    },
    {
      "epoch": 2.742818519770193,
      "grad_norm": 5.2292585372924805,
      "learning_rate": 2e-05,
      "loss": 0.1856,
      "step": 4059
    },
    {
      "epoch": 2.7434944237918213,
      "grad_norm": 5.914241313934326,
      "learning_rate": 2e-05,
      "loss": 0.1522,
      "step": 4060
    },
    {
      "epoch": 2.7441703278134506,
      "grad_norm": 6.22280740737915,
      "learning_rate": 2e-05,
      "loss": 0.2033,
      "step": 4061
    },
    {
      "epoch": 2.7448462318350795,
      "grad_norm": 5.271011829376221,
      "learning_rate": 2e-05,
      "loss": 0.1563,
      "step": 4062
    },
    {
      "epoch": 2.7455221358567083,
      "grad_norm": 9.624288558959961,
      "learning_rate": 2e-05,
      "loss": 0.2032,
      "step": 4063
    },
    {
      "epoch": 2.746198039878337,
      "grad_norm": 2.6709728240966797,
      "learning_rate": 2e-05,
      "loss": 0.1591,
      "step": 4064
    },
    {
      "epoch": 2.746873943899966,
      "grad_norm": 10.65442943572998,
      "learning_rate": 2e-05,
      "loss": 0.1987,
      "step": 4065
    },
    {
      "epoch": 2.7475498479215954,
      "grad_norm": 5.799344539642334,
      "learning_rate": 2e-05,
      "loss": 0.1504,
      "step": 4066
    },
    {
      "epoch": 2.748225751943224,
      "grad_norm": 6.700428009033203,
      "learning_rate": 2e-05,
      "loss": 0.1575,
      "step": 4067
    },
    {
      "epoch": 2.748901655964853,
      "grad_norm": 5.727789878845215,
      "learning_rate": 2e-05,
      "loss": 0.1711,
      "step": 4068
    },
    {
      "epoch": 2.749577559986482,
      "grad_norm": 6.426336765289307,
      "learning_rate": 2e-05,
      "loss": 0.1534,
      "step": 4069
    },
    {
      "epoch": 2.750253464008111,
      "grad_norm": 5.150265693664551,
      "learning_rate": 2e-05,
      "loss": 0.2706,
      "step": 4070
    },
    {
      "epoch": 2.7509293680297398,
      "grad_norm": 4.691850185394287,
      "learning_rate": 2e-05,
      "loss": 0.1707,
      "step": 4071
    },
    {
      "epoch": 2.7516052720513686,
      "grad_norm": 4.839570999145508,
      "learning_rate": 2e-05,
      "loss": 0.2195,
      "step": 4072
    },
    {
      "epoch": 2.7522811760729975,
      "grad_norm": 3.6882617473602295,
      "learning_rate": 2e-05,
      "loss": 0.1826,
      "step": 4073
    },
    {
      "epoch": 2.7529570800946264,
      "grad_norm": 3.987391710281372,
      "learning_rate": 2e-05,
      "loss": 0.1446,
      "step": 4074
    },
    {
      "epoch": 2.7536329841162557,
      "grad_norm": 3.964045286178589,
      "learning_rate": 2e-05,
      "loss": 0.1708,
      "step": 4075
    },
    {
      "epoch": 2.7543088881378845,
      "grad_norm": 5.470864772796631,
      "learning_rate": 2e-05,
      "loss": 0.2262,
      "step": 4076
    },
    {
      "epoch": 2.7549847921595134,
      "grad_norm": 3.217360496520996,
      "learning_rate": 2e-05,
      "loss": 0.1595,
      "step": 4077
    },
    {
      "epoch": 2.7556606961811423,
      "grad_norm": 3.5405492782592773,
      "learning_rate": 2e-05,
      "loss": 0.1248,
      "step": 4078
    },
    {
      "epoch": 2.756336600202771,
      "grad_norm": 3.040837049484253,
      "learning_rate": 2e-05,
      "loss": 0.1946,
      "step": 4079
    },
    {
      "epoch": 2.7570125042244,
      "grad_norm": 4.236926555633545,
      "learning_rate": 2e-05,
      "loss": 0.184,
      "step": 4080
    },
    {
      "epoch": 2.757688408246029,
      "grad_norm": 5.213834762573242,
      "learning_rate": 2e-05,
      "loss": 0.1511,
      "step": 4081
    },
    {
      "epoch": 2.758364312267658,
      "grad_norm": 3.8601720333099365,
      "learning_rate": 2e-05,
      "loss": 0.1765,
      "step": 4082
    },
    {
      "epoch": 2.759040216289287,
      "grad_norm": 3.0374410152435303,
      "learning_rate": 2e-05,
      "loss": 0.1701,
      "step": 4083
    },
    {
      "epoch": 2.759716120310916,
      "grad_norm": 3.6581521034240723,
      "learning_rate": 2e-05,
      "loss": 0.188,
      "step": 4084
    },
    {
      "epoch": 2.760392024332545,
      "grad_norm": 4.224642753601074,
      "learning_rate": 2e-05,
      "loss": 0.187,
      "step": 4085
    },
    {
      "epoch": 2.7610679283541737,
      "grad_norm": 4.065493106842041,
      "learning_rate": 2e-05,
      "loss": 0.1599,
      "step": 4086
    },
    {
      "epoch": 2.7617438323758026,
      "grad_norm": 2.689380407333374,
      "learning_rate": 2e-05,
      "loss": 0.149,
      "step": 4087
    },
    {
      "epoch": 2.7624197363974314,
      "grad_norm": 5.382197856903076,
      "learning_rate": 2e-05,
      "loss": 0.1792,
      "step": 4088
    },
    {
      "epoch": 2.7630956404190608,
      "grad_norm": 3.793755054473877,
      "learning_rate": 2e-05,
      "loss": 0.188,
      "step": 4089
    },
    {
      "epoch": 2.763771544440689,
      "grad_norm": 4.308106899261475,
      "learning_rate": 2e-05,
      "loss": 0.1118,
      "step": 4090
    },
    {
      "epoch": 2.7644474484623185,
      "grad_norm": 4.699409484863281,
      "learning_rate": 2e-05,
      "loss": 0.1618,
      "step": 4091
    },
    {
      "epoch": 2.7651233524839474,
      "grad_norm": 6.520378589630127,
      "learning_rate": 2e-05,
      "loss": 0.2291,
      "step": 4092
    },
    {
      "epoch": 2.7657992565055762,
      "grad_norm": 9.02273941040039,
      "learning_rate": 2e-05,
      "loss": 0.176,
      "step": 4093
    },
    {
      "epoch": 2.766475160527205,
      "grad_norm": 2.8938279151916504,
      "learning_rate": 2e-05,
      "loss": 0.174,
      "step": 4094
    },
    {
      "epoch": 2.767151064548834,
      "grad_norm": 3.2431693077087402,
      "learning_rate": 2e-05,
      "loss": 0.1451,
      "step": 4095
    },
    {
      "epoch": 2.767826968570463,
      "grad_norm": 3.614827871322632,
      "learning_rate": 2e-05,
      "loss": 0.1734,
      "step": 4096
    },
    {
      "epoch": 2.7685028725920917,
      "grad_norm": 6.75798225402832,
      "learning_rate": 2e-05,
      "loss": 0.1843,
      "step": 4097
    },
    {
      "epoch": 2.769178776613721,
      "grad_norm": 5.305551528930664,
      "learning_rate": 2e-05,
      "loss": 0.1488,
      "step": 4098
    },
    {
      "epoch": 2.76985468063535,
      "grad_norm": 3.8440942764282227,
      "learning_rate": 2e-05,
      "loss": 0.2175,
      "step": 4099
    },
    {
      "epoch": 2.7705305846569788,
      "grad_norm": 7.498745441436768,
      "learning_rate": 2e-05,
      "loss": 0.1998,
      "step": 4100
    },
    {
      "epoch": 2.7705305846569788,
      "eval_accuracy": 0.605785472972973,
      "eval_loss": 0.23780125379562378,
      "eval_runtime": 43.0381,
      "eval_samples_per_second": 550.21,
      "eval_steps_per_second": 17.194,
      "step": 4100
    },
    {
      "epoch": 2.7712064886786076,
      "grad_norm": 2.821658134460449,
      "learning_rate": 2e-05,
      "loss": 0.1783,
      "step": 4101
    },
    {
      "epoch": 2.7718823927002365,
      "grad_norm": 9.020559310913086,
      "learning_rate": 2e-05,
      "loss": 0.1541,
      "step": 4102
    },
    {
      "epoch": 2.7725582967218654,
      "grad_norm": 6.930007457733154,
      "learning_rate": 2e-05,
      "loss": 0.1792,
      "step": 4103
    },
    {
      "epoch": 2.7732342007434942,
      "grad_norm": 4.86575174331665,
      "learning_rate": 2e-05,
      "loss": 0.1484,
      "step": 4104
    },
    {
      "epoch": 2.7739101047651236,
      "grad_norm": 4.709434986114502,
      "learning_rate": 2e-05,
      "loss": 0.2069,
      "step": 4105
    },
    {
      "epoch": 2.7745860087867524,
      "grad_norm": 5.053496360778809,
      "learning_rate": 2e-05,
      "loss": 0.1588,
      "step": 4106
    },
    {
      "epoch": 2.7752619128083813,
      "grad_norm": 2.620030403137207,
      "learning_rate": 2e-05,
      "loss": 0.1547,
      "step": 4107
    },
    {
      "epoch": 2.77593781683001,
      "grad_norm": 3.1866023540496826,
      "learning_rate": 2e-05,
      "loss": 0.1185,
      "step": 4108
    },
    {
      "epoch": 2.776613720851639,
      "grad_norm": 2.8862698078155518,
      "learning_rate": 2e-05,
      "loss": 0.1463,
      "step": 4109
    },
    {
      "epoch": 2.777289624873268,
      "grad_norm": 4.912588596343994,
      "learning_rate": 2e-05,
      "loss": 0.1726,
      "step": 4110
    },
    {
      "epoch": 2.777965528894897,
      "grad_norm": 7.861147403717041,
      "learning_rate": 2e-05,
      "loss": 0.1547,
      "step": 4111
    },
    {
      "epoch": 2.778641432916526,
      "grad_norm": 3.405280828475952,
      "learning_rate": 2e-05,
      "loss": 0.175,
      "step": 4112
    },
    {
      "epoch": 2.7793173369381545,
      "grad_norm": 3.5840816497802734,
      "learning_rate": 2e-05,
      "loss": 0.1304,
      "step": 4113
    },
    {
      "epoch": 2.779993240959784,
      "grad_norm": 3.0004944801330566,
      "learning_rate": 2e-05,
      "loss": 0.2028,
      "step": 4114
    },
    {
      "epoch": 2.7806691449814127,
      "grad_norm": 4.404737949371338,
      "learning_rate": 2e-05,
      "loss": 0.2497,
      "step": 4115
    },
    {
      "epoch": 2.7813450490030416,
      "grad_norm": 5.400577545166016,
      "learning_rate": 2e-05,
      "loss": 0.1488,
      "step": 4116
    },
    {
      "epoch": 2.7820209530246705,
      "grad_norm": 6.558976650238037,
      "learning_rate": 2e-05,
      "loss": 0.22,
      "step": 4117
    },
    {
      "epoch": 2.7826968570462993,
      "grad_norm": 5.519373893737793,
      "learning_rate": 2e-05,
      "loss": 0.1941,
      "step": 4118
    },
    {
      "epoch": 2.783372761067928,
      "grad_norm": 3.609247922897339,
      "learning_rate": 2e-05,
      "loss": 0.1907,
      "step": 4119
    },
    {
      "epoch": 2.784048665089557,
      "grad_norm": 2.4847848415374756,
      "learning_rate": 2e-05,
      "loss": 0.117,
      "step": 4120
    },
    {
      "epoch": 2.7847245691111864,
      "grad_norm": 3.270517349243164,
      "learning_rate": 2e-05,
      "loss": 0.1561,
      "step": 4121
    },
    {
      "epoch": 2.7854004731328152,
      "grad_norm": 4.532674312591553,
      "learning_rate": 2e-05,
      "loss": 0.2203,
      "step": 4122
    },
    {
      "epoch": 2.786076377154444,
      "grad_norm": 4.494446277618408,
      "learning_rate": 2e-05,
      "loss": 0.1725,
      "step": 4123
    },
    {
      "epoch": 2.786752281176073,
      "grad_norm": 4.004085063934326,
      "learning_rate": 2e-05,
      "loss": 0.1997,
      "step": 4124
    },
    {
      "epoch": 2.787428185197702,
      "grad_norm": 3.9079482555389404,
      "learning_rate": 2e-05,
      "loss": 0.1114,
      "step": 4125
    },
    {
      "epoch": 2.7881040892193307,
      "grad_norm": 6.391659259796143,
      "learning_rate": 2e-05,
      "loss": 0.118,
      "step": 4126
    },
    {
      "epoch": 2.7887799932409596,
      "grad_norm": 3.7285821437835693,
      "learning_rate": 2e-05,
      "loss": 0.1911,
      "step": 4127
    },
    {
      "epoch": 2.789455897262589,
      "grad_norm": 3.0064074993133545,
      "learning_rate": 2e-05,
      "loss": 0.1757,
      "step": 4128
    },
    {
      "epoch": 2.790131801284218,
      "grad_norm": 4.57772970199585,
      "learning_rate": 2e-05,
      "loss": 0.2292,
      "step": 4129
    },
    {
      "epoch": 2.7908077053058467,
      "grad_norm": 3.7008612155914307,
      "learning_rate": 2e-05,
      "loss": 0.1555,
      "step": 4130
    },
    {
      "epoch": 2.7914836093274755,
      "grad_norm": 3.43204665184021,
      "learning_rate": 2e-05,
      "loss": 0.1488,
      "step": 4131
    },
    {
      "epoch": 2.7921595133491044,
      "grad_norm": 6.406134605407715,
      "learning_rate": 2e-05,
      "loss": 0.1851,
      "step": 4132
    },
    {
      "epoch": 2.7928354173707333,
      "grad_norm": 3.2738559246063232,
      "learning_rate": 2e-05,
      "loss": 0.1768,
      "step": 4133
    },
    {
      "epoch": 2.793511321392362,
      "grad_norm": 5.004574775695801,
      "learning_rate": 2e-05,
      "loss": 0.1598,
      "step": 4134
    },
    {
      "epoch": 2.7941872254139914,
      "grad_norm": 3.4103620052337646,
      "learning_rate": 2e-05,
      "loss": 0.1754,
      "step": 4135
    },
    {
      "epoch": 2.79486312943562,
      "grad_norm": 4.574471473693848,
      "learning_rate": 2e-05,
      "loss": 0.1933,
      "step": 4136
    },
    {
      "epoch": 2.795539033457249,
      "grad_norm": 5.432750225067139,
      "learning_rate": 2e-05,
      "loss": 0.1665,
      "step": 4137
    },
    {
      "epoch": 2.796214937478878,
      "grad_norm": 3.944547653198242,
      "learning_rate": 2e-05,
      "loss": 0.168,
      "step": 4138
    },
    {
      "epoch": 2.796890841500507,
      "grad_norm": 3.356376886367798,
      "learning_rate": 2e-05,
      "loss": 0.1782,
      "step": 4139
    },
    {
      "epoch": 2.797566745522136,
      "grad_norm": 4.4019341468811035,
      "learning_rate": 2e-05,
      "loss": 0.1544,
      "step": 4140
    },
    {
      "epoch": 2.7982426495437647,
      "grad_norm": 5.624101638793945,
      "learning_rate": 2e-05,
      "loss": 0.2312,
      "step": 4141
    },
    {
      "epoch": 2.798918553565394,
      "grad_norm": 4.833008289337158,
      "learning_rate": 2e-05,
      "loss": 0.1763,
      "step": 4142
    },
    {
      "epoch": 2.7995944575870224,
      "grad_norm": 10.78359603881836,
      "learning_rate": 2e-05,
      "loss": 0.1914,
      "step": 4143
    },
    {
      "epoch": 2.8002703616086517,
      "grad_norm": 2.60788631439209,
      "learning_rate": 2e-05,
      "loss": 0.1424,
      "step": 4144
    },
    {
      "epoch": 2.8009462656302806,
      "grad_norm": 8.40326976776123,
      "learning_rate": 2e-05,
      "loss": 0.1846,
      "step": 4145
    },
    {
      "epoch": 2.8016221696519095,
      "grad_norm": 9.79725170135498,
      "learning_rate": 2e-05,
      "loss": 0.1728,
      "step": 4146
    },
    {
      "epoch": 2.8022980736735383,
      "grad_norm": 3.184913396835327,
      "learning_rate": 2e-05,
      "loss": 0.1754,
      "step": 4147
    },
    {
      "epoch": 2.802973977695167,
      "grad_norm": 12.429795265197754,
      "learning_rate": 2e-05,
      "loss": 0.1715,
      "step": 4148
    },
    {
      "epoch": 2.803649881716796,
      "grad_norm": 5.0875468254089355,
      "learning_rate": 2e-05,
      "loss": 0.1459,
      "step": 4149
    },
    {
      "epoch": 2.804325785738425,
      "grad_norm": 3.073892593383789,
      "learning_rate": 2e-05,
      "loss": 0.1845,
      "step": 4150
    },
    {
      "epoch": 2.804325785738425,
      "eval_accuracy": 0.6028716216216217,
      "eval_loss": 0.23610176146030426,
      "eval_runtime": 44.1191,
      "eval_samples_per_second": 536.729,
      "eval_steps_per_second": 16.773,
      "step": 4150
    },
    {
      "epoch": 2.8050016897600543,
      "grad_norm": 8.727582931518555,
      "learning_rate": 2e-05,
      "loss": 0.2419,
      "step": 4151
    },
    {
      "epoch": 2.805677593781683,
      "grad_norm": 5.383760452270508,
      "learning_rate": 2e-05,
      "loss": 0.199,
      "step": 4152
    },
    {
      "epoch": 2.806353497803312,
      "grad_norm": 2.654463291168213,
      "learning_rate": 2e-05,
      "loss": 0.1679,
      "step": 4153
    },
    {
      "epoch": 2.807029401824941,
      "grad_norm": 3.9914345741271973,
      "learning_rate": 2e-05,
      "loss": 0.2178,
      "step": 4154
    },
    {
      "epoch": 2.8077053058465697,
      "grad_norm": 8.280055046081543,
      "learning_rate": 2e-05,
      "loss": 0.114,
      "step": 4155
    },
    {
      "epoch": 2.8083812098681986,
      "grad_norm": 9.76237678527832,
      "learning_rate": 2e-05,
      "loss": 0.1734,
      "step": 4156
    },
    {
      "epoch": 2.8090571138898275,
      "grad_norm": 6.783533573150635,
      "learning_rate": 2e-05,
      "loss": 0.1633,
      "step": 4157
    },
    {
      "epoch": 2.809733017911457,
      "grad_norm": 3.404541015625,
      "learning_rate": 2e-05,
      "loss": 0.1423,
      "step": 4158
    },
    {
      "epoch": 2.8104089219330852,
      "grad_norm": 3.964338541030884,
      "learning_rate": 2e-05,
      "loss": 0.1167,
      "step": 4159
    },
    {
      "epoch": 2.8110848259547145,
      "grad_norm": 7.751773834228516,
      "learning_rate": 2e-05,
      "loss": 0.1896,
      "step": 4160
    },
    {
      "epoch": 2.8117607299763434,
      "grad_norm": 5.675203323364258,
      "learning_rate": 2e-05,
      "loss": 0.159,
      "step": 4161
    },
    {
      "epoch": 2.8124366339979723,
      "grad_norm": 6.808338165283203,
      "learning_rate": 2e-05,
      "loss": 0.1607,
      "step": 4162
    },
    {
      "epoch": 2.813112538019601,
      "grad_norm": 3.2285046577453613,
      "learning_rate": 2e-05,
      "loss": 0.1224,
      "step": 4163
    },
    {
      "epoch": 2.81378844204123,
      "grad_norm": 4.347068786621094,
      "learning_rate": 2e-05,
      "loss": 0.1574,
      "step": 4164
    },
    {
      "epoch": 2.8144643460628593,
      "grad_norm": 4.295374870300293,
      "learning_rate": 2e-05,
      "loss": 0.1515,
      "step": 4165
    },
    {
      "epoch": 2.8151402500844878,
      "grad_norm": 11.831228256225586,
      "learning_rate": 2e-05,
      "loss": 0.2269,
      "step": 4166
    },
    {
      "epoch": 2.815816154106117,
      "grad_norm": 3.063218355178833,
      "learning_rate": 2e-05,
      "loss": 0.1565,
      "step": 4167
    },
    {
      "epoch": 2.816492058127746,
      "grad_norm": 4.9776201248168945,
      "learning_rate": 2e-05,
      "loss": 0.1132,
      "step": 4168
    },
    {
      "epoch": 2.817167962149375,
      "grad_norm": 6.85977029800415,
      "learning_rate": 2e-05,
      "loss": 0.1723,
      "step": 4169
    },
    {
      "epoch": 2.8178438661710037,
      "grad_norm": 3.530789852142334,
      "learning_rate": 2e-05,
      "loss": 0.1677,
      "step": 4170
    },
    {
      "epoch": 2.8185197701926326,
      "grad_norm": 4.916831016540527,
      "learning_rate": 2e-05,
      "loss": 0.188,
      "step": 4171
    },
    {
      "epoch": 2.8191956742142614,
      "grad_norm": 5.6775665283203125,
      "learning_rate": 2e-05,
      "loss": 0.1639,
      "step": 4172
    },
    {
      "epoch": 2.8198715782358903,
      "grad_norm": 6.180357456207275,
      "learning_rate": 2e-05,
      "loss": 0.2056,
      "step": 4173
    },
    {
      "epoch": 2.8205474822575196,
      "grad_norm": 4.673091888427734,
      "learning_rate": 2e-05,
      "loss": 0.2152,
      "step": 4174
    },
    {
      "epoch": 2.8212233862791485,
      "grad_norm": 5.5834479331970215,
      "learning_rate": 2e-05,
      "loss": 0.1492,
      "step": 4175
    },
    {
      "epoch": 2.8218992903007774,
      "grad_norm": 5.108085632324219,
      "learning_rate": 2e-05,
      "loss": 0.1901,
      "step": 4176
    },
    {
      "epoch": 2.8225751943224062,
      "grad_norm": 3.5377113819122314,
      "learning_rate": 2e-05,
      "loss": 0.1874,
      "step": 4177
    },
    {
      "epoch": 2.823251098344035,
      "grad_norm": 4.196198463439941,
      "learning_rate": 2e-05,
      "loss": 0.1728,
      "step": 4178
    },
    {
      "epoch": 2.823927002365664,
      "grad_norm": 4.2683234214782715,
      "learning_rate": 2e-05,
      "loss": 0.1437,
      "step": 4179
    },
    {
      "epoch": 2.824602906387293,
      "grad_norm": 8.633320808410645,
      "learning_rate": 2e-05,
      "loss": 0.204,
      "step": 4180
    },
    {
      "epoch": 2.825278810408922,
      "grad_norm": 3.8583269119262695,
      "learning_rate": 2e-05,
      "loss": 0.1227,
      "step": 4181
    },
    {
      "epoch": 2.825954714430551,
      "grad_norm": 6.54818058013916,
      "learning_rate": 2e-05,
      "loss": 0.163,
      "step": 4182
    },
    {
      "epoch": 2.82663061845218,
      "grad_norm": 4.326067924499512,
      "learning_rate": 2e-05,
      "loss": 0.152,
      "step": 4183
    },
    {
      "epoch": 2.8273065224738088,
      "grad_norm": 4.343260765075684,
      "learning_rate": 2e-05,
      "loss": 0.19,
      "step": 4184
    },
    {
      "epoch": 2.8279824264954376,
      "grad_norm": 7.505638122558594,
      "learning_rate": 2e-05,
      "loss": 0.18,
      "step": 4185
    },
    {
      "epoch": 2.8286583305170665,
      "grad_norm": 4.1911444664001465,
      "learning_rate": 2e-05,
      "loss": 0.1708,
      "step": 4186
    },
    {
      "epoch": 2.8293342345386954,
      "grad_norm": 4.845101833343506,
      "learning_rate": 2e-05,
      "loss": 0.1264,
      "step": 4187
    },
    {
      "epoch": 2.8300101385603247,
      "grad_norm": 4.806951522827148,
      "learning_rate": 2e-05,
      "loss": 0.1914,
      "step": 4188
    },
    {
      "epoch": 2.830686042581953,
      "grad_norm": 3.359830856323242,
      "learning_rate": 2e-05,
      "loss": 0.1578,
      "step": 4189
    },
    {
      "epoch": 2.8313619466035824,
      "grad_norm": 4.707402229309082,
      "learning_rate": 2e-05,
      "loss": 0.1745,
      "step": 4190
    },
    {
      "epoch": 2.8320378506252113,
      "grad_norm": 3.927990436553955,
      "learning_rate": 2e-05,
      "loss": 0.1961,
      "step": 4191
    },
    {
      "epoch": 2.83271375464684,
      "grad_norm": 3.9141829013824463,
      "learning_rate": 2e-05,
      "loss": 0.1714,
      "step": 4192
    },
    {
      "epoch": 2.833389658668469,
      "grad_norm": 7.530604362487793,
      "learning_rate": 2e-05,
      "loss": 0.1724,
      "step": 4193
    },
    {
      "epoch": 2.834065562690098,
      "grad_norm": 5.017278671264648,
      "learning_rate": 2e-05,
      "loss": 0.2296,
      "step": 4194
    },
    {
      "epoch": 2.8347414667117268,
      "grad_norm": 4.928203105926514,
      "learning_rate": 2e-05,
      "loss": 0.1915,
      "step": 4195
    },
    {
      "epoch": 2.8354173707333556,
      "grad_norm": 6.741231441497803,
      "learning_rate": 2e-05,
      "loss": 0.123,
      "step": 4196
    },
    {
      "epoch": 2.836093274754985,
      "grad_norm": 5.402796745300293,
      "learning_rate": 2e-05,
      "loss": 0.1762,
      "step": 4197
    },
    {
      "epoch": 2.836769178776614,
      "grad_norm": 7.062037467956543,
      "learning_rate": 2e-05,
      "loss": 0.1345,
      "step": 4198
    },
    {
      "epoch": 2.8374450827982427,
      "grad_norm": 6.407635688781738,
      "learning_rate": 2e-05,
      "loss": 0.1869,
      "step": 4199
    },
    {
      "epoch": 2.8381209868198716,
      "grad_norm": 4.8336615562438965,
      "learning_rate": 2e-05,
      "loss": 0.1804,
      "step": 4200
    },
    {
      "epoch": 2.8381209868198716,
      "eval_accuracy": 0.6253378378378378,
      "eval_loss": 0.23828846216201782,
      "eval_runtime": 43.1656,
      "eval_samples_per_second": 548.584,
      "eval_steps_per_second": 17.143,
      "step": 4200
    },
    {
      "epoch": 2.8387968908415004,
      "grad_norm": 8.502623558044434,
      "learning_rate": 2e-05,
      "loss": 0.1487,
      "step": 4201
    },
    {
      "epoch": 2.8394727948631293,
      "grad_norm": 4.061350345611572,
      "learning_rate": 2e-05,
      "loss": 0.1321,
      "step": 4202
    },
    {
      "epoch": 2.840148698884758,
      "grad_norm": 7.057444095611572,
      "learning_rate": 2e-05,
      "loss": 0.2159,
      "step": 4203
    },
    {
      "epoch": 2.8408246029063875,
      "grad_norm": 4.489827632904053,
      "learning_rate": 2e-05,
      "loss": 0.2229,
      "step": 4204
    },
    {
      "epoch": 2.8415005069280164,
      "grad_norm": 6.209469318389893,
      "learning_rate": 2e-05,
      "loss": 0.2033,
      "step": 4205
    },
    {
      "epoch": 2.8421764109496452,
      "grad_norm": 10.51617431640625,
      "learning_rate": 2e-05,
      "loss": 0.1982,
      "step": 4206
    },
    {
      "epoch": 2.842852314971274,
      "grad_norm": 4.992805004119873,
      "learning_rate": 2e-05,
      "loss": 0.1847,
      "step": 4207
    },
    {
      "epoch": 2.843528218992903,
      "grad_norm": 11.178023338317871,
      "learning_rate": 2e-05,
      "loss": 0.1476,
      "step": 4208
    },
    {
      "epoch": 2.844204123014532,
      "grad_norm": 2.8976593017578125,
      "learning_rate": 2e-05,
      "loss": 0.1975,
      "step": 4209
    },
    {
      "epoch": 2.8448800270361607,
      "grad_norm": 5.180940628051758,
      "learning_rate": 2e-05,
      "loss": 0.1457,
      "step": 4210
    },
    {
      "epoch": 2.84555593105779,
      "grad_norm": 2.740811347961426,
      "learning_rate": 2e-05,
      "loss": 0.1364,
      "step": 4211
    },
    {
      "epoch": 2.8462318350794185,
      "grad_norm": 3.8852739334106445,
      "learning_rate": 2e-05,
      "loss": 0.1244,
      "step": 4212
    },
    {
      "epoch": 2.8469077391010478,
      "grad_norm": 8.82465934753418,
      "learning_rate": 2e-05,
      "loss": 0.2348,
      "step": 4213
    },
    {
      "epoch": 2.8475836431226766,
      "grad_norm": 3.807239294052124,
      "learning_rate": 2e-05,
      "loss": 0.1397,
      "step": 4214
    },
    {
      "epoch": 2.8482595471443055,
      "grad_norm": 6.492600440979004,
      "learning_rate": 2e-05,
      "loss": 0.1875,
      "step": 4215
    },
    {
      "epoch": 2.8489354511659344,
      "grad_norm": 6.5023627281188965,
      "learning_rate": 2e-05,
      "loss": 0.2008,
      "step": 4216
    },
    {
      "epoch": 2.8496113551875633,
      "grad_norm": 4.439709186553955,
      "learning_rate": 2e-05,
      "loss": 0.1734,
      "step": 4217
    },
    {
      "epoch": 2.8502872592091926,
      "grad_norm": 3.772366762161255,
      "learning_rate": 2e-05,
      "loss": 0.1848,
      "step": 4218
    },
    {
      "epoch": 2.850963163230821,
      "grad_norm": 3.325153112411499,
      "learning_rate": 2e-05,
      "loss": 0.1433,
      "step": 4219
    },
    {
      "epoch": 2.8516390672524503,
      "grad_norm": 2.70759916305542,
      "learning_rate": 2e-05,
      "loss": 0.1186,
      "step": 4220
    },
    {
      "epoch": 2.852314971274079,
      "grad_norm": 6.9762420654296875,
      "learning_rate": 2e-05,
      "loss": 0.2416,
      "step": 4221
    },
    {
      "epoch": 2.852990875295708,
      "grad_norm": 7.8046956062316895,
      "learning_rate": 2e-05,
      "loss": 0.2496,
      "step": 4222
    },
    {
      "epoch": 2.853666779317337,
      "grad_norm": 2.888289213180542,
      "learning_rate": 2e-05,
      "loss": 0.1855,
      "step": 4223
    },
    {
      "epoch": 2.854342683338966,
      "grad_norm": 3.9692299365997314,
      "learning_rate": 2e-05,
      "loss": 0.1469,
      "step": 4224
    },
    {
      "epoch": 2.8550185873605947,
      "grad_norm": 4.275857925415039,
      "learning_rate": 2e-05,
      "loss": 0.197,
      "step": 4225
    },
    {
      "epoch": 2.8556944913822235,
      "grad_norm": 7.840794563293457,
      "learning_rate": 2e-05,
      "loss": 0.1458,
      "step": 4226
    },
    {
      "epoch": 2.856370395403853,
      "grad_norm": 2.904775381088257,
      "learning_rate": 2e-05,
      "loss": 0.1572,
      "step": 4227
    },
    {
      "epoch": 2.8570462994254817,
      "grad_norm": 5.1092658042907715,
      "learning_rate": 2e-05,
      "loss": 0.2481,
      "step": 4228
    },
    {
      "epoch": 2.8577222034471106,
      "grad_norm": 4.31952428817749,
      "learning_rate": 2e-05,
      "loss": 0.1606,
      "step": 4229
    },
    {
      "epoch": 2.8583981074687395,
      "grad_norm": 3.2719435691833496,
      "learning_rate": 2e-05,
      "loss": 0.1429,
      "step": 4230
    },
    {
      "epoch": 2.8590740114903683,
      "grad_norm": 2.719299554824829,
      "learning_rate": 2e-05,
      "loss": 0.1348,
      "step": 4231
    },
    {
      "epoch": 2.859749915511997,
      "grad_norm": 9.789551734924316,
      "learning_rate": 2e-05,
      "loss": 0.1675,
      "step": 4232
    },
    {
      "epoch": 2.860425819533626,
      "grad_norm": 6.339001655578613,
      "learning_rate": 2e-05,
      "loss": 0.1697,
      "step": 4233
    },
    {
      "epoch": 2.8611017235552554,
      "grad_norm": 2.435515880584717,
      "learning_rate": 2e-05,
      "loss": 0.1314,
      "step": 4234
    },
    {
      "epoch": 2.861777627576884,
      "grad_norm": 3.967608690261841,
      "learning_rate": 2e-05,
      "loss": 0.1903,
      "step": 4235
    },
    {
      "epoch": 2.862453531598513,
      "grad_norm": 9.44953441619873,
      "learning_rate": 2e-05,
      "loss": 0.1685,
      "step": 4236
    },
    {
      "epoch": 2.863129435620142,
      "grad_norm": 4.5116963386535645,
      "learning_rate": 2e-05,
      "loss": 0.1429,
      "step": 4237
    },
    {
      "epoch": 2.863805339641771,
      "grad_norm": 13.456419944763184,
      "learning_rate": 2e-05,
      "loss": 0.2035,
      "step": 4238
    },
    {
      "epoch": 2.8644812436633997,
      "grad_norm": 4.205825328826904,
      "learning_rate": 2e-05,
      "loss": 0.198,
      "step": 4239
    },
    {
      "epoch": 2.8651571476850286,
      "grad_norm": 4.674407958984375,
      "learning_rate": 2e-05,
      "loss": 0.1747,
      "step": 4240
    },
    {
      "epoch": 2.865833051706658,
      "grad_norm": 3.7169876098632812,
      "learning_rate": 2e-05,
      "loss": 0.1315,
      "step": 4241
    },
    {
      "epoch": 2.8665089557282863,
      "grad_norm": 6.039844036102295,
      "learning_rate": 2e-05,
      "loss": 0.2051,
      "step": 4242
    },
    {
      "epoch": 2.8671848597499157,
      "grad_norm": 4.344793319702148,
      "learning_rate": 2e-05,
      "loss": 0.1405,
      "step": 4243
    },
    {
      "epoch": 2.8678607637715445,
      "grad_norm": 3.3234548568725586,
      "learning_rate": 2e-05,
      "loss": 0.2186,
      "step": 4244
    },
    {
      "epoch": 2.8685366677931734,
      "grad_norm": 7.757950782775879,
      "learning_rate": 2e-05,
      "loss": 0.2119,
      "step": 4245
    },
    {
      "epoch": 2.8692125718148023,
      "grad_norm": 2.8578717708587646,
      "learning_rate": 2e-05,
      "loss": 0.1507,
      "step": 4246
    },
    {
      "epoch": 2.869888475836431,
      "grad_norm": 3.2482078075408936,
      "learning_rate": 2e-05,
      "loss": 0.1653,
      "step": 4247
    },
    {
      "epoch": 2.87056437985806,
      "grad_norm": 6.805845260620117,
      "learning_rate": 2e-05,
      "loss": 0.1617,
      "step": 4248
    },
    {
      "epoch": 2.871240283879689,
      "grad_norm": 4.262026309967041,
      "learning_rate": 2e-05,
      "loss": 0.2216,
      "step": 4249
    },
    {
      "epoch": 2.871916187901318,
      "grad_norm": 3.2105600833892822,
      "learning_rate": 2e-05,
      "loss": 0.1881,
      "step": 4250
    },
    {
      "epoch": 2.871916187901318,
      "eval_accuracy": 0.6022804054054054,
      "eval_loss": 0.228218674659729,
      "eval_runtime": 43.5918,
      "eval_samples_per_second": 543.222,
      "eval_steps_per_second": 16.976,
      "step": 4250
    },
    {
      "epoch": 2.872592091922947,
      "grad_norm": 2.7526845932006836,
      "learning_rate": 2e-05,
      "loss": 0.19,
      "step": 4251
    },
    {
      "epoch": 2.873267995944576,
      "grad_norm": 3.8498353958129883,
      "learning_rate": 2e-05,
      "loss": 0.1527,
      "step": 4252
    },
    {
      "epoch": 2.873943899966205,
      "grad_norm": 6.406511306762695,
      "learning_rate": 2e-05,
      "loss": 0.17,
      "step": 4253
    },
    {
      "epoch": 2.8746198039878337,
      "grad_norm": 7.366283416748047,
      "learning_rate": 2e-05,
      "loss": 0.1498,
      "step": 4254
    },
    {
      "epoch": 2.8752957080094625,
      "grad_norm": 5.42094087600708,
      "learning_rate": 2e-05,
      "loss": 0.1472,
      "step": 4255
    },
    {
      "epoch": 2.8759716120310914,
      "grad_norm": 6.925508499145508,
      "learning_rate": 2e-05,
      "loss": 0.2252,
      "step": 4256
    },
    {
      "epoch": 2.8766475160527207,
      "grad_norm": 5.87070369720459,
      "learning_rate": 2e-05,
      "loss": 0.196,
      "step": 4257
    },
    {
      "epoch": 2.8773234200743496,
      "grad_norm": 5.571274280548096,
      "learning_rate": 2e-05,
      "loss": 0.1612,
      "step": 4258
    },
    {
      "epoch": 2.8779993240959785,
      "grad_norm": 2.9669992923736572,
      "learning_rate": 2e-05,
      "loss": 0.1571,
      "step": 4259
    },
    {
      "epoch": 2.8786752281176073,
      "grad_norm": 4.918482303619385,
      "learning_rate": 2e-05,
      "loss": 0.2054,
      "step": 4260
    },
    {
      "epoch": 2.879351132139236,
      "grad_norm": 9.723369598388672,
      "learning_rate": 2e-05,
      "loss": 0.2144,
      "step": 4261
    },
    {
      "epoch": 2.880027036160865,
      "grad_norm": 7.514566421508789,
      "learning_rate": 2e-05,
      "loss": 0.1888,
      "step": 4262
    },
    {
      "epoch": 2.880702940182494,
      "grad_norm": 8.816705703735352,
      "learning_rate": 2e-05,
      "loss": 0.1727,
      "step": 4263
    },
    {
      "epoch": 2.8813788442041233,
      "grad_norm": 8.625283241271973,
      "learning_rate": 2e-05,
      "loss": 0.1874,
      "step": 4264
    },
    {
      "epoch": 2.8820547482257517,
      "grad_norm": 3.4474539756774902,
      "learning_rate": 2e-05,
      "loss": 0.2436,
      "step": 4265
    },
    {
      "epoch": 2.882730652247381,
      "grad_norm": 3.8445472717285156,
      "learning_rate": 2e-05,
      "loss": 0.1885,
      "step": 4266
    },
    {
      "epoch": 2.88340655626901,
      "grad_norm": 10.655577659606934,
      "learning_rate": 2e-05,
      "loss": 0.1903,
      "step": 4267
    },
    {
      "epoch": 2.8840824602906387,
      "grad_norm": 8.88461971282959,
      "learning_rate": 2e-05,
      "loss": 0.1491,
      "step": 4268
    },
    {
      "epoch": 2.8847583643122676,
      "grad_norm": 4.226637840270996,
      "learning_rate": 2e-05,
      "loss": 0.1356,
      "step": 4269
    },
    {
      "epoch": 2.8854342683338965,
      "grad_norm": 8.12353801727295,
      "learning_rate": 2e-05,
      "loss": 0.1747,
      "step": 4270
    },
    {
      "epoch": 2.8861101723555254,
      "grad_norm": 4.449465274810791,
      "learning_rate": 2e-05,
      "loss": 0.1725,
      "step": 4271
    },
    {
      "epoch": 2.8867860763771542,
      "grad_norm": 3.3445546627044678,
      "learning_rate": 2e-05,
      "loss": 0.1325,
      "step": 4272
    },
    {
      "epoch": 2.8874619803987835,
      "grad_norm": 2.557526111602783,
      "learning_rate": 2e-05,
      "loss": 0.1436,
      "step": 4273
    },
    {
      "epoch": 2.8881378844204124,
      "grad_norm": 10.014896392822266,
      "learning_rate": 2e-05,
      "loss": 0.1963,
      "step": 4274
    },
    {
      "epoch": 2.8888137884420413,
      "grad_norm": 6.702539443969727,
      "learning_rate": 2e-05,
      "loss": 0.1435,
      "step": 4275
    },
    {
      "epoch": 2.88948969246367,
      "grad_norm": 4.170548915863037,
      "learning_rate": 2e-05,
      "loss": 0.1837,
      "step": 4276
    },
    {
      "epoch": 2.890165596485299,
      "grad_norm": 2.838500738143921,
      "learning_rate": 2e-05,
      "loss": 0.1699,
      "step": 4277
    },
    {
      "epoch": 2.890841500506928,
      "grad_norm": 6.7317399978637695,
      "learning_rate": 2e-05,
      "loss": 0.1648,
      "step": 4278
    },
    {
      "epoch": 2.8915174045285568,
      "grad_norm": 7.846303462982178,
      "learning_rate": 2e-05,
      "loss": 0.1843,
      "step": 4279
    },
    {
      "epoch": 2.892193308550186,
      "grad_norm": 3.382099151611328,
      "learning_rate": 2e-05,
      "loss": 0.1919,
      "step": 4280
    },
    {
      "epoch": 2.892869212571815,
      "grad_norm": 6.242315292358398,
      "learning_rate": 2e-05,
      "loss": 0.1709,
      "step": 4281
    },
    {
      "epoch": 2.893545116593444,
      "grad_norm": 5.104639053344727,
      "learning_rate": 2e-05,
      "loss": 0.1592,
      "step": 4282
    },
    {
      "epoch": 2.8942210206150727,
      "grad_norm": 6.232515335083008,
      "learning_rate": 2e-05,
      "loss": 0.274,
      "step": 4283
    },
    {
      "epoch": 2.8948969246367016,
      "grad_norm": 5.605199337005615,
      "learning_rate": 2e-05,
      "loss": 0.1788,
      "step": 4284
    },
    {
      "epoch": 2.8955728286583304,
      "grad_norm": 4.569427967071533,
      "learning_rate": 2e-05,
      "loss": 0.1223,
      "step": 4285
    },
    {
      "epoch": 2.8962487326799593,
      "grad_norm": 4.776878356933594,
      "learning_rate": 2e-05,
      "loss": 0.239,
      "step": 4286
    },
    {
      "epoch": 2.8969246367015886,
      "grad_norm": 2.952101707458496,
      "learning_rate": 2e-05,
      "loss": 0.1995,
      "step": 4287
    },
    {
      "epoch": 2.897600540723217,
      "grad_norm": 3.351766347885132,
      "learning_rate": 2e-05,
      "loss": 0.1785,
      "step": 4288
    },
    {
      "epoch": 2.8982764447448464,
      "grad_norm": 10.949353218078613,
      "learning_rate": 2e-05,
      "loss": 0.1716,
      "step": 4289
    },
    {
      "epoch": 2.8989523487664752,
      "grad_norm": 4.329030513763428,
      "learning_rate": 2e-05,
      "loss": 0.1365,
      "step": 4290
    },
    {
      "epoch": 2.899628252788104,
      "grad_norm": 2.632316827774048,
      "learning_rate": 2e-05,
      "loss": 0.1614,
      "step": 4291
    },
    {
      "epoch": 2.900304156809733,
      "grad_norm": 3.9197943210601807,
      "learning_rate": 2e-05,
      "loss": 0.1876,
      "step": 4292
    },
    {
      "epoch": 2.900980060831362,
      "grad_norm": 3.023142099380493,
      "learning_rate": 2e-05,
      "loss": 0.1821,
      "step": 4293
    },
    {
      "epoch": 2.9016559648529907,
      "grad_norm": 4.678294658660889,
      "learning_rate": 2e-05,
      "loss": 0.1776,
      "step": 4294
    },
    {
      "epoch": 2.9023318688746196,
      "grad_norm": 3.2191667556762695,
      "learning_rate": 2e-05,
      "loss": 0.1883,
      "step": 4295
    },
    {
      "epoch": 2.903007772896249,
      "grad_norm": 4.799568176269531,
      "learning_rate": 2e-05,
      "loss": 0.2072,
      "step": 4296
    },
    {
      "epoch": 2.9036836769178778,
      "grad_norm": 4.916661262512207,
      "learning_rate": 2e-05,
      "loss": 0.175,
      "step": 4297
    },
    {
      "epoch": 2.9043595809395066,
      "grad_norm": 6.049447536468506,
      "learning_rate": 2e-05,
      "loss": 0.1876,
      "step": 4298
    },
    {
      "epoch": 2.9050354849611355,
      "grad_norm": 5.298351287841797,
      "learning_rate": 2e-05,
      "loss": 0.1667,
      "step": 4299
    },
    {
      "epoch": 2.9057113889827644,
      "grad_norm": 3.9353749752044678,
      "learning_rate": 2e-05,
      "loss": 0.1094,
      "step": 4300
    },
    {
      "epoch": 2.9057113889827644,
      "eval_accuracy": 0.625464527027027,
      "eval_loss": 0.22377727925777435,
      "eval_runtime": 47.5099,
      "eval_samples_per_second": 498.422,
      "eval_steps_per_second": 15.576,
      "step": 4300
    },
    {
      "epoch": 2.9063872930043932,
      "grad_norm": 3.7418837547302246,
      "learning_rate": 2e-05,
      "loss": 0.1854,
      "step": 4301
    },
    {
      "epoch": 2.907063197026022,
      "grad_norm": 5.739346027374268,
      "learning_rate": 2e-05,
      "loss": 0.2202,
      "step": 4302
    },
    {
      "epoch": 2.9077391010476514,
      "grad_norm": 5.965887546539307,
      "learning_rate": 2e-05,
      "loss": 0.1394,
      "step": 4303
    },
    {
      "epoch": 2.9084150050692803,
      "grad_norm": 4.4496259689331055,
      "learning_rate": 2e-05,
      "loss": 0.1494,
      "step": 4304
    },
    {
      "epoch": 2.909090909090909,
      "grad_norm": 3.5675699710845947,
      "learning_rate": 2e-05,
      "loss": 0.2312,
      "step": 4305
    },
    {
      "epoch": 2.909766813112538,
      "grad_norm": 3.261827230453491,
      "learning_rate": 2e-05,
      "loss": 0.167,
      "step": 4306
    },
    {
      "epoch": 2.910442717134167,
      "grad_norm": 6.292448997497559,
      "learning_rate": 2e-05,
      "loss": 0.1579,
      "step": 4307
    },
    {
      "epoch": 2.911118621155796,
      "grad_norm": 3.126659631729126,
      "learning_rate": 2e-05,
      "loss": 0.1676,
      "step": 4308
    },
    {
      "epoch": 2.9117945251774247,
      "grad_norm": 4.143451690673828,
      "learning_rate": 2e-05,
      "loss": 0.1687,
      "step": 4309
    },
    {
      "epoch": 2.912470429199054,
      "grad_norm": 4.890226364135742,
      "learning_rate": 2e-05,
      "loss": 0.1859,
      "step": 4310
    },
    {
      "epoch": 2.9131463332206824,
      "grad_norm": 2.7414941787719727,
      "learning_rate": 2e-05,
      "loss": 0.1695,
      "step": 4311
    },
    {
      "epoch": 2.9138222372423117,
      "grad_norm": 4.473435878753662,
      "learning_rate": 2e-05,
      "loss": 0.235,
      "step": 4312
    },
    {
      "epoch": 2.9144981412639406,
      "grad_norm": 4.3952507972717285,
      "learning_rate": 2e-05,
      "loss": 0.1764,
      "step": 4313
    },
    {
      "epoch": 2.9151740452855694,
      "grad_norm": 3.6870498657226562,
      "learning_rate": 2e-05,
      "loss": 0.1546,
      "step": 4314
    },
    {
      "epoch": 2.9158499493071983,
      "grad_norm": 5.694862365722656,
      "learning_rate": 2e-05,
      "loss": 0.1736,
      "step": 4315
    },
    {
      "epoch": 2.916525853328827,
      "grad_norm": 5.696045398712158,
      "learning_rate": 2e-05,
      "loss": 0.1797,
      "step": 4316
    },
    {
      "epoch": 2.9172017573504565,
      "grad_norm": 2.775691509246826,
      "learning_rate": 2e-05,
      "loss": 0.1435,
      "step": 4317
    },
    {
      "epoch": 2.917877661372085,
      "grad_norm": 6.157948970794678,
      "learning_rate": 2e-05,
      "loss": 0.1238,
      "step": 4318
    },
    {
      "epoch": 2.9185535653937142,
      "grad_norm": 3.0559091567993164,
      "learning_rate": 2e-05,
      "loss": 0.1362,
      "step": 4319
    },
    {
      "epoch": 2.919229469415343,
      "grad_norm": 3.353671073913574,
      "learning_rate": 2e-05,
      "loss": 0.1781,
      "step": 4320
    },
    {
      "epoch": 2.919905373436972,
      "grad_norm": 10.788576126098633,
      "learning_rate": 2e-05,
      "loss": 0.1916,
      "step": 4321
    },
    {
      "epoch": 2.920581277458601,
      "grad_norm": 5.7703471183776855,
      "learning_rate": 2e-05,
      "loss": 0.1549,
      "step": 4322
    },
    {
      "epoch": 2.9212571814802297,
      "grad_norm": 3.890587091445923,
      "learning_rate": 2e-05,
      "loss": 0.1942,
      "step": 4323
    },
    {
      "epoch": 2.9219330855018586,
      "grad_norm": 5.048915386199951,
      "learning_rate": 2e-05,
      "loss": 0.1399,
      "step": 4324
    },
    {
      "epoch": 2.9226089895234875,
      "grad_norm": 5.276831150054932,
      "learning_rate": 2e-05,
      "loss": 0.1489,
      "step": 4325
    },
    {
      "epoch": 2.923284893545117,
      "grad_norm": 3.2404022216796875,
      "learning_rate": 2e-05,
      "loss": 0.1589,
      "step": 4326
    },
    {
      "epoch": 2.9239607975667457,
      "grad_norm": 4.572868347167969,
      "learning_rate": 2e-05,
      "loss": 0.2043,
      "step": 4327
    },
    {
      "epoch": 2.9246367015883745,
      "grad_norm": 3.323272228240967,
      "learning_rate": 2e-05,
      "loss": 0.143,
      "step": 4328
    },
    {
      "epoch": 2.9253126056100034,
      "grad_norm": 6.709671497344971,
      "learning_rate": 2e-05,
      "loss": 0.1844,
      "step": 4329
    },
    {
      "epoch": 2.9259885096316323,
      "grad_norm": 10.60517692565918,
      "learning_rate": 2e-05,
      "loss": 0.285,
      "step": 4330
    },
    {
      "epoch": 2.926664413653261,
      "grad_norm": 3.095698595046997,
      "learning_rate": 2e-05,
      "loss": 0.1308,
      "step": 4331
    },
    {
      "epoch": 2.92734031767489,
      "grad_norm": 3.251227855682373,
      "learning_rate": 2e-05,
      "loss": 0.1962,
      "step": 4332
    },
    {
      "epoch": 2.9280162216965193,
      "grad_norm": 2.575324535369873,
      "learning_rate": 2e-05,
      "loss": 0.1495,
      "step": 4333
    },
    {
      "epoch": 2.928692125718148,
      "grad_norm": 5.4089741706848145,
      "learning_rate": 2e-05,
      "loss": 0.2107,
      "step": 4334
    },
    {
      "epoch": 2.929368029739777,
      "grad_norm": 6.004955768585205,
      "learning_rate": 2e-05,
      "loss": 0.1933,
      "step": 4335
    },
    {
      "epoch": 2.930043933761406,
      "grad_norm": 4.909146308898926,
      "learning_rate": 2e-05,
      "loss": 0.1127,
      "step": 4336
    },
    {
      "epoch": 2.930719837783035,
      "grad_norm": 3.0166714191436768,
      "learning_rate": 2e-05,
      "loss": 0.1387,
      "step": 4337
    },
    {
      "epoch": 2.9313957418046637,
      "grad_norm": 2.4063425064086914,
      "learning_rate": 2e-05,
      "loss": 0.1309,
      "step": 4338
    },
    {
      "epoch": 2.9320716458262925,
      "grad_norm": 4.357141494750977,
      "learning_rate": 2e-05,
      "loss": 0.1979,
      "step": 4339
    },
    {
      "epoch": 2.932747549847922,
      "grad_norm": 6.059750080108643,
      "learning_rate": 2e-05,
      "loss": 0.1628,
      "step": 4340
    },
    {
      "epoch": 2.9334234538695503,
      "grad_norm": 5.077617645263672,
      "learning_rate": 2e-05,
      "loss": 0.2097,
      "step": 4341
    },
    {
      "epoch": 2.9340993578911796,
      "grad_norm": 2.92315673828125,
      "learning_rate": 2e-05,
      "loss": 0.1604,
      "step": 4342
    },
    {
      "epoch": 2.9347752619128085,
      "grad_norm": 3.065535068511963,
      "learning_rate": 2e-05,
      "loss": 0.1403,
      "step": 4343
    },
    {
      "epoch": 2.9354511659344373,
      "grad_norm": 2.7299344539642334,
      "learning_rate": 2e-05,
      "loss": 0.1747,
      "step": 4344
    },
    {
      "epoch": 2.936127069956066,
      "grad_norm": 4.631441116333008,
      "learning_rate": 2e-05,
      "loss": 0.1312,
      "step": 4345
    },
    {
      "epoch": 2.936802973977695,
      "grad_norm": 4.076853275299072,
      "learning_rate": 2e-05,
      "loss": 0.1629,
      "step": 4346
    },
    {
      "epoch": 2.937478877999324,
      "grad_norm": 6.923962593078613,
      "learning_rate": 2e-05,
      "loss": 0.1799,
      "step": 4347
    },
    {
      "epoch": 2.938154782020953,
      "grad_norm": 8.710369110107422,
      "learning_rate": 2e-05,
      "loss": 0.1687,
      "step": 4348
    },
    {
      "epoch": 2.938830686042582,
      "grad_norm": 5.447247505187988,
      "learning_rate": 2e-05,
      "loss": 0.16,
      "step": 4349
    },
    {
      "epoch": 2.939506590064211,
      "grad_norm": 3.112790584564209,
      "learning_rate": 2e-05,
      "loss": 0.0987,
      "step": 4350
    },
    {
      "epoch": 2.939506590064211,
      "eval_accuracy": 0.6108952702702702,
      "eval_loss": 0.22907879948616028,
      "eval_runtime": 43.4999,
      "eval_samples_per_second": 544.369,
      "eval_steps_per_second": 17.012,
      "step": 4350
    },
    {
      "epoch": 2.94018249408584,
      "grad_norm": 4.1327104568481445,
      "learning_rate": 2e-05,
      "loss": 0.1611,
      "step": 4351
    },
    {
      "epoch": 2.9408583981074687,
      "grad_norm": 6.876573085784912,
      "learning_rate": 2e-05,
      "loss": 0.1901,
      "step": 4352
    },
    {
      "epoch": 2.9415343021290976,
      "grad_norm": 3.1776223182678223,
      "learning_rate": 2e-05,
      "loss": 0.1167,
      "step": 4353
    },
    {
      "epoch": 2.9422102061507265,
      "grad_norm": 2.6286988258361816,
      "learning_rate": 2e-05,
      "loss": 0.1299,
      "step": 4354
    },
    {
      "epoch": 2.9428861101723554,
      "grad_norm": 13.630569458007812,
      "learning_rate": 2e-05,
      "loss": 0.2941,
      "step": 4355
    },
    {
      "epoch": 2.9435620141939847,
      "grad_norm": 6.337322235107422,
      "learning_rate": 2e-05,
      "loss": 0.2233,
      "step": 4356
    },
    {
      "epoch": 2.9442379182156135,
      "grad_norm": 4.081108093261719,
      "learning_rate": 2e-05,
      "loss": 0.2098,
      "step": 4357
    },
    {
      "epoch": 2.9449138222372424,
      "grad_norm": 4.980901718139648,
      "learning_rate": 2e-05,
      "loss": 0.2032,
      "step": 4358
    },
    {
      "epoch": 2.9455897262588713,
      "grad_norm": 3.7127394676208496,
      "learning_rate": 2e-05,
      "loss": 0.1446,
      "step": 4359
    },
    {
      "epoch": 2.9462656302805,
      "grad_norm": 2.098464250564575,
      "learning_rate": 2e-05,
      "loss": 0.1249,
      "step": 4360
    },
    {
      "epoch": 2.946941534302129,
      "grad_norm": 4.753969669342041,
      "learning_rate": 2e-05,
      "loss": 0.1139,
      "step": 4361
    },
    {
      "epoch": 2.947617438323758,
      "grad_norm": 13.63505744934082,
      "learning_rate": 2e-05,
      "loss": 0.1895,
      "step": 4362
    },
    {
      "epoch": 2.948293342345387,
      "grad_norm": 7.465898036956787,
      "learning_rate": 2e-05,
      "loss": 0.1743,
      "step": 4363
    },
    {
      "epoch": 2.9489692463670156,
      "grad_norm": 2.0365452766418457,
      "learning_rate": 2e-05,
      "loss": 0.1229,
      "step": 4364
    },
    {
      "epoch": 2.949645150388645,
      "grad_norm": 2.7300024032592773,
      "learning_rate": 2e-05,
      "loss": 0.1566,
      "step": 4365
    },
    {
      "epoch": 2.950321054410274,
      "grad_norm": 3.275742292404175,
      "learning_rate": 2e-05,
      "loss": 0.107,
      "step": 4366
    },
    {
      "epoch": 2.9509969584319027,
      "grad_norm": 4.540854454040527,
      "learning_rate": 2e-05,
      "loss": 0.1144,
      "step": 4367
    },
    {
      "epoch": 2.9516728624535316,
      "grad_norm": 12.635581016540527,
      "learning_rate": 2e-05,
      "loss": 0.2667,
      "step": 4368
    },
    {
      "epoch": 2.9523487664751604,
      "grad_norm": 7.53907585144043,
      "learning_rate": 2e-05,
      "loss": 0.1723,
      "step": 4369
    },
    {
      "epoch": 2.9530246704967893,
      "grad_norm": 4.056133270263672,
      "learning_rate": 2e-05,
      "loss": 0.1473,
      "step": 4370
    },
    {
      "epoch": 2.953700574518418,
      "grad_norm": 3.686338424682617,
      "learning_rate": 2e-05,
      "loss": 0.1424,
      "step": 4371
    },
    {
      "epoch": 2.9543764785400475,
      "grad_norm": 3.3984365463256836,
      "learning_rate": 2e-05,
      "loss": 0.1635,
      "step": 4372
    },
    {
      "epoch": 2.9550523825616763,
      "grad_norm": 3.833489179611206,
      "learning_rate": 2e-05,
      "loss": 0.1272,
      "step": 4373
    },
    {
      "epoch": 2.955728286583305,
      "grad_norm": 3.617868423461914,
      "learning_rate": 2e-05,
      "loss": 0.1098,
      "step": 4374
    },
    {
      "epoch": 2.956404190604934,
      "grad_norm": 6.80121374130249,
      "learning_rate": 2e-05,
      "loss": 0.2042,
      "step": 4375
    },
    {
      "epoch": 2.957080094626563,
      "grad_norm": 4.942742824554443,
      "learning_rate": 2e-05,
      "loss": 0.1618,
      "step": 4376
    },
    {
      "epoch": 2.957755998648192,
      "grad_norm": 12.78822135925293,
      "learning_rate": 2e-05,
      "loss": 0.3058,
      "step": 4377
    },
    {
      "epoch": 2.9584319026698207,
      "grad_norm": 6.460870265960693,
      "learning_rate": 2e-05,
      "loss": 0.2769,
      "step": 4378
    },
    {
      "epoch": 2.95910780669145,
      "grad_norm": 2.903460741043091,
      "learning_rate": 2e-05,
      "loss": 0.1624,
      "step": 4379
    },
    {
      "epoch": 2.959783710713079,
      "grad_norm": 3.497415781021118,
      "learning_rate": 2e-05,
      "loss": 0.163,
      "step": 4380
    },
    {
      "epoch": 2.9604596147347078,
      "grad_norm": 4.17212438583374,
      "learning_rate": 2e-05,
      "loss": 0.1483,
      "step": 4381
    },
    {
      "epoch": 2.9611355187563366,
      "grad_norm": 8.766508102416992,
      "learning_rate": 2e-05,
      "loss": 0.184,
      "step": 4382
    },
    {
      "epoch": 2.9618114227779655,
      "grad_norm": 6.383959770202637,
      "learning_rate": 2e-05,
      "loss": 0.1617,
      "step": 4383
    },
    {
      "epoch": 2.9624873267995944,
      "grad_norm": 10.755457878112793,
      "learning_rate": 2e-05,
      "loss": 0.2434,
      "step": 4384
    },
    {
      "epoch": 2.9631632308212232,
      "grad_norm": 3.714205265045166,
      "learning_rate": 2e-05,
      "loss": 0.1641,
      "step": 4385
    },
    {
      "epoch": 2.9638391348428526,
      "grad_norm": 5.8531880378723145,
      "learning_rate": 2e-05,
      "loss": 0.1452,
      "step": 4386
    },
    {
      "epoch": 2.964515038864481,
      "grad_norm": 4.116839408874512,
      "learning_rate": 2e-05,
      "loss": 0.196,
      "step": 4387
    },
    {
      "epoch": 2.9651909428861103,
      "grad_norm": 5.889173984527588,
      "learning_rate": 2e-05,
      "loss": 0.1699,
      "step": 4388
    },
    {
      "epoch": 2.965866846907739,
      "grad_norm": 4.580231666564941,
      "learning_rate": 2e-05,
      "loss": 0.1855,
      "step": 4389
    },
    {
      "epoch": 2.966542750929368,
      "grad_norm": 3.5486676692962646,
      "learning_rate": 2e-05,
      "loss": 0.2179,
      "step": 4390
    },
    {
      "epoch": 2.967218654950997,
      "grad_norm": 4.680939674377441,
      "learning_rate": 2e-05,
      "loss": 0.1793,
      "step": 4391
    },
    {
      "epoch": 2.9678945589726258,
      "grad_norm": 3.5844130516052246,
      "learning_rate": 2e-05,
      "loss": 0.2118,
      "step": 4392
    },
    {
      "epoch": 2.968570462994255,
      "grad_norm": 12.440610885620117,
      "learning_rate": 2e-05,
      "loss": 0.2568,
      "step": 4393
    },
    {
      "epoch": 2.9692463670158835,
      "grad_norm": 8.257242202758789,
      "learning_rate": 2e-05,
      "loss": 0.2092,
      "step": 4394
    },
    {
      "epoch": 2.969922271037513,
      "grad_norm": 4.547946453094482,
      "learning_rate": 2e-05,
      "loss": 0.2103,
      "step": 4395
    },
    {
      "epoch": 2.9705981750591417,
      "grad_norm": 3.089972972869873,
      "learning_rate": 2e-05,
      "loss": 0.1555,
      "step": 4396
    },
    {
      "epoch": 2.9712740790807706,
      "grad_norm": 8.969686508178711,
      "learning_rate": 2e-05,
      "loss": 0.1667,
      "step": 4397
    },
    {
      "epoch": 2.9719499831023994,
      "grad_norm": 4.333922863006592,
      "learning_rate": 2e-05,
      "loss": 0.1613,
      "step": 4398
    },
    {
      "epoch": 2.9726258871240283,
      "grad_norm": 3.353353500366211,
      "learning_rate": 2e-05,
      "loss": 0.2209,
      "step": 4399
    },
    {
      "epoch": 2.973301791145657,
      "grad_norm": 3.5068821907043457,
      "learning_rate": 2e-05,
      "loss": 0.1755,
      "step": 4400
    },
    {
      "epoch": 2.973301791145657,
      "eval_accuracy": 0.6184121621621622,
      "eval_loss": 0.22386817634105682,
      "eval_runtime": 44.9112,
      "eval_samples_per_second": 527.263,
      "eval_steps_per_second": 16.477,
      "step": 4400
    },
    {
      "epoch": 2.973977695167286,
      "grad_norm": 3.693638324737549,
      "learning_rate": 2e-05,
      "loss": 0.1406,
      "step": 4401
    },
    {
      "epoch": 2.9746535991889154,
      "grad_norm": 12.419256210327148,
      "learning_rate": 2e-05,
      "loss": 0.2027,
      "step": 4402
    },
    {
      "epoch": 2.9753295032105442,
      "grad_norm": 2.9521923065185547,
      "learning_rate": 2e-05,
      "loss": 0.1515,
      "step": 4403
    },
    {
      "epoch": 2.976005407232173,
      "grad_norm": 3.0946600437164307,
      "learning_rate": 2e-05,
      "loss": 0.1765,
      "step": 4404
    },
    {
      "epoch": 2.976681311253802,
      "grad_norm": 3.563199758529663,
      "learning_rate": 2e-05,
      "loss": 0.2471,
      "step": 4405
    },
    {
      "epoch": 2.977357215275431,
      "grad_norm": 3.8496553897857666,
      "learning_rate": 2e-05,
      "loss": 0.1435,
      "step": 4406
    },
    {
      "epoch": 2.9780331192970597,
      "grad_norm": 3.8990728855133057,
      "learning_rate": 2e-05,
      "loss": 0.1586,
      "step": 4407
    },
    {
      "epoch": 2.9787090233186886,
      "grad_norm": 2.8141872882843018,
      "learning_rate": 2e-05,
      "loss": 0.1509,
      "step": 4408
    },
    {
      "epoch": 2.979384927340318,
      "grad_norm": 4.019842624664307,
      "learning_rate": 2e-05,
      "loss": 0.1989,
      "step": 4409
    },
    {
      "epoch": 2.9800608313619463,
      "grad_norm": 3.6447560787200928,
      "learning_rate": 2e-05,
      "loss": 0.1808,
      "step": 4410
    },
    {
      "epoch": 2.9807367353835756,
      "grad_norm": 3.6977853775024414,
      "learning_rate": 2e-05,
      "loss": 0.1358,
      "step": 4411
    },
    {
      "epoch": 2.9814126394052045,
      "grad_norm": 6.4169535636901855,
      "learning_rate": 2e-05,
      "loss": 0.1672,
      "step": 4412
    },
    {
      "epoch": 2.9820885434268334,
      "grad_norm": 5.608850002288818,
      "learning_rate": 2e-05,
      "loss": 0.17,
      "step": 4413
    },
    {
      "epoch": 2.9827644474484623,
      "grad_norm": 3.177699089050293,
      "learning_rate": 2e-05,
      "loss": 0.1229,
      "step": 4414
    },
    {
      "epoch": 2.983440351470091,
      "grad_norm": 5.524549961090088,
      "learning_rate": 2e-05,
      "loss": 0.196,
      "step": 4415
    },
    {
      "epoch": 2.9841162554917204,
      "grad_norm": 3.516306161880493,
      "learning_rate": 2e-05,
      "loss": 0.1488,
      "step": 4416
    },
    {
      "epoch": 2.984792159513349,
      "grad_norm": 4.329484939575195,
      "learning_rate": 2e-05,
      "loss": 0.2204,
      "step": 4417
    },
    {
      "epoch": 2.985468063534978,
      "grad_norm": 4.172278881072998,
      "learning_rate": 2e-05,
      "loss": 0.187,
      "step": 4418
    },
    {
      "epoch": 2.986143967556607,
      "grad_norm": 4.166094779968262,
      "learning_rate": 2e-05,
      "loss": 0.2291,
      "step": 4419
    },
    {
      "epoch": 2.986819871578236,
      "grad_norm": 2.6657633781433105,
      "learning_rate": 2e-05,
      "loss": 0.1292,
      "step": 4420
    },
    {
      "epoch": 2.987495775599865,
      "grad_norm": 6.389969348907471,
      "learning_rate": 2e-05,
      "loss": 0.1542,
      "step": 4421
    },
    {
      "epoch": 2.9881716796214937,
      "grad_norm": 2.738954782485962,
      "learning_rate": 2e-05,
      "loss": 0.1628,
      "step": 4422
    },
    {
      "epoch": 2.9888475836431225,
      "grad_norm": 3.720449209213257,
      "learning_rate": 2e-05,
      "loss": 0.1349,
      "step": 4423
    },
    {
      "epoch": 2.9895234876647514,
      "grad_norm": 2.6195809841156006,
      "learning_rate": 2e-05,
      "loss": 0.1759,
      "step": 4424
    },
    {
      "epoch": 2.9901993916863807,
      "grad_norm": 3.6290674209594727,
      "learning_rate": 2e-05,
      "loss": 0.2468,
      "step": 4425
    },
    {
      "epoch": 2.9908752957080096,
      "grad_norm": 2.7638328075408936,
      "learning_rate": 2e-05,
      "loss": 0.1728,
      "step": 4426
    },
    {
      "epoch": 2.9915511997296385,
      "grad_norm": 5.647416591644287,
      "learning_rate": 2e-05,
      "loss": 0.1743,
      "step": 4427
    },
    {
      "epoch": 2.9922271037512673,
      "grad_norm": 5.467458724975586,
      "learning_rate": 2e-05,
      "loss": 0.1547,
      "step": 4428
    },
    {
      "epoch": 2.992903007772896,
      "grad_norm": 2.2552826404571533,
      "learning_rate": 2e-05,
      "loss": 0.163,
      "step": 4429
    },
    {
      "epoch": 2.993578911794525,
      "grad_norm": 3.6085498332977295,
      "learning_rate": 2e-05,
      "loss": 0.2016,
      "step": 4430
    },
    {
      "epoch": 2.994254815816154,
      "grad_norm": 6.307020664215088,
      "learning_rate": 2e-05,
      "loss": 0.1793,
      "step": 4431
    },
    {
      "epoch": 2.9949307198377833,
      "grad_norm": 2.953916549682617,
      "learning_rate": 2e-05,
      "loss": 0.1144,
      "step": 4432
    },
    {
      "epoch": 2.995606623859412,
      "grad_norm": 2.8970494270324707,
      "learning_rate": 2e-05,
      "loss": 0.1971,
      "step": 4433
    },
    {
      "epoch": 2.996282527881041,
      "grad_norm": 3.757978677749634,
      "learning_rate": 2e-05,
      "loss": 0.1158,
      "step": 4434
    },
    {
      "epoch": 2.99695843190267,
      "grad_norm": 9.651060104370117,
      "learning_rate": 2e-05,
      "loss": 0.2217,
      "step": 4435
    },
    {
      "epoch": 2.9976343359242987,
      "grad_norm": 6.97280740737915,
      "learning_rate": 2e-05,
      "loss": 0.0937,
      "step": 4436
    },
    {
      "epoch": 2.9983102399459276,
      "grad_norm": 3.8623955249786377,
      "learning_rate": 2e-05,
      "loss": 0.2088,
      "step": 4437
    },
    {
      "epoch": 2.9989861439675565,
      "grad_norm": 5.214295864105225,
      "learning_rate": 2e-05,
      "loss": 0.1757,
      "step": 4438
    },
    {
      "epoch": 2.999662047989186,
      "grad_norm": 2.7299489974975586,
      "learning_rate": 2e-05,
      "loss": 0.1364,
      "step": 4439
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.7299489974975586,
      "learning_rate": 2e-05,
      "loss": 0.0912,
      "step": 4440
    },
    {
      "epoch": 3.000675904021629,
      "grad_norm": 2.5701045989990234,
      "learning_rate": 2e-05,
      "loss": 0.1695,
      "step": 4441
    },
    {
      "epoch": 3.0013518080432577,
      "grad_norm": 6.635688304901123,
      "learning_rate": 2e-05,
      "loss": 0.2194,
      "step": 4442
    },
    {
      "epoch": 3.0020277120648866,
      "grad_norm": 4.840435028076172,
      "learning_rate": 2e-05,
      "loss": 0.1576,
      "step": 4443
    },
    {
      "epoch": 3.002703616086516,
      "grad_norm": 3.946706771850586,
      "learning_rate": 2e-05,
      "loss": 0.1738,
      "step": 4444
    },
    {
      "epoch": 3.003379520108145,
      "grad_norm": 3.3302114009857178,
      "learning_rate": 2e-05,
      "loss": 0.1357,
      "step": 4445
    },
    {
      "epoch": 3.0040554241297737,
      "grad_norm": 2.819836378097534,
      "learning_rate": 2e-05,
      "loss": 0.0974,
      "step": 4446
    },
    {
      "epoch": 3.0047313281514025,
      "grad_norm": 2.7940618991851807,
      "learning_rate": 2e-05,
      "loss": 0.206,
      "step": 4447
    },
    {
      "epoch": 3.0054072321730314,
      "grad_norm": 7.287351608276367,
      "learning_rate": 2e-05,
      "loss": 0.1453,
      "step": 4448
    },
    {
      "epoch": 3.0060831361946603,
      "grad_norm": 8.375088691711426,
      "learning_rate": 2e-05,
      "loss": 0.1707,
      "step": 4449
    },
    {
      "epoch": 3.006759040216289,
      "grad_norm": 3.508229970932007,
      "learning_rate": 2e-05,
      "loss": 0.142,
      "step": 4450
    },
    {
      "epoch": 3.006759040216289,
      "eval_accuracy": 0.6137246621621621,
      "eval_loss": 0.2429080307483673,
      "eval_runtime": 44.1099,
      "eval_samples_per_second": 536.841,
      "eval_steps_per_second": 16.776,
      "step": 4450
    },
    {
      "epoch": 3.007434944237918,
      "grad_norm": 2.8719022274017334,
      "learning_rate": 2e-05,
      "loss": 0.121,
      "step": 4451
    },
    {
      "epoch": 3.0081108482595473,
      "grad_norm": 4.200806140899658,
      "learning_rate": 2e-05,
      "loss": 0.1176,
      "step": 4452
    },
    {
      "epoch": 3.008786752281176,
      "grad_norm": 6.822378158569336,
      "learning_rate": 2e-05,
      "loss": 0.2132,
      "step": 4453
    },
    {
      "epoch": 3.009462656302805,
      "grad_norm": 6.771753311157227,
      "learning_rate": 2e-05,
      "loss": 0.2264,
      "step": 4454
    },
    {
      "epoch": 3.010138560324434,
      "grad_norm": 4.818078517913818,
      "learning_rate": 2e-05,
      "loss": 0.1387,
      "step": 4455
    },
    {
      "epoch": 3.010814464346063,
      "grad_norm": 4.632866859436035,
      "learning_rate": 2e-05,
      "loss": 0.1991,
      "step": 4456
    },
    {
      "epoch": 3.0114903683676917,
      "grad_norm": 3.978179454803467,
      "learning_rate": 2e-05,
      "loss": 0.1497,
      "step": 4457
    },
    {
      "epoch": 3.0121662723893206,
      "grad_norm": 4.5451436042785645,
      "learning_rate": 2e-05,
      "loss": 0.1909,
      "step": 4458
    },
    {
      "epoch": 3.0128421764109494,
      "grad_norm": 3.2660045623779297,
      "learning_rate": 2e-05,
      "loss": 0.1783,
      "step": 4459
    },
    {
      "epoch": 3.0135180804325787,
      "grad_norm": 4.692905902862549,
      "learning_rate": 2e-05,
      "loss": 0.1672,
      "step": 4460
    },
    {
      "epoch": 3.0141939844542076,
      "grad_norm": 8.96642017364502,
      "learning_rate": 2e-05,
      "loss": 0.1457,
      "step": 4461
    },
    {
      "epoch": 3.0148698884758365,
      "grad_norm": 2.6122984886169434,
      "learning_rate": 2e-05,
      "loss": 0.1326,
      "step": 4462
    },
    {
      "epoch": 3.0155457924974653,
      "grad_norm": 4.576719760894775,
      "learning_rate": 2e-05,
      "loss": 0.1576,
      "step": 4463
    },
    {
      "epoch": 3.016221696519094,
      "grad_norm": 4.411528587341309,
      "learning_rate": 2e-05,
      "loss": 0.156,
      "step": 4464
    },
    {
      "epoch": 3.016897600540723,
      "grad_norm": 3.353925943374634,
      "learning_rate": 2e-05,
      "loss": 0.1757,
      "step": 4465
    },
    {
      "epoch": 3.017573504562352,
      "grad_norm": 2.91274356842041,
      "learning_rate": 2e-05,
      "loss": 0.2339,
      "step": 4466
    },
    {
      "epoch": 3.0182494085839813,
      "grad_norm": 3.16196870803833,
      "learning_rate": 2e-05,
      "loss": 0.1575,
      "step": 4467
    },
    {
      "epoch": 3.01892531260561,
      "grad_norm": 5.485776901245117,
      "learning_rate": 2e-05,
      "loss": 0.1292,
      "step": 4468
    },
    {
      "epoch": 3.019601216627239,
      "grad_norm": 2.1826541423797607,
      "learning_rate": 2e-05,
      "loss": 0.1329,
      "step": 4469
    },
    {
      "epoch": 3.020277120648868,
      "grad_norm": 3.2992842197418213,
      "learning_rate": 2e-05,
      "loss": 0.1457,
      "step": 4470
    },
    {
      "epoch": 3.0209530246704968,
      "grad_norm": 3.300422430038452,
      "learning_rate": 2e-05,
      "loss": 0.1708,
      "step": 4471
    },
    {
      "epoch": 3.0216289286921256,
      "grad_norm": 4.55975866317749,
      "learning_rate": 2e-05,
      "loss": 0.183,
      "step": 4472
    },
    {
      "epoch": 3.0223048327137545,
      "grad_norm": 4.737203121185303,
      "learning_rate": 2e-05,
      "loss": 0.1502,
      "step": 4473
    },
    {
      "epoch": 3.0229807367353834,
      "grad_norm": 2.233137369155884,
      "learning_rate": 2e-05,
      "loss": 0.137,
      "step": 4474
    },
    {
      "epoch": 3.0236566407570127,
      "grad_norm": 4.309138774871826,
      "learning_rate": 2e-05,
      "loss": 0.1599,
      "step": 4475
    },
    {
      "epoch": 3.0243325447786416,
      "grad_norm": 4.096744537353516,
      "learning_rate": 2e-05,
      "loss": 0.1612,
      "step": 4476
    },
    {
      "epoch": 3.0250084488002704,
      "grad_norm": 4.431070327758789,
      "learning_rate": 2e-05,
      "loss": 0.1811,
      "step": 4477
    },
    {
      "epoch": 3.0256843528218993,
      "grad_norm": 2.6535210609436035,
      "learning_rate": 2e-05,
      "loss": 0.16,
      "step": 4478
    },
    {
      "epoch": 3.026360256843528,
      "grad_norm": 3.4846858978271484,
      "learning_rate": 2e-05,
      "loss": 0.1793,
      "step": 4479
    },
    {
      "epoch": 3.027036160865157,
      "grad_norm": 10.697047233581543,
      "learning_rate": 2e-05,
      "loss": 0.1424,
      "step": 4480
    },
    {
      "epoch": 3.027712064886786,
      "grad_norm": 3.349545955657959,
      "learning_rate": 2e-05,
      "loss": 0.2025,
      "step": 4481
    },
    {
      "epoch": 3.028387968908415,
      "grad_norm": 7.044104099273682,
      "learning_rate": 2e-05,
      "loss": 0.2119,
      "step": 4482
    },
    {
      "epoch": 3.029063872930044,
      "grad_norm": 6.220849514007568,
      "learning_rate": 2e-05,
      "loss": 0.1818,
      "step": 4483
    },
    {
      "epoch": 3.029739776951673,
      "grad_norm": 4.021681308746338,
      "learning_rate": 2e-05,
      "loss": 0.144,
      "step": 4484
    },
    {
      "epoch": 3.030415680973302,
      "grad_norm": 5.93811559677124,
      "learning_rate": 2e-05,
      "loss": 0.1268,
      "step": 4485
    },
    {
      "epoch": 3.0310915849949307,
      "grad_norm": 2.2422006130218506,
      "learning_rate": 2e-05,
      "loss": 0.1246,
      "step": 4486
    },
    {
      "epoch": 3.0317674890165596,
      "grad_norm": 3.674163579940796,
      "learning_rate": 2e-05,
      "loss": 0.1642,
      "step": 4487
    },
    {
      "epoch": 3.0324433930381884,
      "grad_norm": 4.9736809730529785,
      "learning_rate": 2e-05,
      "loss": 0.2059,
      "step": 4488
    },
    {
      "epoch": 3.0331192970598173,
      "grad_norm": 15.503100395202637,
      "learning_rate": 2e-05,
      "loss": 0.0956,
      "step": 4489
    },
    {
      "epoch": 3.0337952010814466,
      "grad_norm": 4.7133965492248535,
      "learning_rate": 2e-05,
      "loss": 0.1569,
      "step": 4490
    },
    {
      "epoch": 3.0344711051030755,
      "grad_norm": 5.7287116050720215,
      "learning_rate": 2e-05,
      "loss": 0.1446,
      "step": 4491
    },
    {
      "epoch": 3.0351470091247044,
      "grad_norm": 4.832435131072998,
      "learning_rate": 2e-05,
      "loss": 0.1472,
      "step": 4492
    },
    {
      "epoch": 3.0358229131463332,
      "grad_norm": 8.585488319396973,
      "learning_rate": 2e-05,
      "loss": 0.1415,
      "step": 4493
    },
    {
      "epoch": 3.036498817167962,
      "grad_norm": 4.3091278076171875,
      "learning_rate": 2e-05,
      "loss": 0.1697,
      "step": 4494
    },
    {
      "epoch": 3.037174721189591,
      "grad_norm": 5.321987152099609,
      "learning_rate": 2e-05,
      "loss": 0.1567,
      "step": 4495
    },
    {
      "epoch": 3.03785062521122,
      "grad_norm": 5.195062637329102,
      "learning_rate": 2e-05,
      "loss": 0.1166,
      "step": 4496
    },
    {
      "epoch": 3.0385265292328487,
      "grad_norm": 3.6612887382507324,
      "learning_rate": 2e-05,
      "loss": 0.2056,
      "step": 4497
    },
    {
      "epoch": 3.039202433254478,
      "grad_norm": 13.554556846618652,
      "learning_rate": 2e-05,
      "loss": 0.1801,
      "step": 4498
    },
    {
      "epoch": 3.039878337276107,
      "grad_norm": 3.420459270477295,
      "learning_rate": 2e-05,
      "loss": 0.1876,
      "step": 4499
    },
    {
      "epoch": 3.0405542412977358,
      "grad_norm": 3.9458727836608887,
      "learning_rate": 2e-05,
      "loss": 0.1559,
      "step": 4500
    },
    {
      "epoch": 3.0405542412977358,
      "eval_accuracy": 0.6336570945945946,
      "eval_loss": 0.2450345903635025,
      "eval_runtime": 44.1782,
      "eval_samples_per_second": 536.011,
      "eval_steps_per_second": 16.75,
      "step": 4500
    },
    {
      "epoch": 3.0412301453193646,
      "grad_norm": 4.106489658355713,
      "learning_rate": 2e-05,
      "loss": 0.2276,
      "step": 4501
    },
    {
      "epoch": 3.0419060493409935,
      "grad_norm": 5.590481281280518,
      "learning_rate": 2e-05,
      "loss": 0.1688,
      "step": 4502
    },
    {
      "epoch": 3.0425819533626224,
      "grad_norm": 3.5595386028289795,
      "learning_rate": 2e-05,
      "loss": 0.1107,
      "step": 4503
    },
    {
      "epoch": 3.0432578573842513,
      "grad_norm": 3.660727024078369,
      "learning_rate": 2e-05,
      "loss": 0.1468,
      "step": 4504
    },
    {
      "epoch": 3.0439337614058806,
      "grad_norm": 2.3472652435302734,
      "learning_rate": 2e-05,
      "loss": 0.1172,
      "step": 4505
    },
    {
      "epoch": 3.0446096654275094,
      "grad_norm": 2.7880141735076904,
      "learning_rate": 2e-05,
      "loss": 0.2128,
      "step": 4506
    },
    {
      "epoch": 3.0452855694491383,
      "grad_norm": 3.5192131996154785,
      "learning_rate": 2e-05,
      "loss": 0.1716,
      "step": 4507
    },
    {
      "epoch": 3.045961473470767,
      "grad_norm": 2.467113971710205,
      "learning_rate": 2e-05,
      "loss": 0.1555,
      "step": 4508
    },
    {
      "epoch": 3.046637377492396,
      "grad_norm": 3.2380917072296143,
      "learning_rate": 2e-05,
      "loss": 0.1616,
      "step": 4509
    },
    {
      "epoch": 3.047313281514025,
      "grad_norm": 4.682806491851807,
      "learning_rate": 2e-05,
      "loss": 0.1445,
      "step": 4510
    },
    {
      "epoch": 3.047989185535654,
      "grad_norm": 2.2056353092193604,
      "learning_rate": 2e-05,
      "loss": 0.1059,
      "step": 4511
    },
    {
      "epoch": 3.0486650895572827,
      "grad_norm": 4.663336753845215,
      "learning_rate": 2e-05,
      "loss": 0.1282,
      "step": 4512
    },
    {
      "epoch": 3.049340993578912,
      "grad_norm": 7.012662887573242,
      "learning_rate": 2e-05,
      "loss": 0.1592,
      "step": 4513
    },
    {
      "epoch": 3.050016897600541,
      "grad_norm": 2.8418922424316406,
      "learning_rate": 2e-05,
      "loss": 0.2008,
      "step": 4514
    },
    {
      "epoch": 3.0506928016221697,
      "grad_norm": 7.258177280426025,
      "learning_rate": 2e-05,
      "loss": 0.1732,
      "step": 4515
    },
    {
      "epoch": 3.0513687056437986,
      "grad_norm": 8.192657470703125,
      "learning_rate": 2e-05,
      "loss": 0.2106,
      "step": 4516
    },
    {
      "epoch": 3.0520446096654275,
      "grad_norm": 3.199262857437134,
      "learning_rate": 2e-05,
      "loss": 0.1799,
      "step": 4517
    },
    {
      "epoch": 3.0527205136870563,
      "grad_norm": 6.632867813110352,
      "learning_rate": 2e-05,
      "loss": 0.0947,
      "step": 4518
    },
    {
      "epoch": 3.053396417708685,
      "grad_norm": 3.3292295932769775,
      "learning_rate": 2e-05,
      "loss": 0.1574,
      "step": 4519
    },
    {
      "epoch": 3.0540723217303145,
      "grad_norm": 4.404890060424805,
      "learning_rate": 2e-05,
      "loss": 0.1338,
      "step": 4520
    },
    {
      "epoch": 3.0547482257519434,
      "grad_norm": 4.122292995452881,
      "learning_rate": 2e-05,
      "loss": 0.1509,
      "step": 4521
    },
    {
      "epoch": 3.0554241297735723,
      "grad_norm": 2.2262556552886963,
      "learning_rate": 2e-05,
      "loss": 0.0995,
      "step": 4522
    },
    {
      "epoch": 3.056100033795201,
      "grad_norm": 4.288898468017578,
      "learning_rate": 2e-05,
      "loss": 0.1936,
      "step": 4523
    },
    {
      "epoch": 3.05677593781683,
      "grad_norm": 4.437625408172607,
      "learning_rate": 2e-05,
      "loss": 0.1548,
      "step": 4524
    },
    {
      "epoch": 3.057451841838459,
      "grad_norm": 4.182445049285889,
      "learning_rate": 2e-05,
      "loss": 0.2051,
      "step": 4525
    },
    {
      "epoch": 3.0581277458600877,
      "grad_norm": 2.457735300064087,
      "learning_rate": 2e-05,
      "loss": 0.1164,
      "step": 4526
    },
    {
      "epoch": 3.0588036498817166,
      "grad_norm": 3.9583523273468018,
      "learning_rate": 2e-05,
      "loss": 0.1822,
      "step": 4527
    },
    {
      "epoch": 3.059479553903346,
      "grad_norm": 8.37673568725586,
      "learning_rate": 2e-05,
      "loss": 0.1386,
      "step": 4528
    },
    {
      "epoch": 3.060155457924975,
      "grad_norm": 7.010115146636963,
      "learning_rate": 2e-05,
      "loss": 0.2424,
      "step": 4529
    },
    {
      "epoch": 3.0608313619466037,
      "grad_norm": 5.639681816101074,
      "learning_rate": 2e-05,
      "loss": 0.1355,
      "step": 4530
    },
    {
      "epoch": 3.0615072659682325,
      "grad_norm": 5.24914026260376,
      "learning_rate": 2e-05,
      "loss": 0.1724,
      "step": 4531
    },
    {
      "epoch": 3.0621831699898614,
      "grad_norm": 5.562801837921143,
      "learning_rate": 2e-05,
      "loss": 0.1598,
      "step": 4532
    },
    {
      "epoch": 3.0628590740114903,
      "grad_norm": 3.7856736183166504,
      "learning_rate": 2e-05,
      "loss": 0.224,
      "step": 4533
    },
    {
      "epoch": 3.063534978033119,
      "grad_norm": 4.11444616317749,
      "learning_rate": 2e-05,
      "loss": 0.1194,
      "step": 4534
    },
    {
      "epoch": 3.0642108820547485,
      "grad_norm": 5.276716709136963,
      "learning_rate": 2e-05,
      "loss": 0.2123,
      "step": 4535
    },
    {
      "epoch": 3.0648867860763773,
      "grad_norm": 3.7661805152893066,
      "learning_rate": 2e-05,
      "loss": 0.1463,
      "step": 4536
    },
    {
      "epoch": 3.065562690098006,
      "grad_norm": 2.690678358078003,
      "learning_rate": 2e-05,
      "loss": 0.1264,
      "step": 4537
    },
    {
      "epoch": 3.066238594119635,
      "grad_norm": 3.0459682941436768,
      "learning_rate": 2e-05,
      "loss": 0.149,
      "step": 4538
    },
    {
      "epoch": 3.066914498141264,
      "grad_norm": 5.194199562072754,
      "learning_rate": 2e-05,
      "loss": 0.1497,
      "step": 4539
    },
    {
      "epoch": 3.067590402162893,
      "grad_norm": 4.438039302825928,
      "learning_rate": 2e-05,
      "loss": 0.2134,
      "step": 4540
    },
    {
      "epoch": 3.0682663061845217,
      "grad_norm": 3.73629093170166,
      "learning_rate": 2e-05,
      "loss": 0.1823,
      "step": 4541
    },
    {
      "epoch": 3.0689422102061505,
      "grad_norm": 5.341899394989014,
      "learning_rate": 2e-05,
      "loss": 0.2827,
      "step": 4542
    },
    {
      "epoch": 3.06961811422778,
      "grad_norm": 4.219468116760254,
      "learning_rate": 2e-05,
      "loss": 0.2475,
      "step": 4543
    },
    {
      "epoch": 3.0702940182494087,
      "grad_norm": 2.7711181640625,
      "learning_rate": 2e-05,
      "loss": 0.1929,
      "step": 4544
    },
    {
      "epoch": 3.0709699222710376,
      "grad_norm": 2.481503963470459,
      "learning_rate": 2e-05,
      "loss": 0.146,
      "step": 4545
    },
    {
      "epoch": 3.0716458262926665,
      "grad_norm": 2.12373423576355,
      "learning_rate": 2e-05,
      "loss": 0.1308,
      "step": 4546
    },
    {
      "epoch": 3.0723217303142953,
      "grad_norm": 2.671748161315918,
      "learning_rate": 2e-05,
      "loss": 0.2058,
      "step": 4547
    },
    {
      "epoch": 3.072997634335924,
      "grad_norm": 7.398271083831787,
      "learning_rate": 2e-05,
      "loss": 0.1364,
      "step": 4548
    },
    {
      "epoch": 3.073673538357553,
      "grad_norm": 3.721162796020508,
      "learning_rate": 2e-05,
      "loss": 0.2076,
      "step": 4549
    },
    {
      "epoch": 3.074349442379182,
      "grad_norm": 8.384413719177246,
      "learning_rate": 2e-05,
      "loss": 0.1756,
      "step": 4550
    },
    {
      "epoch": 3.074349442379182,
      "eval_accuracy": 0.6918918918918919,
      "eval_loss": 0.22328519821166992,
      "eval_runtime": 43.5926,
      "eval_samples_per_second": 543.211,
      "eval_steps_per_second": 16.975,
      "step": 4550
    },
    {
      "epoch": 3.0750253464008113,
      "grad_norm": 3.602733612060547,
      "learning_rate": 2e-05,
      "loss": 0.1868,
      "step": 4551
    },
    {
      "epoch": 3.07570125042244,
      "grad_norm": 6.311629295349121,
      "learning_rate": 2e-05,
      "loss": 0.1341,
      "step": 4552
    },
    {
      "epoch": 3.076377154444069,
      "grad_norm": 7.732621192932129,
      "learning_rate": 2e-05,
      "loss": 0.1504,
      "step": 4553
    },
    {
      "epoch": 3.077053058465698,
      "grad_norm": 2.771376609802246,
      "learning_rate": 2e-05,
      "loss": 0.1326,
      "step": 4554
    },
    {
      "epoch": 3.0777289624873267,
      "grad_norm": 2.313241958618164,
      "learning_rate": 2e-05,
      "loss": 0.1448,
      "step": 4555
    },
    {
      "epoch": 3.0784048665089556,
      "grad_norm": 6.959941387176514,
      "learning_rate": 2e-05,
      "loss": 0.1913,
      "step": 4556
    },
    {
      "epoch": 3.0790807705305845,
      "grad_norm": 4.822881698608398,
      "learning_rate": 2e-05,
      "loss": 0.1511,
      "step": 4557
    },
    {
      "epoch": 3.079756674552214,
      "grad_norm": 4.442361354827881,
      "learning_rate": 2e-05,
      "loss": 0.13,
      "step": 4558
    },
    {
      "epoch": 3.0804325785738427,
      "grad_norm": 2.596949338912964,
      "learning_rate": 2e-05,
      "loss": 0.121,
      "step": 4559
    },
    {
      "epoch": 3.0811084825954715,
      "grad_norm": 5.498781681060791,
      "learning_rate": 2e-05,
      "loss": 0.2295,
      "step": 4560
    },
    {
      "epoch": 3.0817843866171004,
      "grad_norm": 9.523430824279785,
      "learning_rate": 2e-05,
      "loss": 0.1957,
      "step": 4561
    },
    {
      "epoch": 3.0824602906387293,
      "grad_norm": 5.127994060516357,
      "learning_rate": 2e-05,
      "loss": 0.1842,
      "step": 4562
    },
    {
      "epoch": 3.083136194660358,
      "grad_norm": 6.953667640686035,
      "learning_rate": 2e-05,
      "loss": 0.1524,
      "step": 4563
    },
    {
      "epoch": 3.083812098681987,
      "grad_norm": 5.583928108215332,
      "learning_rate": 2e-05,
      "loss": 0.1699,
      "step": 4564
    },
    {
      "epoch": 3.084488002703616,
      "grad_norm": 4.910676956176758,
      "learning_rate": 2e-05,
      "loss": 0.1335,
      "step": 4565
    },
    {
      "epoch": 3.085163906725245,
      "grad_norm": 3.941115617752075,
      "learning_rate": 2e-05,
      "loss": 0.1777,
      "step": 4566
    },
    {
      "epoch": 3.085839810746874,
      "grad_norm": 5.319295883178711,
      "learning_rate": 2e-05,
      "loss": 0.1564,
      "step": 4567
    },
    {
      "epoch": 3.086515714768503,
      "grad_norm": 3.6890056133270264,
      "learning_rate": 2e-05,
      "loss": 0.2405,
      "step": 4568
    },
    {
      "epoch": 3.087191618790132,
      "grad_norm": 3.917071580886841,
      "learning_rate": 2e-05,
      "loss": 0.1974,
      "step": 4569
    },
    {
      "epoch": 3.0878675228117607,
      "grad_norm": 3.101820230484009,
      "learning_rate": 2e-05,
      "loss": 0.1596,
      "step": 4570
    },
    {
      "epoch": 3.0885434268333896,
      "grad_norm": 3.4253456592559814,
      "learning_rate": 2e-05,
      "loss": 0.1743,
      "step": 4571
    },
    {
      "epoch": 3.0892193308550184,
      "grad_norm": 3.477412700653076,
      "learning_rate": 2e-05,
      "loss": 0.2072,
      "step": 4572
    },
    {
      "epoch": 3.0898952348766473,
      "grad_norm": 3.94240140914917,
      "learning_rate": 2e-05,
      "loss": 0.164,
      "step": 4573
    },
    {
      "epoch": 3.0905711388982766,
      "grad_norm": 3.042628288269043,
      "learning_rate": 2e-05,
      "loss": 0.1032,
      "step": 4574
    },
    {
      "epoch": 3.0912470429199055,
      "grad_norm": 3.3572521209716797,
      "learning_rate": 2e-05,
      "loss": 0.1687,
      "step": 4575
    },
    {
      "epoch": 3.0919229469415344,
      "grad_norm": 2.7968294620513916,
      "learning_rate": 2e-05,
      "loss": 0.1445,
      "step": 4576
    },
    {
      "epoch": 3.0925988509631632,
      "grad_norm": 7.817874431610107,
      "learning_rate": 2e-05,
      "loss": 0.1875,
      "step": 4577
    },
    {
      "epoch": 3.093274754984792,
      "grad_norm": 6.223893165588379,
      "learning_rate": 2e-05,
      "loss": 0.1487,
      "step": 4578
    },
    {
      "epoch": 3.093950659006421,
      "grad_norm": 8.13761043548584,
      "learning_rate": 2e-05,
      "loss": 0.1615,
      "step": 4579
    },
    {
      "epoch": 3.09462656302805,
      "grad_norm": 3.307000160217285,
      "learning_rate": 2e-05,
      "loss": 0.1997,
      "step": 4580
    },
    {
      "epoch": 3.095302467049679,
      "grad_norm": 2.8847203254699707,
      "learning_rate": 2e-05,
      "loss": 0.1553,
      "step": 4581
    },
    {
      "epoch": 3.095978371071308,
      "grad_norm": 3.942613363265991,
      "learning_rate": 2e-05,
      "loss": 0.2044,
      "step": 4582
    },
    {
      "epoch": 3.096654275092937,
      "grad_norm": 3.6861321926116943,
      "learning_rate": 2e-05,
      "loss": 0.205,
      "step": 4583
    },
    {
      "epoch": 3.0973301791145658,
      "grad_norm": 5.346112251281738,
      "learning_rate": 2e-05,
      "loss": 0.2054,
      "step": 4584
    },
    {
      "epoch": 3.0980060831361946,
      "grad_norm": 7.069784641265869,
      "learning_rate": 2e-05,
      "loss": 0.1659,
      "step": 4585
    },
    {
      "epoch": 3.0986819871578235,
      "grad_norm": 5.253779411315918,
      "learning_rate": 2e-05,
      "loss": 0.1902,
      "step": 4586
    },
    {
      "epoch": 3.0993578911794524,
      "grad_norm": 6.7149810791015625,
      "learning_rate": 2e-05,
      "loss": 0.2449,
      "step": 4587
    },
    {
      "epoch": 3.1000337952010812,
      "grad_norm": 2.475088119506836,
      "learning_rate": 2e-05,
      "loss": 0.1142,
      "step": 4588
    },
    {
      "epoch": 3.1007096992227106,
      "grad_norm": 5.4755754470825195,
      "learning_rate": 2e-05,
      "loss": 0.1262,
      "step": 4589
    },
    {
      "epoch": 3.1013856032443394,
      "grad_norm": 2.1907527446746826,
      "learning_rate": 2e-05,
      "loss": 0.1416,
      "step": 4590
    },
    {
      "epoch": 3.1020615072659683,
      "grad_norm": 3.110926866531372,
      "learning_rate": 2e-05,
      "loss": 0.176,
      "step": 4591
    },
    {
      "epoch": 3.102737411287597,
      "grad_norm": 3.674893379211426,
      "learning_rate": 2e-05,
      "loss": 0.1549,
      "step": 4592
    },
    {
      "epoch": 3.103413315309226,
      "grad_norm": 2.964597463607788,
      "learning_rate": 2e-05,
      "loss": 0.1589,
      "step": 4593
    },
    {
      "epoch": 3.104089219330855,
      "grad_norm": 3.7876946926116943,
      "learning_rate": 2e-05,
      "loss": 0.1402,
      "step": 4594
    },
    {
      "epoch": 3.104765123352484,
      "grad_norm": 3.789639472961426,
      "learning_rate": 2e-05,
      "loss": 0.2279,
      "step": 4595
    },
    {
      "epoch": 3.1054410273741127,
      "grad_norm": 5.789622783660889,
      "learning_rate": 2e-05,
      "loss": 0.1492,
      "step": 4596
    },
    {
      "epoch": 3.106116931395742,
      "grad_norm": 3.7121615409851074,
      "learning_rate": 2e-05,
      "loss": 0.1441,
      "step": 4597
    },
    {
      "epoch": 3.106792835417371,
      "grad_norm": 2.9149510860443115,
      "learning_rate": 2e-05,
      "loss": 0.1423,
      "step": 4598
    },
    {
      "epoch": 3.1074687394389997,
      "grad_norm": 6.9625701904296875,
      "learning_rate": 2e-05,
      "loss": 0.1808,
      "step": 4599
    },
    {
      "epoch": 3.1081446434606286,
      "grad_norm": 5.8970513343811035,
      "learning_rate": 2e-05,
      "loss": 0.1466,
      "step": 4600
    },
    {
      "epoch": 3.1081446434606286,
      "eval_accuracy": 0.6875422297297298,
      "eval_loss": 0.23672881722450256,
      "eval_runtime": 43.4974,
      "eval_samples_per_second": 544.401,
      "eval_steps_per_second": 17.013,
      "step": 4600
    },
    {
      "epoch": 3.1088205474822574,
      "grad_norm": 4.674081802368164,
      "learning_rate": 2e-05,
      "loss": 0.1584,
      "step": 4601
    },
    {
      "epoch": 3.1094964515038863,
      "grad_norm": 4.364163875579834,
      "learning_rate": 2e-05,
      "loss": 0.17,
      "step": 4602
    },
    {
      "epoch": 3.110172355525515,
      "grad_norm": 3.476311683654785,
      "learning_rate": 2e-05,
      "loss": 0.1511,
      "step": 4603
    },
    {
      "epoch": 3.1108482595471445,
      "grad_norm": 5.310286998748779,
      "learning_rate": 2e-05,
      "loss": 0.2119,
      "step": 4604
    },
    {
      "epoch": 3.1115241635687734,
      "grad_norm": 3.36671781539917,
      "learning_rate": 2e-05,
      "loss": 0.1361,
      "step": 4605
    },
    {
      "epoch": 3.1122000675904022,
      "grad_norm": 4.089176654815674,
      "learning_rate": 2e-05,
      "loss": 0.197,
      "step": 4606
    },
    {
      "epoch": 3.112875971612031,
      "grad_norm": 7.715679168701172,
      "learning_rate": 2e-05,
      "loss": 0.1573,
      "step": 4607
    },
    {
      "epoch": 3.11355187563366,
      "grad_norm": 5.0763678550720215,
      "learning_rate": 2e-05,
      "loss": 0.249,
      "step": 4608
    },
    {
      "epoch": 3.114227779655289,
      "grad_norm": 2.640749216079712,
      "learning_rate": 2e-05,
      "loss": 0.134,
      "step": 4609
    },
    {
      "epoch": 3.1149036836769177,
      "grad_norm": 4.6169562339782715,
      "learning_rate": 2e-05,
      "loss": 0.1661,
      "step": 4610
    },
    {
      "epoch": 3.1155795876985466,
      "grad_norm": 2.8031792640686035,
      "learning_rate": 2e-05,
      "loss": 0.1461,
      "step": 4611
    },
    {
      "epoch": 3.116255491720176,
      "grad_norm": 2.8081321716308594,
      "learning_rate": 2e-05,
      "loss": 0.1277,
      "step": 4612
    },
    {
      "epoch": 3.1169313957418048,
      "grad_norm": 1.9681322574615479,
      "learning_rate": 2e-05,
      "loss": 0.1584,
      "step": 4613
    },
    {
      "epoch": 3.1176072997634336,
      "grad_norm": 3.5861780643463135,
      "learning_rate": 2e-05,
      "loss": 0.1575,
      "step": 4614
    },
    {
      "epoch": 3.1182832037850625,
      "grad_norm": 2.7930686473846436,
      "learning_rate": 2e-05,
      "loss": 0.2158,
      "step": 4615
    },
    {
      "epoch": 3.1189591078066914,
      "grad_norm": 3.5573761463165283,
      "learning_rate": 2e-05,
      "loss": 0.2033,
      "step": 4616
    },
    {
      "epoch": 3.1196350118283203,
      "grad_norm": 5.20182466506958,
      "learning_rate": 2e-05,
      "loss": 0.198,
      "step": 4617
    },
    {
      "epoch": 3.120310915849949,
      "grad_norm": 3.660456418991089,
      "learning_rate": 2e-05,
      "loss": 0.1411,
      "step": 4618
    },
    {
      "epoch": 3.1209868198715784,
      "grad_norm": 4.062540531158447,
      "learning_rate": 2e-05,
      "loss": 0.1597,
      "step": 4619
    },
    {
      "epoch": 3.1216627238932073,
      "grad_norm": 3.169297933578491,
      "learning_rate": 2e-05,
      "loss": 0.1259,
      "step": 4620
    },
    {
      "epoch": 3.122338627914836,
      "grad_norm": 2.158219337463379,
      "learning_rate": 2e-05,
      "loss": 0.1851,
      "step": 4621
    },
    {
      "epoch": 3.123014531936465,
      "grad_norm": 4.324291706085205,
      "learning_rate": 2e-05,
      "loss": 0.1893,
      "step": 4622
    },
    {
      "epoch": 3.123690435958094,
      "grad_norm": 6.684196472167969,
      "learning_rate": 2e-05,
      "loss": 0.1833,
      "step": 4623
    },
    {
      "epoch": 3.124366339979723,
      "grad_norm": 3.2098467350006104,
      "learning_rate": 2e-05,
      "loss": 0.1286,
      "step": 4624
    },
    {
      "epoch": 3.1250422440013517,
      "grad_norm": 4.84719181060791,
      "learning_rate": 2e-05,
      "loss": 0.1372,
      "step": 4625
    },
    {
      "epoch": 3.1257181480229805,
      "grad_norm": 6.563335418701172,
      "learning_rate": 2e-05,
      "loss": 0.1831,
      "step": 4626
    },
    {
      "epoch": 3.12639405204461,
      "grad_norm": 8.60922908782959,
      "learning_rate": 2e-05,
      "loss": 0.1602,
      "step": 4627
    },
    {
      "epoch": 3.1270699560662387,
      "grad_norm": 3.5775651931762695,
      "learning_rate": 2e-05,
      "loss": 0.141,
      "step": 4628
    },
    {
      "epoch": 3.1277458600878676,
      "grad_norm": 2.190286874771118,
      "learning_rate": 2e-05,
      "loss": 0.1464,
      "step": 4629
    },
    {
      "epoch": 3.1284217641094965,
      "grad_norm": 3.4653286933898926,
      "learning_rate": 2e-05,
      "loss": 0.1803,
      "step": 4630
    },
    {
      "epoch": 3.1290976681311253,
      "grad_norm": 6.528054237365723,
      "learning_rate": 2e-05,
      "loss": 0.2368,
      "step": 4631
    },
    {
      "epoch": 3.129773572152754,
      "grad_norm": 8.93786334991455,
      "learning_rate": 2e-05,
      "loss": 0.1217,
      "step": 4632
    },
    {
      "epoch": 3.130449476174383,
      "grad_norm": 3.4026331901550293,
      "learning_rate": 2e-05,
      "loss": 0.152,
      "step": 4633
    },
    {
      "epoch": 3.1311253801960124,
      "grad_norm": 3.417797088623047,
      "learning_rate": 2e-05,
      "loss": 0.201,
      "step": 4634
    },
    {
      "epoch": 3.1318012842176413,
      "grad_norm": 3.552384853363037,
      "learning_rate": 2e-05,
      "loss": 0.1676,
      "step": 4635
    },
    {
      "epoch": 3.13247718823927,
      "grad_norm": 5.442955493927002,
      "learning_rate": 2e-05,
      "loss": 0.1521,
      "step": 4636
    },
    {
      "epoch": 3.133153092260899,
      "grad_norm": 4.569618225097656,
      "learning_rate": 2e-05,
      "loss": 0.1687,
      "step": 4637
    },
    {
      "epoch": 3.133828996282528,
      "grad_norm": 3.559675931930542,
      "learning_rate": 2e-05,
      "loss": 0.1918,
      "step": 4638
    },
    {
      "epoch": 3.1345049003041567,
      "grad_norm": 2.8223907947540283,
      "learning_rate": 2e-05,
      "loss": 0.1657,
      "step": 4639
    },
    {
      "epoch": 3.1351808043257856,
      "grad_norm": 7.634295463562012,
      "learning_rate": 2e-05,
      "loss": 0.2177,
      "step": 4640
    },
    {
      "epoch": 3.1358567083474145,
      "grad_norm": 3.448239803314209,
      "learning_rate": 2e-05,
      "loss": 0.1192,
      "step": 4641
    },
    {
      "epoch": 3.136532612369044,
      "grad_norm": 3.266571283340454,
      "learning_rate": 2e-05,
      "loss": 0.1819,
      "step": 4642
    },
    {
      "epoch": 3.1372085163906727,
      "grad_norm": 4.0470805168151855,
      "learning_rate": 2e-05,
      "loss": 0.2031,
      "step": 4643
    },
    {
      "epoch": 3.1378844204123015,
      "grad_norm": 9.048531532287598,
      "learning_rate": 2e-05,
      "loss": 0.1132,
      "step": 4644
    },
    {
      "epoch": 3.1385603244339304,
      "grad_norm": 2.773056983947754,
      "learning_rate": 2e-05,
      "loss": 0.181,
      "step": 4645
    },
    {
      "epoch": 3.1392362284555593,
      "grad_norm": 4.118150234222412,
      "learning_rate": 2e-05,
      "loss": 0.1032,
      "step": 4646
    },
    {
      "epoch": 3.139912132477188,
      "grad_norm": 6.675115585327148,
      "learning_rate": 2e-05,
      "loss": 0.1832,
      "step": 4647
    },
    {
      "epoch": 3.140588036498817,
      "grad_norm": 2.0793333053588867,
      "learning_rate": 2e-05,
      "loss": 0.108,
      "step": 4648
    },
    {
      "epoch": 3.141263940520446,
      "grad_norm": 2.516632318496704,
      "learning_rate": 2e-05,
      "loss": 0.1335,
      "step": 4649
    },
    {
      "epoch": 3.141939844542075,
      "grad_norm": 5.1779704093933105,
      "learning_rate": 2e-05,
      "loss": 0.2055,
      "step": 4650
    },
    {
      "epoch": 3.141939844542075,
      "eval_accuracy": 0.666722972972973,
      "eval_loss": 0.24565130472183228,
      "eval_runtime": 43.1969,
      "eval_samples_per_second": 548.188,
      "eval_steps_per_second": 17.131,
      "step": 4650
    },
    {
      "epoch": 3.142615748563704,
      "grad_norm": 7.977149963378906,
      "learning_rate": 2e-05,
      "loss": 0.1505,
      "step": 4651
    },
    {
      "epoch": 3.143291652585333,
      "grad_norm": 4.164057731628418,
      "learning_rate": 2e-05,
      "loss": 0.1584,
      "step": 4652
    },
    {
      "epoch": 3.143967556606962,
      "grad_norm": 3.6418826580047607,
      "learning_rate": 2e-05,
      "loss": 0.154,
      "step": 4653
    },
    {
      "epoch": 3.1446434606285907,
      "grad_norm": 3.3270509243011475,
      "learning_rate": 2e-05,
      "loss": 0.1237,
      "step": 4654
    },
    {
      "epoch": 3.1453193646502196,
      "grad_norm": 4.996715545654297,
      "learning_rate": 2e-05,
      "loss": 0.2582,
      "step": 4655
    },
    {
      "epoch": 3.1459952686718484,
      "grad_norm": 10.385043144226074,
      "learning_rate": 2e-05,
      "loss": 0.141,
      "step": 4656
    },
    {
      "epoch": 3.1466711726934777,
      "grad_norm": 2.9244892597198486,
      "learning_rate": 2e-05,
      "loss": 0.1191,
      "step": 4657
    },
    {
      "epoch": 3.1473470767151066,
      "grad_norm": 3.0397586822509766,
      "learning_rate": 2e-05,
      "loss": 0.1485,
      "step": 4658
    },
    {
      "epoch": 3.1480229807367355,
      "grad_norm": 3.7158427238464355,
      "learning_rate": 2e-05,
      "loss": 0.1595,
      "step": 4659
    },
    {
      "epoch": 3.1486988847583643,
      "grad_norm": 2.418213367462158,
      "learning_rate": 2e-05,
      "loss": 0.1259,
      "step": 4660
    },
    {
      "epoch": 3.149374788779993,
      "grad_norm": 4.474905014038086,
      "learning_rate": 2e-05,
      "loss": 0.1147,
      "step": 4661
    },
    {
      "epoch": 3.150050692801622,
      "grad_norm": 4.549851894378662,
      "learning_rate": 2e-05,
      "loss": 0.2339,
      "step": 4662
    },
    {
      "epoch": 3.150726596823251,
      "grad_norm": 5.292133331298828,
      "learning_rate": 2e-05,
      "loss": 0.1345,
      "step": 4663
    },
    {
      "epoch": 3.15140250084488,
      "grad_norm": 3.9113681316375732,
      "learning_rate": 2e-05,
      "loss": 0.1375,
      "step": 4664
    },
    {
      "epoch": 3.152078404866509,
      "grad_norm": 2.7917590141296387,
      "learning_rate": 2e-05,
      "loss": 0.185,
      "step": 4665
    },
    {
      "epoch": 3.152754308888138,
      "grad_norm": 4.2090744972229,
      "learning_rate": 2e-05,
      "loss": 0.1456,
      "step": 4666
    },
    {
      "epoch": 3.153430212909767,
      "grad_norm": 3.5527186393737793,
      "learning_rate": 2e-05,
      "loss": 0.1267,
      "step": 4667
    },
    {
      "epoch": 3.1541061169313958,
      "grad_norm": 4.65567684173584,
      "learning_rate": 2e-05,
      "loss": 0.1019,
      "step": 4668
    },
    {
      "epoch": 3.1547820209530246,
      "grad_norm": 3.7026095390319824,
      "learning_rate": 2e-05,
      "loss": 0.085,
      "step": 4669
    },
    {
      "epoch": 3.1554579249746535,
      "grad_norm": 7.123610496520996,
      "learning_rate": 2e-05,
      "loss": 0.1552,
      "step": 4670
    },
    {
      "epoch": 3.1561338289962824,
      "grad_norm": 2.3289287090301514,
      "learning_rate": 2e-05,
      "loss": 0.1496,
      "step": 4671
    },
    {
      "epoch": 3.1568097330179112,
      "grad_norm": 6.8559794425964355,
      "learning_rate": 2e-05,
      "loss": 0.1774,
      "step": 4672
    },
    {
      "epoch": 3.1574856370395405,
      "grad_norm": 5.28338098526001,
      "learning_rate": 2e-05,
      "loss": 0.1576,
      "step": 4673
    },
    {
      "epoch": 3.1581615410611694,
      "grad_norm": 3.2735867500305176,
      "learning_rate": 2e-05,
      "loss": 0.1446,
      "step": 4674
    },
    {
      "epoch": 3.1588374450827983,
      "grad_norm": 3.0894737243652344,
      "learning_rate": 2e-05,
      "loss": 0.1482,
      "step": 4675
    },
    {
      "epoch": 3.159513349104427,
      "grad_norm": 4.279919147491455,
      "learning_rate": 2e-05,
      "loss": 0.1647,
      "step": 4676
    },
    {
      "epoch": 3.160189253126056,
      "grad_norm": 5.99785041809082,
      "learning_rate": 2e-05,
      "loss": 0.1704,
      "step": 4677
    },
    {
      "epoch": 3.160865157147685,
      "grad_norm": 7.739467144012451,
      "learning_rate": 2e-05,
      "loss": 0.1659,
      "step": 4678
    },
    {
      "epoch": 3.1615410611693138,
      "grad_norm": 7.025681018829346,
      "learning_rate": 2e-05,
      "loss": 0.1711,
      "step": 4679
    },
    {
      "epoch": 3.162216965190943,
      "grad_norm": 4.500668525695801,
      "learning_rate": 2e-05,
      "loss": 0.1886,
      "step": 4680
    },
    {
      "epoch": 3.162892869212572,
      "grad_norm": 3.192917823791504,
      "learning_rate": 2e-05,
      "loss": 0.1795,
      "step": 4681
    },
    {
      "epoch": 3.163568773234201,
      "grad_norm": 2.60766339302063,
      "learning_rate": 2e-05,
      "loss": 0.1257,
      "step": 4682
    },
    {
      "epoch": 3.1642446772558297,
      "grad_norm": 2.66562819480896,
      "learning_rate": 2e-05,
      "loss": 0.1538,
      "step": 4683
    },
    {
      "epoch": 3.1649205812774586,
      "grad_norm": 6.161983966827393,
      "learning_rate": 2e-05,
      "loss": 0.2552,
      "step": 4684
    },
    {
      "epoch": 3.1655964852990874,
      "grad_norm": 7.103938102722168,
      "learning_rate": 2e-05,
      "loss": 0.1822,
      "step": 4685
    },
    {
      "epoch": 3.1662723893207163,
      "grad_norm": 6.945894718170166,
      "learning_rate": 2e-05,
      "loss": 0.1527,
      "step": 4686
    },
    {
      "epoch": 3.1669482933423456,
      "grad_norm": 9.93480110168457,
      "learning_rate": 2e-05,
      "loss": 0.2031,
      "step": 4687
    },
    {
      "epoch": 3.1676241973639745,
      "grad_norm": 2.974353790283203,
      "learning_rate": 2e-05,
      "loss": 0.2211,
      "step": 4688
    },
    {
      "epoch": 3.1683001013856034,
      "grad_norm": 4.838623523712158,
      "learning_rate": 2e-05,
      "loss": 0.2176,
      "step": 4689
    },
    {
      "epoch": 3.1689760054072322,
      "grad_norm": 5.547061443328857,
      "learning_rate": 2e-05,
      "loss": 0.2186,
      "step": 4690
    },
    {
      "epoch": 3.169651909428861,
      "grad_norm": 7.319096565246582,
      "learning_rate": 2e-05,
      "loss": 0.1983,
      "step": 4691
    },
    {
      "epoch": 3.17032781345049,
      "grad_norm": 10.203378677368164,
      "learning_rate": 2e-05,
      "loss": 0.1978,
      "step": 4692
    },
    {
      "epoch": 3.171003717472119,
      "grad_norm": 9.099552154541016,
      "learning_rate": 2e-05,
      "loss": 0.2019,
      "step": 4693
    },
    {
      "epoch": 3.1716796214937477,
      "grad_norm": 5.420621395111084,
      "learning_rate": 2e-05,
      "loss": 0.1596,
      "step": 4694
    },
    {
      "epoch": 3.1723555255153766,
      "grad_norm": 2.4237477779388428,
      "learning_rate": 2e-05,
      "loss": 0.1521,
      "step": 4695
    },
    {
      "epoch": 3.173031429537006,
      "grad_norm": 2.571101665496826,
      "learning_rate": 2e-05,
      "loss": 0.1968,
      "step": 4696
    },
    {
      "epoch": 3.1737073335586348,
      "grad_norm": 9.222325325012207,
      "learning_rate": 2e-05,
      "loss": 0.1784,
      "step": 4697
    },
    {
      "epoch": 3.1743832375802636,
      "grad_norm": 4.129406452178955,
      "learning_rate": 2e-05,
      "loss": 0.0965,
      "step": 4698
    },
    {
      "epoch": 3.1750591416018925,
      "grad_norm": 7.222982883453369,
      "learning_rate": 2e-05,
      "loss": 0.1553,
      "step": 4699
    },
    {
      "epoch": 3.1757350456235214,
      "grad_norm": 3.9501073360443115,
      "learning_rate": 2e-05,
      "loss": 0.1809,
      "step": 4700
    },
    {
      "epoch": 3.1757350456235214,
      "eval_accuracy": 0.7340371621621622,
      "eval_loss": 0.24260114133358002,
      "eval_runtime": 43.3261,
      "eval_samples_per_second": 546.553,
      "eval_steps_per_second": 17.08,
      "step": 4700
    },
    {
      "epoch": 3.1764109496451503,
      "grad_norm": 2.516392946243286,
      "learning_rate": 2e-05,
      "loss": 0.1641,
      "step": 4701
    },
    {
      "epoch": 3.177086853666779,
      "grad_norm": 4.2964911460876465,
      "learning_rate": 2e-05,
      "loss": 0.1702,
      "step": 4702
    },
    {
      "epoch": 3.1777627576884084,
      "grad_norm": 7.716554164886475,
      "learning_rate": 2e-05,
      "loss": 0.1425,
      "step": 4703
    },
    {
      "epoch": 3.1784386617100373,
      "grad_norm": 3.6200006008148193,
      "learning_rate": 2e-05,
      "loss": 0.149,
      "step": 4704
    },
    {
      "epoch": 3.179114565731666,
      "grad_norm": 8.90368366241455,
      "learning_rate": 2e-05,
      "loss": 0.1817,
      "step": 4705
    },
    {
      "epoch": 3.179790469753295,
      "grad_norm": 3.618197441101074,
      "learning_rate": 2e-05,
      "loss": 0.177,
      "step": 4706
    },
    {
      "epoch": 3.180466373774924,
      "grad_norm": 3.950077533721924,
      "learning_rate": 2e-05,
      "loss": 0.1173,
      "step": 4707
    },
    {
      "epoch": 3.181142277796553,
      "grad_norm": 9.842316627502441,
      "learning_rate": 2e-05,
      "loss": 0.147,
      "step": 4708
    },
    {
      "epoch": 3.1818181818181817,
      "grad_norm": 6.714404106140137,
      "learning_rate": 2e-05,
      "loss": 0.1911,
      "step": 4709
    },
    {
      "epoch": 3.182494085839811,
      "grad_norm": 4.029579162597656,
      "learning_rate": 2e-05,
      "loss": 0.1692,
      "step": 4710
    },
    {
      "epoch": 3.18316998986144,
      "grad_norm": 4.239700794219971,
      "learning_rate": 2e-05,
      "loss": 0.1889,
      "step": 4711
    },
    {
      "epoch": 3.1838458938830687,
      "grad_norm": 3.74059796333313,
      "learning_rate": 2e-05,
      "loss": 0.1221,
      "step": 4712
    },
    {
      "epoch": 3.1845217979046976,
      "grad_norm": 3.5573744773864746,
      "learning_rate": 2e-05,
      "loss": 0.2342,
      "step": 4713
    },
    {
      "epoch": 3.1851977019263265,
      "grad_norm": 4.3249430656433105,
      "learning_rate": 2e-05,
      "loss": 0.1801,
      "step": 4714
    },
    {
      "epoch": 3.1858736059479553,
      "grad_norm": 5.035532474517822,
      "learning_rate": 2e-05,
      "loss": 0.1855,
      "step": 4715
    },
    {
      "epoch": 3.186549509969584,
      "grad_norm": 5.230154037475586,
      "learning_rate": 2e-05,
      "loss": 0.2174,
      "step": 4716
    },
    {
      "epoch": 3.187225413991213,
      "grad_norm": 7.46312141418457,
      "learning_rate": 2e-05,
      "loss": 0.0902,
      "step": 4717
    },
    {
      "epoch": 3.1879013180128424,
      "grad_norm": 4.225657939910889,
      "learning_rate": 2e-05,
      "loss": 0.1793,
      "step": 4718
    },
    {
      "epoch": 3.1885772220344712,
      "grad_norm": 9.236992835998535,
      "learning_rate": 2e-05,
      "loss": 0.1402,
      "step": 4719
    },
    {
      "epoch": 3.1892531260561,
      "grad_norm": 2.875502586364746,
      "learning_rate": 2e-05,
      "loss": 0.1762,
      "step": 4720
    },
    {
      "epoch": 3.189929030077729,
      "grad_norm": 4.246948719024658,
      "learning_rate": 2e-05,
      "loss": 0.2704,
      "step": 4721
    },
    {
      "epoch": 3.190604934099358,
      "grad_norm": 4.219547271728516,
      "learning_rate": 2e-05,
      "loss": 0.198,
      "step": 4722
    },
    {
      "epoch": 3.1912808381209867,
      "grad_norm": 2.9871866703033447,
      "learning_rate": 2e-05,
      "loss": 0.1795,
      "step": 4723
    },
    {
      "epoch": 3.1919567421426156,
      "grad_norm": 5.325039386749268,
      "learning_rate": 2e-05,
      "loss": 0.1493,
      "step": 4724
    },
    {
      "epoch": 3.1926326461642445,
      "grad_norm": 7.266176223754883,
      "learning_rate": 2e-05,
      "loss": 0.2127,
      "step": 4725
    },
    {
      "epoch": 3.193308550185874,
      "grad_norm": 3.6359450817108154,
      "learning_rate": 2e-05,
      "loss": 0.1628,
      "step": 4726
    },
    {
      "epoch": 3.1939844542075027,
      "grad_norm": 3.615445852279663,
      "learning_rate": 2e-05,
      "loss": 0.1643,
      "step": 4727
    },
    {
      "epoch": 3.1946603582291315,
      "grad_norm": 4.408514976501465,
      "learning_rate": 2e-05,
      "loss": 0.1741,
      "step": 4728
    },
    {
      "epoch": 3.1953362622507604,
      "grad_norm": 7.852147102355957,
      "learning_rate": 2e-05,
      "loss": 0.2029,
      "step": 4729
    },
    {
      "epoch": 3.1960121662723893,
      "grad_norm": 8.06386661529541,
      "learning_rate": 2e-05,
      "loss": 0.1921,
      "step": 4730
    },
    {
      "epoch": 3.196688070294018,
      "grad_norm": 10.51245403289795,
      "learning_rate": 2e-05,
      "loss": 0.2008,
      "step": 4731
    },
    {
      "epoch": 3.197363974315647,
      "grad_norm": 4.7568511962890625,
      "learning_rate": 2e-05,
      "loss": 0.1394,
      "step": 4732
    },
    {
      "epoch": 3.1980398783372763,
      "grad_norm": 5.857522964477539,
      "learning_rate": 2e-05,
      "loss": 0.1925,
      "step": 4733
    },
    {
      "epoch": 3.198715782358905,
      "grad_norm": 6.957383632659912,
      "learning_rate": 2e-05,
      "loss": 0.1874,
      "step": 4734
    },
    {
      "epoch": 3.199391686380534,
      "grad_norm": 4.317417144775391,
      "learning_rate": 2e-05,
      "loss": 0.1974,
      "step": 4735
    },
    {
      "epoch": 3.200067590402163,
      "grad_norm": 2.6813113689422607,
      "learning_rate": 2e-05,
      "loss": 0.1523,
      "step": 4736
    },
    {
      "epoch": 3.200743494423792,
      "grad_norm": 2.6136014461517334,
      "learning_rate": 2e-05,
      "loss": 0.1536,
      "step": 4737
    },
    {
      "epoch": 3.2014193984454207,
      "grad_norm": 5.749530792236328,
      "learning_rate": 2e-05,
      "loss": 0.1408,
      "step": 4738
    },
    {
      "epoch": 3.2020953024670495,
      "grad_norm": 6.520163536071777,
      "learning_rate": 2e-05,
      "loss": 0.1811,
      "step": 4739
    },
    {
      "epoch": 3.2027712064886784,
      "grad_norm": 5.358052730560303,
      "learning_rate": 2e-05,
      "loss": 0.1646,
      "step": 4740
    },
    {
      "epoch": 3.2034471105103077,
      "grad_norm": 13.17647933959961,
      "learning_rate": 2e-05,
      "loss": 0.2073,
      "step": 4741
    },
    {
      "epoch": 3.2041230145319366,
      "grad_norm": 3.173050880432129,
      "learning_rate": 2e-05,
      "loss": 0.2033,
      "step": 4742
    },
    {
      "epoch": 3.2047989185535655,
      "grad_norm": 6.3750176429748535,
      "learning_rate": 2e-05,
      "loss": 0.1889,
      "step": 4743
    },
    {
      "epoch": 3.2054748225751943,
      "grad_norm": 9.134383201599121,
      "learning_rate": 2e-05,
      "loss": 0.1125,
      "step": 4744
    },
    {
      "epoch": 3.206150726596823,
      "grad_norm": 6.183806419372559,
      "learning_rate": 2e-05,
      "loss": 0.149,
      "step": 4745
    },
    {
      "epoch": 3.206826630618452,
      "grad_norm": 7.275829315185547,
      "learning_rate": 2e-05,
      "loss": 0.1689,
      "step": 4746
    },
    {
      "epoch": 3.207502534640081,
      "grad_norm": 3.8284361362457275,
      "learning_rate": 2e-05,
      "loss": 0.1491,
      "step": 4747
    },
    {
      "epoch": 3.20817843866171,
      "grad_norm": 5.58668327331543,
      "learning_rate": 2e-05,
      "loss": 0.1762,
      "step": 4748
    },
    {
      "epoch": 3.208854342683339,
      "grad_norm": 3.194458246231079,
      "learning_rate": 2e-05,
      "loss": 0.1758,
      "step": 4749
    },
    {
      "epoch": 3.209530246704968,
      "grad_norm": 6.067990779876709,
      "learning_rate": 2e-05,
      "loss": 0.2142,
      "step": 4750
    },
    {
      "epoch": 3.209530246704968,
      "eval_accuracy": 0.7229307432432432,
      "eval_loss": 0.2532256543636322,
      "eval_runtime": 43.4665,
      "eval_samples_per_second": 544.787,
      "eval_steps_per_second": 17.025,
      "step": 4750
    },
    {
      "epoch": 3.210206150726597,
      "grad_norm": 4.431941986083984,
      "learning_rate": 2e-05,
      "loss": 0.1909,
      "step": 4751
    },
    {
      "epoch": 3.2108820547482257,
      "grad_norm": 6.405493259429932,
      "learning_rate": 2e-05,
      "loss": 0.1401,
      "step": 4752
    },
    {
      "epoch": 3.2115579587698546,
      "grad_norm": 5.848548412322998,
      "learning_rate": 2e-05,
      "loss": 0.1675,
      "step": 4753
    },
    {
      "epoch": 3.2122338627914835,
      "grad_norm": 4.216307163238525,
      "learning_rate": 2e-05,
      "loss": 0.2021,
      "step": 4754
    },
    {
      "epoch": 3.2129097668131124,
      "grad_norm": 4.814270496368408,
      "learning_rate": 2e-05,
      "loss": 0.2129,
      "step": 4755
    },
    {
      "epoch": 3.2135856708347417,
      "grad_norm": 3.591844081878662,
      "learning_rate": 2e-05,
      "loss": 0.1437,
      "step": 4756
    },
    {
      "epoch": 3.2142615748563705,
      "grad_norm": 2.9623823165893555,
      "learning_rate": 2e-05,
      "loss": 0.2077,
      "step": 4757
    },
    {
      "epoch": 3.2149374788779994,
      "grad_norm": 3.718191623687744,
      "learning_rate": 2e-05,
      "loss": 0.1873,
      "step": 4758
    },
    {
      "epoch": 3.2156133828996283,
      "grad_norm": 2.3246195316314697,
      "learning_rate": 2e-05,
      "loss": 0.147,
      "step": 4759
    },
    {
      "epoch": 3.216289286921257,
      "grad_norm": 4.489903926849365,
      "learning_rate": 2e-05,
      "loss": 0.1285,
      "step": 4760
    },
    {
      "epoch": 3.216965190942886,
      "grad_norm": 2.5921823978424072,
      "learning_rate": 2e-05,
      "loss": 0.1395,
      "step": 4761
    },
    {
      "epoch": 3.217641094964515,
      "grad_norm": 4.790523529052734,
      "learning_rate": 2e-05,
      "loss": 0.1989,
      "step": 4762
    },
    {
      "epoch": 3.2183169989861438,
      "grad_norm": 3.1578543186187744,
      "learning_rate": 2e-05,
      "loss": 0.2213,
      "step": 4763
    },
    {
      "epoch": 3.218992903007773,
      "grad_norm": 3.8059654235839844,
      "learning_rate": 2e-05,
      "loss": 0.1208,
      "step": 4764
    },
    {
      "epoch": 3.219668807029402,
      "grad_norm": 6.007192611694336,
      "learning_rate": 2e-05,
      "loss": 0.1545,
      "step": 4765
    },
    {
      "epoch": 3.220344711051031,
      "grad_norm": 4.225831985473633,
      "learning_rate": 2e-05,
      "loss": 0.1709,
      "step": 4766
    },
    {
      "epoch": 3.2210206150726597,
      "grad_norm": 5.134098052978516,
      "learning_rate": 2e-05,
      "loss": 0.1278,
      "step": 4767
    },
    {
      "epoch": 3.2216965190942886,
      "grad_norm": 4.977596282958984,
      "learning_rate": 2e-05,
      "loss": 0.1201,
      "step": 4768
    },
    {
      "epoch": 3.2223724231159174,
      "grad_norm": 2.54329252243042,
      "learning_rate": 2e-05,
      "loss": 0.178,
      "step": 4769
    },
    {
      "epoch": 3.2230483271375463,
      "grad_norm": 3.937161684036255,
      "learning_rate": 2e-05,
      "loss": 0.223,
      "step": 4770
    },
    {
      "epoch": 3.223724231159175,
      "grad_norm": 4.628030776977539,
      "learning_rate": 2e-05,
      "loss": 0.1128,
      "step": 4771
    },
    {
      "epoch": 3.2244001351808045,
      "grad_norm": 2.789734125137329,
      "learning_rate": 2e-05,
      "loss": 0.1725,
      "step": 4772
    },
    {
      "epoch": 3.2250760392024334,
      "grad_norm": 5.5802483558654785,
      "learning_rate": 2e-05,
      "loss": 0.1144,
      "step": 4773
    },
    {
      "epoch": 3.2257519432240622,
      "grad_norm": 2.3869271278381348,
      "learning_rate": 2e-05,
      "loss": 0.1436,
      "step": 4774
    },
    {
      "epoch": 3.226427847245691,
      "grad_norm": 3.3115642070770264,
      "learning_rate": 2e-05,
      "loss": 0.1609,
      "step": 4775
    },
    {
      "epoch": 3.22710375126732,
      "grad_norm": 5.0943732261657715,
      "learning_rate": 2e-05,
      "loss": 0.1464,
      "step": 4776
    },
    {
      "epoch": 3.227779655288949,
      "grad_norm": 4.599772930145264,
      "learning_rate": 2e-05,
      "loss": 0.157,
      "step": 4777
    },
    {
      "epoch": 3.2284555593105777,
      "grad_norm": 3.592033624649048,
      "learning_rate": 2e-05,
      "loss": 0.1382,
      "step": 4778
    },
    {
      "epoch": 3.229131463332207,
      "grad_norm": 5.1632399559021,
      "learning_rate": 2e-05,
      "loss": 0.1615,
      "step": 4779
    },
    {
      "epoch": 3.229807367353836,
      "grad_norm": 4.535800457000732,
      "learning_rate": 2e-05,
      "loss": 0.1635,
      "step": 4780
    },
    {
      "epoch": 3.2304832713754648,
      "grad_norm": 5.965167999267578,
      "learning_rate": 2e-05,
      "loss": 0.1989,
      "step": 4781
    },
    {
      "epoch": 3.2311591753970936,
      "grad_norm": 4.768083572387695,
      "learning_rate": 2e-05,
      "loss": 0.1914,
      "step": 4782
    },
    {
      "epoch": 3.2318350794187225,
      "grad_norm": 5.204427242279053,
      "learning_rate": 2e-05,
      "loss": 0.1632,
      "step": 4783
    },
    {
      "epoch": 3.2325109834403514,
      "grad_norm": 3.861450433731079,
      "learning_rate": 2e-05,
      "loss": 0.1982,
      "step": 4784
    },
    {
      "epoch": 3.2331868874619802,
      "grad_norm": 3.1964991092681885,
      "learning_rate": 2e-05,
      "loss": 0.1184,
      "step": 4785
    },
    {
      "epoch": 3.2338627914836096,
      "grad_norm": 5.257052898406982,
      "learning_rate": 2e-05,
      "loss": 0.1675,
      "step": 4786
    },
    {
      "epoch": 3.2345386955052384,
      "grad_norm": 3.023545980453491,
      "learning_rate": 2e-05,
      "loss": 0.15,
      "step": 4787
    },
    {
      "epoch": 3.2352145995268673,
      "grad_norm": 3.4760313034057617,
      "learning_rate": 2e-05,
      "loss": 0.1257,
      "step": 4788
    },
    {
      "epoch": 3.235890503548496,
      "grad_norm": 2.581986665725708,
      "learning_rate": 2e-05,
      "loss": 0.1484,
      "step": 4789
    },
    {
      "epoch": 3.236566407570125,
      "grad_norm": 2.761653423309326,
      "learning_rate": 2e-05,
      "loss": 0.1374,
      "step": 4790
    },
    {
      "epoch": 3.237242311591754,
      "grad_norm": 3.408958911895752,
      "learning_rate": 2e-05,
      "loss": 0.1802,
      "step": 4791
    },
    {
      "epoch": 3.2379182156133828,
      "grad_norm": 4.963140487670898,
      "learning_rate": 2e-05,
      "loss": 0.144,
      "step": 4792
    },
    {
      "epoch": 3.2385941196350116,
      "grad_norm": 2.4329614639282227,
      "learning_rate": 2e-05,
      "loss": 0.2095,
      "step": 4793
    },
    {
      "epoch": 3.239270023656641,
      "grad_norm": 8.348962783813477,
      "learning_rate": 2e-05,
      "loss": 0.1652,
      "step": 4794
    },
    {
      "epoch": 3.23994592767827,
      "grad_norm": 7.179321765899658,
      "learning_rate": 2e-05,
      "loss": 0.19,
      "step": 4795
    },
    {
      "epoch": 3.2406218316998987,
      "grad_norm": 6.649045467376709,
      "learning_rate": 2e-05,
      "loss": 0.1952,
      "step": 4796
    },
    {
      "epoch": 3.2412977357215276,
      "grad_norm": 4.078974723815918,
      "learning_rate": 2e-05,
      "loss": 0.1857,
      "step": 4797
    },
    {
      "epoch": 3.2419736397431564,
      "grad_norm": 3.349693775177002,
      "learning_rate": 2e-05,
      "loss": 0.1664,
      "step": 4798
    },
    {
      "epoch": 3.2426495437647853,
      "grad_norm": 2.533881664276123,
      "learning_rate": 2e-05,
      "loss": 0.1417,
      "step": 4799
    },
    {
      "epoch": 3.243325447786414,
      "grad_norm": 3.7489962577819824,
      "learning_rate": 2e-05,
      "loss": 0.1787,
      "step": 4800
    },
    {
      "epoch": 3.243325447786414,
      "eval_accuracy": 0.7234797297297297,
      "eval_loss": 0.2382824867963791,
      "eval_runtime": 44.3559,
      "eval_samples_per_second": 533.864,
      "eval_steps_per_second": 16.683,
      "step": 4800
    },
    {
      "epoch": 3.244001351808043,
      "grad_norm": 3.361509323120117,
      "learning_rate": 2e-05,
      "loss": 0.2314,
      "step": 4801
    },
    {
      "epoch": 3.2446772558296724,
      "grad_norm": 2.1337804794311523,
      "learning_rate": 2e-05,
      "loss": 0.1493,
      "step": 4802
    },
    {
      "epoch": 3.2453531598513012,
      "grad_norm": 4.9066667556762695,
      "learning_rate": 2e-05,
      "loss": 0.1328,
      "step": 4803
    },
    {
      "epoch": 3.24602906387293,
      "grad_norm": 4.360440254211426,
      "learning_rate": 2e-05,
      "loss": 0.1679,
      "step": 4804
    },
    {
      "epoch": 3.246704967894559,
      "grad_norm": 2.691720962524414,
      "learning_rate": 2e-05,
      "loss": 0.148,
      "step": 4805
    },
    {
      "epoch": 3.247380871916188,
      "grad_norm": 3.2050833702087402,
      "learning_rate": 2e-05,
      "loss": 0.0939,
      "step": 4806
    },
    {
      "epoch": 3.2480567759378167,
      "grad_norm": 3.3336617946624756,
      "learning_rate": 2e-05,
      "loss": 0.1795,
      "step": 4807
    },
    {
      "epoch": 3.2487326799594456,
      "grad_norm": 2.519118070602417,
      "learning_rate": 2e-05,
      "loss": 0.1712,
      "step": 4808
    },
    {
      "epoch": 3.249408583981075,
      "grad_norm": 3.2413649559020996,
      "learning_rate": 2e-05,
      "loss": 0.1915,
      "step": 4809
    },
    {
      "epoch": 3.2500844880027038,
      "grad_norm": 6.206714153289795,
      "learning_rate": 2e-05,
      "loss": 0.1996,
      "step": 4810
    },
    {
      "epoch": 3.2507603920243326,
      "grad_norm": 3.0772311687469482,
      "learning_rate": 2e-05,
      "loss": 0.1552,
      "step": 4811
    },
    {
      "epoch": 3.2514362960459615,
      "grad_norm": 7.5213093757629395,
      "learning_rate": 2e-05,
      "loss": 0.1564,
      "step": 4812
    },
    {
      "epoch": 3.2521122000675904,
      "grad_norm": 2.657291889190674,
      "learning_rate": 2e-05,
      "loss": 0.1259,
      "step": 4813
    },
    {
      "epoch": 3.2527881040892193,
      "grad_norm": 3.201171636581421,
      "learning_rate": 2e-05,
      "loss": 0.1368,
      "step": 4814
    },
    {
      "epoch": 3.253464008110848,
      "grad_norm": 3.8674380779266357,
      "learning_rate": 2e-05,
      "loss": 0.1972,
      "step": 4815
    },
    {
      "epoch": 3.254139912132477,
      "grad_norm": 2.064941167831421,
      "learning_rate": 2e-05,
      "loss": 0.1435,
      "step": 4816
    },
    {
      "epoch": 3.254815816154106,
      "grad_norm": 2.9185173511505127,
      "learning_rate": 2e-05,
      "loss": 0.1336,
      "step": 4817
    },
    {
      "epoch": 3.255491720175735,
      "grad_norm": 3.97576642036438,
      "learning_rate": 2e-05,
      "loss": 0.0994,
      "step": 4818
    },
    {
      "epoch": 3.256167624197364,
      "grad_norm": 3.0177712440490723,
      "learning_rate": 2e-05,
      "loss": 0.1794,
      "step": 4819
    },
    {
      "epoch": 3.256843528218993,
      "grad_norm": 4.422573566436768,
      "learning_rate": 2e-05,
      "loss": 0.2027,
      "step": 4820
    },
    {
      "epoch": 3.257519432240622,
      "grad_norm": 6.406856536865234,
      "learning_rate": 2e-05,
      "loss": 0.1893,
      "step": 4821
    },
    {
      "epoch": 3.2581953362622507,
      "grad_norm": 9.281767845153809,
      "learning_rate": 2e-05,
      "loss": 0.2308,
      "step": 4822
    },
    {
      "epoch": 3.2588712402838795,
      "grad_norm": 6.603646278381348,
      "learning_rate": 2e-05,
      "loss": 0.1229,
      "step": 4823
    },
    {
      "epoch": 3.2595471443055084,
      "grad_norm": 4.929388523101807,
      "learning_rate": 2e-05,
      "loss": 0.1365,
      "step": 4824
    },
    {
      "epoch": 3.2602230483271377,
      "grad_norm": 2.473125696182251,
      "learning_rate": 2e-05,
      "loss": 0.1304,
      "step": 4825
    },
    {
      "epoch": 3.2608989523487666,
      "grad_norm": 5.214893817901611,
      "learning_rate": 2e-05,
      "loss": 0.1892,
      "step": 4826
    },
    {
      "epoch": 3.2615748563703955,
      "grad_norm": 3.355135917663574,
      "learning_rate": 2e-05,
      "loss": 0.1181,
      "step": 4827
    },
    {
      "epoch": 3.2622507603920243,
      "grad_norm": 7.87064790725708,
      "learning_rate": 2e-05,
      "loss": 0.1378,
      "step": 4828
    },
    {
      "epoch": 3.262926664413653,
      "grad_norm": 3.7231099605560303,
      "learning_rate": 2e-05,
      "loss": 0.1244,
      "step": 4829
    },
    {
      "epoch": 3.263602568435282,
      "grad_norm": 7.237569808959961,
      "learning_rate": 2e-05,
      "loss": 0.1466,
      "step": 4830
    },
    {
      "epoch": 3.264278472456911,
      "grad_norm": 4.339483737945557,
      "learning_rate": 2e-05,
      "loss": 0.1494,
      "step": 4831
    },
    {
      "epoch": 3.2649543764785403,
      "grad_norm": 6.254413604736328,
      "learning_rate": 2e-05,
      "loss": 0.2435,
      "step": 4832
    },
    {
      "epoch": 3.265630280500169,
      "grad_norm": 4.059866428375244,
      "learning_rate": 2e-05,
      "loss": 0.1655,
      "step": 4833
    },
    {
      "epoch": 3.266306184521798,
      "grad_norm": 8.491233825683594,
      "learning_rate": 2e-05,
      "loss": 0.2098,
      "step": 4834
    },
    {
      "epoch": 3.266982088543427,
      "grad_norm": 9.239238739013672,
      "learning_rate": 2e-05,
      "loss": 0.126,
      "step": 4835
    },
    {
      "epoch": 3.2676579925650557,
      "grad_norm": 2.9209702014923096,
      "learning_rate": 2e-05,
      "loss": 0.1974,
      "step": 4836
    },
    {
      "epoch": 3.2683338965866846,
      "grad_norm": 4.211286544799805,
      "learning_rate": 2e-05,
      "loss": 0.1797,
      "step": 4837
    },
    {
      "epoch": 3.2690098006083135,
      "grad_norm": 3.4821813106536865,
      "learning_rate": 2e-05,
      "loss": 0.2062,
      "step": 4838
    },
    {
      "epoch": 3.269685704629943,
      "grad_norm": 7.902518272399902,
      "learning_rate": 2e-05,
      "loss": 0.1433,
      "step": 4839
    },
    {
      "epoch": 3.2703616086515717,
      "grad_norm": 3.6441986560821533,
      "learning_rate": 2e-05,
      "loss": 0.1696,
      "step": 4840
    },
    {
      "epoch": 3.2710375126732005,
      "grad_norm": 4.243021011352539,
      "learning_rate": 2e-05,
      "loss": 0.16,
      "step": 4841
    },
    {
      "epoch": 3.2717134166948294,
      "grad_norm": 5.251236915588379,
      "learning_rate": 2e-05,
      "loss": 0.2077,
      "step": 4842
    },
    {
      "epoch": 3.2723893207164583,
      "grad_norm": 5.330360412597656,
      "learning_rate": 2e-05,
      "loss": 0.1852,
      "step": 4843
    },
    {
      "epoch": 3.273065224738087,
      "grad_norm": 4.896707534790039,
      "learning_rate": 2e-05,
      "loss": 0.1502,
      "step": 4844
    },
    {
      "epoch": 3.273741128759716,
      "grad_norm": 4.630577087402344,
      "learning_rate": 2e-05,
      "loss": 0.1702,
      "step": 4845
    },
    {
      "epoch": 3.274417032781345,
      "grad_norm": 3.2240443229675293,
      "learning_rate": 2e-05,
      "loss": 0.0962,
      "step": 4846
    },
    {
      "epoch": 3.2750929368029738,
      "grad_norm": 1.8542141914367676,
      "learning_rate": 2e-05,
      "loss": 0.1654,
      "step": 4847
    },
    {
      "epoch": 3.275768840824603,
      "grad_norm": 5.63132905960083,
      "learning_rate": 2e-05,
      "loss": 0.1755,
      "step": 4848
    },
    {
      "epoch": 3.276444744846232,
      "grad_norm": 2.4814116954803467,
      "learning_rate": 2e-05,
      "loss": 0.1362,
      "step": 4849
    },
    {
      "epoch": 3.277120648867861,
      "grad_norm": 2.8097493648529053,
      "learning_rate": 2e-05,
      "loss": 0.127,
      "step": 4850
    },
    {
      "epoch": 3.277120648867861,
      "eval_accuracy": 0.7146537162162162,
      "eval_loss": 0.23632875084877014,
      "eval_runtime": 43.6085,
      "eval_samples_per_second": 543.013,
      "eval_steps_per_second": 16.969,
      "step": 4850
    },
    {
      "epoch": 3.2777965528894897,
      "grad_norm": 3.5086276531219482,
      "learning_rate": 2e-05,
      "loss": 0.142,
      "step": 4851
    },
    {
      "epoch": 3.2784724569111185,
      "grad_norm": 2.2983808517456055,
      "learning_rate": 2e-05,
      "loss": 0.1907,
      "step": 4852
    },
    {
      "epoch": 3.2791483609327474,
      "grad_norm": 5.699300765991211,
      "learning_rate": 2e-05,
      "loss": 0.2168,
      "step": 4853
    },
    {
      "epoch": 3.2798242649543763,
      "grad_norm": 6.12921667098999,
      "learning_rate": 2e-05,
      "loss": 0.2056,
      "step": 4854
    },
    {
      "epoch": 3.2805001689760056,
      "grad_norm": 4.019576072692871,
      "learning_rate": 2e-05,
      "loss": 0.1443,
      "step": 4855
    },
    {
      "epoch": 3.2811760729976345,
      "grad_norm": 2.210205316543579,
      "learning_rate": 2e-05,
      "loss": 0.1286,
      "step": 4856
    },
    {
      "epoch": 3.2818519770192633,
      "grad_norm": 6.689401626586914,
      "learning_rate": 2e-05,
      "loss": 0.1346,
      "step": 4857
    },
    {
      "epoch": 3.282527881040892,
      "grad_norm": 5.464079856872559,
      "learning_rate": 2e-05,
      "loss": 0.1835,
      "step": 4858
    },
    {
      "epoch": 3.283203785062521,
      "grad_norm": 6.370802879333496,
      "learning_rate": 2e-05,
      "loss": 0.2035,
      "step": 4859
    },
    {
      "epoch": 3.28387968908415,
      "grad_norm": 7.210939407348633,
      "learning_rate": 2e-05,
      "loss": 0.1502,
      "step": 4860
    },
    {
      "epoch": 3.284555593105779,
      "grad_norm": 3.832209348678589,
      "learning_rate": 2e-05,
      "loss": 0.1356,
      "step": 4861
    },
    {
      "epoch": 3.285231497127408,
      "grad_norm": 2.889664649963379,
      "learning_rate": 2e-05,
      "loss": 0.1326,
      "step": 4862
    },
    {
      "epoch": 3.285907401149037,
      "grad_norm": 3.990245819091797,
      "learning_rate": 2e-05,
      "loss": 0.091,
      "step": 4863
    },
    {
      "epoch": 3.286583305170666,
      "grad_norm": 3.179633140563965,
      "learning_rate": 2e-05,
      "loss": 0.1971,
      "step": 4864
    },
    {
      "epoch": 3.2872592091922948,
      "grad_norm": 4.1279473304748535,
      "learning_rate": 2e-05,
      "loss": 0.1511,
      "step": 4865
    },
    {
      "epoch": 3.2879351132139236,
      "grad_norm": 6.4097771644592285,
      "learning_rate": 2e-05,
      "loss": 0.1435,
      "step": 4866
    },
    {
      "epoch": 3.2886110172355525,
      "grad_norm": 3.1767399311065674,
      "learning_rate": 2e-05,
      "loss": 0.1659,
      "step": 4867
    },
    {
      "epoch": 3.2892869212571814,
      "grad_norm": 3.702737331390381,
      "learning_rate": 2e-05,
      "loss": 0.1828,
      "step": 4868
    },
    {
      "epoch": 3.2899628252788102,
      "grad_norm": 3.965782642364502,
      "learning_rate": 2e-05,
      "loss": 0.1167,
      "step": 4869
    },
    {
      "epoch": 3.290638729300439,
      "grad_norm": 5.230532646179199,
      "learning_rate": 2e-05,
      "loss": 0.1655,
      "step": 4870
    },
    {
      "epoch": 3.2913146333220684,
      "grad_norm": 4.569159984588623,
      "learning_rate": 2e-05,
      "loss": 0.1472,
      "step": 4871
    },
    {
      "epoch": 3.2919905373436973,
      "grad_norm": 6.821109294891357,
      "learning_rate": 2e-05,
      "loss": 0.1812,
      "step": 4872
    },
    {
      "epoch": 3.292666441365326,
      "grad_norm": 9.264505386352539,
      "learning_rate": 2e-05,
      "loss": 0.1454,
      "step": 4873
    },
    {
      "epoch": 3.293342345386955,
      "grad_norm": 3.895969867706299,
      "learning_rate": 2e-05,
      "loss": 0.2522,
      "step": 4874
    },
    {
      "epoch": 3.294018249408584,
      "grad_norm": 3.718428373336792,
      "learning_rate": 2e-05,
      "loss": 0.1578,
      "step": 4875
    },
    {
      "epoch": 3.2946941534302128,
      "grad_norm": 7.059120178222656,
      "learning_rate": 2e-05,
      "loss": 0.2167,
      "step": 4876
    },
    {
      "epoch": 3.2953700574518416,
      "grad_norm": 6.658481121063232,
      "learning_rate": 2e-05,
      "loss": 0.1877,
      "step": 4877
    },
    {
      "epoch": 3.296045961473471,
      "grad_norm": 6.128091335296631,
      "learning_rate": 2e-05,
      "loss": 0.2568,
      "step": 4878
    },
    {
      "epoch": 3.2967218654951,
      "grad_norm": 7.972533226013184,
      "learning_rate": 2e-05,
      "loss": 0.1816,
      "step": 4879
    },
    {
      "epoch": 3.2973977695167287,
      "grad_norm": 2.755584478378296,
      "learning_rate": 2e-05,
      "loss": 0.2118,
      "step": 4880
    },
    {
      "epoch": 3.2980736735383576,
      "grad_norm": 5.254836559295654,
      "learning_rate": 2e-05,
      "loss": 0.1641,
      "step": 4881
    },
    {
      "epoch": 3.2987495775599864,
      "grad_norm": 2.3997690677642822,
      "learning_rate": 2e-05,
      "loss": 0.1459,
      "step": 4882
    },
    {
      "epoch": 3.2994254815816153,
      "grad_norm": 7.0695695877075195,
      "learning_rate": 2e-05,
      "loss": 0.1261,
      "step": 4883
    },
    {
      "epoch": 3.300101385603244,
      "grad_norm": 3.5530433654785156,
      "learning_rate": 2e-05,
      "loss": 0.1307,
      "step": 4884
    },
    {
      "epoch": 3.3007772896248735,
      "grad_norm": 2.424116611480713,
      "learning_rate": 2e-05,
      "loss": 0.1766,
      "step": 4885
    },
    {
      "epoch": 3.3014531936465024,
      "grad_norm": 4.0853657722473145,
      "learning_rate": 2e-05,
      "loss": 0.1686,
      "step": 4886
    },
    {
      "epoch": 3.3021290976681312,
      "grad_norm": 2.290283203125,
      "learning_rate": 2e-05,
      "loss": 0.0753,
      "step": 4887
    },
    {
      "epoch": 3.30280500168976,
      "grad_norm": 5.430014133453369,
      "learning_rate": 2e-05,
      "loss": 0.173,
      "step": 4888
    },
    {
      "epoch": 3.303480905711389,
      "grad_norm": 5.793393611907959,
      "learning_rate": 2e-05,
      "loss": 0.1895,
      "step": 4889
    },
    {
      "epoch": 3.304156809733018,
      "grad_norm": 5.341001510620117,
      "learning_rate": 2e-05,
      "loss": 0.1867,
      "step": 4890
    },
    {
      "epoch": 3.3048327137546467,
      "grad_norm": 4.678966999053955,
      "learning_rate": 2e-05,
      "loss": 0.193,
      "step": 4891
    },
    {
      "epoch": 3.3055086177762756,
      "grad_norm": 5.080401420593262,
      "learning_rate": 2e-05,
      "loss": 0.1815,
      "step": 4892
    },
    {
      "epoch": 3.3061845217979045,
      "grad_norm": 3.8499932289123535,
      "learning_rate": 2e-05,
      "loss": 0.1672,
      "step": 4893
    },
    {
      "epoch": 3.3068604258195338,
      "grad_norm": 7.8032965660095215,
      "learning_rate": 2e-05,
      "loss": 0.1248,
      "step": 4894
    },
    {
      "epoch": 3.3075363298411626,
      "grad_norm": 8.922210693359375,
      "learning_rate": 2e-05,
      "loss": 0.198,
      "step": 4895
    },
    {
      "epoch": 3.3082122338627915,
      "grad_norm": 4.385180950164795,
      "learning_rate": 2e-05,
      "loss": 0.174,
      "step": 4896
    },
    {
      "epoch": 3.3088881378844204,
      "grad_norm": 5.383145332336426,
      "learning_rate": 2e-05,
      "loss": 0.1673,
      "step": 4897
    },
    {
      "epoch": 3.3095640419060492,
      "grad_norm": 2.448655128479004,
      "learning_rate": 2e-05,
      "loss": 0.2254,
      "step": 4898
    },
    {
      "epoch": 3.310239945927678,
      "grad_norm": 11.599344253540039,
      "learning_rate": 2e-05,
      "loss": 0.1643,
      "step": 4899
    },
    {
      "epoch": 3.310915849949307,
      "grad_norm": 5.706470489501953,
      "learning_rate": 2e-05,
      "loss": 0.1094,
      "step": 4900
    },
    {
      "epoch": 3.310915849949307,
      "eval_accuracy": 0.7377111486486486,
      "eval_loss": 0.2576860189437866,
      "eval_runtime": 44.2818,
      "eval_samples_per_second": 534.757,
      "eval_steps_per_second": 16.711,
      "step": 4900
    },
    {
      "epoch": 3.3115917539709363,
      "grad_norm": 1.5244061946868896,
      "learning_rate": 2e-05,
      "loss": 0.1168,
      "step": 4901
    },
    {
      "epoch": 3.312267657992565,
      "grad_norm": 7.588718414306641,
      "learning_rate": 2e-05,
      "loss": 0.1638,
      "step": 4902
    },
    {
      "epoch": 3.312943562014194,
      "grad_norm": 4.517731189727783,
      "learning_rate": 2e-05,
      "loss": 0.1194,
      "step": 4903
    },
    {
      "epoch": 3.313619466035823,
      "grad_norm": 4.96787166595459,
      "learning_rate": 2e-05,
      "loss": 0.1896,
      "step": 4904
    },
    {
      "epoch": 3.314295370057452,
      "grad_norm": 4.087845802307129,
      "learning_rate": 2e-05,
      "loss": 0.1673,
      "step": 4905
    },
    {
      "epoch": 3.3149712740790807,
      "grad_norm": 2.783810615539551,
      "learning_rate": 2e-05,
      "loss": 0.1138,
      "step": 4906
    },
    {
      "epoch": 3.3156471781007095,
      "grad_norm": 3.579270839691162,
      "learning_rate": 2e-05,
      "loss": 0.1576,
      "step": 4907
    },
    {
      "epoch": 3.316323082122339,
      "grad_norm": 2.8417069911956787,
      "learning_rate": 2e-05,
      "loss": 0.1397,
      "step": 4908
    },
    {
      "epoch": 3.3169989861439677,
      "grad_norm": 4.995669364929199,
      "learning_rate": 2e-05,
      "loss": 0.1885,
      "step": 4909
    },
    {
      "epoch": 3.3176748901655966,
      "grad_norm": 6.976171493530273,
      "learning_rate": 2e-05,
      "loss": 0.1368,
      "step": 4910
    },
    {
      "epoch": 3.3183507941872254,
      "grad_norm": 8.395267486572266,
      "learning_rate": 2e-05,
      "loss": 0.2062,
      "step": 4911
    },
    {
      "epoch": 3.3190266982088543,
      "grad_norm": 4.412419319152832,
      "learning_rate": 2e-05,
      "loss": 0.1162,
      "step": 4912
    },
    {
      "epoch": 3.319702602230483,
      "grad_norm": 4.263030052185059,
      "learning_rate": 2e-05,
      "loss": 0.1413,
      "step": 4913
    },
    {
      "epoch": 3.320378506252112,
      "grad_norm": 2.5392212867736816,
      "learning_rate": 2e-05,
      "loss": 0.1541,
      "step": 4914
    },
    {
      "epoch": 3.3210544102737414,
      "grad_norm": 7.715529441833496,
      "learning_rate": 2e-05,
      "loss": 0.1759,
      "step": 4915
    },
    {
      "epoch": 3.3217303142953702,
      "grad_norm": 3.849393606185913,
      "learning_rate": 2e-05,
      "loss": 0.1439,
      "step": 4916
    },
    {
      "epoch": 3.322406218316999,
      "grad_norm": 3.121464729309082,
      "learning_rate": 2e-05,
      "loss": 0.1558,
      "step": 4917
    },
    {
      "epoch": 3.323082122338628,
      "grad_norm": 9.053426742553711,
      "learning_rate": 2e-05,
      "loss": 0.1558,
      "step": 4918
    },
    {
      "epoch": 3.323758026360257,
      "grad_norm": 4.215582847595215,
      "learning_rate": 2e-05,
      "loss": 0.1675,
      "step": 4919
    },
    {
      "epoch": 3.3244339303818857,
      "grad_norm": 4.000611782073975,
      "learning_rate": 2e-05,
      "loss": 0.1367,
      "step": 4920
    },
    {
      "epoch": 3.3251098344035146,
      "grad_norm": 4.958078861236572,
      "learning_rate": 2e-05,
      "loss": 0.186,
      "step": 4921
    },
    {
      "epoch": 3.3257857384251435,
      "grad_norm": 7.38039493560791,
      "learning_rate": 2e-05,
      "loss": 0.1488,
      "step": 4922
    },
    {
      "epoch": 3.3264616424467723,
      "grad_norm": 4.980011940002441,
      "learning_rate": 2e-05,
      "loss": 0.1478,
      "step": 4923
    },
    {
      "epoch": 3.3271375464684017,
      "grad_norm": 4.644046783447266,
      "learning_rate": 2e-05,
      "loss": 0.1681,
      "step": 4924
    },
    {
      "epoch": 3.3278134504900305,
      "grad_norm": 5.417985439300537,
      "learning_rate": 2e-05,
      "loss": 0.1401,
      "step": 4925
    },
    {
      "epoch": 3.3284893545116594,
      "grad_norm": 3.7351577281951904,
      "learning_rate": 2e-05,
      "loss": 0.127,
      "step": 4926
    },
    {
      "epoch": 3.3291652585332883,
      "grad_norm": 7.799246311187744,
      "learning_rate": 2e-05,
      "loss": 0.1616,
      "step": 4927
    },
    {
      "epoch": 3.329841162554917,
      "grad_norm": 5.142565727233887,
      "learning_rate": 2e-05,
      "loss": 0.1162,
      "step": 4928
    },
    {
      "epoch": 3.330517066576546,
      "grad_norm": 2.314727544784546,
      "learning_rate": 2e-05,
      "loss": 0.1233,
      "step": 4929
    },
    {
      "epoch": 3.331192970598175,
      "grad_norm": 14.430474281311035,
      "learning_rate": 2e-05,
      "loss": 0.2149,
      "step": 4930
    },
    {
      "epoch": 3.331868874619804,
      "grad_norm": 8.857850074768066,
      "learning_rate": 2e-05,
      "loss": 0.1867,
      "step": 4931
    },
    {
      "epoch": 3.332544778641433,
      "grad_norm": 5.68023157119751,
      "learning_rate": 2e-05,
      "loss": 0.144,
      "step": 4932
    },
    {
      "epoch": 3.333220682663062,
      "grad_norm": 2.8469924926757812,
      "learning_rate": 2e-05,
      "loss": 0.15,
      "step": 4933
    },
    {
      "epoch": 3.333896586684691,
      "grad_norm": 6.605942726135254,
      "learning_rate": 2e-05,
      "loss": 0.0987,
      "step": 4934
    },
    {
      "epoch": 3.3345724907063197,
      "grad_norm": 2.340982675552368,
      "learning_rate": 2e-05,
      "loss": 0.1096,
      "step": 4935
    },
    {
      "epoch": 3.3352483947279485,
      "grad_norm": 5.692279815673828,
      "learning_rate": 2e-05,
      "loss": 0.1405,
      "step": 4936
    },
    {
      "epoch": 3.3359242987495774,
      "grad_norm": 6.808485984802246,
      "learning_rate": 2e-05,
      "loss": 0.1123,
      "step": 4937
    },
    {
      "epoch": 3.3366002027712067,
      "grad_norm": 3.4109690189361572,
      "learning_rate": 2e-05,
      "loss": 0.1548,
      "step": 4938
    },
    {
      "epoch": 3.3372761067928356,
      "grad_norm": 3.8633193969726562,
      "learning_rate": 2e-05,
      "loss": 0.1274,
      "step": 4939
    },
    {
      "epoch": 3.3379520108144645,
      "grad_norm": 7.326228141784668,
      "learning_rate": 2e-05,
      "loss": 0.1152,
      "step": 4940
    },
    {
      "epoch": 3.3386279148360933,
      "grad_norm": 4.920999050140381,
      "learning_rate": 2e-05,
      "loss": 0.1194,
      "step": 4941
    },
    {
      "epoch": 3.339303818857722,
      "grad_norm": 7.590628623962402,
      "learning_rate": 2e-05,
      "loss": 0.1855,
      "step": 4942
    },
    {
      "epoch": 3.339979722879351,
      "grad_norm": 5.851323127746582,
      "learning_rate": 2e-05,
      "loss": 0.2008,
      "step": 4943
    },
    {
      "epoch": 3.34065562690098,
      "grad_norm": 8.65151309967041,
      "learning_rate": 2e-05,
      "loss": 0.2602,
      "step": 4944
    },
    {
      "epoch": 3.341331530922609,
      "grad_norm": 6.0336127281188965,
      "learning_rate": 2e-05,
      "loss": 0.2805,
      "step": 4945
    },
    {
      "epoch": 3.3420074349442377,
      "grad_norm": 3.8404197692871094,
      "learning_rate": 2e-05,
      "loss": 0.1513,
      "step": 4946
    },
    {
      "epoch": 3.342683338965867,
      "grad_norm": 4.299644947052002,
      "learning_rate": 2e-05,
      "loss": 0.194,
      "step": 4947
    },
    {
      "epoch": 3.343359242987496,
      "grad_norm": 2.5860164165496826,
      "learning_rate": 2e-05,
      "loss": 0.1526,
      "step": 4948
    },
    {
      "epoch": 3.3440351470091247,
      "grad_norm": 2.7203187942504883,
      "learning_rate": 2e-05,
      "loss": 0.1518,
      "step": 4949
    },
    {
      "epoch": 3.3447110510307536,
      "grad_norm": 5.718898773193359,
      "learning_rate": 2e-05,
      "loss": 0.2178,
      "step": 4950
    },
    {
      "epoch": 3.3447110510307536,
      "eval_accuracy": 0.7282094594594595,
      "eval_loss": 0.2313668131828308,
      "eval_runtime": 44.2812,
      "eval_samples_per_second": 534.764,
      "eval_steps_per_second": 16.711,
      "step": 4950
    },
    {
      "epoch": 3.3453869550523825,
      "grad_norm": 7.051306247711182,
      "learning_rate": 2e-05,
      "loss": 0.1575,
      "step": 4951
    },
    {
      "epoch": 3.3460628590740114,
      "grad_norm": 5.688163757324219,
      "learning_rate": 2e-05,
      "loss": 0.211,
      "step": 4952
    },
    {
      "epoch": 3.3467387630956402,
      "grad_norm": 4.391408920288086,
      "learning_rate": 2e-05,
      "loss": 0.1304,
      "step": 4953
    },
    {
      "epoch": 3.3474146671172695,
      "grad_norm": 5.358067989349365,
      "learning_rate": 2e-05,
      "loss": 0.2028,
      "step": 4954
    },
    {
      "epoch": 3.3480905711388984,
      "grad_norm": 3.19608998298645,
      "learning_rate": 2e-05,
      "loss": 0.1713,
      "step": 4955
    },
    {
      "epoch": 3.3487664751605273,
      "grad_norm": 3.5568041801452637,
      "learning_rate": 2e-05,
      "loss": 0.1501,
      "step": 4956
    },
    {
      "epoch": 3.349442379182156,
      "grad_norm": 3.118767023086548,
      "learning_rate": 2e-05,
      "loss": 0.1436,
      "step": 4957
    },
    {
      "epoch": 3.350118283203785,
      "grad_norm": 4.471249103546143,
      "learning_rate": 2e-05,
      "loss": 0.1705,
      "step": 4958
    },
    {
      "epoch": 3.350794187225414,
      "grad_norm": 6.469296455383301,
      "learning_rate": 2e-05,
      "loss": 0.1832,
      "step": 4959
    },
    {
      "epoch": 3.3514700912470428,
      "grad_norm": 3.1338045597076416,
      "learning_rate": 2e-05,
      "loss": 0.2417,
      "step": 4960
    },
    {
      "epoch": 3.352145995268672,
      "grad_norm": 6.346375465393066,
      "learning_rate": 2e-05,
      "loss": 0.172,
      "step": 4961
    },
    {
      "epoch": 3.352821899290301,
      "grad_norm": 2.5622992515563965,
      "learning_rate": 2e-05,
      "loss": 0.1527,
      "step": 4962
    },
    {
      "epoch": 3.35349780331193,
      "grad_norm": 2.3699233531951904,
      "learning_rate": 2e-05,
      "loss": 0.2102,
      "step": 4963
    },
    {
      "epoch": 3.3541737073335587,
      "grad_norm": 2.7008163928985596,
      "learning_rate": 2e-05,
      "loss": 0.157,
      "step": 4964
    },
    {
      "epoch": 3.3548496113551876,
      "grad_norm": 4.009364128112793,
      "learning_rate": 2e-05,
      "loss": 0.1326,
      "step": 4965
    },
    {
      "epoch": 3.3555255153768164,
      "grad_norm": 2.050665855407715,
      "learning_rate": 2e-05,
      "loss": 0.1328,
      "step": 4966
    },
    {
      "epoch": 3.3562014193984453,
      "grad_norm": 2.734391212463379,
      "learning_rate": 2e-05,
      "loss": 0.1586,
      "step": 4967
    },
    {
      "epoch": 3.356877323420074,
      "grad_norm": 3.20218825340271,
      "learning_rate": 2e-05,
      "loss": 0.1906,
      "step": 4968
    },
    {
      "epoch": 3.357553227441703,
      "grad_norm": 7.725566387176514,
      "learning_rate": 2e-05,
      "loss": 0.1756,
      "step": 4969
    },
    {
      "epoch": 3.3582291314633324,
      "grad_norm": 2.9443888664245605,
      "learning_rate": 2e-05,
      "loss": 0.1697,
      "step": 4970
    },
    {
      "epoch": 3.358905035484961,
      "grad_norm": 2.179471731185913,
      "learning_rate": 2e-05,
      "loss": 0.1431,
      "step": 4971
    },
    {
      "epoch": 3.35958093950659,
      "grad_norm": 3.748501777648926,
      "learning_rate": 2e-05,
      "loss": 0.1484,
      "step": 4972
    },
    {
      "epoch": 3.360256843528219,
      "grad_norm": 5.785654544830322,
      "learning_rate": 2e-05,
      "loss": 0.2376,
      "step": 4973
    },
    {
      "epoch": 3.360932747549848,
      "grad_norm": 12.219841003417969,
      "learning_rate": 2e-05,
      "loss": 0.2532,
      "step": 4974
    },
    {
      "epoch": 3.3616086515714767,
      "grad_norm": 7.633777618408203,
      "learning_rate": 2e-05,
      "loss": 0.1612,
      "step": 4975
    },
    {
      "epoch": 3.3622845555931056,
      "grad_norm": 7.773499965667725,
      "learning_rate": 2e-05,
      "loss": 0.1002,
      "step": 4976
    },
    {
      "epoch": 3.362960459614735,
      "grad_norm": 3.433001756668091,
      "learning_rate": 2e-05,
      "loss": 0.1802,
      "step": 4977
    },
    {
      "epoch": 3.3636363636363638,
      "grad_norm": 2.675271511077881,
      "learning_rate": 2e-05,
      "loss": 0.1834,
      "step": 4978
    },
    {
      "epoch": 3.3643122676579926,
      "grad_norm": 4.117567539215088,
      "learning_rate": 2e-05,
      "loss": 0.1733,
      "step": 4979
    },
    {
      "epoch": 3.3649881716796215,
      "grad_norm": 5.581009864807129,
      "learning_rate": 2e-05,
      "loss": 0.1514,
      "step": 4980
    },
    {
      "epoch": 3.3656640757012504,
      "grad_norm": 8.819046020507812,
      "learning_rate": 2e-05,
      "loss": 0.1964,
      "step": 4981
    },
    {
      "epoch": 3.3663399797228792,
      "grad_norm": 10.313074111938477,
      "learning_rate": 2e-05,
      "loss": 0.1763,
      "step": 4982
    },
    {
      "epoch": 3.367015883744508,
      "grad_norm": 3.814495801925659,
      "learning_rate": 2e-05,
      "loss": 0.2007,
      "step": 4983
    },
    {
      "epoch": 3.3676917877661374,
      "grad_norm": 5.322821140289307,
      "learning_rate": 2e-05,
      "loss": 0.1672,
      "step": 4984
    },
    {
      "epoch": 3.3683676917877663,
      "grad_norm": 2.7262415885925293,
      "learning_rate": 2e-05,
      "loss": 0.1329,
      "step": 4985
    },
    {
      "epoch": 3.369043595809395,
      "grad_norm": 7.237770080566406,
      "learning_rate": 2e-05,
      "loss": 0.1561,
      "step": 4986
    },
    {
      "epoch": 3.369719499831024,
      "grad_norm": 5.121662616729736,
      "learning_rate": 2e-05,
      "loss": 0.1835,
      "step": 4987
    },
    {
      "epoch": 3.370395403852653,
      "grad_norm": 4.618360996246338,
      "learning_rate": 2e-05,
      "loss": 0.1125,
      "step": 4988
    },
    {
      "epoch": 3.3710713078742818,
      "grad_norm": 3.1756560802459717,
      "learning_rate": 2e-05,
      "loss": 0.1498,
      "step": 4989
    },
    {
      "epoch": 3.3717472118959106,
      "grad_norm": 3.362417459487915,
      "learning_rate": 2e-05,
      "loss": 0.2116,
      "step": 4990
    },
    {
      "epoch": 3.3724231159175395,
      "grad_norm": 3.6826257705688477,
      "learning_rate": 2e-05,
      "loss": 0.1911,
      "step": 4991
    },
    {
      "epoch": 3.373099019939169,
      "grad_norm": 4.744004726409912,
      "learning_rate": 2e-05,
      "loss": 0.1454,
      "step": 4992
    },
    {
      "epoch": 3.3737749239607977,
      "grad_norm": 2.6151323318481445,
      "learning_rate": 2e-05,
      "loss": 0.1482,
      "step": 4993
    },
    {
      "epoch": 3.3744508279824266,
      "grad_norm": 3.2839443683624268,
      "learning_rate": 2e-05,
      "loss": 0.1845,
      "step": 4994
    },
    {
      "epoch": 3.3751267320040554,
      "grad_norm": 2.6427102088928223,
      "learning_rate": 2e-05,
      "loss": 0.1358,
      "step": 4995
    },
    {
      "epoch": 3.3758026360256843,
      "grad_norm": 2.8550784587860107,
      "learning_rate": 2e-05,
      "loss": 0.1519,
      "step": 4996
    },
    {
      "epoch": 3.376478540047313,
      "grad_norm": 2.641817331314087,
      "learning_rate": 2e-05,
      "loss": 0.174,
      "step": 4997
    },
    {
      "epoch": 3.377154444068942,
      "grad_norm": 2.4064018726348877,
      "learning_rate": 2e-05,
      "loss": 0.1706,
      "step": 4998
    },
    {
      "epoch": 3.377830348090571,
      "grad_norm": 2.909121036529541,
      "learning_rate": 2e-05,
      "loss": 0.1371,
      "step": 4999
    },
    {
      "epoch": 3.3785062521122002,
      "grad_norm": 3.5771775245666504,
      "learning_rate": 2e-05,
      "loss": 0.166,
      "step": 5000
    },
    {
      "epoch": 3.3785062521122002,
      "eval_accuracy": 0.706418918918919,
      "eval_loss": 0.24980895221233368,
      "eval_runtime": 44.8253,
      "eval_samples_per_second": 528.274,
      "eval_steps_per_second": 16.509,
      "step": 5000
    },
    {
      "epoch": 3.379182156133829,
      "grad_norm": 3.6229074001312256,
      "learning_rate": 2e-05,
      "loss": 0.1913,
      "step": 5001
    },
    {
      "epoch": 3.379858060155458,
      "grad_norm": 4.0349602699279785,
      "learning_rate": 2e-05,
      "loss": 0.1536,
      "step": 5002
    },
    {
      "epoch": 3.380533964177087,
      "grad_norm": 2.8903415203094482,
      "learning_rate": 2e-05,
      "loss": 0.1595,
      "step": 5003
    },
    {
      "epoch": 3.3812098681987157,
      "grad_norm": 8.140375137329102,
      "learning_rate": 2e-05,
      "loss": 0.1731,
      "step": 5004
    },
    {
      "epoch": 3.3818857722203446,
      "grad_norm": 3.439587116241455,
      "learning_rate": 2e-05,
      "loss": 0.1322,
      "step": 5005
    },
    {
      "epoch": 3.3825616762419735,
      "grad_norm": 6.424832344055176,
      "learning_rate": 2e-05,
      "loss": 0.1671,
      "step": 5006
    },
    {
      "epoch": 3.3832375802636028,
      "grad_norm": 2.786506414413452,
      "learning_rate": 2e-05,
      "loss": 0.1654,
      "step": 5007
    },
    {
      "epoch": 3.3839134842852316,
      "grad_norm": 7.056422233581543,
      "learning_rate": 2e-05,
      "loss": 0.2153,
      "step": 5008
    },
    {
      "epoch": 3.3845893883068605,
      "grad_norm": 4.390847682952881,
      "learning_rate": 2e-05,
      "loss": 0.1369,
      "step": 5009
    },
    {
      "epoch": 3.3852652923284894,
      "grad_norm": 5.905240535736084,
      "learning_rate": 2e-05,
      "loss": 0.1566,
      "step": 5010
    },
    {
      "epoch": 3.3859411963501183,
      "grad_norm": 3.4437899589538574,
      "learning_rate": 2e-05,
      "loss": 0.1636,
      "step": 5011
    },
    {
      "epoch": 3.386617100371747,
      "grad_norm": 4.137582302093506,
      "learning_rate": 2e-05,
      "loss": 0.1215,
      "step": 5012
    },
    {
      "epoch": 3.387293004393376,
      "grad_norm": 3.0143003463745117,
      "learning_rate": 2e-05,
      "loss": 0.1812,
      "step": 5013
    },
    {
      "epoch": 3.3879689084150053,
      "grad_norm": 2.4623775482177734,
      "learning_rate": 2e-05,
      "loss": 0.1545,
      "step": 5014
    },
    {
      "epoch": 3.388644812436634,
      "grad_norm": 5.225471496582031,
      "learning_rate": 2e-05,
      "loss": 0.1702,
      "step": 5015
    },
    {
      "epoch": 3.389320716458263,
      "grad_norm": 2.5978410243988037,
      "learning_rate": 2e-05,
      "loss": 0.1472,
      "step": 5016
    },
    {
      "epoch": 3.389996620479892,
      "grad_norm": 2.2469725608825684,
      "learning_rate": 2e-05,
      "loss": 0.1498,
      "step": 5017
    },
    {
      "epoch": 3.390672524501521,
      "grad_norm": 5.287197113037109,
      "learning_rate": 2e-05,
      "loss": 0.1248,
      "step": 5018
    },
    {
      "epoch": 3.3913484285231497,
      "grad_norm": 5.3906660079956055,
      "learning_rate": 2e-05,
      "loss": 0.2163,
      "step": 5019
    },
    {
      "epoch": 3.3920243325447785,
      "grad_norm": 5.256158828735352,
      "learning_rate": 2e-05,
      "loss": 0.1706,
      "step": 5020
    },
    {
      "epoch": 3.3927002365664074,
      "grad_norm": 2.8985018730163574,
      "learning_rate": 2e-05,
      "loss": 0.1669,
      "step": 5021
    },
    {
      "epoch": 3.3933761405880363,
      "grad_norm": 3.19985294342041,
      "learning_rate": 2e-05,
      "loss": 0.1441,
      "step": 5022
    },
    {
      "epoch": 3.3940520446096656,
      "grad_norm": 2.4715962409973145,
      "learning_rate": 2e-05,
      "loss": 0.1301,
      "step": 5023
    },
    {
      "epoch": 3.3947279486312945,
      "grad_norm": 3.181875228881836,
      "learning_rate": 2e-05,
      "loss": 0.2047,
      "step": 5024
    },
    {
      "epoch": 3.3954038526529233,
      "grad_norm": 8.566337585449219,
      "learning_rate": 2e-05,
      "loss": 0.1443,
      "step": 5025
    },
    {
      "epoch": 3.396079756674552,
      "grad_norm": 2.8206448554992676,
      "learning_rate": 2e-05,
      "loss": 0.1535,
      "step": 5026
    },
    {
      "epoch": 3.396755660696181,
      "grad_norm": 6.564590930938721,
      "learning_rate": 2e-05,
      "loss": 0.1807,
      "step": 5027
    },
    {
      "epoch": 3.39743156471781,
      "grad_norm": 6.643883228302002,
      "learning_rate": 2e-05,
      "loss": 0.154,
      "step": 5028
    },
    {
      "epoch": 3.398107468739439,
      "grad_norm": 3.2476861476898193,
      "learning_rate": 2e-05,
      "loss": 0.1168,
      "step": 5029
    },
    {
      "epoch": 3.398783372761068,
      "grad_norm": 5.111917495727539,
      "learning_rate": 2e-05,
      "loss": 0.1439,
      "step": 5030
    },
    {
      "epoch": 3.399459276782697,
      "grad_norm": 6.829251766204834,
      "learning_rate": 2e-05,
      "loss": 0.2013,
      "step": 5031
    },
    {
      "epoch": 3.400135180804326,
      "grad_norm": 4.377017974853516,
      "learning_rate": 2e-05,
      "loss": 0.1916,
      "step": 5032
    },
    {
      "epoch": 3.4008110848259547,
      "grad_norm": 3.0534586906433105,
      "learning_rate": 2e-05,
      "loss": 0.1885,
      "step": 5033
    },
    {
      "epoch": 3.4014869888475836,
      "grad_norm": 6.59726095199585,
      "learning_rate": 2e-05,
      "loss": 0.1749,
      "step": 5034
    },
    {
      "epoch": 3.4021628928692125,
      "grad_norm": 4.4116950035095215,
      "learning_rate": 2e-05,
      "loss": 0.2201,
      "step": 5035
    },
    {
      "epoch": 3.4028387968908413,
      "grad_norm": 4.856156349182129,
      "learning_rate": 2e-05,
      "loss": 0.1958,
      "step": 5036
    },
    {
      "epoch": 3.4035147009124707,
      "grad_norm": 6.616055488586426,
      "learning_rate": 2e-05,
      "loss": 0.1865,
      "step": 5037
    },
    {
      "epoch": 3.4041906049340995,
      "grad_norm": 2.928694248199463,
      "learning_rate": 2e-05,
      "loss": 0.1325,
      "step": 5038
    },
    {
      "epoch": 3.4048665089557284,
      "grad_norm": 10.697937965393066,
      "learning_rate": 2e-05,
      "loss": 0.1855,
      "step": 5039
    },
    {
      "epoch": 3.4055424129773573,
      "grad_norm": 7.813797950744629,
      "learning_rate": 2e-05,
      "loss": 0.1435,
      "step": 5040
    },
    {
      "epoch": 3.406218316998986,
      "grad_norm": 4.564904689788818,
      "learning_rate": 2e-05,
      "loss": 0.1749,
      "step": 5041
    },
    {
      "epoch": 3.406894221020615,
      "grad_norm": 3.534733533859253,
      "learning_rate": 2e-05,
      "loss": 0.1518,
      "step": 5042
    },
    {
      "epoch": 3.407570125042244,
      "grad_norm": 9.089198112487793,
      "learning_rate": 2e-05,
      "loss": 0.1455,
      "step": 5043
    },
    {
      "epoch": 3.4082460290638728,
      "grad_norm": 13.026521682739258,
      "learning_rate": 2e-05,
      "loss": 0.2066,
      "step": 5044
    },
    {
      "epoch": 3.4089219330855016,
      "grad_norm": 16.114953994750977,
      "learning_rate": 2e-05,
      "loss": 0.1579,
      "step": 5045
    },
    {
      "epoch": 3.409597837107131,
      "grad_norm": 8.299908638000488,
      "learning_rate": 2e-05,
      "loss": 0.1475,
      "step": 5046
    },
    {
      "epoch": 3.41027374112876,
      "grad_norm": 2.881021022796631,
      "learning_rate": 2e-05,
      "loss": 0.1447,
      "step": 5047
    },
    {
      "epoch": 3.4109496451503887,
      "grad_norm": 3.7141165733337402,
      "learning_rate": 2e-05,
      "loss": 0.1695,
      "step": 5048
    },
    {
      "epoch": 3.4116255491720175,
      "grad_norm": 4.230886459350586,
      "learning_rate": 2e-05,
      "loss": 0.1446,
      "step": 5049
    },
    {
      "epoch": 3.4123014531936464,
      "grad_norm": 4.702147960662842,
      "learning_rate": 2e-05,
      "loss": 0.2108,
      "step": 5050
    },
    {
      "epoch": 3.4123014531936464,
      "eval_accuracy": 0.6581503378378378,
      "eval_loss": 0.28131964802742004,
      "eval_runtime": 43.2974,
      "eval_samples_per_second": 546.915,
      "eval_steps_per_second": 17.091,
      "step": 5050
    },
    {
      "epoch": 3.4129773572152753,
      "grad_norm": 10.89028549194336,
      "learning_rate": 2e-05,
      "loss": 0.196,
      "step": 5051
    },
    {
      "epoch": 3.413653261236904,
      "grad_norm": 11.106606483459473,
      "learning_rate": 2e-05,
      "loss": 0.1825,
      "step": 5052
    },
    {
      "epoch": 3.4143291652585335,
      "grad_norm": 4.489277362823486,
      "learning_rate": 2e-05,
      "loss": 0.1648,
      "step": 5053
    },
    {
      "epoch": 3.4150050692801623,
      "grad_norm": 4.186264991760254,
      "learning_rate": 2e-05,
      "loss": 0.1579,
      "step": 5054
    },
    {
      "epoch": 3.415680973301791,
      "grad_norm": 4.366604328155518,
      "learning_rate": 2e-05,
      "loss": 0.1543,
      "step": 5055
    },
    {
      "epoch": 3.41635687732342,
      "grad_norm": 8.980498313903809,
      "learning_rate": 2e-05,
      "loss": 0.1764,
      "step": 5056
    },
    {
      "epoch": 3.417032781345049,
      "grad_norm": 4.829426288604736,
      "learning_rate": 2e-05,
      "loss": 0.1496,
      "step": 5057
    },
    {
      "epoch": 3.417708685366678,
      "grad_norm": 3.8017871379852295,
      "learning_rate": 2e-05,
      "loss": 0.1344,
      "step": 5058
    },
    {
      "epoch": 3.4183845893883067,
      "grad_norm": 10.617935180664062,
      "learning_rate": 2e-05,
      "loss": 0.1533,
      "step": 5059
    },
    {
      "epoch": 3.419060493409936,
      "grad_norm": 6.9605302810668945,
      "learning_rate": 2e-05,
      "loss": 0.1629,
      "step": 5060
    },
    {
      "epoch": 3.419736397431565,
      "grad_norm": 2.293869972229004,
      "learning_rate": 2e-05,
      "loss": 0.1178,
      "step": 5061
    },
    {
      "epoch": 3.4204123014531937,
      "grad_norm": 3.343555450439453,
      "learning_rate": 2e-05,
      "loss": 0.1794,
      "step": 5062
    },
    {
      "epoch": 3.4210882054748226,
      "grad_norm": 3.5351474285125732,
      "learning_rate": 2e-05,
      "loss": 0.2192,
      "step": 5063
    },
    {
      "epoch": 3.4217641094964515,
      "grad_norm": 9.736130714416504,
      "learning_rate": 2e-05,
      "loss": 0.2116,
      "step": 5064
    },
    {
      "epoch": 3.4224400135180804,
      "grad_norm": 5.499204635620117,
      "learning_rate": 2e-05,
      "loss": 0.2239,
      "step": 5065
    },
    {
      "epoch": 3.4231159175397092,
      "grad_norm": 4.829761981964111,
      "learning_rate": 2e-05,
      "loss": 0.1617,
      "step": 5066
    },
    {
      "epoch": 3.423791821561338,
      "grad_norm": 4.786393165588379,
      "learning_rate": 2e-05,
      "loss": 0.1793,
      "step": 5067
    },
    {
      "epoch": 3.424467725582967,
      "grad_norm": 6.039772987365723,
      "learning_rate": 2e-05,
      "loss": 0.2184,
      "step": 5068
    },
    {
      "epoch": 3.4251436296045963,
      "grad_norm": 2.571697235107422,
      "learning_rate": 2e-05,
      "loss": 0.1281,
      "step": 5069
    },
    {
      "epoch": 3.425819533626225,
      "grad_norm": 3.595581293106079,
      "learning_rate": 2e-05,
      "loss": 0.2146,
      "step": 5070
    },
    {
      "epoch": 3.426495437647854,
      "grad_norm": 5.74290657043457,
      "learning_rate": 2e-05,
      "loss": 0.1351,
      "step": 5071
    },
    {
      "epoch": 3.427171341669483,
      "grad_norm": 4.833536624908447,
      "learning_rate": 2e-05,
      "loss": 0.2113,
      "step": 5072
    },
    {
      "epoch": 3.4278472456911118,
      "grad_norm": 8.434996604919434,
      "learning_rate": 2e-05,
      "loss": 0.142,
      "step": 5073
    },
    {
      "epoch": 3.4285231497127406,
      "grad_norm": 3.4305107593536377,
      "learning_rate": 2e-05,
      "loss": 0.1334,
      "step": 5074
    },
    {
      "epoch": 3.4291990537343695,
      "grad_norm": 2.211566925048828,
      "learning_rate": 2e-05,
      "loss": 0.1951,
      "step": 5075
    },
    {
      "epoch": 3.429874957755999,
      "grad_norm": 4.040179252624512,
      "learning_rate": 2e-05,
      "loss": 0.1438,
      "step": 5076
    },
    {
      "epoch": 3.4305508617776277,
      "grad_norm": 5.825564861297607,
      "learning_rate": 2e-05,
      "loss": 0.2011,
      "step": 5077
    },
    {
      "epoch": 3.4312267657992566,
      "grad_norm": 6.562072277069092,
      "learning_rate": 2e-05,
      "loss": 0.1924,
      "step": 5078
    },
    {
      "epoch": 3.4319026698208854,
      "grad_norm": 3.351198196411133,
      "learning_rate": 2e-05,
      "loss": 0.1644,
      "step": 5079
    },
    {
      "epoch": 3.4325785738425143,
      "grad_norm": 3.278870105743408,
      "learning_rate": 2e-05,
      "loss": 0.1916,
      "step": 5080
    },
    {
      "epoch": 3.433254477864143,
      "grad_norm": 8.423399925231934,
      "learning_rate": 2e-05,
      "loss": 0.1803,
      "step": 5081
    },
    {
      "epoch": 3.433930381885772,
      "grad_norm": 7.329078674316406,
      "learning_rate": 2e-05,
      "loss": 0.2431,
      "step": 5082
    },
    {
      "epoch": 3.4346062859074014,
      "grad_norm": 9.012709617614746,
      "learning_rate": 2e-05,
      "loss": 0.2121,
      "step": 5083
    },
    {
      "epoch": 3.4352821899290302,
      "grad_norm": 4.040212154388428,
      "learning_rate": 2e-05,
      "loss": 0.1413,
      "step": 5084
    },
    {
      "epoch": 3.435958093950659,
      "grad_norm": 4.20123291015625,
      "learning_rate": 2e-05,
      "loss": 0.2315,
      "step": 5085
    },
    {
      "epoch": 3.436633997972288,
      "grad_norm": 10.293111801147461,
      "learning_rate": 2e-05,
      "loss": 0.1901,
      "step": 5086
    },
    {
      "epoch": 3.437309901993917,
      "grad_norm": 12.014921188354492,
      "learning_rate": 2e-05,
      "loss": 0.2408,
      "step": 5087
    },
    {
      "epoch": 3.4379858060155457,
      "grad_norm": 6.612142086029053,
      "learning_rate": 2e-05,
      "loss": 0.1163,
      "step": 5088
    },
    {
      "epoch": 3.4386617100371746,
      "grad_norm": 2.365570306777954,
      "learning_rate": 2e-05,
      "loss": 0.1416,
      "step": 5089
    },
    {
      "epoch": 3.439337614058804,
      "grad_norm": 2.7324423789978027,
      "learning_rate": 2e-05,
      "loss": 0.1826,
      "step": 5090
    },
    {
      "epoch": 3.4400135180804328,
      "grad_norm": 2.79211688041687,
      "learning_rate": 2e-05,
      "loss": 0.1273,
      "step": 5091
    },
    {
      "epoch": 3.4406894221020616,
      "grad_norm": 5.950274467468262,
      "learning_rate": 2e-05,
      "loss": 0.1419,
      "step": 5092
    },
    {
      "epoch": 3.4413653261236905,
      "grad_norm": 2.143707752227783,
      "learning_rate": 2e-05,
      "loss": 0.2135,
      "step": 5093
    },
    {
      "epoch": 3.4420412301453194,
      "grad_norm": 9.862853050231934,
      "learning_rate": 2e-05,
      "loss": 0.1705,
      "step": 5094
    },
    {
      "epoch": 3.4427171341669482,
      "grad_norm": 4.802044868469238,
      "learning_rate": 2e-05,
      "loss": 0.21,
      "step": 5095
    },
    {
      "epoch": 3.443393038188577,
      "grad_norm": 5.572488784790039,
      "learning_rate": 2e-05,
      "loss": 0.2033,
      "step": 5096
    },
    {
      "epoch": 3.444068942210206,
      "grad_norm": 2.8475546836853027,
      "learning_rate": 2e-05,
      "loss": 0.2396,
      "step": 5097
    },
    {
      "epoch": 3.444744846231835,
      "grad_norm": 9.759305000305176,
      "learning_rate": 2e-05,
      "loss": 0.1516,
      "step": 5098
    },
    {
      "epoch": 3.445420750253464,
      "grad_norm": 1.8114852905273438,
      "learning_rate": 2e-05,
      "loss": 0.1851,
      "step": 5099
    },
    {
      "epoch": 3.446096654275093,
      "grad_norm": 3.795611619949341,
      "learning_rate": 2e-05,
      "loss": 0.1868,
      "step": 5100
    },
    {
      "epoch": 3.446096654275093,
      "eval_accuracy": 0.730616554054054,
      "eval_loss": 0.23986326158046722,
      "eval_runtime": 44.6403,
      "eval_samples_per_second": 530.463,
      "eval_steps_per_second": 16.577,
      "step": 5100
    },
    {
      "epoch": 3.446772558296722,
      "grad_norm": 4.225226402282715,
      "learning_rate": 2e-05,
      "loss": 0.1205,
      "step": 5101
    },
    {
      "epoch": 3.447448462318351,
      "grad_norm": 2.6175055503845215,
      "learning_rate": 2e-05,
      "loss": 0.126,
      "step": 5102
    },
    {
      "epoch": 3.4481243663399797,
      "grad_norm": 4.986509799957275,
      "learning_rate": 2e-05,
      "loss": 0.1812,
      "step": 5103
    },
    {
      "epoch": 3.4488002703616085,
      "grad_norm": 6.048786640167236,
      "learning_rate": 2e-05,
      "loss": 0.1619,
      "step": 5104
    },
    {
      "epoch": 3.4494761743832374,
      "grad_norm": 2.3079826831817627,
      "learning_rate": 2e-05,
      "loss": 0.1849,
      "step": 5105
    },
    {
      "epoch": 3.4501520784048667,
      "grad_norm": 2.450388193130493,
      "learning_rate": 2e-05,
      "loss": 0.1407,
      "step": 5106
    },
    {
      "epoch": 3.4508279824264956,
      "grad_norm": 3.371323585510254,
      "learning_rate": 2e-05,
      "loss": 0.1361,
      "step": 5107
    },
    {
      "epoch": 3.4515038864481244,
      "grad_norm": 2.3989319801330566,
      "learning_rate": 2e-05,
      "loss": 0.1815,
      "step": 5108
    },
    {
      "epoch": 3.4521797904697533,
      "grad_norm": 3.376514434814453,
      "learning_rate": 2e-05,
      "loss": 0.1703,
      "step": 5109
    },
    {
      "epoch": 3.452855694491382,
      "grad_norm": 2.39506196975708,
      "learning_rate": 2e-05,
      "loss": 0.1757,
      "step": 5110
    },
    {
      "epoch": 3.453531598513011,
      "grad_norm": 2.9343209266662598,
      "learning_rate": 2e-05,
      "loss": 0.1932,
      "step": 5111
    },
    {
      "epoch": 3.45420750253464,
      "grad_norm": 4.237273216247559,
      "learning_rate": 2e-05,
      "loss": 0.1738,
      "step": 5112
    },
    {
      "epoch": 3.4548834065562692,
      "grad_norm": 4.0722551345825195,
      "learning_rate": 2e-05,
      "loss": 0.1398,
      "step": 5113
    },
    {
      "epoch": 3.455559310577898,
      "grad_norm": 2.5163426399230957,
      "learning_rate": 2e-05,
      "loss": 0.161,
      "step": 5114
    },
    {
      "epoch": 3.456235214599527,
      "grad_norm": 2.963005781173706,
      "learning_rate": 2e-05,
      "loss": 0.1431,
      "step": 5115
    },
    {
      "epoch": 3.456911118621156,
      "grad_norm": 4.653918743133545,
      "learning_rate": 2e-05,
      "loss": 0.1523,
      "step": 5116
    },
    {
      "epoch": 3.4575870226427847,
      "grad_norm": 4.094677925109863,
      "learning_rate": 2e-05,
      "loss": 0.1818,
      "step": 5117
    },
    {
      "epoch": 3.4582629266644136,
      "grad_norm": 2.758991003036499,
      "learning_rate": 2e-05,
      "loss": 0.099,
      "step": 5118
    },
    {
      "epoch": 3.4589388306860425,
      "grad_norm": 2.748034715652466,
      "learning_rate": 2e-05,
      "loss": 0.2229,
      "step": 5119
    },
    {
      "epoch": 3.4596147347076713,
      "grad_norm": 3.695469617843628,
      "learning_rate": 2e-05,
      "loss": 0.1288,
      "step": 5120
    },
    {
      "epoch": 3.4602906387293,
      "grad_norm": 7.137129306793213,
      "learning_rate": 2e-05,
      "loss": 0.169,
      "step": 5121
    },
    {
      "epoch": 3.4609665427509295,
      "grad_norm": 3.20766282081604,
      "learning_rate": 2e-05,
      "loss": 0.1211,
      "step": 5122
    },
    {
      "epoch": 3.4616424467725584,
      "grad_norm": 4.429872035980225,
      "learning_rate": 2e-05,
      "loss": 0.1496,
      "step": 5123
    },
    {
      "epoch": 3.4623183507941873,
      "grad_norm": 3.611034631729126,
      "learning_rate": 2e-05,
      "loss": 0.2232,
      "step": 5124
    },
    {
      "epoch": 3.462994254815816,
      "grad_norm": 11.183202743530273,
      "learning_rate": 2e-05,
      "loss": 0.1862,
      "step": 5125
    },
    {
      "epoch": 3.463670158837445,
      "grad_norm": 4.250548362731934,
      "learning_rate": 2e-05,
      "loss": 0.1429,
      "step": 5126
    },
    {
      "epoch": 3.464346062859074,
      "grad_norm": 5.758704662322998,
      "learning_rate": 2e-05,
      "loss": 0.1833,
      "step": 5127
    },
    {
      "epoch": 3.4650219668807027,
      "grad_norm": 4.073819637298584,
      "learning_rate": 2e-05,
      "loss": 0.1403,
      "step": 5128
    },
    {
      "epoch": 3.465697870902332,
      "grad_norm": 3.998004198074341,
      "learning_rate": 2e-05,
      "loss": 0.2923,
      "step": 5129
    },
    {
      "epoch": 3.466373774923961,
      "grad_norm": 7.948611259460449,
      "learning_rate": 2e-05,
      "loss": 0.1215,
      "step": 5130
    },
    {
      "epoch": 3.46704967894559,
      "grad_norm": 4.125249862670898,
      "learning_rate": 2e-05,
      "loss": 0.236,
      "step": 5131
    },
    {
      "epoch": 3.4677255829672187,
      "grad_norm": 8.315879821777344,
      "learning_rate": 2e-05,
      "loss": 0.2315,
      "step": 5132
    },
    {
      "epoch": 3.4684014869888475,
      "grad_norm": 5.897200107574463,
      "learning_rate": 2e-05,
      "loss": 0.1579,
      "step": 5133
    },
    {
      "epoch": 3.4690773910104764,
      "grad_norm": 2.4445559978485107,
      "learning_rate": 2e-05,
      "loss": 0.1235,
      "step": 5134
    },
    {
      "epoch": 3.4697532950321053,
      "grad_norm": 7.698793888092041,
      "learning_rate": 2e-05,
      "loss": 0.2503,
      "step": 5135
    },
    {
      "epoch": 3.4704291990537346,
      "grad_norm": 6.219704627990723,
      "learning_rate": 2e-05,
      "loss": 0.2459,
      "step": 5136
    },
    {
      "epoch": 3.4711051030753635,
      "grad_norm": 4.676562786102295,
      "learning_rate": 2e-05,
      "loss": 0.156,
      "step": 5137
    },
    {
      "epoch": 3.4717810070969923,
      "grad_norm": 3.951385974884033,
      "learning_rate": 2e-05,
      "loss": 0.1753,
      "step": 5138
    },
    {
      "epoch": 3.472456911118621,
      "grad_norm": 2.242922782897949,
      "learning_rate": 2e-05,
      "loss": 0.1536,
      "step": 5139
    },
    {
      "epoch": 3.47313281514025,
      "grad_norm": 5.794315814971924,
      "learning_rate": 2e-05,
      "loss": 0.1372,
      "step": 5140
    },
    {
      "epoch": 3.473808719161879,
      "grad_norm": 2.2659056186676025,
      "learning_rate": 2e-05,
      "loss": 0.1923,
      "step": 5141
    },
    {
      "epoch": 3.474484623183508,
      "grad_norm": 8.366515159606934,
      "learning_rate": 2e-05,
      "loss": 0.1472,
      "step": 5142
    },
    {
      "epoch": 3.4751605272051367,
      "grad_norm": 2.205219268798828,
      "learning_rate": 2e-05,
      "loss": 0.1505,
      "step": 5143
    },
    {
      "epoch": 3.4758364312267656,
      "grad_norm": 8.422141075134277,
      "learning_rate": 2e-05,
      "loss": 0.1799,
      "step": 5144
    },
    {
      "epoch": 3.476512335248395,
      "grad_norm": 2.2391269207000732,
      "learning_rate": 2e-05,
      "loss": 0.1912,
      "step": 5145
    },
    {
      "epoch": 3.4771882392700237,
      "grad_norm": 2.1893322467803955,
      "learning_rate": 2e-05,
      "loss": 0.1405,
      "step": 5146
    },
    {
      "epoch": 3.4778641432916526,
      "grad_norm": 3.060938596725464,
      "learning_rate": 2e-05,
      "loss": 0.1689,
      "step": 5147
    },
    {
      "epoch": 3.4785400473132815,
      "grad_norm": 2.283292531967163,
      "learning_rate": 2e-05,
      "loss": 0.1818,
      "step": 5148
    },
    {
      "epoch": 3.4792159513349104,
      "grad_norm": 3.470184326171875,
      "learning_rate": 2e-05,
      "loss": 0.1817,
      "step": 5149
    },
    {
      "epoch": 3.479891855356539,
      "grad_norm": 5.593623161315918,
      "learning_rate": 2e-05,
      "loss": 0.1506,
      "step": 5150
    },
    {
      "epoch": 3.479891855356539,
      "eval_accuracy": 0.7124155405405406,
      "eval_loss": 0.23403747379779816,
      "eval_runtime": 44.6312,
      "eval_samples_per_second": 530.571,
      "eval_steps_per_second": 16.58,
      "step": 5150
    },
    {
      "epoch": 3.480567759378168,
      "grad_norm": 5.860019683837891,
      "learning_rate": 2e-05,
      "loss": 0.1984,
      "step": 5151
    },
    {
      "epoch": 3.4812436633997974,
      "grad_norm": 3.4932098388671875,
      "learning_rate": 2e-05,
      "loss": 0.1187,
      "step": 5152
    },
    {
      "epoch": 3.4819195674214263,
      "grad_norm": 4.159267902374268,
      "learning_rate": 2e-05,
      "loss": 0.1342,
      "step": 5153
    },
    {
      "epoch": 3.482595471443055,
      "grad_norm": 2.927067518234253,
      "learning_rate": 2e-05,
      "loss": 0.1728,
      "step": 5154
    },
    {
      "epoch": 3.483271375464684,
      "grad_norm": 4.779613018035889,
      "learning_rate": 2e-05,
      "loss": 0.1383,
      "step": 5155
    },
    {
      "epoch": 3.483947279486313,
      "grad_norm": 2.0518951416015625,
      "learning_rate": 2e-05,
      "loss": 0.1001,
      "step": 5156
    },
    {
      "epoch": 3.4846231835079418,
      "grad_norm": 5.684352397918701,
      "learning_rate": 2e-05,
      "loss": 0.2409,
      "step": 5157
    },
    {
      "epoch": 3.4852990875295706,
      "grad_norm": 5.46646785736084,
      "learning_rate": 2e-05,
      "loss": 0.1841,
      "step": 5158
    },
    {
      "epoch": 3.4859749915512,
      "grad_norm": 3.8388266563415527,
      "learning_rate": 2e-05,
      "loss": 0.2374,
      "step": 5159
    },
    {
      "epoch": 3.486650895572829,
      "grad_norm": 5.141568660736084,
      "learning_rate": 2e-05,
      "loss": 0.1522,
      "step": 5160
    },
    {
      "epoch": 3.4873267995944577,
      "grad_norm": 2.687657117843628,
      "learning_rate": 2e-05,
      "loss": 0.1315,
      "step": 5161
    },
    {
      "epoch": 3.4880027036160866,
      "grad_norm": 3.793118953704834,
      "learning_rate": 2e-05,
      "loss": 0.1919,
      "step": 5162
    },
    {
      "epoch": 3.4886786076377154,
      "grad_norm": 4.163122653961182,
      "learning_rate": 2e-05,
      "loss": 0.1778,
      "step": 5163
    },
    {
      "epoch": 3.4893545116593443,
      "grad_norm": 5.555816650390625,
      "learning_rate": 2e-05,
      "loss": 0.1368,
      "step": 5164
    },
    {
      "epoch": 3.490030415680973,
      "grad_norm": 3.970592975616455,
      "learning_rate": 2e-05,
      "loss": 0.1538,
      "step": 5165
    },
    {
      "epoch": 3.4907063197026025,
      "grad_norm": 2.931356906890869,
      "learning_rate": 2e-05,
      "loss": 0.2032,
      "step": 5166
    },
    {
      "epoch": 3.4913822237242313,
      "grad_norm": 2.3739471435546875,
      "learning_rate": 2e-05,
      "loss": 0.1371,
      "step": 5167
    },
    {
      "epoch": 3.49205812774586,
      "grad_norm": 5.370301246643066,
      "learning_rate": 2e-05,
      "loss": 0.1898,
      "step": 5168
    },
    {
      "epoch": 3.492734031767489,
      "grad_norm": 2.4394569396972656,
      "learning_rate": 2e-05,
      "loss": 0.1639,
      "step": 5169
    },
    {
      "epoch": 3.493409935789118,
      "grad_norm": 4.433583736419678,
      "learning_rate": 2e-05,
      "loss": 0.1724,
      "step": 5170
    },
    {
      "epoch": 3.494085839810747,
      "grad_norm": 3.9833762645721436,
      "learning_rate": 2e-05,
      "loss": 0.1766,
      "step": 5171
    },
    {
      "epoch": 3.4947617438323757,
      "grad_norm": 3.5234153270721436,
      "learning_rate": 2e-05,
      "loss": 0.2215,
      "step": 5172
    },
    {
      "epoch": 3.4954376478540046,
      "grad_norm": 4.513398170471191,
      "learning_rate": 2e-05,
      "loss": 0.1343,
      "step": 5173
    },
    {
      "epoch": 3.4961135518756334,
      "grad_norm": 2.1810379028320312,
      "learning_rate": 2e-05,
      "loss": 0.1775,
      "step": 5174
    },
    {
      "epoch": 3.4967894558972628,
      "grad_norm": 5.1967692375183105,
      "learning_rate": 2e-05,
      "loss": 0.1809,
      "step": 5175
    },
    {
      "epoch": 3.4974653599188916,
      "grad_norm": 3.421245574951172,
      "learning_rate": 2e-05,
      "loss": 0.1937,
      "step": 5176
    },
    {
      "epoch": 3.4981412639405205,
      "grad_norm": 3.8574280738830566,
      "learning_rate": 2e-05,
      "loss": 0.1808,
      "step": 5177
    },
    {
      "epoch": 3.4988171679621494,
      "grad_norm": 3.9923365116119385,
      "learning_rate": 2e-05,
      "loss": 0.2032,
      "step": 5178
    },
    {
      "epoch": 3.4994930719837782,
      "grad_norm": 7.59091854095459,
      "learning_rate": 2e-05,
      "loss": 0.1407,
      "step": 5179
    },
    {
      "epoch": 3.500168976005407,
      "grad_norm": 5.519775867462158,
      "learning_rate": 2e-05,
      "loss": 0.2084,
      "step": 5180
    },
    {
      "epoch": 3.500844880027036,
      "grad_norm": 4.269474983215332,
      "learning_rate": 2e-05,
      "loss": 0.2076,
      "step": 5181
    },
    {
      "epoch": 3.5015207840486653,
      "grad_norm": 4.973507881164551,
      "learning_rate": 2e-05,
      "loss": 0.1773,
      "step": 5182
    },
    {
      "epoch": 3.502196688070294,
      "grad_norm": 4.469421863555908,
      "learning_rate": 2e-05,
      "loss": 0.2169,
      "step": 5183
    },
    {
      "epoch": 3.502872592091923,
      "grad_norm": 6.516108989715576,
      "learning_rate": 2e-05,
      "loss": 0.2,
      "step": 5184
    },
    {
      "epoch": 3.503548496113552,
      "grad_norm": 5.176827907562256,
      "learning_rate": 2e-05,
      "loss": 0.2061,
      "step": 5185
    },
    {
      "epoch": 3.5042244001351808,
      "grad_norm": 2.765911817550659,
      "learning_rate": 2e-05,
      "loss": 0.1487,
      "step": 5186
    },
    {
      "epoch": 3.5049003041568096,
      "grad_norm": 2.1882710456848145,
      "learning_rate": 2e-05,
      "loss": 0.1908,
      "step": 5187
    },
    {
      "epoch": 3.5055762081784385,
      "grad_norm": 2.0281569957733154,
      "learning_rate": 2e-05,
      "loss": 0.1897,
      "step": 5188
    },
    {
      "epoch": 3.506252112200068,
      "grad_norm": 5.109952449798584,
      "learning_rate": 2e-05,
      "loss": 0.1495,
      "step": 5189
    },
    {
      "epoch": 3.5069280162216963,
      "grad_norm": 3.3071200847625732,
      "learning_rate": 2e-05,
      "loss": 0.1606,
      "step": 5190
    },
    {
      "epoch": 3.5076039202433256,
      "grad_norm": 2.3027021884918213,
      "learning_rate": 2e-05,
      "loss": 0.1451,
      "step": 5191
    },
    {
      "epoch": 3.5082798242649544,
      "grad_norm": 3.505110025405884,
      "learning_rate": 2e-05,
      "loss": 0.1718,
      "step": 5192
    },
    {
      "epoch": 3.5089557282865833,
      "grad_norm": 6.433834552764893,
      "learning_rate": 2e-05,
      "loss": 0.1466,
      "step": 5193
    },
    {
      "epoch": 3.509631632308212,
      "grad_norm": 4.647143363952637,
      "learning_rate": 2e-05,
      "loss": 0.1291,
      "step": 5194
    },
    {
      "epoch": 3.510307536329841,
      "grad_norm": 4.372971534729004,
      "learning_rate": 2e-05,
      "loss": 0.2009,
      "step": 5195
    },
    {
      "epoch": 3.5109834403514704,
      "grad_norm": 3.180213451385498,
      "learning_rate": 2e-05,
      "loss": 0.1789,
      "step": 5196
    },
    {
      "epoch": 3.511659344373099,
      "grad_norm": 2.2352278232574463,
      "learning_rate": 2e-05,
      "loss": 0.163,
      "step": 5197
    },
    {
      "epoch": 3.512335248394728,
      "grad_norm": 2.4491376876831055,
      "learning_rate": 2e-05,
      "loss": 0.1323,
      "step": 5198
    },
    {
      "epoch": 3.513011152416357,
      "grad_norm": 2.701141357421875,
      "learning_rate": 2e-05,
      "loss": 0.1991,
      "step": 5199
    },
    {
      "epoch": 3.513687056437986,
      "grad_norm": 5.075976371765137,
      "learning_rate": 2e-05,
      "loss": 0.1714,
      "step": 5200
    },
    {
      "epoch": 3.513687056437986,
      "eval_accuracy": 0.6849239864864864,
      "eval_loss": 0.24099090695381165,
      "eval_runtime": 53.1582,
      "eval_samples_per_second": 445.463,
      "eval_steps_per_second": 13.921,
      "step": 5200
    },
    {
      "epoch": 3.5143629604596147,
      "grad_norm": 4.456894874572754,
      "learning_rate": 2e-05,
      "loss": 0.1993,
      "step": 5201
    },
    {
      "epoch": 3.5150388644812436,
      "grad_norm": 3.4141106605529785,
      "learning_rate": 2e-05,
      "loss": 0.1553,
      "step": 5202
    },
    {
      "epoch": 3.5157147685028725,
      "grad_norm": 5.0775957107543945,
      "learning_rate": 2e-05,
      "loss": 0.1504,
      "step": 5203
    },
    {
      "epoch": 3.5163906725245013,
      "grad_norm": 5.237554550170898,
      "learning_rate": 2e-05,
      "loss": 0.2173,
      "step": 5204
    },
    {
      "epoch": 3.5170665765461306,
      "grad_norm": 7.383360385894775,
      "learning_rate": 2e-05,
      "loss": 0.1835,
      "step": 5205
    },
    {
      "epoch": 3.5177424805677595,
      "grad_norm": 3.9435741901397705,
      "learning_rate": 2e-05,
      "loss": 0.2145,
      "step": 5206
    },
    {
      "epoch": 3.5184183845893884,
      "grad_norm": 8.087639808654785,
      "learning_rate": 2e-05,
      "loss": 0.199,
      "step": 5207
    },
    {
      "epoch": 3.5190942886110173,
      "grad_norm": 5.021901607513428,
      "learning_rate": 2e-05,
      "loss": 0.1381,
      "step": 5208
    },
    {
      "epoch": 3.519770192632646,
      "grad_norm": 6.310368061065674,
      "learning_rate": 2e-05,
      "loss": 0.2057,
      "step": 5209
    },
    {
      "epoch": 3.520446096654275,
      "grad_norm": 5.203811168670654,
      "learning_rate": 2e-05,
      "loss": 0.1954,
      "step": 5210
    },
    {
      "epoch": 3.521122000675904,
      "grad_norm": 3.1383564472198486,
      "learning_rate": 2e-05,
      "loss": 0.1387,
      "step": 5211
    },
    {
      "epoch": 3.521797904697533,
      "grad_norm": 5.328390598297119,
      "learning_rate": 2e-05,
      "loss": 0.1757,
      "step": 5212
    },
    {
      "epoch": 3.5224738087191616,
      "grad_norm": 5.774386405944824,
      "learning_rate": 2e-05,
      "loss": 0.2449,
      "step": 5213
    },
    {
      "epoch": 3.523149712740791,
      "grad_norm": 12.565771102905273,
      "learning_rate": 2e-05,
      "loss": 0.1962,
      "step": 5214
    },
    {
      "epoch": 3.52382561676242,
      "grad_norm": 2.7098042964935303,
      "learning_rate": 2e-05,
      "loss": 0.1269,
      "step": 5215
    },
    {
      "epoch": 3.5245015207840487,
      "grad_norm": 2.1062633991241455,
      "learning_rate": 2e-05,
      "loss": 0.1238,
      "step": 5216
    },
    {
      "epoch": 3.5251774248056775,
      "grad_norm": 4.176229476928711,
      "learning_rate": 2e-05,
      "loss": 0.175,
      "step": 5217
    },
    {
      "epoch": 3.5258533288273064,
      "grad_norm": 6.378198146820068,
      "learning_rate": 2e-05,
      "loss": 0.1526,
      "step": 5218
    },
    {
      "epoch": 3.5265292328489357,
      "grad_norm": 5.028080463409424,
      "learning_rate": 2e-05,
      "loss": 0.1378,
      "step": 5219
    },
    {
      "epoch": 3.527205136870564,
      "grad_norm": 2.3860085010528564,
      "learning_rate": 2e-05,
      "loss": 0.115,
      "step": 5220
    },
    {
      "epoch": 3.5278810408921935,
      "grad_norm": 2.6432676315307617,
      "learning_rate": 2e-05,
      "loss": 0.1662,
      "step": 5221
    },
    {
      "epoch": 3.5285569449138223,
      "grad_norm": 1.9227782487869263,
      "learning_rate": 2e-05,
      "loss": 0.1272,
      "step": 5222
    },
    {
      "epoch": 3.529232848935451,
      "grad_norm": 6.567150115966797,
      "learning_rate": 2e-05,
      "loss": 0.175,
      "step": 5223
    },
    {
      "epoch": 3.52990875295708,
      "grad_norm": 2.843987226486206,
      "learning_rate": 2e-05,
      "loss": 0.2038,
      "step": 5224
    },
    {
      "epoch": 3.530584656978709,
      "grad_norm": 4.37555456161499,
      "learning_rate": 2e-05,
      "loss": 0.1508,
      "step": 5225
    },
    {
      "epoch": 3.531260561000338,
      "grad_norm": 3.333091974258423,
      "learning_rate": 2e-05,
      "loss": 0.1882,
      "step": 5226
    },
    {
      "epoch": 3.5319364650219667,
      "grad_norm": 5.545870780944824,
      "learning_rate": 2e-05,
      "loss": 0.1252,
      "step": 5227
    },
    {
      "epoch": 3.532612369043596,
      "grad_norm": 3.5540010929107666,
      "learning_rate": 2e-05,
      "loss": 0.2109,
      "step": 5228
    },
    {
      "epoch": 3.533288273065225,
      "grad_norm": 3.2949554920196533,
      "learning_rate": 2e-05,
      "loss": 0.1089,
      "step": 5229
    },
    {
      "epoch": 3.5339641770868537,
      "grad_norm": 4.744385719299316,
      "learning_rate": 2e-05,
      "loss": 0.1575,
      "step": 5230
    },
    {
      "epoch": 3.5346400811084826,
      "grad_norm": 5.796182155609131,
      "learning_rate": 2e-05,
      "loss": 0.1744,
      "step": 5231
    },
    {
      "epoch": 3.5353159851301115,
      "grad_norm": 5.031807899475098,
      "learning_rate": 2e-05,
      "loss": 0.1625,
      "step": 5232
    },
    {
      "epoch": 3.5359918891517403,
      "grad_norm": 4.33152961730957,
      "learning_rate": 2e-05,
      "loss": 0.1428,
      "step": 5233
    },
    {
      "epoch": 3.536667793173369,
      "grad_norm": 2.316042184829712,
      "learning_rate": 2e-05,
      "loss": 0.164,
      "step": 5234
    },
    {
      "epoch": 3.5373436971949985,
      "grad_norm": 4.205415725708008,
      "learning_rate": 2e-05,
      "loss": 0.1774,
      "step": 5235
    },
    {
      "epoch": 3.5380196012166274,
      "grad_norm": 6.821885585784912,
      "learning_rate": 2e-05,
      "loss": 0.136,
      "step": 5236
    },
    {
      "epoch": 3.5386955052382563,
      "grad_norm": 6.051962375640869,
      "learning_rate": 2e-05,
      "loss": 0.1978,
      "step": 5237
    },
    {
      "epoch": 3.539371409259885,
      "grad_norm": 3.805877447128296,
      "learning_rate": 2e-05,
      "loss": 0.1793,
      "step": 5238
    },
    {
      "epoch": 3.540047313281514,
      "grad_norm": 5.813775539398193,
      "learning_rate": 2e-05,
      "loss": 0.2128,
      "step": 5239
    },
    {
      "epoch": 3.540723217303143,
      "grad_norm": 4.4111151695251465,
      "learning_rate": 2e-05,
      "loss": 0.1067,
      "step": 5240
    },
    {
      "epoch": 3.5413991213247717,
      "grad_norm": 2.126260757446289,
      "learning_rate": 2e-05,
      "loss": 0.1691,
      "step": 5241
    },
    {
      "epoch": 3.542075025346401,
      "grad_norm": 3.5302932262420654,
      "learning_rate": 2e-05,
      "loss": 0.1835,
      "step": 5242
    },
    {
      "epoch": 3.5427509293680295,
      "grad_norm": 2.5885934829711914,
      "learning_rate": 2e-05,
      "loss": 0.2239,
      "step": 5243
    },
    {
      "epoch": 3.543426833389659,
      "grad_norm": 3.116459369659424,
      "learning_rate": 2e-05,
      "loss": 0.1882,
      "step": 5244
    },
    {
      "epoch": 3.5441027374112877,
      "grad_norm": 2.337223768234253,
      "learning_rate": 2e-05,
      "loss": 0.1181,
      "step": 5245
    },
    {
      "epoch": 3.5447786414329165,
      "grad_norm": 2.698560953140259,
      "learning_rate": 2e-05,
      "loss": 0.1938,
      "step": 5246
    },
    {
      "epoch": 3.5454545454545454,
      "grad_norm": 3.5496666431427,
      "learning_rate": 2e-05,
      "loss": 0.1387,
      "step": 5247
    },
    {
      "epoch": 3.5461304494761743,
      "grad_norm": 2.650345802307129,
      "learning_rate": 2e-05,
      "loss": 0.1559,
      "step": 5248
    },
    {
      "epoch": 3.546806353497803,
      "grad_norm": 6.456459999084473,
      "learning_rate": 2e-05,
      "loss": 0.2078,
      "step": 5249
    },
    {
      "epoch": 3.547482257519432,
      "grad_norm": 4.842037200927734,
      "learning_rate": 2e-05,
      "loss": 0.1845,
      "step": 5250
    },
    {
      "epoch": 3.547482257519432,
      "eval_accuracy": 0.7328969594594594,
      "eval_loss": 0.2414552867412567,
      "eval_runtime": 49.5513,
      "eval_samples_per_second": 477.889,
      "eval_steps_per_second": 14.934,
      "step": 5250
    },
    {
      "epoch": 3.5481581615410613,
      "grad_norm": 2.505640983581543,
      "learning_rate": 2e-05,
      "loss": 0.1626,
      "step": 5251
    },
    {
      "epoch": 3.54883406556269,
      "grad_norm": 4.098002910614014,
      "learning_rate": 2e-05,
      "loss": 0.1437,
      "step": 5252
    },
    {
      "epoch": 3.549509969584319,
      "grad_norm": 3.5513150691986084,
      "learning_rate": 2e-05,
      "loss": 0.1922,
      "step": 5253
    },
    {
      "epoch": 3.550185873605948,
      "grad_norm": 2.2661311626434326,
      "learning_rate": 2e-05,
      "loss": 0.1611,
      "step": 5254
    },
    {
      "epoch": 3.550861777627577,
      "grad_norm": 3.1176517009735107,
      "learning_rate": 2e-05,
      "loss": 0.2016,
      "step": 5255
    },
    {
      "epoch": 3.5515376816492057,
      "grad_norm": 9.087153434753418,
      "learning_rate": 2e-05,
      "loss": 0.205,
      "step": 5256
    },
    {
      "epoch": 3.5522135856708346,
      "grad_norm": 8.055842399597168,
      "learning_rate": 2e-05,
      "loss": 0.1915,
      "step": 5257
    },
    {
      "epoch": 3.552889489692464,
      "grad_norm": 7.858943462371826,
      "learning_rate": 2e-05,
      "loss": 0.1489,
      "step": 5258
    },
    {
      "epoch": 3.5535653937140927,
      "grad_norm": 2.4502642154693604,
      "learning_rate": 2e-05,
      "loss": 0.1558,
      "step": 5259
    },
    {
      "epoch": 3.5542412977357216,
      "grad_norm": 2.802452564239502,
      "learning_rate": 2e-05,
      "loss": 0.1922,
      "step": 5260
    },
    {
      "epoch": 3.5549172017573505,
      "grad_norm": 3.915968656539917,
      "learning_rate": 2e-05,
      "loss": 0.1516,
      "step": 5261
    },
    {
      "epoch": 3.5555931057789794,
      "grad_norm": 2.200908660888672,
      "learning_rate": 2e-05,
      "loss": 0.1021,
      "step": 5262
    },
    {
      "epoch": 3.5562690098006082,
      "grad_norm": 6.950832843780518,
      "learning_rate": 2e-05,
      "loss": 0.1571,
      "step": 5263
    },
    {
      "epoch": 3.556944913822237,
      "grad_norm": 7.549816131591797,
      "learning_rate": 2e-05,
      "loss": 0.2186,
      "step": 5264
    },
    {
      "epoch": 3.5576208178438664,
      "grad_norm": 10.618569374084473,
      "learning_rate": 2e-05,
      "loss": 0.1559,
      "step": 5265
    },
    {
      "epoch": 3.558296721865495,
      "grad_norm": 6.485898494720459,
      "learning_rate": 2e-05,
      "loss": 0.1781,
      "step": 5266
    },
    {
      "epoch": 3.558972625887124,
      "grad_norm": 6.3655619621276855,
      "learning_rate": 2e-05,
      "loss": 0.1375,
      "step": 5267
    },
    {
      "epoch": 3.559648529908753,
      "grad_norm": 3.6870012283325195,
      "learning_rate": 2e-05,
      "loss": 0.0895,
      "step": 5268
    },
    {
      "epoch": 3.560324433930382,
      "grad_norm": 4.897733688354492,
      "learning_rate": 2e-05,
      "loss": 0.1235,
      "step": 5269
    },
    {
      "epoch": 3.5610003379520108,
      "grad_norm": 6.01665735244751,
      "learning_rate": 2e-05,
      "loss": 0.1803,
      "step": 5270
    },
    {
      "epoch": 3.5616762419736396,
      "grad_norm": 10.99584674835205,
      "learning_rate": 2e-05,
      "loss": 0.1886,
      "step": 5271
    },
    {
      "epoch": 3.562352145995269,
      "grad_norm": 3.5600502490997314,
      "learning_rate": 2e-05,
      "loss": 0.1588,
      "step": 5272
    },
    {
      "epoch": 3.5630280500168974,
      "grad_norm": 5.683593273162842,
      "learning_rate": 2e-05,
      "loss": 0.1561,
      "step": 5273
    },
    {
      "epoch": 3.5637039540385267,
      "grad_norm": 9.51378059387207,
      "learning_rate": 2e-05,
      "loss": 0.1328,
      "step": 5274
    },
    {
      "epoch": 3.5643798580601556,
      "grad_norm": 6.027655601501465,
      "learning_rate": 2e-05,
      "loss": 0.1556,
      "step": 5275
    },
    {
      "epoch": 3.5650557620817844,
      "grad_norm": 7.499930381774902,
      "learning_rate": 2e-05,
      "loss": 0.1173,
      "step": 5276
    },
    {
      "epoch": 3.5657316661034133,
      "grad_norm": 2.958716630935669,
      "learning_rate": 2e-05,
      "loss": 0.1783,
      "step": 5277
    },
    {
      "epoch": 3.566407570125042,
      "grad_norm": 4.525477886199951,
      "learning_rate": 2e-05,
      "loss": 0.1339,
      "step": 5278
    },
    {
      "epoch": 3.567083474146671,
      "grad_norm": 1.8301900625228882,
      "learning_rate": 2e-05,
      "loss": 0.1836,
      "step": 5279
    },
    {
      "epoch": 3.5677593781683,
      "grad_norm": 5.265141487121582,
      "learning_rate": 2e-05,
      "loss": 0.0715,
      "step": 5280
    },
    {
      "epoch": 3.5684352821899292,
      "grad_norm": 3.9477295875549316,
      "learning_rate": 2e-05,
      "loss": 0.161,
      "step": 5281
    },
    {
      "epoch": 3.569111186211558,
      "grad_norm": 2.352635622024536,
      "learning_rate": 2e-05,
      "loss": 0.185,
      "step": 5282
    },
    {
      "epoch": 3.569787090233187,
      "grad_norm": 3.33182692527771,
      "learning_rate": 2e-05,
      "loss": 0.1547,
      "step": 5283
    },
    {
      "epoch": 3.570462994254816,
      "grad_norm": 2.824582815170288,
      "learning_rate": 2e-05,
      "loss": 0.2055,
      "step": 5284
    },
    {
      "epoch": 3.5711388982764447,
      "grad_norm": 3.0800681114196777,
      "learning_rate": 2e-05,
      "loss": 0.1811,
      "step": 5285
    },
    {
      "epoch": 3.5718148022980736,
      "grad_norm": 3.245086908340454,
      "learning_rate": 2e-05,
      "loss": 0.1348,
      "step": 5286
    },
    {
      "epoch": 3.5724907063197024,
      "grad_norm": 6.961956977844238,
      "learning_rate": 2e-05,
      "loss": 0.174,
      "step": 5287
    },
    {
      "epoch": 3.5731666103413318,
      "grad_norm": 3.9503703117370605,
      "learning_rate": 2e-05,
      "loss": 0.2171,
      "step": 5288
    },
    {
      "epoch": 3.57384251436296,
      "grad_norm": 7.051729202270508,
      "learning_rate": 2e-05,
      "loss": 0.1207,
      "step": 5289
    },
    {
      "epoch": 3.5745184183845895,
      "grad_norm": 2.5178792476654053,
      "learning_rate": 2e-05,
      "loss": 0.1277,
      "step": 5290
    },
    {
      "epoch": 3.5751943224062184,
      "grad_norm": 4.497530460357666,
      "learning_rate": 2e-05,
      "loss": 0.1486,
      "step": 5291
    },
    {
      "epoch": 3.5758702264278472,
      "grad_norm": 8.199091911315918,
      "learning_rate": 2e-05,
      "loss": 0.135,
      "step": 5292
    },
    {
      "epoch": 3.576546130449476,
      "grad_norm": 2.67620587348938,
      "learning_rate": 2e-05,
      "loss": 0.1356,
      "step": 5293
    },
    {
      "epoch": 3.577222034471105,
      "grad_norm": 1.977970838546753,
      "learning_rate": 2e-05,
      "loss": 0.1637,
      "step": 5294
    },
    {
      "epoch": 3.5778979384927343,
      "grad_norm": 5.355981349945068,
      "learning_rate": 2e-05,
      "loss": 0.158,
      "step": 5295
    },
    {
      "epoch": 3.5785738425143627,
      "grad_norm": 5.096035480499268,
      "learning_rate": 2e-05,
      "loss": 0.1528,
      "step": 5296
    },
    {
      "epoch": 3.579249746535992,
      "grad_norm": 4.915134429931641,
      "learning_rate": 2e-05,
      "loss": 0.1758,
      "step": 5297
    },
    {
      "epoch": 3.579925650557621,
      "grad_norm": 2.482661724090576,
      "learning_rate": 2e-05,
      "loss": 0.1336,
      "step": 5298
    },
    {
      "epoch": 3.58060155457925,
      "grad_norm": 5.877342700958252,
      "learning_rate": 2e-05,
      "loss": 0.2035,
      "step": 5299
    },
    {
      "epoch": 3.5812774586008786,
      "grad_norm": 5.035378932952881,
      "learning_rate": 2e-05,
      "loss": 0.1764,
      "step": 5300
    },
    {
      "epoch": 3.5812774586008786,
      "eval_accuracy": 0.7135135135135136,
      "eval_loss": 0.24472835659980774,
      "eval_runtime": 50.0092,
      "eval_samples_per_second": 473.513,
      "eval_steps_per_second": 14.797,
      "step": 5300
    },
    {
      "epoch": 3.5819533626225075,
      "grad_norm": 3.968423366546631,
      "learning_rate": 2e-05,
      "loss": 0.1295,
      "step": 5301
    },
    {
      "epoch": 3.5826292666441364,
      "grad_norm": 2.571499824523926,
      "learning_rate": 2e-05,
      "loss": 0.1255,
      "step": 5302
    },
    {
      "epoch": 3.5833051706657653,
      "grad_norm": 6.603121280670166,
      "learning_rate": 2e-05,
      "loss": 0.1525,
      "step": 5303
    },
    {
      "epoch": 3.5839810746873946,
      "grad_norm": 4.0738372802734375,
      "learning_rate": 2e-05,
      "loss": 0.1881,
      "step": 5304
    },
    {
      "epoch": 3.5846569787090234,
      "grad_norm": 4.186938285827637,
      "learning_rate": 2e-05,
      "loss": 0.1765,
      "step": 5305
    },
    {
      "epoch": 3.5853328827306523,
      "grad_norm": 8.532149314880371,
      "learning_rate": 2e-05,
      "loss": 0.1457,
      "step": 5306
    },
    {
      "epoch": 3.586008786752281,
      "grad_norm": 4.566595077514648,
      "learning_rate": 2e-05,
      "loss": 0.1582,
      "step": 5307
    },
    {
      "epoch": 3.58668469077391,
      "grad_norm": 4.4693379402160645,
      "learning_rate": 2e-05,
      "loss": 0.2549,
      "step": 5308
    },
    {
      "epoch": 3.587360594795539,
      "grad_norm": 6.613735675811768,
      "learning_rate": 2e-05,
      "loss": 0.1431,
      "step": 5309
    },
    {
      "epoch": 3.588036498817168,
      "grad_norm": 4.202228546142578,
      "learning_rate": 2e-05,
      "loss": 0.1972,
      "step": 5310
    },
    {
      "epoch": 3.588712402838797,
      "grad_norm": 4.542100429534912,
      "learning_rate": 2e-05,
      "loss": 0.1211,
      "step": 5311
    },
    {
      "epoch": 3.589388306860426,
      "grad_norm": 4.39876651763916,
      "learning_rate": 2e-05,
      "loss": 0.1165,
      "step": 5312
    },
    {
      "epoch": 3.590064210882055,
      "grad_norm": 5.257306098937988,
      "learning_rate": 2e-05,
      "loss": 0.153,
      "step": 5313
    },
    {
      "epoch": 3.5907401149036837,
      "grad_norm": 3.580122947692871,
      "learning_rate": 2e-05,
      "loss": 0.2048,
      "step": 5314
    },
    {
      "epoch": 3.5914160189253126,
      "grad_norm": 2.6939589977264404,
      "learning_rate": 2e-05,
      "loss": 0.2112,
      "step": 5315
    },
    {
      "epoch": 3.5920919229469415,
      "grad_norm": 3.9259021282196045,
      "learning_rate": 2e-05,
      "loss": 0.2182,
      "step": 5316
    },
    {
      "epoch": 3.5927678269685703,
      "grad_norm": 5.821676254272461,
      "learning_rate": 2e-05,
      "loss": 0.1523,
      "step": 5317
    },
    {
      "epoch": 3.5934437309901996,
      "grad_norm": 5.272063732147217,
      "learning_rate": 2e-05,
      "loss": 0.1133,
      "step": 5318
    },
    {
      "epoch": 3.594119635011828,
      "grad_norm": 1.9904643297195435,
      "learning_rate": 2e-05,
      "loss": 0.1547,
      "step": 5319
    },
    {
      "epoch": 3.5947955390334574,
      "grad_norm": 6.351142883300781,
      "learning_rate": 2e-05,
      "loss": 0.1025,
      "step": 5320
    },
    {
      "epoch": 3.5954714430550863,
      "grad_norm": 3.985820770263672,
      "learning_rate": 2e-05,
      "loss": 0.1606,
      "step": 5321
    },
    {
      "epoch": 3.596147347076715,
      "grad_norm": 3.2738037109375,
      "learning_rate": 2e-05,
      "loss": 0.1259,
      "step": 5322
    },
    {
      "epoch": 3.596823251098344,
      "grad_norm": 2.57224178314209,
      "learning_rate": 2e-05,
      "loss": 0.1095,
      "step": 5323
    },
    {
      "epoch": 3.597499155119973,
      "grad_norm": 4.542603015899658,
      "learning_rate": 2e-05,
      "loss": 0.2096,
      "step": 5324
    },
    {
      "epoch": 3.5981750591416017,
      "grad_norm": 4.565358638763428,
      "learning_rate": 2e-05,
      "loss": 0.1625,
      "step": 5325
    },
    {
      "epoch": 3.5988509631632306,
      "grad_norm": 3.5854272842407227,
      "learning_rate": 2e-05,
      "loss": 0.1348,
      "step": 5326
    },
    {
      "epoch": 3.59952686718486,
      "grad_norm": 6.363953590393066,
      "learning_rate": 2e-05,
      "loss": 0.2096,
      "step": 5327
    },
    {
      "epoch": 3.600202771206489,
      "grad_norm": 3.8018221855163574,
      "learning_rate": 2e-05,
      "loss": 0.1449,
      "step": 5328
    },
    {
      "epoch": 3.6008786752281177,
      "grad_norm": 5.3858232498168945,
      "learning_rate": 2e-05,
      "loss": 0.1478,
      "step": 5329
    },
    {
      "epoch": 3.6015545792497465,
      "grad_norm": 3.2161004543304443,
      "learning_rate": 2e-05,
      "loss": 0.1461,
      "step": 5330
    },
    {
      "epoch": 3.6022304832713754,
      "grad_norm": 3.0288610458374023,
      "learning_rate": 2e-05,
      "loss": 0.1511,
      "step": 5331
    },
    {
      "epoch": 3.6029063872930043,
      "grad_norm": 3.0804708003997803,
      "learning_rate": 2e-05,
      "loss": 0.1965,
      "step": 5332
    },
    {
      "epoch": 3.603582291314633,
      "grad_norm": 6.146119594573975,
      "learning_rate": 2e-05,
      "loss": 0.1819,
      "step": 5333
    },
    {
      "epoch": 3.6042581953362625,
      "grad_norm": 2.855759382247925,
      "learning_rate": 2e-05,
      "loss": 0.1978,
      "step": 5334
    },
    {
      "epoch": 3.6049340993578913,
      "grad_norm": 4.197208404541016,
      "learning_rate": 2e-05,
      "loss": 0.1466,
      "step": 5335
    },
    {
      "epoch": 3.60561000337952,
      "grad_norm": 2.541797161102295,
      "learning_rate": 2e-05,
      "loss": 0.0985,
      "step": 5336
    },
    {
      "epoch": 3.606285907401149,
      "grad_norm": 2.5347280502319336,
      "learning_rate": 2e-05,
      "loss": 0.1778,
      "step": 5337
    },
    {
      "epoch": 3.606961811422778,
      "grad_norm": 3.415205955505371,
      "learning_rate": 2e-05,
      "loss": 0.1918,
      "step": 5338
    },
    {
      "epoch": 3.607637715444407,
      "grad_norm": 7.408230304718018,
      "learning_rate": 2e-05,
      "loss": 0.1819,
      "step": 5339
    },
    {
      "epoch": 3.6083136194660357,
      "grad_norm": 4.402916431427002,
      "learning_rate": 2e-05,
      "loss": 0.2227,
      "step": 5340
    },
    {
      "epoch": 3.608989523487665,
      "grad_norm": 6.6884541511535645,
      "learning_rate": 2e-05,
      "loss": 0.142,
      "step": 5341
    },
    {
      "epoch": 3.6096654275092934,
      "grad_norm": 8.596169471740723,
      "learning_rate": 2e-05,
      "loss": 0.2235,
      "step": 5342
    },
    {
      "epoch": 3.6103413315309227,
      "grad_norm": 3.0617153644561768,
      "learning_rate": 2e-05,
      "loss": 0.1506,
      "step": 5343
    },
    {
      "epoch": 3.6110172355525516,
      "grad_norm": 2.983968734741211,
      "learning_rate": 2e-05,
      "loss": 0.1496,
      "step": 5344
    },
    {
      "epoch": 3.6116931395741805,
      "grad_norm": 2.0219333171844482,
      "learning_rate": 2e-05,
      "loss": 0.2078,
      "step": 5345
    },
    {
      "epoch": 3.6123690435958093,
      "grad_norm": 2.9435534477233887,
      "learning_rate": 2e-05,
      "loss": 0.2385,
      "step": 5346
    },
    {
      "epoch": 3.613044947617438,
      "grad_norm": 7.467973709106445,
      "learning_rate": 2e-05,
      "loss": 0.2264,
      "step": 5347
    },
    {
      "epoch": 3.613720851639067,
      "grad_norm": 2.55777645111084,
      "learning_rate": 2e-05,
      "loss": 0.1667,
      "step": 5348
    },
    {
      "epoch": 3.614396755660696,
      "grad_norm": 5.514954566955566,
      "learning_rate": 2e-05,
      "loss": 0.2063,
      "step": 5349
    },
    {
      "epoch": 3.6150726596823253,
      "grad_norm": 9.218352317810059,
      "learning_rate": 2e-05,
      "loss": 0.1718,
      "step": 5350
    },
    {
      "epoch": 3.6150726596823253,
      "eval_accuracy": 0.7339527027027027,
      "eval_loss": 0.2258172333240509,
      "eval_runtime": 45.608,
      "eval_samples_per_second": 519.207,
      "eval_steps_per_second": 16.225,
      "step": 5350
    },
    {
      "epoch": 3.615748563703954,
      "grad_norm": 2.7961432933807373,
      "learning_rate": 2e-05,
      "loss": 0.1903,
      "step": 5351
    },
    {
      "epoch": 3.616424467725583,
      "grad_norm": 6.381292343139648,
      "learning_rate": 2e-05,
      "loss": 0.1525,
      "step": 5352
    },
    {
      "epoch": 3.617100371747212,
      "grad_norm": 7.796747207641602,
      "learning_rate": 2e-05,
      "loss": 0.1617,
      "step": 5353
    },
    {
      "epoch": 3.6177762757688408,
      "grad_norm": 3.5176308155059814,
      "learning_rate": 2e-05,
      "loss": 0.1368,
      "step": 5354
    },
    {
      "epoch": 3.6184521797904696,
      "grad_norm": 2.299320697784424,
      "learning_rate": 2e-05,
      "loss": 0.1253,
      "step": 5355
    },
    {
      "epoch": 3.6191280838120985,
      "grad_norm": 8.047762870788574,
      "learning_rate": 2e-05,
      "loss": 0.2309,
      "step": 5356
    },
    {
      "epoch": 3.619803987833728,
      "grad_norm": 3.8018383979797363,
      "learning_rate": 2e-05,
      "loss": 0.1161,
      "step": 5357
    },
    {
      "epoch": 3.6204798918553567,
      "grad_norm": 4.859368324279785,
      "learning_rate": 2e-05,
      "loss": 0.171,
      "step": 5358
    },
    {
      "epoch": 3.6211557958769856,
      "grad_norm": 5.773250579833984,
      "learning_rate": 2e-05,
      "loss": 0.1612,
      "step": 5359
    },
    {
      "epoch": 3.6218316998986144,
      "grad_norm": 2.143646478652954,
      "learning_rate": 2e-05,
      "loss": 0.1374,
      "step": 5360
    },
    {
      "epoch": 3.6225076039202433,
      "grad_norm": 5.833280563354492,
      "learning_rate": 2e-05,
      "loss": 0.1561,
      "step": 5361
    },
    {
      "epoch": 3.623183507941872,
      "grad_norm": 7.256464004516602,
      "learning_rate": 2e-05,
      "loss": 0.1953,
      "step": 5362
    },
    {
      "epoch": 3.623859411963501,
      "grad_norm": 4.770035266876221,
      "learning_rate": 2e-05,
      "loss": 0.1611,
      "step": 5363
    },
    {
      "epoch": 3.6245353159851303,
      "grad_norm": 4.73997163772583,
      "learning_rate": 2e-05,
      "loss": 0.1323,
      "step": 5364
    },
    {
      "epoch": 3.6252112200067588,
      "grad_norm": 4.691354274749756,
      "learning_rate": 2e-05,
      "loss": 0.1298,
      "step": 5365
    },
    {
      "epoch": 3.625887124028388,
      "grad_norm": 9.01979923248291,
      "learning_rate": 2e-05,
      "loss": 0.1692,
      "step": 5366
    },
    {
      "epoch": 3.626563028050017,
      "grad_norm": 2.3389461040496826,
      "learning_rate": 2e-05,
      "loss": 0.1863,
      "step": 5367
    },
    {
      "epoch": 3.627238932071646,
      "grad_norm": 4.082907199859619,
      "learning_rate": 2e-05,
      "loss": 0.1904,
      "step": 5368
    },
    {
      "epoch": 3.6279148360932747,
      "grad_norm": 3.0290610790252686,
      "learning_rate": 2e-05,
      "loss": 0.211,
      "step": 5369
    },
    {
      "epoch": 3.6285907401149036,
      "grad_norm": 2.6365346908569336,
      "learning_rate": 2e-05,
      "loss": 0.1592,
      "step": 5370
    },
    {
      "epoch": 3.629266644136533,
      "grad_norm": 1.846373200416565,
      "learning_rate": 2e-05,
      "loss": 0.1588,
      "step": 5371
    },
    {
      "epoch": 3.6299425481581613,
      "grad_norm": 5.463951110839844,
      "learning_rate": 2e-05,
      "loss": 0.1054,
      "step": 5372
    },
    {
      "epoch": 3.6306184521797906,
      "grad_norm": 3.428457736968994,
      "learning_rate": 2e-05,
      "loss": 0.1299,
      "step": 5373
    },
    {
      "epoch": 3.6312943562014195,
      "grad_norm": 5.939024448394775,
      "learning_rate": 2e-05,
      "loss": 0.1238,
      "step": 5374
    },
    {
      "epoch": 3.6319702602230484,
      "grad_norm": 2.680893898010254,
      "learning_rate": 2e-05,
      "loss": 0.1697,
      "step": 5375
    },
    {
      "epoch": 3.6326461642446772,
      "grad_norm": 7.440643787384033,
      "learning_rate": 2e-05,
      "loss": 0.1377,
      "step": 5376
    },
    {
      "epoch": 3.633322068266306,
      "grad_norm": 3.662519931793213,
      "learning_rate": 2e-05,
      "loss": 0.1803,
      "step": 5377
    },
    {
      "epoch": 3.633997972287935,
      "grad_norm": 4.211389541625977,
      "learning_rate": 2e-05,
      "loss": 0.1501,
      "step": 5378
    },
    {
      "epoch": 3.634673876309564,
      "grad_norm": 4.589566707611084,
      "learning_rate": 2e-05,
      "loss": 0.1871,
      "step": 5379
    },
    {
      "epoch": 3.635349780331193,
      "grad_norm": 10.038708686828613,
      "learning_rate": 2e-05,
      "loss": 0.2035,
      "step": 5380
    },
    {
      "epoch": 3.636025684352822,
      "grad_norm": 11.392895698547363,
      "learning_rate": 2e-05,
      "loss": 0.185,
      "step": 5381
    },
    {
      "epoch": 3.636701588374451,
      "grad_norm": 8.733683586120605,
      "learning_rate": 2e-05,
      "loss": 0.187,
      "step": 5382
    },
    {
      "epoch": 3.6373774923960798,
      "grad_norm": 4.574251174926758,
      "learning_rate": 2e-05,
      "loss": 0.147,
      "step": 5383
    },
    {
      "epoch": 3.6380533964177086,
      "grad_norm": 2.900472640991211,
      "learning_rate": 2e-05,
      "loss": 0.1527,
      "step": 5384
    },
    {
      "epoch": 3.6387293004393375,
      "grad_norm": 3.1836905479431152,
      "learning_rate": 2e-05,
      "loss": 0.1275,
      "step": 5385
    },
    {
      "epoch": 3.6394052044609664,
      "grad_norm": 1.755094289779663,
      "learning_rate": 2e-05,
      "loss": 0.1018,
      "step": 5386
    },
    {
      "epoch": 3.6400811084825957,
      "grad_norm": 10.216047286987305,
      "learning_rate": 2e-05,
      "loss": 0.1914,
      "step": 5387
    },
    {
      "epoch": 3.6407570125042246,
      "grad_norm": 14.518362998962402,
      "learning_rate": 2e-05,
      "loss": 0.2382,
      "step": 5388
    },
    {
      "epoch": 3.6414329165258534,
      "grad_norm": 5.7449164390563965,
      "learning_rate": 2e-05,
      "loss": 0.2241,
      "step": 5389
    },
    {
      "epoch": 3.6421088205474823,
      "grad_norm": 13.490306854248047,
      "learning_rate": 2e-05,
      "loss": 0.1767,
      "step": 5390
    },
    {
      "epoch": 3.642784724569111,
      "grad_norm": 5.075697422027588,
      "learning_rate": 2e-05,
      "loss": 0.1488,
      "step": 5391
    },
    {
      "epoch": 3.64346062859074,
      "grad_norm": 3.0088257789611816,
      "learning_rate": 2e-05,
      "loss": 0.1846,
      "step": 5392
    },
    {
      "epoch": 3.644136532612369,
      "grad_norm": 3.482966423034668,
      "learning_rate": 2e-05,
      "loss": 0.1662,
      "step": 5393
    },
    {
      "epoch": 3.6448124366339982,
      "grad_norm": 5.84253454208374,
      "learning_rate": 2e-05,
      "loss": 0.1998,
      "step": 5394
    },
    {
      "epoch": 3.6454883406556267,
      "grad_norm": 6.156309127807617,
      "learning_rate": 2e-05,
      "loss": 0.1438,
      "step": 5395
    },
    {
      "epoch": 3.646164244677256,
      "grad_norm": 5.631312847137451,
      "learning_rate": 2e-05,
      "loss": 0.1717,
      "step": 5396
    },
    {
      "epoch": 3.646840148698885,
      "grad_norm": 3.6511199474334717,
      "learning_rate": 2e-05,
      "loss": 0.1538,
      "step": 5397
    },
    {
      "epoch": 3.6475160527205137,
      "grad_norm": 5.019963264465332,
      "learning_rate": 2e-05,
      "loss": 0.1569,
      "step": 5398
    },
    {
      "epoch": 3.6481919567421426,
      "grad_norm": 4.710596084594727,
      "learning_rate": 2e-05,
      "loss": 0.1063,
      "step": 5399
    },
    {
      "epoch": 3.6488678607637715,
      "grad_norm": 1.8504118919372559,
      "learning_rate": 2e-05,
      "loss": 0.1047,
      "step": 5400
    },
    {
      "epoch": 3.6488678607637715,
      "eval_accuracy": 0.7366976351351351,
      "eval_loss": 0.24096830189228058,
      "eval_runtime": 46.8662,
      "eval_samples_per_second": 505.268,
      "eval_steps_per_second": 15.79,
      "step": 5400
    },
    {
      "epoch": 3.6495437647854003,
      "grad_norm": 3.2394156455993652,
      "learning_rate": 2e-05,
      "loss": 0.156,
      "step": 5401
    },
    {
      "epoch": 3.650219668807029,
      "grad_norm": 7.616469383239746,
      "learning_rate": 2e-05,
      "loss": 0.2275,
      "step": 5402
    },
    {
      "epoch": 3.6508955728286585,
      "grad_norm": 10.817818641662598,
      "learning_rate": 2e-05,
      "loss": 0.2017,
      "step": 5403
    },
    {
      "epoch": 3.6515714768502874,
      "grad_norm": 9.425456047058105,
      "learning_rate": 2e-05,
      "loss": 0.144,
      "step": 5404
    },
    {
      "epoch": 3.6522473808719162,
      "grad_norm": 2.1585428714752197,
      "learning_rate": 2e-05,
      "loss": 0.1515,
      "step": 5405
    },
    {
      "epoch": 3.652923284893545,
      "grad_norm": 9.17733383178711,
      "learning_rate": 2e-05,
      "loss": 0.1774,
      "step": 5406
    },
    {
      "epoch": 3.653599188915174,
      "grad_norm": 4.680342674255371,
      "learning_rate": 2e-05,
      "loss": 0.1789,
      "step": 5407
    },
    {
      "epoch": 3.654275092936803,
      "grad_norm": 4.508382797241211,
      "learning_rate": 2e-05,
      "loss": 0.1932,
      "step": 5408
    },
    {
      "epoch": 3.6549509969584317,
      "grad_norm": 3.271103620529175,
      "learning_rate": 2e-05,
      "loss": 0.1313,
      "step": 5409
    },
    {
      "epoch": 3.655626900980061,
      "grad_norm": 6.048834323883057,
      "learning_rate": 2e-05,
      "loss": 0.2187,
      "step": 5410
    },
    {
      "epoch": 3.65630280500169,
      "grad_norm": 12.358418464660645,
      "learning_rate": 2e-05,
      "loss": 0.2428,
      "step": 5411
    },
    {
      "epoch": 3.656978709023319,
      "grad_norm": 7.3177571296691895,
      "learning_rate": 2e-05,
      "loss": 0.218,
      "step": 5412
    },
    {
      "epoch": 3.6576546130449477,
      "grad_norm": 4.195733547210693,
      "learning_rate": 2e-05,
      "loss": 0.148,
      "step": 5413
    },
    {
      "epoch": 3.6583305170665765,
      "grad_norm": 2.3116703033447266,
      "learning_rate": 2e-05,
      "loss": 0.0983,
      "step": 5414
    },
    {
      "epoch": 3.6590064210882054,
      "grad_norm": 4.395430564880371,
      "learning_rate": 2e-05,
      "loss": 0.1211,
      "step": 5415
    },
    {
      "epoch": 3.6596823251098343,
      "grad_norm": 3.595926284790039,
      "learning_rate": 2e-05,
      "loss": 0.1738,
      "step": 5416
    },
    {
      "epoch": 3.6603582291314636,
      "grad_norm": 3.380878210067749,
      "learning_rate": 2e-05,
      "loss": 0.2158,
      "step": 5417
    },
    {
      "epoch": 3.661034133153092,
      "grad_norm": 4.398932456970215,
      "learning_rate": 2e-05,
      "loss": 0.1438,
      "step": 5418
    },
    {
      "epoch": 3.6617100371747213,
      "grad_norm": 3.1073195934295654,
      "learning_rate": 2e-05,
      "loss": 0.2013,
      "step": 5419
    },
    {
      "epoch": 3.66238594119635,
      "grad_norm": 7.499815940856934,
      "learning_rate": 2e-05,
      "loss": 0.1389,
      "step": 5420
    },
    {
      "epoch": 3.663061845217979,
      "grad_norm": 4.653838157653809,
      "learning_rate": 2e-05,
      "loss": 0.1871,
      "step": 5421
    },
    {
      "epoch": 3.663737749239608,
      "grad_norm": 6.9918742179870605,
      "learning_rate": 2e-05,
      "loss": 0.158,
      "step": 5422
    },
    {
      "epoch": 3.664413653261237,
      "grad_norm": 2.531989097595215,
      "learning_rate": 2e-05,
      "loss": 0.1868,
      "step": 5423
    },
    {
      "epoch": 3.6650895572828657,
      "grad_norm": 2.885899782180786,
      "learning_rate": 2e-05,
      "loss": 0.2182,
      "step": 5424
    },
    {
      "epoch": 3.6657654613044945,
      "grad_norm": 2.7396066188812256,
      "learning_rate": 2e-05,
      "loss": 0.1677,
      "step": 5425
    },
    {
      "epoch": 3.666441365326124,
      "grad_norm": 2.244330644607544,
      "learning_rate": 2e-05,
      "loss": 0.1464,
      "step": 5426
    },
    {
      "epoch": 3.6671172693477527,
      "grad_norm": 2.7298951148986816,
      "learning_rate": 2e-05,
      "loss": 0.1822,
      "step": 5427
    },
    {
      "epoch": 3.6677931733693816,
      "grad_norm": 4.527921199798584,
      "learning_rate": 2e-05,
      "loss": 0.161,
      "step": 5428
    },
    {
      "epoch": 3.6684690773910105,
      "grad_norm": 2.7534122467041016,
      "learning_rate": 2e-05,
      "loss": 0.1757,
      "step": 5429
    },
    {
      "epoch": 3.6691449814126393,
      "grad_norm": 3.726963758468628,
      "learning_rate": 2e-05,
      "loss": 0.1713,
      "step": 5430
    },
    {
      "epoch": 3.669820885434268,
      "grad_norm": 5.593324184417725,
      "learning_rate": 2e-05,
      "loss": 0.1762,
      "step": 5431
    },
    {
      "epoch": 3.670496789455897,
      "grad_norm": 2.8234689235687256,
      "learning_rate": 2e-05,
      "loss": 0.1698,
      "step": 5432
    },
    {
      "epoch": 3.6711726934775264,
      "grad_norm": 3.0681116580963135,
      "learning_rate": 2e-05,
      "loss": 0.1667,
      "step": 5433
    },
    {
      "epoch": 3.6718485974991553,
      "grad_norm": 2.5113470554351807,
      "learning_rate": 2e-05,
      "loss": 0.1634,
      "step": 5434
    },
    {
      "epoch": 3.672524501520784,
      "grad_norm": 10.717145919799805,
      "learning_rate": 2e-05,
      "loss": 0.1953,
      "step": 5435
    },
    {
      "epoch": 3.673200405542413,
      "grad_norm": 2.8538498878479004,
      "learning_rate": 2e-05,
      "loss": 0.1512,
      "step": 5436
    },
    {
      "epoch": 3.673876309564042,
      "grad_norm": 5.869643688201904,
      "learning_rate": 2e-05,
      "loss": 0.1963,
      "step": 5437
    },
    {
      "epoch": 3.6745522135856707,
      "grad_norm": 15.432609558105469,
      "learning_rate": 2e-05,
      "loss": 0.1849,
      "step": 5438
    },
    {
      "epoch": 3.6752281176072996,
      "grad_norm": 4.133533954620361,
      "learning_rate": 2e-05,
      "loss": 0.1809,
      "step": 5439
    },
    {
      "epoch": 3.675904021628929,
      "grad_norm": 6.1811933517456055,
      "learning_rate": 2e-05,
      "loss": 0.1437,
      "step": 5440
    },
    {
      "epoch": 3.6765799256505574,
      "grad_norm": 5.997272968292236,
      "learning_rate": 2e-05,
      "loss": 0.1801,
      "step": 5441
    },
    {
      "epoch": 3.6772558296721867,
      "grad_norm": 3.3622617721557617,
      "learning_rate": 2e-05,
      "loss": 0.1502,
      "step": 5442
    },
    {
      "epoch": 3.6779317336938155,
      "grad_norm": 2.824657440185547,
      "learning_rate": 2e-05,
      "loss": 0.2009,
      "step": 5443
    },
    {
      "epoch": 3.6786076377154444,
      "grad_norm": 6.51162052154541,
      "learning_rate": 2e-05,
      "loss": 0.1653,
      "step": 5444
    },
    {
      "epoch": 3.6792835417370733,
      "grad_norm": 3.2063324451446533,
      "learning_rate": 2e-05,
      "loss": 0.1355,
      "step": 5445
    },
    {
      "epoch": 3.679959445758702,
      "grad_norm": 2.220686674118042,
      "learning_rate": 2e-05,
      "loss": 0.1665,
      "step": 5446
    },
    {
      "epoch": 3.6806353497803315,
      "grad_norm": 6.615819454193115,
      "learning_rate": 2e-05,
      "loss": 0.2308,
      "step": 5447
    },
    {
      "epoch": 3.68131125380196,
      "grad_norm": 7.382824420928955,
      "learning_rate": 2e-05,
      "loss": 0.1891,
      "step": 5448
    },
    {
      "epoch": 3.681987157823589,
      "grad_norm": 3.9041454792022705,
      "learning_rate": 2e-05,
      "loss": 0.167,
      "step": 5449
    },
    {
      "epoch": 3.682663061845218,
      "grad_norm": 2.2621665000915527,
      "learning_rate": 2e-05,
      "loss": 0.1145,
      "step": 5450
    },
    {
      "epoch": 3.682663061845218,
      "eval_accuracy": 0.7316722972972973,
      "eval_loss": 0.23839005827903748,
      "eval_runtime": 44.7639,
      "eval_samples_per_second": 528.997,
      "eval_steps_per_second": 16.531,
      "step": 5450
    },
    {
      "epoch": 3.683338965866847,
      "grad_norm": 6.085796356201172,
      "learning_rate": 2e-05,
      "loss": 0.2052,
      "step": 5451
    },
    {
      "epoch": 3.684014869888476,
      "grad_norm": 2.0812156200408936,
      "learning_rate": 2e-05,
      "loss": 0.1189,
      "step": 5452
    },
    {
      "epoch": 3.6846907739101047,
      "grad_norm": 6.431249618530273,
      "learning_rate": 2e-05,
      "loss": 0.167,
      "step": 5453
    },
    {
      "epoch": 3.6853666779317336,
      "grad_norm": 2.0256710052490234,
      "learning_rate": 2e-05,
      "loss": 0.1095,
      "step": 5454
    },
    {
      "epoch": 3.6860425819533624,
      "grad_norm": 1.9973706007003784,
      "learning_rate": 2e-05,
      "loss": 0.1005,
      "step": 5455
    },
    {
      "epoch": 3.6867184859749917,
      "grad_norm": 5.045362949371338,
      "learning_rate": 2e-05,
      "loss": 0.1555,
      "step": 5456
    },
    {
      "epoch": 3.6873943899966206,
      "grad_norm": 2.784858226776123,
      "learning_rate": 2e-05,
      "loss": 0.158,
      "step": 5457
    },
    {
      "epoch": 3.6880702940182495,
      "grad_norm": 3.480037212371826,
      "learning_rate": 2e-05,
      "loss": 0.1429,
      "step": 5458
    },
    {
      "epoch": 3.6887461980398784,
      "grad_norm": 3.184875726699829,
      "learning_rate": 2e-05,
      "loss": 0.1918,
      "step": 5459
    },
    {
      "epoch": 3.6894221020615072,
      "grad_norm": 2.6428091526031494,
      "learning_rate": 2e-05,
      "loss": 0.1661,
      "step": 5460
    },
    {
      "epoch": 3.690098006083136,
      "grad_norm": 3.176898241043091,
      "learning_rate": 2e-05,
      "loss": 0.1814,
      "step": 5461
    },
    {
      "epoch": 3.690773910104765,
      "grad_norm": 7.995889663696289,
      "learning_rate": 2e-05,
      "loss": 0.1656,
      "step": 5462
    },
    {
      "epoch": 3.6914498141263943,
      "grad_norm": 5.779033660888672,
      "learning_rate": 2e-05,
      "loss": 0.1682,
      "step": 5463
    },
    {
      "epoch": 3.6921257181480227,
      "grad_norm": 3.8250234127044678,
      "learning_rate": 2e-05,
      "loss": 0.1975,
      "step": 5464
    },
    {
      "epoch": 3.692801622169652,
      "grad_norm": 4.240795612335205,
      "learning_rate": 2e-05,
      "loss": 0.2382,
      "step": 5465
    },
    {
      "epoch": 3.693477526191281,
      "grad_norm": 4.794029712677002,
      "learning_rate": 2e-05,
      "loss": 0.1819,
      "step": 5466
    },
    {
      "epoch": 3.6941534302129098,
      "grad_norm": 3.3847830295562744,
      "learning_rate": 2e-05,
      "loss": 0.1272,
      "step": 5467
    },
    {
      "epoch": 3.6948293342345386,
      "grad_norm": 7.366349220275879,
      "learning_rate": 2e-05,
      "loss": 0.1662,
      "step": 5468
    },
    {
      "epoch": 3.6955052382561675,
      "grad_norm": 3.770686149597168,
      "learning_rate": 2e-05,
      "loss": 0.1873,
      "step": 5469
    },
    {
      "epoch": 3.696181142277797,
      "grad_norm": 16.243846893310547,
      "learning_rate": 2e-05,
      "loss": 0.2812,
      "step": 5470
    },
    {
      "epoch": 3.6968570462994252,
      "grad_norm": 2.479111433029175,
      "learning_rate": 2e-05,
      "loss": 0.1856,
      "step": 5471
    },
    {
      "epoch": 3.6975329503210546,
      "grad_norm": 2.7504665851593018,
      "learning_rate": 2e-05,
      "loss": 0.148,
      "step": 5472
    },
    {
      "epoch": 3.6982088543426834,
      "grad_norm": 2.10137677192688,
      "learning_rate": 2e-05,
      "loss": 0.1431,
      "step": 5473
    },
    {
      "epoch": 3.6988847583643123,
      "grad_norm": 3.589125156402588,
      "learning_rate": 2e-05,
      "loss": 0.1285,
      "step": 5474
    },
    {
      "epoch": 3.699560662385941,
      "grad_norm": 2.2089850902557373,
      "learning_rate": 2e-05,
      "loss": 0.1179,
      "step": 5475
    },
    {
      "epoch": 3.70023656640757,
      "grad_norm": 2.9548442363739014,
      "learning_rate": 2e-05,
      "loss": 0.1418,
      "step": 5476
    },
    {
      "epoch": 3.700912470429199,
      "grad_norm": 2.7171590328216553,
      "learning_rate": 2e-05,
      "loss": 0.1904,
      "step": 5477
    },
    {
      "epoch": 3.701588374450828,
      "grad_norm": 2.276737928390503,
      "learning_rate": 2e-05,
      "loss": 0.1689,
      "step": 5478
    },
    {
      "epoch": 3.702264278472457,
      "grad_norm": 5.007747173309326,
      "learning_rate": 2e-05,
      "loss": 0.1576,
      "step": 5479
    },
    {
      "epoch": 3.702940182494086,
      "grad_norm": 7.244895935058594,
      "learning_rate": 2e-05,
      "loss": 0.1826,
      "step": 5480
    },
    {
      "epoch": 3.703616086515715,
      "grad_norm": 5.557367324829102,
      "learning_rate": 2e-05,
      "loss": 0.1791,
      "step": 5481
    },
    {
      "epoch": 3.7042919905373437,
      "grad_norm": 4.98173713684082,
      "learning_rate": 2e-05,
      "loss": 0.2023,
      "step": 5482
    },
    {
      "epoch": 3.7049678945589726,
      "grad_norm": 5.575046539306641,
      "learning_rate": 2e-05,
      "loss": 0.1892,
      "step": 5483
    },
    {
      "epoch": 3.7056437985806014,
      "grad_norm": 5.598419189453125,
      "learning_rate": 2e-05,
      "loss": 0.1017,
      "step": 5484
    },
    {
      "epoch": 3.7063197026022303,
      "grad_norm": 3.398878812789917,
      "learning_rate": 2e-05,
      "loss": 0.1228,
      "step": 5485
    },
    {
      "epoch": 3.7069956066238596,
      "grad_norm": 3.1187219619750977,
      "learning_rate": 2e-05,
      "loss": 0.1575,
      "step": 5486
    },
    {
      "epoch": 3.7076715106454885,
      "grad_norm": 2.3810391426086426,
      "learning_rate": 2e-05,
      "loss": 0.151,
      "step": 5487
    },
    {
      "epoch": 3.7083474146671174,
      "grad_norm": 7.382774829864502,
      "learning_rate": 2e-05,
      "loss": 0.0905,
      "step": 5488
    },
    {
      "epoch": 3.7090233186887462,
      "grad_norm": 2.785849094390869,
      "learning_rate": 2e-05,
      "loss": 0.1631,
      "step": 5489
    },
    {
      "epoch": 3.709699222710375,
      "grad_norm": 3.4070868492126465,
      "learning_rate": 2e-05,
      "loss": 0.1192,
      "step": 5490
    },
    {
      "epoch": 3.710375126732004,
      "grad_norm": 5.580560684204102,
      "learning_rate": 2e-05,
      "loss": 0.2191,
      "step": 5491
    },
    {
      "epoch": 3.711051030753633,
      "grad_norm": 4.233524799346924,
      "learning_rate": 2e-05,
      "loss": 0.1526,
      "step": 5492
    },
    {
      "epoch": 3.711726934775262,
      "grad_norm": 3.5840039253234863,
      "learning_rate": 2e-05,
      "loss": 0.1503,
      "step": 5493
    },
    {
      "epoch": 3.7124028387968906,
      "grad_norm": 4.720475673675537,
      "learning_rate": 2e-05,
      "loss": 0.1857,
      "step": 5494
    },
    {
      "epoch": 3.71307874281852,
      "grad_norm": 6.584447860717773,
      "learning_rate": 2e-05,
      "loss": 0.1415,
      "step": 5495
    },
    {
      "epoch": 3.7137546468401488,
      "grad_norm": 4.983138084411621,
      "learning_rate": 2e-05,
      "loss": 0.1344,
      "step": 5496
    },
    {
      "epoch": 3.7144305508617776,
      "grad_norm": 4.726710319519043,
      "learning_rate": 2e-05,
      "loss": 0.203,
      "step": 5497
    },
    {
      "epoch": 3.7151064548834065,
      "grad_norm": 3.4405083656311035,
      "learning_rate": 2e-05,
      "loss": 0.1427,
      "step": 5498
    },
    {
      "epoch": 3.7157823589050354,
      "grad_norm": 7.276168346405029,
      "learning_rate": 2e-05,
      "loss": 0.1256,
      "step": 5499
    },
    {
      "epoch": 3.7164582629266643,
      "grad_norm": 10.116206169128418,
      "learning_rate": 2e-05,
      "loss": 0.1777,
      "step": 5500
    },
    {
      "epoch": 3.7164582629266643,
      "eval_accuracy": 0.7395692567567568,
      "eval_loss": 0.2424560785293579,
      "eval_runtime": 45.9796,
      "eval_samples_per_second": 515.011,
      "eval_steps_per_second": 16.094,
      "step": 5500
    },
    {
      "epoch": 3.717134166948293,
      "grad_norm": 4.443328857421875,
      "learning_rate": 2e-05,
      "loss": 0.1783,
      "step": 5501
    },
    {
      "epoch": 3.7178100709699224,
      "grad_norm": 2.7796976566314697,
      "learning_rate": 2e-05,
      "loss": 0.2468,
      "step": 5502
    },
    {
      "epoch": 3.7184859749915513,
      "grad_norm": 8.203929901123047,
      "learning_rate": 2e-05,
      "loss": 0.1907,
      "step": 5503
    },
    {
      "epoch": 3.71916187901318,
      "grad_norm": 3.8151798248291016,
      "learning_rate": 2e-05,
      "loss": 0.1814,
      "step": 5504
    },
    {
      "epoch": 3.719837783034809,
      "grad_norm": 6.747115135192871,
      "learning_rate": 2e-05,
      "loss": 0.1244,
      "step": 5505
    },
    {
      "epoch": 3.720513687056438,
      "grad_norm": 4.105491638183594,
      "learning_rate": 2e-05,
      "loss": 0.248,
      "step": 5506
    },
    {
      "epoch": 3.721189591078067,
      "grad_norm": 13.681949615478516,
      "learning_rate": 2e-05,
      "loss": 0.2243,
      "step": 5507
    },
    {
      "epoch": 3.7218654950996957,
      "grad_norm": 9.122276306152344,
      "learning_rate": 2e-05,
      "loss": 0.2019,
      "step": 5508
    },
    {
      "epoch": 3.722541399121325,
      "grad_norm": 5.696601867675781,
      "learning_rate": 2e-05,
      "loss": 0.1565,
      "step": 5509
    },
    {
      "epoch": 3.723217303142954,
      "grad_norm": 5.712381839752197,
      "learning_rate": 2e-05,
      "loss": 0.1486,
      "step": 5510
    },
    {
      "epoch": 3.7238932071645827,
      "grad_norm": 7.226929187774658,
      "learning_rate": 2e-05,
      "loss": 0.1694,
      "step": 5511
    },
    {
      "epoch": 3.7245691111862116,
      "grad_norm": 8.258482933044434,
      "learning_rate": 2e-05,
      "loss": 0.2484,
      "step": 5512
    },
    {
      "epoch": 3.7252450152078405,
      "grad_norm": 10.284956932067871,
      "learning_rate": 2e-05,
      "loss": 0.2761,
      "step": 5513
    },
    {
      "epoch": 3.7259209192294693,
      "grad_norm": 4.011879920959473,
      "learning_rate": 2e-05,
      "loss": 0.185,
      "step": 5514
    },
    {
      "epoch": 3.726596823251098,
      "grad_norm": 5.778975486755371,
      "learning_rate": 2e-05,
      "loss": 0.1548,
      "step": 5515
    },
    {
      "epoch": 3.7272727272727275,
      "grad_norm": 2.186678409576416,
      "learning_rate": 2e-05,
      "loss": 0.1617,
      "step": 5516
    },
    {
      "epoch": 3.727948631294356,
      "grad_norm": 3.874242067337036,
      "learning_rate": 2e-05,
      "loss": 0.1427,
      "step": 5517
    },
    {
      "epoch": 3.7286245353159853,
      "grad_norm": 1.9584476947784424,
      "learning_rate": 2e-05,
      "loss": 0.1828,
      "step": 5518
    },
    {
      "epoch": 3.729300439337614,
      "grad_norm": 5.048618793487549,
      "learning_rate": 2e-05,
      "loss": 0.162,
      "step": 5519
    },
    {
      "epoch": 3.729976343359243,
      "grad_norm": 2.263502836227417,
      "learning_rate": 2e-05,
      "loss": 0.2001,
      "step": 5520
    },
    {
      "epoch": 3.730652247380872,
      "grad_norm": 3.6358883380889893,
      "learning_rate": 2e-05,
      "loss": 0.1927,
      "step": 5521
    },
    {
      "epoch": 3.7313281514025007,
      "grad_norm": 2.3581764698028564,
      "learning_rate": 2e-05,
      "loss": 0.1415,
      "step": 5522
    },
    {
      "epoch": 3.7320040554241296,
      "grad_norm": 5.345638751983643,
      "learning_rate": 2e-05,
      "loss": 0.1622,
      "step": 5523
    },
    {
      "epoch": 3.7326799594457585,
      "grad_norm": 3.743532657623291,
      "learning_rate": 2e-05,
      "loss": 0.2417,
      "step": 5524
    },
    {
      "epoch": 3.733355863467388,
      "grad_norm": 3.3994381427764893,
      "learning_rate": 2e-05,
      "loss": 0.2008,
      "step": 5525
    },
    {
      "epoch": 3.7340317674890167,
      "grad_norm": 3.901966094970703,
      "learning_rate": 2e-05,
      "loss": 0.1377,
      "step": 5526
    },
    {
      "epoch": 3.7347076715106455,
      "grad_norm": 4.593057155609131,
      "learning_rate": 2e-05,
      "loss": 0.1469,
      "step": 5527
    },
    {
      "epoch": 3.7353835755322744,
      "grad_norm": 2.0405004024505615,
      "learning_rate": 2e-05,
      "loss": 0.1635,
      "step": 5528
    },
    {
      "epoch": 3.7360594795539033,
      "grad_norm": 5.619697570800781,
      "learning_rate": 2e-05,
      "loss": 0.1985,
      "step": 5529
    },
    {
      "epoch": 3.736735383575532,
      "grad_norm": 3.677414894104004,
      "learning_rate": 2e-05,
      "loss": 0.1822,
      "step": 5530
    },
    {
      "epoch": 3.737411287597161,
      "grad_norm": 5.813183307647705,
      "learning_rate": 2e-05,
      "loss": 0.1666,
      "step": 5531
    },
    {
      "epoch": 3.7380871916187903,
      "grad_norm": 2.847853183746338,
      "learning_rate": 2e-05,
      "loss": 0.1722,
      "step": 5532
    },
    {
      "epoch": 3.738763095640419,
      "grad_norm": 4.298083782196045,
      "learning_rate": 2e-05,
      "loss": 0.1154,
      "step": 5533
    },
    {
      "epoch": 3.739438999662048,
      "grad_norm": 2.096067428588867,
      "learning_rate": 2e-05,
      "loss": 0.1543,
      "step": 5534
    },
    {
      "epoch": 3.740114903683677,
      "grad_norm": 2.24698543548584,
      "learning_rate": 2e-05,
      "loss": 0.2055,
      "step": 5535
    },
    {
      "epoch": 3.740790807705306,
      "grad_norm": 3.979240894317627,
      "learning_rate": 2e-05,
      "loss": 0.1682,
      "step": 5536
    },
    {
      "epoch": 3.7414667117269347,
      "grad_norm": 5.53845739364624,
      "learning_rate": 2e-05,
      "loss": 0.1421,
      "step": 5537
    },
    {
      "epoch": 3.7421426157485635,
      "grad_norm": 2.7879598140716553,
      "learning_rate": 2e-05,
      "loss": 0.2172,
      "step": 5538
    },
    {
      "epoch": 3.742818519770193,
      "grad_norm": 4.784994125366211,
      "learning_rate": 2e-05,
      "loss": 0.2018,
      "step": 5539
    },
    {
      "epoch": 3.7434944237918213,
      "grad_norm": 6.589082717895508,
      "learning_rate": 2e-05,
      "loss": 0.1603,
      "step": 5540
    },
    {
      "epoch": 3.7441703278134506,
      "grad_norm": 2.4354734420776367,
      "learning_rate": 2e-05,
      "loss": 0.1377,
      "step": 5541
    },
    {
      "epoch": 3.7448462318350795,
      "grad_norm": 2.689826726913452,
      "learning_rate": 2e-05,
      "loss": 0.1598,
      "step": 5542
    },
    {
      "epoch": 3.7455221358567083,
      "grad_norm": 10.419938087463379,
      "learning_rate": 2e-05,
      "loss": 0.1864,
      "step": 5543
    },
    {
      "epoch": 3.746198039878337,
      "grad_norm": 6.221328258514404,
      "learning_rate": 2e-05,
      "loss": 0.1158,
      "step": 5544
    },
    {
      "epoch": 3.746873943899966,
      "grad_norm": 8.386383056640625,
      "learning_rate": 2e-05,
      "loss": 0.2134,
      "step": 5545
    },
    {
      "epoch": 3.7475498479215954,
      "grad_norm": 4.483159065246582,
      "learning_rate": 2e-05,
      "loss": 0.1887,
      "step": 5546
    },
    {
      "epoch": 3.748225751943224,
      "grad_norm": 5.520960807800293,
      "learning_rate": 2e-05,
      "loss": 0.172,
      "step": 5547
    },
    {
      "epoch": 3.748901655964853,
      "grad_norm": 3.0502872467041016,
      "learning_rate": 2e-05,
      "loss": 0.1407,
      "step": 5548
    },
    {
      "epoch": 3.749577559986482,
      "grad_norm": 4.638334274291992,
      "learning_rate": 2e-05,
      "loss": 0.1713,
      "step": 5549
    },
    {
      "epoch": 3.750253464008111,
      "grad_norm": 4.785128116607666,
      "learning_rate": 2e-05,
      "loss": 0.1812,
      "step": 5550
    },
    {
      "epoch": 3.750253464008111,
      "eval_accuracy": 0.7226773648648649,
      "eval_loss": 0.24083715677261353,
      "eval_runtime": 47.7279,
      "eval_samples_per_second": 496.146,
      "eval_steps_per_second": 15.505,
      "step": 5550
    },
    {
      "epoch": 3.7509293680297398,
      "grad_norm": 3.173689365386963,
      "learning_rate": 2e-05,
      "loss": 0.1336,
      "step": 5551
    },
    {
      "epoch": 3.7516052720513686,
      "grad_norm": 3.028743267059326,
      "learning_rate": 2e-05,
      "loss": 0.2129,
      "step": 5552
    },
    {
      "epoch": 3.7522811760729975,
      "grad_norm": 3.6677374839782715,
      "learning_rate": 2e-05,
      "loss": 0.1332,
      "step": 5553
    },
    {
      "epoch": 3.7529570800946264,
      "grad_norm": 2.5796563625335693,
      "learning_rate": 2e-05,
      "loss": 0.1354,
      "step": 5554
    },
    {
      "epoch": 3.7536329841162557,
      "grad_norm": 5.59733247756958,
      "learning_rate": 2e-05,
      "loss": 0.1636,
      "step": 5555
    },
    {
      "epoch": 3.7543088881378845,
      "grad_norm": 7.67853307723999,
      "learning_rate": 2e-05,
      "loss": 0.1759,
      "step": 5556
    },
    {
      "epoch": 3.7549847921595134,
      "grad_norm": 2.8800084590911865,
      "learning_rate": 2e-05,
      "loss": 0.1851,
      "step": 5557
    },
    {
      "epoch": 3.7556606961811423,
      "grad_norm": 3.4054462909698486,
      "learning_rate": 2e-05,
      "loss": 0.1537,
      "step": 5558
    },
    {
      "epoch": 3.756336600202771,
      "grad_norm": 2.6236164569854736,
      "learning_rate": 2e-05,
      "loss": 0.1663,
      "step": 5559
    },
    {
      "epoch": 3.7570125042244,
      "grad_norm": 4.216082572937012,
      "learning_rate": 2e-05,
      "loss": 0.2099,
      "step": 5560
    },
    {
      "epoch": 3.757688408246029,
      "grad_norm": 4.368841648101807,
      "learning_rate": 2e-05,
      "loss": 0.1307,
      "step": 5561
    },
    {
      "epoch": 3.758364312267658,
      "grad_norm": 7.167003631591797,
      "learning_rate": 2e-05,
      "loss": 0.1354,
      "step": 5562
    },
    {
      "epoch": 3.759040216289287,
      "grad_norm": 8.533295631408691,
      "learning_rate": 2e-05,
      "loss": 0.1616,
      "step": 5563
    },
    {
      "epoch": 3.759716120310916,
      "grad_norm": 2.82564640045166,
      "learning_rate": 2e-05,
      "loss": 0.145,
      "step": 5564
    },
    {
      "epoch": 3.760392024332545,
      "grad_norm": 2.660717010498047,
      "learning_rate": 2e-05,
      "loss": 0.1634,
      "step": 5565
    },
    {
      "epoch": 3.7610679283541737,
      "grad_norm": 5.199586868286133,
      "learning_rate": 2e-05,
      "loss": 0.1427,
      "step": 5566
    },
    {
      "epoch": 3.7617438323758026,
      "grad_norm": 3.1122734546661377,
      "learning_rate": 2e-05,
      "loss": 0.2241,
      "step": 5567
    },
    {
      "epoch": 3.7624197363974314,
      "grad_norm": 4.9316630363464355,
      "learning_rate": 2e-05,
      "loss": 0.1823,
      "step": 5568
    },
    {
      "epoch": 3.7630956404190608,
      "grad_norm": 3.5665924549102783,
      "learning_rate": 2e-05,
      "loss": 0.1871,
      "step": 5569
    },
    {
      "epoch": 3.763771544440689,
      "grad_norm": 3.774451732635498,
      "learning_rate": 2e-05,
      "loss": 0.168,
      "step": 5570
    },
    {
      "epoch": 3.7644474484623185,
      "grad_norm": 3.002134084701538,
      "learning_rate": 2e-05,
      "loss": 0.173,
      "step": 5571
    },
    {
      "epoch": 3.7651233524839474,
      "grad_norm": 2.0757806301116943,
      "learning_rate": 2e-05,
      "loss": 0.1418,
      "step": 5572
    },
    {
      "epoch": 3.7657992565055762,
      "grad_norm": 4.243719577789307,
      "learning_rate": 2e-05,
      "loss": 0.1716,
      "step": 5573
    },
    {
      "epoch": 3.766475160527205,
      "grad_norm": 3.7978360652923584,
      "learning_rate": 2e-05,
      "loss": 0.1244,
      "step": 5574
    },
    {
      "epoch": 3.767151064548834,
      "grad_norm": 6.05989933013916,
      "learning_rate": 2e-05,
      "loss": 0.1887,
      "step": 5575
    },
    {
      "epoch": 3.767826968570463,
      "grad_norm": 6.615500450134277,
      "learning_rate": 2e-05,
      "loss": 0.1851,
      "step": 5576
    },
    {
      "epoch": 3.7685028725920917,
      "grad_norm": 7.112826347351074,
      "learning_rate": 2e-05,
      "loss": 0.1748,
      "step": 5577
    },
    {
      "epoch": 3.769178776613721,
      "grad_norm": 5.6395440101623535,
      "learning_rate": 2e-05,
      "loss": 0.214,
      "step": 5578
    },
    {
      "epoch": 3.76985468063535,
      "grad_norm": 11.390191078186035,
      "learning_rate": 2e-05,
      "loss": 0.2716,
      "step": 5579
    },
    {
      "epoch": 3.7705305846569788,
      "grad_norm": 9.824883460998535,
      "learning_rate": 2e-05,
      "loss": 0.1868,
      "step": 5580
    },
    {
      "epoch": 3.7712064886786076,
      "grad_norm": 3.076751708984375,
      "learning_rate": 2e-05,
      "loss": 0.1664,
      "step": 5581
    },
    {
      "epoch": 3.7718823927002365,
      "grad_norm": 6.1371965408325195,
      "learning_rate": 2e-05,
      "loss": 0.1615,
      "step": 5582
    },
    {
      "epoch": 3.7725582967218654,
      "grad_norm": 4.808010101318359,
      "learning_rate": 2e-05,
      "loss": 0.1402,
      "step": 5583
    },
    {
      "epoch": 3.7732342007434942,
      "grad_norm": 10.260967254638672,
      "learning_rate": 2e-05,
      "loss": 0.2753,
      "step": 5584
    },
    {
      "epoch": 3.7739101047651236,
      "grad_norm": 9.53116512298584,
      "learning_rate": 2e-05,
      "loss": 0.1643,
      "step": 5585
    },
    {
      "epoch": 3.7745860087867524,
      "grad_norm": 6.394683361053467,
      "learning_rate": 2e-05,
      "loss": 0.1726,
      "step": 5586
    },
    {
      "epoch": 3.7752619128083813,
      "grad_norm": 2.6822125911712646,
      "learning_rate": 2e-05,
      "loss": 0.2265,
      "step": 5587
    },
    {
      "epoch": 3.77593781683001,
      "grad_norm": 4.099122524261475,
      "learning_rate": 2e-05,
      "loss": 0.1735,
      "step": 5588
    },
    {
      "epoch": 3.776613720851639,
      "grad_norm": 5.885971546173096,
      "learning_rate": 2e-05,
      "loss": 0.1698,
      "step": 5589
    },
    {
      "epoch": 3.777289624873268,
      "grad_norm": 5.427668571472168,
      "learning_rate": 2e-05,
      "loss": 0.1762,
      "step": 5590
    },
    {
      "epoch": 3.777965528894897,
      "grad_norm": 3.3985698223114014,
      "learning_rate": 2e-05,
      "loss": 0.1362,
      "step": 5591
    },
    {
      "epoch": 3.778641432916526,
      "grad_norm": 6.095481872558594,
      "learning_rate": 2e-05,
      "loss": 0.2588,
      "step": 5592
    },
    {
      "epoch": 3.7793173369381545,
      "grad_norm": 10.438968658447266,
      "learning_rate": 2e-05,
      "loss": 0.1858,
      "step": 5593
    },
    {
      "epoch": 3.779993240959784,
      "grad_norm": 7.920506477355957,
      "learning_rate": 2e-05,
      "loss": 0.1848,
      "step": 5594
    },
    {
      "epoch": 3.7806691449814127,
      "grad_norm": 2.628777265548706,
      "learning_rate": 2e-05,
      "loss": 0.1873,
      "step": 5595
    },
    {
      "epoch": 3.7813450490030416,
      "grad_norm": 2.7718329429626465,
      "learning_rate": 2e-05,
      "loss": 0.1896,
      "step": 5596
    },
    {
      "epoch": 3.7820209530246705,
      "grad_norm": 6.691939830780029,
      "learning_rate": 2e-05,
      "loss": 0.1852,
      "step": 5597
    },
    {
      "epoch": 3.7826968570462993,
      "grad_norm": 4.4716715812683105,
      "learning_rate": 2e-05,
      "loss": 0.1488,
      "step": 5598
    },
    {
      "epoch": 3.783372761067928,
      "grad_norm": 5.375904083251953,
      "learning_rate": 2e-05,
      "loss": 0.1416,
      "step": 5599
    },
    {
      "epoch": 3.784048665089557,
      "grad_norm": 4.9015727043151855,
      "learning_rate": 2e-05,
      "loss": 0.132,
      "step": 5600
    },
    {
      "epoch": 3.784048665089557,
      "eval_accuracy": 0.671875,
      "eval_loss": 0.23395214974880219,
      "eval_runtime": 49.471,
      "eval_samples_per_second": 478.665,
      "eval_steps_per_second": 14.958,
      "step": 5600
    },
    {
      "epoch": 3.7847245691111864,
      "grad_norm": 5.040184020996094,
      "learning_rate": 2e-05,
      "loss": 0.2232,
      "step": 5601
    },
    {
      "epoch": 3.7854004731328152,
      "grad_norm": 3.6073169708251953,
      "learning_rate": 2e-05,
      "loss": 0.1578,
      "step": 5602
    },
    {
      "epoch": 3.786076377154444,
      "grad_norm": 2.3679659366607666,
      "learning_rate": 2e-05,
      "loss": 0.2059,
      "step": 5603
    },
    {
      "epoch": 3.786752281176073,
      "grad_norm": 2.8214759826660156,
      "learning_rate": 2e-05,
      "loss": 0.1865,
      "step": 5604
    },
    {
      "epoch": 3.787428185197702,
      "grad_norm": 3.0279150009155273,
      "learning_rate": 2e-05,
      "loss": 0.1467,
      "step": 5605
    },
    {
      "epoch": 3.7881040892193307,
      "grad_norm": 2.5903143882751465,
      "learning_rate": 2e-05,
      "loss": 0.1451,
      "step": 5606
    },
    {
      "epoch": 3.7887799932409596,
      "grad_norm": 8.735489845275879,
      "learning_rate": 2e-05,
      "loss": 0.2439,
      "step": 5607
    },
    {
      "epoch": 3.789455897262589,
      "grad_norm": 2.4026997089385986,
      "learning_rate": 2e-05,
      "loss": 0.148,
      "step": 5608
    },
    {
      "epoch": 3.790131801284218,
      "grad_norm": 3.107203960418701,
      "learning_rate": 2e-05,
      "loss": 0.2273,
      "step": 5609
    },
    {
      "epoch": 3.7908077053058467,
      "grad_norm": 6.423748970031738,
      "learning_rate": 2e-05,
      "loss": 0.1729,
      "step": 5610
    },
    {
      "epoch": 3.7914836093274755,
      "grad_norm": 3.3737106323242188,
      "learning_rate": 2e-05,
      "loss": 0.1524,
      "step": 5611
    },
    {
      "epoch": 3.7921595133491044,
      "grad_norm": 4.697040557861328,
      "learning_rate": 2e-05,
      "loss": 0.1762,
      "step": 5612
    },
    {
      "epoch": 3.7928354173707333,
      "grad_norm": 1.9626059532165527,
      "learning_rate": 2e-05,
      "loss": 0.2132,
      "step": 5613
    },
    {
      "epoch": 3.793511321392362,
      "grad_norm": 4.053718090057373,
      "learning_rate": 2e-05,
      "loss": 0.1546,
      "step": 5614
    },
    {
      "epoch": 3.7941872254139914,
      "grad_norm": 5.126800060272217,
      "learning_rate": 2e-05,
      "loss": 0.1378,
      "step": 5615
    },
    {
      "epoch": 3.79486312943562,
      "grad_norm": 4.141162872314453,
      "learning_rate": 2e-05,
      "loss": 0.1627,
      "step": 5616
    },
    {
      "epoch": 3.795539033457249,
      "grad_norm": 2.4319229125976562,
      "learning_rate": 2e-05,
      "loss": 0.1452,
      "step": 5617
    },
    {
      "epoch": 3.796214937478878,
      "grad_norm": 2.0171353816986084,
      "learning_rate": 2e-05,
      "loss": 0.1791,
      "step": 5618
    },
    {
      "epoch": 3.796890841500507,
      "grad_norm": 8.556910514831543,
      "learning_rate": 2e-05,
      "loss": 0.1931,
      "step": 5619
    },
    {
      "epoch": 3.797566745522136,
      "grad_norm": 2.327991485595703,
      "learning_rate": 2e-05,
      "loss": 0.1228,
      "step": 5620
    },
    {
      "epoch": 3.7982426495437647,
      "grad_norm": 3.104430675506592,
      "learning_rate": 2e-05,
      "loss": 0.1255,
      "step": 5621
    },
    {
      "epoch": 3.798918553565394,
      "grad_norm": 3.0176842212677,
      "learning_rate": 2e-05,
      "loss": 0.1497,
      "step": 5622
    },
    {
      "epoch": 3.7995944575870224,
      "grad_norm": 3.0166401863098145,
      "learning_rate": 2e-05,
      "loss": 0.1919,
      "step": 5623
    },
    {
      "epoch": 3.8002703616086517,
      "grad_norm": 4.449126720428467,
      "learning_rate": 2e-05,
      "loss": 0.1463,
      "step": 5624
    },
    {
      "epoch": 3.8009462656302806,
      "grad_norm": 2.9433841705322266,
      "learning_rate": 2e-05,
      "loss": 0.2062,
      "step": 5625
    },
    {
      "epoch": 3.8016221696519095,
      "grad_norm": 2.8989264965057373,
      "learning_rate": 2e-05,
      "loss": 0.1495,
      "step": 5626
    },
    {
      "epoch": 3.8022980736735383,
      "grad_norm": 6.433858394622803,
      "learning_rate": 2e-05,
      "loss": 0.1353,
      "step": 5627
    },
    {
      "epoch": 3.802973977695167,
      "grad_norm": 5.968183517456055,
      "learning_rate": 2e-05,
      "loss": 0.1493,
      "step": 5628
    },
    {
      "epoch": 3.803649881716796,
      "grad_norm": 2.344021797180176,
      "learning_rate": 2e-05,
      "loss": 0.1804,
      "step": 5629
    },
    {
      "epoch": 3.804325785738425,
      "grad_norm": 3.153200387954712,
      "learning_rate": 2e-05,
      "loss": 0.1696,
      "step": 5630
    },
    {
      "epoch": 3.8050016897600543,
      "grad_norm": 2.8074796199798584,
      "learning_rate": 2e-05,
      "loss": 0.132,
      "step": 5631
    },
    {
      "epoch": 3.805677593781683,
      "grad_norm": 2.8015406131744385,
      "learning_rate": 2e-05,
      "loss": 0.153,
      "step": 5632
    },
    {
      "epoch": 3.806353497803312,
      "grad_norm": 2.050065517425537,
      "learning_rate": 2e-05,
      "loss": 0.1616,
      "step": 5633
    },
    {
      "epoch": 3.807029401824941,
      "grad_norm": 5.835668563842773,
      "learning_rate": 2e-05,
      "loss": 0.1928,
      "step": 5634
    },
    {
      "epoch": 3.8077053058465697,
      "grad_norm": 4.328991889953613,
      "learning_rate": 2e-05,
      "loss": 0.1612,
      "step": 5635
    },
    {
      "epoch": 3.8083812098681986,
      "grad_norm": 3.8041810989379883,
      "learning_rate": 2e-05,
      "loss": 0.2149,
      "step": 5636
    },
    {
      "epoch": 3.8090571138898275,
      "grad_norm": 3.7471725940704346,
      "learning_rate": 2e-05,
      "loss": 0.1471,
      "step": 5637
    },
    {
      "epoch": 3.809733017911457,
      "grad_norm": 5.3610944747924805,
      "learning_rate": 2e-05,
      "loss": 0.2169,
      "step": 5638
    },
    {
      "epoch": 3.8104089219330852,
      "grad_norm": 4.694716453552246,
      "learning_rate": 2e-05,
      "loss": 0.1599,
      "step": 5639
    },
    {
      "epoch": 3.8110848259547145,
      "grad_norm": 3.4921252727508545,
      "learning_rate": 2e-05,
      "loss": 0.1562,
      "step": 5640
    },
    {
      "epoch": 3.8117607299763434,
      "grad_norm": 2.4167394638061523,
      "learning_rate": 2e-05,
      "loss": 0.1764,
      "step": 5641
    },
    {
      "epoch": 3.8124366339979723,
      "grad_norm": 2.682955741882324,
      "learning_rate": 2e-05,
      "loss": 0.1886,
      "step": 5642
    },
    {
      "epoch": 3.813112538019601,
      "grad_norm": 6.318434715270996,
      "learning_rate": 2e-05,
      "loss": 0.1305,
      "step": 5643
    },
    {
      "epoch": 3.81378844204123,
      "grad_norm": 4.318447589874268,
      "learning_rate": 2e-05,
      "loss": 0.2031,
      "step": 5644
    },
    {
      "epoch": 3.8144643460628593,
      "grad_norm": 8.276395797729492,
      "learning_rate": 2e-05,
      "loss": 0.1856,
      "step": 5645
    },
    {
      "epoch": 3.8151402500844878,
      "grad_norm": 4.36600399017334,
      "learning_rate": 2e-05,
      "loss": 0.1534,
      "step": 5646
    },
    {
      "epoch": 3.815816154106117,
      "grad_norm": 3.873584508895874,
      "learning_rate": 2e-05,
      "loss": 0.1523,
      "step": 5647
    },
    {
      "epoch": 3.816492058127746,
      "grad_norm": 4.336427688598633,
      "learning_rate": 2e-05,
      "loss": 0.0914,
      "step": 5648
    },
    {
      "epoch": 3.817167962149375,
      "grad_norm": 6.744791507720947,
      "learning_rate": 2e-05,
      "loss": 0.2128,
      "step": 5649
    },
    {
      "epoch": 3.8178438661710037,
      "grad_norm": 4.141669273376465,
      "learning_rate": 2e-05,
      "loss": 0.2017,
      "step": 5650
    },
    {
      "epoch": 3.8178438661710037,
      "eval_accuracy": 0.7364020270270271,
      "eval_loss": 0.24354617297649384,
      "eval_runtime": 50.4857,
      "eval_samples_per_second": 469.044,
      "eval_steps_per_second": 14.658,
      "step": 5650
    },
    {
      "epoch": 3.8185197701926326,
      "grad_norm": 9.513885498046875,
      "learning_rate": 2e-05,
      "loss": 0.1416,
      "step": 5651
    },
    {
      "epoch": 3.8191956742142614,
      "grad_norm": 3.335390329360962,
      "learning_rate": 2e-05,
      "loss": 0.1507,
      "step": 5652
    },
    {
      "epoch": 3.8198715782358903,
      "grad_norm": 2.1856276988983154,
      "learning_rate": 2e-05,
      "loss": 0.13,
      "step": 5653
    },
    {
      "epoch": 3.8205474822575196,
      "grad_norm": 2.800313711166382,
      "learning_rate": 2e-05,
      "loss": 0.0929,
      "step": 5654
    },
    {
      "epoch": 3.8212233862791485,
      "grad_norm": 5.578930854797363,
      "learning_rate": 2e-05,
      "loss": 0.2058,
      "step": 5655
    },
    {
      "epoch": 3.8218992903007774,
      "grad_norm": 3.0266458988189697,
      "learning_rate": 2e-05,
      "loss": 0.2059,
      "step": 5656
    },
    {
      "epoch": 3.8225751943224062,
      "grad_norm": 10.055424690246582,
      "learning_rate": 2e-05,
      "loss": 0.2135,
      "step": 5657
    },
    {
      "epoch": 3.823251098344035,
      "grad_norm": 10.517975807189941,
      "learning_rate": 2e-05,
      "loss": 0.2005,
      "step": 5658
    },
    {
      "epoch": 3.823927002365664,
      "grad_norm": 2.5231525897979736,
      "learning_rate": 2e-05,
      "loss": 0.1379,
      "step": 5659
    },
    {
      "epoch": 3.824602906387293,
      "grad_norm": 2.683443307876587,
      "learning_rate": 2e-05,
      "loss": 0.1577,
      "step": 5660
    },
    {
      "epoch": 3.825278810408922,
      "grad_norm": 3.693960666656494,
      "learning_rate": 2e-05,
      "loss": 0.1887,
      "step": 5661
    },
    {
      "epoch": 3.825954714430551,
      "grad_norm": 8.750907897949219,
      "learning_rate": 2e-05,
      "loss": 0.1494,
      "step": 5662
    },
    {
      "epoch": 3.82663061845218,
      "grad_norm": 7.523889064788818,
      "learning_rate": 2e-05,
      "loss": 0.121,
      "step": 5663
    },
    {
      "epoch": 3.8273065224738088,
      "grad_norm": 3.5814738273620605,
      "learning_rate": 2e-05,
      "loss": 0.1817,
      "step": 5664
    },
    {
      "epoch": 3.8279824264954376,
      "grad_norm": 2.6326191425323486,
      "learning_rate": 2e-05,
      "loss": 0.1369,
      "step": 5665
    },
    {
      "epoch": 3.8286583305170665,
      "grad_norm": 3.2224245071411133,
      "learning_rate": 2e-05,
      "loss": 0.154,
      "step": 5666
    },
    {
      "epoch": 3.8293342345386954,
      "grad_norm": 5.295842170715332,
      "learning_rate": 2e-05,
      "loss": 0.1653,
      "step": 5667
    },
    {
      "epoch": 3.8300101385603247,
      "grad_norm": 3.8523244857788086,
      "learning_rate": 2e-05,
      "loss": 0.1013,
      "step": 5668
    },
    {
      "epoch": 3.830686042581953,
      "grad_norm": 9.036855697631836,
      "learning_rate": 2e-05,
      "loss": 0.275,
      "step": 5669
    },
    {
      "epoch": 3.8313619466035824,
      "grad_norm": 5.579338550567627,
      "learning_rate": 2e-05,
      "loss": 0.2292,
      "step": 5670
    },
    {
      "epoch": 3.8320378506252113,
      "grad_norm": 5.501974105834961,
      "learning_rate": 2e-05,
      "loss": 0.1916,
      "step": 5671
    },
    {
      "epoch": 3.83271375464684,
      "grad_norm": 6.699617385864258,
      "learning_rate": 2e-05,
      "loss": 0.1718,
      "step": 5672
    },
    {
      "epoch": 3.833389658668469,
      "grad_norm": 2.787403106689453,
      "learning_rate": 2e-05,
      "loss": 0.1812,
      "step": 5673
    },
    {
      "epoch": 3.834065562690098,
      "grad_norm": 2.32252836227417,
      "learning_rate": 2e-05,
      "loss": 0.1698,
      "step": 5674
    },
    {
      "epoch": 3.8347414667117268,
      "grad_norm": 11.525341033935547,
      "learning_rate": 2e-05,
      "loss": 0.1979,
      "step": 5675
    },
    {
      "epoch": 3.8354173707333556,
      "grad_norm": 2.108736991882324,
      "learning_rate": 2e-05,
      "loss": 0.1412,
      "step": 5676
    },
    {
      "epoch": 3.836093274754985,
      "grad_norm": 6.514559745788574,
      "learning_rate": 2e-05,
      "loss": 0.1501,
      "step": 5677
    },
    {
      "epoch": 3.836769178776614,
      "grad_norm": 2.698448657989502,
      "learning_rate": 2e-05,
      "loss": 0.122,
      "step": 5678
    },
    {
      "epoch": 3.8374450827982427,
      "grad_norm": 1.9345006942749023,
      "learning_rate": 2e-05,
      "loss": 0.1483,
      "step": 5679
    },
    {
      "epoch": 3.8381209868198716,
      "grad_norm": 2.9722023010253906,
      "learning_rate": 2e-05,
      "loss": 0.1498,
      "step": 5680
    },
    {
      "epoch": 3.8387968908415004,
      "grad_norm": 1.981825590133667,
      "learning_rate": 2e-05,
      "loss": 0.1787,
      "step": 5681
    },
    {
      "epoch": 3.8394727948631293,
      "grad_norm": 1.897842526435852,
      "learning_rate": 2e-05,
      "loss": 0.1338,
      "step": 5682
    },
    {
      "epoch": 3.840148698884758,
      "grad_norm": 4.657876491546631,
      "learning_rate": 2e-05,
      "loss": 0.2101,
      "step": 5683
    },
    {
      "epoch": 3.8408246029063875,
      "grad_norm": 4.506203651428223,
      "learning_rate": 2e-05,
      "loss": 0.2479,
      "step": 5684
    },
    {
      "epoch": 3.8415005069280164,
      "grad_norm": 6.260153293609619,
      "learning_rate": 2e-05,
      "loss": 0.1249,
      "step": 5685
    },
    {
      "epoch": 3.8421764109496452,
      "grad_norm": 2.533811330795288,
      "learning_rate": 2e-05,
      "loss": 0.1482,
      "step": 5686
    },
    {
      "epoch": 3.842852314971274,
      "grad_norm": 3.0415432453155518,
      "learning_rate": 2e-05,
      "loss": 0.1873,
      "step": 5687
    },
    {
      "epoch": 3.843528218992903,
      "grad_norm": 2.624488592147827,
      "learning_rate": 2e-05,
      "loss": 0.1953,
      "step": 5688
    },
    {
      "epoch": 3.844204123014532,
      "grad_norm": 13.369369506835938,
      "learning_rate": 2e-05,
      "loss": 0.1879,
      "step": 5689
    },
    {
      "epoch": 3.8448800270361607,
      "grad_norm": 4.937232971191406,
      "learning_rate": 2e-05,
      "loss": 0.1795,
      "step": 5690
    },
    {
      "epoch": 3.84555593105779,
      "grad_norm": 3.723703622817993,
      "learning_rate": 2e-05,
      "loss": 0.1714,
      "step": 5691
    },
    {
      "epoch": 3.8462318350794185,
      "grad_norm": 3.476933717727661,
      "learning_rate": 2e-05,
      "loss": 0.1788,
      "step": 5692
    },
    {
      "epoch": 3.8469077391010478,
      "grad_norm": 5.877937316894531,
      "learning_rate": 2e-05,
      "loss": 0.1407,
      "step": 5693
    },
    {
      "epoch": 3.8475836431226766,
      "grad_norm": 9.5213041305542,
      "learning_rate": 2e-05,
      "loss": 0.1286,
      "step": 5694
    },
    {
      "epoch": 3.8482595471443055,
      "grad_norm": 7.075664043426514,
      "learning_rate": 2e-05,
      "loss": 0.1679,
      "step": 5695
    },
    {
      "epoch": 3.8489354511659344,
      "grad_norm": 2.5249619483947754,
      "learning_rate": 2e-05,
      "loss": 0.1517,
      "step": 5696
    },
    {
      "epoch": 3.8496113551875633,
      "grad_norm": 2.3462588787078857,
      "learning_rate": 2e-05,
      "loss": 0.2132,
      "step": 5697
    },
    {
      "epoch": 3.8502872592091926,
      "grad_norm": 7.990818023681641,
      "learning_rate": 2e-05,
      "loss": 0.1816,
      "step": 5698
    },
    {
      "epoch": 3.850963163230821,
      "grad_norm": 7.597602367401123,
      "learning_rate": 2e-05,
      "loss": 0.1432,
      "step": 5699
    },
    {
      "epoch": 3.8516390672524503,
      "grad_norm": 6.010751724243164,
      "learning_rate": 2e-05,
      "loss": 0.1913,
      "step": 5700
    },
    {
      "epoch": 3.8516390672524503,
      "eval_accuracy": 0.7353462837837837,
      "eval_loss": 0.2444404810667038,
      "eval_runtime": 47.9088,
      "eval_samples_per_second": 494.272,
      "eval_steps_per_second": 15.446,
      "step": 5700
    },
    {
      "epoch": 3.852314971274079,
      "grad_norm": 5.740936279296875,
      "learning_rate": 2e-05,
      "loss": 0.1547,
      "step": 5701
    },
    {
      "epoch": 3.852990875295708,
      "grad_norm": 3.657909393310547,
      "learning_rate": 2e-05,
      "loss": 0.1597,
      "step": 5702
    },
    {
      "epoch": 3.853666779317337,
      "grad_norm": 5.504480361938477,
      "learning_rate": 2e-05,
      "loss": 0.1902,
      "step": 5703
    },
    {
      "epoch": 3.854342683338966,
      "grad_norm": 6.276440620422363,
      "learning_rate": 2e-05,
      "loss": 0.2386,
      "step": 5704
    },
    {
      "epoch": 3.8550185873605947,
      "grad_norm": 2.821690559387207,
      "learning_rate": 2e-05,
      "loss": 0.15,
      "step": 5705
    },
    {
      "epoch": 3.8556944913822235,
      "grad_norm": 3.3567726612091064,
      "learning_rate": 2e-05,
      "loss": 0.2511,
      "step": 5706
    },
    {
      "epoch": 3.856370395403853,
      "grad_norm": 3.0726091861724854,
      "learning_rate": 2e-05,
      "loss": 0.2527,
      "step": 5707
    },
    {
      "epoch": 3.8570462994254817,
      "grad_norm": 4.235030651092529,
      "learning_rate": 2e-05,
      "loss": 0.1692,
      "step": 5708
    },
    {
      "epoch": 3.8577222034471106,
      "grad_norm": 2.493828773498535,
      "learning_rate": 2e-05,
      "loss": 0.1361,
      "step": 5709
    },
    {
      "epoch": 3.8583981074687395,
      "grad_norm": 3.52876615524292,
      "learning_rate": 2e-05,
      "loss": 0.2202,
      "step": 5710
    },
    {
      "epoch": 3.8590740114903683,
      "grad_norm": 2.69331693649292,
      "learning_rate": 2e-05,
      "loss": 0.1962,
      "step": 5711
    },
    {
      "epoch": 3.859749915511997,
      "grad_norm": 6.78504753112793,
      "learning_rate": 2e-05,
      "loss": 0.1158,
      "step": 5712
    },
    {
      "epoch": 3.860425819533626,
      "grad_norm": 2.381026029586792,
      "learning_rate": 2e-05,
      "loss": 0.2599,
      "step": 5713
    },
    {
      "epoch": 3.8611017235552554,
      "grad_norm": 6.739652633666992,
      "learning_rate": 2e-05,
      "loss": 0.1475,
      "step": 5714
    },
    {
      "epoch": 3.861777627576884,
      "grad_norm": 4.243177890777588,
      "learning_rate": 2e-05,
      "loss": 0.2008,
      "step": 5715
    },
    {
      "epoch": 3.862453531598513,
      "grad_norm": 5.960952281951904,
      "learning_rate": 2e-05,
      "loss": 0.1592,
      "step": 5716
    },
    {
      "epoch": 3.863129435620142,
      "grad_norm": 2.905592918395996,
      "learning_rate": 2e-05,
      "loss": 0.2313,
      "step": 5717
    },
    {
      "epoch": 3.863805339641771,
      "grad_norm": 6.513103008270264,
      "learning_rate": 2e-05,
      "loss": 0.1849,
      "step": 5718
    },
    {
      "epoch": 3.8644812436633997,
      "grad_norm": 5.819515228271484,
      "learning_rate": 2e-05,
      "loss": 0.2212,
      "step": 5719
    },
    {
      "epoch": 3.8651571476850286,
      "grad_norm": 4.013911724090576,
      "learning_rate": 2e-05,
      "loss": 0.1251,
      "step": 5720
    },
    {
      "epoch": 3.865833051706658,
      "grad_norm": 3.590052843093872,
      "learning_rate": 2e-05,
      "loss": 0.1282,
      "step": 5721
    },
    {
      "epoch": 3.8665089557282863,
      "grad_norm": 6.80195426940918,
      "learning_rate": 2e-05,
      "loss": 0.189,
      "step": 5722
    },
    {
      "epoch": 3.8671848597499157,
      "grad_norm": 2.6624982357025146,
      "learning_rate": 2e-05,
      "loss": 0.158,
      "step": 5723
    },
    {
      "epoch": 3.8678607637715445,
      "grad_norm": 1.7099113464355469,
      "learning_rate": 2e-05,
      "loss": 0.1349,
      "step": 5724
    },
    {
      "epoch": 3.8685366677931734,
      "grad_norm": 3.4036190509796143,
      "learning_rate": 2e-05,
      "loss": 0.1385,
      "step": 5725
    },
    {
      "epoch": 3.8692125718148023,
      "grad_norm": 2.445984363555908,
      "learning_rate": 2e-05,
      "loss": 0.2277,
      "step": 5726
    },
    {
      "epoch": 3.869888475836431,
      "grad_norm": 8.93860149383545,
      "learning_rate": 2e-05,
      "loss": 0.2201,
      "step": 5727
    },
    {
      "epoch": 3.87056437985806,
      "grad_norm": 3.684041976928711,
      "learning_rate": 2e-05,
      "loss": 0.1468,
      "step": 5728
    },
    {
      "epoch": 3.871240283879689,
      "grad_norm": 8.037984848022461,
      "learning_rate": 2e-05,
      "loss": 0.2204,
      "step": 5729
    },
    {
      "epoch": 3.871916187901318,
      "grad_norm": 4.304277420043945,
      "learning_rate": 2e-05,
      "loss": 0.1668,
      "step": 5730
    },
    {
      "epoch": 3.872592091922947,
      "grad_norm": 3.4956212043762207,
      "learning_rate": 2e-05,
      "loss": 0.253,
      "step": 5731
    },
    {
      "epoch": 3.873267995944576,
      "grad_norm": 2.4128618240356445,
      "learning_rate": 2e-05,
      "loss": 0.1894,
      "step": 5732
    },
    {
      "epoch": 3.873943899966205,
      "grad_norm": 2.3132872581481934,
      "learning_rate": 2e-05,
      "loss": 0.1711,
      "step": 5733
    },
    {
      "epoch": 3.8746198039878337,
      "grad_norm": 4.592403888702393,
      "learning_rate": 2e-05,
      "loss": 0.1724,
      "step": 5734
    },
    {
      "epoch": 3.8752957080094625,
      "grad_norm": 6.283106803894043,
      "learning_rate": 2e-05,
      "loss": 0.2012,
      "step": 5735
    },
    {
      "epoch": 3.8759716120310914,
      "grad_norm": 10.576272010803223,
      "learning_rate": 2e-05,
      "loss": 0.1926,
      "step": 5736
    },
    {
      "epoch": 3.8766475160527207,
      "grad_norm": 4.744899749755859,
      "learning_rate": 2e-05,
      "loss": 0.1468,
      "step": 5737
    },
    {
      "epoch": 3.8773234200743496,
      "grad_norm": 6.916989326477051,
      "learning_rate": 2e-05,
      "loss": 0.1739,
      "step": 5738
    },
    {
      "epoch": 3.8779993240959785,
      "grad_norm": 2.236283779144287,
      "learning_rate": 2e-05,
      "loss": 0.1862,
      "step": 5739
    },
    {
      "epoch": 3.8786752281176073,
      "grad_norm": 2.555872917175293,
      "learning_rate": 2e-05,
      "loss": 0.1747,
      "step": 5740
    },
    {
      "epoch": 3.879351132139236,
      "grad_norm": 1.7132920026779175,
      "learning_rate": 2e-05,
      "loss": 0.109,
      "step": 5741
    },
    {
      "epoch": 3.880027036160865,
      "grad_norm": 4.707760334014893,
      "learning_rate": 2e-05,
      "loss": 0.135,
      "step": 5742
    },
    {
      "epoch": 3.880702940182494,
      "grad_norm": 2.3328189849853516,
      "learning_rate": 2e-05,
      "loss": 0.1853,
      "step": 5743
    },
    {
      "epoch": 3.8813788442041233,
      "grad_norm": 7.06760311126709,
      "learning_rate": 2e-05,
      "loss": 0.1416,
      "step": 5744
    },
    {
      "epoch": 3.8820547482257517,
      "grad_norm": 3.2310714721679688,
      "learning_rate": 2e-05,
      "loss": 0.2121,
      "step": 5745
    },
    {
      "epoch": 3.882730652247381,
      "grad_norm": 5.5370330810546875,
      "learning_rate": 2e-05,
      "loss": 0.1066,
      "step": 5746
    },
    {
      "epoch": 3.88340655626901,
      "grad_norm": 3.889495611190796,
      "learning_rate": 2e-05,
      "loss": 0.2096,
      "step": 5747
    },
    {
      "epoch": 3.8840824602906387,
      "grad_norm": 2.122300624847412,
      "learning_rate": 2e-05,
      "loss": 0.1713,
      "step": 5748
    },
    {
      "epoch": 3.8847583643122676,
      "grad_norm": 4.503293991088867,
      "learning_rate": 2e-05,
      "loss": 0.1735,
      "step": 5749
    },
    {
      "epoch": 3.8854342683338965,
      "grad_norm": 4.463473320007324,
      "learning_rate": 2e-05,
      "loss": 0.1625,
      "step": 5750
    },
    {
      "epoch": 3.8854342683338965,
      "eval_accuracy": 0.6616554054054054,
      "eval_loss": 0.24359776079654694,
      "eval_runtime": 58.9507,
      "eval_samples_per_second": 401.692,
      "eval_steps_per_second": 12.553,
      "step": 5750
    },
    {
      "epoch": 3.8861101723555254,
      "grad_norm": 3.6970629692077637,
      "learning_rate": 2e-05,
      "loss": 0.2544,
      "step": 5751
    },
    {
      "epoch": 3.8867860763771542,
      "grad_norm": 14.733530044555664,
      "learning_rate": 2e-05,
      "loss": 0.272,
      "step": 5752
    },
    {
      "epoch": 3.8874619803987835,
      "grad_norm": 3.3533337116241455,
      "learning_rate": 2e-05,
      "loss": 0.1751,
      "step": 5753
    },
    {
      "epoch": 3.8881378844204124,
      "grad_norm": 3.717526435852051,
      "learning_rate": 2e-05,
      "loss": 0.1584,
      "step": 5754
    },
    {
      "epoch": 3.8888137884420413,
      "grad_norm": 7.995940208435059,
      "learning_rate": 2e-05,
      "loss": 0.1673,
      "step": 5755
    },
    {
      "epoch": 3.88948969246367,
      "grad_norm": 1.6178697347640991,
      "learning_rate": 2e-05,
      "loss": 0.1269,
      "step": 5756
    },
    {
      "epoch": 3.890165596485299,
      "grad_norm": 4.630681037902832,
      "learning_rate": 2e-05,
      "loss": 0.2532,
      "step": 5757
    },
    {
      "epoch": 3.890841500506928,
      "grad_norm": 7.3389387130737305,
      "learning_rate": 2e-05,
      "loss": 0.2013,
      "step": 5758
    },
    {
      "epoch": 3.8915174045285568,
      "grad_norm": 8.188896179199219,
      "learning_rate": 2e-05,
      "loss": 0.1545,
      "step": 5759
    },
    {
      "epoch": 3.892193308550186,
      "grad_norm": 2.1525492668151855,
      "learning_rate": 2e-05,
      "loss": 0.1188,
      "step": 5760
    },
    {
      "epoch": 3.892869212571815,
      "grad_norm": 2.6920740604400635,
      "learning_rate": 2e-05,
      "loss": 0.1564,
      "step": 5761
    },
    {
      "epoch": 3.893545116593444,
      "grad_norm": 6.738633632659912,
      "learning_rate": 2e-05,
      "loss": 0.2129,
      "step": 5762
    },
    {
      "epoch": 3.8942210206150727,
      "grad_norm": 4.944591045379639,
      "learning_rate": 2e-05,
      "loss": 0.2804,
      "step": 5763
    },
    {
      "epoch": 3.8948969246367016,
      "grad_norm": 7.0334858894348145,
      "learning_rate": 2e-05,
      "loss": 0.2315,
      "step": 5764
    },
    {
      "epoch": 3.8955728286583304,
      "grad_norm": 2.3782706260681152,
      "learning_rate": 2e-05,
      "loss": 0.1682,
      "step": 5765
    },
    {
      "epoch": 3.8962487326799593,
      "grad_norm": 3.7437126636505127,
      "learning_rate": 2e-05,
      "loss": 0.1795,
      "step": 5766
    },
    {
      "epoch": 3.8969246367015886,
      "grad_norm": 2.8713512420654297,
      "learning_rate": 2e-05,
      "loss": 0.1553,
      "step": 5767
    },
    {
      "epoch": 3.897600540723217,
      "grad_norm": 4.259865760803223,
      "learning_rate": 2e-05,
      "loss": 0.1439,
      "step": 5768
    },
    {
      "epoch": 3.8982764447448464,
      "grad_norm": 1.910573124885559,
      "learning_rate": 2e-05,
      "loss": 0.2074,
      "step": 5769
    },
    {
      "epoch": 3.8989523487664752,
      "grad_norm": 4.288532257080078,
      "learning_rate": 2e-05,
      "loss": 0.1654,
      "step": 5770
    },
    {
      "epoch": 3.899628252788104,
      "grad_norm": 1.7777173519134521,
      "learning_rate": 2e-05,
      "loss": 0.152,
      "step": 5771
    },
    {
      "epoch": 3.900304156809733,
      "grad_norm": 3.738985300064087,
      "learning_rate": 2e-05,
      "loss": 0.1495,
      "step": 5772
    },
    {
      "epoch": 3.900980060831362,
      "grad_norm": 3.5182642936706543,
      "learning_rate": 2e-05,
      "loss": 0.1019,
      "step": 5773
    },
    {
      "epoch": 3.9016559648529907,
      "grad_norm": 2.8880198001861572,
      "learning_rate": 2e-05,
      "loss": 0.1698,
      "step": 5774
    },
    {
      "epoch": 3.9023318688746196,
      "grad_norm": 4.299959659576416,
      "learning_rate": 2e-05,
      "loss": 0.1533,
      "step": 5775
    },
    {
      "epoch": 3.903007772896249,
      "grad_norm": 1.840120792388916,
      "learning_rate": 2e-05,
      "loss": 0.1498,
      "step": 5776
    },
    {
      "epoch": 3.9036836769178778,
      "grad_norm": 3.1124253273010254,
      "learning_rate": 2e-05,
      "loss": 0.1259,
      "step": 5777
    },
    {
      "epoch": 3.9043595809395066,
      "grad_norm": 3.0232479572296143,
      "learning_rate": 2e-05,
      "loss": 0.1716,
      "step": 5778
    },
    {
      "epoch": 3.9050354849611355,
      "grad_norm": 3.023501396179199,
      "learning_rate": 2e-05,
      "loss": 0.108,
      "step": 5779
    },
    {
      "epoch": 3.9057113889827644,
      "grad_norm": 3.3342268466949463,
      "learning_rate": 2e-05,
      "loss": 0.1517,
      "step": 5780
    },
    {
      "epoch": 3.9063872930043932,
      "grad_norm": 3.6894497871398926,
      "learning_rate": 2e-05,
      "loss": 0.1329,
      "step": 5781
    },
    {
      "epoch": 3.907063197026022,
      "grad_norm": 2.560276508331299,
      "learning_rate": 2e-05,
      "loss": 0.1681,
      "step": 5782
    },
    {
      "epoch": 3.9077391010476514,
      "grad_norm": 2.0417377948760986,
      "learning_rate": 2e-05,
      "loss": 0.1686,
      "step": 5783
    },
    {
      "epoch": 3.9084150050692803,
      "grad_norm": 3.873534917831421,
      "learning_rate": 2e-05,
      "loss": 0.2308,
      "step": 5784
    },
    {
      "epoch": 3.909090909090909,
      "grad_norm": 6.275103569030762,
      "learning_rate": 2e-05,
      "loss": 0.1621,
      "step": 5785
    },
    {
      "epoch": 3.909766813112538,
      "grad_norm": 1.9719719886779785,
      "learning_rate": 2e-05,
      "loss": 0.1633,
      "step": 5786
    },
    {
      "epoch": 3.910442717134167,
      "grad_norm": 5.768387794494629,
      "learning_rate": 2e-05,
      "loss": 0.1791,
      "step": 5787
    },
    {
      "epoch": 3.911118621155796,
      "grad_norm": 3.8229830265045166,
      "learning_rate": 2e-05,
      "loss": 0.1422,
      "step": 5788
    },
    {
      "epoch": 3.9117945251774247,
      "grad_norm": 4.996890544891357,
      "learning_rate": 2e-05,
      "loss": 0.1291,
      "step": 5789
    },
    {
      "epoch": 3.912470429199054,
      "grad_norm": 3.0447962284088135,
      "learning_rate": 2e-05,
      "loss": 0.2008,
      "step": 5790
    },
    {
      "epoch": 3.9131463332206824,
      "grad_norm": 7.146088600158691,
      "learning_rate": 2e-05,
      "loss": 0.1591,
      "step": 5791
    },
    {
      "epoch": 3.9138222372423117,
      "grad_norm": 3.811136484146118,
      "learning_rate": 2e-05,
      "loss": 0.1307,
      "step": 5792
    },
    {
      "epoch": 3.9144981412639406,
      "grad_norm": 5.977417945861816,
      "learning_rate": 2e-05,
      "loss": 0.1745,
      "step": 5793
    },
    {
      "epoch": 3.9151740452855694,
      "grad_norm": 4.188277721405029,
      "learning_rate": 2e-05,
      "loss": 0.184,
      "step": 5794
    },
    {
      "epoch": 3.9158499493071983,
      "grad_norm": 7.912221908569336,
      "learning_rate": 2e-05,
      "loss": 0.1639,
      "step": 5795
    },
    {
      "epoch": 3.916525853328827,
      "grad_norm": 5.3748459815979,
      "learning_rate": 2e-05,
      "loss": 0.1655,
      "step": 5796
    },
    {
      "epoch": 3.9172017573504565,
      "grad_norm": 2.43931245803833,
      "learning_rate": 2e-05,
      "loss": 0.1563,
      "step": 5797
    },
    {
      "epoch": 3.917877661372085,
      "grad_norm": 6.042451858520508,
      "learning_rate": 2e-05,
      "loss": 0.2027,
      "step": 5798
    },
    {
      "epoch": 3.9185535653937142,
      "grad_norm": 3.4596478939056396,
      "learning_rate": 2e-05,
      "loss": 0.1502,
      "step": 5799
    },
    {
      "epoch": 3.919229469415343,
      "grad_norm": 2.7858033180236816,
      "learning_rate": 2e-05,
      "loss": 0.1439,
      "step": 5800
    },
    {
      "epoch": 3.919229469415343,
      "eval_accuracy": 0.7075168918918919,
      "eval_loss": 0.22150936722755432,
      "eval_runtime": 63.2073,
      "eval_samples_per_second": 374.641,
      "eval_steps_per_second": 11.708,
      "step": 5800
    },
    {
      "epoch": 3.919905373436972,
      "grad_norm": 4.989176273345947,
      "learning_rate": 2e-05,
      "loss": 0.1508,
      "step": 5801
    },
    {
      "epoch": 3.920581277458601,
      "grad_norm": 2.1687748432159424,
      "learning_rate": 2e-05,
      "loss": 0.1541,
      "step": 5802
    },
    {
      "epoch": 3.9212571814802297,
      "grad_norm": 3.282290458679199,
      "learning_rate": 2e-05,
      "loss": 0.2096,
      "step": 5803
    },
    {
      "epoch": 3.9219330855018586,
      "grad_norm": 4.224964618682861,
      "learning_rate": 2e-05,
      "loss": 0.2102,
      "step": 5804
    },
    {
      "epoch": 3.9226089895234875,
      "grad_norm": 4.883566379547119,
      "learning_rate": 2e-05,
      "loss": 0.1213,
      "step": 5805
    },
    {
      "epoch": 3.923284893545117,
      "grad_norm": 7.953604698181152,
      "learning_rate": 2e-05,
      "loss": 0.1655,
      "step": 5806
    },
    {
      "epoch": 3.9239607975667457,
      "grad_norm": 2.1634202003479004,
      "learning_rate": 2e-05,
      "loss": 0.1717,
      "step": 5807
    },
    {
      "epoch": 3.9246367015883745,
      "grad_norm": 2.0296006202697754,
      "learning_rate": 2e-05,
      "loss": 0.1346,
      "step": 5808
    },
    {
      "epoch": 3.9253126056100034,
      "grad_norm": 5.452413558959961,
      "learning_rate": 2e-05,
      "loss": 0.2204,
      "step": 5809
    },
    {
      "epoch": 3.9259885096316323,
      "grad_norm": 7.056030750274658,
      "learning_rate": 2e-05,
      "loss": 0.1619,
      "step": 5810
    },
    {
      "epoch": 3.926664413653261,
      "grad_norm": 7.182498931884766,
      "learning_rate": 2e-05,
      "loss": 0.2285,
      "step": 5811
    },
    {
      "epoch": 3.92734031767489,
      "grad_norm": 4.875061988830566,
      "learning_rate": 2e-05,
      "loss": 0.1206,
      "step": 5812
    },
    {
      "epoch": 3.9280162216965193,
      "grad_norm": 3.963660717010498,
      "learning_rate": 2e-05,
      "loss": 0.1483,
      "step": 5813
    },
    {
      "epoch": 3.928692125718148,
      "grad_norm": 2.7260732650756836,
      "learning_rate": 2e-05,
      "loss": 0.111,
      "step": 5814
    },
    {
      "epoch": 3.929368029739777,
      "grad_norm": 3.0561392307281494,
      "learning_rate": 2e-05,
      "loss": 0.1695,
      "step": 5815
    },
    {
      "epoch": 3.930043933761406,
      "grad_norm": 2.2434396743774414,
      "learning_rate": 2e-05,
      "loss": 0.1799,
      "step": 5816
    },
    {
      "epoch": 3.930719837783035,
      "grad_norm": 2.2586729526519775,
      "learning_rate": 2e-05,
      "loss": 0.1592,
      "step": 5817
    },
    {
      "epoch": 3.9313957418046637,
      "grad_norm": 5.863203048706055,
      "learning_rate": 2e-05,
      "loss": 0.1883,
      "step": 5818
    },
    {
      "epoch": 3.9320716458262925,
      "grad_norm": 4.0152201652526855,
      "learning_rate": 2e-05,
      "loss": 0.122,
      "step": 5819
    },
    {
      "epoch": 3.932747549847922,
      "grad_norm": 2.07881498336792,
      "learning_rate": 2e-05,
      "loss": 0.1601,
      "step": 5820
    },
    {
      "epoch": 3.9334234538695503,
      "grad_norm": 4.682923316955566,
      "learning_rate": 2e-05,
      "loss": 0.1148,
      "step": 5821
    },
    {
      "epoch": 3.9340993578911796,
      "grad_norm": 2.835958480834961,
      "learning_rate": 2e-05,
      "loss": 0.186,
      "step": 5822
    },
    {
      "epoch": 3.9347752619128085,
      "grad_norm": 7.254109859466553,
      "learning_rate": 2e-05,
      "loss": 0.1729,
      "step": 5823
    },
    {
      "epoch": 3.9354511659344373,
      "grad_norm": 2.590785264968872,
      "learning_rate": 2e-05,
      "loss": 0.1567,
      "step": 5824
    },
    {
      "epoch": 3.936127069956066,
      "grad_norm": 2.6141703128814697,
      "learning_rate": 2e-05,
      "loss": 0.1567,
      "step": 5825
    },
    {
      "epoch": 3.936802973977695,
      "grad_norm": 3.501641273498535,
      "learning_rate": 2e-05,
      "loss": 0.1303,
      "step": 5826
    },
    {
      "epoch": 3.937478877999324,
      "grad_norm": 1.6611804962158203,
      "learning_rate": 2e-05,
      "loss": 0.1639,
      "step": 5827
    },
    {
      "epoch": 3.938154782020953,
      "grad_norm": 4.9098124504089355,
      "learning_rate": 2e-05,
      "loss": 0.1379,
      "step": 5828
    },
    {
      "epoch": 3.938830686042582,
      "grad_norm": 2.39231276512146,
      "learning_rate": 2e-05,
      "loss": 0.214,
      "step": 5829
    },
    {
      "epoch": 3.939506590064211,
      "grad_norm": 8.389201164245605,
      "learning_rate": 2e-05,
      "loss": 0.1966,
      "step": 5830
    },
    {
      "epoch": 3.94018249408584,
      "grad_norm": 2.818854808807373,
      "learning_rate": 2e-05,
      "loss": 0.1297,
      "step": 5831
    },
    {
      "epoch": 3.9408583981074687,
      "grad_norm": 3.3658177852630615,
      "learning_rate": 2e-05,
      "loss": 0.1931,
      "step": 5832
    },
    {
      "epoch": 3.9415343021290976,
      "grad_norm": 4.119118690490723,
      "learning_rate": 2e-05,
      "loss": 0.1982,
      "step": 5833
    },
    {
      "epoch": 3.9422102061507265,
      "grad_norm": 3.7378480434417725,
      "learning_rate": 2e-05,
      "loss": 0.1949,
      "step": 5834
    },
    {
      "epoch": 3.9428861101723554,
      "grad_norm": 5.160818576812744,
      "learning_rate": 2e-05,
      "loss": 0.155,
      "step": 5835
    },
    {
      "epoch": 3.9435620141939847,
      "grad_norm": 2.6472175121307373,
      "learning_rate": 2e-05,
      "loss": 0.1748,
      "step": 5836
    },
    {
      "epoch": 3.9442379182156135,
      "grad_norm": 7.083293437957764,
      "learning_rate": 2e-05,
      "loss": 0.1678,
      "step": 5837
    },
    {
      "epoch": 3.9449138222372424,
      "grad_norm": 6.947316646575928,
      "learning_rate": 2e-05,
      "loss": 0.1969,
      "step": 5838
    },
    {
      "epoch": 3.9455897262588713,
      "grad_norm": 4.896764278411865,
      "learning_rate": 2e-05,
      "loss": 0.1751,
      "step": 5839
    },
    {
      "epoch": 3.9462656302805,
      "grad_norm": 4.193535804748535,
      "learning_rate": 2e-05,
      "loss": 0.1471,
      "step": 5840
    },
    {
      "epoch": 3.946941534302129,
      "grad_norm": 6.428092956542969,
      "learning_rate": 2e-05,
      "loss": 0.2245,
      "step": 5841
    },
    {
      "epoch": 3.947617438323758,
      "grad_norm": 4.107319355010986,
      "learning_rate": 2e-05,
      "loss": 0.1687,
      "step": 5842
    },
    {
      "epoch": 3.948293342345387,
      "grad_norm": 5.5815558433532715,
      "learning_rate": 2e-05,
      "loss": 0.1515,
      "step": 5843
    },
    {
      "epoch": 3.9489692463670156,
      "grad_norm": 3.30216908454895,
      "learning_rate": 2e-05,
      "loss": 0.1585,
      "step": 5844
    },
    {
      "epoch": 3.949645150388645,
      "grad_norm": 4.815157890319824,
      "learning_rate": 2e-05,
      "loss": 0.1933,
      "step": 5845
    },
    {
      "epoch": 3.950321054410274,
      "grad_norm": 4.184268474578857,
      "learning_rate": 2e-05,
      "loss": 0.1471,
      "step": 5846
    },
    {
      "epoch": 3.9509969584319027,
      "grad_norm": 1.8934928178787231,
      "learning_rate": 2e-05,
      "loss": 0.16,
      "step": 5847
    },
    {
      "epoch": 3.9516728624535316,
      "grad_norm": 6.584310054779053,
      "learning_rate": 2e-05,
      "loss": 0.1817,
      "step": 5848
    },
    {
      "epoch": 3.9523487664751604,
      "grad_norm": 2.1588854789733887,
      "learning_rate": 2e-05,
      "loss": 0.1916,
      "step": 5849
    },
    {
      "epoch": 3.9530246704967893,
      "grad_norm": 1.8811520338058472,
      "learning_rate": 2e-05,
      "loss": 0.1146,
      "step": 5850
    },
    {
      "epoch": 3.9530246704967893,
      "eval_accuracy": 0.7235219594594594,
      "eval_loss": 0.2295803427696228,
      "eval_runtime": 58.6636,
      "eval_samples_per_second": 403.658,
      "eval_steps_per_second": 12.614,
      "step": 5850
    },
    {
      "epoch": 3.953700574518418,
      "grad_norm": 5.478398323059082,
      "learning_rate": 2e-05,
      "loss": 0.2415,
      "step": 5851
    },
    {
      "epoch": 3.9543764785400475,
      "grad_norm": 2.1006104946136475,
      "learning_rate": 2e-05,
      "loss": 0.1663,
      "step": 5852
    },
    {
      "epoch": 3.9550523825616763,
      "grad_norm": 4.197693347930908,
      "learning_rate": 2e-05,
      "loss": 0.169,
      "step": 5853
    },
    {
      "epoch": 3.955728286583305,
      "grad_norm": 2.779141664505005,
      "learning_rate": 2e-05,
      "loss": 0.1741,
      "step": 5854
    },
    {
      "epoch": 3.956404190604934,
      "grad_norm": 4.015800952911377,
      "learning_rate": 2e-05,
      "loss": 0.1794,
      "step": 5855
    },
    {
      "epoch": 3.957080094626563,
      "grad_norm": 1.9453644752502441,
      "learning_rate": 2e-05,
      "loss": 0.1363,
      "step": 5856
    },
    {
      "epoch": 3.957755998648192,
      "grad_norm": 3.6242759227752686,
      "learning_rate": 2e-05,
      "loss": 0.1225,
      "step": 5857
    },
    {
      "epoch": 3.9584319026698207,
      "grad_norm": 2.0799176692962646,
      "learning_rate": 2e-05,
      "loss": 0.158,
      "step": 5858
    },
    {
      "epoch": 3.95910780669145,
      "grad_norm": 2.5391173362731934,
      "learning_rate": 2e-05,
      "loss": 0.1798,
      "step": 5859
    },
    {
      "epoch": 3.959783710713079,
      "grad_norm": 6.131653308868408,
      "learning_rate": 2e-05,
      "loss": 0.1754,
      "step": 5860
    },
    {
      "epoch": 3.9604596147347078,
      "grad_norm": 2.1152265071868896,
      "learning_rate": 2e-05,
      "loss": 0.1568,
      "step": 5861
    },
    {
      "epoch": 3.9611355187563366,
      "grad_norm": 4.695295810699463,
      "learning_rate": 2e-05,
      "loss": 0.1159,
      "step": 5862
    },
    {
      "epoch": 3.9618114227779655,
      "grad_norm": 2.395559310913086,
      "learning_rate": 2e-05,
      "loss": 0.1997,
      "step": 5863
    },
    {
      "epoch": 3.9624873267995944,
      "grad_norm": 3.9653196334838867,
      "learning_rate": 2e-05,
      "loss": 0.1635,
      "step": 5864
    },
    {
      "epoch": 3.9631632308212232,
      "grad_norm": 7.716347694396973,
      "learning_rate": 2e-05,
      "loss": 0.1976,
      "step": 5865
    },
    {
      "epoch": 3.9638391348428526,
      "grad_norm": 2.1239571571350098,
      "learning_rate": 2e-05,
      "loss": 0.1457,
      "step": 5866
    },
    {
      "epoch": 3.964515038864481,
      "grad_norm": 2.7195394039154053,
      "learning_rate": 2e-05,
      "loss": 0.1434,
      "step": 5867
    },
    {
      "epoch": 3.9651909428861103,
      "grad_norm": 2.927277088165283,
      "learning_rate": 2e-05,
      "loss": 0.177,
      "step": 5868
    },
    {
      "epoch": 3.965866846907739,
      "grad_norm": 5.098531723022461,
      "learning_rate": 2e-05,
      "loss": 0.0841,
      "step": 5869
    },
    {
      "epoch": 3.966542750929368,
      "grad_norm": 3.5897042751312256,
      "learning_rate": 2e-05,
      "loss": 0.1595,
      "step": 5870
    },
    {
      "epoch": 3.967218654950997,
      "grad_norm": 5.063659191131592,
      "learning_rate": 2e-05,
      "loss": 0.137,
      "step": 5871
    },
    {
      "epoch": 3.9678945589726258,
      "grad_norm": 2.1732211112976074,
      "learning_rate": 2e-05,
      "loss": 0.1734,
      "step": 5872
    },
    {
      "epoch": 3.968570462994255,
      "grad_norm": 4.322139263153076,
      "learning_rate": 2e-05,
      "loss": 0.1501,
      "step": 5873
    },
    {
      "epoch": 3.9692463670158835,
      "grad_norm": 3.479583263397217,
      "learning_rate": 2e-05,
      "loss": 0.1218,
      "step": 5874
    },
    {
      "epoch": 3.969922271037513,
      "grad_norm": 2.4653537273406982,
      "learning_rate": 2e-05,
      "loss": 0.1455,
      "step": 5875
    },
    {
      "epoch": 3.9705981750591417,
      "grad_norm": 7.397743225097656,
      "learning_rate": 2e-05,
      "loss": 0.1973,
      "step": 5876
    },
    {
      "epoch": 3.9712740790807706,
      "grad_norm": 9.434165000915527,
      "learning_rate": 2e-05,
      "loss": 0.1564,
      "step": 5877
    },
    {
      "epoch": 3.9719499831023994,
      "grad_norm": 4.7489237785339355,
      "learning_rate": 2e-05,
      "loss": 0.1498,
      "step": 5878
    },
    {
      "epoch": 3.9726258871240283,
      "grad_norm": 2.8160243034362793,
      "learning_rate": 2e-05,
      "loss": 0.1634,
      "step": 5879
    },
    {
      "epoch": 3.973301791145657,
      "grad_norm": 6.142129898071289,
      "learning_rate": 2e-05,
      "loss": 0.1975,
      "step": 5880
    },
    {
      "epoch": 3.973977695167286,
      "grad_norm": 8.268302917480469,
      "learning_rate": 2e-05,
      "loss": 0.2294,
      "step": 5881
    },
    {
      "epoch": 3.9746535991889154,
      "grad_norm": 5.631998062133789,
      "learning_rate": 2e-05,
      "loss": 0.1356,
      "step": 5882
    },
    {
      "epoch": 3.9753295032105442,
      "grad_norm": 2.763558864593506,
      "learning_rate": 2e-05,
      "loss": 0.1805,
      "step": 5883
    },
    {
      "epoch": 3.976005407232173,
      "grad_norm": 8.22624683380127,
      "learning_rate": 2e-05,
      "loss": 0.2121,
      "step": 5884
    },
    {
      "epoch": 3.976681311253802,
      "grad_norm": 2.302248001098633,
      "learning_rate": 2e-05,
      "loss": 0.1806,
      "step": 5885
    },
    {
      "epoch": 3.977357215275431,
      "grad_norm": 5.116126537322998,
      "learning_rate": 2e-05,
      "loss": 0.163,
      "step": 5886
    },
    {
      "epoch": 3.9780331192970597,
      "grad_norm": 2.579554796218872,
      "learning_rate": 2e-05,
      "loss": 0.1951,
      "step": 5887
    },
    {
      "epoch": 3.9787090233186886,
      "grad_norm": 5.44020414352417,
      "learning_rate": 2e-05,
      "loss": 0.1519,
      "step": 5888
    },
    {
      "epoch": 3.979384927340318,
      "grad_norm": 11.134465217590332,
      "learning_rate": 2e-05,
      "loss": 0.2404,
      "step": 5889
    },
    {
      "epoch": 3.9800608313619463,
      "grad_norm": 5.071147918701172,
      "learning_rate": 2e-05,
      "loss": 0.1925,
      "step": 5890
    },
    {
      "epoch": 3.9807367353835756,
      "grad_norm": 6.1875104904174805,
      "learning_rate": 2e-05,
      "loss": 0.1566,
      "step": 5891
    },
    {
      "epoch": 3.9814126394052045,
      "grad_norm": 4.4117255210876465,
      "learning_rate": 2e-05,
      "loss": 0.2219,
      "step": 5892
    },
    {
      "epoch": 3.9820885434268334,
      "grad_norm": 2.7367682456970215,
      "learning_rate": 2e-05,
      "loss": 0.1415,
      "step": 5893
    },
    {
      "epoch": 3.9827644474484623,
      "grad_norm": 1.905557632446289,
      "learning_rate": 2e-05,
      "loss": 0.1676,
      "step": 5894
    },
    {
      "epoch": 3.983440351470091,
      "grad_norm": 7.18945837020874,
      "learning_rate": 2e-05,
      "loss": 0.1534,
      "step": 5895
    },
    {
      "epoch": 3.9841162554917204,
      "grad_norm": 4.219326019287109,
      "learning_rate": 2e-05,
      "loss": 0.1619,
      "step": 5896
    },
    {
      "epoch": 3.984792159513349,
      "grad_norm": 1.697547197341919,
      "learning_rate": 2e-05,
      "loss": 0.1632,
      "step": 5897
    },
    {
      "epoch": 3.985468063534978,
      "grad_norm": 3.0403831005096436,
      "learning_rate": 2e-05,
      "loss": 0.1299,
      "step": 5898
    },
    {
      "epoch": 3.986143967556607,
      "grad_norm": 2.1039135456085205,
      "learning_rate": 2e-05,
      "loss": 0.2107,
      "step": 5899
    },
    {
      "epoch": 3.986819871578236,
      "grad_norm": 4.23249626159668,
      "learning_rate": 2e-05,
      "loss": 0.1899,
      "step": 5900
    },
    {
      "epoch": 3.986819871578236,
      "eval_accuracy": 0.7434543918918919,
      "eval_loss": 0.2372904121875763,
      "eval_runtime": 77.3598,
      "eval_samples_per_second": 306.102,
      "eval_steps_per_second": 9.566,
      "step": 5900
    },
    {
      "epoch": 3.987495775599865,
      "grad_norm": 2.2446413040161133,
      "learning_rate": 2e-05,
      "loss": 0.1257,
      "step": 5901
    },
    {
      "epoch": 3.9881716796214937,
      "grad_norm": 3.926029682159424,
      "learning_rate": 2e-05,
      "loss": 0.1114,
      "step": 5902
    },
    {
      "epoch": 3.9888475836431225,
      "grad_norm": 2.5736892223358154,
      "learning_rate": 2e-05,
      "loss": 0.1746,
      "step": 5903
    },
    {
      "epoch": 3.9895234876647514,
      "grad_norm": 5.370956897735596,
      "learning_rate": 2e-05,
      "loss": 0.1596,
      "step": 5904
    },
    {
      "epoch": 3.9901993916863807,
      "grad_norm": 5.794821262359619,
      "learning_rate": 2e-05,
      "loss": 0.1691,
      "step": 5905
    },
    {
      "epoch": 3.9908752957080096,
      "grad_norm": 3.9191694259643555,
      "learning_rate": 2e-05,
      "loss": 0.1229,
      "step": 5906
    },
    {
      "epoch": 3.9915511997296385,
      "grad_norm": 2.868987798690796,
      "learning_rate": 2e-05,
      "loss": 0.1655,
      "step": 5907
    },
    {
      "epoch": 3.9922271037512673,
      "grad_norm": 4.992239475250244,
      "learning_rate": 2e-05,
      "loss": 0.1885,
      "step": 5908
    },
    {
      "epoch": 3.992903007772896,
      "grad_norm": 3.0221474170684814,
      "learning_rate": 2e-05,
      "loss": 0.1434,
      "step": 5909
    },
    {
      "epoch": 3.993578911794525,
      "grad_norm": 7.593080997467041,
      "learning_rate": 2e-05,
      "loss": 0.1823,
      "step": 5910
    },
    {
      "epoch": 3.994254815816154,
      "grad_norm": 4.044813632965088,
      "learning_rate": 2e-05,
      "loss": 0.1883,
      "step": 5911
    },
    {
      "epoch": 3.9949307198377833,
      "grad_norm": 2.971304416656494,
      "learning_rate": 2e-05,
      "loss": 0.174,
      "step": 5912
    },
    {
      "epoch": 3.995606623859412,
      "grad_norm": 5.808455944061279,
      "learning_rate": 2e-05,
      "loss": 0.2117,
      "step": 5913
    },
    {
      "epoch": 3.996282527881041,
      "grad_norm": 4.283097743988037,
      "learning_rate": 2e-05,
      "loss": 0.1091,
      "step": 5914
    },
    {
      "epoch": 3.99695843190267,
      "grad_norm": 3.461533784866333,
      "learning_rate": 2e-05,
      "loss": 0.214,
      "step": 5915
    },
    {
      "epoch": 3.9976343359242987,
      "grad_norm": 5.802422523498535,
      "learning_rate": 2e-05,
      "loss": 0.1855,
      "step": 5916
    },
    {
      "epoch": 3.9983102399459276,
      "grad_norm": 3.5770504474639893,
      "learning_rate": 2e-05,
      "loss": 0.1875,
      "step": 5917
    },
    {
      "epoch": 3.9989861439675565,
      "grad_norm": 3.4503793716430664,
      "learning_rate": 2e-05,
      "loss": 0.1538,
      "step": 5918
    },
    {
      "epoch": 3.999662047989186,
      "grad_norm": 2.848845958709717,
      "learning_rate": 2e-05,
      "loss": 0.1386,
      "step": 5919
    },
    {
      "epoch": 4.0,
      "grad_norm": 4.336761951446533,
      "learning_rate": 2e-05,
      "loss": 0.0988,
      "step": 5920
    },
    {
      "epoch": 4.000675904021629,
      "grad_norm": 3.877772808074951,
      "learning_rate": 2e-05,
      "loss": 0.1545,
      "step": 5921
    },
    {
      "epoch": 4.001351808043258,
      "grad_norm": 4.792400360107422,
      "learning_rate": 2e-05,
      "loss": 0.1576,
      "step": 5922
    },
    {
      "epoch": 4.002027712064887,
      "grad_norm": 3.562535285949707,
      "learning_rate": 2e-05,
      "loss": 0.1982,
      "step": 5923
    },
    {
      "epoch": 4.0027036160865155,
      "grad_norm": 6.37260103225708,
      "learning_rate": 2e-05,
      "loss": 0.1416,
      "step": 5924
    },
    {
      "epoch": 4.003379520108145,
      "grad_norm": 6.273654937744141,
      "learning_rate": 2e-05,
      "loss": 0.1455,
      "step": 5925
    },
    {
      "epoch": 4.004055424129773,
      "grad_norm": 1.9614921808242798,
      "learning_rate": 2e-05,
      "loss": 0.1575,
      "step": 5926
    },
    {
      "epoch": 4.0047313281514025,
      "grad_norm": 2.297823429107666,
      "learning_rate": 2e-05,
      "loss": 0.1774,
      "step": 5927
    },
    {
      "epoch": 4.005407232173032,
      "grad_norm": 1.3403681516647339,
      "learning_rate": 2e-05,
      "loss": 0.1278,
      "step": 5928
    },
    {
      "epoch": 4.00608313619466,
      "grad_norm": 3.544326066970825,
      "learning_rate": 2e-05,
      "loss": 0.1696,
      "step": 5929
    },
    {
      "epoch": 4.00675904021629,
      "grad_norm": 2.932507038116455,
      "learning_rate": 2e-05,
      "loss": 0.1596,
      "step": 5930
    },
    {
      "epoch": 4.007434944237918,
      "grad_norm": 2.760758399963379,
      "learning_rate": 2e-05,
      "loss": 0.1415,
      "step": 5931
    },
    {
      "epoch": 4.008110848259547,
      "grad_norm": 1.9079585075378418,
      "learning_rate": 2e-05,
      "loss": 0.1156,
      "step": 5932
    },
    {
      "epoch": 4.008786752281176,
      "grad_norm": 1.9709793329238892,
      "learning_rate": 2e-05,
      "loss": 0.106,
      "step": 5933
    },
    {
      "epoch": 4.009462656302805,
      "grad_norm": 3.079359531402588,
      "learning_rate": 2e-05,
      "loss": 0.1105,
      "step": 5934
    },
    {
      "epoch": 4.0101385603244335,
      "grad_norm": 6.332680702209473,
      "learning_rate": 2e-05,
      "loss": 0.2107,
      "step": 5935
    },
    {
      "epoch": 4.010814464346063,
      "grad_norm": 2.427941083908081,
      "learning_rate": 2e-05,
      "loss": 0.1185,
      "step": 5936
    },
    {
      "epoch": 4.011490368367692,
      "grad_norm": 2.5436620712280273,
      "learning_rate": 2e-05,
      "loss": 0.1503,
      "step": 5937
    },
    {
      "epoch": 4.0121662723893206,
      "grad_norm": 2.3454554080963135,
      "learning_rate": 2e-05,
      "loss": 0.1633,
      "step": 5938
    },
    {
      "epoch": 4.01284217641095,
      "grad_norm": 2.514228105545044,
      "learning_rate": 2e-05,
      "loss": 0.1895,
      "step": 5939
    },
    {
      "epoch": 4.013518080432578,
      "grad_norm": 4.016822338104248,
      "learning_rate": 2e-05,
      "loss": 0.1866,
      "step": 5940
    },
    {
      "epoch": 4.014193984454208,
      "grad_norm": 3.7148780822753906,
      "learning_rate": 2e-05,
      "loss": 0.1827,
      "step": 5941
    },
    {
      "epoch": 4.014869888475836,
      "grad_norm": 2.3364968299865723,
      "learning_rate": 2e-05,
      "loss": 0.0804,
      "step": 5942
    },
    {
      "epoch": 4.015545792497465,
      "grad_norm": 4.564971923828125,
      "learning_rate": 2e-05,
      "loss": 0.1488,
      "step": 5943
    },
    {
      "epoch": 4.016221696519095,
      "grad_norm": 1.9026312828063965,
      "learning_rate": 2e-05,
      "loss": 0.105,
      "step": 5944
    },
    {
      "epoch": 4.016897600540723,
      "grad_norm": 3.5459089279174805,
      "learning_rate": 2e-05,
      "loss": 0.1719,
      "step": 5945
    },
    {
      "epoch": 4.017573504562352,
      "grad_norm": 4.348086357116699,
      "learning_rate": 2e-05,
      "loss": 0.2135,
      "step": 5946
    },
    {
      "epoch": 4.018249408583981,
      "grad_norm": 2.3813319206237793,
      "learning_rate": 2e-05,
      "loss": 0.1655,
      "step": 5947
    },
    {
      "epoch": 4.01892531260561,
      "grad_norm": 1.9477592706680298,
      "learning_rate": 2e-05,
      "loss": 0.099,
      "step": 5948
    },
    {
      "epoch": 4.019601216627239,
      "grad_norm": 2.757025718688965,
      "learning_rate": 2e-05,
      "loss": 0.18,
      "step": 5949
    },
    {
      "epoch": 4.020277120648868,
      "grad_norm": 5.596302032470703,
      "learning_rate": 2e-05,
      "loss": 0.1646,
      "step": 5950
    },
    {
      "epoch": 4.020277120648868,
      "eval_accuracy": 0.7182010135135135,
      "eval_loss": 0.23785313963890076,
      "eval_runtime": 78.7541,
      "eval_samples_per_second": 300.683,
      "eval_steps_per_second": 9.396,
      "step": 5950
    },
    {
      "epoch": 4.020953024670497,
      "grad_norm": 4.479350566864014,
      "learning_rate": 2e-05,
      "loss": 0.1083,
      "step": 5951
    },
    {
      "epoch": 4.021628928692126,
      "grad_norm": 3.2534523010253906,
      "learning_rate": 2e-05,
      "loss": 0.1463,
      "step": 5952
    },
    {
      "epoch": 4.022304832713755,
      "grad_norm": 4.296679973602295,
      "learning_rate": 2e-05,
      "loss": 0.1637,
      "step": 5953
    },
    {
      "epoch": 4.022980736735383,
      "grad_norm": 9.372695922851562,
      "learning_rate": 2e-05,
      "loss": 0.2234,
      "step": 5954
    },
    {
      "epoch": 4.023656640757013,
      "grad_norm": 2.6631433963775635,
      "learning_rate": 2e-05,
      "loss": 0.1795,
      "step": 5955
    },
    {
      "epoch": 4.024332544778641,
      "grad_norm": 8.92272663116455,
      "learning_rate": 2e-05,
      "loss": 0.2015,
      "step": 5956
    },
    {
      "epoch": 4.02500844880027,
      "grad_norm": 3.430333137512207,
      "learning_rate": 2e-05,
      "loss": 0.1293,
      "step": 5957
    },
    {
      "epoch": 4.025684352821899,
      "grad_norm": 10.561985969543457,
      "learning_rate": 2e-05,
      "loss": 0.2165,
      "step": 5958
    },
    {
      "epoch": 4.026360256843528,
      "grad_norm": 4.735415935516357,
      "learning_rate": 2e-05,
      "loss": 0.155,
      "step": 5959
    },
    {
      "epoch": 4.0270361608651575,
      "grad_norm": 4.962104797363281,
      "learning_rate": 2e-05,
      "loss": 0.1403,
      "step": 5960
    },
    {
      "epoch": 4.027712064886786,
      "grad_norm": 2.8353793621063232,
      "learning_rate": 2e-05,
      "loss": 0.1159,
      "step": 5961
    },
    {
      "epoch": 4.028387968908415,
      "grad_norm": 2.9620392322540283,
      "learning_rate": 2e-05,
      "loss": 0.1287,
      "step": 5962
    },
    {
      "epoch": 4.029063872930044,
      "grad_norm": 3.3764026165008545,
      "learning_rate": 2e-05,
      "loss": 0.1481,
      "step": 5963
    },
    {
      "epoch": 4.029739776951673,
      "grad_norm": 5.8362226486206055,
      "learning_rate": 2e-05,
      "loss": 0.1377,
      "step": 5964
    },
    {
      "epoch": 4.030415680973301,
      "grad_norm": 4.514745712280273,
      "learning_rate": 2e-05,
      "loss": 0.1669,
      "step": 5965
    },
    {
      "epoch": 4.031091584994931,
      "grad_norm": 4.699632167816162,
      "learning_rate": 2e-05,
      "loss": 0.2107,
      "step": 5966
    },
    {
      "epoch": 4.03176748901656,
      "grad_norm": 2.7389097213745117,
      "learning_rate": 2e-05,
      "loss": 0.1316,
      "step": 5967
    },
    {
      "epoch": 4.032443393038188,
      "grad_norm": 2.1578452587127686,
      "learning_rate": 2e-05,
      "loss": 0.137,
      "step": 5968
    },
    {
      "epoch": 4.033119297059818,
      "grad_norm": 5.181255340576172,
      "learning_rate": 2e-05,
      "loss": 0.1648,
      "step": 5969
    },
    {
      "epoch": 4.033795201081446,
      "grad_norm": 7.628348350524902,
      "learning_rate": 2e-05,
      "loss": 0.1857,
      "step": 5970
    },
    {
      "epoch": 4.0344711051030755,
      "grad_norm": 3.057807683944702,
      "learning_rate": 2e-05,
      "loss": 0.2384,
      "step": 5971
    },
    {
      "epoch": 4.035147009124704,
      "grad_norm": 2.8357298374176025,
      "learning_rate": 2e-05,
      "loss": 0.132,
      "step": 5972
    },
    {
      "epoch": 4.035822913146333,
      "grad_norm": 3.0287957191467285,
      "learning_rate": 2e-05,
      "loss": 0.1935,
      "step": 5973
    },
    {
      "epoch": 4.0364988171679625,
      "grad_norm": 1.9146722555160522,
      "learning_rate": 2e-05,
      "loss": 0.1187,
      "step": 5974
    },
    {
      "epoch": 4.037174721189591,
      "grad_norm": 6.217987060546875,
      "learning_rate": 2e-05,
      "loss": 0.1491,
      "step": 5975
    },
    {
      "epoch": 4.03785062521122,
      "grad_norm": 3.25813889503479,
      "learning_rate": 2e-05,
      "loss": 0.1213,
      "step": 5976
    },
    {
      "epoch": 4.038526529232849,
      "grad_norm": 3.542292594909668,
      "learning_rate": 2e-05,
      "loss": 0.1276,
      "step": 5977
    },
    {
      "epoch": 4.039202433254478,
      "grad_norm": 2.3480794429779053,
      "learning_rate": 2e-05,
      "loss": 0.1535,
      "step": 5978
    },
    {
      "epoch": 4.0398783372761065,
      "grad_norm": 2.6554572582244873,
      "learning_rate": 2e-05,
      "loss": 0.1439,
      "step": 5979
    },
    {
      "epoch": 4.040554241297736,
      "grad_norm": 4.29343318939209,
      "learning_rate": 2e-05,
      "loss": 0.1322,
      "step": 5980
    },
    {
      "epoch": 4.041230145319365,
      "grad_norm": 6.0417609214782715,
      "learning_rate": 2e-05,
      "loss": 0.1606,
      "step": 5981
    },
    {
      "epoch": 4.0419060493409935,
      "grad_norm": 4.625161170959473,
      "learning_rate": 2e-05,
      "loss": 0.1398,
      "step": 5982
    },
    {
      "epoch": 4.042581953362623,
      "grad_norm": 7.543738842010498,
      "learning_rate": 2e-05,
      "loss": 0.165,
      "step": 5983
    },
    {
      "epoch": 4.043257857384251,
      "grad_norm": 2.317776918411255,
      "learning_rate": 2e-05,
      "loss": 0.1027,
      "step": 5984
    },
    {
      "epoch": 4.043933761405881,
      "grad_norm": 4.8585615158081055,
      "learning_rate": 2e-05,
      "loss": 0.1759,
      "step": 5985
    },
    {
      "epoch": 4.044609665427509,
      "grad_norm": 3.072545051574707,
      "learning_rate": 2e-05,
      "loss": 0.1677,
      "step": 5986
    },
    {
      "epoch": 4.045285569449138,
      "grad_norm": 3.950263261795044,
      "learning_rate": 2e-05,
      "loss": 0.1862,
      "step": 5987
    },
    {
      "epoch": 4.045961473470767,
      "grad_norm": 4.257156848907471,
      "learning_rate": 2e-05,
      "loss": 0.152,
      "step": 5988
    },
    {
      "epoch": 4.046637377492396,
      "grad_norm": 3.766530752182007,
      "learning_rate": 2e-05,
      "loss": 0.1377,
      "step": 5989
    },
    {
      "epoch": 4.047313281514025,
      "grad_norm": 2.521242618560791,
      "learning_rate": 2e-05,
      "loss": 0.124,
      "step": 5990
    },
    {
      "epoch": 4.047989185535654,
      "grad_norm": 2.929335594177246,
      "learning_rate": 2e-05,
      "loss": 0.1712,
      "step": 5991
    },
    {
      "epoch": 4.048665089557283,
      "grad_norm": 4.414095401763916,
      "learning_rate": 2e-05,
      "loss": 0.2189,
      "step": 5992
    },
    {
      "epoch": 4.0493409935789115,
      "grad_norm": 2.1936094760894775,
      "learning_rate": 2e-05,
      "loss": 0.1277,
      "step": 5993
    },
    {
      "epoch": 4.050016897600541,
      "grad_norm": 5.972534656524658,
      "learning_rate": 2e-05,
      "loss": 0.1123,
      "step": 5994
    },
    {
      "epoch": 4.050692801622169,
      "grad_norm": 5.282381534576416,
      "learning_rate": 2e-05,
      "loss": 0.1313,
      "step": 5995
    },
    {
      "epoch": 4.051368705643799,
      "grad_norm": 2.339155435562134,
      "learning_rate": 2e-05,
      "loss": 0.1635,
      "step": 5996
    },
    {
      "epoch": 4.052044609665428,
      "grad_norm": 5.121767997741699,
      "learning_rate": 2e-05,
      "loss": 0.2121,
      "step": 5997
    },
    {
      "epoch": 4.052720513687056,
      "grad_norm": 4.315727233886719,
      "learning_rate": 2e-05,
      "loss": 0.1364,
      "step": 5998
    },
    {
      "epoch": 4.053396417708686,
      "grad_norm": 2.73087739944458,
      "learning_rate": 2e-05,
      "loss": 0.1188,
      "step": 5999
    },
    {
      "epoch": 4.054072321730314,
      "grad_norm": 6.181706428527832,
      "learning_rate": 2e-05,
      "loss": 0.1459,
      "step": 6000
    },
    {
      "epoch": 4.054072321730314,
      "eval_accuracy": 0.7385135135135135,
      "eval_loss": 0.2321215271949768,
      "eval_runtime": 67.3209,
      "eval_samples_per_second": 351.748,
      "eval_steps_per_second": 10.992,
      "step": 6000
    },
    {
      "epoch": 4.054748225751943,
      "grad_norm": 2.757004499435425,
      "learning_rate": 2e-05,
      "loss": 0.0903,
      "step": 6001
    },
    {
      "epoch": 4.055424129773572,
      "grad_norm": 2.2059521675109863,
      "learning_rate": 2e-05,
      "loss": 0.0991,
      "step": 6002
    },
    {
      "epoch": 4.056100033795201,
      "grad_norm": 4.6335296630859375,
      "learning_rate": 2e-05,
      "loss": 0.1188,
      "step": 6003
    },
    {
      "epoch": 4.05677593781683,
      "grad_norm": 3.9778025150299072,
      "learning_rate": 2e-05,
      "loss": 0.1597,
      "step": 6004
    },
    {
      "epoch": 4.057451841838459,
      "grad_norm": 4.150561809539795,
      "learning_rate": 2e-05,
      "loss": 0.1514,
      "step": 6005
    },
    {
      "epoch": 4.058127745860088,
      "grad_norm": 3.1944684982299805,
      "learning_rate": 2e-05,
      "loss": 0.2088,
      "step": 6006
    },
    {
      "epoch": 4.058803649881717,
      "grad_norm": 1.4415786266326904,
      "learning_rate": 2e-05,
      "loss": 0.0854,
      "step": 6007
    },
    {
      "epoch": 4.059479553903346,
      "grad_norm": 2.0330278873443604,
      "learning_rate": 2e-05,
      "loss": 0.1091,
      "step": 6008
    },
    {
      "epoch": 4.060155457924974,
      "grad_norm": 6.337888717651367,
      "learning_rate": 2e-05,
      "loss": 0.1583,
      "step": 6009
    },
    {
      "epoch": 4.060831361946604,
      "grad_norm": 3.038358211517334,
      "learning_rate": 2e-05,
      "loss": 0.1725,
      "step": 6010
    },
    {
      "epoch": 4.061507265968232,
      "grad_norm": 3.440408229827881,
      "learning_rate": 2e-05,
      "loss": 0.1732,
      "step": 6011
    },
    {
      "epoch": 4.062183169989861,
      "grad_norm": 3.4864349365234375,
      "learning_rate": 2e-05,
      "loss": 0.1653,
      "step": 6012
    },
    {
      "epoch": 4.062859074011491,
      "grad_norm": 7.566119194030762,
      "learning_rate": 2e-05,
      "loss": 0.1811,
      "step": 6013
    },
    {
      "epoch": 4.063534978033119,
      "grad_norm": 3.540668487548828,
      "learning_rate": 2e-05,
      "loss": 0.1409,
      "step": 6014
    },
    {
      "epoch": 4.0642108820547485,
      "grad_norm": 2.9883999824523926,
      "learning_rate": 2e-05,
      "loss": 0.1502,
      "step": 6015
    },
    {
      "epoch": 4.064886786076377,
      "grad_norm": 2.1131598949432373,
      "learning_rate": 2e-05,
      "loss": 0.1124,
      "step": 6016
    },
    {
      "epoch": 4.065562690098006,
      "grad_norm": 3.30395770072937,
      "learning_rate": 2e-05,
      "loss": 0.1306,
      "step": 6017
    },
    {
      "epoch": 4.066238594119635,
      "grad_norm": 4.42984676361084,
      "learning_rate": 2e-05,
      "loss": 0.1588,
      "step": 6018
    },
    {
      "epoch": 4.066914498141264,
      "grad_norm": 3.4905481338500977,
      "learning_rate": 2e-05,
      "loss": 0.1482,
      "step": 6019
    },
    {
      "epoch": 4.067590402162893,
      "grad_norm": 5.592382907867432,
      "learning_rate": 2e-05,
      "loss": 0.1828,
      "step": 6020
    },
    {
      "epoch": 4.068266306184522,
      "grad_norm": 2.214325189590454,
      "learning_rate": 2e-05,
      "loss": 0.1768,
      "step": 6021
    },
    {
      "epoch": 4.068942210206151,
      "grad_norm": 2.451348304748535,
      "learning_rate": 2e-05,
      "loss": 0.1159,
      "step": 6022
    },
    {
      "epoch": 4.069618114227779,
      "grad_norm": 3.090230703353882,
      "learning_rate": 2e-05,
      "loss": 0.1538,
      "step": 6023
    },
    {
      "epoch": 4.070294018249409,
      "grad_norm": 3.860980272293091,
      "learning_rate": 2e-05,
      "loss": 0.1197,
      "step": 6024
    },
    {
      "epoch": 4.070969922271037,
      "grad_norm": 1.8067511320114136,
      "learning_rate": 2e-05,
      "loss": 0.1243,
      "step": 6025
    },
    {
      "epoch": 4.0716458262926665,
      "grad_norm": 5.026554584503174,
      "learning_rate": 2e-05,
      "loss": 0.1401,
      "step": 6026
    },
    {
      "epoch": 4.072321730314296,
      "grad_norm": 1.9641401767730713,
      "learning_rate": 2e-05,
      "loss": 0.1174,
      "step": 6027
    },
    {
      "epoch": 4.072997634335924,
      "grad_norm": 2.4053781032562256,
      "learning_rate": 2e-05,
      "loss": 0.1453,
      "step": 6028
    },
    {
      "epoch": 4.0736735383575535,
      "grad_norm": 2.314420223236084,
      "learning_rate": 2e-05,
      "loss": 0.1327,
      "step": 6029
    },
    {
      "epoch": 4.074349442379182,
      "grad_norm": 4.51899528503418,
      "learning_rate": 2e-05,
      "loss": 0.1583,
      "step": 6030
    },
    {
      "epoch": 4.075025346400811,
      "grad_norm": 1.8079265356063843,
      "learning_rate": 2e-05,
      "loss": 0.09,
      "step": 6031
    },
    {
      "epoch": 4.07570125042244,
      "grad_norm": 1.9722028970718384,
      "learning_rate": 2e-05,
      "loss": 0.0986,
      "step": 6032
    },
    {
      "epoch": 4.076377154444069,
      "grad_norm": 3.478182792663574,
      "learning_rate": 2e-05,
      "loss": 0.1662,
      "step": 6033
    },
    {
      "epoch": 4.077053058465697,
      "grad_norm": 3.0452284812927246,
      "learning_rate": 2e-05,
      "loss": 0.1398,
      "step": 6034
    },
    {
      "epoch": 4.077728962487327,
      "grad_norm": 2.5909805297851562,
      "learning_rate": 2e-05,
      "loss": 0.1013,
      "step": 6035
    },
    {
      "epoch": 4.078404866508956,
      "grad_norm": 5.149347305297852,
      "learning_rate": 2e-05,
      "loss": 0.1554,
      "step": 6036
    },
    {
      "epoch": 4.0790807705305845,
      "grad_norm": 2.8887641429901123,
      "learning_rate": 2e-05,
      "loss": 0.1333,
      "step": 6037
    },
    {
      "epoch": 4.079756674552214,
      "grad_norm": 6.897005081176758,
      "learning_rate": 2e-05,
      "loss": 0.1834,
      "step": 6038
    },
    {
      "epoch": 4.080432578573842,
      "grad_norm": 8.193465232849121,
      "learning_rate": 2e-05,
      "loss": 0.2169,
      "step": 6039
    },
    {
      "epoch": 4.0811084825954715,
      "grad_norm": 5.322359561920166,
      "learning_rate": 2e-05,
      "loss": 0.2088,
      "step": 6040
    },
    {
      "epoch": 4.0817843866171,
      "grad_norm": 2.739659547805786,
      "learning_rate": 2e-05,
      "loss": 0.1855,
      "step": 6041
    },
    {
      "epoch": 4.082460290638729,
      "grad_norm": 7.457422256469727,
      "learning_rate": 2e-05,
      "loss": 0.2072,
      "step": 6042
    },
    {
      "epoch": 4.083136194660359,
      "grad_norm": 8.422574996948242,
      "learning_rate": 2e-05,
      "loss": 0.1848,
      "step": 6043
    },
    {
      "epoch": 4.083812098681987,
      "grad_norm": 6.9970245361328125,
      "learning_rate": 2e-05,
      "loss": 0.1167,
      "step": 6044
    },
    {
      "epoch": 4.084488002703616,
      "grad_norm": 6.996243000030518,
      "learning_rate": 2e-05,
      "loss": 0.1654,
      "step": 6045
    },
    {
      "epoch": 4.085163906725245,
      "grad_norm": 2.9135618209838867,
      "learning_rate": 2e-05,
      "loss": 0.1672,
      "step": 6046
    },
    {
      "epoch": 4.085839810746874,
      "grad_norm": 3.231106996536255,
      "learning_rate": 2e-05,
      "loss": 0.196,
      "step": 6047
    },
    {
      "epoch": 4.0865157147685025,
      "grad_norm": 1.969253659248352,
      "learning_rate": 2e-05,
      "loss": 0.1568,
      "step": 6048
    },
    {
      "epoch": 4.087191618790132,
      "grad_norm": 2.228097438812256,
      "learning_rate": 2e-05,
      "loss": 0.1346,
      "step": 6049
    },
    {
      "epoch": 4.087867522811761,
      "grad_norm": 3.059337854385376,
      "learning_rate": 2e-05,
      "loss": 0.1701,
      "step": 6050
    },
    {
      "epoch": 4.087867522811761,
      "eval_accuracy": 0.7353462837837837,
      "eval_loss": 0.2262779176235199,
      "eval_runtime": 56.8289,
      "eval_samples_per_second": 416.689,
      "eval_steps_per_second": 13.022,
      "step": 6050
    },
    {
      "epoch": 4.08854342683339,
      "grad_norm": 3.097503900527954,
      "learning_rate": 2e-05,
      "loss": 0.1162,
      "step": 6051
    },
    {
      "epoch": 4.089219330855019,
      "grad_norm": 8.0562744140625,
      "learning_rate": 2e-05,
      "loss": 0.1546,
      "step": 6052
    },
    {
      "epoch": 4.089895234876647,
      "grad_norm": 5.991288661956787,
      "learning_rate": 2e-05,
      "loss": 0.1728,
      "step": 6053
    },
    {
      "epoch": 4.090571138898277,
      "grad_norm": 6.34078311920166,
      "learning_rate": 2e-05,
      "loss": 0.1701,
      "step": 6054
    },
    {
      "epoch": 4.091247042919905,
      "grad_norm": 6.219736099243164,
      "learning_rate": 2e-05,
      "loss": 0.1636,
      "step": 6055
    },
    {
      "epoch": 4.091922946941534,
      "grad_norm": 3.811692714691162,
      "learning_rate": 2e-05,
      "loss": 0.1583,
      "step": 6056
    },
    {
      "epoch": 4.092598850963164,
      "grad_norm": 2.815898895263672,
      "learning_rate": 2e-05,
      "loss": 0.1815,
      "step": 6057
    },
    {
      "epoch": 4.093274754984792,
      "grad_norm": 6.943788051605225,
      "learning_rate": 2e-05,
      "loss": 0.1687,
      "step": 6058
    },
    {
      "epoch": 4.093950659006421,
      "grad_norm": 3.2690422534942627,
      "learning_rate": 2e-05,
      "loss": 0.1712,
      "step": 6059
    },
    {
      "epoch": 4.09462656302805,
      "grad_norm": 2.696180582046509,
      "learning_rate": 2e-05,
      "loss": 0.1297,
      "step": 6060
    },
    {
      "epoch": 4.095302467049679,
      "grad_norm": 3.6414480209350586,
      "learning_rate": 2e-05,
      "loss": 0.1301,
      "step": 6061
    },
    {
      "epoch": 4.095978371071308,
      "grad_norm": 8.419443130493164,
      "learning_rate": 2e-05,
      "loss": 0.1834,
      "step": 6062
    },
    {
      "epoch": 4.096654275092937,
      "grad_norm": 4.403066635131836,
      "learning_rate": 2e-05,
      "loss": 0.1324,
      "step": 6063
    },
    {
      "epoch": 4.097330179114565,
      "grad_norm": 4.7545905113220215,
      "learning_rate": 2e-05,
      "loss": 0.222,
      "step": 6064
    },
    {
      "epoch": 4.098006083136195,
      "grad_norm": 5.7205705642700195,
      "learning_rate": 2e-05,
      "loss": 0.2201,
      "step": 6065
    },
    {
      "epoch": 4.098681987157824,
      "grad_norm": 7.724461555480957,
      "learning_rate": 2e-05,
      "loss": 0.217,
      "step": 6066
    },
    {
      "epoch": 4.099357891179452,
      "grad_norm": 7.724101543426514,
      "learning_rate": 2e-05,
      "loss": 0.2101,
      "step": 6067
    },
    {
      "epoch": 4.100033795201082,
      "grad_norm": 3.0401391983032227,
      "learning_rate": 2e-05,
      "loss": 0.1583,
      "step": 6068
    },
    {
      "epoch": 4.10070969922271,
      "grad_norm": 3.115269660949707,
      "learning_rate": 2e-05,
      "loss": 0.187,
      "step": 6069
    },
    {
      "epoch": 4.101385603244339,
      "grad_norm": 1.550874948501587,
      "learning_rate": 2e-05,
      "loss": 0.0675,
      "step": 6070
    },
    {
      "epoch": 4.102061507265968,
      "grad_norm": 3.1656196117401123,
      "learning_rate": 2e-05,
      "loss": 0.2047,
      "step": 6071
    },
    {
      "epoch": 4.102737411287597,
      "grad_norm": 3.730518102645874,
      "learning_rate": 2e-05,
      "loss": 0.2093,
      "step": 6072
    },
    {
      "epoch": 4.1034133153092265,
      "grad_norm": 5.427579879760742,
      "learning_rate": 2e-05,
      "loss": 0.1765,
      "step": 6073
    },
    {
      "epoch": 4.104089219330855,
      "grad_norm": 4.6324687004089355,
      "learning_rate": 2e-05,
      "loss": 0.2099,
      "step": 6074
    },
    {
      "epoch": 4.104765123352484,
      "grad_norm": 3.595536708831787,
      "learning_rate": 2e-05,
      "loss": 0.1626,
      "step": 6075
    },
    {
      "epoch": 4.105441027374113,
      "grad_norm": 2.039138078689575,
      "learning_rate": 2e-05,
      "loss": 0.1213,
      "step": 6076
    },
    {
      "epoch": 4.106116931395742,
      "grad_norm": 3.0348691940307617,
      "learning_rate": 2e-05,
      "loss": 0.1842,
      "step": 6077
    },
    {
      "epoch": 4.10679283541737,
      "grad_norm": 2.760173797607422,
      "learning_rate": 2e-05,
      "loss": 0.1493,
      "step": 6078
    },
    {
      "epoch": 4.107468739439,
      "grad_norm": 2.340419292449951,
      "learning_rate": 2e-05,
      "loss": 0.1525,
      "step": 6079
    },
    {
      "epoch": 4.108144643460629,
      "grad_norm": 4.913007736206055,
      "learning_rate": 2e-05,
      "loss": 0.1283,
      "step": 6080
    },
    {
      "epoch": 4.1088205474822574,
      "grad_norm": 7.013321399688721,
      "learning_rate": 2e-05,
      "loss": 0.1705,
      "step": 6081
    },
    {
      "epoch": 4.109496451503887,
      "grad_norm": 8.176398277282715,
      "learning_rate": 2e-05,
      "loss": 0.1864,
      "step": 6082
    },
    {
      "epoch": 4.110172355525515,
      "grad_norm": 1.9102208614349365,
      "learning_rate": 2e-05,
      "loss": 0.1354,
      "step": 6083
    },
    {
      "epoch": 4.1108482595471445,
      "grad_norm": 2.5853824615478516,
      "learning_rate": 2e-05,
      "loss": 0.1213,
      "step": 6084
    },
    {
      "epoch": 4.111524163568773,
      "grad_norm": 2.18259859085083,
      "learning_rate": 2e-05,
      "loss": 0.1234,
      "step": 6085
    },
    {
      "epoch": 4.112200067590402,
      "grad_norm": 4.01481294631958,
      "learning_rate": 2e-05,
      "loss": 0.1653,
      "step": 6086
    },
    {
      "epoch": 4.112875971612031,
      "grad_norm": 5.346920967102051,
      "learning_rate": 2e-05,
      "loss": 0.1462,
      "step": 6087
    },
    {
      "epoch": 4.11355187563366,
      "grad_norm": 6.183706283569336,
      "learning_rate": 2e-05,
      "loss": 0.1495,
      "step": 6088
    },
    {
      "epoch": 4.114227779655289,
      "grad_norm": 3.9923346042633057,
      "learning_rate": 2e-05,
      "loss": 0.1541,
      "step": 6089
    },
    {
      "epoch": 4.114903683676918,
      "grad_norm": 4.631706237792969,
      "learning_rate": 2e-05,
      "loss": 0.1663,
      "step": 6090
    },
    {
      "epoch": 4.115579587698547,
      "grad_norm": 3.2256417274475098,
      "learning_rate": 2e-05,
      "loss": 0.135,
      "step": 6091
    },
    {
      "epoch": 4.1162554917201755,
      "grad_norm": 2.7804486751556396,
      "learning_rate": 2e-05,
      "loss": 0.1661,
      "step": 6092
    },
    {
      "epoch": 4.116931395741805,
      "grad_norm": 5.014955043792725,
      "learning_rate": 2e-05,
      "loss": 0.2005,
      "step": 6093
    },
    {
      "epoch": 4.117607299763433,
      "grad_norm": 2.250911235809326,
      "learning_rate": 2e-05,
      "loss": 0.0735,
      "step": 6094
    },
    {
      "epoch": 4.1182832037850625,
      "grad_norm": 3.986591100692749,
      "learning_rate": 2e-05,
      "loss": 0.1971,
      "step": 6095
    },
    {
      "epoch": 4.118959107806692,
      "grad_norm": 3.187697649002075,
      "learning_rate": 2e-05,
      "loss": 0.1498,
      "step": 6096
    },
    {
      "epoch": 4.11963501182832,
      "grad_norm": 7.33013391494751,
      "learning_rate": 2e-05,
      "loss": 0.2088,
      "step": 6097
    },
    {
      "epoch": 4.12031091584995,
      "grad_norm": 2.559391736984253,
      "learning_rate": 2e-05,
      "loss": 0.1252,
      "step": 6098
    },
    {
      "epoch": 4.120986819871578,
      "grad_norm": 5.512972831726074,
      "learning_rate": 2e-05,
      "loss": 0.1614,
      "step": 6099
    },
    {
      "epoch": 4.121662723893207,
      "grad_norm": 2.453488349914551,
      "learning_rate": 2e-05,
      "loss": 0.1597,
      "step": 6100
    },
    {
      "epoch": 4.121662723893207,
      "eval_accuracy": 0.7005067567567568,
      "eval_loss": 0.24151690304279327,
      "eval_runtime": 59.5861,
      "eval_samples_per_second": 397.408,
      "eval_steps_per_second": 12.419,
      "step": 6100
    },
    {
      "epoch": 4.122338627914836,
      "grad_norm": 4.599640846252441,
      "learning_rate": 2e-05,
      "loss": 0.1299,
      "step": 6101
    },
    {
      "epoch": 4.123014531936465,
      "grad_norm": 6.1339945793151855,
      "learning_rate": 2e-05,
      "loss": 0.1355,
      "step": 6102
    },
    {
      "epoch": 4.123690435958094,
      "grad_norm": 4.014218807220459,
      "learning_rate": 2e-05,
      "loss": 0.1441,
      "step": 6103
    },
    {
      "epoch": 4.124366339979723,
      "grad_norm": 3.761927366256714,
      "learning_rate": 2e-05,
      "loss": 0.1386,
      "step": 6104
    },
    {
      "epoch": 4.125042244001352,
      "grad_norm": 3.955191135406494,
      "learning_rate": 2e-05,
      "loss": 0.1719,
      "step": 6105
    },
    {
      "epoch": 4.1257181480229805,
      "grad_norm": 6.08704137802124,
      "learning_rate": 2e-05,
      "loss": 0.1286,
      "step": 6106
    },
    {
      "epoch": 4.12639405204461,
      "grad_norm": 7.219403266906738,
      "learning_rate": 2e-05,
      "loss": 0.206,
      "step": 6107
    },
    {
      "epoch": 4.127069956066238,
      "grad_norm": 2.60709285736084,
      "learning_rate": 2e-05,
      "loss": 0.1641,
      "step": 6108
    },
    {
      "epoch": 4.127745860087868,
      "grad_norm": 2.7817187309265137,
      "learning_rate": 2e-05,
      "loss": 0.1361,
      "step": 6109
    },
    {
      "epoch": 4.128421764109497,
      "grad_norm": 1.9536246061325073,
      "learning_rate": 2e-05,
      "loss": 0.1168,
      "step": 6110
    },
    {
      "epoch": 4.129097668131125,
      "grad_norm": 3.0383460521698,
      "learning_rate": 2e-05,
      "loss": 0.1685,
      "step": 6111
    },
    {
      "epoch": 4.129773572152755,
      "grad_norm": 2.0315558910369873,
      "learning_rate": 2e-05,
      "loss": 0.1349,
      "step": 6112
    },
    {
      "epoch": 4.130449476174383,
      "grad_norm": 2.7339491844177246,
      "learning_rate": 2e-05,
      "loss": 0.1698,
      "step": 6113
    },
    {
      "epoch": 4.131125380196012,
      "grad_norm": 4.383904457092285,
      "learning_rate": 2e-05,
      "loss": 0.138,
      "step": 6114
    },
    {
      "epoch": 4.131801284217641,
      "grad_norm": 3.280054807662964,
      "learning_rate": 2e-05,
      "loss": 0.1625,
      "step": 6115
    },
    {
      "epoch": 4.13247718823927,
      "grad_norm": 2.7890336513519287,
      "learning_rate": 2e-05,
      "loss": 0.1351,
      "step": 6116
    },
    {
      "epoch": 4.1331530922608986,
      "grad_norm": 5.347922325134277,
      "learning_rate": 2e-05,
      "loss": 0.1432,
      "step": 6117
    },
    {
      "epoch": 4.133828996282528,
      "grad_norm": 3.6584854125976562,
      "learning_rate": 2e-05,
      "loss": 0.1234,
      "step": 6118
    },
    {
      "epoch": 4.134504900304157,
      "grad_norm": 4.877383708953857,
      "learning_rate": 2e-05,
      "loss": 0.2053,
      "step": 6119
    },
    {
      "epoch": 4.135180804325786,
      "grad_norm": 2.776392698287964,
      "learning_rate": 2e-05,
      "loss": 0.1691,
      "step": 6120
    },
    {
      "epoch": 4.135856708347415,
      "grad_norm": 3.263359308242798,
      "learning_rate": 2e-05,
      "loss": 0.2159,
      "step": 6121
    },
    {
      "epoch": 4.136532612369043,
      "grad_norm": 2.670645236968994,
      "learning_rate": 2e-05,
      "loss": 0.1461,
      "step": 6122
    },
    {
      "epoch": 4.137208516390673,
      "grad_norm": 8.445600509643555,
      "learning_rate": 2e-05,
      "loss": 0.2628,
      "step": 6123
    },
    {
      "epoch": 4.137884420412301,
      "grad_norm": 3.883256435394287,
      "learning_rate": 2e-05,
      "loss": 0.146,
      "step": 6124
    },
    {
      "epoch": 4.13856032443393,
      "grad_norm": 2.7139437198638916,
      "learning_rate": 2e-05,
      "loss": 0.1829,
      "step": 6125
    },
    {
      "epoch": 4.13923622845556,
      "grad_norm": 2.3576250076293945,
      "learning_rate": 2e-05,
      "loss": 0.0929,
      "step": 6126
    },
    {
      "epoch": 4.139912132477188,
      "grad_norm": 2.0886781215667725,
      "learning_rate": 2e-05,
      "loss": 0.1204,
      "step": 6127
    },
    {
      "epoch": 4.1405880364988175,
      "grad_norm": 2.6614091396331787,
      "learning_rate": 2e-05,
      "loss": 0.1306,
      "step": 6128
    },
    {
      "epoch": 4.141263940520446,
      "grad_norm": 3.399962902069092,
      "learning_rate": 2e-05,
      "loss": 0.1191,
      "step": 6129
    },
    {
      "epoch": 4.141939844542075,
      "grad_norm": 2.7641100883483887,
      "learning_rate": 2e-05,
      "loss": 0.1286,
      "step": 6130
    },
    {
      "epoch": 4.142615748563704,
      "grad_norm": 4.239030838012695,
      "learning_rate": 2e-05,
      "loss": 0.1181,
      "step": 6131
    },
    {
      "epoch": 4.143291652585333,
      "grad_norm": 2.9794204235076904,
      "learning_rate": 2e-05,
      "loss": 0.1516,
      "step": 6132
    },
    {
      "epoch": 4.143967556606961,
      "grad_norm": 4.384451389312744,
      "learning_rate": 2e-05,
      "loss": 0.1702,
      "step": 6133
    },
    {
      "epoch": 4.144643460628591,
      "grad_norm": 4.148662567138672,
      "learning_rate": 2e-05,
      "loss": 0.1997,
      "step": 6134
    },
    {
      "epoch": 4.14531936465022,
      "grad_norm": 3.129305601119995,
      "learning_rate": 2e-05,
      "loss": 0.1623,
      "step": 6135
    },
    {
      "epoch": 4.145995268671848,
      "grad_norm": 4.315300464630127,
      "learning_rate": 2e-05,
      "loss": 0.2179,
      "step": 6136
    },
    {
      "epoch": 4.146671172693478,
      "grad_norm": 3.976933717727661,
      "learning_rate": 2e-05,
      "loss": 0.1425,
      "step": 6137
    },
    {
      "epoch": 4.147347076715106,
      "grad_norm": 3.0151114463806152,
      "learning_rate": 2e-05,
      "loss": 0.1535,
      "step": 6138
    },
    {
      "epoch": 4.1480229807367355,
      "grad_norm": 4.709417819976807,
      "learning_rate": 2e-05,
      "loss": 0.1325,
      "step": 6139
    },
    {
      "epoch": 4.148698884758364,
      "grad_norm": 4.973837852478027,
      "learning_rate": 2e-05,
      "loss": 0.14,
      "step": 6140
    },
    {
      "epoch": 4.149374788779993,
      "grad_norm": 2.853055477142334,
      "learning_rate": 2e-05,
      "loss": 0.1875,
      "step": 6141
    },
    {
      "epoch": 4.1500506928016225,
      "grad_norm": 2.8199245929718018,
      "learning_rate": 2e-05,
      "loss": 0.1819,
      "step": 6142
    },
    {
      "epoch": 4.150726596823251,
      "grad_norm": 3.3903114795684814,
      "learning_rate": 2e-05,
      "loss": 0.1506,
      "step": 6143
    },
    {
      "epoch": 4.15140250084488,
      "grad_norm": 4.044972896575928,
      "learning_rate": 2e-05,
      "loss": 0.1189,
      "step": 6144
    },
    {
      "epoch": 4.152078404866509,
      "grad_norm": 1.6796895265579224,
      "learning_rate": 2e-05,
      "loss": 0.1199,
      "step": 6145
    },
    {
      "epoch": 4.152754308888138,
      "grad_norm": 2.9901607036590576,
      "learning_rate": 2e-05,
      "loss": 0.157,
      "step": 6146
    },
    {
      "epoch": 4.153430212909766,
      "grad_norm": 5.980422019958496,
      "learning_rate": 2e-05,
      "loss": 0.1069,
      "step": 6147
    },
    {
      "epoch": 4.154106116931396,
      "grad_norm": 3.1898093223571777,
      "learning_rate": 2e-05,
      "loss": 0.1648,
      "step": 6148
    },
    {
      "epoch": 4.154782020953025,
      "grad_norm": 1.7890381813049316,
      "learning_rate": 2e-05,
      "loss": 0.1157,
      "step": 6149
    },
    {
      "epoch": 4.1554579249746535,
      "grad_norm": 2.317837715148926,
      "learning_rate": 2e-05,
      "loss": 0.1392,
      "step": 6150
    },
    {
      "epoch": 4.1554579249746535,
      "eval_accuracy": 0.7430320945945946,
      "eval_loss": 0.2309252768754959,
      "eval_runtime": 60.4075,
      "eval_samples_per_second": 392.004,
      "eval_steps_per_second": 12.25,
      "step": 6150
    },
    {
      "epoch": 4.156133828996283,
      "grad_norm": 4.233912944793701,
      "learning_rate": 2e-05,
      "loss": 0.145,
      "step": 6151
    },
    {
      "epoch": 4.156809733017911,
      "grad_norm": 2.888629198074341,
      "learning_rate": 2e-05,
      "loss": 0.1858,
      "step": 6152
    },
    {
      "epoch": 4.1574856370395405,
      "grad_norm": 3.7472307682037354,
      "learning_rate": 2e-05,
      "loss": 0.1387,
      "step": 6153
    },
    {
      "epoch": 4.158161541061169,
      "grad_norm": 6.225857734680176,
      "learning_rate": 2e-05,
      "loss": 0.1611,
      "step": 6154
    },
    {
      "epoch": 4.158837445082798,
      "grad_norm": 5.633131504058838,
      "learning_rate": 2e-05,
      "loss": 0.2181,
      "step": 6155
    },
    {
      "epoch": 4.159513349104428,
      "grad_norm": 2.4953794479370117,
      "learning_rate": 2e-05,
      "loss": 0.1376,
      "step": 6156
    },
    {
      "epoch": 4.160189253126056,
      "grad_norm": 3.4902193546295166,
      "learning_rate": 2e-05,
      "loss": 0.2451,
      "step": 6157
    },
    {
      "epoch": 4.160865157147685,
      "grad_norm": 8.868794441223145,
      "learning_rate": 2e-05,
      "loss": 0.1555,
      "step": 6158
    },
    {
      "epoch": 4.161541061169314,
      "grad_norm": 3.2319753170013428,
      "learning_rate": 2e-05,
      "loss": 0.1133,
      "step": 6159
    },
    {
      "epoch": 4.162216965190943,
      "grad_norm": 6.249331474304199,
      "learning_rate": 2e-05,
      "loss": 0.1526,
      "step": 6160
    },
    {
      "epoch": 4.1628928692125715,
      "grad_norm": 5.563086986541748,
      "learning_rate": 2e-05,
      "loss": 0.1673,
      "step": 6161
    },
    {
      "epoch": 4.163568773234201,
      "grad_norm": 3.1149981021881104,
      "learning_rate": 2e-05,
      "loss": 0.1485,
      "step": 6162
    },
    {
      "epoch": 4.164244677255829,
      "grad_norm": 4.8729634284973145,
      "learning_rate": 2e-05,
      "loss": 0.1546,
      "step": 6163
    },
    {
      "epoch": 4.164920581277459,
      "grad_norm": 12.991255760192871,
      "learning_rate": 2e-05,
      "loss": 0.2151,
      "step": 6164
    },
    {
      "epoch": 4.165596485299088,
      "grad_norm": 5.533585548400879,
      "learning_rate": 2e-05,
      "loss": 0.1339,
      "step": 6165
    },
    {
      "epoch": 4.166272389320716,
      "grad_norm": 2.442362070083618,
      "learning_rate": 2e-05,
      "loss": 0.1265,
      "step": 6166
    },
    {
      "epoch": 4.166948293342346,
      "grad_norm": 2.246105432510376,
      "learning_rate": 2e-05,
      "loss": 0.1026,
      "step": 6167
    },
    {
      "epoch": 4.167624197363974,
      "grad_norm": 5.6920857429504395,
      "learning_rate": 2e-05,
      "loss": 0.125,
      "step": 6168
    },
    {
      "epoch": 4.168300101385603,
      "grad_norm": 2.116575241088867,
      "learning_rate": 2e-05,
      "loss": 0.1496,
      "step": 6169
    },
    {
      "epoch": 4.168976005407232,
      "grad_norm": 2.242928981781006,
      "learning_rate": 2e-05,
      "loss": 0.1171,
      "step": 6170
    },
    {
      "epoch": 4.169651909428861,
      "grad_norm": 2.793071985244751,
      "learning_rate": 2e-05,
      "loss": 0.1442,
      "step": 6171
    },
    {
      "epoch": 4.17032781345049,
      "grad_norm": 3.044152021408081,
      "learning_rate": 2e-05,
      "loss": 0.1836,
      "step": 6172
    },
    {
      "epoch": 4.171003717472119,
      "grad_norm": 3.6980652809143066,
      "learning_rate": 2e-05,
      "loss": 0.1562,
      "step": 6173
    },
    {
      "epoch": 4.171679621493748,
      "grad_norm": 2.2021679878234863,
      "learning_rate": 2e-05,
      "loss": 0.1495,
      "step": 6174
    },
    {
      "epoch": 4.172355525515377,
      "grad_norm": 2.5273382663726807,
      "learning_rate": 2e-05,
      "loss": 0.1256,
      "step": 6175
    },
    {
      "epoch": 4.173031429537006,
      "grad_norm": 3.4007160663604736,
      "learning_rate": 2e-05,
      "loss": 0.1686,
      "step": 6176
    },
    {
      "epoch": 4.173707333558634,
      "grad_norm": 3.2590744495391846,
      "learning_rate": 2e-05,
      "loss": 0.1474,
      "step": 6177
    },
    {
      "epoch": 4.174383237580264,
      "grad_norm": 4.91434383392334,
      "learning_rate": 2e-05,
      "loss": 0.1852,
      "step": 6178
    },
    {
      "epoch": 4.175059141601893,
      "grad_norm": 3.7803969383239746,
      "learning_rate": 2e-05,
      "loss": 0.135,
      "step": 6179
    },
    {
      "epoch": 4.175735045623521,
      "grad_norm": 3.2313926219940186,
      "learning_rate": 2e-05,
      "loss": 0.1277,
      "step": 6180
    },
    {
      "epoch": 4.176410949645151,
      "grad_norm": 6.105226516723633,
      "learning_rate": 2e-05,
      "loss": 0.1711,
      "step": 6181
    },
    {
      "epoch": 4.177086853666779,
      "grad_norm": 3.350430727005005,
      "learning_rate": 2e-05,
      "loss": 0.1065,
      "step": 6182
    },
    {
      "epoch": 4.177762757688408,
      "grad_norm": 7.344993591308594,
      "learning_rate": 2e-05,
      "loss": 0.1358,
      "step": 6183
    },
    {
      "epoch": 4.178438661710037,
      "grad_norm": 3.989229440689087,
      "learning_rate": 2e-05,
      "loss": 0.1462,
      "step": 6184
    },
    {
      "epoch": 4.179114565731666,
      "grad_norm": 4.18123722076416,
      "learning_rate": 2e-05,
      "loss": 0.1216,
      "step": 6185
    },
    {
      "epoch": 4.179790469753295,
      "grad_norm": 1.9022855758666992,
      "learning_rate": 2e-05,
      "loss": 0.1282,
      "step": 6186
    },
    {
      "epoch": 4.180466373774924,
      "grad_norm": 2.927896022796631,
      "learning_rate": 2e-05,
      "loss": 0.1336,
      "step": 6187
    },
    {
      "epoch": 4.181142277796553,
      "grad_norm": 3.871670961380005,
      "learning_rate": 2e-05,
      "loss": 0.1187,
      "step": 6188
    },
    {
      "epoch": 4.181818181818182,
      "grad_norm": 5.506327152252197,
      "learning_rate": 2e-05,
      "loss": 0.2551,
      "step": 6189
    },
    {
      "epoch": 4.182494085839811,
      "grad_norm": 4.418001651763916,
      "learning_rate": 2e-05,
      "loss": 0.1574,
      "step": 6190
    },
    {
      "epoch": 4.183169989861439,
      "grad_norm": 2.0426197052001953,
      "learning_rate": 2e-05,
      "loss": 0.1348,
      "step": 6191
    },
    {
      "epoch": 4.183845893883069,
      "grad_norm": 6.373532772064209,
      "learning_rate": 2e-05,
      "loss": 0.2179,
      "step": 6192
    },
    {
      "epoch": 4.184521797904697,
      "grad_norm": 7.969865798950195,
      "learning_rate": 2e-05,
      "loss": 0.1626,
      "step": 6193
    },
    {
      "epoch": 4.1851977019263265,
      "grad_norm": 2.712327003479004,
      "learning_rate": 2e-05,
      "loss": 0.1266,
      "step": 6194
    },
    {
      "epoch": 4.185873605947956,
      "grad_norm": 5.601124286651611,
      "learning_rate": 2e-05,
      "loss": 0.2484,
      "step": 6195
    },
    {
      "epoch": 4.186549509969584,
      "grad_norm": 5.1215596199035645,
      "learning_rate": 2e-05,
      "loss": 0.1687,
      "step": 6196
    },
    {
      "epoch": 4.1872254139912135,
      "grad_norm": 6.839081764221191,
      "learning_rate": 2e-05,
      "loss": 0.2199,
      "step": 6197
    },
    {
      "epoch": 4.187901318012842,
      "grad_norm": 7.855785369873047,
      "learning_rate": 2e-05,
      "loss": 0.1674,
      "step": 6198
    },
    {
      "epoch": 4.188577222034471,
      "grad_norm": 4.710515975952148,
      "learning_rate": 2e-05,
      "loss": 0.1811,
      "step": 6199
    },
    {
      "epoch": 4.1892531260561,
      "grad_norm": 3.6472463607788086,
      "learning_rate": 2e-05,
      "loss": 0.1906,
      "step": 6200
    },
    {
      "epoch": 4.1892531260561,
      "eval_accuracy": 0.7266047297297298,
      "eval_loss": 0.22272951900959015,
      "eval_runtime": 59.8944,
      "eval_samples_per_second": 395.363,
      "eval_steps_per_second": 12.355,
      "step": 6200
    },
    {
      "epoch": 4.189929030077729,
      "grad_norm": 4.151423454284668,
      "learning_rate": 2e-05,
      "loss": 0.1471,
      "step": 6201
    },
    {
      "epoch": 4.190604934099358,
      "grad_norm": 2.7161216735839844,
      "learning_rate": 2e-05,
      "loss": 0.155,
      "step": 6202
    },
    {
      "epoch": 4.191280838120987,
      "grad_norm": 2.0319364070892334,
      "learning_rate": 2e-05,
      "loss": 0.1694,
      "step": 6203
    },
    {
      "epoch": 4.191956742142616,
      "grad_norm": 3.797396659851074,
      "learning_rate": 2e-05,
      "loss": 0.1458,
      "step": 6204
    },
    {
      "epoch": 4.1926326461642445,
      "grad_norm": 3.8116188049316406,
      "learning_rate": 2e-05,
      "loss": 0.1788,
      "step": 6205
    },
    {
      "epoch": 4.193308550185874,
      "grad_norm": 6.784358978271484,
      "learning_rate": 2e-05,
      "loss": 0.1677,
      "step": 6206
    },
    {
      "epoch": 4.193984454207502,
      "grad_norm": 7.31110954284668,
      "learning_rate": 2e-05,
      "loss": 0.1909,
      "step": 6207
    },
    {
      "epoch": 4.1946603582291315,
      "grad_norm": 6.285653114318848,
      "learning_rate": 2e-05,
      "loss": 0.207,
      "step": 6208
    },
    {
      "epoch": 4.195336262250761,
      "grad_norm": 2.2258896827697754,
      "learning_rate": 2e-05,
      "loss": 0.1761,
      "step": 6209
    },
    {
      "epoch": 4.196012166272389,
      "grad_norm": 1.9113695621490479,
      "learning_rate": 2e-05,
      "loss": 0.1408,
      "step": 6210
    },
    {
      "epoch": 4.196688070294019,
      "grad_norm": 4.498875141143799,
      "learning_rate": 2e-05,
      "loss": 0.1788,
      "step": 6211
    },
    {
      "epoch": 4.197363974315647,
      "grad_norm": 7.024548530578613,
      "learning_rate": 2e-05,
      "loss": 0.1509,
      "step": 6212
    },
    {
      "epoch": 4.198039878337276,
      "grad_norm": 3.506108522415161,
      "learning_rate": 2e-05,
      "loss": 0.1817,
      "step": 6213
    },
    {
      "epoch": 4.198715782358905,
      "grad_norm": 2.0298233032226562,
      "learning_rate": 2e-05,
      "loss": 0.1086,
      "step": 6214
    },
    {
      "epoch": 4.199391686380534,
      "grad_norm": 4.773645877838135,
      "learning_rate": 2e-05,
      "loss": 0.1607,
      "step": 6215
    },
    {
      "epoch": 4.2000675904021625,
      "grad_norm": 4.244959831237793,
      "learning_rate": 2e-05,
      "loss": 0.1379,
      "step": 6216
    },
    {
      "epoch": 4.200743494423792,
      "grad_norm": 4.728854656219482,
      "learning_rate": 2e-05,
      "loss": 0.1614,
      "step": 6217
    },
    {
      "epoch": 4.201419398445421,
      "grad_norm": 8.650041580200195,
      "learning_rate": 2e-05,
      "loss": 0.2074,
      "step": 6218
    },
    {
      "epoch": 4.2020953024670495,
      "grad_norm": 3.160224199295044,
      "learning_rate": 2e-05,
      "loss": 0.1332,
      "step": 6219
    },
    {
      "epoch": 4.202771206488679,
      "grad_norm": 2.189758062362671,
      "learning_rate": 2e-05,
      "loss": 0.137,
      "step": 6220
    },
    {
      "epoch": 4.203447110510307,
      "grad_norm": 3.1668756008148193,
      "learning_rate": 2e-05,
      "loss": 0.112,
      "step": 6221
    },
    {
      "epoch": 4.204123014531937,
      "grad_norm": 4.332049369812012,
      "learning_rate": 2e-05,
      "loss": 0.1492,
      "step": 6222
    },
    {
      "epoch": 4.204798918553565,
      "grad_norm": 4.899263858795166,
      "learning_rate": 2e-05,
      "loss": 0.2492,
      "step": 6223
    },
    {
      "epoch": 4.205474822575194,
      "grad_norm": 2.3105759620666504,
      "learning_rate": 2e-05,
      "loss": 0.1148,
      "step": 6224
    },
    {
      "epoch": 4.206150726596824,
      "grad_norm": 3.5571889877319336,
      "learning_rate": 2e-05,
      "loss": 0.1595,
      "step": 6225
    },
    {
      "epoch": 4.206826630618452,
      "grad_norm": 3.5122056007385254,
      "learning_rate": 2e-05,
      "loss": 0.2142,
      "step": 6226
    },
    {
      "epoch": 4.207502534640081,
      "grad_norm": 4.0021891593933105,
      "learning_rate": 2e-05,
      "loss": 0.1383,
      "step": 6227
    },
    {
      "epoch": 4.20817843866171,
      "grad_norm": 3.2303950786590576,
      "learning_rate": 2e-05,
      "loss": 0.1389,
      "step": 6228
    },
    {
      "epoch": 4.208854342683339,
      "grad_norm": 1.98818838596344,
      "learning_rate": 2e-05,
      "loss": 0.1097,
      "step": 6229
    },
    {
      "epoch": 4.209530246704968,
      "grad_norm": 3.716153621673584,
      "learning_rate": 2e-05,
      "loss": 0.1713,
      "step": 6230
    },
    {
      "epoch": 4.210206150726597,
      "grad_norm": 4.711181163787842,
      "learning_rate": 2e-05,
      "loss": 0.1494,
      "step": 6231
    },
    {
      "epoch": 4.210882054748225,
      "grad_norm": 6.365814685821533,
      "learning_rate": 2e-05,
      "loss": 0.2384,
      "step": 6232
    },
    {
      "epoch": 4.211557958769855,
      "grad_norm": 2.3826816082000732,
      "learning_rate": 2e-05,
      "loss": 0.1189,
      "step": 6233
    },
    {
      "epoch": 4.212233862791484,
      "grad_norm": 8.769266128540039,
      "learning_rate": 2e-05,
      "loss": 0.218,
      "step": 6234
    },
    {
      "epoch": 4.212909766813112,
      "grad_norm": 10.643434524536133,
      "learning_rate": 2e-05,
      "loss": 0.1591,
      "step": 6235
    },
    {
      "epoch": 4.213585670834742,
      "grad_norm": 2.2679474353790283,
      "learning_rate": 2e-05,
      "loss": 0.08,
      "step": 6236
    },
    {
      "epoch": 4.21426157485637,
      "grad_norm": 10.383645057678223,
      "learning_rate": 2e-05,
      "loss": 0.2003,
      "step": 6237
    },
    {
      "epoch": 4.214937478877999,
      "grad_norm": 3.6710572242736816,
      "learning_rate": 2e-05,
      "loss": 0.1523,
      "step": 6238
    },
    {
      "epoch": 4.215613382899628,
      "grad_norm": 5.640382289886475,
      "learning_rate": 2e-05,
      "loss": 0.1812,
      "step": 6239
    },
    {
      "epoch": 4.216289286921257,
      "grad_norm": 7.724191188812256,
      "learning_rate": 2e-05,
      "loss": 0.2181,
      "step": 6240
    },
    {
      "epoch": 4.2169651909428865,
      "grad_norm": 4.718608379364014,
      "learning_rate": 2e-05,
      "loss": 0.1196,
      "step": 6241
    },
    {
      "epoch": 4.217641094964515,
      "grad_norm": 8.241337776184082,
      "learning_rate": 2e-05,
      "loss": 0.163,
      "step": 6242
    },
    {
      "epoch": 4.218316998986144,
      "grad_norm": 4.382613182067871,
      "learning_rate": 2e-05,
      "loss": 0.1071,
      "step": 6243
    },
    {
      "epoch": 4.218992903007773,
      "grad_norm": 1.781718373298645,
      "learning_rate": 2e-05,
      "loss": 0.1223,
      "step": 6244
    },
    {
      "epoch": 4.219668807029402,
      "grad_norm": 6.858415126800537,
      "learning_rate": 2e-05,
      "loss": 0.2733,
      "step": 6245
    },
    {
      "epoch": 4.22034471105103,
      "grad_norm": 3.505146026611328,
      "learning_rate": 2e-05,
      "loss": 0.1142,
      "step": 6246
    },
    {
      "epoch": 4.22102061507266,
      "grad_norm": 3.6452269554138184,
      "learning_rate": 2e-05,
      "loss": 0.1564,
      "step": 6247
    },
    {
      "epoch": 4.221696519094289,
      "grad_norm": 3.8925397396087646,
      "learning_rate": 2e-05,
      "loss": 0.1453,
      "step": 6248
    },
    {
      "epoch": 4.222372423115917,
      "grad_norm": 2.901346206665039,
      "learning_rate": 2e-05,
      "loss": 0.1717,
      "step": 6249
    },
    {
      "epoch": 4.223048327137547,
      "grad_norm": 1.7630319595336914,
      "learning_rate": 2e-05,
      "loss": 0.1153,
      "step": 6250
    },
    {
      "epoch": 4.223048327137547,
      "eval_accuracy": 0.7349662162162162,
      "eval_loss": 0.22886045277118683,
      "eval_runtime": 60.8388,
      "eval_samples_per_second": 389.225,
      "eval_steps_per_second": 12.163,
      "step": 6250
    },
    {
      "epoch": 4.223724231159175,
      "grad_norm": 4.457579612731934,
      "learning_rate": 2e-05,
      "loss": 0.1321,
      "step": 6251
    },
    {
      "epoch": 4.2244001351808045,
      "grad_norm": 2.9045870304107666,
      "learning_rate": 2e-05,
      "loss": 0.1607,
      "step": 6252
    },
    {
      "epoch": 4.225076039202433,
      "grad_norm": 4.352908611297607,
      "learning_rate": 2e-05,
      "loss": 0.1518,
      "step": 6253
    },
    {
      "epoch": 4.225751943224062,
      "grad_norm": 1.9341622591018677,
      "learning_rate": 2e-05,
      "loss": 0.1354,
      "step": 6254
    },
    {
      "epoch": 4.2264278472456915,
      "grad_norm": 2.6578209400177,
      "learning_rate": 2e-05,
      "loss": 0.1766,
      "step": 6255
    },
    {
      "epoch": 4.22710375126732,
      "grad_norm": 4.066863536834717,
      "learning_rate": 2e-05,
      "loss": 0.1651,
      "step": 6256
    },
    {
      "epoch": 4.227779655288949,
      "grad_norm": 2.199343204498291,
      "learning_rate": 2e-05,
      "loss": 0.1418,
      "step": 6257
    },
    {
      "epoch": 4.228455559310578,
      "grad_norm": 4.155644416809082,
      "learning_rate": 2e-05,
      "loss": 0.2119,
      "step": 6258
    },
    {
      "epoch": 4.229131463332207,
      "grad_norm": 3.864563226699829,
      "learning_rate": 2e-05,
      "loss": 0.1439,
      "step": 6259
    },
    {
      "epoch": 4.2298073673538354,
      "grad_norm": 4.810790061950684,
      "learning_rate": 2e-05,
      "loss": 0.1058,
      "step": 6260
    },
    {
      "epoch": 4.230483271375465,
      "grad_norm": 4.936160087585449,
      "learning_rate": 2e-05,
      "loss": 0.1557,
      "step": 6261
    },
    {
      "epoch": 4.231159175397093,
      "grad_norm": 5.872967720031738,
      "learning_rate": 2e-05,
      "loss": 0.1736,
      "step": 6262
    },
    {
      "epoch": 4.2318350794187225,
      "grad_norm": 3.546112537384033,
      "learning_rate": 2e-05,
      "loss": 0.2044,
      "step": 6263
    },
    {
      "epoch": 4.232510983440352,
      "grad_norm": 4.26315450668335,
      "learning_rate": 2e-05,
      "loss": 0.1208,
      "step": 6264
    },
    {
      "epoch": 4.23318688746198,
      "grad_norm": 3.1268808841705322,
      "learning_rate": 2e-05,
      "loss": 0.1678,
      "step": 6265
    },
    {
      "epoch": 4.2338627914836096,
      "grad_norm": 2.0658323764801025,
      "learning_rate": 2e-05,
      "loss": 0.1155,
      "step": 6266
    },
    {
      "epoch": 4.234538695505238,
      "grad_norm": 2.764418363571167,
      "learning_rate": 2e-05,
      "loss": 0.1568,
      "step": 6267
    },
    {
      "epoch": 4.235214599526867,
      "grad_norm": 9.875589370727539,
      "learning_rate": 2e-05,
      "loss": 0.2331,
      "step": 6268
    },
    {
      "epoch": 4.235890503548496,
      "grad_norm": 3.5120975971221924,
      "learning_rate": 2e-05,
      "loss": 0.1342,
      "step": 6269
    },
    {
      "epoch": 4.236566407570125,
      "grad_norm": 4.698256969451904,
      "learning_rate": 2e-05,
      "loss": 0.1086,
      "step": 6270
    },
    {
      "epoch": 4.237242311591754,
      "grad_norm": 6.287431240081787,
      "learning_rate": 2e-05,
      "loss": 0.2198,
      "step": 6271
    },
    {
      "epoch": 4.237918215613383,
      "grad_norm": 2.710313320159912,
      "learning_rate": 2e-05,
      "loss": 0.207,
      "step": 6272
    },
    {
      "epoch": 4.238594119635012,
      "grad_norm": 5.98323917388916,
      "learning_rate": 2e-05,
      "loss": 0.1203,
      "step": 6273
    },
    {
      "epoch": 4.2392700236566405,
      "grad_norm": 4.527634620666504,
      "learning_rate": 2e-05,
      "loss": 0.1244,
      "step": 6274
    },
    {
      "epoch": 4.23994592767827,
      "grad_norm": 7.557475566864014,
      "learning_rate": 2e-05,
      "loss": 0.1397,
      "step": 6275
    },
    {
      "epoch": 4.240621831699898,
      "grad_norm": 5.967245578765869,
      "learning_rate": 2e-05,
      "loss": 0.1537,
      "step": 6276
    },
    {
      "epoch": 4.241297735721528,
      "grad_norm": 2.1880624294281006,
      "learning_rate": 2e-05,
      "loss": 0.1384,
      "step": 6277
    },
    {
      "epoch": 4.241973639743157,
      "grad_norm": 8.486559867858887,
      "learning_rate": 2e-05,
      "loss": 0.1971,
      "step": 6278
    },
    {
      "epoch": 4.242649543764785,
      "grad_norm": 8.058314323425293,
      "learning_rate": 2e-05,
      "loss": 0.2299,
      "step": 6279
    },
    {
      "epoch": 4.243325447786415,
      "grad_norm": 3.577575445175171,
      "learning_rate": 2e-05,
      "loss": 0.1364,
      "step": 6280
    },
    {
      "epoch": 4.244001351808043,
      "grad_norm": 2.4404499530792236,
      "learning_rate": 2e-05,
      "loss": 0.1578,
      "step": 6281
    },
    {
      "epoch": 4.244677255829672,
      "grad_norm": 4.768744468688965,
      "learning_rate": 2e-05,
      "loss": 0.1632,
      "step": 6282
    },
    {
      "epoch": 4.245353159851301,
      "grad_norm": 1.9310773611068726,
      "learning_rate": 2e-05,
      "loss": 0.1195,
      "step": 6283
    },
    {
      "epoch": 4.24602906387293,
      "grad_norm": 2.0303077697753906,
      "learning_rate": 2e-05,
      "loss": 0.1373,
      "step": 6284
    },
    {
      "epoch": 4.2467049678945585,
      "grad_norm": 1.657588005065918,
      "learning_rate": 2e-05,
      "loss": 0.1438,
      "step": 6285
    },
    {
      "epoch": 4.247380871916188,
      "grad_norm": 3.04363751411438,
      "learning_rate": 2e-05,
      "loss": 0.2003,
      "step": 6286
    },
    {
      "epoch": 4.248056775937817,
      "grad_norm": 2.5097854137420654,
      "learning_rate": 2e-05,
      "loss": 0.1277,
      "step": 6287
    },
    {
      "epoch": 4.248732679959446,
      "grad_norm": 2.11625599861145,
      "learning_rate": 2e-05,
      "loss": 0.1442,
      "step": 6288
    },
    {
      "epoch": 4.249408583981075,
      "grad_norm": 2.2927463054656982,
      "learning_rate": 2e-05,
      "loss": 0.1269,
      "step": 6289
    },
    {
      "epoch": 4.250084488002703,
      "grad_norm": 3.4299137592315674,
      "learning_rate": 2e-05,
      "loss": 0.1433,
      "step": 6290
    },
    {
      "epoch": 4.250760392024333,
      "grad_norm": 2.156268835067749,
      "learning_rate": 2e-05,
      "loss": 0.1663,
      "step": 6291
    },
    {
      "epoch": 4.251436296045961,
      "grad_norm": 2.637941360473633,
      "learning_rate": 2e-05,
      "loss": 0.1342,
      "step": 6292
    },
    {
      "epoch": 4.25211220006759,
      "grad_norm": 7.476767539978027,
      "learning_rate": 2e-05,
      "loss": 0.2304,
      "step": 6293
    },
    {
      "epoch": 4.25278810408922,
      "grad_norm": 4.724040508270264,
      "learning_rate": 2e-05,
      "loss": 0.1283,
      "step": 6294
    },
    {
      "epoch": 4.253464008110848,
      "grad_norm": 2.0013720989227295,
      "learning_rate": 2e-05,
      "loss": 0.1261,
      "step": 6295
    },
    {
      "epoch": 4.254139912132477,
      "grad_norm": 2.6530847549438477,
      "learning_rate": 2e-05,
      "loss": 0.1098,
      "step": 6296
    },
    {
      "epoch": 4.254815816154106,
      "grad_norm": 2.836857318878174,
      "learning_rate": 2e-05,
      "loss": 0.1257,
      "step": 6297
    },
    {
      "epoch": 4.255491720175735,
      "grad_norm": 4.687417984008789,
      "learning_rate": 2e-05,
      "loss": 0.106,
      "step": 6298
    },
    {
      "epoch": 4.256167624197364,
      "grad_norm": 6.497343063354492,
      "learning_rate": 2e-05,
      "loss": 0.1879,
      "step": 6299
    },
    {
      "epoch": 4.256843528218993,
      "grad_norm": 3.355236053466797,
      "learning_rate": 2e-05,
      "loss": 0.1514,
      "step": 6300
    },
    {
      "epoch": 4.256843528218993,
      "eval_accuracy": 0.7444679054054054,
      "eval_loss": 0.25060296058654785,
      "eval_runtime": 61.4685,
      "eval_samples_per_second": 385.238,
      "eval_steps_per_second": 12.039,
      "step": 6300
    },
    {
      "epoch": 4.257519432240622,
      "grad_norm": 3.2010152339935303,
      "learning_rate": 2e-05,
      "loss": 0.1677,
      "step": 6301
    },
    {
      "epoch": 4.258195336262251,
      "grad_norm": 3.0971462726593018,
      "learning_rate": 2e-05,
      "loss": 0.107,
      "step": 6302
    },
    {
      "epoch": 4.25887124028388,
      "grad_norm": 2.7723209857940674,
      "learning_rate": 2e-05,
      "loss": 0.1477,
      "step": 6303
    },
    {
      "epoch": 4.259547144305508,
      "grad_norm": 4.167843341827393,
      "learning_rate": 2e-05,
      "loss": 0.1511,
      "step": 6304
    },
    {
      "epoch": 4.260223048327138,
      "grad_norm": 6.610359191894531,
      "learning_rate": 2e-05,
      "loss": 0.1022,
      "step": 6305
    },
    {
      "epoch": 4.260898952348766,
      "grad_norm": 3.6166574954986572,
      "learning_rate": 2e-05,
      "loss": 0.1294,
      "step": 6306
    },
    {
      "epoch": 4.2615748563703955,
      "grad_norm": 3.291466236114502,
      "learning_rate": 2e-05,
      "loss": 0.1376,
      "step": 6307
    },
    {
      "epoch": 4.262250760392025,
      "grad_norm": 11.106415748596191,
      "learning_rate": 2e-05,
      "loss": 0.219,
      "step": 6308
    },
    {
      "epoch": 4.262926664413653,
      "grad_norm": 6.2704243659973145,
      "learning_rate": 2e-05,
      "loss": 0.1366,
      "step": 6309
    },
    {
      "epoch": 4.2636025684352825,
      "grad_norm": 3.157630205154419,
      "learning_rate": 2e-05,
      "loss": 0.1271,
      "step": 6310
    },
    {
      "epoch": 4.264278472456911,
      "grad_norm": 6.6869730949401855,
      "learning_rate": 2e-05,
      "loss": 0.1956,
      "step": 6311
    },
    {
      "epoch": 4.26495437647854,
      "grad_norm": 4.046499729156494,
      "learning_rate": 2e-05,
      "loss": 0.1357,
      "step": 6312
    },
    {
      "epoch": 4.265630280500169,
      "grad_norm": 3.3995604515075684,
      "learning_rate": 2e-05,
      "loss": 0.1197,
      "step": 6313
    },
    {
      "epoch": 4.266306184521798,
      "grad_norm": 3.899919271469116,
      "learning_rate": 2e-05,
      "loss": 0.1213,
      "step": 6314
    },
    {
      "epoch": 4.266982088543426,
      "grad_norm": 3.3012125492095947,
      "learning_rate": 2e-05,
      "loss": 0.1749,
      "step": 6315
    },
    {
      "epoch": 4.267657992565056,
      "grad_norm": 6.381699085235596,
      "learning_rate": 2e-05,
      "loss": 0.2764,
      "step": 6316
    },
    {
      "epoch": 4.268333896586685,
      "grad_norm": 8.678458213806152,
      "learning_rate": 2e-05,
      "loss": 0.1585,
      "step": 6317
    },
    {
      "epoch": 4.2690098006083135,
      "grad_norm": 5.160369873046875,
      "learning_rate": 2e-05,
      "loss": 0.1904,
      "step": 6318
    },
    {
      "epoch": 4.269685704629943,
      "grad_norm": 1.9873143434524536,
      "learning_rate": 2e-05,
      "loss": 0.1157,
      "step": 6319
    },
    {
      "epoch": 4.270361608651571,
      "grad_norm": 3.8704535961151123,
      "learning_rate": 2e-05,
      "loss": 0.1681,
      "step": 6320
    },
    {
      "epoch": 4.2710375126732005,
      "grad_norm": 2.3213744163513184,
      "learning_rate": 2e-05,
      "loss": 0.1421,
      "step": 6321
    },
    {
      "epoch": 4.271713416694829,
      "grad_norm": 3.735292673110962,
      "learning_rate": 2e-05,
      "loss": 0.163,
      "step": 6322
    },
    {
      "epoch": 4.272389320716458,
      "grad_norm": 2.184831380844116,
      "learning_rate": 2e-05,
      "loss": 0.1642,
      "step": 6323
    },
    {
      "epoch": 4.273065224738088,
      "grad_norm": 1.9949991703033447,
      "learning_rate": 2e-05,
      "loss": 0.1106,
      "step": 6324
    },
    {
      "epoch": 4.273741128759716,
      "grad_norm": 3.0041449069976807,
      "learning_rate": 2e-05,
      "loss": 0.1907,
      "step": 6325
    },
    {
      "epoch": 4.274417032781345,
      "grad_norm": 7.438480377197266,
      "learning_rate": 2e-05,
      "loss": 0.2324,
      "step": 6326
    },
    {
      "epoch": 4.275092936802974,
      "grad_norm": 3.772860050201416,
      "learning_rate": 2e-05,
      "loss": 0.1401,
      "step": 6327
    },
    {
      "epoch": 4.275768840824603,
      "grad_norm": 4.770651340484619,
      "learning_rate": 2e-05,
      "loss": 0.175,
      "step": 6328
    },
    {
      "epoch": 4.2764447448462315,
      "grad_norm": 3.2590243816375732,
      "learning_rate": 2e-05,
      "loss": 0.1658,
      "step": 6329
    },
    {
      "epoch": 4.277120648867861,
      "grad_norm": 12.320571899414062,
      "learning_rate": 2e-05,
      "loss": 0.1939,
      "step": 6330
    },
    {
      "epoch": 4.277796552889489,
      "grad_norm": 6.653195381164551,
      "learning_rate": 2e-05,
      "loss": 0.1466,
      "step": 6331
    },
    {
      "epoch": 4.2784724569111185,
      "grad_norm": 2.2682998180389404,
      "learning_rate": 2e-05,
      "loss": 0.12,
      "step": 6332
    },
    {
      "epoch": 4.279148360932748,
      "grad_norm": 2.354823112487793,
      "learning_rate": 2e-05,
      "loss": 0.1879,
      "step": 6333
    },
    {
      "epoch": 4.279824264954376,
      "grad_norm": 3.449767589569092,
      "learning_rate": 2e-05,
      "loss": 0.1402,
      "step": 6334
    },
    {
      "epoch": 4.280500168976006,
      "grad_norm": 2.1937015056610107,
      "learning_rate": 2e-05,
      "loss": 0.1391,
      "step": 6335
    },
    {
      "epoch": 4.281176072997634,
      "grad_norm": 2.9294064044952393,
      "learning_rate": 2e-05,
      "loss": 0.2099,
      "step": 6336
    },
    {
      "epoch": 4.281851977019263,
      "grad_norm": 3.5962491035461426,
      "learning_rate": 2e-05,
      "loss": 0.0665,
      "step": 6337
    },
    {
      "epoch": 4.282527881040892,
      "grad_norm": 5.070093631744385,
      "learning_rate": 2e-05,
      "loss": 0.1252,
      "step": 6338
    },
    {
      "epoch": 4.283203785062521,
      "grad_norm": 1.684593677520752,
      "learning_rate": 2e-05,
      "loss": 0.1153,
      "step": 6339
    },
    {
      "epoch": 4.28387968908415,
      "grad_norm": 1.789787769317627,
      "learning_rate": 2e-05,
      "loss": 0.1077,
      "step": 6340
    },
    {
      "epoch": 4.284555593105779,
      "grad_norm": 7.409839153289795,
      "learning_rate": 2e-05,
      "loss": 0.1462,
      "step": 6341
    },
    {
      "epoch": 4.285231497127408,
      "grad_norm": 5.897705554962158,
      "learning_rate": 2e-05,
      "loss": 0.1673,
      "step": 6342
    },
    {
      "epoch": 4.285907401149037,
      "grad_norm": 4.544023513793945,
      "learning_rate": 2e-05,
      "loss": 0.1686,
      "step": 6343
    },
    {
      "epoch": 4.286583305170666,
      "grad_norm": 3.8346877098083496,
      "learning_rate": 2e-05,
      "loss": 0.1664,
      "step": 6344
    },
    {
      "epoch": 4.287259209192294,
      "grad_norm": 4.428942680358887,
      "learning_rate": 2e-05,
      "loss": 0.1886,
      "step": 6345
    },
    {
      "epoch": 4.287935113213924,
      "grad_norm": 4.9024128913879395,
      "learning_rate": 2e-05,
      "loss": 0.1912,
      "step": 6346
    },
    {
      "epoch": 4.288611017235553,
      "grad_norm": 2.218890905380249,
      "learning_rate": 2e-05,
      "loss": 0.1123,
      "step": 6347
    },
    {
      "epoch": 4.289286921257181,
      "grad_norm": 3.7790310382843018,
      "learning_rate": 2e-05,
      "loss": 0.2448,
      "step": 6348
    },
    {
      "epoch": 4.289962825278811,
      "grad_norm": 3.6292028427124023,
      "learning_rate": 2e-05,
      "loss": 0.1581,
      "step": 6349
    },
    {
      "epoch": 4.290638729300439,
      "grad_norm": 4.355430603027344,
      "learning_rate": 2e-05,
      "loss": 0.1291,
      "step": 6350
    },
    {
      "epoch": 4.290638729300439,
      "eval_accuracy": 0.7247466216216216,
      "eval_loss": 0.2413736879825592,
      "eval_runtime": 64.2319,
      "eval_samples_per_second": 368.664,
      "eval_steps_per_second": 11.521,
      "step": 6350
    },
    {
      "epoch": 4.291314633322068,
      "grad_norm": 4.551562309265137,
      "learning_rate": 2e-05,
      "loss": 0.2239,
      "step": 6351
    },
    {
      "epoch": 4.291990537343697,
      "grad_norm": 2.4661574363708496,
      "learning_rate": 2e-05,
      "loss": 0.1549,
      "step": 6352
    },
    {
      "epoch": 4.292666441365326,
      "grad_norm": 2.6178834438323975,
      "learning_rate": 2e-05,
      "loss": 0.1614,
      "step": 6353
    },
    {
      "epoch": 4.2933423453869555,
      "grad_norm": 3.4534411430358887,
      "learning_rate": 2e-05,
      "loss": 0.1401,
      "step": 6354
    },
    {
      "epoch": 4.294018249408584,
      "grad_norm": 4.011856555938721,
      "learning_rate": 2e-05,
      "loss": 0.1492,
      "step": 6355
    },
    {
      "epoch": 4.294694153430213,
      "grad_norm": 2.9575629234313965,
      "learning_rate": 2e-05,
      "loss": 0.0988,
      "step": 6356
    },
    {
      "epoch": 4.295370057451842,
      "grad_norm": 5.611600875854492,
      "learning_rate": 2e-05,
      "loss": 0.195,
      "step": 6357
    },
    {
      "epoch": 4.296045961473471,
      "grad_norm": 4.627816200256348,
      "learning_rate": 2e-05,
      "loss": 0.1146,
      "step": 6358
    },
    {
      "epoch": 4.296721865495099,
      "grad_norm": 6.069965362548828,
      "learning_rate": 2e-05,
      "loss": 0.2332,
      "step": 6359
    },
    {
      "epoch": 4.297397769516729,
      "grad_norm": 4.252265453338623,
      "learning_rate": 2e-05,
      "loss": 0.168,
      "step": 6360
    },
    {
      "epoch": 4.298073673538358,
      "grad_norm": 1.731764316558838,
      "learning_rate": 2e-05,
      "loss": 0.1117,
      "step": 6361
    },
    {
      "epoch": 4.298749577559986,
      "grad_norm": 4.257668972015381,
      "learning_rate": 2e-05,
      "loss": 0.1968,
      "step": 6362
    },
    {
      "epoch": 4.299425481581616,
      "grad_norm": 2.0639965534210205,
      "learning_rate": 2e-05,
      "loss": 0.1821,
      "step": 6363
    },
    {
      "epoch": 4.300101385603244,
      "grad_norm": 4.739585876464844,
      "learning_rate": 2e-05,
      "loss": 0.1458,
      "step": 6364
    },
    {
      "epoch": 4.3007772896248735,
      "grad_norm": 2.3721628189086914,
      "learning_rate": 2e-05,
      "loss": 0.1593,
      "step": 6365
    },
    {
      "epoch": 4.301453193646502,
      "grad_norm": 4.923586845397949,
      "learning_rate": 2e-05,
      "loss": 0.132,
      "step": 6366
    },
    {
      "epoch": 4.302129097668131,
      "grad_norm": 2.357184648513794,
      "learning_rate": 2e-05,
      "loss": 0.1533,
      "step": 6367
    },
    {
      "epoch": 4.30280500168976,
      "grad_norm": 5.144529819488525,
      "learning_rate": 2e-05,
      "loss": 0.1692,
      "step": 6368
    },
    {
      "epoch": 4.303480905711389,
      "grad_norm": 6.40391206741333,
      "learning_rate": 2e-05,
      "loss": 0.142,
      "step": 6369
    },
    {
      "epoch": 4.304156809733018,
      "grad_norm": 4.593046188354492,
      "learning_rate": 2e-05,
      "loss": 0.1979,
      "step": 6370
    },
    {
      "epoch": 4.304832713754647,
      "grad_norm": 3.503392457962036,
      "learning_rate": 2e-05,
      "loss": 0.1358,
      "step": 6371
    },
    {
      "epoch": 4.305508617776276,
      "grad_norm": 2.807002305984497,
      "learning_rate": 2e-05,
      "loss": 0.1757,
      "step": 6372
    },
    {
      "epoch": 4.3061845217979045,
      "grad_norm": 3.415738105773926,
      "learning_rate": 2e-05,
      "loss": 0.2215,
      "step": 6373
    },
    {
      "epoch": 4.306860425819534,
      "grad_norm": 2.2964820861816406,
      "learning_rate": 2e-05,
      "loss": 0.1641,
      "step": 6374
    },
    {
      "epoch": 4.307536329841162,
      "grad_norm": 3.43096661567688,
      "learning_rate": 2e-05,
      "loss": 0.201,
      "step": 6375
    },
    {
      "epoch": 4.3082122338627915,
      "grad_norm": 2.40718936920166,
      "learning_rate": 2e-05,
      "loss": 0.1329,
      "step": 6376
    },
    {
      "epoch": 4.308888137884421,
      "grad_norm": 2.766655683517456,
      "learning_rate": 2e-05,
      "loss": 0.1298,
      "step": 6377
    },
    {
      "epoch": 4.309564041906049,
      "grad_norm": 6.827731132507324,
      "learning_rate": 2e-05,
      "loss": 0.1736,
      "step": 6378
    },
    {
      "epoch": 4.310239945927679,
      "grad_norm": 1.9158296585083008,
      "learning_rate": 2e-05,
      "loss": 0.1347,
      "step": 6379
    },
    {
      "epoch": 4.310915849949307,
      "grad_norm": 3.325432538986206,
      "learning_rate": 2e-05,
      "loss": 0.1843,
      "step": 6380
    },
    {
      "epoch": 4.311591753970936,
      "grad_norm": 3.584296941757202,
      "learning_rate": 2e-05,
      "loss": 0.2533,
      "step": 6381
    },
    {
      "epoch": 4.312267657992565,
      "grad_norm": 3.094247341156006,
      "learning_rate": 2e-05,
      "loss": 0.1963,
      "step": 6382
    },
    {
      "epoch": 4.312943562014194,
      "grad_norm": 2.172438144683838,
      "learning_rate": 2e-05,
      "loss": 0.1525,
      "step": 6383
    },
    {
      "epoch": 4.3136194660358225,
      "grad_norm": 2.3058276176452637,
      "learning_rate": 2e-05,
      "loss": 0.1625,
      "step": 6384
    },
    {
      "epoch": 4.314295370057452,
      "grad_norm": 2.643976926803589,
      "learning_rate": 2e-05,
      "loss": 0.1555,
      "step": 6385
    },
    {
      "epoch": 4.314971274079081,
      "grad_norm": 2.7898592948913574,
      "learning_rate": 2e-05,
      "loss": 0.1618,
      "step": 6386
    },
    {
      "epoch": 4.3156471781007095,
      "grad_norm": 3.3077125549316406,
      "learning_rate": 2e-05,
      "loss": 0.1632,
      "step": 6387
    },
    {
      "epoch": 4.316323082122339,
      "grad_norm": 3.304051399230957,
      "learning_rate": 2e-05,
      "loss": 0.1965,
      "step": 6388
    },
    {
      "epoch": 4.316998986143967,
      "grad_norm": 2.133599281311035,
      "learning_rate": 2e-05,
      "loss": 0.204,
      "step": 6389
    },
    {
      "epoch": 4.317674890165597,
      "grad_norm": 2.7835164070129395,
      "learning_rate": 2e-05,
      "loss": 0.1644,
      "step": 6390
    },
    {
      "epoch": 4.318350794187225,
      "grad_norm": 5.838906764984131,
      "learning_rate": 2e-05,
      "loss": 0.1715,
      "step": 6391
    },
    {
      "epoch": 4.319026698208854,
      "grad_norm": 3.78214430809021,
      "learning_rate": 2e-05,
      "loss": 0.2181,
      "step": 6392
    },
    {
      "epoch": 4.319702602230484,
      "grad_norm": 5.189573287963867,
      "learning_rate": 2e-05,
      "loss": 0.1831,
      "step": 6393
    },
    {
      "epoch": 4.320378506252112,
      "grad_norm": 4.432830810546875,
      "learning_rate": 2e-05,
      "loss": 0.1676,
      "step": 6394
    },
    {
      "epoch": 4.321054410273741,
      "grad_norm": 2.3511993885040283,
      "learning_rate": 2e-05,
      "loss": 0.156,
      "step": 6395
    },
    {
      "epoch": 4.32173031429537,
      "grad_norm": 2.8094465732574463,
      "learning_rate": 2e-05,
      "loss": 0.1288,
      "step": 6396
    },
    {
      "epoch": 4.322406218316999,
      "grad_norm": 7.155848026275635,
      "learning_rate": 2e-05,
      "loss": 0.2088,
      "step": 6397
    },
    {
      "epoch": 4.3230821223386275,
      "grad_norm": 2.6292026042938232,
      "learning_rate": 2e-05,
      "loss": 0.1541,
      "step": 6398
    },
    {
      "epoch": 4.323758026360257,
      "grad_norm": 5.116791248321533,
      "learning_rate": 2e-05,
      "loss": 0.1389,
      "step": 6399
    },
    {
      "epoch": 4.324433930381886,
      "grad_norm": 4.447173118591309,
      "learning_rate": 2e-05,
      "loss": 0.1477,
      "step": 6400
    },
    {
      "epoch": 4.324433930381886,
      "eval_accuracy": 0.7051097972972973,
      "eval_loss": 0.23107048869132996,
      "eval_runtime": 56.1921,
      "eval_samples_per_second": 421.411,
      "eval_steps_per_second": 13.169,
      "step": 6400
    },
    {
      "epoch": 4.325109834403515,
      "grad_norm": 5.419342994689941,
      "learning_rate": 2e-05,
      "loss": 0.1856,
      "step": 6401
    },
    {
      "epoch": 4.325785738425144,
      "grad_norm": 5.398038387298584,
      "learning_rate": 2e-05,
      "loss": 0.1312,
      "step": 6402
    },
    {
      "epoch": 4.326461642446772,
      "grad_norm": 2.1435041427612305,
      "learning_rate": 2e-05,
      "loss": 0.1701,
      "step": 6403
    },
    {
      "epoch": 4.327137546468402,
      "grad_norm": 8.747164726257324,
      "learning_rate": 2e-05,
      "loss": 0.1596,
      "step": 6404
    },
    {
      "epoch": 4.32781345049003,
      "grad_norm": 2.2230563163757324,
      "learning_rate": 2e-05,
      "loss": 0.1325,
      "step": 6405
    },
    {
      "epoch": 4.328489354511659,
      "grad_norm": 5.6497321128845215,
      "learning_rate": 2e-05,
      "loss": 0.147,
      "step": 6406
    },
    {
      "epoch": 4.329165258533289,
      "grad_norm": 2.4256410598754883,
      "learning_rate": 2e-05,
      "loss": 0.1388,
      "step": 6407
    },
    {
      "epoch": 4.329841162554917,
      "grad_norm": 3.764725685119629,
      "learning_rate": 2e-05,
      "loss": 0.1455,
      "step": 6408
    },
    {
      "epoch": 4.3305170665765464,
      "grad_norm": 2.5897576808929443,
      "learning_rate": 2e-05,
      "loss": 0.1467,
      "step": 6409
    },
    {
      "epoch": 4.331192970598175,
      "grad_norm": 5.361947536468506,
      "learning_rate": 2e-05,
      "loss": 0.2208,
      "step": 6410
    },
    {
      "epoch": 4.331868874619804,
      "grad_norm": 3.1437551975250244,
      "learning_rate": 2e-05,
      "loss": 0.1311,
      "step": 6411
    },
    {
      "epoch": 4.332544778641433,
      "grad_norm": 4.72548246383667,
      "learning_rate": 2e-05,
      "loss": 0.2159,
      "step": 6412
    },
    {
      "epoch": 4.333220682663062,
      "grad_norm": 5.126175880432129,
      "learning_rate": 2e-05,
      "loss": 0.1329,
      "step": 6413
    },
    {
      "epoch": 4.333896586684691,
      "grad_norm": 2.9760944843292236,
      "learning_rate": 2e-05,
      "loss": 0.1639,
      "step": 6414
    },
    {
      "epoch": 4.33457249070632,
      "grad_norm": 3.7497575283050537,
      "learning_rate": 2e-05,
      "loss": 0.1475,
      "step": 6415
    },
    {
      "epoch": 4.335248394727949,
      "grad_norm": 4.943502902984619,
      "learning_rate": 2e-05,
      "loss": 0.1187,
      "step": 6416
    },
    {
      "epoch": 4.335924298749577,
      "grad_norm": 3.554804563522339,
      "learning_rate": 2e-05,
      "loss": 0.2177,
      "step": 6417
    },
    {
      "epoch": 4.336600202771207,
      "grad_norm": 3.726170063018799,
      "learning_rate": 2e-05,
      "loss": 0.2054,
      "step": 6418
    },
    {
      "epoch": 4.337276106792835,
      "grad_norm": 2.5158603191375732,
      "learning_rate": 2e-05,
      "loss": 0.1189,
      "step": 6419
    },
    {
      "epoch": 4.3379520108144645,
      "grad_norm": 3.184405565261841,
      "learning_rate": 2e-05,
      "loss": 0.1506,
      "step": 6420
    },
    {
      "epoch": 4.338627914836093,
      "grad_norm": 2.5135130882263184,
      "learning_rate": 2e-05,
      "loss": 0.1669,
      "step": 6421
    },
    {
      "epoch": 4.339303818857722,
      "grad_norm": 2.7867069244384766,
      "learning_rate": 2e-05,
      "loss": 0.1583,
      "step": 6422
    },
    {
      "epoch": 4.3399797228793515,
      "grad_norm": 3.7407827377319336,
      "learning_rate": 2e-05,
      "loss": 0.1764,
      "step": 6423
    },
    {
      "epoch": 4.34065562690098,
      "grad_norm": 4.0060954093933105,
      "learning_rate": 2e-05,
      "loss": 0.2276,
      "step": 6424
    },
    {
      "epoch": 4.341331530922609,
      "grad_norm": 3.941894292831421,
      "learning_rate": 2e-05,
      "loss": 0.214,
      "step": 6425
    },
    {
      "epoch": 4.342007434944238,
      "grad_norm": 3.3528735637664795,
      "learning_rate": 2e-05,
      "loss": 0.1672,
      "step": 6426
    },
    {
      "epoch": 4.342683338965867,
      "grad_norm": 6.632089138031006,
      "learning_rate": 2e-05,
      "loss": 0.1543,
      "step": 6427
    },
    {
      "epoch": 4.343359242987495,
      "grad_norm": 3.6761956214904785,
      "learning_rate": 2e-05,
      "loss": 0.1617,
      "step": 6428
    },
    {
      "epoch": 4.344035147009125,
      "grad_norm": 1.8392963409423828,
      "learning_rate": 2e-05,
      "loss": 0.1065,
      "step": 6429
    },
    {
      "epoch": 4.344711051030753,
      "grad_norm": 4.293346881866455,
      "learning_rate": 2e-05,
      "loss": 0.2385,
      "step": 6430
    },
    {
      "epoch": 4.3453869550523825,
      "grad_norm": 4.772795677185059,
      "learning_rate": 2e-05,
      "loss": 0.1239,
      "step": 6431
    },
    {
      "epoch": 4.346062859074012,
      "grad_norm": 4.118333339691162,
      "learning_rate": 2e-05,
      "loss": 0.174,
      "step": 6432
    },
    {
      "epoch": 4.34673876309564,
      "grad_norm": 3.891658306121826,
      "learning_rate": 2e-05,
      "loss": 0.162,
      "step": 6433
    },
    {
      "epoch": 4.3474146671172695,
      "grad_norm": 2.1462318897247314,
      "learning_rate": 2e-05,
      "loss": 0.2065,
      "step": 6434
    },
    {
      "epoch": 4.348090571138898,
      "grad_norm": 2.850698709487915,
      "learning_rate": 2e-05,
      "loss": 0.18,
      "step": 6435
    },
    {
      "epoch": 4.348766475160527,
      "grad_norm": 3.048020124435425,
      "learning_rate": 2e-05,
      "loss": 0.1328,
      "step": 6436
    },
    {
      "epoch": 4.349442379182156,
      "grad_norm": 2.78159499168396,
      "learning_rate": 2e-05,
      "loss": 0.1841,
      "step": 6437
    },
    {
      "epoch": 4.350118283203785,
      "grad_norm": 6.369277000427246,
      "learning_rate": 2e-05,
      "loss": 0.1582,
      "step": 6438
    },
    {
      "epoch": 4.350794187225414,
      "grad_norm": 3.5200629234313965,
      "learning_rate": 2e-05,
      "loss": 0.154,
      "step": 6439
    },
    {
      "epoch": 4.351470091247043,
      "grad_norm": 3.7728569507598877,
      "learning_rate": 2e-05,
      "loss": 0.1773,
      "step": 6440
    },
    {
      "epoch": 4.352145995268672,
      "grad_norm": 8.868468284606934,
      "learning_rate": 2e-05,
      "loss": 0.2037,
      "step": 6441
    },
    {
      "epoch": 4.3528218992903005,
      "grad_norm": 3.842400074005127,
      "learning_rate": 2e-05,
      "loss": 0.1642,
      "step": 6442
    },
    {
      "epoch": 4.35349780331193,
      "grad_norm": 4.612904071807861,
      "learning_rate": 2e-05,
      "loss": 0.1696,
      "step": 6443
    },
    {
      "epoch": 4.354173707333558,
      "grad_norm": 2.115467071533203,
      "learning_rate": 2e-05,
      "loss": 0.173,
      "step": 6444
    },
    {
      "epoch": 4.3548496113551876,
      "grad_norm": 3.3368780612945557,
      "learning_rate": 2e-05,
      "loss": 0.1418,
      "step": 6445
    },
    {
      "epoch": 4.355525515376817,
      "grad_norm": 2.298999547958374,
      "learning_rate": 2e-05,
      "loss": 0.1876,
      "step": 6446
    },
    {
      "epoch": 4.356201419398445,
      "grad_norm": 4.487280368804932,
      "learning_rate": 2e-05,
      "loss": 0.1756,
      "step": 6447
    },
    {
      "epoch": 4.356877323420075,
      "grad_norm": 2.6059324741363525,
      "learning_rate": 2e-05,
      "loss": 0.1994,
      "step": 6448
    },
    {
      "epoch": 4.357553227441703,
      "grad_norm": 8.945523262023926,
      "learning_rate": 2e-05,
      "loss": 0.1695,
      "step": 6449
    },
    {
      "epoch": 4.358229131463332,
      "grad_norm": 6.379693031311035,
      "learning_rate": 2e-05,
      "loss": 0.2201,
      "step": 6450
    },
    {
      "epoch": 4.358229131463332,
      "eval_accuracy": 0.7116131756756757,
      "eval_loss": 0.23011982440948486,
      "eval_runtime": 57.4365,
      "eval_samples_per_second": 412.282,
      "eval_steps_per_second": 12.884,
      "step": 6450
    },
    {
      "epoch": 4.358905035484961,
      "grad_norm": 2.4231324195861816,
      "learning_rate": 2e-05,
      "loss": 0.1454,
      "step": 6451
    },
    {
      "epoch": 4.35958093950659,
      "grad_norm": 3.702319622039795,
      "learning_rate": 2e-05,
      "loss": 0.1594,
      "step": 6452
    },
    {
      "epoch": 4.360256843528219,
      "grad_norm": 2.2108070850372314,
      "learning_rate": 2e-05,
      "loss": 0.1979,
      "step": 6453
    },
    {
      "epoch": 4.360932747549848,
      "grad_norm": 1.7959365844726562,
      "learning_rate": 2e-05,
      "loss": 0.1367,
      "step": 6454
    },
    {
      "epoch": 4.361608651571477,
      "grad_norm": 2.093634843826294,
      "learning_rate": 2e-05,
      "loss": 0.1322,
      "step": 6455
    },
    {
      "epoch": 4.362284555593106,
      "grad_norm": 2.193892002105713,
      "learning_rate": 2e-05,
      "loss": 0.1346,
      "step": 6456
    },
    {
      "epoch": 4.362960459614735,
      "grad_norm": 2.5986781120300293,
      "learning_rate": 2e-05,
      "loss": 0.0954,
      "step": 6457
    },
    {
      "epoch": 4.363636363636363,
      "grad_norm": 4.345341682434082,
      "learning_rate": 2e-05,
      "loss": 0.1692,
      "step": 6458
    },
    {
      "epoch": 4.364312267657993,
      "grad_norm": 1.9360722303390503,
      "learning_rate": 2e-05,
      "loss": 0.1105,
      "step": 6459
    },
    {
      "epoch": 4.364988171679622,
      "grad_norm": 4.413333892822266,
      "learning_rate": 2e-05,
      "loss": 0.1546,
      "step": 6460
    },
    {
      "epoch": 4.36566407570125,
      "grad_norm": 5.969544410705566,
      "learning_rate": 2e-05,
      "loss": 0.1916,
      "step": 6461
    },
    {
      "epoch": 4.36633997972288,
      "grad_norm": 3.0931875705718994,
      "learning_rate": 2e-05,
      "loss": 0.1398,
      "step": 6462
    },
    {
      "epoch": 4.367015883744508,
      "grad_norm": 4.944656848907471,
      "learning_rate": 2e-05,
      "loss": 0.1564,
      "step": 6463
    },
    {
      "epoch": 4.367691787766137,
      "grad_norm": 3.8119022846221924,
      "learning_rate": 2e-05,
      "loss": 0.143,
      "step": 6464
    },
    {
      "epoch": 4.368367691787766,
      "grad_norm": 3.495459794998169,
      "learning_rate": 2e-05,
      "loss": 0.1274,
      "step": 6465
    },
    {
      "epoch": 4.369043595809395,
      "grad_norm": 3.599719762802124,
      "learning_rate": 2e-05,
      "loss": 0.1642,
      "step": 6466
    },
    {
      "epoch": 4.369719499831024,
      "grad_norm": 3.968644142150879,
      "learning_rate": 2e-05,
      "loss": 0.1648,
      "step": 6467
    },
    {
      "epoch": 4.370395403852653,
      "grad_norm": 6.771524906158447,
      "learning_rate": 2e-05,
      "loss": 0.1654,
      "step": 6468
    },
    {
      "epoch": 4.371071307874282,
      "grad_norm": 5.315611362457275,
      "learning_rate": 2e-05,
      "loss": 0.1427,
      "step": 6469
    },
    {
      "epoch": 4.371747211895911,
      "grad_norm": 3.335501194000244,
      "learning_rate": 2e-05,
      "loss": 0.1432,
      "step": 6470
    },
    {
      "epoch": 4.37242311591754,
      "grad_norm": 2.9818289279937744,
      "learning_rate": 2e-05,
      "loss": 0.0924,
      "step": 6471
    },
    {
      "epoch": 4.373099019939168,
      "grad_norm": 7.898040771484375,
      "learning_rate": 2e-05,
      "loss": 0.2009,
      "step": 6472
    },
    {
      "epoch": 4.373774923960798,
      "grad_norm": 5.718786716461182,
      "learning_rate": 2e-05,
      "loss": 0.1836,
      "step": 6473
    },
    {
      "epoch": 4.374450827982426,
      "grad_norm": 5.88620662689209,
      "learning_rate": 2e-05,
      "loss": 0.1138,
      "step": 6474
    },
    {
      "epoch": 4.375126732004055,
      "grad_norm": 3.5165703296661377,
      "learning_rate": 2e-05,
      "loss": 0.0938,
      "step": 6475
    },
    {
      "epoch": 4.375802636025685,
      "grad_norm": 4.3072638511657715,
      "learning_rate": 2e-05,
      "loss": 0.1677,
      "step": 6476
    },
    {
      "epoch": 4.376478540047313,
      "grad_norm": 3.9218108654022217,
      "learning_rate": 2e-05,
      "loss": 0.1514,
      "step": 6477
    },
    {
      "epoch": 4.3771544440689425,
      "grad_norm": 4.443016529083252,
      "learning_rate": 2e-05,
      "loss": 0.1319,
      "step": 6478
    },
    {
      "epoch": 4.377830348090571,
      "grad_norm": 12.0404634475708,
      "learning_rate": 2e-05,
      "loss": 0.2414,
      "step": 6479
    },
    {
      "epoch": 4.3785062521122,
      "grad_norm": 2.782980442047119,
      "learning_rate": 2e-05,
      "loss": 0.1147,
      "step": 6480
    },
    {
      "epoch": 4.379182156133829,
      "grad_norm": 3.088489532470703,
      "learning_rate": 2e-05,
      "loss": 0.1228,
      "step": 6481
    },
    {
      "epoch": 4.379858060155458,
      "grad_norm": 7.6337995529174805,
      "learning_rate": 2e-05,
      "loss": 0.2563,
      "step": 6482
    },
    {
      "epoch": 4.380533964177086,
      "grad_norm": 8.131879806518555,
      "learning_rate": 2e-05,
      "loss": 0.1508,
      "step": 6483
    },
    {
      "epoch": 4.381209868198716,
      "grad_norm": 7.578224182128906,
      "learning_rate": 2e-05,
      "loss": 0.1679,
      "step": 6484
    },
    {
      "epoch": 4.381885772220345,
      "grad_norm": 7.6753339767456055,
      "learning_rate": 2e-05,
      "loss": 0.1593,
      "step": 6485
    },
    {
      "epoch": 4.3825616762419735,
      "grad_norm": 3.745915651321411,
      "learning_rate": 2e-05,
      "loss": 0.1882,
      "step": 6486
    },
    {
      "epoch": 4.383237580263603,
      "grad_norm": 2.959418296813965,
      "learning_rate": 2e-05,
      "loss": 0.1411,
      "step": 6487
    },
    {
      "epoch": 4.383913484285231,
      "grad_norm": 1.8714497089385986,
      "learning_rate": 2e-05,
      "loss": 0.0954,
      "step": 6488
    },
    {
      "epoch": 4.3845893883068605,
      "grad_norm": 1.6931649446487427,
      "learning_rate": 2e-05,
      "loss": 0.1023,
      "step": 6489
    },
    {
      "epoch": 4.385265292328489,
      "grad_norm": 3.787843704223633,
      "learning_rate": 2e-05,
      "loss": 0.215,
      "step": 6490
    },
    {
      "epoch": 4.385941196350118,
      "grad_norm": 2.0924072265625,
      "learning_rate": 2e-05,
      "loss": 0.1317,
      "step": 6491
    },
    {
      "epoch": 4.386617100371748,
      "grad_norm": 3.394308567047119,
      "learning_rate": 2e-05,
      "loss": 0.1939,
      "step": 6492
    },
    {
      "epoch": 4.387293004393376,
      "grad_norm": 4.180185317993164,
      "learning_rate": 2e-05,
      "loss": 0.145,
      "step": 6493
    },
    {
      "epoch": 4.387968908415005,
      "grad_norm": 2.907573699951172,
      "learning_rate": 2e-05,
      "loss": 0.0896,
      "step": 6494
    },
    {
      "epoch": 4.388644812436634,
      "grad_norm": 3.522991895675659,
      "learning_rate": 2e-05,
      "loss": 0.201,
      "step": 6495
    },
    {
      "epoch": 4.389320716458263,
      "grad_norm": 1.8020836114883423,
      "learning_rate": 2e-05,
      "loss": 0.1377,
      "step": 6496
    },
    {
      "epoch": 4.3899966204798915,
      "grad_norm": 3.592444896697998,
      "learning_rate": 2e-05,
      "loss": 0.2384,
      "step": 6497
    },
    {
      "epoch": 4.390672524501521,
      "grad_norm": 2.381391763687134,
      "learning_rate": 2e-05,
      "loss": 0.1338,
      "step": 6498
    },
    {
      "epoch": 4.39134842852315,
      "grad_norm": 5.525143623352051,
      "learning_rate": 2e-05,
      "loss": 0.1765,
      "step": 6499
    },
    {
      "epoch": 4.3920243325447785,
      "grad_norm": 2.381915807723999,
      "learning_rate": 2e-05,
      "loss": 0.107,
      "step": 6500
    },
    {
      "epoch": 4.3920243325447785,
      "eval_accuracy": 0.733910472972973,
      "eval_loss": 0.23830172419548035,
      "eval_runtime": 61.8529,
      "eval_samples_per_second": 382.844,
      "eval_steps_per_second": 11.964,
      "step": 6500
    },
    {
      "epoch": 4.392700236566408,
      "grad_norm": 5.6059699058532715,
      "learning_rate": 2e-05,
      "loss": 0.2101,
      "step": 6501
    },
    {
      "epoch": 4.393376140588036,
      "grad_norm": 1.8617867231369019,
      "learning_rate": 2e-05,
      "loss": 0.1404,
      "step": 6502
    },
    {
      "epoch": 4.394052044609666,
      "grad_norm": 3.3954803943634033,
      "learning_rate": 2e-05,
      "loss": 0.1417,
      "step": 6503
    },
    {
      "epoch": 4.394727948631294,
      "grad_norm": 1.6981620788574219,
      "learning_rate": 2e-05,
      "loss": 0.1492,
      "step": 6504
    },
    {
      "epoch": 4.395403852652923,
      "grad_norm": 1.7463746070861816,
      "learning_rate": 2e-05,
      "loss": 0.1561,
      "step": 6505
    },
    {
      "epoch": 4.396079756674553,
      "grad_norm": 11.235346794128418,
      "learning_rate": 2e-05,
      "loss": 0.1527,
      "step": 6506
    },
    {
      "epoch": 4.396755660696181,
      "grad_norm": 3.696848154067993,
      "learning_rate": 2e-05,
      "loss": 0.1828,
      "step": 6507
    },
    {
      "epoch": 4.39743156471781,
      "grad_norm": 2.948324203491211,
      "learning_rate": 2e-05,
      "loss": 0.1292,
      "step": 6508
    },
    {
      "epoch": 4.398107468739439,
      "grad_norm": 5.659432888031006,
      "learning_rate": 2e-05,
      "loss": 0.1993,
      "step": 6509
    },
    {
      "epoch": 4.398783372761068,
      "grad_norm": 1.960572361946106,
      "learning_rate": 2e-05,
      "loss": 0.1249,
      "step": 6510
    },
    {
      "epoch": 4.3994592767826965,
      "grad_norm": 4.201564311981201,
      "learning_rate": 2e-05,
      "loss": 0.1229,
      "step": 6511
    },
    {
      "epoch": 4.400135180804326,
      "grad_norm": 2.2754480838775635,
      "learning_rate": 2e-05,
      "loss": 0.1215,
      "step": 6512
    },
    {
      "epoch": 4.400811084825955,
      "grad_norm": 3.171355962753296,
      "learning_rate": 2e-05,
      "loss": 0.1864,
      "step": 6513
    },
    {
      "epoch": 4.401486988847584,
      "grad_norm": 2.64829158782959,
      "learning_rate": 2e-05,
      "loss": 0.2252,
      "step": 6514
    },
    {
      "epoch": 4.402162892869213,
      "grad_norm": 5.350667476654053,
      "learning_rate": 2e-05,
      "loss": 0.1831,
      "step": 6515
    },
    {
      "epoch": 4.402838796890841,
      "grad_norm": 10.658886909484863,
      "learning_rate": 2e-05,
      "loss": 0.2607,
      "step": 6516
    },
    {
      "epoch": 4.403514700912471,
      "grad_norm": 4.7017717361450195,
      "learning_rate": 2e-05,
      "loss": 0.1508,
      "step": 6517
    },
    {
      "epoch": 4.404190604934099,
      "grad_norm": 1.8104079961776733,
      "learning_rate": 2e-05,
      "loss": 0.1303,
      "step": 6518
    },
    {
      "epoch": 4.404866508955728,
      "grad_norm": 3.9667627811431885,
      "learning_rate": 2e-05,
      "loss": 0.1899,
      "step": 6519
    },
    {
      "epoch": 4.405542412977357,
      "grad_norm": 2.432354211807251,
      "learning_rate": 2e-05,
      "loss": 0.1871,
      "step": 6520
    },
    {
      "epoch": 4.406218316998986,
      "grad_norm": 4.19553804397583,
      "learning_rate": 2e-05,
      "loss": 0.1768,
      "step": 6521
    },
    {
      "epoch": 4.4068942210206155,
      "grad_norm": 1.9177417755126953,
      "learning_rate": 2e-05,
      "loss": 0.1772,
      "step": 6522
    },
    {
      "epoch": 4.407570125042244,
      "grad_norm": 2.8811073303222656,
      "learning_rate": 2e-05,
      "loss": 0.1371,
      "step": 6523
    },
    {
      "epoch": 4.408246029063873,
      "grad_norm": 2.3005833625793457,
      "learning_rate": 2e-05,
      "loss": 0.1643,
      "step": 6524
    },
    {
      "epoch": 4.408921933085502,
      "grad_norm": 5.198184490203857,
      "learning_rate": 2e-05,
      "loss": 0.1689,
      "step": 6525
    },
    {
      "epoch": 4.409597837107131,
      "grad_norm": 2.4861319065093994,
      "learning_rate": 2e-05,
      "loss": 0.188,
      "step": 6526
    },
    {
      "epoch": 4.410273741128759,
      "grad_norm": 2.9760701656341553,
      "learning_rate": 2e-05,
      "loss": 0.2307,
      "step": 6527
    },
    {
      "epoch": 4.410949645150389,
      "grad_norm": 4.210541248321533,
      "learning_rate": 2e-05,
      "loss": 0.1257,
      "step": 6528
    },
    {
      "epoch": 4.411625549172017,
      "grad_norm": 5.37472677230835,
      "learning_rate": 2e-05,
      "loss": 0.2329,
      "step": 6529
    },
    {
      "epoch": 4.412301453193646,
      "grad_norm": 1.8868653774261475,
      "learning_rate": 2e-05,
      "loss": 0.1269,
      "step": 6530
    },
    {
      "epoch": 4.412977357215276,
      "grad_norm": 6.625517845153809,
      "learning_rate": 2e-05,
      "loss": 0.1838,
      "step": 6531
    },
    {
      "epoch": 4.413653261236904,
      "grad_norm": 5.332554340362549,
      "learning_rate": 2e-05,
      "loss": 0.1419,
      "step": 6532
    },
    {
      "epoch": 4.4143291652585335,
      "grad_norm": 2.161543369293213,
      "learning_rate": 2e-05,
      "loss": 0.1767,
      "step": 6533
    },
    {
      "epoch": 4.415005069280162,
      "grad_norm": 2.108013391494751,
      "learning_rate": 2e-05,
      "loss": 0.1449,
      "step": 6534
    },
    {
      "epoch": 4.415680973301791,
      "grad_norm": 2.5528557300567627,
      "learning_rate": 2e-05,
      "loss": 0.1667,
      "step": 6535
    },
    {
      "epoch": 4.41635687732342,
      "grad_norm": 3.525322198867798,
      "learning_rate": 2e-05,
      "loss": 0.1186,
      "step": 6536
    },
    {
      "epoch": 4.417032781345049,
      "grad_norm": 4.603446960449219,
      "learning_rate": 2e-05,
      "loss": 0.1343,
      "step": 6537
    },
    {
      "epoch": 4.417708685366678,
      "grad_norm": 3.384279727935791,
      "learning_rate": 2e-05,
      "loss": 0.1308,
      "step": 6538
    },
    {
      "epoch": 4.418384589388307,
      "grad_norm": 3.0623810291290283,
      "learning_rate": 2e-05,
      "loss": 0.178,
      "step": 6539
    },
    {
      "epoch": 4.419060493409936,
      "grad_norm": 4.242415904998779,
      "learning_rate": 2e-05,
      "loss": 0.251,
      "step": 6540
    },
    {
      "epoch": 4.419736397431564,
      "grad_norm": 4.2331132888793945,
      "learning_rate": 2e-05,
      "loss": 0.2053,
      "step": 6541
    },
    {
      "epoch": 4.420412301453194,
      "grad_norm": 2.481804370880127,
      "learning_rate": 2e-05,
      "loss": 0.124,
      "step": 6542
    },
    {
      "epoch": 4.421088205474822,
      "grad_norm": 3.3198235034942627,
      "learning_rate": 2e-05,
      "loss": 0.1737,
      "step": 6543
    },
    {
      "epoch": 4.4217641094964515,
      "grad_norm": 3.9505293369293213,
      "learning_rate": 2e-05,
      "loss": 0.1517,
      "step": 6544
    },
    {
      "epoch": 4.422440013518081,
      "grad_norm": 9.030109405517578,
      "learning_rate": 2e-05,
      "loss": 0.2027,
      "step": 6545
    },
    {
      "epoch": 4.423115917539709,
      "grad_norm": 6.583612442016602,
      "learning_rate": 2e-05,
      "loss": 0.1786,
      "step": 6546
    },
    {
      "epoch": 4.4237918215613385,
      "grad_norm": 5.891360759735107,
      "learning_rate": 2e-05,
      "loss": 0.1227,
      "step": 6547
    },
    {
      "epoch": 4.424467725582967,
      "grad_norm": 2.803679943084717,
      "learning_rate": 2e-05,
      "loss": 0.1731,
      "step": 6548
    },
    {
      "epoch": 4.425143629604596,
      "grad_norm": 3.4456586837768555,
      "learning_rate": 2e-05,
      "loss": 0.1851,
      "step": 6549
    },
    {
      "epoch": 4.425819533626225,
      "grad_norm": 8.632447242736816,
      "learning_rate": 2e-05,
      "loss": 0.1706,
      "step": 6550
    },
    {
      "epoch": 4.425819533626225,
      "eval_accuracy": 0.7491131756756757,
      "eval_loss": 0.24115540087223053,
      "eval_runtime": 54.9416,
      "eval_samples_per_second": 431.003,
      "eval_steps_per_second": 13.469,
      "step": 6550
    },
    {
      "epoch": 4.426495437647854,
      "grad_norm": 6.261250972747803,
      "learning_rate": 2e-05,
      "loss": 0.1759,
      "step": 6551
    },
    {
      "epoch": 4.427171341669483,
      "grad_norm": 2.039259195327759,
      "learning_rate": 2e-05,
      "loss": 0.1671,
      "step": 6552
    },
    {
      "epoch": 4.427847245691112,
      "grad_norm": 8.379960060119629,
      "learning_rate": 2e-05,
      "loss": 0.257,
      "step": 6553
    },
    {
      "epoch": 4.428523149712741,
      "grad_norm": 5.079673767089844,
      "learning_rate": 2e-05,
      "loss": 0.1763,
      "step": 6554
    },
    {
      "epoch": 4.4291990537343695,
      "grad_norm": 2.780133008956909,
      "learning_rate": 2e-05,
      "loss": 0.2177,
      "step": 6555
    },
    {
      "epoch": 4.429874957755999,
      "grad_norm": 7.0878753662109375,
      "learning_rate": 2e-05,
      "loss": 0.1829,
      "step": 6556
    },
    {
      "epoch": 4.430550861777627,
      "grad_norm": 11.638708114624023,
      "learning_rate": 2e-05,
      "loss": 0.2075,
      "step": 6557
    },
    {
      "epoch": 4.431226765799257,
      "grad_norm": 5.403242588043213,
      "learning_rate": 2e-05,
      "loss": 0.1666,
      "step": 6558
    },
    {
      "epoch": 4.431902669820886,
      "grad_norm": 5.093428611755371,
      "learning_rate": 2e-05,
      "loss": 0.1756,
      "step": 6559
    },
    {
      "epoch": 4.432578573842514,
      "grad_norm": 4.840706825256348,
      "learning_rate": 2e-05,
      "loss": 0.1689,
      "step": 6560
    },
    {
      "epoch": 4.433254477864144,
      "grad_norm": 5.082194805145264,
      "learning_rate": 2e-05,
      "loss": 0.1795,
      "step": 6561
    },
    {
      "epoch": 4.433930381885772,
      "grad_norm": 2.8099570274353027,
      "learning_rate": 2e-05,
      "loss": 0.1585,
      "step": 6562
    },
    {
      "epoch": 4.434606285907401,
      "grad_norm": 3.3945846557617188,
      "learning_rate": 2e-05,
      "loss": 0.1914,
      "step": 6563
    },
    {
      "epoch": 4.43528218992903,
      "grad_norm": 9.684135437011719,
      "learning_rate": 2e-05,
      "loss": 0.1778,
      "step": 6564
    },
    {
      "epoch": 4.435958093950659,
      "grad_norm": 7.801479339599609,
      "learning_rate": 2e-05,
      "loss": 0.1993,
      "step": 6565
    },
    {
      "epoch": 4.4366339979722875,
      "grad_norm": 6.040948867797852,
      "learning_rate": 2e-05,
      "loss": 0.2146,
      "step": 6566
    },
    {
      "epoch": 4.437309901993917,
      "grad_norm": 1.7525156736373901,
      "learning_rate": 2e-05,
      "loss": 0.1507,
      "step": 6567
    },
    {
      "epoch": 4.437985806015546,
      "grad_norm": 3.3140869140625,
      "learning_rate": 2e-05,
      "loss": 0.131,
      "step": 6568
    },
    {
      "epoch": 4.438661710037175,
      "grad_norm": 4.954311370849609,
      "learning_rate": 2e-05,
      "loss": 0.1832,
      "step": 6569
    },
    {
      "epoch": 4.439337614058804,
      "grad_norm": 4.780386447906494,
      "learning_rate": 2e-05,
      "loss": 0.12,
      "step": 6570
    },
    {
      "epoch": 4.440013518080432,
      "grad_norm": 6.996568202972412,
      "learning_rate": 2e-05,
      "loss": 0.1511,
      "step": 6571
    },
    {
      "epoch": 4.440689422102062,
      "grad_norm": 3.0926568508148193,
      "learning_rate": 2e-05,
      "loss": 0.1741,
      "step": 6572
    },
    {
      "epoch": 4.44136532612369,
      "grad_norm": 2.505613327026367,
      "learning_rate": 2e-05,
      "loss": 0.1636,
      "step": 6573
    },
    {
      "epoch": 4.442041230145319,
      "grad_norm": 2.338900089263916,
      "learning_rate": 2e-05,
      "loss": 0.1203,
      "step": 6574
    },
    {
      "epoch": 4.442717134166949,
      "grad_norm": 3.5978877544403076,
      "learning_rate": 2e-05,
      "loss": 0.1408,
      "step": 6575
    },
    {
      "epoch": 4.443393038188577,
      "grad_norm": 4.78387975692749,
      "learning_rate": 2e-05,
      "loss": 0.1612,
      "step": 6576
    },
    {
      "epoch": 4.444068942210206,
      "grad_norm": 4.747165679931641,
      "learning_rate": 2e-05,
      "loss": 0.1714,
      "step": 6577
    },
    {
      "epoch": 4.444744846231835,
      "grad_norm": 3.73539137840271,
      "learning_rate": 2e-05,
      "loss": 0.1121,
      "step": 6578
    },
    {
      "epoch": 4.445420750253464,
      "grad_norm": 4.520723342895508,
      "learning_rate": 2e-05,
      "loss": 0.1149,
      "step": 6579
    },
    {
      "epoch": 4.446096654275093,
      "grad_norm": 4.797768592834473,
      "learning_rate": 2e-05,
      "loss": 0.1455,
      "step": 6580
    },
    {
      "epoch": 4.446772558296722,
      "grad_norm": 2.2980620861053467,
      "learning_rate": 2e-05,
      "loss": 0.1149,
      "step": 6581
    },
    {
      "epoch": 4.44744846231835,
      "grad_norm": 2.8378052711486816,
      "learning_rate": 2e-05,
      "loss": 0.1751,
      "step": 6582
    },
    {
      "epoch": 4.44812436633998,
      "grad_norm": 5.394678115844727,
      "learning_rate": 2e-05,
      "loss": 0.112,
      "step": 6583
    },
    {
      "epoch": 4.448800270361609,
      "grad_norm": 2.7513368129730225,
      "learning_rate": 2e-05,
      "loss": 0.1018,
      "step": 6584
    },
    {
      "epoch": 4.449476174383237,
      "grad_norm": 3.3705408573150635,
      "learning_rate": 2e-05,
      "loss": 0.115,
      "step": 6585
    },
    {
      "epoch": 4.450152078404867,
      "grad_norm": 3.6378047466278076,
      "learning_rate": 2e-05,
      "loss": 0.1141,
      "step": 6586
    },
    {
      "epoch": 4.450827982426495,
      "grad_norm": 4.576470375061035,
      "learning_rate": 2e-05,
      "loss": 0.1436,
      "step": 6587
    },
    {
      "epoch": 4.4515038864481244,
      "grad_norm": 4.20199728012085,
      "learning_rate": 2e-05,
      "loss": 0.1533,
      "step": 6588
    },
    {
      "epoch": 4.452179790469753,
      "grad_norm": 5.299299240112305,
      "learning_rate": 2e-05,
      "loss": 0.1531,
      "step": 6589
    },
    {
      "epoch": 4.452855694491382,
      "grad_norm": 2.8410186767578125,
      "learning_rate": 2e-05,
      "loss": 0.1613,
      "step": 6590
    },
    {
      "epoch": 4.4535315985130115,
      "grad_norm": 4.2266974449157715,
      "learning_rate": 2e-05,
      "loss": 0.2177,
      "step": 6591
    },
    {
      "epoch": 4.45420750253464,
      "grad_norm": 2.22892165184021,
      "learning_rate": 2e-05,
      "loss": 0.1269,
      "step": 6592
    },
    {
      "epoch": 4.454883406556269,
      "grad_norm": 4.987698554992676,
      "learning_rate": 2e-05,
      "loss": 0.1807,
      "step": 6593
    },
    {
      "epoch": 4.455559310577898,
      "grad_norm": 4.889943599700928,
      "learning_rate": 2e-05,
      "loss": 0.1591,
      "step": 6594
    },
    {
      "epoch": 4.456235214599527,
      "grad_norm": 3.3884902000427246,
      "learning_rate": 2e-05,
      "loss": 0.1549,
      "step": 6595
    },
    {
      "epoch": 4.456911118621155,
      "grad_norm": 3.922400951385498,
      "learning_rate": 2e-05,
      "loss": 0.1693,
      "step": 6596
    },
    {
      "epoch": 4.457587022642785,
      "grad_norm": 11.035252571105957,
      "learning_rate": 2e-05,
      "loss": 0.2139,
      "step": 6597
    },
    {
      "epoch": 4.458262926664414,
      "grad_norm": 6.683803558349609,
      "learning_rate": 2e-05,
      "loss": 0.1787,
      "step": 6598
    },
    {
      "epoch": 4.4589388306860425,
      "grad_norm": 4.084479808807373,
      "learning_rate": 2e-05,
      "loss": 0.1828,
      "step": 6599
    },
    {
      "epoch": 4.459614734707672,
      "grad_norm": 4.664865970611572,
      "learning_rate": 2e-05,
      "loss": 0.1491,
      "step": 6600
    },
    {
      "epoch": 4.459614734707672,
      "eval_accuracy": 0.7232263513513514,
      "eval_loss": 0.24183592200279236,
      "eval_runtime": 66.8851,
      "eval_samples_per_second": 354.04,
      "eval_steps_per_second": 11.064,
      "step": 6600
    },
    {
      "epoch": 4.4602906387293,
      "grad_norm": 4.161423206329346,
      "learning_rate": 2e-05,
      "loss": 0.1805,
      "step": 6601
    },
    {
      "epoch": 4.4609665427509295,
      "grad_norm": 2.6403167247772217,
      "learning_rate": 2e-05,
      "loss": 0.1681,
      "step": 6602
    },
    {
      "epoch": 4.461642446772558,
      "grad_norm": 3.29962158203125,
      "learning_rate": 2e-05,
      "loss": 0.1883,
      "step": 6603
    },
    {
      "epoch": 4.462318350794187,
      "grad_norm": 2.7678418159484863,
      "learning_rate": 2e-05,
      "loss": 0.181,
      "step": 6604
    },
    {
      "epoch": 4.462994254815817,
      "grad_norm": 1.5706722736358643,
      "learning_rate": 2e-05,
      "loss": 0.1832,
      "step": 6605
    },
    {
      "epoch": 4.463670158837445,
      "grad_norm": 2.216583013534546,
      "learning_rate": 2e-05,
      "loss": 0.1764,
      "step": 6606
    },
    {
      "epoch": 4.464346062859074,
      "grad_norm": 5.2252278327941895,
      "learning_rate": 2e-05,
      "loss": 0.2043,
      "step": 6607
    },
    {
      "epoch": 4.465021966880703,
      "grad_norm": 9.224177360534668,
      "learning_rate": 2e-05,
      "loss": 0.2227,
      "step": 6608
    },
    {
      "epoch": 4.465697870902332,
      "grad_norm": 3.630967617034912,
      "learning_rate": 2e-05,
      "loss": 0.1785,
      "step": 6609
    },
    {
      "epoch": 4.4663737749239605,
      "grad_norm": 2.2354071140289307,
      "learning_rate": 2e-05,
      "loss": 0.1362,
      "step": 6610
    },
    {
      "epoch": 4.46704967894559,
      "grad_norm": 2.004608154296875,
      "learning_rate": 2e-05,
      "loss": 0.1683,
      "step": 6611
    },
    {
      "epoch": 4.467725582967219,
      "grad_norm": 3.381943464279175,
      "learning_rate": 2e-05,
      "loss": 0.1458,
      "step": 6612
    },
    {
      "epoch": 4.4684014869888475,
      "grad_norm": 4.385499477386475,
      "learning_rate": 2e-05,
      "loss": 0.1554,
      "step": 6613
    },
    {
      "epoch": 4.469077391010477,
      "grad_norm": 8.344043731689453,
      "learning_rate": 2e-05,
      "loss": 0.2086,
      "step": 6614
    },
    {
      "epoch": 4.469753295032105,
      "grad_norm": 6.022549629211426,
      "learning_rate": 2e-05,
      "loss": 0.1245,
      "step": 6615
    },
    {
      "epoch": 4.470429199053735,
      "grad_norm": 3.128512144088745,
      "learning_rate": 2e-05,
      "loss": 0.1705,
      "step": 6616
    },
    {
      "epoch": 4.471105103075363,
      "grad_norm": 3.3224034309387207,
      "learning_rate": 2e-05,
      "loss": 0.1562,
      "step": 6617
    },
    {
      "epoch": 4.471781007096992,
      "grad_norm": 2.2695541381835938,
      "learning_rate": 2e-05,
      "loss": 0.1904,
      "step": 6618
    },
    {
      "epoch": 4.472456911118621,
      "grad_norm": 6.64061164855957,
      "learning_rate": 2e-05,
      "loss": 0.2055,
      "step": 6619
    },
    {
      "epoch": 4.47313281514025,
      "grad_norm": 6.9872517585754395,
      "learning_rate": 2e-05,
      "loss": 0.1683,
      "step": 6620
    },
    {
      "epoch": 4.473808719161879,
      "grad_norm": 4.889660358428955,
      "learning_rate": 2e-05,
      "loss": 0.1277,
      "step": 6621
    },
    {
      "epoch": 4.474484623183508,
      "grad_norm": 4.156866073608398,
      "learning_rate": 2e-05,
      "loss": 0.1801,
      "step": 6622
    },
    {
      "epoch": 4.475160527205137,
      "grad_norm": 1.9225172996520996,
      "learning_rate": 2e-05,
      "loss": 0.1185,
      "step": 6623
    },
    {
      "epoch": 4.4758364312267656,
      "grad_norm": 3.604471445083618,
      "learning_rate": 2e-05,
      "loss": 0.1886,
      "step": 6624
    },
    {
      "epoch": 4.476512335248395,
      "grad_norm": 4.878686904907227,
      "learning_rate": 2e-05,
      "loss": 0.1741,
      "step": 6625
    },
    {
      "epoch": 4.477188239270023,
      "grad_norm": 3.428546190261841,
      "learning_rate": 2e-05,
      "loss": 0.1295,
      "step": 6626
    },
    {
      "epoch": 4.477864143291653,
      "grad_norm": 6.853039741516113,
      "learning_rate": 2e-05,
      "loss": 0.2241,
      "step": 6627
    },
    {
      "epoch": 4.478540047313282,
      "grad_norm": 2.820347309112549,
      "learning_rate": 2e-05,
      "loss": 0.1406,
      "step": 6628
    },
    {
      "epoch": 4.47921595133491,
      "grad_norm": 2.977778673171997,
      "learning_rate": 2e-05,
      "loss": 0.1553,
      "step": 6629
    },
    {
      "epoch": 4.47989185535654,
      "grad_norm": 7.551669597625732,
      "learning_rate": 2e-05,
      "loss": 0.2114,
      "step": 6630
    },
    {
      "epoch": 4.480567759378168,
      "grad_norm": 3.570266008377075,
      "learning_rate": 2e-05,
      "loss": 0.1348,
      "step": 6631
    },
    {
      "epoch": 4.481243663399797,
      "grad_norm": 2.413139581680298,
      "learning_rate": 2e-05,
      "loss": 0.1587,
      "step": 6632
    },
    {
      "epoch": 4.481919567421426,
      "grad_norm": 1.8741849660873413,
      "learning_rate": 2e-05,
      "loss": 0.1448,
      "step": 6633
    },
    {
      "epoch": 4.482595471443055,
      "grad_norm": 3.2628116607666016,
      "learning_rate": 2e-05,
      "loss": 0.2024,
      "step": 6634
    },
    {
      "epoch": 4.483271375464684,
      "grad_norm": 2.21744704246521,
      "learning_rate": 2e-05,
      "loss": 0.1087,
      "step": 6635
    },
    {
      "epoch": 4.483947279486313,
      "grad_norm": 6.171612739562988,
      "learning_rate": 2e-05,
      "loss": 0.2624,
      "step": 6636
    },
    {
      "epoch": 4.484623183507942,
      "grad_norm": 3.5415377616882324,
      "learning_rate": 2e-05,
      "loss": 0.2141,
      "step": 6637
    },
    {
      "epoch": 4.485299087529571,
      "grad_norm": 8.404831886291504,
      "learning_rate": 2e-05,
      "loss": 0.1544,
      "step": 6638
    },
    {
      "epoch": 4.4859749915512,
      "grad_norm": 7.055545806884766,
      "learning_rate": 2e-05,
      "loss": 0.1436,
      "step": 6639
    },
    {
      "epoch": 4.486650895572828,
      "grad_norm": 2.849137544631958,
      "learning_rate": 2e-05,
      "loss": 0.1785,
      "step": 6640
    },
    {
      "epoch": 4.487326799594458,
      "grad_norm": 5.08564567565918,
      "learning_rate": 2e-05,
      "loss": 0.1844,
      "step": 6641
    },
    {
      "epoch": 4.488002703616086,
      "grad_norm": 9.482903480529785,
      "learning_rate": 2e-05,
      "loss": 0.1798,
      "step": 6642
    },
    {
      "epoch": 4.488678607637715,
      "grad_norm": 2.029952049255371,
      "learning_rate": 2e-05,
      "loss": 0.1298,
      "step": 6643
    },
    {
      "epoch": 4.489354511659345,
      "grad_norm": 4.261765480041504,
      "learning_rate": 2e-05,
      "loss": 0.1823,
      "step": 6644
    },
    {
      "epoch": 4.490030415680973,
      "grad_norm": 5.9345927238464355,
      "learning_rate": 2e-05,
      "loss": 0.1558,
      "step": 6645
    },
    {
      "epoch": 4.4907063197026025,
      "grad_norm": 4.891417026519775,
      "learning_rate": 2e-05,
      "loss": 0.1941,
      "step": 6646
    },
    {
      "epoch": 4.491382223724231,
      "grad_norm": 2.7171952724456787,
      "learning_rate": 2e-05,
      "loss": 0.1032,
      "step": 6647
    },
    {
      "epoch": 4.49205812774586,
      "grad_norm": 6.695376396179199,
      "learning_rate": 2e-05,
      "loss": 0.1556,
      "step": 6648
    },
    {
      "epoch": 4.492734031767489,
      "grad_norm": 2.336717128753662,
      "learning_rate": 2e-05,
      "loss": 0.1515,
      "step": 6649
    },
    {
      "epoch": 4.493409935789118,
      "grad_norm": 2.0826871395111084,
      "learning_rate": 2e-05,
      "loss": 0.1454,
      "step": 6650
    },
    {
      "epoch": 4.493409935789118,
      "eval_accuracy": 0.725,
      "eval_loss": 0.233652263879776,
      "eval_runtime": 65.3353,
      "eval_samples_per_second": 362.438,
      "eval_steps_per_second": 11.326,
      "step": 6650
    },
    {
      "epoch": 4.494085839810747,
      "grad_norm": 2.751805305480957,
      "learning_rate": 2e-05,
      "loss": 0.137,
      "step": 6651
    },
    {
      "epoch": 4.494761743832376,
      "grad_norm": 4.212330341339111,
      "learning_rate": 2e-05,
      "loss": 0.216,
      "step": 6652
    },
    {
      "epoch": 4.495437647854005,
      "grad_norm": 5.994387149810791,
      "learning_rate": 2e-05,
      "loss": 0.1464,
      "step": 6653
    },
    {
      "epoch": 4.496113551875633,
      "grad_norm": 1.8331682682037354,
      "learning_rate": 2e-05,
      "loss": 0.1099,
      "step": 6654
    },
    {
      "epoch": 4.496789455897263,
      "grad_norm": 2.317741632461548,
      "learning_rate": 2e-05,
      "loss": 0.1171,
      "step": 6655
    },
    {
      "epoch": 4.497465359918891,
      "grad_norm": 3.501518964767456,
      "learning_rate": 2e-05,
      "loss": 0.1258,
      "step": 6656
    },
    {
      "epoch": 4.4981412639405205,
      "grad_norm": 4.192080497741699,
      "learning_rate": 2e-05,
      "loss": 0.1823,
      "step": 6657
    },
    {
      "epoch": 4.49881716796215,
      "grad_norm": 2.728412628173828,
      "learning_rate": 2e-05,
      "loss": 0.1686,
      "step": 6658
    },
    {
      "epoch": 4.499493071983778,
      "grad_norm": 5.00947380065918,
      "learning_rate": 2e-05,
      "loss": 0.1608,
      "step": 6659
    },
    {
      "epoch": 4.5001689760054076,
      "grad_norm": 6.636612892150879,
      "learning_rate": 2e-05,
      "loss": 0.1875,
      "step": 6660
    },
    {
      "epoch": 4.500844880027036,
      "grad_norm": 7.913212776184082,
      "learning_rate": 2e-05,
      "loss": 0.2233,
      "step": 6661
    },
    {
      "epoch": 4.501520784048665,
      "grad_norm": 4.506720066070557,
      "learning_rate": 2e-05,
      "loss": 0.1533,
      "step": 6662
    },
    {
      "epoch": 4.502196688070294,
      "grad_norm": 3.6490683555603027,
      "learning_rate": 2e-05,
      "loss": 0.1784,
      "step": 6663
    },
    {
      "epoch": 4.502872592091923,
      "grad_norm": 6.282339572906494,
      "learning_rate": 2e-05,
      "loss": 0.1709,
      "step": 6664
    },
    {
      "epoch": 4.503548496113552,
      "grad_norm": 3.0183348655700684,
      "learning_rate": 2e-05,
      "loss": 0.167,
      "step": 6665
    },
    {
      "epoch": 4.504224400135181,
      "grad_norm": 4.057502746582031,
      "learning_rate": 2e-05,
      "loss": 0.1645,
      "step": 6666
    },
    {
      "epoch": 4.50490030415681,
      "grad_norm": 5.431028842926025,
      "learning_rate": 2e-05,
      "loss": 0.1537,
      "step": 6667
    },
    {
      "epoch": 4.5055762081784385,
      "grad_norm": 2.173617124557495,
      "learning_rate": 2e-05,
      "loss": 0.1203,
      "step": 6668
    },
    {
      "epoch": 4.506252112200068,
      "grad_norm": 4.223822593688965,
      "learning_rate": 2e-05,
      "loss": 0.1847,
      "step": 6669
    },
    {
      "epoch": 4.506928016221696,
      "grad_norm": 14.385774612426758,
      "learning_rate": 2e-05,
      "loss": 0.2486,
      "step": 6670
    },
    {
      "epoch": 4.507603920243326,
      "grad_norm": 5.164773464202881,
      "learning_rate": 2e-05,
      "loss": 0.1653,
      "step": 6671
    },
    {
      "epoch": 4.508279824264954,
      "grad_norm": 4.057539939880371,
      "learning_rate": 2e-05,
      "loss": 0.186,
      "step": 6672
    },
    {
      "epoch": 4.508955728286583,
      "grad_norm": 3.4531898498535156,
      "learning_rate": 2e-05,
      "loss": 0.1552,
      "step": 6673
    },
    {
      "epoch": 4.509631632308212,
      "grad_norm": 2.681138515472412,
      "learning_rate": 2e-05,
      "loss": 0.133,
      "step": 6674
    },
    {
      "epoch": 4.510307536329841,
      "grad_norm": 3.447171688079834,
      "learning_rate": 2e-05,
      "loss": 0.1467,
      "step": 6675
    },
    {
      "epoch": 4.51098344035147,
      "grad_norm": 3.043381929397583,
      "learning_rate": 2e-05,
      "loss": 0.1295,
      "step": 6676
    },
    {
      "epoch": 4.511659344373099,
      "grad_norm": 7.235108375549316,
      "learning_rate": 2e-05,
      "loss": 0.1674,
      "step": 6677
    },
    {
      "epoch": 4.512335248394728,
      "grad_norm": 2.594996452331543,
      "learning_rate": 2e-05,
      "loss": 0.1524,
      "step": 6678
    },
    {
      "epoch": 4.5130111524163565,
      "grad_norm": 3.591775417327881,
      "learning_rate": 2e-05,
      "loss": 0.1414,
      "step": 6679
    },
    {
      "epoch": 4.513687056437986,
      "grad_norm": 2.5523321628570557,
      "learning_rate": 2e-05,
      "loss": 0.151,
      "step": 6680
    },
    {
      "epoch": 4.514362960459614,
      "grad_norm": 2.6306684017181396,
      "learning_rate": 2e-05,
      "loss": 0.1704,
      "step": 6681
    },
    {
      "epoch": 4.515038864481244,
      "grad_norm": 2.103947639465332,
      "learning_rate": 2e-05,
      "loss": 0.1553,
      "step": 6682
    },
    {
      "epoch": 4.515714768502873,
      "grad_norm": 4.428382873535156,
      "learning_rate": 2e-05,
      "loss": 0.1884,
      "step": 6683
    },
    {
      "epoch": 4.516390672524501,
      "grad_norm": 2.2347326278686523,
      "learning_rate": 2e-05,
      "loss": 0.1235,
      "step": 6684
    },
    {
      "epoch": 4.517066576546131,
      "grad_norm": 2.475717306137085,
      "learning_rate": 2e-05,
      "loss": 0.1441,
      "step": 6685
    },
    {
      "epoch": 4.517742480567759,
      "grad_norm": 2.126366138458252,
      "learning_rate": 2e-05,
      "loss": 0.1435,
      "step": 6686
    },
    {
      "epoch": 4.518418384589388,
      "grad_norm": 8.68636703491211,
      "learning_rate": 2e-05,
      "loss": 0.2533,
      "step": 6687
    },
    {
      "epoch": 4.519094288611017,
      "grad_norm": 7.111759662628174,
      "learning_rate": 2e-05,
      "loss": 0.1718,
      "step": 6688
    },
    {
      "epoch": 4.519770192632646,
      "grad_norm": 3.8556926250457764,
      "learning_rate": 2e-05,
      "loss": 0.1714,
      "step": 6689
    },
    {
      "epoch": 4.520446096654275,
      "grad_norm": 5.32113790512085,
      "learning_rate": 2e-05,
      "loss": 0.1763,
      "step": 6690
    },
    {
      "epoch": 4.521122000675904,
      "grad_norm": 2.046074628829956,
      "learning_rate": 2e-05,
      "loss": 0.1553,
      "step": 6691
    },
    {
      "epoch": 4.521797904697533,
      "grad_norm": 9.824214935302734,
      "learning_rate": 2e-05,
      "loss": 0.1685,
      "step": 6692
    },
    {
      "epoch": 4.522473808719162,
      "grad_norm": 4.876579761505127,
      "learning_rate": 2e-05,
      "loss": 0.2054,
      "step": 6693
    },
    {
      "epoch": 4.523149712740791,
      "grad_norm": 5.066339015960693,
      "learning_rate": 2e-05,
      "loss": 0.2213,
      "step": 6694
    },
    {
      "epoch": 4.523825616762419,
      "grad_norm": 5.757728576660156,
      "learning_rate": 2e-05,
      "loss": 0.1635,
      "step": 6695
    },
    {
      "epoch": 4.524501520784049,
      "grad_norm": 2.1838974952697754,
      "learning_rate": 2e-05,
      "loss": 0.2094,
      "step": 6696
    },
    {
      "epoch": 4.525177424805678,
      "grad_norm": 5.10828161239624,
      "learning_rate": 2e-05,
      "loss": 0.1614,
      "step": 6697
    },
    {
      "epoch": 4.525853328827306,
      "grad_norm": 1.9002255201339722,
      "learning_rate": 2e-05,
      "loss": 0.1867,
      "step": 6698
    },
    {
      "epoch": 4.526529232848936,
      "grad_norm": 6.079162120819092,
      "learning_rate": 2e-05,
      "loss": 0.2358,
      "step": 6699
    },
    {
      "epoch": 4.527205136870564,
      "grad_norm": 2.9183928966522217,
      "learning_rate": 2e-05,
      "loss": 0.1469,
      "step": 6700
    },
    {
      "epoch": 4.527205136870564,
      "eval_accuracy": 0.7214949324324325,
      "eval_loss": 0.23264990746974945,
      "eval_runtime": 68.8532,
      "eval_samples_per_second": 343.92,
      "eval_steps_per_second": 10.748,
      "step": 6700
    },
    {
      "epoch": 4.5278810408921935,
      "grad_norm": 1.935644268989563,
      "learning_rate": 2e-05,
      "loss": 0.1718,
      "step": 6701
    },
    {
      "epoch": 4.528556944913822,
      "grad_norm": 2.3913490772247314,
      "learning_rate": 2e-05,
      "loss": 0.1523,
      "step": 6702
    },
    {
      "epoch": 4.529232848935451,
      "grad_norm": 2.427157402038574,
      "learning_rate": 2e-05,
      "loss": 0.204,
      "step": 6703
    },
    {
      "epoch": 4.5299087529570805,
      "grad_norm": 3.8263323307037354,
      "learning_rate": 2e-05,
      "loss": 0.1266,
      "step": 6704
    },
    {
      "epoch": 4.530584656978709,
      "grad_norm": 1.6846672296524048,
      "learning_rate": 2e-05,
      "loss": 0.1574,
      "step": 6705
    },
    {
      "epoch": 4.531260561000338,
      "grad_norm": 1.9746904373168945,
      "learning_rate": 2e-05,
      "loss": 0.1432,
      "step": 6706
    },
    {
      "epoch": 4.531936465021967,
      "grad_norm": 2.266838550567627,
      "learning_rate": 2e-05,
      "loss": 0.1612,
      "step": 6707
    },
    {
      "epoch": 4.532612369043596,
      "grad_norm": 6.969661235809326,
      "learning_rate": 2e-05,
      "loss": 0.1536,
      "step": 6708
    },
    {
      "epoch": 4.533288273065224,
      "grad_norm": 4.9895148277282715,
      "learning_rate": 2e-05,
      "loss": 0.1372,
      "step": 6709
    },
    {
      "epoch": 4.533964177086854,
      "grad_norm": 6.9073262214660645,
      "learning_rate": 2e-05,
      "loss": 0.1937,
      "step": 6710
    },
    {
      "epoch": 4.534640081108483,
      "grad_norm": 7.3112969398498535,
      "learning_rate": 2e-05,
      "loss": 0.1763,
      "step": 6711
    },
    {
      "epoch": 4.5353159851301115,
      "grad_norm": 3.3715338706970215,
      "learning_rate": 2e-05,
      "loss": 0.1883,
      "step": 6712
    },
    {
      "epoch": 4.535991889151741,
      "grad_norm": 3.761998176574707,
      "learning_rate": 2e-05,
      "loss": 0.2322,
      "step": 6713
    },
    {
      "epoch": 4.536667793173369,
      "grad_norm": 5.8430070877075195,
      "learning_rate": 2e-05,
      "loss": 0.1847,
      "step": 6714
    },
    {
      "epoch": 4.5373436971949985,
      "grad_norm": 2.9412841796875,
      "learning_rate": 2e-05,
      "loss": 0.1748,
      "step": 6715
    },
    {
      "epoch": 4.538019601216627,
      "grad_norm": 3.393683433532715,
      "learning_rate": 2e-05,
      "loss": 0.2622,
      "step": 6716
    },
    {
      "epoch": 4.538695505238256,
      "grad_norm": 6.652483940124512,
      "learning_rate": 2e-05,
      "loss": 0.1633,
      "step": 6717
    },
    {
      "epoch": 4.539371409259886,
      "grad_norm": 7.600559234619141,
      "learning_rate": 2e-05,
      "loss": 0.1533,
      "step": 6718
    },
    {
      "epoch": 4.540047313281514,
      "grad_norm": 2.073091983795166,
      "learning_rate": 2e-05,
      "loss": 0.1938,
      "step": 6719
    },
    {
      "epoch": 4.540723217303143,
      "grad_norm": 2.2175443172454834,
      "learning_rate": 2e-05,
      "loss": 0.1052,
      "step": 6720
    },
    {
      "epoch": 4.541399121324772,
      "grad_norm": 4.408407688140869,
      "learning_rate": 2e-05,
      "loss": 0.1622,
      "step": 6721
    },
    {
      "epoch": 4.542075025346401,
      "grad_norm": 3.755817413330078,
      "learning_rate": 2e-05,
      "loss": 0.2049,
      "step": 6722
    },
    {
      "epoch": 4.5427509293680295,
      "grad_norm": 5.5011091232299805,
      "learning_rate": 2e-05,
      "loss": 0.1606,
      "step": 6723
    },
    {
      "epoch": 4.543426833389659,
      "grad_norm": 2.8339216709136963,
      "learning_rate": 2e-05,
      "loss": 0.1688,
      "step": 6724
    },
    {
      "epoch": 4.544102737411287,
      "grad_norm": 4.473662853240967,
      "learning_rate": 2e-05,
      "loss": 0.1507,
      "step": 6725
    },
    {
      "epoch": 4.5447786414329165,
      "grad_norm": 7.369375228881836,
      "learning_rate": 2e-05,
      "loss": 0.2461,
      "step": 6726
    },
    {
      "epoch": 4.545454545454545,
      "grad_norm": 2.782167434692383,
      "learning_rate": 2e-05,
      "loss": 0.2495,
      "step": 6727
    },
    {
      "epoch": 4.546130449476174,
      "grad_norm": 3.1586127281188965,
      "learning_rate": 2e-05,
      "loss": 0.1512,
      "step": 6728
    },
    {
      "epoch": 4.546806353497804,
      "grad_norm": 3.6024351119995117,
      "learning_rate": 2e-05,
      "loss": 0.15,
      "step": 6729
    },
    {
      "epoch": 4.547482257519432,
      "grad_norm": 5.736011028289795,
      "learning_rate": 2e-05,
      "loss": 0.2197,
      "step": 6730
    },
    {
      "epoch": 4.548158161541061,
      "grad_norm": 7.434205532073975,
      "learning_rate": 2e-05,
      "loss": 0.142,
      "step": 6731
    },
    {
      "epoch": 4.54883406556269,
      "grad_norm": 1.6591075658798218,
      "learning_rate": 2e-05,
      "loss": 0.169,
      "step": 6732
    },
    {
      "epoch": 4.549509969584319,
      "grad_norm": 2.8473074436187744,
      "learning_rate": 2e-05,
      "loss": 0.1349,
      "step": 6733
    },
    {
      "epoch": 4.5501858736059475,
      "grad_norm": 5.412266254425049,
      "learning_rate": 2e-05,
      "loss": 0.1676,
      "step": 6734
    },
    {
      "epoch": 4.550861777627577,
      "grad_norm": 2.638619899749756,
      "learning_rate": 2e-05,
      "loss": 0.162,
      "step": 6735
    },
    {
      "epoch": 4.551537681649206,
      "grad_norm": 1.7260265350341797,
      "learning_rate": 2e-05,
      "loss": 0.1286,
      "step": 6736
    },
    {
      "epoch": 4.552213585670835,
      "grad_norm": 2.196676015853882,
      "learning_rate": 2e-05,
      "loss": 0.1113,
      "step": 6737
    },
    {
      "epoch": 4.552889489692464,
      "grad_norm": 8.31379222869873,
      "learning_rate": 2e-05,
      "loss": 0.1752,
      "step": 6738
    },
    {
      "epoch": 4.553565393714092,
      "grad_norm": 4.309295177459717,
      "learning_rate": 2e-05,
      "loss": 0.2107,
      "step": 6739
    },
    {
      "epoch": 4.554241297735722,
      "grad_norm": 2.380683422088623,
      "learning_rate": 2e-05,
      "loss": 0.1238,
      "step": 6740
    },
    {
      "epoch": 4.55491720175735,
      "grad_norm": 5.653498649597168,
      "learning_rate": 2e-05,
      "loss": 0.1755,
      "step": 6741
    },
    {
      "epoch": 4.555593105778979,
      "grad_norm": 1.8827919960021973,
      "learning_rate": 2e-05,
      "loss": 0.1363,
      "step": 6742
    },
    {
      "epoch": 4.556269009800609,
      "grad_norm": 2.699348211288452,
      "learning_rate": 2e-05,
      "loss": 0.1754,
      "step": 6743
    },
    {
      "epoch": 4.556944913822237,
      "grad_norm": 5.367449760437012,
      "learning_rate": 2e-05,
      "loss": 0.2078,
      "step": 6744
    },
    {
      "epoch": 4.557620817843866,
      "grad_norm": 5.448431968688965,
      "learning_rate": 2e-05,
      "loss": 0.2732,
      "step": 6745
    },
    {
      "epoch": 4.558296721865495,
      "grad_norm": 3.0454459190368652,
      "learning_rate": 2e-05,
      "loss": 0.1406,
      "step": 6746
    },
    {
      "epoch": 4.558972625887124,
      "grad_norm": 3.483022451400757,
      "learning_rate": 2e-05,
      "loss": 0.1268,
      "step": 6747
    },
    {
      "epoch": 4.559648529908753,
      "grad_norm": 3.7419612407684326,
      "learning_rate": 2e-05,
      "loss": 0.1712,
      "step": 6748
    },
    {
      "epoch": 4.560324433930382,
      "grad_norm": 2.180119752883911,
      "learning_rate": 2e-05,
      "loss": 0.1458,
      "step": 6749
    },
    {
      "epoch": 4.561000337952011,
      "grad_norm": 3.3825271129608154,
      "learning_rate": 2e-05,
      "loss": 0.1461,
      "step": 6750
    },
    {
      "epoch": 4.561000337952011,
      "eval_accuracy": 0.7364020270270271,
      "eval_loss": 0.22693021595478058,
      "eval_runtime": 87.3485,
      "eval_samples_per_second": 271.098,
      "eval_steps_per_second": 8.472,
      "step": 6750
    },
    {
      "epoch": 4.56167624197364,
      "grad_norm": 3.0033528804779053,
      "learning_rate": 2e-05,
      "loss": 0.1342,
      "step": 6751
    },
    {
      "epoch": 4.562352145995269,
      "grad_norm": 2.362168788909912,
      "learning_rate": 2e-05,
      "loss": 0.1693,
      "step": 6752
    },
    {
      "epoch": 4.563028050016897,
      "grad_norm": 2.7632031440734863,
      "learning_rate": 2e-05,
      "loss": 0.142,
      "step": 6753
    },
    {
      "epoch": 4.563703954038527,
      "grad_norm": 2.1999659538269043,
      "learning_rate": 2e-05,
      "loss": 0.1591,
      "step": 6754
    },
    {
      "epoch": 4.564379858060155,
      "grad_norm": 5.77324104309082,
      "learning_rate": 2e-05,
      "loss": 0.156,
      "step": 6755
    },
    {
      "epoch": 4.565055762081784,
      "grad_norm": 4.025196075439453,
      "learning_rate": 2e-05,
      "loss": 0.1778,
      "step": 6756
    },
    {
      "epoch": 4.565731666103414,
      "grad_norm": 8.791045188903809,
      "learning_rate": 2e-05,
      "loss": 0.2351,
      "step": 6757
    },
    {
      "epoch": 4.566407570125042,
      "grad_norm": 5.977962970733643,
      "learning_rate": 2e-05,
      "loss": 0.1229,
      "step": 6758
    },
    {
      "epoch": 4.5670834741466715,
      "grad_norm": 1.9078601598739624,
      "learning_rate": 2e-05,
      "loss": 0.1189,
      "step": 6759
    },
    {
      "epoch": 4.5677593781683,
      "grad_norm": 4.20589542388916,
      "learning_rate": 2e-05,
      "loss": 0.1744,
      "step": 6760
    },
    {
      "epoch": 4.568435282189929,
      "grad_norm": 4.3977813720703125,
      "learning_rate": 2e-05,
      "loss": 0.1137,
      "step": 6761
    },
    {
      "epoch": 4.569111186211558,
      "grad_norm": 2.636345386505127,
      "learning_rate": 2e-05,
      "loss": 0.2033,
      "step": 6762
    },
    {
      "epoch": 4.569787090233187,
      "grad_norm": 6.03747034072876,
      "learning_rate": 2e-05,
      "loss": 0.2231,
      "step": 6763
    },
    {
      "epoch": 4.570462994254816,
      "grad_norm": 7.227992534637451,
      "learning_rate": 2e-05,
      "loss": 0.2173,
      "step": 6764
    },
    {
      "epoch": 4.571138898276445,
      "grad_norm": 5.664833068847656,
      "learning_rate": 2e-05,
      "loss": 0.135,
      "step": 6765
    },
    {
      "epoch": 4.571814802298074,
      "grad_norm": 2.7618050575256348,
      "learning_rate": 2e-05,
      "loss": 0.1075,
      "step": 6766
    },
    {
      "epoch": 4.5724907063197024,
      "grad_norm": 1.9812854528427124,
      "learning_rate": 2e-05,
      "loss": 0.0965,
      "step": 6767
    },
    {
      "epoch": 4.573166610341332,
      "grad_norm": 3.0803298950195312,
      "learning_rate": 2e-05,
      "loss": 0.2321,
      "step": 6768
    },
    {
      "epoch": 4.57384251436296,
      "grad_norm": 8.471466064453125,
      "learning_rate": 2e-05,
      "loss": 0.2089,
      "step": 6769
    },
    {
      "epoch": 4.5745184183845895,
      "grad_norm": 2.647693634033203,
      "learning_rate": 2e-05,
      "loss": 0.1582,
      "step": 6770
    },
    {
      "epoch": 4.575194322406219,
      "grad_norm": 3.545879364013672,
      "learning_rate": 2e-05,
      "loss": 0.1266,
      "step": 6771
    },
    {
      "epoch": 4.575870226427847,
      "grad_norm": 2.179187536239624,
      "learning_rate": 2e-05,
      "loss": 0.1183,
      "step": 6772
    },
    {
      "epoch": 4.576546130449477,
      "grad_norm": 3.309091091156006,
      "learning_rate": 2e-05,
      "loss": 0.1933,
      "step": 6773
    },
    {
      "epoch": 4.577222034471105,
      "grad_norm": 1.7909111976623535,
      "learning_rate": 2e-05,
      "loss": 0.1428,
      "step": 6774
    },
    {
      "epoch": 4.577897938492734,
      "grad_norm": 6.610849380493164,
      "learning_rate": 2e-05,
      "loss": 0.1478,
      "step": 6775
    },
    {
      "epoch": 4.578573842514363,
      "grad_norm": 3.7724392414093018,
      "learning_rate": 2e-05,
      "loss": 0.1596,
      "step": 6776
    },
    {
      "epoch": 4.579249746535992,
      "grad_norm": 3.938400983810425,
      "learning_rate": 2e-05,
      "loss": 0.1315,
      "step": 6777
    },
    {
      "epoch": 4.5799256505576205,
      "grad_norm": 2.772965431213379,
      "learning_rate": 2e-05,
      "loss": 0.1557,
      "step": 6778
    },
    {
      "epoch": 4.58060155457925,
      "grad_norm": 2.676136016845703,
      "learning_rate": 2e-05,
      "loss": 0.1351,
      "step": 6779
    },
    {
      "epoch": 4.581277458600878,
      "grad_norm": 2.3197803497314453,
      "learning_rate": 2e-05,
      "loss": 0.1512,
      "step": 6780
    },
    {
      "epoch": 4.5819533626225075,
      "grad_norm": 2.7629292011260986,
      "learning_rate": 2e-05,
      "loss": 0.1208,
      "step": 6781
    },
    {
      "epoch": 4.582629266644137,
      "grad_norm": 3.3199167251586914,
      "learning_rate": 2e-05,
      "loss": 0.1448,
      "step": 6782
    },
    {
      "epoch": 4.583305170665765,
      "grad_norm": 2.8131656646728516,
      "learning_rate": 2e-05,
      "loss": 0.1196,
      "step": 6783
    },
    {
      "epoch": 4.583981074687395,
      "grad_norm": 4.274287700653076,
      "learning_rate": 2e-05,
      "loss": 0.1912,
      "step": 6784
    },
    {
      "epoch": 4.584656978709023,
      "grad_norm": 9.58735179901123,
      "learning_rate": 2e-05,
      "loss": 0.177,
      "step": 6785
    },
    {
      "epoch": 4.585332882730652,
      "grad_norm": 2.980490207672119,
      "learning_rate": 2e-05,
      "loss": 0.1729,
      "step": 6786
    },
    {
      "epoch": 4.586008786752281,
      "grad_norm": 3.119788885116577,
      "learning_rate": 2e-05,
      "loss": 0.1697,
      "step": 6787
    },
    {
      "epoch": 4.58668469077391,
      "grad_norm": 6.150162220001221,
      "learning_rate": 2e-05,
      "loss": 0.2339,
      "step": 6788
    },
    {
      "epoch": 4.587360594795539,
      "grad_norm": 5.211350917816162,
      "learning_rate": 2e-05,
      "loss": 0.1347,
      "step": 6789
    },
    {
      "epoch": 4.588036498817168,
      "grad_norm": 6.690974235534668,
      "learning_rate": 2e-05,
      "loss": 0.2355,
      "step": 6790
    },
    {
      "epoch": 4.588712402838797,
      "grad_norm": 5.3064141273498535,
      "learning_rate": 2e-05,
      "loss": 0.1639,
      "step": 6791
    },
    {
      "epoch": 4.5893883068604255,
      "grad_norm": 3.2774770259857178,
      "learning_rate": 2e-05,
      "loss": 0.221,
      "step": 6792
    },
    {
      "epoch": 4.590064210882055,
      "grad_norm": 2.4357821941375732,
      "learning_rate": 2e-05,
      "loss": 0.1182,
      "step": 6793
    },
    {
      "epoch": 4.590740114903683,
      "grad_norm": 3.4154603481292725,
      "learning_rate": 2e-05,
      "loss": 0.2199,
      "step": 6794
    },
    {
      "epoch": 4.591416018925313,
      "grad_norm": 6.276726245880127,
      "learning_rate": 2e-05,
      "loss": 0.1886,
      "step": 6795
    },
    {
      "epoch": 4.592091922946942,
      "grad_norm": 1.7599725723266602,
      "learning_rate": 2e-05,
      "loss": 0.1238,
      "step": 6796
    },
    {
      "epoch": 4.59276782696857,
      "grad_norm": 5.572595119476318,
      "learning_rate": 2e-05,
      "loss": 0.1551,
      "step": 6797
    },
    {
      "epoch": 4.5934437309902,
      "grad_norm": 2.484423875808716,
      "learning_rate": 2e-05,
      "loss": 0.1799,
      "step": 6798
    },
    {
      "epoch": 4.594119635011828,
      "grad_norm": 6.1682844161987305,
      "learning_rate": 2e-05,
      "loss": 0.1604,
      "step": 6799
    },
    {
      "epoch": 4.594795539033457,
      "grad_norm": 3.8647890090942383,
      "learning_rate": 2e-05,
      "loss": 0.1765,
      "step": 6800
    },
    {
      "epoch": 4.594795539033457,
      "eval_accuracy": 0.7311655405405405,
      "eval_loss": 0.22704830765724182,
      "eval_runtime": 74.6775,
      "eval_samples_per_second": 317.097,
      "eval_steps_per_second": 9.909,
      "step": 6800
    },
    {
      "epoch": 4.595471443055086,
      "grad_norm": 2.9994540214538574,
      "learning_rate": 2e-05,
      "loss": 0.1079,
      "step": 6801
    },
    {
      "epoch": 4.596147347076715,
      "grad_norm": 7.889893531799316,
      "learning_rate": 2e-05,
      "loss": 0.1919,
      "step": 6802
    },
    {
      "epoch": 4.596823251098344,
      "grad_norm": 2.3198962211608887,
      "learning_rate": 2e-05,
      "loss": 0.1786,
      "step": 6803
    },
    {
      "epoch": 4.597499155119973,
      "grad_norm": 3.402505397796631,
      "learning_rate": 2e-05,
      "loss": 0.1207,
      "step": 6804
    },
    {
      "epoch": 4.598175059141602,
      "grad_norm": 3.292012929916382,
      "learning_rate": 2e-05,
      "loss": 0.1553,
      "step": 6805
    },
    {
      "epoch": 4.598850963163231,
      "grad_norm": 3.498296022415161,
      "learning_rate": 2e-05,
      "loss": 0.1855,
      "step": 6806
    },
    {
      "epoch": 4.59952686718486,
      "grad_norm": 3.854365587234497,
      "learning_rate": 2e-05,
      "loss": 0.1526,
      "step": 6807
    },
    {
      "epoch": 4.600202771206488,
      "grad_norm": 4.966411590576172,
      "learning_rate": 2e-05,
      "loss": 0.1687,
      "step": 6808
    },
    {
      "epoch": 4.600878675228118,
      "grad_norm": 5.021853923797607,
      "learning_rate": 2e-05,
      "loss": 0.0927,
      "step": 6809
    },
    {
      "epoch": 4.601554579249747,
      "grad_norm": 4.508306503295898,
      "learning_rate": 2e-05,
      "loss": 0.1582,
      "step": 6810
    },
    {
      "epoch": 4.602230483271375,
      "grad_norm": 3.1172213554382324,
      "learning_rate": 2e-05,
      "loss": 0.124,
      "step": 6811
    },
    {
      "epoch": 4.602906387293005,
      "grad_norm": 3.5172204971313477,
      "learning_rate": 2e-05,
      "loss": 0.2551,
      "step": 6812
    },
    {
      "epoch": 4.603582291314633,
      "grad_norm": 7.521477699279785,
      "learning_rate": 2e-05,
      "loss": 0.2023,
      "step": 6813
    },
    {
      "epoch": 4.6042581953362625,
      "grad_norm": 5.051335334777832,
      "learning_rate": 2e-05,
      "loss": 0.1787,
      "step": 6814
    },
    {
      "epoch": 4.604934099357891,
      "grad_norm": 4.0239434242248535,
      "learning_rate": 2e-05,
      "loss": 0.1302,
      "step": 6815
    },
    {
      "epoch": 4.60561000337952,
      "grad_norm": 3.444854736328125,
      "learning_rate": 2e-05,
      "loss": 0.1856,
      "step": 6816
    },
    {
      "epoch": 4.6062859074011495,
      "grad_norm": 3.6826460361480713,
      "learning_rate": 2e-05,
      "loss": 0.2214,
      "step": 6817
    },
    {
      "epoch": 4.606961811422778,
      "grad_norm": 4.35515022277832,
      "learning_rate": 2e-05,
      "loss": 0.1794,
      "step": 6818
    },
    {
      "epoch": 4.607637715444407,
      "grad_norm": 5.871730327606201,
      "learning_rate": 2e-05,
      "loss": 0.1638,
      "step": 6819
    },
    {
      "epoch": 4.608313619466036,
      "grad_norm": 8.081541061401367,
      "learning_rate": 2e-05,
      "loss": 0.1914,
      "step": 6820
    },
    {
      "epoch": 4.608989523487665,
      "grad_norm": 1.4274545907974243,
      "learning_rate": 2e-05,
      "loss": 0.106,
      "step": 6821
    },
    {
      "epoch": 4.609665427509293,
      "grad_norm": 5.864448070526123,
      "learning_rate": 2e-05,
      "loss": 0.1917,
      "step": 6822
    },
    {
      "epoch": 4.610341331530923,
      "grad_norm": 2.237647294998169,
      "learning_rate": 2e-05,
      "loss": 0.1539,
      "step": 6823
    },
    {
      "epoch": 4.611017235552551,
      "grad_norm": 4.998645305633545,
      "learning_rate": 2e-05,
      "loss": 0.1554,
      "step": 6824
    },
    {
      "epoch": 4.6116931395741805,
      "grad_norm": 7.253979206085205,
      "learning_rate": 2e-05,
      "loss": 0.1781,
      "step": 6825
    },
    {
      "epoch": 4.612369043595809,
      "grad_norm": 7.123815536499023,
      "learning_rate": 2e-05,
      "loss": 0.183,
      "step": 6826
    },
    {
      "epoch": 4.613044947617438,
      "grad_norm": 2.9315619468688965,
      "learning_rate": 2e-05,
      "loss": 0.1162,
      "step": 6827
    },
    {
      "epoch": 4.6137208516390675,
      "grad_norm": 2.5407140254974365,
      "learning_rate": 2e-05,
      "loss": 0.1485,
      "step": 6828
    },
    {
      "epoch": 4.614396755660696,
      "grad_norm": 1.615835189819336,
      "learning_rate": 2e-05,
      "loss": 0.1012,
      "step": 6829
    },
    {
      "epoch": 4.615072659682325,
      "grad_norm": 2.3681273460388184,
      "learning_rate": 2e-05,
      "loss": 0.191,
      "step": 6830
    },
    {
      "epoch": 4.615748563703954,
      "grad_norm": 9.98496150970459,
      "learning_rate": 2e-05,
      "loss": 0.1728,
      "step": 6831
    },
    {
      "epoch": 4.616424467725583,
      "grad_norm": 6.613104343414307,
      "learning_rate": 2e-05,
      "loss": 0.1951,
      "step": 6832
    },
    {
      "epoch": 4.617100371747211,
      "grad_norm": 2.454094648361206,
      "learning_rate": 2e-05,
      "loss": 0.1523,
      "step": 6833
    },
    {
      "epoch": 4.617776275768841,
      "grad_norm": 2.328418493270874,
      "learning_rate": 2e-05,
      "loss": 0.1772,
      "step": 6834
    },
    {
      "epoch": 4.61845217979047,
      "grad_norm": 2.3355681896209717,
      "learning_rate": 2e-05,
      "loss": 0.1196,
      "step": 6835
    },
    {
      "epoch": 4.6191280838120985,
      "grad_norm": 4.831709861755371,
      "learning_rate": 2e-05,
      "loss": 0.2253,
      "step": 6836
    },
    {
      "epoch": 4.619803987833728,
      "grad_norm": 4.365566253662109,
      "learning_rate": 2e-05,
      "loss": 0.1284,
      "step": 6837
    },
    {
      "epoch": 4.620479891855356,
      "grad_norm": 2.1921534538269043,
      "learning_rate": 2e-05,
      "loss": 0.1231,
      "step": 6838
    },
    {
      "epoch": 4.6211557958769856,
      "grad_norm": 3.9905025959014893,
      "learning_rate": 2e-05,
      "loss": 0.1529,
      "step": 6839
    },
    {
      "epoch": 4.621831699898614,
      "grad_norm": 7.0273637771606445,
      "learning_rate": 2e-05,
      "loss": 0.1894,
      "step": 6840
    },
    {
      "epoch": 4.622507603920243,
      "grad_norm": 2.3571252822875977,
      "learning_rate": 2e-05,
      "loss": 0.1787,
      "step": 6841
    },
    {
      "epoch": 4.623183507941873,
      "grad_norm": 4.317793369293213,
      "learning_rate": 2e-05,
      "loss": 0.2142,
      "step": 6842
    },
    {
      "epoch": 4.623859411963501,
      "grad_norm": 2.4637417793273926,
      "learning_rate": 2e-05,
      "loss": 0.0914,
      "step": 6843
    },
    {
      "epoch": 4.62453531598513,
      "grad_norm": 2.8081343173980713,
      "learning_rate": 2e-05,
      "loss": 0.1871,
      "step": 6844
    },
    {
      "epoch": 4.625211220006759,
      "grad_norm": 4.253475666046143,
      "learning_rate": 2e-05,
      "loss": 0.1771,
      "step": 6845
    },
    {
      "epoch": 4.625887124028388,
      "grad_norm": 2.2432007789611816,
      "learning_rate": 2e-05,
      "loss": 0.1865,
      "step": 6846
    },
    {
      "epoch": 4.6265630280500165,
      "grad_norm": 3.7322328090667725,
      "learning_rate": 2e-05,
      "loss": 0.1172,
      "step": 6847
    },
    {
      "epoch": 4.627238932071646,
      "grad_norm": 1.8065848350524902,
      "learning_rate": 2e-05,
      "loss": 0.154,
      "step": 6848
    },
    {
      "epoch": 4.627914836093275,
      "grad_norm": 2.4964687824249268,
      "learning_rate": 2e-05,
      "loss": 0.1381,
      "step": 6849
    },
    {
      "epoch": 4.628590740114904,
      "grad_norm": 2.740401029586792,
      "learning_rate": 2e-05,
      "loss": 0.2144,
      "step": 6850
    },
    {
      "epoch": 4.628590740114904,
      "eval_accuracy": 0.718285472972973,
      "eval_loss": 0.22786945104599,
      "eval_runtime": 84.2361,
      "eval_samples_per_second": 281.115,
      "eval_steps_per_second": 8.785,
      "step": 6850
    },
    {
      "epoch": 4.629266644136533,
      "grad_norm": 1.9507471323013306,
      "learning_rate": 2e-05,
      "loss": 0.1799,
      "step": 6851
    },
    {
      "epoch": 4.629942548158161,
      "grad_norm": 2.2508649826049805,
      "learning_rate": 2e-05,
      "loss": 0.1604,
      "step": 6852
    },
    {
      "epoch": 4.630618452179791,
      "grad_norm": 7.814982891082764,
      "learning_rate": 2e-05,
      "loss": 0.1924,
      "step": 6853
    },
    {
      "epoch": 4.631294356201419,
      "grad_norm": 1.739836573600769,
      "learning_rate": 2e-05,
      "loss": 0.1533,
      "step": 6854
    },
    {
      "epoch": 4.631970260223048,
      "grad_norm": 7.18069314956665,
      "learning_rate": 2e-05,
      "loss": 0.2054,
      "step": 6855
    },
    {
      "epoch": 4.632646164244678,
      "grad_norm": 8.190669059753418,
      "learning_rate": 2e-05,
      "loss": 0.1714,
      "step": 6856
    },
    {
      "epoch": 4.633322068266306,
      "grad_norm": 2.4630203247070312,
      "learning_rate": 2e-05,
      "loss": 0.1632,
      "step": 6857
    },
    {
      "epoch": 4.633997972287935,
      "grad_norm": 1.7136108875274658,
      "learning_rate": 2e-05,
      "loss": 0.1545,
      "step": 6858
    },
    {
      "epoch": 4.634673876309564,
      "grad_norm": 2.3414134979248047,
      "learning_rate": 2e-05,
      "loss": 0.1326,
      "step": 6859
    },
    {
      "epoch": 4.635349780331193,
      "grad_norm": 1.5831667184829712,
      "learning_rate": 2e-05,
      "loss": 0.1336,
      "step": 6860
    },
    {
      "epoch": 4.636025684352822,
      "grad_norm": 3.5200798511505127,
      "learning_rate": 2e-05,
      "loss": 0.1576,
      "step": 6861
    },
    {
      "epoch": 4.636701588374451,
      "grad_norm": 6.327441215515137,
      "learning_rate": 2e-05,
      "loss": 0.1337,
      "step": 6862
    },
    {
      "epoch": 4.63737749239608,
      "grad_norm": 3.158012866973877,
      "learning_rate": 2e-05,
      "loss": 0.1536,
      "step": 6863
    },
    {
      "epoch": 4.638053396417709,
      "grad_norm": 2.212235927581787,
      "learning_rate": 2e-05,
      "loss": 0.1585,
      "step": 6864
    },
    {
      "epoch": 4.638729300439338,
      "grad_norm": 2.060326337814331,
      "learning_rate": 2e-05,
      "loss": 0.1396,
      "step": 6865
    },
    {
      "epoch": 4.639405204460966,
      "grad_norm": 1.619863510131836,
      "learning_rate": 2e-05,
      "loss": 0.1113,
      "step": 6866
    },
    {
      "epoch": 4.640081108482596,
      "grad_norm": 2.5023622512817383,
      "learning_rate": 2e-05,
      "loss": 0.1905,
      "step": 6867
    },
    {
      "epoch": 4.640757012504224,
      "grad_norm": 3.063429594039917,
      "learning_rate": 2e-05,
      "loss": 0.1496,
      "step": 6868
    },
    {
      "epoch": 4.641432916525853,
      "grad_norm": 2.684736967086792,
      "learning_rate": 2e-05,
      "loss": 0.1756,
      "step": 6869
    },
    {
      "epoch": 4.642108820547483,
      "grad_norm": 5.675417423248291,
      "learning_rate": 2e-05,
      "loss": 0.1911,
      "step": 6870
    },
    {
      "epoch": 4.642784724569111,
      "grad_norm": 4.373174667358398,
      "learning_rate": 2e-05,
      "loss": 0.1548,
      "step": 6871
    },
    {
      "epoch": 4.6434606285907405,
      "grad_norm": 5.174712181091309,
      "learning_rate": 2e-05,
      "loss": 0.1662,
      "step": 6872
    },
    {
      "epoch": 4.644136532612369,
      "grad_norm": 7.059999465942383,
      "learning_rate": 2e-05,
      "loss": 0.1563,
      "step": 6873
    },
    {
      "epoch": 4.644812436633998,
      "grad_norm": 4.245145320892334,
      "learning_rate": 2e-05,
      "loss": 0.1371,
      "step": 6874
    },
    {
      "epoch": 4.645488340655627,
      "grad_norm": 1.8594075441360474,
      "learning_rate": 2e-05,
      "loss": 0.1621,
      "step": 6875
    },
    {
      "epoch": 4.646164244677256,
      "grad_norm": 6.4442315101623535,
      "learning_rate": 2e-05,
      "loss": 0.1684,
      "step": 6876
    },
    {
      "epoch": 4.646840148698884,
      "grad_norm": 3.108337640762329,
      "learning_rate": 2e-05,
      "loss": 0.1237,
      "step": 6877
    },
    {
      "epoch": 4.647516052720514,
      "grad_norm": 2.0232489109039307,
      "learning_rate": 2e-05,
      "loss": 0.1089,
      "step": 6878
    },
    {
      "epoch": 4.648191956742142,
      "grad_norm": 2.6868090629577637,
      "learning_rate": 2e-05,
      "loss": 0.1444,
      "step": 6879
    },
    {
      "epoch": 4.6488678607637715,
      "grad_norm": 4.284171104431152,
      "learning_rate": 2e-05,
      "loss": 0.1632,
      "step": 6880
    },
    {
      "epoch": 4.649543764785401,
      "grad_norm": 8.127676010131836,
      "learning_rate": 2e-05,
      "loss": 0.2099,
      "step": 6881
    },
    {
      "epoch": 4.650219668807029,
      "grad_norm": 7.376944065093994,
      "learning_rate": 2e-05,
      "loss": 0.2708,
      "step": 6882
    },
    {
      "epoch": 4.6508955728286585,
      "grad_norm": 2.616464853286743,
      "learning_rate": 2e-05,
      "loss": 0.1818,
      "step": 6883
    },
    {
      "epoch": 4.651571476850287,
      "grad_norm": 4.105311393737793,
      "learning_rate": 2e-05,
      "loss": 0.2062,
      "step": 6884
    },
    {
      "epoch": 4.652247380871916,
      "grad_norm": 14.050629615783691,
      "learning_rate": 2e-05,
      "loss": 0.2886,
      "step": 6885
    },
    {
      "epoch": 4.652923284893545,
      "grad_norm": 3.575850486755371,
      "learning_rate": 2e-05,
      "loss": 0.2175,
      "step": 6886
    },
    {
      "epoch": 4.653599188915174,
      "grad_norm": 8.569762229919434,
      "learning_rate": 2e-05,
      "loss": 0.1828,
      "step": 6887
    },
    {
      "epoch": 4.654275092936803,
      "grad_norm": 8.994558334350586,
      "learning_rate": 2e-05,
      "loss": 0.1705,
      "step": 6888
    },
    {
      "epoch": 4.654950996958432,
      "grad_norm": 7.7436652183532715,
      "learning_rate": 2e-05,
      "loss": 0.1793,
      "step": 6889
    },
    {
      "epoch": 4.655626900980061,
      "grad_norm": 2.929455041885376,
      "learning_rate": 2e-05,
      "loss": 0.2176,
      "step": 6890
    },
    {
      "epoch": 4.6563028050016895,
      "grad_norm": 5.397281169891357,
      "learning_rate": 2e-05,
      "loss": 0.1383,
      "step": 6891
    },
    {
      "epoch": 4.656978709023319,
      "grad_norm": 3.8931071758270264,
      "learning_rate": 2e-05,
      "loss": 0.1731,
      "step": 6892
    },
    {
      "epoch": 4.657654613044947,
      "grad_norm": 4.594730854034424,
      "learning_rate": 2e-05,
      "loss": 0.1883,
      "step": 6893
    },
    {
      "epoch": 4.6583305170665765,
      "grad_norm": 7.681121826171875,
      "learning_rate": 2e-05,
      "loss": 0.2159,
      "step": 6894
    },
    {
      "epoch": 4.659006421088206,
      "grad_norm": 6.228323936462402,
      "learning_rate": 2e-05,
      "loss": 0.19,
      "step": 6895
    },
    {
      "epoch": 4.659682325109834,
      "grad_norm": 2.834243059158325,
      "learning_rate": 2e-05,
      "loss": 0.1628,
      "step": 6896
    },
    {
      "epoch": 4.660358229131464,
      "grad_norm": 1.4797759056091309,
      "learning_rate": 2e-05,
      "loss": 0.1739,
      "step": 6897
    },
    {
      "epoch": 4.661034133153092,
      "grad_norm": 2.28169322013855,
      "learning_rate": 2e-05,
      "loss": 0.1465,
      "step": 6898
    },
    {
      "epoch": 4.661710037174721,
      "grad_norm": 4.542727947235107,
      "learning_rate": 2e-05,
      "loss": 0.1535,
      "step": 6899
    },
    {
      "epoch": 4.66238594119635,
      "grad_norm": 3.1822476387023926,
      "learning_rate": 2e-05,
      "loss": 0.2463,
      "step": 6900
    },
    {
      "epoch": 4.66238594119635,
      "eval_accuracy": 0.7210726351351351,
      "eval_loss": 0.22153182327747345,
      "eval_runtime": 82.1404,
      "eval_samples_per_second": 288.287,
      "eval_steps_per_second": 9.009,
      "step": 6900
    },
    {
      "epoch": 4.663061845217979,
      "grad_norm": 3.076699733734131,
      "learning_rate": 2e-05,
      "loss": 0.1751,
      "step": 6901
    },
    {
      "epoch": 4.663737749239608,
      "grad_norm": 6.156479835510254,
      "learning_rate": 2e-05,
      "loss": 0.2358,
      "step": 6902
    },
    {
      "epoch": 4.664413653261237,
      "grad_norm": 4.570157051086426,
      "learning_rate": 2e-05,
      "loss": 0.177,
      "step": 6903
    },
    {
      "epoch": 4.665089557282866,
      "grad_norm": 2.2710773944854736,
      "learning_rate": 2e-05,
      "loss": 0.1202,
      "step": 6904
    },
    {
      "epoch": 4.6657654613044945,
      "grad_norm": 2.230480432510376,
      "learning_rate": 2e-05,
      "loss": 0.1979,
      "step": 6905
    },
    {
      "epoch": 4.666441365326124,
      "grad_norm": 4.17747688293457,
      "learning_rate": 2e-05,
      "loss": 0.1637,
      "step": 6906
    },
    {
      "epoch": 4.667117269347752,
      "grad_norm": 2.9045186042785645,
      "learning_rate": 2e-05,
      "loss": 0.1637,
      "step": 6907
    },
    {
      "epoch": 4.667793173369382,
      "grad_norm": 1.9130500555038452,
      "learning_rate": 2e-05,
      "loss": 0.17,
      "step": 6908
    },
    {
      "epoch": 4.668469077391011,
      "grad_norm": 7.369197845458984,
      "learning_rate": 2e-05,
      "loss": 0.1541,
      "step": 6909
    },
    {
      "epoch": 4.669144981412639,
      "grad_norm": 3.6780292987823486,
      "learning_rate": 2e-05,
      "loss": 0.0947,
      "step": 6910
    },
    {
      "epoch": 4.669820885434269,
      "grad_norm": 2.396036148071289,
      "learning_rate": 2e-05,
      "loss": 0.1445,
      "step": 6911
    },
    {
      "epoch": 4.670496789455897,
      "grad_norm": 3.2852160930633545,
      "learning_rate": 2e-05,
      "loss": 0.1405,
      "step": 6912
    },
    {
      "epoch": 4.671172693477526,
      "grad_norm": 7.969626426696777,
      "learning_rate": 2e-05,
      "loss": 0.1831,
      "step": 6913
    },
    {
      "epoch": 4.671848597499155,
      "grad_norm": 2.8499953746795654,
      "learning_rate": 2e-05,
      "loss": 0.1259,
      "step": 6914
    },
    {
      "epoch": 4.672524501520784,
      "grad_norm": 2.9077258110046387,
      "learning_rate": 2e-05,
      "loss": 0.22,
      "step": 6915
    },
    {
      "epoch": 4.6732004055424134,
      "grad_norm": 2.005695104598999,
      "learning_rate": 2e-05,
      "loss": 0.1366,
      "step": 6916
    },
    {
      "epoch": 4.673876309564042,
      "grad_norm": 4.886782169342041,
      "learning_rate": 2e-05,
      "loss": 0.1581,
      "step": 6917
    },
    {
      "epoch": 4.674552213585671,
      "grad_norm": 2.0739994049072266,
      "learning_rate": 2e-05,
      "loss": 0.1309,
      "step": 6918
    },
    {
      "epoch": 4.6752281176073,
      "grad_norm": 6.322997570037842,
      "learning_rate": 2e-05,
      "loss": 0.2533,
      "step": 6919
    },
    {
      "epoch": 4.675904021628929,
      "grad_norm": 5.9264817237854,
      "learning_rate": 2e-05,
      "loss": 0.1695,
      "step": 6920
    },
    {
      "epoch": 4.676579925650557,
      "grad_norm": 2.8349084854125977,
      "learning_rate": 2e-05,
      "loss": 0.1146,
      "step": 6921
    },
    {
      "epoch": 4.677255829672187,
      "grad_norm": 3.862192153930664,
      "learning_rate": 2e-05,
      "loss": 0.1645,
      "step": 6922
    },
    {
      "epoch": 4.677931733693815,
      "grad_norm": 5.496240139007568,
      "learning_rate": 2e-05,
      "loss": 0.1759,
      "step": 6923
    },
    {
      "epoch": 4.678607637715444,
      "grad_norm": 2.9162075519561768,
      "learning_rate": 2e-05,
      "loss": 0.1289,
      "step": 6924
    },
    {
      "epoch": 4.679283541737073,
      "grad_norm": 8.737896919250488,
      "learning_rate": 2e-05,
      "loss": 0.1712,
      "step": 6925
    },
    {
      "epoch": 4.679959445758702,
      "grad_norm": 2.2764909267425537,
      "learning_rate": 2e-05,
      "loss": 0.1539,
      "step": 6926
    },
    {
      "epoch": 4.6806353497803315,
      "grad_norm": 3.6565654277801514,
      "learning_rate": 2e-05,
      "loss": 0.1179,
      "step": 6927
    },
    {
      "epoch": 4.68131125380196,
      "grad_norm": 1.9109504222869873,
      "learning_rate": 2e-05,
      "loss": 0.1851,
      "step": 6928
    },
    {
      "epoch": 4.681987157823589,
      "grad_norm": 2.7769947052001953,
      "learning_rate": 2e-05,
      "loss": 0.174,
      "step": 6929
    },
    {
      "epoch": 4.682663061845218,
      "grad_norm": 1.5542269945144653,
      "learning_rate": 2e-05,
      "loss": 0.1376,
      "step": 6930
    },
    {
      "epoch": 4.683338965866847,
      "grad_norm": 2.056983709335327,
      "learning_rate": 2e-05,
      "loss": 0.1587,
      "step": 6931
    },
    {
      "epoch": 4.684014869888475,
      "grad_norm": 2.949127674102783,
      "learning_rate": 2e-05,
      "loss": 0.0915,
      "step": 6932
    },
    {
      "epoch": 4.684690773910105,
      "grad_norm": 1.6760447025299072,
      "learning_rate": 2e-05,
      "loss": 0.138,
      "step": 6933
    },
    {
      "epoch": 4.685366677931734,
      "grad_norm": 4.389612674713135,
      "learning_rate": 2e-05,
      "loss": 0.1548,
      "step": 6934
    },
    {
      "epoch": 4.686042581953362,
      "grad_norm": 6.231406211853027,
      "learning_rate": 2e-05,
      "loss": 0.1666,
      "step": 6935
    },
    {
      "epoch": 4.686718485974992,
      "grad_norm": 2.7618823051452637,
      "learning_rate": 2e-05,
      "loss": 0.1525,
      "step": 6936
    },
    {
      "epoch": 4.68739438999662,
      "grad_norm": 6.3954644203186035,
      "learning_rate": 2e-05,
      "loss": 0.1687,
      "step": 6937
    },
    {
      "epoch": 4.6880702940182495,
      "grad_norm": 4.370234489440918,
      "learning_rate": 2e-05,
      "loss": 0.1681,
      "step": 6938
    },
    {
      "epoch": 4.688746198039878,
      "grad_norm": 2.388615369796753,
      "learning_rate": 2e-05,
      "loss": 0.1572,
      "step": 6939
    },
    {
      "epoch": 4.689422102061507,
      "grad_norm": 2.746002674102783,
      "learning_rate": 2e-05,
      "loss": 0.1594,
      "step": 6940
    },
    {
      "epoch": 4.6900980060831365,
      "grad_norm": 3.5504467487335205,
      "learning_rate": 2e-05,
      "loss": 0.2029,
      "step": 6941
    },
    {
      "epoch": 4.690773910104765,
      "grad_norm": 5.401784420013428,
      "learning_rate": 2e-05,
      "loss": 0.1606,
      "step": 6942
    },
    {
      "epoch": 4.691449814126394,
      "grad_norm": 2.4599578380584717,
      "learning_rate": 2e-05,
      "loss": 0.1559,
      "step": 6943
    },
    {
      "epoch": 4.692125718148023,
      "grad_norm": 3.608961343765259,
      "learning_rate": 2e-05,
      "loss": 0.1509,
      "step": 6944
    },
    {
      "epoch": 4.692801622169652,
      "grad_norm": 3.6577556133270264,
      "learning_rate": 2e-05,
      "loss": 0.1234,
      "step": 6945
    },
    {
      "epoch": 4.6934775261912804,
      "grad_norm": 4.182014465332031,
      "learning_rate": 2e-05,
      "loss": 0.189,
      "step": 6946
    },
    {
      "epoch": 4.69415343021291,
      "grad_norm": 3.0080525875091553,
      "learning_rate": 2e-05,
      "loss": 0.1339,
      "step": 6947
    },
    {
      "epoch": 4.694829334234539,
      "grad_norm": 2.0154407024383545,
      "learning_rate": 2e-05,
      "loss": 0.1606,
      "step": 6948
    },
    {
      "epoch": 4.6955052382561675,
      "grad_norm": 3.1667542457580566,
      "learning_rate": 2e-05,
      "loss": 0.1582,
      "step": 6949
    },
    {
      "epoch": 4.696181142277797,
      "grad_norm": 2.3478283882141113,
      "learning_rate": 2e-05,
      "loss": 0.0859,
      "step": 6950
    },
    {
      "epoch": 4.696181142277797,
      "eval_accuracy": 0.7420608108108108,
      "eval_loss": 0.24122178554534912,
      "eval_runtime": 74.2834,
      "eval_samples_per_second": 318.779,
      "eval_steps_per_second": 9.962,
      "step": 6950
    },
    {
      "epoch": 4.696857046299425,
      "grad_norm": 3.215951681137085,
      "learning_rate": 2e-05,
      "loss": 0.1257,
      "step": 6951
    },
    {
      "epoch": 4.697532950321055,
      "grad_norm": 5.451211929321289,
      "learning_rate": 2e-05,
      "loss": 0.2038,
      "step": 6952
    },
    {
      "epoch": 4.698208854342683,
      "grad_norm": 7.18816614151001,
      "learning_rate": 2e-05,
      "loss": 0.2201,
      "step": 6953
    },
    {
      "epoch": 4.698884758364312,
      "grad_norm": 4.679708480834961,
      "learning_rate": 2e-05,
      "loss": 0.2602,
      "step": 6954
    },
    {
      "epoch": 4.699560662385942,
      "grad_norm": 2.9920029640197754,
      "learning_rate": 2e-05,
      "loss": 0.1424,
      "step": 6955
    },
    {
      "epoch": 4.70023656640757,
      "grad_norm": 2.4234888553619385,
      "learning_rate": 2e-05,
      "loss": 0.1548,
      "step": 6956
    },
    {
      "epoch": 4.700912470429199,
      "grad_norm": 9.270777702331543,
      "learning_rate": 2e-05,
      "loss": 0.1825,
      "step": 6957
    },
    {
      "epoch": 4.701588374450828,
      "grad_norm": 2.048539400100708,
      "learning_rate": 2e-05,
      "loss": 0.1662,
      "step": 6958
    },
    {
      "epoch": 4.702264278472457,
      "grad_norm": 3.2400104999542236,
      "learning_rate": 2e-05,
      "loss": 0.1766,
      "step": 6959
    },
    {
      "epoch": 4.7029401824940855,
      "grad_norm": 3.878714084625244,
      "learning_rate": 2e-05,
      "loss": 0.1559,
      "step": 6960
    },
    {
      "epoch": 4.703616086515715,
      "grad_norm": 2.030090808868408,
      "learning_rate": 2e-05,
      "loss": 0.1137,
      "step": 6961
    },
    {
      "epoch": 4.704291990537344,
      "grad_norm": 2.059547185897827,
      "learning_rate": 2e-05,
      "loss": 0.1546,
      "step": 6962
    },
    {
      "epoch": 4.704967894558973,
      "grad_norm": 3.8585991859436035,
      "learning_rate": 2e-05,
      "loss": 0.1659,
      "step": 6963
    },
    {
      "epoch": 4.705643798580602,
      "grad_norm": 1.831348180770874,
      "learning_rate": 2e-05,
      "loss": 0.1642,
      "step": 6964
    },
    {
      "epoch": 4.70631970260223,
      "grad_norm": 2.8371999263763428,
      "learning_rate": 2e-05,
      "loss": 0.1348,
      "step": 6965
    },
    {
      "epoch": 4.70699560662386,
      "grad_norm": 11.813593864440918,
      "learning_rate": 2e-05,
      "loss": 0.2235,
      "step": 6966
    },
    {
      "epoch": 4.707671510645488,
      "grad_norm": 5.2821760177612305,
      "learning_rate": 2e-05,
      "loss": 0.2026,
      "step": 6967
    },
    {
      "epoch": 4.708347414667117,
      "grad_norm": 7.4265828132629395,
      "learning_rate": 2e-05,
      "loss": 0.2227,
      "step": 6968
    },
    {
      "epoch": 4.709023318688747,
      "grad_norm": 2.0451416969299316,
      "learning_rate": 2e-05,
      "loss": 0.1751,
      "step": 6969
    },
    {
      "epoch": 4.709699222710375,
      "grad_norm": 5.876232147216797,
      "learning_rate": 2e-05,
      "loss": 0.1551,
      "step": 6970
    },
    {
      "epoch": 4.710375126732004,
      "grad_norm": 5.575299263000488,
      "learning_rate": 2e-05,
      "loss": 0.1703,
      "step": 6971
    },
    {
      "epoch": 4.711051030753633,
      "grad_norm": 3.954850196838379,
      "learning_rate": 2e-05,
      "loss": 0.2212,
      "step": 6972
    },
    {
      "epoch": 4.711726934775262,
      "grad_norm": 2.044768810272217,
      "learning_rate": 2e-05,
      "loss": 0.146,
      "step": 6973
    },
    {
      "epoch": 4.712402838796891,
      "grad_norm": 4.483190059661865,
      "learning_rate": 2e-05,
      "loss": 0.1909,
      "step": 6974
    },
    {
      "epoch": 4.71307874281852,
      "grad_norm": 6.076782703399658,
      "learning_rate": 2e-05,
      "loss": 0.1357,
      "step": 6975
    },
    {
      "epoch": 4.713754646840148,
      "grad_norm": 5.999468803405762,
      "learning_rate": 2e-05,
      "loss": 0.1729,
      "step": 6976
    },
    {
      "epoch": 4.714430550861778,
      "grad_norm": 5.745473384857178,
      "learning_rate": 2e-05,
      "loss": 0.1476,
      "step": 6977
    },
    {
      "epoch": 4.715106454883406,
      "grad_norm": 2.2996349334716797,
      "learning_rate": 2e-05,
      "loss": 0.1109,
      "step": 6978
    },
    {
      "epoch": 4.715782358905035,
      "grad_norm": 4.4828410148620605,
      "learning_rate": 2e-05,
      "loss": 0.133,
      "step": 6979
    },
    {
      "epoch": 4.716458262926665,
      "grad_norm": 3.3437082767486572,
      "learning_rate": 2e-05,
      "loss": 0.1509,
      "step": 6980
    },
    {
      "epoch": 4.717134166948293,
      "grad_norm": 5.585807800292969,
      "learning_rate": 2e-05,
      "loss": 0.1793,
      "step": 6981
    },
    {
      "epoch": 4.717810070969922,
      "grad_norm": 3.5466041564941406,
      "learning_rate": 2e-05,
      "loss": 0.1306,
      "step": 6982
    },
    {
      "epoch": 4.718485974991551,
      "grad_norm": 7.915553092956543,
      "learning_rate": 2e-05,
      "loss": 0.1781,
      "step": 6983
    },
    {
      "epoch": 4.71916187901318,
      "grad_norm": 4.087377548217773,
      "learning_rate": 2e-05,
      "loss": 0.2152,
      "step": 6984
    },
    {
      "epoch": 4.719837783034809,
      "grad_norm": 3.300783634185791,
      "learning_rate": 2e-05,
      "loss": 0.1642,
      "step": 6985
    },
    {
      "epoch": 4.720513687056438,
      "grad_norm": 4.0137553215026855,
      "learning_rate": 2e-05,
      "loss": 0.1363,
      "step": 6986
    },
    {
      "epoch": 4.721189591078067,
      "grad_norm": 5.284428596496582,
      "learning_rate": 2e-05,
      "loss": 0.176,
      "step": 6987
    },
    {
      "epoch": 4.721865495099696,
      "grad_norm": 6.886884689331055,
      "learning_rate": 2e-05,
      "loss": 0.1327,
      "step": 6988
    },
    {
      "epoch": 4.722541399121325,
      "grad_norm": 6.411003112792969,
      "learning_rate": 2e-05,
      "loss": 0.2028,
      "step": 6989
    },
    {
      "epoch": 4.723217303142953,
      "grad_norm": 5.1047258377075195,
      "learning_rate": 2e-05,
      "loss": 0.2542,
      "step": 6990
    },
    {
      "epoch": 4.723893207164583,
      "grad_norm": 2.081336498260498,
      "learning_rate": 2e-05,
      "loss": 0.1148,
      "step": 6991
    },
    {
      "epoch": 4.724569111186211,
      "grad_norm": 4.5318803787231445,
      "learning_rate": 2e-05,
      "loss": 0.1707,
      "step": 6992
    },
    {
      "epoch": 4.7252450152078405,
      "grad_norm": 5.308281421661377,
      "learning_rate": 2e-05,
      "loss": 0.1303,
      "step": 6993
    },
    {
      "epoch": 4.72592091922947,
      "grad_norm": 3.714707136154175,
      "learning_rate": 2e-05,
      "loss": 0.163,
      "step": 6994
    },
    {
      "epoch": 4.726596823251098,
      "grad_norm": 2.9690346717834473,
      "learning_rate": 2e-05,
      "loss": 0.1312,
      "step": 6995
    },
    {
      "epoch": 4.7272727272727275,
      "grad_norm": 4.687417030334473,
      "learning_rate": 2e-05,
      "loss": 0.1557,
      "step": 6996
    },
    {
      "epoch": 4.727948631294356,
      "grad_norm": 3.0576415061950684,
      "learning_rate": 2e-05,
      "loss": 0.1316,
      "step": 6997
    },
    {
      "epoch": 4.728624535315985,
      "grad_norm": 4.104924201965332,
      "learning_rate": 2e-05,
      "loss": 0.1633,
      "step": 6998
    },
    {
      "epoch": 4.729300439337614,
      "grad_norm": 2.740114212036133,
      "learning_rate": 2e-05,
      "loss": 0.1379,
      "step": 6999
    },
    {
      "epoch": 4.729976343359243,
      "grad_norm": 3.137800693511963,
      "learning_rate": 2e-05,
      "loss": 0.1364,
      "step": 7000
    },
    {
      "epoch": 4.729976343359243,
      "eval_accuracy": 0.6913429054054054,
      "eval_loss": 0.2303849458694458,
      "eval_runtime": 49.0699,
      "eval_samples_per_second": 482.576,
      "eval_steps_per_second": 15.081,
      "step": 7000
    },
    {
      "epoch": 4.730652247380872,
      "grad_norm": 4.320187568664551,
      "learning_rate": 2e-05,
      "loss": 0.1595,
      "step": 7001
    },
    {
      "epoch": 4.731328151402501,
      "grad_norm": 5.9151225090026855,
      "learning_rate": 2e-05,
      "loss": 0.2214,
      "step": 7002
    },
    {
      "epoch": 4.73200405542413,
      "grad_norm": 2.334266424179077,
      "learning_rate": 2e-05,
      "loss": 0.1556,
      "step": 7003
    },
    {
      "epoch": 4.7326799594457585,
      "grad_norm": 1.6778051853179932,
      "learning_rate": 2e-05,
      "loss": 0.1375,
      "step": 7004
    },
    {
      "epoch": 4.733355863467388,
      "grad_norm": 5.056000232696533,
      "learning_rate": 2e-05,
      "loss": 0.2103,
      "step": 7005
    },
    {
      "epoch": 4.734031767489016,
      "grad_norm": 4.412537097930908,
      "learning_rate": 2e-05,
      "loss": 0.0891,
      "step": 7006
    },
    {
      "epoch": 4.7347076715106455,
      "grad_norm": 3.343648910522461,
      "learning_rate": 2e-05,
      "loss": 0.1593,
      "step": 7007
    },
    {
      "epoch": 4.735383575532275,
      "grad_norm": 6.090826034545898,
      "learning_rate": 2e-05,
      "loss": 0.182,
      "step": 7008
    },
    {
      "epoch": 4.736059479553903,
      "grad_norm": 5.039658546447754,
      "learning_rate": 2e-05,
      "loss": 0.1989,
      "step": 7009
    },
    {
      "epoch": 4.736735383575533,
      "grad_norm": 1.8900657892227173,
      "learning_rate": 2e-05,
      "loss": 0.1222,
      "step": 7010
    },
    {
      "epoch": 4.737411287597161,
      "grad_norm": 2.68585467338562,
      "learning_rate": 2e-05,
      "loss": 0.1478,
      "step": 7011
    },
    {
      "epoch": 4.73808719161879,
      "grad_norm": 1.8970310688018799,
      "learning_rate": 2e-05,
      "loss": 0.1179,
      "step": 7012
    },
    {
      "epoch": 4.738763095640419,
      "grad_norm": 4.2598347663879395,
      "learning_rate": 2e-05,
      "loss": 0.1387,
      "step": 7013
    },
    {
      "epoch": 4.739438999662048,
      "grad_norm": 3.66182804107666,
      "learning_rate": 2e-05,
      "loss": 0.101,
      "step": 7014
    },
    {
      "epoch": 4.740114903683677,
      "grad_norm": 3.464280366897583,
      "learning_rate": 2e-05,
      "loss": 0.1106,
      "step": 7015
    },
    {
      "epoch": 4.740790807705306,
      "grad_norm": 2.668043375015259,
      "learning_rate": 2e-05,
      "loss": 0.1994,
      "step": 7016
    },
    {
      "epoch": 4.741466711726935,
      "grad_norm": 2.8558011054992676,
      "learning_rate": 2e-05,
      "loss": 0.1625,
      "step": 7017
    },
    {
      "epoch": 4.7421426157485635,
      "grad_norm": 2.1265618801116943,
      "learning_rate": 2e-05,
      "loss": 0.1353,
      "step": 7018
    },
    {
      "epoch": 4.742818519770193,
      "grad_norm": 3.0915157794952393,
      "learning_rate": 2e-05,
      "loss": 0.1416,
      "step": 7019
    },
    {
      "epoch": 4.743494423791821,
      "grad_norm": 6.439059734344482,
      "learning_rate": 2e-05,
      "loss": 0.1298,
      "step": 7020
    },
    {
      "epoch": 4.744170327813451,
      "grad_norm": 3.3082034587860107,
      "learning_rate": 2e-05,
      "loss": 0.1332,
      "step": 7021
    },
    {
      "epoch": 4.744846231835079,
      "grad_norm": 3.4525909423828125,
      "learning_rate": 2e-05,
      "loss": 0.1784,
      "step": 7022
    },
    {
      "epoch": 4.745522135856708,
      "grad_norm": 2.9003565311431885,
      "learning_rate": 2e-05,
      "loss": 0.1669,
      "step": 7023
    },
    {
      "epoch": 4.746198039878338,
      "grad_norm": 4.582150936126709,
      "learning_rate": 2e-05,
      "loss": 0.1982,
      "step": 7024
    },
    {
      "epoch": 4.746873943899966,
      "grad_norm": 4.2883100509643555,
      "learning_rate": 2e-05,
      "loss": 0.1595,
      "step": 7025
    },
    {
      "epoch": 4.747549847921595,
      "grad_norm": 4.98576545715332,
      "learning_rate": 2e-05,
      "loss": 0.1798,
      "step": 7026
    },
    {
      "epoch": 4.748225751943224,
      "grad_norm": 3.714721918106079,
      "learning_rate": 2e-05,
      "loss": 0.201,
      "step": 7027
    },
    {
      "epoch": 4.748901655964853,
      "grad_norm": 4.031825065612793,
      "learning_rate": 2e-05,
      "loss": 0.137,
      "step": 7028
    },
    {
      "epoch": 4.749577559986482,
      "grad_norm": 3.5741002559661865,
      "learning_rate": 2e-05,
      "loss": 0.1561,
      "step": 7029
    },
    {
      "epoch": 4.750253464008111,
      "grad_norm": 8.511181831359863,
      "learning_rate": 2e-05,
      "loss": 0.1595,
      "step": 7030
    },
    {
      "epoch": 4.750929368029739,
      "grad_norm": 10.154016494750977,
      "learning_rate": 2e-05,
      "loss": 0.1723,
      "step": 7031
    },
    {
      "epoch": 4.751605272051369,
      "grad_norm": 4.0682759284973145,
      "learning_rate": 2e-05,
      "loss": 0.1615,
      "step": 7032
    },
    {
      "epoch": 4.752281176072998,
      "grad_norm": 3.0151541233062744,
      "learning_rate": 2e-05,
      "loss": 0.1444,
      "step": 7033
    },
    {
      "epoch": 4.752957080094626,
      "grad_norm": 6.5109758377075195,
      "learning_rate": 2e-05,
      "loss": 0.1981,
      "step": 7034
    },
    {
      "epoch": 4.753632984116256,
      "grad_norm": 9.449956893920898,
      "learning_rate": 2e-05,
      "loss": 0.2018,
      "step": 7035
    },
    {
      "epoch": 4.754308888137884,
      "grad_norm": 4.026897430419922,
      "learning_rate": 2e-05,
      "loss": 0.1771,
      "step": 7036
    },
    {
      "epoch": 4.754984792159513,
      "grad_norm": 3.1893980503082275,
      "learning_rate": 2e-05,
      "loss": 0.1325,
      "step": 7037
    },
    {
      "epoch": 4.755660696181142,
      "grad_norm": 4.8790717124938965,
      "learning_rate": 2e-05,
      "loss": 0.1985,
      "step": 7038
    },
    {
      "epoch": 4.756336600202771,
      "grad_norm": 2.2543883323669434,
      "learning_rate": 2e-05,
      "loss": 0.1777,
      "step": 7039
    },
    {
      "epoch": 4.7570125042244005,
      "grad_norm": 3.8161227703094482,
      "learning_rate": 2e-05,
      "loss": 0.1844,
      "step": 7040
    },
    {
      "epoch": 4.757688408246029,
      "grad_norm": 5.935304164886475,
      "learning_rate": 2e-05,
      "loss": 0.134,
      "step": 7041
    },
    {
      "epoch": 4.758364312267658,
      "grad_norm": 3.43159818649292,
      "learning_rate": 2e-05,
      "loss": 0.1838,
      "step": 7042
    },
    {
      "epoch": 4.759040216289287,
      "grad_norm": 6.746768951416016,
      "learning_rate": 2e-05,
      "loss": 0.1874,
      "step": 7043
    },
    {
      "epoch": 4.759716120310916,
      "grad_norm": 6.78346586227417,
      "learning_rate": 2e-05,
      "loss": 0.1382,
      "step": 7044
    },
    {
      "epoch": 4.760392024332544,
      "grad_norm": 1.335221529006958,
      "learning_rate": 2e-05,
      "loss": 0.103,
      "step": 7045
    },
    {
      "epoch": 4.761067928354174,
      "grad_norm": 5.045212745666504,
      "learning_rate": 2e-05,
      "loss": 0.1829,
      "step": 7046
    },
    {
      "epoch": 4.761743832375803,
      "grad_norm": 3.6248526573181152,
      "learning_rate": 2e-05,
      "loss": 0.144,
      "step": 7047
    },
    {
      "epoch": 4.762419736397431,
      "grad_norm": 1.7092103958129883,
      "learning_rate": 2e-05,
      "loss": 0.092,
      "step": 7048
    },
    {
      "epoch": 4.763095640419061,
      "grad_norm": 3.4890761375427246,
      "learning_rate": 2e-05,
      "loss": 0.2101,
      "step": 7049
    },
    {
      "epoch": 4.763771544440689,
      "grad_norm": 3.099322557449341,
      "learning_rate": 2e-05,
      "loss": 0.1956,
      "step": 7050
    },
    {
      "epoch": 4.763771544440689,
      "eval_accuracy": 0.6828547297297297,
      "eval_loss": 0.24820628762245178,
      "eval_runtime": 49.8286,
      "eval_samples_per_second": 475.229,
      "eval_steps_per_second": 14.851,
      "step": 7050
    },
    {
      "epoch": 4.7644474484623185,
      "grad_norm": 4.081798553466797,
      "learning_rate": 2e-05,
      "loss": 0.1881,
      "step": 7051
    },
    {
      "epoch": 4.765123352483947,
      "grad_norm": 3.3819074630737305,
      "learning_rate": 2e-05,
      "loss": 0.159,
      "step": 7052
    },
    {
      "epoch": 4.765799256505576,
      "grad_norm": 7.205258846282959,
      "learning_rate": 2e-05,
      "loss": 0.1199,
      "step": 7053
    },
    {
      "epoch": 4.7664751605272055,
      "grad_norm": 7.166323661804199,
      "learning_rate": 2e-05,
      "loss": 0.22,
      "step": 7054
    },
    {
      "epoch": 4.767151064548834,
      "grad_norm": 5.161471366882324,
      "learning_rate": 2e-05,
      "loss": 0.1293,
      "step": 7055
    },
    {
      "epoch": 4.767826968570463,
      "grad_norm": 4.132156848907471,
      "learning_rate": 2e-05,
      "loss": 0.1315,
      "step": 7056
    },
    {
      "epoch": 4.768502872592092,
      "grad_norm": 1.9661953449249268,
      "learning_rate": 2e-05,
      "loss": 0.143,
      "step": 7057
    },
    {
      "epoch": 4.769178776613721,
      "grad_norm": 3.325778007507324,
      "learning_rate": 2e-05,
      "loss": 0.143,
      "step": 7058
    },
    {
      "epoch": 4.7698546806353495,
      "grad_norm": 7.710371971130371,
      "learning_rate": 2e-05,
      "loss": 0.1944,
      "step": 7059
    },
    {
      "epoch": 4.770530584656979,
      "grad_norm": 8.3777494430542,
      "learning_rate": 2e-05,
      "loss": 0.1859,
      "step": 7060
    },
    {
      "epoch": 4.771206488678608,
      "grad_norm": 2.4988021850585938,
      "learning_rate": 2e-05,
      "loss": 0.1246,
      "step": 7061
    },
    {
      "epoch": 4.7718823927002365,
      "grad_norm": 7.142056465148926,
      "learning_rate": 2e-05,
      "loss": 0.1497,
      "step": 7062
    },
    {
      "epoch": 4.772558296721866,
      "grad_norm": 4.42257833480835,
      "learning_rate": 2e-05,
      "loss": 0.2003,
      "step": 7063
    },
    {
      "epoch": 4.773234200743494,
      "grad_norm": 5.545375347137451,
      "learning_rate": 2e-05,
      "loss": 0.2063,
      "step": 7064
    },
    {
      "epoch": 4.773910104765124,
      "grad_norm": 2.4677727222442627,
      "learning_rate": 2e-05,
      "loss": 0.112,
      "step": 7065
    },
    {
      "epoch": 4.774586008786752,
      "grad_norm": 6.1584601402282715,
      "learning_rate": 2e-05,
      "loss": 0.2379,
      "step": 7066
    },
    {
      "epoch": 4.775261912808381,
      "grad_norm": 4.353444576263428,
      "learning_rate": 2e-05,
      "loss": 0.1553,
      "step": 7067
    },
    {
      "epoch": 4.775937816830011,
      "grad_norm": 5.036253452301025,
      "learning_rate": 2e-05,
      "loss": 0.1824,
      "step": 7068
    },
    {
      "epoch": 4.776613720851639,
      "grad_norm": 5.2259297370910645,
      "learning_rate": 2e-05,
      "loss": 0.1632,
      "step": 7069
    },
    {
      "epoch": 4.777289624873268,
      "grad_norm": 2.4645681381225586,
      "learning_rate": 2e-05,
      "loss": 0.1799,
      "step": 7070
    },
    {
      "epoch": 4.777965528894897,
      "grad_norm": 3.021923065185547,
      "learning_rate": 2e-05,
      "loss": 0.1511,
      "step": 7071
    },
    {
      "epoch": 4.778641432916526,
      "grad_norm": 2.7660439014434814,
      "learning_rate": 2e-05,
      "loss": 0.2112,
      "step": 7072
    },
    {
      "epoch": 4.7793173369381545,
      "grad_norm": 3.3660290241241455,
      "learning_rate": 2e-05,
      "loss": 0.2098,
      "step": 7073
    },
    {
      "epoch": 4.779993240959784,
      "grad_norm": 8.690129280090332,
      "learning_rate": 2e-05,
      "loss": 0.1961,
      "step": 7074
    },
    {
      "epoch": 4.780669144981412,
      "grad_norm": 5.655040264129639,
      "learning_rate": 2e-05,
      "loss": 0.1362,
      "step": 7075
    },
    {
      "epoch": 4.781345049003042,
      "grad_norm": 2.8664534091949463,
      "learning_rate": 2e-05,
      "loss": 0.1272,
      "step": 7076
    },
    {
      "epoch": 4.78202095302467,
      "grad_norm": 5.73947286605835,
      "learning_rate": 2e-05,
      "loss": 0.2146,
      "step": 7077
    },
    {
      "epoch": 4.782696857046299,
      "grad_norm": 4.470032691955566,
      "learning_rate": 2e-05,
      "loss": 0.2037,
      "step": 7078
    },
    {
      "epoch": 4.783372761067929,
      "grad_norm": 7.841040134429932,
      "learning_rate": 2e-05,
      "loss": 0.1704,
      "step": 7079
    },
    {
      "epoch": 4.784048665089557,
      "grad_norm": 1.8764617443084717,
      "learning_rate": 2e-05,
      "loss": 0.1194,
      "step": 7080
    },
    {
      "epoch": 4.784724569111186,
      "grad_norm": 1.8565666675567627,
      "learning_rate": 2e-05,
      "loss": 0.1388,
      "step": 7081
    },
    {
      "epoch": 4.785400473132815,
      "grad_norm": 2.720381259918213,
      "learning_rate": 2e-05,
      "loss": 0.2103,
      "step": 7082
    },
    {
      "epoch": 4.786076377154444,
      "grad_norm": 2.3938493728637695,
      "learning_rate": 2e-05,
      "loss": 0.1416,
      "step": 7083
    },
    {
      "epoch": 4.7867522811760725,
      "grad_norm": 2.105276107788086,
      "learning_rate": 2e-05,
      "loss": 0.1761,
      "step": 7084
    },
    {
      "epoch": 4.787428185197702,
      "grad_norm": 5.180670261383057,
      "learning_rate": 2e-05,
      "loss": 0.1726,
      "step": 7085
    },
    {
      "epoch": 4.788104089219331,
      "grad_norm": 3.0642333030700684,
      "learning_rate": 2e-05,
      "loss": 0.1205,
      "step": 7086
    },
    {
      "epoch": 4.78877999324096,
      "grad_norm": 5.26179838180542,
      "learning_rate": 2e-05,
      "loss": 0.1206,
      "step": 7087
    },
    {
      "epoch": 4.789455897262589,
      "grad_norm": 6.406681060791016,
      "learning_rate": 2e-05,
      "loss": 0.181,
      "step": 7088
    },
    {
      "epoch": 4.790131801284217,
      "grad_norm": 5.854868412017822,
      "learning_rate": 2e-05,
      "loss": 0.1394,
      "step": 7089
    },
    {
      "epoch": 4.790807705305847,
      "grad_norm": 3.8926186561584473,
      "learning_rate": 2e-05,
      "loss": 0.131,
      "step": 7090
    },
    {
      "epoch": 4.791483609327475,
      "grad_norm": 5.44577693939209,
      "learning_rate": 2e-05,
      "loss": 0.1831,
      "step": 7091
    },
    {
      "epoch": 4.792159513349104,
      "grad_norm": 2.3476617336273193,
      "learning_rate": 2e-05,
      "loss": 0.1447,
      "step": 7092
    },
    {
      "epoch": 4.792835417370734,
      "grad_norm": 4.195927143096924,
      "learning_rate": 2e-05,
      "loss": 0.1101,
      "step": 7093
    },
    {
      "epoch": 4.793511321392362,
      "grad_norm": 5.013359069824219,
      "learning_rate": 2e-05,
      "loss": 0.154,
      "step": 7094
    },
    {
      "epoch": 4.7941872254139914,
      "grad_norm": 3.089158296585083,
      "learning_rate": 2e-05,
      "loss": 0.1625,
      "step": 7095
    },
    {
      "epoch": 4.79486312943562,
      "grad_norm": 4.529845237731934,
      "learning_rate": 2e-05,
      "loss": 0.1733,
      "step": 7096
    },
    {
      "epoch": 4.795539033457249,
      "grad_norm": 2.857215166091919,
      "learning_rate": 2e-05,
      "loss": 0.1704,
      "step": 7097
    },
    {
      "epoch": 4.796214937478878,
      "grad_norm": 4.198948860168457,
      "learning_rate": 2e-05,
      "loss": 0.1383,
      "step": 7098
    },
    {
      "epoch": 4.796890841500507,
      "grad_norm": 4.2696404457092285,
      "learning_rate": 2e-05,
      "loss": 0.147,
      "step": 7099
    },
    {
      "epoch": 4.797566745522136,
      "grad_norm": 2.432346820831299,
      "learning_rate": 2e-05,
      "loss": 0.1997,
      "step": 7100
    },
    {
      "epoch": 4.797566745522136,
      "eval_accuracy": 0.7203125,
      "eval_loss": 0.2392282485961914,
      "eval_runtime": 47.4256,
      "eval_samples_per_second": 499.309,
      "eval_steps_per_second": 15.603,
      "step": 7100
    },
    {
      "epoch": 4.798242649543765,
      "grad_norm": 5.820883274078369,
      "learning_rate": 2e-05,
      "loss": 0.1388,
      "step": 7101
    },
    {
      "epoch": 4.798918553565394,
      "grad_norm": 5.848119258880615,
      "learning_rate": 2e-05,
      "loss": 0.3047,
      "step": 7102
    },
    {
      "epoch": 4.799594457587022,
      "grad_norm": 2.4915692806243896,
      "learning_rate": 2e-05,
      "loss": 0.1908,
      "step": 7103
    },
    {
      "epoch": 4.800270361608652,
      "grad_norm": 2.499473810195923,
      "learning_rate": 2e-05,
      "loss": 0.1909,
      "step": 7104
    },
    {
      "epoch": 4.80094626563028,
      "grad_norm": 4.394671440124512,
      "learning_rate": 2e-05,
      "loss": 0.1928,
      "step": 7105
    },
    {
      "epoch": 4.8016221696519095,
      "grad_norm": 8.073238372802734,
      "learning_rate": 2e-05,
      "loss": 0.1778,
      "step": 7106
    },
    {
      "epoch": 4.802298073673539,
      "grad_norm": 1.359328031539917,
      "learning_rate": 2e-05,
      "loss": 0.0979,
      "step": 7107
    },
    {
      "epoch": 4.802973977695167,
      "grad_norm": 3.098660469055176,
      "learning_rate": 2e-05,
      "loss": 0.1963,
      "step": 7108
    },
    {
      "epoch": 4.8036498817167965,
      "grad_norm": 2.236111879348755,
      "learning_rate": 2e-05,
      "loss": 0.2164,
      "step": 7109
    },
    {
      "epoch": 4.804325785738425,
      "grad_norm": 2.497978448867798,
      "learning_rate": 2e-05,
      "loss": 0.2137,
      "step": 7110
    },
    {
      "epoch": 4.805001689760054,
      "grad_norm": 3.887554407119751,
      "learning_rate": 2e-05,
      "loss": 0.1831,
      "step": 7111
    },
    {
      "epoch": 4.805677593781683,
      "grad_norm": 5.904563903808594,
      "learning_rate": 2e-05,
      "loss": 0.1816,
      "step": 7112
    },
    {
      "epoch": 4.806353497803312,
      "grad_norm": 5.247188568115234,
      "learning_rate": 2e-05,
      "loss": 0.1676,
      "step": 7113
    },
    {
      "epoch": 4.807029401824941,
      "grad_norm": 2.274181365966797,
      "learning_rate": 2e-05,
      "loss": 0.1631,
      "step": 7114
    },
    {
      "epoch": 4.80770530584657,
      "grad_norm": 2.3655948638916016,
      "learning_rate": 2e-05,
      "loss": 0.1759,
      "step": 7115
    },
    {
      "epoch": 4.808381209868199,
      "grad_norm": 4.675170421600342,
      "learning_rate": 2e-05,
      "loss": 0.1414,
      "step": 7116
    },
    {
      "epoch": 4.8090571138898275,
      "grad_norm": 2.264516592025757,
      "learning_rate": 2e-05,
      "loss": 0.1973,
      "step": 7117
    },
    {
      "epoch": 4.809733017911457,
      "grad_norm": 4.0637922286987305,
      "learning_rate": 2e-05,
      "loss": 0.1419,
      "step": 7118
    },
    {
      "epoch": 4.810408921933085,
      "grad_norm": 2.0107333660125732,
      "learning_rate": 2e-05,
      "loss": 0.1534,
      "step": 7119
    },
    {
      "epoch": 4.8110848259547145,
      "grad_norm": 5.411266326904297,
      "learning_rate": 2e-05,
      "loss": 0.2113,
      "step": 7120
    },
    {
      "epoch": 4.811760729976344,
      "grad_norm": 2.185458183288574,
      "learning_rate": 2e-05,
      "loss": 0.1412,
      "step": 7121
    },
    {
      "epoch": 4.812436633997972,
      "grad_norm": 1.7374682426452637,
      "learning_rate": 2e-05,
      "loss": 0.1646,
      "step": 7122
    },
    {
      "epoch": 4.813112538019602,
      "grad_norm": 2.2719149589538574,
      "learning_rate": 2e-05,
      "loss": 0.1145,
      "step": 7123
    },
    {
      "epoch": 4.81378844204123,
      "grad_norm": 3.399148464202881,
      "learning_rate": 2e-05,
      "loss": 0.1842,
      "step": 7124
    },
    {
      "epoch": 4.814464346062859,
      "grad_norm": 2.3772225379943848,
      "learning_rate": 2e-05,
      "loss": 0.1716,
      "step": 7125
    },
    {
      "epoch": 4.815140250084488,
      "grad_norm": 2.2047195434570312,
      "learning_rate": 2e-05,
      "loss": 0.1308,
      "step": 7126
    },
    {
      "epoch": 4.815816154106117,
      "grad_norm": 1.570086121559143,
      "learning_rate": 2e-05,
      "loss": 0.1256,
      "step": 7127
    },
    {
      "epoch": 4.8164920581277455,
      "grad_norm": 3.183957815170288,
      "learning_rate": 2e-05,
      "loss": 0.1674,
      "step": 7128
    },
    {
      "epoch": 4.817167962149375,
      "grad_norm": 2.4626684188842773,
      "learning_rate": 2e-05,
      "loss": 0.1907,
      "step": 7129
    },
    {
      "epoch": 4.817843866171003,
      "grad_norm": 2.856490135192871,
      "learning_rate": 2e-05,
      "loss": 0.2203,
      "step": 7130
    },
    {
      "epoch": 4.818519770192633,
      "grad_norm": 5.629586696624756,
      "learning_rate": 2e-05,
      "loss": 0.1694,
      "step": 7131
    },
    {
      "epoch": 4.819195674214262,
      "grad_norm": 1.6946123838424683,
      "learning_rate": 2e-05,
      "loss": 0.1204,
      "step": 7132
    },
    {
      "epoch": 4.81987157823589,
      "grad_norm": 2.7945196628570557,
      "learning_rate": 2e-05,
      "loss": 0.1634,
      "step": 7133
    },
    {
      "epoch": 4.82054748225752,
      "grad_norm": 3.862372636795044,
      "learning_rate": 2e-05,
      "loss": 0.187,
      "step": 7134
    },
    {
      "epoch": 4.821223386279148,
      "grad_norm": 1.6351957321166992,
      "learning_rate": 2e-05,
      "loss": 0.1425,
      "step": 7135
    },
    {
      "epoch": 4.821899290300777,
      "grad_norm": 1.8227486610412598,
      "learning_rate": 2e-05,
      "loss": 0.1417,
      "step": 7136
    },
    {
      "epoch": 4.822575194322406,
      "grad_norm": 7.498658657073975,
      "learning_rate": 2e-05,
      "loss": 0.1648,
      "step": 7137
    },
    {
      "epoch": 4.823251098344035,
      "grad_norm": 7.315643787384033,
      "learning_rate": 2e-05,
      "loss": 0.2398,
      "step": 7138
    },
    {
      "epoch": 4.823927002365664,
      "grad_norm": 6.091081619262695,
      "learning_rate": 2e-05,
      "loss": 0.1234,
      "step": 7139
    },
    {
      "epoch": 4.824602906387293,
      "grad_norm": 3.6426706314086914,
      "learning_rate": 2e-05,
      "loss": 0.1283,
      "step": 7140
    },
    {
      "epoch": 4.825278810408922,
      "grad_norm": 1.7953505516052246,
      "learning_rate": 2e-05,
      "loss": 0.1444,
      "step": 7141
    },
    {
      "epoch": 4.825954714430551,
      "grad_norm": 3.453782081604004,
      "learning_rate": 2e-05,
      "loss": 0.1768,
      "step": 7142
    },
    {
      "epoch": 4.82663061845218,
      "grad_norm": 4.234476566314697,
      "learning_rate": 2e-05,
      "loss": 0.12,
      "step": 7143
    },
    {
      "epoch": 4.827306522473808,
      "grad_norm": 2.3982226848602295,
      "learning_rate": 2e-05,
      "loss": 0.1602,
      "step": 7144
    },
    {
      "epoch": 4.827982426495438,
      "grad_norm": 3.5013046264648438,
      "learning_rate": 2e-05,
      "loss": 0.1844,
      "step": 7145
    },
    {
      "epoch": 4.828658330517067,
      "grad_norm": 6.04900598526001,
      "learning_rate": 2e-05,
      "loss": 0.2574,
      "step": 7146
    },
    {
      "epoch": 4.829334234538695,
      "grad_norm": 3.319005250930786,
      "learning_rate": 2e-05,
      "loss": 0.1808,
      "step": 7147
    },
    {
      "epoch": 4.830010138560325,
      "grad_norm": 1.8871773481369019,
      "learning_rate": 2e-05,
      "loss": 0.1424,
      "step": 7148
    },
    {
      "epoch": 4.830686042581953,
      "grad_norm": 1.7801928520202637,
      "learning_rate": 2e-05,
      "loss": 0.1583,
      "step": 7149
    },
    {
      "epoch": 4.831361946603582,
      "grad_norm": 1.6453841924667358,
      "learning_rate": 2e-05,
      "loss": 0.1566,
      "step": 7150
    },
    {
      "epoch": 4.831361946603582,
      "eval_accuracy": 0.7010557432432433,
      "eval_loss": 0.23441343009471893,
      "eval_runtime": 48.8511,
      "eval_samples_per_second": 484.738,
      "eval_steps_per_second": 15.148,
      "step": 7150
    },
    {
      "epoch": 4.832037850625211,
      "grad_norm": 3.19771146774292,
      "learning_rate": 2e-05,
      "loss": 0.1865,
      "step": 7151
    },
    {
      "epoch": 4.83271375464684,
      "grad_norm": 2.2757062911987305,
      "learning_rate": 2e-05,
      "loss": 0.1491,
      "step": 7152
    },
    {
      "epoch": 4.8333896586684695,
      "grad_norm": 5.632377624511719,
      "learning_rate": 2e-05,
      "loss": 0.1623,
      "step": 7153
    },
    {
      "epoch": 4.834065562690098,
      "grad_norm": 3.487635612487793,
      "learning_rate": 2e-05,
      "loss": 0.1299,
      "step": 7154
    },
    {
      "epoch": 4.834741466711727,
      "grad_norm": 1.8878395557403564,
      "learning_rate": 2e-05,
      "loss": 0.1379,
      "step": 7155
    },
    {
      "epoch": 4.835417370733356,
      "grad_norm": 2.6605682373046875,
      "learning_rate": 2e-05,
      "loss": 0.1398,
      "step": 7156
    },
    {
      "epoch": 4.836093274754985,
      "grad_norm": 1.9836010932922363,
      "learning_rate": 2e-05,
      "loss": 0.106,
      "step": 7157
    },
    {
      "epoch": 4.836769178776613,
      "grad_norm": 2.896219253540039,
      "learning_rate": 2e-05,
      "loss": 0.1397,
      "step": 7158
    },
    {
      "epoch": 4.837445082798243,
      "grad_norm": 3.7973906993865967,
      "learning_rate": 2e-05,
      "loss": 0.1458,
      "step": 7159
    },
    {
      "epoch": 4.838120986819872,
      "grad_norm": 5.490700721740723,
      "learning_rate": 2e-05,
      "loss": 0.1821,
      "step": 7160
    },
    {
      "epoch": 4.8387968908415,
      "grad_norm": 2.6688690185546875,
      "learning_rate": 2e-05,
      "loss": 0.1447,
      "step": 7161
    },
    {
      "epoch": 4.83947279486313,
      "grad_norm": 4.225487232208252,
      "learning_rate": 2e-05,
      "loss": 0.1239,
      "step": 7162
    },
    {
      "epoch": 4.840148698884758,
      "grad_norm": 2.7802653312683105,
      "learning_rate": 2e-05,
      "loss": 0.1968,
      "step": 7163
    },
    {
      "epoch": 4.8408246029063875,
      "grad_norm": 1.4072669744491577,
      "learning_rate": 2e-05,
      "loss": 0.1161,
      "step": 7164
    },
    {
      "epoch": 4.841500506928016,
      "grad_norm": 1.6032426357269287,
      "learning_rate": 2e-05,
      "loss": 0.1087,
      "step": 7165
    },
    {
      "epoch": 4.842176410949645,
      "grad_norm": 3.0095787048339844,
      "learning_rate": 2e-05,
      "loss": 0.1038,
      "step": 7166
    },
    {
      "epoch": 4.8428523149712746,
      "grad_norm": 1.7921334505081177,
      "learning_rate": 2e-05,
      "loss": 0.1051,
      "step": 7167
    },
    {
      "epoch": 4.843528218992903,
      "grad_norm": 5.244153022766113,
      "learning_rate": 2e-05,
      "loss": 0.175,
      "step": 7168
    },
    {
      "epoch": 4.844204123014532,
      "grad_norm": 2.8771064281463623,
      "learning_rate": 2e-05,
      "loss": 0.1517,
      "step": 7169
    },
    {
      "epoch": 4.844880027036161,
      "grad_norm": 6.779643535614014,
      "learning_rate": 2e-05,
      "loss": 0.2157,
      "step": 7170
    },
    {
      "epoch": 4.84555593105779,
      "grad_norm": 3.8796274662017822,
      "learning_rate": 2e-05,
      "loss": 0.2127,
      "step": 7171
    },
    {
      "epoch": 4.8462318350794185,
      "grad_norm": 8.310722351074219,
      "learning_rate": 2e-05,
      "loss": 0.141,
      "step": 7172
    },
    {
      "epoch": 4.846907739101048,
      "grad_norm": 3.533329486846924,
      "learning_rate": 2e-05,
      "loss": 0.2258,
      "step": 7173
    },
    {
      "epoch": 4.847583643122676,
      "grad_norm": 3.321589469909668,
      "learning_rate": 2e-05,
      "loss": 0.1317,
      "step": 7174
    },
    {
      "epoch": 4.8482595471443055,
      "grad_norm": 3.1678600311279297,
      "learning_rate": 2e-05,
      "loss": 0.1661,
      "step": 7175
    },
    {
      "epoch": 4.848935451165934,
      "grad_norm": 1.7493551969528198,
      "learning_rate": 2e-05,
      "loss": 0.1151,
      "step": 7176
    },
    {
      "epoch": 4.849611355187563,
      "grad_norm": 5.2323408126831055,
      "learning_rate": 2e-05,
      "loss": 0.179,
      "step": 7177
    },
    {
      "epoch": 4.850287259209193,
      "grad_norm": 2.9670794010162354,
      "learning_rate": 2e-05,
      "loss": 0.1576,
      "step": 7178
    },
    {
      "epoch": 4.850963163230821,
      "grad_norm": 3.199995517730713,
      "learning_rate": 2e-05,
      "loss": 0.1451,
      "step": 7179
    },
    {
      "epoch": 4.85163906725245,
      "grad_norm": 1.4717271327972412,
      "learning_rate": 2e-05,
      "loss": 0.1274,
      "step": 7180
    },
    {
      "epoch": 4.852314971274079,
      "grad_norm": 1.877245545387268,
      "learning_rate": 2e-05,
      "loss": 0.1759,
      "step": 7181
    },
    {
      "epoch": 4.852990875295708,
      "grad_norm": 2.704585552215576,
      "learning_rate": 2e-05,
      "loss": 0.177,
      "step": 7182
    },
    {
      "epoch": 4.8536667793173365,
      "grad_norm": 2.2283430099487305,
      "learning_rate": 2e-05,
      "loss": 0.1819,
      "step": 7183
    },
    {
      "epoch": 4.854342683338966,
      "grad_norm": 3.443814992904663,
      "learning_rate": 2e-05,
      "loss": 0.1299,
      "step": 7184
    },
    {
      "epoch": 4.855018587360595,
      "grad_norm": 2.226334810256958,
      "learning_rate": 2e-05,
      "loss": 0.147,
      "step": 7185
    },
    {
      "epoch": 4.8556944913822235,
      "grad_norm": 1.2692893743515015,
      "learning_rate": 2e-05,
      "loss": 0.1053,
      "step": 7186
    },
    {
      "epoch": 4.856370395403853,
      "grad_norm": 1.9186853170394897,
      "learning_rate": 2e-05,
      "loss": 0.1495,
      "step": 7187
    },
    {
      "epoch": 4.857046299425481,
      "grad_norm": 3.988396644592285,
      "learning_rate": 2e-05,
      "loss": 0.136,
      "step": 7188
    },
    {
      "epoch": 4.857722203447111,
      "grad_norm": 2.42484712600708,
      "learning_rate": 2e-05,
      "loss": 0.1357,
      "step": 7189
    },
    {
      "epoch": 4.858398107468739,
      "grad_norm": 5.056849956512451,
      "learning_rate": 2e-05,
      "loss": 0.1633,
      "step": 7190
    },
    {
      "epoch": 4.859074011490368,
      "grad_norm": 4.01808500289917,
      "learning_rate": 2e-05,
      "loss": 0.1464,
      "step": 7191
    },
    {
      "epoch": 4.859749915511998,
      "grad_norm": 7.8870415687561035,
      "learning_rate": 2e-05,
      "loss": 0.2379,
      "step": 7192
    },
    {
      "epoch": 4.860425819533626,
      "grad_norm": 5.418152332305908,
      "learning_rate": 2e-05,
      "loss": 0.1176,
      "step": 7193
    },
    {
      "epoch": 4.861101723555255,
      "grad_norm": 6.221909999847412,
      "learning_rate": 2e-05,
      "loss": 0.19,
      "step": 7194
    },
    {
      "epoch": 4.861777627576884,
      "grad_norm": 4.573555946350098,
      "learning_rate": 2e-05,
      "loss": 0.1828,
      "step": 7195
    },
    {
      "epoch": 4.862453531598513,
      "grad_norm": 2.3850653171539307,
      "learning_rate": 2e-05,
      "loss": 0.1806,
      "step": 7196
    },
    {
      "epoch": 4.8631294356201415,
      "grad_norm": 2.4987051486968994,
      "learning_rate": 2e-05,
      "loss": 0.1276,
      "step": 7197
    },
    {
      "epoch": 4.863805339641771,
      "grad_norm": 5.879251956939697,
      "learning_rate": 2e-05,
      "loss": 0.1448,
      "step": 7198
    },
    {
      "epoch": 4.8644812436634,
      "grad_norm": 4.970425128936768,
      "learning_rate": 2e-05,
      "loss": 0.1981,
      "step": 7199
    },
    {
      "epoch": 4.865157147685029,
      "grad_norm": 5.387129306793213,
      "learning_rate": 2e-05,
      "loss": 0.2077,
      "step": 7200
    },
    {
      "epoch": 4.865157147685029,
      "eval_accuracy": 0.737964527027027,
      "eval_loss": 0.24116544425487518,
      "eval_runtime": 51.8136,
      "eval_samples_per_second": 457.023,
      "eval_steps_per_second": 14.282,
      "step": 7200
    },
    {
      "epoch": 4.865833051706658,
      "grad_norm": 2.638207197189331,
      "learning_rate": 2e-05,
      "loss": 0.1518,
      "step": 7201
    },
    {
      "epoch": 4.866508955728286,
      "grad_norm": 2.8460609912872314,
      "learning_rate": 2e-05,
      "loss": 0.14,
      "step": 7202
    },
    {
      "epoch": 4.867184859749916,
      "grad_norm": 6.6611833572387695,
      "learning_rate": 2e-05,
      "loss": 0.1582,
      "step": 7203
    },
    {
      "epoch": 4.867860763771544,
      "grad_norm": 2.1069564819335938,
      "learning_rate": 2e-05,
      "loss": 0.1456,
      "step": 7204
    },
    {
      "epoch": 4.868536667793173,
      "grad_norm": 2.0327768325805664,
      "learning_rate": 2e-05,
      "loss": 0.1732,
      "step": 7205
    },
    {
      "epoch": 4.869212571814803,
      "grad_norm": 1.7992037534713745,
      "learning_rate": 2e-05,
      "loss": 0.1587,
      "step": 7206
    },
    {
      "epoch": 4.869888475836431,
      "grad_norm": 4.234808444976807,
      "learning_rate": 2e-05,
      "loss": 0.1948,
      "step": 7207
    },
    {
      "epoch": 4.8705643798580605,
      "grad_norm": 2.0658528804779053,
      "learning_rate": 2e-05,
      "loss": 0.1458,
      "step": 7208
    },
    {
      "epoch": 4.871240283879689,
      "grad_norm": 2.333954334259033,
      "learning_rate": 2e-05,
      "loss": 0.1391,
      "step": 7209
    },
    {
      "epoch": 4.871916187901318,
      "grad_norm": 5.572136878967285,
      "learning_rate": 2e-05,
      "loss": 0.1723,
      "step": 7210
    },
    {
      "epoch": 4.872592091922947,
      "grad_norm": 4.074207305908203,
      "learning_rate": 2e-05,
      "loss": 0.123,
      "step": 7211
    },
    {
      "epoch": 4.873267995944576,
      "grad_norm": 3.1010258197784424,
      "learning_rate": 2e-05,
      "loss": 0.2078,
      "step": 7212
    },
    {
      "epoch": 4.873943899966205,
      "grad_norm": 7.404003143310547,
      "learning_rate": 2e-05,
      "loss": 0.1603,
      "step": 7213
    },
    {
      "epoch": 4.874619803987834,
      "grad_norm": 3.6353847980499268,
      "learning_rate": 2e-05,
      "loss": 0.1277,
      "step": 7214
    },
    {
      "epoch": 4.875295708009463,
      "grad_norm": 1.515444040298462,
      "learning_rate": 2e-05,
      "loss": 0.1065,
      "step": 7215
    },
    {
      "epoch": 4.875971612031091,
      "grad_norm": 6.275171279907227,
      "learning_rate": 2e-05,
      "loss": 0.1986,
      "step": 7216
    },
    {
      "epoch": 4.876647516052721,
      "grad_norm": 2.212820529937744,
      "learning_rate": 2e-05,
      "loss": 0.1404,
      "step": 7217
    },
    {
      "epoch": 4.877323420074349,
      "grad_norm": 7.379702091217041,
      "learning_rate": 2e-05,
      "loss": 0.1402,
      "step": 7218
    },
    {
      "epoch": 4.8779993240959785,
      "grad_norm": 2.15902042388916,
      "learning_rate": 2e-05,
      "loss": 0.1821,
      "step": 7219
    },
    {
      "epoch": 4.878675228117608,
      "grad_norm": 2.0791866779327393,
      "learning_rate": 2e-05,
      "loss": 0.1798,
      "step": 7220
    },
    {
      "epoch": 4.879351132139236,
      "grad_norm": 3.043151378631592,
      "learning_rate": 2e-05,
      "loss": 0.0932,
      "step": 7221
    },
    {
      "epoch": 4.8800270361608655,
      "grad_norm": 5.938875198364258,
      "learning_rate": 2e-05,
      "loss": 0.1285,
      "step": 7222
    },
    {
      "epoch": 4.880702940182494,
      "grad_norm": 6.218297481536865,
      "learning_rate": 2e-05,
      "loss": 0.1859,
      "step": 7223
    },
    {
      "epoch": 4.881378844204123,
      "grad_norm": 3.0474984645843506,
      "learning_rate": 2e-05,
      "loss": 0.1498,
      "step": 7224
    },
    {
      "epoch": 4.882054748225752,
      "grad_norm": 4.09666633605957,
      "learning_rate": 2e-05,
      "loss": 0.1473,
      "step": 7225
    },
    {
      "epoch": 4.882730652247381,
      "grad_norm": 2.6711831092834473,
      "learning_rate": 2e-05,
      "loss": 0.1708,
      "step": 7226
    },
    {
      "epoch": 4.883406556269009,
      "grad_norm": 5.892667293548584,
      "learning_rate": 2e-05,
      "loss": 0.2111,
      "step": 7227
    },
    {
      "epoch": 4.884082460290639,
      "grad_norm": 7.069108963012695,
      "learning_rate": 2e-05,
      "loss": 0.1572,
      "step": 7228
    },
    {
      "epoch": 4.884758364312267,
      "grad_norm": 4.4032769203186035,
      "learning_rate": 2e-05,
      "loss": 0.1751,
      "step": 7229
    },
    {
      "epoch": 4.8854342683338965,
      "grad_norm": 4.964427471160889,
      "learning_rate": 2e-05,
      "loss": 0.1426,
      "step": 7230
    },
    {
      "epoch": 4.886110172355526,
      "grad_norm": 1.728680968284607,
      "learning_rate": 2e-05,
      "loss": 0.1292,
      "step": 7231
    },
    {
      "epoch": 4.886786076377154,
      "grad_norm": 3.7069718837738037,
      "learning_rate": 2e-05,
      "loss": 0.1725,
      "step": 7232
    },
    {
      "epoch": 4.8874619803987835,
      "grad_norm": 4.474459648132324,
      "learning_rate": 2e-05,
      "loss": 0.1783,
      "step": 7233
    },
    {
      "epoch": 4.888137884420412,
      "grad_norm": 2.8219447135925293,
      "learning_rate": 2e-05,
      "loss": 0.09,
      "step": 7234
    },
    {
      "epoch": 4.888813788442041,
      "grad_norm": 1.661925196647644,
      "learning_rate": 2e-05,
      "loss": 0.1727,
      "step": 7235
    },
    {
      "epoch": 4.88948969246367,
      "grad_norm": 3.229573965072632,
      "learning_rate": 2e-05,
      "loss": 0.122,
      "step": 7236
    },
    {
      "epoch": 4.890165596485299,
      "grad_norm": 3.8113462924957275,
      "learning_rate": 2e-05,
      "loss": 0.2158,
      "step": 7237
    },
    {
      "epoch": 4.890841500506928,
      "grad_norm": 2.2738873958587646,
      "learning_rate": 2e-05,
      "loss": 0.1725,
      "step": 7238
    },
    {
      "epoch": 4.891517404528557,
      "grad_norm": 4.307225704193115,
      "learning_rate": 2e-05,
      "loss": 0.1549,
      "step": 7239
    },
    {
      "epoch": 4.892193308550186,
      "grad_norm": 2.102233409881592,
      "learning_rate": 2e-05,
      "loss": 0.1261,
      "step": 7240
    },
    {
      "epoch": 4.8928692125718145,
      "grad_norm": 2.361741065979004,
      "learning_rate": 2e-05,
      "loss": 0.1953,
      "step": 7241
    },
    {
      "epoch": 4.893545116593444,
      "grad_norm": 1.9440882205963135,
      "learning_rate": 2e-05,
      "loss": 0.1566,
      "step": 7242
    },
    {
      "epoch": 4.894221020615072,
      "grad_norm": 7.094907283782959,
      "learning_rate": 2e-05,
      "loss": 0.1963,
      "step": 7243
    },
    {
      "epoch": 4.894896924636702,
      "grad_norm": 4.796937465667725,
      "learning_rate": 2e-05,
      "loss": 0.2228,
      "step": 7244
    },
    {
      "epoch": 4.895572828658331,
      "grad_norm": 2.519773244857788,
      "learning_rate": 2e-05,
      "loss": 0.2054,
      "step": 7245
    },
    {
      "epoch": 4.896248732679959,
      "grad_norm": 4.511626243591309,
      "learning_rate": 2e-05,
      "loss": 0.187,
      "step": 7246
    },
    {
      "epoch": 4.896924636701589,
      "grad_norm": 1.854715347290039,
      "learning_rate": 2e-05,
      "loss": 0.1505,
      "step": 7247
    },
    {
      "epoch": 4.897600540723217,
      "grad_norm": 4.071605205535889,
      "learning_rate": 2e-05,
      "loss": 0.1488,
      "step": 7248
    },
    {
      "epoch": 4.898276444744846,
      "grad_norm": 2.4968981742858887,
      "learning_rate": 2e-05,
      "loss": 0.1346,
      "step": 7249
    },
    {
      "epoch": 4.898952348766475,
      "grad_norm": 2.875994920730591,
      "learning_rate": 2e-05,
      "loss": 0.1796,
      "step": 7250
    },
    {
      "epoch": 4.898952348766475,
      "eval_accuracy": 0.7802787162162163,
      "eval_loss": 0.23581118881702423,
      "eval_runtime": 48.5936,
      "eval_samples_per_second": 487.307,
      "eval_steps_per_second": 15.228,
      "step": 7250
    },
    {
      "epoch": 4.899628252788104,
      "grad_norm": 2.703223943710327,
      "learning_rate": 2e-05,
      "loss": 0.1592,
      "step": 7251
    },
    {
      "epoch": 4.900304156809733,
      "grad_norm": 4.2997660636901855,
      "learning_rate": 2e-05,
      "loss": 0.2007,
      "step": 7252
    },
    {
      "epoch": 4.900980060831362,
      "grad_norm": 2.0527877807617188,
      "learning_rate": 2e-05,
      "loss": 0.1161,
      "step": 7253
    },
    {
      "epoch": 4.901655964852991,
      "grad_norm": 2.4071927070617676,
      "learning_rate": 2e-05,
      "loss": 0.1198,
      "step": 7254
    },
    {
      "epoch": 4.90233186887462,
      "grad_norm": 1.9555562734603882,
      "learning_rate": 2e-05,
      "loss": 0.1489,
      "step": 7255
    },
    {
      "epoch": 4.903007772896249,
      "grad_norm": 4.782604217529297,
      "learning_rate": 2e-05,
      "loss": 0.1547,
      "step": 7256
    },
    {
      "epoch": 4.903683676917877,
      "grad_norm": 4.372509002685547,
      "learning_rate": 2e-05,
      "loss": 0.1432,
      "step": 7257
    },
    {
      "epoch": 4.904359580939507,
      "grad_norm": 7.338765621185303,
      "learning_rate": 2e-05,
      "loss": 0.1917,
      "step": 7258
    },
    {
      "epoch": 4.905035484961136,
      "grad_norm": 3.368175745010376,
      "learning_rate": 2e-05,
      "loss": 0.1665,
      "step": 7259
    },
    {
      "epoch": 4.905711388982764,
      "grad_norm": 2.380571126937866,
      "learning_rate": 2e-05,
      "loss": 0.1944,
      "step": 7260
    },
    {
      "epoch": 4.906387293004394,
      "grad_norm": 5.323369026184082,
      "learning_rate": 2e-05,
      "loss": 0.2155,
      "step": 7261
    },
    {
      "epoch": 4.907063197026022,
      "grad_norm": 6.87929105758667,
      "learning_rate": 2e-05,
      "loss": 0.176,
      "step": 7262
    },
    {
      "epoch": 4.907739101047651,
      "grad_norm": 1.946049451828003,
      "learning_rate": 2e-05,
      "loss": 0.1264,
      "step": 7263
    },
    {
      "epoch": 4.90841500506928,
      "grad_norm": 2.544370174407959,
      "learning_rate": 2e-05,
      "loss": 0.2519,
      "step": 7264
    },
    {
      "epoch": 4.909090909090909,
      "grad_norm": 4.589902877807617,
      "learning_rate": 2e-05,
      "loss": 0.1678,
      "step": 7265
    },
    {
      "epoch": 4.9097668131125385,
      "grad_norm": 3.5709073543548584,
      "learning_rate": 2e-05,
      "loss": 0.0943,
      "step": 7266
    },
    {
      "epoch": 4.910442717134167,
      "grad_norm": 2.1278038024902344,
      "learning_rate": 2e-05,
      "loss": 0.1588,
      "step": 7267
    },
    {
      "epoch": 4.911118621155796,
      "grad_norm": 4.653007507324219,
      "learning_rate": 2e-05,
      "loss": 0.17,
      "step": 7268
    },
    {
      "epoch": 4.911794525177425,
      "grad_norm": 19.258394241333008,
      "learning_rate": 2e-05,
      "loss": 0.1141,
      "step": 7269
    },
    {
      "epoch": 4.912470429199054,
      "grad_norm": 5.303653717041016,
      "learning_rate": 2e-05,
      "loss": 0.1556,
      "step": 7270
    },
    {
      "epoch": 4.913146333220682,
      "grad_norm": 1.7933279275894165,
      "learning_rate": 2e-05,
      "loss": 0.122,
      "step": 7271
    },
    {
      "epoch": 4.913822237242312,
      "grad_norm": 3.3211679458618164,
      "learning_rate": 2e-05,
      "loss": 0.097,
      "step": 7272
    },
    {
      "epoch": 4.91449814126394,
      "grad_norm": 1.5910848379135132,
      "learning_rate": 2e-05,
      "loss": 0.1614,
      "step": 7273
    },
    {
      "epoch": 4.9151740452855694,
      "grad_norm": 3.907193422317505,
      "learning_rate": 2e-05,
      "loss": 0.1687,
      "step": 7274
    },
    {
      "epoch": 4.915849949307199,
      "grad_norm": 2.021003246307373,
      "learning_rate": 2e-05,
      "loss": 0.1557,
      "step": 7275
    },
    {
      "epoch": 4.916525853328827,
      "grad_norm": 4.161309242248535,
      "learning_rate": 2e-05,
      "loss": 0.1777,
      "step": 7276
    },
    {
      "epoch": 4.9172017573504565,
      "grad_norm": 1.7565932273864746,
      "learning_rate": 2e-05,
      "loss": 0.1319,
      "step": 7277
    },
    {
      "epoch": 4.917877661372085,
      "grad_norm": 2.8397200107574463,
      "learning_rate": 2e-05,
      "loss": 0.1364,
      "step": 7278
    },
    {
      "epoch": 4.918553565393714,
      "grad_norm": 5.628444671630859,
      "learning_rate": 2e-05,
      "loss": 0.2276,
      "step": 7279
    },
    {
      "epoch": 4.919229469415343,
      "grad_norm": 3.1690549850463867,
      "learning_rate": 2e-05,
      "loss": 0.1981,
      "step": 7280
    },
    {
      "epoch": 4.919905373436972,
      "grad_norm": 6.204672813415527,
      "learning_rate": 2e-05,
      "loss": 0.2363,
      "step": 7281
    },
    {
      "epoch": 4.9205812774586,
      "grad_norm": 3.1102261543273926,
      "learning_rate": 2e-05,
      "loss": 0.2076,
      "step": 7282
    },
    {
      "epoch": 4.92125718148023,
      "grad_norm": 4.139063835144043,
      "learning_rate": 2e-05,
      "loss": 0.1137,
      "step": 7283
    },
    {
      "epoch": 4.921933085501859,
      "grad_norm": 1.7150421142578125,
      "learning_rate": 2e-05,
      "loss": 0.1346,
      "step": 7284
    },
    {
      "epoch": 4.9226089895234875,
      "grad_norm": 2.774620771408081,
      "learning_rate": 2e-05,
      "loss": 0.1913,
      "step": 7285
    },
    {
      "epoch": 4.923284893545117,
      "grad_norm": 1.8826398849487305,
      "learning_rate": 2e-05,
      "loss": 0.1157,
      "step": 7286
    },
    {
      "epoch": 4.923960797566745,
      "grad_norm": 5.5587873458862305,
      "learning_rate": 2e-05,
      "loss": 0.128,
      "step": 7287
    },
    {
      "epoch": 4.9246367015883745,
      "grad_norm": 2.90144681930542,
      "learning_rate": 2e-05,
      "loss": 0.1843,
      "step": 7288
    },
    {
      "epoch": 4.925312605610003,
      "grad_norm": 2.3048486709594727,
      "learning_rate": 2e-05,
      "loss": 0.1154,
      "step": 7289
    },
    {
      "epoch": 4.925988509631632,
      "grad_norm": 3.6832497119903564,
      "learning_rate": 2e-05,
      "loss": 0.1435,
      "step": 7290
    },
    {
      "epoch": 4.926664413653262,
      "grad_norm": 7.270464897155762,
      "learning_rate": 2e-05,
      "loss": 0.1906,
      "step": 7291
    },
    {
      "epoch": 4.92734031767489,
      "grad_norm": 3.095099687576294,
      "learning_rate": 2e-05,
      "loss": 0.1579,
      "step": 7292
    },
    {
      "epoch": 4.928016221696519,
      "grad_norm": 3.068082809448242,
      "learning_rate": 2e-05,
      "loss": 0.1598,
      "step": 7293
    },
    {
      "epoch": 4.928692125718148,
      "grad_norm": 6.157203674316406,
      "learning_rate": 2e-05,
      "loss": 0.1713,
      "step": 7294
    },
    {
      "epoch": 4.929368029739777,
      "grad_norm": 3.8276193141937256,
      "learning_rate": 2e-05,
      "loss": 0.1092,
      "step": 7295
    },
    {
      "epoch": 4.9300439337614055,
      "grad_norm": 5.6388654708862305,
      "learning_rate": 2e-05,
      "loss": 0.2117,
      "step": 7296
    },
    {
      "epoch": 4.930719837783035,
      "grad_norm": 8.629498481750488,
      "learning_rate": 2e-05,
      "loss": 0.1298,
      "step": 7297
    },
    {
      "epoch": 4.931395741804664,
      "grad_norm": 7.71702766418457,
      "learning_rate": 2e-05,
      "loss": 0.1999,
      "step": 7298
    },
    {
      "epoch": 4.9320716458262925,
      "grad_norm": 4.966789245605469,
      "learning_rate": 2e-05,
      "loss": 0.1725,
      "step": 7299
    },
    {
      "epoch": 4.932747549847922,
      "grad_norm": 9.165950775146484,
      "learning_rate": 2e-05,
      "loss": 0.1241,
      "step": 7300
    },
    {
      "epoch": 4.932747549847922,
      "eval_accuracy": 0.737964527027027,
      "eval_loss": 0.271952360868454,
      "eval_runtime": 50.8471,
      "eval_samples_per_second": 465.71,
      "eval_steps_per_second": 14.553,
      "step": 7300
    },
    {
      "epoch": 4.93342345386955,
      "grad_norm": 2.1708462238311768,
      "learning_rate": 2e-05,
      "loss": 0.1492,
      "step": 7301
    },
    {
      "epoch": 4.93409935789118,
      "grad_norm": 4.940573215484619,
      "learning_rate": 2e-05,
      "loss": 0.165,
      "step": 7302
    },
    {
      "epoch": 4.934775261912808,
      "grad_norm": 9.798130989074707,
      "learning_rate": 2e-05,
      "loss": 0.167,
      "step": 7303
    },
    {
      "epoch": 4.935451165934437,
      "grad_norm": 12.221875190734863,
      "learning_rate": 2e-05,
      "loss": 0.194,
      "step": 7304
    },
    {
      "epoch": 4.936127069956067,
      "grad_norm": 10.442641258239746,
      "learning_rate": 2e-05,
      "loss": 0.1576,
      "step": 7305
    },
    {
      "epoch": 4.936802973977695,
      "grad_norm": 3.561617612838745,
      "learning_rate": 2e-05,
      "loss": 0.1032,
      "step": 7306
    },
    {
      "epoch": 4.937478877999324,
      "grad_norm": 4.97029447555542,
      "learning_rate": 2e-05,
      "loss": 0.1575,
      "step": 7307
    },
    {
      "epoch": 4.938154782020953,
      "grad_norm": 9.93654727935791,
      "learning_rate": 2e-05,
      "loss": 0.2104,
      "step": 7308
    },
    {
      "epoch": 4.938830686042582,
      "grad_norm": 8.49761962890625,
      "learning_rate": 2e-05,
      "loss": 0.1823,
      "step": 7309
    },
    {
      "epoch": 4.939506590064211,
      "grad_norm": 9.677934646606445,
      "learning_rate": 2e-05,
      "loss": 0.2505,
      "step": 7310
    },
    {
      "epoch": 4.94018249408584,
      "grad_norm": 4.906760215759277,
      "learning_rate": 2e-05,
      "loss": 0.1717,
      "step": 7311
    },
    {
      "epoch": 4.940858398107469,
      "grad_norm": 2.702286958694458,
      "learning_rate": 2e-05,
      "loss": 0.1583,
      "step": 7312
    },
    {
      "epoch": 4.941534302129098,
      "grad_norm": 5.408538341522217,
      "learning_rate": 2e-05,
      "loss": 0.1736,
      "step": 7313
    },
    {
      "epoch": 4.942210206150727,
      "grad_norm": 4.803462505340576,
      "learning_rate": 2e-05,
      "loss": 0.1473,
      "step": 7314
    },
    {
      "epoch": 4.942886110172355,
      "grad_norm": 5.3275957107543945,
      "learning_rate": 2e-05,
      "loss": 0.1537,
      "step": 7315
    },
    {
      "epoch": 4.943562014193985,
      "grad_norm": 5.596399307250977,
      "learning_rate": 2e-05,
      "loss": 0.1461,
      "step": 7316
    },
    {
      "epoch": 4.944237918215613,
      "grad_norm": 7.493070602416992,
      "learning_rate": 2e-05,
      "loss": 0.1825,
      "step": 7317
    },
    {
      "epoch": 4.944913822237242,
      "grad_norm": 6.647232532501221,
      "learning_rate": 2e-05,
      "loss": 0.1761,
      "step": 7318
    },
    {
      "epoch": 4.945589726258872,
      "grad_norm": 4.2901506423950195,
      "learning_rate": 2e-05,
      "loss": 0.1423,
      "step": 7319
    },
    {
      "epoch": 4.9462656302805,
      "grad_norm": 2.6016910076141357,
      "learning_rate": 2e-05,
      "loss": 0.1587,
      "step": 7320
    },
    {
      "epoch": 4.9469415343021295,
      "grad_norm": 3.396334648132324,
      "learning_rate": 2e-05,
      "loss": 0.1896,
      "step": 7321
    },
    {
      "epoch": 4.947617438323758,
      "grad_norm": 2.104485273361206,
      "learning_rate": 2e-05,
      "loss": 0.1268,
      "step": 7322
    },
    {
      "epoch": 4.948293342345387,
      "grad_norm": 4.743696689605713,
      "learning_rate": 2e-05,
      "loss": 0.1489,
      "step": 7323
    },
    {
      "epoch": 4.948969246367016,
      "grad_norm": 9.326696395874023,
      "learning_rate": 2e-05,
      "loss": 0.2323,
      "step": 7324
    },
    {
      "epoch": 4.949645150388645,
      "grad_norm": 8.48007869720459,
      "learning_rate": 2e-05,
      "loss": 0.1781,
      "step": 7325
    },
    {
      "epoch": 4.950321054410273,
      "grad_norm": 5.14724588394165,
      "learning_rate": 2e-05,
      "loss": 0.2126,
      "step": 7326
    },
    {
      "epoch": 4.950996958431903,
      "grad_norm": 3.763169527053833,
      "learning_rate": 2e-05,
      "loss": 0.1448,
      "step": 7327
    },
    {
      "epoch": 4.951672862453531,
      "grad_norm": 4.504483222961426,
      "learning_rate": 2e-05,
      "loss": 0.2021,
      "step": 7328
    },
    {
      "epoch": 4.95234876647516,
      "grad_norm": 1.9096896648406982,
      "learning_rate": 2e-05,
      "loss": 0.1117,
      "step": 7329
    },
    {
      "epoch": 4.95302467049679,
      "grad_norm": 3.2305185794830322,
      "learning_rate": 2e-05,
      "loss": 0.2448,
      "step": 7330
    },
    {
      "epoch": 4.953700574518418,
      "grad_norm": 1.7409415245056152,
      "learning_rate": 2e-05,
      "loss": 0.1402,
      "step": 7331
    },
    {
      "epoch": 4.9543764785400475,
      "grad_norm": 3.4440102577209473,
      "learning_rate": 2e-05,
      "loss": 0.146,
      "step": 7332
    },
    {
      "epoch": 4.955052382561676,
      "grad_norm": 4.753054618835449,
      "learning_rate": 2e-05,
      "loss": 0.1605,
      "step": 7333
    },
    {
      "epoch": 4.955728286583305,
      "grad_norm": 3.6991360187530518,
      "learning_rate": 2e-05,
      "loss": 0.1649,
      "step": 7334
    },
    {
      "epoch": 4.956404190604934,
      "grad_norm": 3.506728172302246,
      "learning_rate": 2e-05,
      "loss": 0.1537,
      "step": 7335
    },
    {
      "epoch": 4.957080094626563,
      "grad_norm": 9.854984283447266,
      "learning_rate": 2e-05,
      "loss": 0.2498,
      "step": 7336
    },
    {
      "epoch": 4.957755998648192,
      "grad_norm": 2.1439173221588135,
      "learning_rate": 2e-05,
      "loss": 0.1617,
      "step": 7337
    },
    {
      "epoch": 4.958431902669821,
      "grad_norm": 3.401005983352661,
      "learning_rate": 2e-05,
      "loss": 0.1534,
      "step": 7338
    },
    {
      "epoch": 4.95910780669145,
      "grad_norm": 3.537031650543213,
      "learning_rate": 2e-05,
      "loss": 0.1539,
      "step": 7339
    },
    {
      "epoch": 4.959783710713078,
      "grad_norm": 4.8287129402160645,
      "learning_rate": 2e-05,
      "loss": 0.162,
      "step": 7340
    },
    {
      "epoch": 4.960459614734708,
      "grad_norm": 5.278522491455078,
      "learning_rate": 2e-05,
      "loss": 0.1636,
      "step": 7341
    },
    {
      "epoch": 4.961135518756336,
      "grad_norm": 8.437423706054688,
      "learning_rate": 2e-05,
      "loss": 0.1685,
      "step": 7342
    },
    {
      "epoch": 4.9618114227779655,
      "grad_norm": 3.6771955490112305,
      "learning_rate": 2e-05,
      "loss": 0.1845,
      "step": 7343
    },
    {
      "epoch": 4.962487326799595,
      "grad_norm": 3.360666036605835,
      "learning_rate": 2e-05,
      "loss": 0.2356,
      "step": 7344
    },
    {
      "epoch": 4.963163230821223,
      "grad_norm": 3.589380979537964,
      "learning_rate": 2e-05,
      "loss": 0.2001,
      "step": 7345
    },
    {
      "epoch": 4.9638391348428526,
      "grad_norm": 1.9216183423995972,
      "learning_rate": 2e-05,
      "loss": 0.1219,
      "step": 7346
    },
    {
      "epoch": 4.964515038864481,
      "grad_norm": 4.306602954864502,
      "learning_rate": 2e-05,
      "loss": 0.1501,
      "step": 7347
    },
    {
      "epoch": 4.96519094288611,
      "grad_norm": 1.8908621072769165,
      "learning_rate": 2e-05,
      "loss": 0.1489,
      "step": 7348
    },
    {
      "epoch": 4.965866846907739,
      "grad_norm": 2.64565372467041,
      "learning_rate": 2e-05,
      "loss": 0.1213,
      "step": 7349
    },
    {
      "epoch": 4.966542750929368,
      "grad_norm": 5.022712230682373,
      "learning_rate": 2e-05,
      "loss": 0.1417,
      "step": 7350
    },
    {
      "epoch": 4.966542750929368,
      "eval_accuracy": 0.7414273648648648,
      "eval_loss": 0.24143347144126892,
      "eval_runtime": 45.6743,
      "eval_samples_per_second": 518.453,
      "eval_steps_per_second": 16.202,
      "step": 7350
    },
    {
      "epoch": 4.967218654950997,
      "grad_norm": 4.788304805755615,
      "learning_rate": 2e-05,
      "loss": 0.1709,
      "step": 7351
    },
    {
      "epoch": 4.967894558972626,
      "grad_norm": 3.7794342041015625,
      "learning_rate": 2e-05,
      "loss": 0.1258,
      "step": 7352
    },
    {
      "epoch": 4.968570462994255,
      "grad_norm": 6.388289928436279,
      "learning_rate": 2e-05,
      "loss": 0.1151,
      "step": 7353
    },
    {
      "epoch": 4.9692463670158835,
      "grad_norm": 2.360379457473755,
      "learning_rate": 2e-05,
      "loss": 0.2036,
      "step": 7354
    },
    {
      "epoch": 4.969922271037513,
      "grad_norm": 5.111299514770508,
      "learning_rate": 2e-05,
      "loss": 0.1278,
      "step": 7355
    },
    {
      "epoch": 4.970598175059141,
      "grad_norm": 4.192529678344727,
      "learning_rate": 2e-05,
      "loss": 0.2232,
      "step": 7356
    },
    {
      "epoch": 4.971274079080771,
      "grad_norm": 8.086542129516602,
      "learning_rate": 2e-05,
      "loss": 0.2489,
      "step": 7357
    },
    {
      "epoch": 4.9719499831024,
      "grad_norm": 4.19054651260376,
      "learning_rate": 2e-05,
      "loss": 0.2352,
      "step": 7358
    },
    {
      "epoch": 4.972625887124028,
      "grad_norm": 3.3088977336883545,
      "learning_rate": 2e-05,
      "loss": 0.161,
      "step": 7359
    },
    {
      "epoch": 4.973301791145658,
      "grad_norm": 3.9382824897766113,
      "learning_rate": 2e-05,
      "loss": 0.1746,
      "step": 7360
    },
    {
      "epoch": 4.973977695167286,
      "grad_norm": 1.7304474115371704,
      "learning_rate": 2e-05,
      "loss": 0.088,
      "step": 7361
    },
    {
      "epoch": 4.974653599188915,
      "grad_norm": 7.398748874664307,
      "learning_rate": 2e-05,
      "loss": 0.2454,
      "step": 7362
    },
    {
      "epoch": 4.975329503210544,
      "grad_norm": 6.551055908203125,
      "learning_rate": 2e-05,
      "loss": 0.1695,
      "step": 7363
    },
    {
      "epoch": 4.976005407232173,
      "grad_norm": 4.557241439819336,
      "learning_rate": 2e-05,
      "loss": 0.2229,
      "step": 7364
    },
    {
      "epoch": 4.976681311253802,
      "grad_norm": 3.60357403755188,
      "learning_rate": 2e-05,
      "loss": 0.1723,
      "step": 7365
    },
    {
      "epoch": 4.977357215275431,
      "grad_norm": 2.677700996398926,
      "learning_rate": 2e-05,
      "loss": 0.1944,
      "step": 7366
    },
    {
      "epoch": 4.97803311929706,
      "grad_norm": 1.7326856851577759,
      "learning_rate": 2e-05,
      "loss": 0.1895,
      "step": 7367
    },
    {
      "epoch": 4.978709023318689,
      "grad_norm": 4.345920562744141,
      "learning_rate": 2e-05,
      "loss": 0.2138,
      "step": 7368
    },
    {
      "epoch": 4.979384927340318,
      "grad_norm": 3.7954938411712646,
      "learning_rate": 2e-05,
      "loss": 0.1482,
      "step": 7369
    },
    {
      "epoch": 4.980060831361946,
      "grad_norm": 6.224590301513672,
      "learning_rate": 2e-05,
      "loss": 0.1645,
      "step": 7370
    },
    {
      "epoch": 4.980736735383576,
      "grad_norm": 1.6597157716751099,
      "learning_rate": 2e-05,
      "loss": 0.2022,
      "step": 7371
    },
    {
      "epoch": 4.981412639405205,
      "grad_norm": 3.0273830890655518,
      "learning_rate": 2e-05,
      "loss": 0.1887,
      "step": 7372
    },
    {
      "epoch": 4.982088543426833,
      "grad_norm": 3.2334787845611572,
      "learning_rate": 2e-05,
      "loss": 0.1814,
      "step": 7373
    },
    {
      "epoch": 4.982764447448463,
      "grad_norm": 5.83894157409668,
      "learning_rate": 2e-05,
      "loss": 0.1762,
      "step": 7374
    },
    {
      "epoch": 4.983440351470091,
      "grad_norm": 5.201065540313721,
      "learning_rate": 2e-05,
      "loss": 0.1374,
      "step": 7375
    },
    {
      "epoch": 4.98411625549172,
      "grad_norm": 4.799150466918945,
      "learning_rate": 2e-05,
      "loss": 0.2051,
      "step": 7376
    },
    {
      "epoch": 4.984792159513349,
      "grad_norm": 1.9672695398330688,
      "learning_rate": 2e-05,
      "loss": 0.1271,
      "step": 7377
    },
    {
      "epoch": 4.985468063534978,
      "grad_norm": 1.7897030115127563,
      "learning_rate": 2e-05,
      "loss": 0.1284,
      "step": 7378
    },
    {
      "epoch": 4.986143967556607,
      "grad_norm": 2.4189705848693848,
      "learning_rate": 2e-05,
      "loss": 0.1697,
      "step": 7379
    },
    {
      "epoch": 4.986819871578236,
      "grad_norm": 2.700472831726074,
      "learning_rate": 2e-05,
      "loss": 0.1192,
      "step": 7380
    },
    {
      "epoch": 4.987495775599864,
      "grad_norm": 11.542924880981445,
      "learning_rate": 2e-05,
      "loss": 0.2339,
      "step": 7381
    },
    {
      "epoch": 4.988171679621494,
      "grad_norm": 2.163823127746582,
      "learning_rate": 2e-05,
      "loss": 0.1625,
      "step": 7382
    },
    {
      "epoch": 4.988847583643123,
      "grad_norm": 2.8103561401367188,
      "learning_rate": 2e-05,
      "loss": 0.1093,
      "step": 7383
    },
    {
      "epoch": 4.989523487664751,
      "grad_norm": 2.5633111000061035,
      "learning_rate": 2e-05,
      "loss": 0.1668,
      "step": 7384
    },
    {
      "epoch": 4.990199391686381,
      "grad_norm": 9.175567626953125,
      "learning_rate": 2e-05,
      "loss": 0.2162,
      "step": 7385
    },
    {
      "epoch": 4.990875295708009,
      "grad_norm": 2.3744356632232666,
      "learning_rate": 2e-05,
      "loss": 0.148,
      "step": 7386
    },
    {
      "epoch": 4.9915511997296385,
      "grad_norm": 2.2030422687530518,
      "learning_rate": 2e-05,
      "loss": 0.1628,
      "step": 7387
    },
    {
      "epoch": 4.992227103751267,
      "grad_norm": 6.377627372741699,
      "learning_rate": 2e-05,
      "loss": 0.1886,
      "step": 7388
    },
    {
      "epoch": 4.992903007772896,
      "grad_norm": 3.283524990081787,
      "learning_rate": 2e-05,
      "loss": 0.2114,
      "step": 7389
    },
    {
      "epoch": 4.9935789117945255,
      "grad_norm": 7.6637678146362305,
      "learning_rate": 2e-05,
      "loss": 0.1511,
      "step": 7390
    },
    {
      "epoch": 4.994254815816154,
      "grad_norm": 3.582052230834961,
      "learning_rate": 2e-05,
      "loss": 0.1662,
      "step": 7391
    },
    {
      "epoch": 4.994930719837783,
      "grad_norm": 4.145167350769043,
      "learning_rate": 2e-05,
      "loss": 0.2089,
      "step": 7392
    },
    {
      "epoch": 4.995606623859412,
      "grad_norm": 3.9028189182281494,
      "learning_rate": 2e-05,
      "loss": 0.1327,
      "step": 7393
    },
    {
      "epoch": 4.996282527881041,
      "grad_norm": 5.35150671005249,
      "learning_rate": 2e-05,
      "loss": 0.2131,
      "step": 7394
    },
    {
      "epoch": 4.996958431902669,
      "grad_norm": 2.275482416152954,
      "learning_rate": 2e-05,
      "loss": 0.1508,
      "step": 7395
    },
    {
      "epoch": 4.997634335924299,
      "grad_norm": 3.357776641845703,
      "learning_rate": 2e-05,
      "loss": 0.1493,
      "step": 7396
    },
    {
      "epoch": 4.998310239945928,
      "grad_norm": 4.025153636932373,
      "learning_rate": 2e-05,
      "loss": 0.1351,
      "step": 7397
    },
    {
      "epoch": 4.9989861439675565,
      "grad_norm": 2.093395948410034,
      "learning_rate": 2e-05,
      "loss": 0.164,
      "step": 7398
    },
    {
      "epoch": 4.999662047989186,
      "grad_norm": 2.0572822093963623,
      "learning_rate": 2e-05,
      "loss": 0.1228,
      "step": 7399
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.0572822093963623,
      "learning_rate": 2e-05,
      "loss": 0.1065,
      "step": 7400
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7455658783783784,
      "eval_loss": 0.23498788475990295,
      "eval_runtime": 51.1613,
      "eval_samples_per_second": 462.849,
      "eval_steps_per_second": 14.464,
      "step": 7400
    },
    {
      "epoch": 5.000675904021629,
      "grad_norm": 3.455070972442627,
      "learning_rate": 2e-05,
      "loss": 0.1725,
      "step": 7401
    },
    {
      "epoch": 5.001351808043258,
      "grad_norm": 4.594246864318848,
      "learning_rate": 2e-05,
      "loss": 0.1576,
      "step": 7402
    },
    {
      "epoch": 5.002027712064887,
      "grad_norm": 2.096808671951294,
      "learning_rate": 2e-05,
      "loss": 0.1382,
      "step": 7403
    },
    {
      "epoch": 5.0027036160865155,
      "grad_norm": 3.5333786010742188,
      "learning_rate": 2e-05,
      "loss": 0.1616,
      "step": 7404
    },
    {
      "epoch": 5.003379520108145,
      "grad_norm": 2.1993248462677,
      "learning_rate": 2e-05,
      "loss": 0.1475,
      "step": 7405
    },
    {
      "epoch": 5.004055424129773,
      "grad_norm": 2.6949501037597656,
      "learning_rate": 2e-05,
      "loss": 0.1562,
      "step": 7406
    },
    {
      "epoch": 5.0047313281514025,
      "grad_norm": 2.5778255462646484,
      "learning_rate": 2e-05,
      "loss": 0.1189,
      "step": 7407
    },
    {
      "epoch": 5.005407232173032,
      "grad_norm": 2.847114086151123,
      "learning_rate": 2e-05,
      "loss": 0.1314,
      "step": 7408
    },
    {
      "epoch": 5.00608313619466,
      "grad_norm": 2.5106379985809326,
      "learning_rate": 2e-05,
      "loss": 0.1456,
      "step": 7409
    },
    {
      "epoch": 5.00675904021629,
      "grad_norm": 3.660008192062378,
      "learning_rate": 2e-05,
      "loss": 0.1377,
      "step": 7410
    },
    {
      "epoch": 5.007434944237918,
      "grad_norm": 3.0908350944519043,
      "learning_rate": 2e-05,
      "loss": 0.1169,
      "step": 7411
    },
    {
      "epoch": 5.008110848259547,
      "grad_norm": 3.3736324310302734,
      "learning_rate": 2e-05,
      "loss": 0.2035,
      "step": 7412
    },
    {
      "epoch": 5.008786752281176,
      "grad_norm": 3.6075000762939453,
      "learning_rate": 2e-05,
      "loss": 0.1553,
      "step": 7413
    },
    {
      "epoch": 5.009462656302805,
      "grad_norm": 3.208522319793701,
      "learning_rate": 2e-05,
      "loss": 0.1236,
      "step": 7414
    },
    {
      "epoch": 5.0101385603244335,
      "grad_norm": 2.9924476146698,
      "learning_rate": 2e-05,
      "loss": 0.1436,
      "step": 7415
    },
    {
      "epoch": 5.010814464346063,
      "grad_norm": 3.1232433319091797,
      "learning_rate": 2e-05,
      "loss": 0.1481,
      "step": 7416
    },
    {
      "epoch": 5.011490368367692,
      "grad_norm": 3.990950584411621,
      "learning_rate": 2e-05,
      "loss": 0.1346,
      "step": 7417
    },
    {
      "epoch": 5.0121662723893206,
      "grad_norm": 2.9352972507476807,
      "learning_rate": 2e-05,
      "loss": 0.1839,
      "step": 7418
    },
    {
      "epoch": 5.01284217641095,
      "grad_norm": 5.428167819976807,
      "learning_rate": 2e-05,
      "loss": 0.1895,
      "step": 7419
    },
    {
      "epoch": 5.013518080432578,
      "grad_norm": 4.874518871307373,
      "learning_rate": 2e-05,
      "loss": 0.1382,
      "step": 7420
    },
    {
      "epoch": 5.014193984454208,
      "grad_norm": 2.9151370525360107,
      "learning_rate": 2e-05,
      "loss": 0.177,
      "step": 7421
    },
    {
      "epoch": 5.014869888475836,
      "grad_norm": 1.9294573068618774,
      "learning_rate": 2e-05,
      "loss": 0.0804,
      "step": 7422
    },
    {
      "epoch": 5.015545792497465,
      "grad_norm": 2.6187591552734375,
      "learning_rate": 2e-05,
      "loss": 0.1567,
      "step": 7423
    },
    {
      "epoch": 5.016221696519095,
      "grad_norm": 2.9142916202545166,
      "learning_rate": 2e-05,
      "loss": 0.1603,
      "step": 7424
    },
    {
      "epoch": 5.016897600540723,
      "grad_norm": 3.167005777359009,
      "learning_rate": 2e-05,
      "loss": 0.1517,
      "step": 7425
    },
    {
      "epoch": 5.017573504562352,
      "grad_norm": 2.5214922428131104,
      "learning_rate": 2e-05,
      "loss": 0.1279,
      "step": 7426
    },
    {
      "epoch": 5.018249408583981,
      "grad_norm": 2.471663475036621,
      "learning_rate": 2e-05,
      "loss": 0.1739,
      "step": 7427
    },
    {
      "epoch": 5.01892531260561,
      "grad_norm": 4.3346052169799805,
      "learning_rate": 2e-05,
      "loss": 0.171,
      "step": 7428
    },
    {
      "epoch": 5.019601216627239,
      "grad_norm": 1.9789386987686157,
      "learning_rate": 2e-05,
      "loss": 0.2004,
      "step": 7429
    },
    {
      "epoch": 5.020277120648868,
      "grad_norm": 4.669678211212158,
      "learning_rate": 2e-05,
      "loss": 0.1795,
      "step": 7430
    },
    {
      "epoch": 5.020953024670497,
      "grad_norm": 5.618382453918457,
      "learning_rate": 2e-05,
      "loss": 0.2238,
      "step": 7431
    },
    {
      "epoch": 5.021628928692126,
      "grad_norm": 2.6098759174346924,
      "learning_rate": 2e-05,
      "loss": 0.1245,
      "step": 7432
    },
    {
      "epoch": 5.022304832713755,
      "grad_norm": 2.101518154144287,
      "learning_rate": 2e-05,
      "loss": 0.1748,
      "step": 7433
    },
    {
      "epoch": 5.022980736735383,
      "grad_norm": 3.2446835041046143,
      "learning_rate": 2e-05,
      "loss": 0.1533,
      "step": 7434
    },
    {
      "epoch": 5.023656640757013,
      "grad_norm": 4.415116786956787,
      "learning_rate": 2e-05,
      "loss": 0.1528,
      "step": 7435
    },
    {
      "epoch": 5.024332544778641,
      "grad_norm": 2.5302000045776367,
      "learning_rate": 2e-05,
      "loss": 0.1307,
      "step": 7436
    },
    {
      "epoch": 5.02500844880027,
      "grad_norm": 3.412398338317871,
      "learning_rate": 2e-05,
      "loss": 0.1625,
      "step": 7437
    },
    {
      "epoch": 5.025684352821899,
      "grad_norm": 3.849611520767212,
      "learning_rate": 2e-05,
      "loss": 0.1245,
      "step": 7438
    },
    {
      "epoch": 5.026360256843528,
      "grad_norm": 2.5864670276641846,
      "learning_rate": 2e-05,
      "loss": 0.1933,
      "step": 7439
    },
    {
      "epoch": 5.0270361608651575,
      "grad_norm": 2.253234624862671,
      "learning_rate": 2e-05,
      "loss": 0.1644,
      "step": 7440
    },
    {
      "epoch": 5.027712064886786,
      "grad_norm": 3.2211496829986572,
      "learning_rate": 2e-05,
      "loss": 0.1315,
      "step": 7441
    },
    {
      "epoch": 5.028387968908415,
      "grad_norm": 2.5149567127227783,
      "learning_rate": 2e-05,
      "loss": 0.1504,
      "step": 7442
    },
    {
      "epoch": 5.029063872930044,
      "grad_norm": 6.2827935218811035,
      "learning_rate": 2e-05,
      "loss": 0.1219,
      "step": 7443
    },
    {
      "epoch": 5.029739776951673,
      "grad_norm": 2.322582960128784,
      "learning_rate": 2e-05,
      "loss": 0.1481,
      "step": 7444
    },
    {
      "epoch": 5.030415680973301,
      "grad_norm": 2.462961435317993,
      "learning_rate": 2e-05,
      "loss": 0.0772,
      "step": 7445
    },
    {
      "epoch": 5.031091584994931,
      "grad_norm": 1.567392110824585,
      "learning_rate": 2e-05,
      "loss": 0.1394,
      "step": 7446
    },
    {
      "epoch": 5.03176748901656,
      "grad_norm": 4.4968743324279785,
      "learning_rate": 2e-05,
      "loss": 0.1863,
      "step": 7447
    },
    {
      "epoch": 5.032443393038188,
      "grad_norm": 5.772333145141602,
      "learning_rate": 2e-05,
      "loss": 0.1738,
      "step": 7448
    },
    {
      "epoch": 5.033119297059818,
      "grad_norm": 2.807926893234253,
      "learning_rate": 2e-05,
      "loss": 0.1315,
      "step": 7449
    },
    {
      "epoch": 5.033795201081446,
      "grad_norm": 2.275479555130005,
      "learning_rate": 2e-05,
      "loss": 0.1176,
      "step": 7450
    },
    {
      "epoch": 5.033795201081446,
      "eval_accuracy": 0.7436233108108108,
      "eval_loss": 0.2602056860923767,
      "eval_runtime": 47.0229,
      "eval_samples_per_second": 503.585,
      "eval_steps_per_second": 15.737,
      "step": 7450
    },
    {
      "epoch": 5.0344711051030755,
      "grad_norm": 2.4149725437164307,
      "learning_rate": 2e-05,
      "loss": 0.1751,
      "step": 7451
    },
    {
      "epoch": 5.035147009124704,
      "grad_norm": 8.332110404968262,
      "learning_rate": 2e-05,
      "loss": 0.1426,
      "step": 7452
    },
    {
      "epoch": 5.035822913146333,
      "grad_norm": 6.783546447753906,
      "learning_rate": 2e-05,
      "loss": 0.1765,
      "step": 7453
    },
    {
      "epoch": 5.0364988171679625,
      "grad_norm": 2.927685499191284,
      "learning_rate": 2e-05,
      "loss": 0.1435,
      "step": 7454
    },
    {
      "epoch": 5.037174721189591,
      "grad_norm": 2.5868515968322754,
      "learning_rate": 2e-05,
      "loss": 0.1637,
      "step": 7455
    },
    {
      "epoch": 5.03785062521122,
      "grad_norm": 2.882571220397949,
      "learning_rate": 2e-05,
      "loss": 0.1166,
      "step": 7456
    },
    {
      "epoch": 5.038526529232849,
      "grad_norm": 2.5391480922698975,
      "learning_rate": 2e-05,
      "loss": 0.1239,
      "step": 7457
    },
    {
      "epoch": 5.039202433254478,
      "grad_norm": 1.89618980884552,
      "learning_rate": 2e-05,
      "loss": 0.1766,
      "step": 7458
    },
    {
      "epoch": 5.0398783372761065,
      "grad_norm": 2.999394416809082,
      "learning_rate": 2e-05,
      "loss": 0.1936,
      "step": 7459
    },
    {
      "epoch": 5.040554241297736,
      "grad_norm": 5.354795932769775,
      "learning_rate": 2e-05,
      "loss": 0.1246,
      "step": 7460
    },
    {
      "epoch": 5.041230145319365,
      "grad_norm": 2.987395763397217,
      "learning_rate": 2e-05,
      "loss": 0.2244,
      "step": 7461
    },
    {
      "epoch": 5.0419060493409935,
      "grad_norm": 2.882411479949951,
      "learning_rate": 2e-05,
      "loss": 0.1626,
      "step": 7462
    },
    {
      "epoch": 5.042581953362623,
      "grad_norm": 8.121099472045898,
      "learning_rate": 2e-05,
      "loss": 0.1532,
      "step": 7463
    },
    {
      "epoch": 5.043257857384251,
      "grad_norm": 6.0330352783203125,
      "learning_rate": 2e-05,
      "loss": 0.1402,
      "step": 7464
    },
    {
      "epoch": 5.043933761405881,
      "grad_norm": 3.1205079555511475,
      "learning_rate": 2e-05,
      "loss": 0.1928,
      "step": 7465
    },
    {
      "epoch": 5.044609665427509,
      "grad_norm": 2.565986156463623,
      "learning_rate": 2e-05,
      "loss": 0.1127,
      "step": 7466
    },
    {
      "epoch": 5.045285569449138,
      "grad_norm": 1.609315037727356,
      "learning_rate": 2e-05,
      "loss": 0.1136,
      "step": 7467
    },
    {
      "epoch": 5.045961473470767,
      "grad_norm": 5.6548614501953125,
      "learning_rate": 2e-05,
      "loss": 0.1559,
      "step": 7468
    },
    {
      "epoch": 5.046637377492396,
      "grad_norm": 3.2807493209838867,
      "learning_rate": 2e-05,
      "loss": 0.12,
      "step": 7469
    },
    {
      "epoch": 5.047313281514025,
      "grad_norm": 2.686129093170166,
      "learning_rate": 2e-05,
      "loss": 0.138,
      "step": 7470
    },
    {
      "epoch": 5.047989185535654,
      "grad_norm": 1.7365601062774658,
      "learning_rate": 2e-05,
      "loss": 0.1394,
      "step": 7471
    },
    {
      "epoch": 5.048665089557283,
      "grad_norm": 2.953791856765747,
      "learning_rate": 2e-05,
      "loss": 0.1211,
      "step": 7472
    },
    {
      "epoch": 5.0493409935789115,
      "grad_norm": 2.6480064392089844,
      "learning_rate": 2e-05,
      "loss": 0.146,
      "step": 7473
    },
    {
      "epoch": 5.050016897600541,
      "grad_norm": 2.645437240600586,
      "learning_rate": 2e-05,
      "loss": 0.1228,
      "step": 7474
    },
    {
      "epoch": 5.050692801622169,
      "grad_norm": 2.6811704635620117,
      "learning_rate": 2e-05,
      "loss": 0.1356,
      "step": 7475
    },
    {
      "epoch": 5.051368705643799,
      "grad_norm": 2.9374682903289795,
      "learning_rate": 2e-05,
      "loss": 0.1073,
      "step": 7476
    },
    {
      "epoch": 5.052044609665428,
      "grad_norm": 5.388411521911621,
      "learning_rate": 2e-05,
      "loss": 0.1532,
      "step": 7477
    },
    {
      "epoch": 5.052720513687056,
      "grad_norm": 3.30356764793396,
      "learning_rate": 2e-05,
      "loss": 0.1298,
      "step": 7478
    },
    {
      "epoch": 5.053396417708686,
      "grad_norm": 3.4130828380584717,
      "learning_rate": 2e-05,
      "loss": 0.1596,
      "step": 7479
    },
    {
      "epoch": 5.054072321730314,
      "grad_norm": 5.3659210205078125,
      "learning_rate": 2e-05,
      "loss": 0.2041,
      "step": 7480
    },
    {
      "epoch": 5.054748225751943,
      "grad_norm": 6.778443336486816,
      "learning_rate": 2e-05,
      "loss": 0.2055,
      "step": 7481
    },
    {
      "epoch": 5.055424129773572,
      "grad_norm": 2.887321710586548,
      "learning_rate": 2e-05,
      "loss": 0.1984,
      "step": 7482
    },
    {
      "epoch": 5.056100033795201,
      "grad_norm": 5.387823581695557,
      "learning_rate": 2e-05,
      "loss": 0.1157,
      "step": 7483
    },
    {
      "epoch": 5.05677593781683,
      "grad_norm": 5.940553665161133,
      "learning_rate": 2e-05,
      "loss": 0.1878,
      "step": 7484
    },
    {
      "epoch": 5.057451841838459,
      "grad_norm": 2.185269832611084,
      "learning_rate": 2e-05,
      "loss": 0.1474,
      "step": 7485
    },
    {
      "epoch": 5.058127745860088,
      "grad_norm": 6.432172775268555,
      "learning_rate": 2e-05,
      "loss": 0.1189,
      "step": 7486
    },
    {
      "epoch": 5.058803649881717,
      "grad_norm": 5.039313316345215,
      "learning_rate": 2e-05,
      "loss": 0.1464,
      "step": 7487
    },
    {
      "epoch": 5.059479553903346,
      "grad_norm": 5.896623611450195,
      "learning_rate": 2e-05,
      "loss": 0.1468,
      "step": 7488
    },
    {
      "epoch": 5.060155457924974,
      "grad_norm": 1.9824917316436768,
      "learning_rate": 2e-05,
      "loss": 0.1519,
      "step": 7489
    },
    {
      "epoch": 5.060831361946604,
      "grad_norm": 4.014425754547119,
      "learning_rate": 2e-05,
      "loss": 0.1131,
      "step": 7490
    },
    {
      "epoch": 5.061507265968232,
      "grad_norm": 2.278265953063965,
      "learning_rate": 2e-05,
      "loss": 0.086,
      "step": 7491
    },
    {
      "epoch": 5.062183169989861,
      "grad_norm": 1.80195152759552,
      "learning_rate": 2e-05,
      "loss": 0.1403,
      "step": 7492
    },
    {
      "epoch": 5.062859074011491,
      "grad_norm": 2.3825671672821045,
      "learning_rate": 2e-05,
      "loss": 0.1328,
      "step": 7493
    },
    {
      "epoch": 5.063534978033119,
      "grad_norm": 2.6876091957092285,
      "learning_rate": 2e-05,
      "loss": 0.1611,
      "step": 7494
    },
    {
      "epoch": 5.0642108820547485,
      "grad_norm": 7.3367390632629395,
      "learning_rate": 2e-05,
      "loss": 0.1576,
      "step": 7495
    },
    {
      "epoch": 5.064886786076377,
      "grad_norm": 5.12595272064209,
      "learning_rate": 2e-05,
      "loss": 0.1558,
      "step": 7496
    },
    {
      "epoch": 5.065562690098006,
      "grad_norm": 5.39927864074707,
      "learning_rate": 2e-05,
      "loss": 0.1424,
      "step": 7497
    },
    {
      "epoch": 5.066238594119635,
      "grad_norm": 2.640666961669922,
      "learning_rate": 2e-05,
      "loss": 0.1473,
      "step": 7498
    },
    {
      "epoch": 5.066914498141264,
      "grad_norm": 5.038501262664795,
      "learning_rate": 2e-05,
      "loss": 0.1453,
      "step": 7499
    },
    {
      "epoch": 5.067590402162893,
      "grad_norm": 3.120713233947754,
      "learning_rate": 2e-05,
      "loss": 0.1051,
      "step": 7500
    },
    {
      "epoch": 5.067590402162893,
      "eval_accuracy": 0.7233952702702703,
      "eval_loss": 0.24130430817604065,
      "eval_runtime": 56.7324,
      "eval_samples_per_second": 417.398,
      "eval_steps_per_second": 13.044,
      "step": 7500
    },
    {
      "epoch": 5.068266306184522,
      "grad_norm": 1.737207293510437,
      "learning_rate": 2e-05,
      "loss": 0.1093,
      "step": 7501
    },
    {
      "epoch": 5.068942210206151,
      "grad_norm": 3.1430108547210693,
      "learning_rate": 2e-05,
      "loss": 0.1884,
      "step": 7502
    },
    {
      "epoch": 5.069618114227779,
      "grad_norm": 2.2559921741485596,
      "learning_rate": 2e-05,
      "loss": 0.1649,
      "step": 7503
    },
    {
      "epoch": 5.070294018249409,
      "grad_norm": 1.6710927486419678,
      "learning_rate": 2e-05,
      "loss": 0.1139,
      "step": 7504
    },
    {
      "epoch": 5.070969922271037,
      "grad_norm": 2.843458652496338,
      "learning_rate": 2e-05,
      "loss": 0.2107,
      "step": 7505
    },
    {
      "epoch": 5.0716458262926665,
      "grad_norm": 2.8729805946350098,
      "learning_rate": 2e-05,
      "loss": 0.0957,
      "step": 7506
    },
    {
      "epoch": 5.072321730314296,
      "grad_norm": 1.826790690422058,
      "learning_rate": 2e-05,
      "loss": 0.1141,
      "step": 7507
    },
    {
      "epoch": 5.072997634335924,
      "grad_norm": 5.102361679077148,
      "learning_rate": 2e-05,
      "loss": 0.1431,
      "step": 7508
    },
    {
      "epoch": 5.0736735383575535,
      "grad_norm": 3.5379350185394287,
      "learning_rate": 2e-05,
      "loss": 0.0725,
      "step": 7509
    },
    {
      "epoch": 5.074349442379182,
      "grad_norm": 4.219025135040283,
      "learning_rate": 2e-05,
      "loss": 0.2558,
      "step": 7510
    },
    {
      "epoch": 5.075025346400811,
      "grad_norm": 3.991584539413452,
      "learning_rate": 2e-05,
      "loss": 0.1385,
      "step": 7511
    },
    {
      "epoch": 5.07570125042244,
      "grad_norm": 2.893838405609131,
      "learning_rate": 2e-05,
      "loss": 0.1428,
      "step": 7512
    },
    {
      "epoch": 5.076377154444069,
      "grad_norm": 3.899163246154785,
      "learning_rate": 2e-05,
      "loss": 0.1668,
      "step": 7513
    },
    {
      "epoch": 5.077053058465697,
      "grad_norm": 5.661072731018066,
      "learning_rate": 2e-05,
      "loss": 0.1974,
      "step": 7514
    },
    {
      "epoch": 5.077728962487327,
      "grad_norm": 9.250579833984375,
      "learning_rate": 2e-05,
      "loss": 0.1702,
      "step": 7515
    },
    {
      "epoch": 5.078404866508956,
      "grad_norm": 4.378829479217529,
      "learning_rate": 2e-05,
      "loss": 0.1091,
      "step": 7516
    },
    {
      "epoch": 5.0790807705305845,
      "grad_norm": 1.5249274969100952,
      "learning_rate": 2e-05,
      "loss": 0.1299,
      "step": 7517
    },
    {
      "epoch": 5.079756674552214,
      "grad_norm": 4.1659111976623535,
      "learning_rate": 2e-05,
      "loss": 0.2174,
      "step": 7518
    },
    {
      "epoch": 5.080432578573842,
      "grad_norm": 8.53628921508789,
      "learning_rate": 2e-05,
      "loss": 0.1943,
      "step": 7519
    },
    {
      "epoch": 5.0811084825954715,
      "grad_norm": 3.7901432514190674,
      "learning_rate": 2e-05,
      "loss": 0.1463,
      "step": 7520
    },
    {
      "epoch": 5.0817843866171,
      "grad_norm": 5.538915634155273,
      "learning_rate": 2e-05,
      "loss": 0.1695,
      "step": 7521
    },
    {
      "epoch": 5.082460290638729,
      "grad_norm": 2.2562448978424072,
      "learning_rate": 2e-05,
      "loss": 0.1217,
      "step": 7522
    },
    {
      "epoch": 5.083136194660359,
      "grad_norm": 1.5259582996368408,
      "learning_rate": 2e-05,
      "loss": 0.1598,
      "step": 7523
    },
    {
      "epoch": 5.083812098681987,
      "grad_norm": 2.6344010829925537,
      "learning_rate": 2e-05,
      "loss": 0.1141,
      "step": 7524
    },
    {
      "epoch": 5.084488002703616,
      "grad_norm": 2.1965110301971436,
      "learning_rate": 2e-05,
      "loss": 0.1817,
      "step": 7525
    },
    {
      "epoch": 5.085163906725245,
      "grad_norm": 2.0155344009399414,
      "learning_rate": 2e-05,
      "loss": 0.1568,
      "step": 7526
    },
    {
      "epoch": 5.085839810746874,
      "grad_norm": 2.08682918548584,
      "learning_rate": 2e-05,
      "loss": 0.1385,
      "step": 7527
    },
    {
      "epoch": 5.0865157147685025,
      "grad_norm": 7.6946492195129395,
      "learning_rate": 2e-05,
      "loss": 0.1372,
      "step": 7528
    },
    {
      "epoch": 5.087191618790132,
      "grad_norm": 3.9944913387298584,
      "learning_rate": 2e-05,
      "loss": 0.1233,
      "step": 7529
    },
    {
      "epoch": 5.087867522811761,
      "grad_norm": 2.2238476276397705,
      "learning_rate": 2e-05,
      "loss": 0.167,
      "step": 7530
    },
    {
      "epoch": 5.08854342683339,
      "grad_norm": 2.819249153137207,
      "learning_rate": 2e-05,
      "loss": 0.1641,
      "step": 7531
    },
    {
      "epoch": 5.089219330855019,
      "grad_norm": 2.5727105140686035,
      "learning_rate": 2e-05,
      "loss": 0.1681,
      "step": 7532
    },
    {
      "epoch": 5.089895234876647,
      "grad_norm": 1.5024816989898682,
      "learning_rate": 2e-05,
      "loss": 0.1123,
      "step": 7533
    },
    {
      "epoch": 5.090571138898277,
      "grad_norm": 1.5883477926254272,
      "learning_rate": 2e-05,
      "loss": 0.1218,
      "step": 7534
    },
    {
      "epoch": 5.091247042919905,
      "grad_norm": 2.8000307083129883,
      "learning_rate": 2e-05,
      "loss": 0.2085,
      "step": 7535
    },
    {
      "epoch": 5.091922946941534,
      "grad_norm": 6.346426486968994,
      "learning_rate": 2e-05,
      "loss": 0.1215,
      "step": 7536
    },
    {
      "epoch": 5.092598850963164,
      "grad_norm": 3.7828757762908936,
      "learning_rate": 2e-05,
      "loss": 0.1476,
      "step": 7537
    },
    {
      "epoch": 5.093274754984792,
      "grad_norm": 4.668438911437988,
      "learning_rate": 2e-05,
      "loss": 0.16,
      "step": 7538
    },
    {
      "epoch": 5.093950659006421,
      "grad_norm": 2.249469518661499,
      "learning_rate": 2e-05,
      "loss": 0.1358,
      "step": 7539
    },
    {
      "epoch": 5.09462656302805,
      "grad_norm": 2.1814913749694824,
      "learning_rate": 2e-05,
      "loss": 0.1204,
      "step": 7540
    },
    {
      "epoch": 5.095302467049679,
      "grad_norm": 2.6845076084136963,
      "learning_rate": 2e-05,
      "loss": 0.147,
      "step": 7541
    },
    {
      "epoch": 5.095978371071308,
      "grad_norm": 2.1967129707336426,
      "learning_rate": 2e-05,
      "loss": 0.226,
      "step": 7542
    },
    {
      "epoch": 5.096654275092937,
      "grad_norm": 5.631228446960449,
      "learning_rate": 2e-05,
      "loss": 0.1423,
      "step": 7543
    },
    {
      "epoch": 5.097330179114565,
      "grad_norm": 1.8589578866958618,
      "learning_rate": 2e-05,
      "loss": 0.1177,
      "step": 7544
    },
    {
      "epoch": 5.098006083136195,
      "grad_norm": 4.196567535400391,
      "learning_rate": 2e-05,
      "loss": 0.1559,
      "step": 7545
    },
    {
      "epoch": 5.098681987157824,
      "grad_norm": 2.9830679893493652,
      "learning_rate": 2e-05,
      "loss": 0.1644,
      "step": 7546
    },
    {
      "epoch": 5.099357891179452,
      "grad_norm": 3.5198943614959717,
      "learning_rate": 2e-05,
      "loss": 0.1567,
      "step": 7547
    },
    {
      "epoch": 5.100033795201082,
      "grad_norm": 2.9958908557891846,
      "learning_rate": 2e-05,
      "loss": 0.1949,
      "step": 7548
    },
    {
      "epoch": 5.10070969922271,
      "grad_norm": 2.9151668548583984,
      "learning_rate": 2e-05,
      "loss": 0.149,
      "step": 7549
    },
    {
      "epoch": 5.101385603244339,
      "grad_norm": 1.8412355184555054,
      "learning_rate": 2e-05,
      "loss": 0.1842,
      "step": 7550
    },
    {
      "epoch": 5.101385603244339,
      "eval_accuracy": 0.7518158783783784,
      "eval_loss": 0.230007603764534,
      "eval_runtime": 49.5116,
      "eval_samples_per_second": 478.272,
      "eval_steps_per_second": 14.946,
      "step": 7550
    },
    {
      "epoch": 5.102061507265968,
      "grad_norm": 3.40234375,
      "learning_rate": 2e-05,
      "loss": 0.1448,
      "step": 7551
    },
    {
      "epoch": 5.102737411287597,
      "grad_norm": 1.7480721473693848,
      "learning_rate": 2e-05,
      "loss": 0.1891,
      "step": 7552
    },
    {
      "epoch": 5.1034133153092265,
      "grad_norm": 2.3939785957336426,
      "learning_rate": 2e-05,
      "loss": 0.1266,
      "step": 7553
    },
    {
      "epoch": 5.104089219330855,
      "grad_norm": 3.116842269897461,
      "learning_rate": 2e-05,
      "loss": 0.1088,
      "step": 7554
    },
    {
      "epoch": 5.104765123352484,
      "grad_norm": 2.5791618824005127,
      "learning_rate": 2e-05,
      "loss": 0.1401,
      "step": 7555
    },
    {
      "epoch": 5.105441027374113,
      "grad_norm": 1.9465590715408325,
      "learning_rate": 2e-05,
      "loss": 0.1737,
      "step": 7556
    },
    {
      "epoch": 5.106116931395742,
      "grad_norm": 2.2820229530334473,
      "learning_rate": 2e-05,
      "loss": 0.1904,
      "step": 7557
    },
    {
      "epoch": 5.10679283541737,
      "grad_norm": 3.9349138736724854,
      "learning_rate": 2e-05,
      "loss": 0.1273,
      "step": 7558
    },
    {
      "epoch": 5.107468739439,
      "grad_norm": 3.210052967071533,
      "learning_rate": 2e-05,
      "loss": 0.1722,
      "step": 7559
    },
    {
      "epoch": 5.108144643460629,
      "grad_norm": 2.7178359031677246,
      "learning_rate": 2e-05,
      "loss": 0.0859,
      "step": 7560
    },
    {
      "epoch": 5.1088205474822574,
      "grad_norm": 4.331723213195801,
      "learning_rate": 2e-05,
      "loss": 0.1693,
      "step": 7561
    },
    {
      "epoch": 5.109496451503887,
      "grad_norm": 3.370319128036499,
      "learning_rate": 2e-05,
      "loss": 0.1576,
      "step": 7562
    },
    {
      "epoch": 5.110172355525515,
      "grad_norm": 2.292850971221924,
      "learning_rate": 2e-05,
      "loss": 0.1121,
      "step": 7563
    },
    {
      "epoch": 5.1108482595471445,
      "grad_norm": 2.749521017074585,
      "learning_rate": 2e-05,
      "loss": 0.1239,
      "step": 7564
    },
    {
      "epoch": 5.111524163568773,
      "grad_norm": 2.5143911838531494,
      "learning_rate": 2e-05,
      "loss": 0.1175,
      "step": 7565
    },
    {
      "epoch": 5.112200067590402,
      "grad_norm": 2.1260688304901123,
      "learning_rate": 2e-05,
      "loss": 0.1259,
      "step": 7566
    },
    {
      "epoch": 5.112875971612031,
      "grad_norm": 2.6917152404785156,
      "learning_rate": 2e-05,
      "loss": 0.1355,
      "step": 7567
    },
    {
      "epoch": 5.11355187563366,
      "grad_norm": 6.176032543182373,
      "learning_rate": 2e-05,
      "loss": 0.155,
      "step": 7568
    },
    {
      "epoch": 5.114227779655289,
      "grad_norm": 2.7556447982788086,
      "learning_rate": 2e-05,
      "loss": 0.1279,
      "step": 7569
    },
    {
      "epoch": 5.114903683676918,
      "grad_norm": 9.401896476745605,
      "learning_rate": 2e-05,
      "loss": 0.1973,
      "step": 7570
    },
    {
      "epoch": 5.115579587698547,
      "grad_norm": 5.622865676879883,
      "learning_rate": 2e-05,
      "loss": 0.1638,
      "step": 7571
    },
    {
      "epoch": 5.1162554917201755,
      "grad_norm": 3.989877223968506,
      "learning_rate": 2e-05,
      "loss": 0.2926,
      "step": 7572
    },
    {
      "epoch": 5.116931395741805,
      "grad_norm": 6.057412624359131,
      "learning_rate": 2e-05,
      "loss": 0.1667,
      "step": 7573
    },
    {
      "epoch": 5.117607299763433,
      "grad_norm": 3.2634177207946777,
      "learning_rate": 2e-05,
      "loss": 0.1584,
      "step": 7574
    },
    {
      "epoch": 5.1182832037850625,
      "grad_norm": 10.328509330749512,
      "learning_rate": 2e-05,
      "loss": 0.1744,
      "step": 7575
    },
    {
      "epoch": 5.118959107806692,
      "grad_norm": 2.613283395767212,
      "learning_rate": 2e-05,
      "loss": 0.1218,
      "step": 7576
    },
    {
      "epoch": 5.11963501182832,
      "grad_norm": 6.973835468292236,
      "learning_rate": 2e-05,
      "loss": 0.1613,
      "step": 7577
    },
    {
      "epoch": 5.12031091584995,
      "grad_norm": 7.9069108963012695,
      "learning_rate": 2e-05,
      "loss": 0.2094,
      "step": 7578
    },
    {
      "epoch": 5.120986819871578,
      "grad_norm": 3.0472891330718994,
      "learning_rate": 2e-05,
      "loss": 0.2155,
      "step": 7579
    },
    {
      "epoch": 5.121662723893207,
      "grad_norm": 2.2730886936187744,
      "learning_rate": 2e-05,
      "loss": 0.1347,
      "step": 7580
    },
    {
      "epoch": 5.122338627914836,
      "grad_norm": 1.4328604936599731,
      "learning_rate": 2e-05,
      "loss": 0.1126,
      "step": 7581
    },
    {
      "epoch": 5.123014531936465,
      "grad_norm": 2.920602560043335,
      "learning_rate": 2e-05,
      "loss": 0.1659,
      "step": 7582
    },
    {
      "epoch": 5.123690435958094,
      "grad_norm": 9.211099624633789,
      "learning_rate": 2e-05,
      "loss": 0.182,
      "step": 7583
    },
    {
      "epoch": 5.124366339979723,
      "grad_norm": 8.676933288574219,
      "learning_rate": 2e-05,
      "loss": 0.138,
      "step": 7584
    },
    {
      "epoch": 5.125042244001352,
      "grad_norm": 7.468955039978027,
      "learning_rate": 2e-05,
      "loss": 0.2353,
      "step": 7585
    },
    {
      "epoch": 5.1257181480229805,
      "grad_norm": 7.990413188934326,
      "learning_rate": 2e-05,
      "loss": 0.1783,
      "step": 7586
    },
    {
      "epoch": 5.12639405204461,
      "grad_norm": 3.5021817684173584,
      "learning_rate": 2e-05,
      "loss": 0.1223,
      "step": 7587
    },
    {
      "epoch": 5.127069956066238,
      "grad_norm": 2.0605311393737793,
      "learning_rate": 2e-05,
      "loss": 0.1158,
      "step": 7588
    },
    {
      "epoch": 5.127745860087868,
      "grad_norm": 4.687828063964844,
      "learning_rate": 2e-05,
      "loss": 0.1256,
      "step": 7589
    },
    {
      "epoch": 5.128421764109497,
      "grad_norm": 1.886339783668518,
      "learning_rate": 2e-05,
      "loss": 0.1382,
      "step": 7590
    },
    {
      "epoch": 5.129097668131125,
      "grad_norm": 5.3101396560668945,
      "learning_rate": 2e-05,
      "loss": 0.1613,
      "step": 7591
    },
    {
      "epoch": 5.129773572152755,
      "grad_norm": 4.766539573669434,
      "learning_rate": 2e-05,
      "loss": 0.1426,
      "step": 7592
    },
    {
      "epoch": 5.130449476174383,
      "grad_norm": 5.621750354766846,
      "learning_rate": 2e-05,
      "loss": 0.1228,
      "step": 7593
    },
    {
      "epoch": 5.131125380196012,
      "grad_norm": 2.471735954284668,
      "learning_rate": 2e-05,
      "loss": 0.1463,
      "step": 7594
    },
    {
      "epoch": 5.131801284217641,
      "grad_norm": 2.8987529277801514,
      "learning_rate": 2e-05,
      "loss": 0.208,
      "step": 7595
    },
    {
      "epoch": 5.13247718823927,
      "grad_norm": 6.724513053894043,
      "learning_rate": 2e-05,
      "loss": 0.132,
      "step": 7596
    },
    {
      "epoch": 5.1331530922608986,
      "grad_norm": 2.770871639251709,
      "learning_rate": 2e-05,
      "loss": 0.1322,
      "step": 7597
    },
    {
      "epoch": 5.133828996282528,
      "grad_norm": 2.274305582046509,
      "learning_rate": 2e-05,
      "loss": 0.0973,
      "step": 7598
    },
    {
      "epoch": 5.134504900304157,
      "grad_norm": 2.1650102138519287,
      "learning_rate": 2e-05,
      "loss": 0.1312,
      "step": 7599
    },
    {
      "epoch": 5.135180804325786,
      "grad_norm": 5.224236488342285,
      "learning_rate": 2e-05,
      "loss": 0.1506,
      "step": 7600
    },
    {
      "epoch": 5.135180804325786,
      "eval_accuracy": 0.7569256756756757,
      "eval_loss": 0.2823445200920105,
      "eval_runtime": 46.7544,
      "eval_samples_per_second": 506.477,
      "eval_steps_per_second": 15.827,
      "step": 7600
    },
    {
      "epoch": 5.135856708347415,
      "grad_norm": 7.080615043640137,
      "learning_rate": 2e-05,
      "loss": 0.163,
      "step": 7601
    },
    {
      "epoch": 5.136532612369043,
      "grad_norm": 6.393982410430908,
      "learning_rate": 2e-05,
      "loss": 0.186,
      "step": 7602
    },
    {
      "epoch": 5.137208516390673,
      "grad_norm": 7.941885948181152,
      "learning_rate": 2e-05,
      "loss": 0.2399,
      "step": 7603
    },
    {
      "epoch": 5.137884420412301,
      "grad_norm": 3.2376275062561035,
      "learning_rate": 2e-05,
      "loss": 0.1371,
      "step": 7604
    },
    {
      "epoch": 5.13856032443393,
      "grad_norm": 4.61256217956543,
      "learning_rate": 2e-05,
      "loss": 0.1761,
      "step": 7605
    },
    {
      "epoch": 5.13923622845556,
      "grad_norm": 4.079240798950195,
      "learning_rate": 2e-05,
      "loss": 0.1643,
      "step": 7606
    },
    {
      "epoch": 5.139912132477188,
      "grad_norm": 3.2530813217163086,
      "learning_rate": 2e-05,
      "loss": 0.1873,
      "step": 7607
    },
    {
      "epoch": 5.1405880364988175,
      "grad_norm": 3.949270486831665,
      "learning_rate": 2e-05,
      "loss": 0.2072,
      "step": 7608
    },
    {
      "epoch": 5.141263940520446,
      "grad_norm": 4.3838348388671875,
      "learning_rate": 2e-05,
      "loss": 0.1653,
      "step": 7609
    },
    {
      "epoch": 5.141939844542075,
      "grad_norm": 10.50023365020752,
      "learning_rate": 2e-05,
      "loss": 0.2414,
      "step": 7610
    },
    {
      "epoch": 5.142615748563704,
      "grad_norm": 8.922839164733887,
      "learning_rate": 2e-05,
      "loss": 0.2395,
      "step": 7611
    },
    {
      "epoch": 5.143291652585333,
      "grad_norm": 6.080845355987549,
      "learning_rate": 2e-05,
      "loss": 0.1639,
      "step": 7612
    },
    {
      "epoch": 5.143967556606961,
      "grad_norm": 3.3306684494018555,
      "learning_rate": 2e-05,
      "loss": 0.2014,
      "step": 7613
    },
    {
      "epoch": 5.144643460628591,
      "grad_norm": 1.9486799240112305,
      "learning_rate": 2e-05,
      "loss": 0.1753,
      "step": 7614
    },
    {
      "epoch": 5.14531936465022,
      "grad_norm": 3.2526934146881104,
      "learning_rate": 2e-05,
      "loss": 0.183,
      "step": 7615
    },
    {
      "epoch": 5.145995268671848,
      "grad_norm": 5.198669910430908,
      "learning_rate": 2e-05,
      "loss": 0.1725,
      "step": 7616
    },
    {
      "epoch": 5.146671172693478,
      "grad_norm": 2.4808156490325928,
      "learning_rate": 2e-05,
      "loss": 0.1499,
      "step": 7617
    },
    {
      "epoch": 5.147347076715106,
      "grad_norm": 2.8296210765838623,
      "learning_rate": 2e-05,
      "loss": 0.1412,
      "step": 7618
    },
    {
      "epoch": 5.1480229807367355,
      "grad_norm": 5.3029279708862305,
      "learning_rate": 2e-05,
      "loss": 0.1801,
      "step": 7619
    },
    {
      "epoch": 5.148698884758364,
      "grad_norm": 6.354578495025635,
      "learning_rate": 2e-05,
      "loss": 0.1823,
      "step": 7620
    },
    {
      "epoch": 5.149374788779993,
      "grad_norm": 1.8675174713134766,
      "learning_rate": 2e-05,
      "loss": 0.1262,
      "step": 7621
    },
    {
      "epoch": 5.1500506928016225,
      "grad_norm": 4.814053535461426,
      "learning_rate": 2e-05,
      "loss": 0.1535,
      "step": 7622
    },
    {
      "epoch": 5.150726596823251,
      "grad_norm": 3.8052947521209717,
      "learning_rate": 2e-05,
      "loss": 0.1475,
      "step": 7623
    },
    {
      "epoch": 5.15140250084488,
      "grad_norm": 1.732933521270752,
      "learning_rate": 2e-05,
      "loss": 0.1435,
      "step": 7624
    },
    {
      "epoch": 5.152078404866509,
      "grad_norm": 2.287597179412842,
      "learning_rate": 2e-05,
      "loss": 0.1214,
      "step": 7625
    },
    {
      "epoch": 5.152754308888138,
      "grad_norm": 9.676032066345215,
      "learning_rate": 2e-05,
      "loss": 0.1799,
      "step": 7626
    },
    {
      "epoch": 5.153430212909766,
      "grad_norm": 4.222305774688721,
      "learning_rate": 2e-05,
      "loss": 0.1578,
      "step": 7627
    },
    {
      "epoch": 5.154106116931396,
      "grad_norm": 6.869168281555176,
      "learning_rate": 2e-05,
      "loss": 0.1381,
      "step": 7628
    },
    {
      "epoch": 5.154782020953025,
      "grad_norm": 2.366380453109741,
      "learning_rate": 2e-05,
      "loss": 0.146,
      "step": 7629
    },
    {
      "epoch": 5.1554579249746535,
      "grad_norm": 7.788395881652832,
      "learning_rate": 2e-05,
      "loss": 0.2067,
      "step": 7630
    },
    {
      "epoch": 5.156133828996283,
      "grad_norm": 1.8782222270965576,
      "learning_rate": 2e-05,
      "loss": 0.1083,
      "step": 7631
    },
    {
      "epoch": 5.156809733017911,
      "grad_norm": 2.5205886363983154,
      "learning_rate": 2e-05,
      "loss": 0.1531,
      "step": 7632
    },
    {
      "epoch": 5.1574856370395405,
      "grad_norm": 2.534795045852661,
      "learning_rate": 2e-05,
      "loss": 0.2068,
      "step": 7633
    },
    {
      "epoch": 5.158161541061169,
      "grad_norm": 6.432212829589844,
      "learning_rate": 2e-05,
      "loss": 0.189,
      "step": 7634
    },
    {
      "epoch": 5.158837445082798,
      "grad_norm": 2.797034502029419,
      "learning_rate": 2e-05,
      "loss": 0.1098,
      "step": 7635
    },
    {
      "epoch": 5.159513349104428,
      "grad_norm": 6.3025360107421875,
      "learning_rate": 2e-05,
      "loss": 0.209,
      "step": 7636
    },
    {
      "epoch": 5.160189253126056,
      "grad_norm": 4.2118425369262695,
      "learning_rate": 2e-05,
      "loss": 0.1483,
      "step": 7637
    },
    {
      "epoch": 5.160865157147685,
      "grad_norm": 4.291739463806152,
      "learning_rate": 2e-05,
      "loss": 0.1925,
      "step": 7638
    },
    {
      "epoch": 5.161541061169314,
      "grad_norm": 2.757368326187134,
      "learning_rate": 2e-05,
      "loss": 0.1879,
      "step": 7639
    },
    {
      "epoch": 5.162216965190943,
      "grad_norm": 6.091044902801514,
      "learning_rate": 2e-05,
      "loss": 0.1322,
      "step": 7640
    },
    {
      "epoch": 5.1628928692125715,
      "grad_norm": 4.062774658203125,
      "learning_rate": 2e-05,
      "loss": 0.1551,
      "step": 7641
    },
    {
      "epoch": 5.163568773234201,
      "grad_norm": 7.436765193939209,
      "learning_rate": 2e-05,
      "loss": 0.1621,
      "step": 7642
    },
    {
      "epoch": 5.164244677255829,
      "grad_norm": 8.138161659240723,
      "learning_rate": 2e-05,
      "loss": 0.2054,
      "step": 7643
    },
    {
      "epoch": 5.164920581277459,
      "grad_norm": 2.1529276371002197,
      "learning_rate": 2e-05,
      "loss": 0.1514,
      "step": 7644
    },
    {
      "epoch": 5.165596485299088,
      "grad_norm": 2.069042921066284,
      "learning_rate": 2e-05,
      "loss": 0.1647,
      "step": 7645
    },
    {
      "epoch": 5.166272389320716,
      "grad_norm": 7.005636215209961,
      "learning_rate": 2e-05,
      "loss": 0.165,
      "step": 7646
    },
    {
      "epoch": 5.166948293342346,
      "grad_norm": 2.818666696548462,
      "learning_rate": 2e-05,
      "loss": 0.1555,
      "step": 7647
    },
    {
      "epoch": 5.167624197363974,
      "grad_norm": 8.058354377746582,
      "learning_rate": 2e-05,
      "loss": 0.1806,
      "step": 7648
    },
    {
      "epoch": 5.168300101385603,
      "grad_norm": 9.074392318725586,
      "learning_rate": 2e-05,
      "loss": 0.1827,
      "step": 7649
    },
    {
      "epoch": 5.168976005407232,
      "grad_norm": 6.634652614593506,
      "learning_rate": 2e-05,
      "loss": 0.1757,
      "step": 7650
    },
    {
      "epoch": 5.168976005407232,
      "eval_accuracy": 0.7835726351351351,
      "eval_loss": 0.23568978905677795,
      "eval_runtime": 45.9171,
      "eval_samples_per_second": 515.712,
      "eval_steps_per_second": 16.116,
      "step": 7650
    },
    {
      "epoch": 5.169651909428861,
      "grad_norm": 2.756027936935425,
      "learning_rate": 2e-05,
      "loss": 0.1439,
      "step": 7651
    },
    {
      "epoch": 5.17032781345049,
      "grad_norm": 3.9145829677581787,
      "learning_rate": 2e-05,
      "loss": 0.1357,
      "step": 7652
    },
    {
      "epoch": 5.171003717472119,
      "grad_norm": 2.3709959983825684,
      "learning_rate": 2e-05,
      "loss": 0.1919,
      "step": 7653
    },
    {
      "epoch": 5.171679621493748,
      "grad_norm": 4.858457565307617,
      "learning_rate": 2e-05,
      "loss": 0.1511,
      "step": 7654
    },
    {
      "epoch": 5.172355525515377,
      "grad_norm": 6.0871686935424805,
      "learning_rate": 2e-05,
      "loss": 0.1587,
      "step": 7655
    },
    {
      "epoch": 5.173031429537006,
      "grad_norm": 2.7608585357666016,
      "learning_rate": 2e-05,
      "loss": 0.161,
      "step": 7656
    },
    {
      "epoch": 5.173707333558634,
      "grad_norm": 3.302018165588379,
      "learning_rate": 2e-05,
      "loss": 0.1685,
      "step": 7657
    },
    {
      "epoch": 5.174383237580264,
      "grad_norm": 7.469437599182129,
      "learning_rate": 2e-05,
      "loss": 0.1464,
      "step": 7658
    },
    {
      "epoch": 5.175059141601893,
      "grad_norm": 2.237062931060791,
      "learning_rate": 2e-05,
      "loss": 0.1329,
      "step": 7659
    },
    {
      "epoch": 5.175735045623521,
      "grad_norm": 5.377345085144043,
      "learning_rate": 2e-05,
      "loss": 0.1162,
      "step": 7660
    },
    {
      "epoch": 5.176410949645151,
      "grad_norm": 2.4558069705963135,
      "learning_rate": 2e-05,
      "loss": 0.1571,
      "step": 7661
    },
    {
      "epoch": 5.177086853666779,
      "grad_norm": 2.645296573638916,
      "learning_rate": 2e-05,
      "loss": 0.1876,
      "step": 7662
    },
    {
      "epoch": 5.177762757688408,
      "grad_norm": 3.978898048400879,
      "learning_rate": 2e-05,
      "loss": 0.1783,
      "step": 7663
    },
    {
      "epoch": 5.178438661710037,
      "grad_norm": 3.816423177719116,
      "learning_rate": 2e-05,
      "loss": 0.1779,
      "step": 7664
    },
    {
      "epoch": 5.179114565731666,
      "grad_norm": 2.7956128120422363,
      "learning_rate": 2e-05,
      "loss": 0.1718,
      "step": 7665
    },
    {
      "epoch": 5.179790469753295,
      "grad_norm": 6.248970985412598,
      "learning_rate": 2e-05,
      "loss": 0.1333,
      "step": 7666
    },
    {
      "epoch": 5.180466373774924,
      "grad_norm": 1.440669059753418,
      "learning_rate": 2e-05,
      "loss": 0.1831,
      "step": 7667
    },
    {
      "epoch": 5.181142277796553,
      "grad_norm": 6.7226386070251465,
      "learning_rate": 2e-05,
      "loss": 0.1431,
      "step": 7668
    },
    {
      "epoch": 5.181818181818182,
      "grad_norm": 4.016459941864014,
      "learning_rate": 2e-05,
      "loss": 0.1577,
      "step": 7669
    },
    {
      "epoch": 5.182494085839811,
      "grad_norm": 3.485485553741455,
      "learning_rate": 2e-05,
      "loss": 0.171,
      "step": 7670
    },
    {
      "epoch": 5.183169989861439,
      "grad_norm": 2.9946506023406982,
      "learning_rate": 2e-05,
      "loss": 0.1372,
      "step": 7671
    },
    {
      "epoch": 5.183845893883069,
      "grad_norm": 4.152590751647949,
      "learning_rate": 2e-05,
      "loss": 0.1972,
      "step": 7672
    },
    {
      "epoch": 5.184521797904697,
      "grad_norm": 5.079774379730225,
      "learning_rate": 2e-05,
      "loss": 0.2359,
      "step": 7673
    },
    {
      "epoch": 5.1851977019263265,
      "grad_norm": 4.4336724281311035,
      "learning_rate": 2e-05,
      "loss": 0.1188,
      "step": 7674
    },
    {
      "epoch": 5.185873605947956,
      "grad_norm": 3.77850604057312,
      "learning_rate": 2e-05,
      "loss": 0.1608,
      "step": 7675
    },
    {
      "epoch": 5.186549509969584,
      "grad_norm": 3.0610527992248535,
      "learning_rate": 2e-05,
      "loss": 0.1477,
      "step": 7676
    },
    {
      "epoch": 5.1872254139912135,
      "grad_norm": 2.2420618534088135,
      "learning_rate": 2e-05,
      "loss": 0.1351,
      "step": 7677
    },
    {
      "epoch": 5.187901318012842,
      "grad_norm": 3.4540586471557617,
      "learning_rate": 2e-05,
      "loss": 0.2069,
      "step": 7678
    },
    {
      "epoch": 5.188577222034471,
      "grad_norm": 4.023447036743164,
      "learning_rate": 2e-05,
      "loss": 0.1855,
      "step": 7679
    },
    {
      "epoch": 5.1892531260561,
      "grad_norm": 2.8462045192718506,
      "learning_rate": 2e-05,
      "loss": 0.168,
      "step": 7680
    },
    {
      "epoch": 5.189929030077729,
      "grad_norm": 2.0468952655792236,
      "learning_rate": 2e-05,
      "loss": 0.1557,
      "step": 7681
    },
    {
      "epoch": 5.190604934099358,
      "grad_norm": 1.7476967573165894,
      "learning_rate": 2e-05,
      "loss": 0.1209,
      "step": 7682
    },
    {
      "epoch": 5.191280838120987,
      "grad_norm": 3.41312837600708,
      "learning_rate": 2e-05,
      "loss": 0.1496,
      "step": 7683
    },
    {
      "epoch": 5.191956742142616,
      "grad_norm": 1.745936393737793,
      "learning_rate": 2e-05,
      "loss": 0.1906,
      "step": 7684
    },
    {
      "epoch": 5.1926326461642445,
      "grad_norm": 4.768367290496826,
      "learning_rate": 2e-05,
      "loss": 0.1782,
      "step": 7685
    },
    {
      "epoch": 5.193308550185874,
      "grad_norm": 3.4811978340148926,
      "learning_rate": 2e-05,
      "loss": 0.0817,
      "step": 7686
    },
    {
      "epoch": 5.193984454207502,
      "grad_norm": 1.4556931257247925,
      "learning_rate": 2e-05,
      "loss": 0.1504,
      "step": 7687
    },
    {
      "epoch": 5.1946603582291315,
      "grad_norm": 3.320396900177002,
      "learning_rate": 2e-05,
      "loss": 0.1719,
      "step": 7688
    },
    {
      "epoch": 5.195336262250761,
      "grad_norm": 3.2905373573303223,
      "learning_rate": 2e-05,
      "loss": 0.1349,
      "step": 7689
    },
    {
      "epoch": 5.196012166272389,
      "grad_norm": 2.3605029582977295,
      "learning_rate": 2e-05,
      "loss": 0.1501,
      "step": 7690
    },
    {
      "epoch": 5.196688070294019,
      "grad_norm": 1.9377374649047852,
      "learning_rate": 2e-05,
      "loss": 0.1159,
      "step": 7691
    },
    {
      "epoch": 5.197363974315647,
      "grad_norm": 2.2668142318725586,
      "learning_rate": 2e-05,
      "loss": 0.0939,
      "step": 7692
    },
    {
      "epoch": 5.198039878337276,
      "grad_norm": 2.7213973999023438,
      "learning_rate": 2e-05,
      "loss": 0.1353,
      "step": 7693
    },
    {
      "epoch": 5.198715782358905,
      "grad_norm": 1.528742790222168,
      "learning_rate": 2e-05,
      "loss": 0.1224,
      "step": 7694
    },
    {
      "epoch": 5.199391686380534,
      "grad_norm": 3.151033639907837,
      "learning_rate": 2e-05,
      "loss": 0.1571,
      "step": 7695
    },
    {
      "epoch": 5.2000675904021625,
      "grad_norm": 2.1961610317230225,
      "learning_rate": 2e-05,
      "loss": 0.1055,
      "step": 7696
    },
    {
      "epoch": 5.200743494423792,
      "grad_norm": 2.453498125076294,
      "learning_rate": 2e-05,
      "loss": 0.1836,
      "step": 7697
    },
    {
      "epoch": 5.201419398445421,
      "grad_norm": 6.001178741455078,
      "learning_rate": 2e-05,
      "loss": 0.1923,
      "step": 7698
    },
    {
      "epoch": 5.2020953024670495,
      "grad_norm": 4.635928630828857,
      "learning_rate": 2e-05,
      "loss": 0.178,
      "step": 7699
    },
    {
      "epoch": 5.202771206488679,
      "grad_norm": 3.9158737659454346,
      "learning_rate": 2e-05,
      "loss": 0.1847,
      "step": 7700
    },
    {
      "epoch": 5.202771206488679,
      "eval_accuracy": 0.7792229729729729,
      "eval_loss": 0.23292586207389832,
      "eval_runtime": 46.0134,
      "eval_samples_per_second": 514.632,
      "eval_steps_per_second": 16.082,
      "step": 7700
    },
    {
      "epoch": 5.203447110510307,
      "grad_norm": 2.607198476791382,
      "learning_rate": 2e-05,
      "loss": 0.1683,
      "step": 7701
    },
    {
      "epoch": 5.204123014531937,
      "grad_norm": 4.988353729248047,
      "learning_rate": 2e-05,
      "loss": 0.1635,
      "step": 7702
    },
    {
      "epoch": 5.204798918553565,
      "grad_norm": 4.452739238739014,
      "learning_rate": 2e-05,
      "loss": 0.1569,
      "step": 7703
    },
    {
      "epoch": 5.205474822575194,
      "grad_norm": 2.1073319911956787,
      "learning_rate": 2e-05,
      "loss": 0.1421,
      "step": 7704
    },
    {
      "epoch": 5.206150726596824,
      "grad_norm": 6.001819133758545,
      "learning_rate": 2e-05,
      "loss": 0.1945,
      "step": 7705
    },
    {
      "epoch": 5.206826630618452,
      "grad_norm": 2.9234039783477783,
      "learning_rate": 2e-05,
      "loss": 0.1996,
      "step": 7706
    },
    {
      "epoch": 5.207502534640081,
      "grad_norm": 2.1493899822235107,
      "learning_rate": 2e-05,
      "loss": 0.1341,
      "step": 7707
    },
    {
      "epoch": 5.20817843866171,
      "grad_norm": 6.074041843414307,
      "learning_rate": 2e-05,
      "loss": 0.1437,
      "step": 7708
    },
    {
      "epoch": 5.208854342683339,
      "grad_norm": 1.829634666442871,
      "learning_rate": 2e-05,
      "loss": 0.1871,
      "step": 7709
    },
    {
      "epoch": 5.209530246704968,
      "grad_norm": 4.157654285430908,
      "learning_rate": 2e-05,
      "loss": 0.1428,
      "step": 7710
    },
    {
      "epoch": 5.210206150726597,
      "grad_norm": 2.2117156982421875,
      "learning_rate": 2e-05,
      "loss": 0.2157,
      "step": 7711
    },
    {
      "epoch": 5.210882054748225,
      "grad_norm": 5.623961925506592,
      "learning_rate": 2e-05,
      "loss": 0.2092,
      "step": 7712
    },
    {
      "epoch": 5.211557958769855,
      "grad_norm": 2.681910276412964,
      "learning_rate": 2e-05,
      "loss": 0.1187,
      "step": 7713
    },
    {
      "epoch": 5.212233862791484,
      "grad_norm": 2.7120375633239746,
      "learning_rate": 2e-05,
      "loss": 0.1478,
      "step": 7714
    },
    {
      "epoch": 5.212909766813112,
      "grad_norm": 2.5515620708465576,
      "learning_rate": 2e-05,
      "loss": 0.1993,
      "step": 7715
    },
    {
      "epoch": 5.213585670834742,
      "grad_norm": 3.0786142349243164,
      "learning_rate": 2e-05,
      "loss": 0.1164,
      "step": 7716
    },
    {
      "epoch": 5.21426157485637,
      "grad_norm": 1.7009739875793457,
      "learning_rate": 2e-05,
      "loss": 0.1279,
      "step": 7717
    },
    {
      "epoch": 5.214937478877999,
      "grad_norm": 2.8872344493865967,
      "learning_rate": 2e-05,
      "loss": 0.2059,
      "step": 7718
    },
    {
      "epoch": 5.215613382899628,
      "grad_norm": 2.7623000144958496,
      "learning_rate": 2e-05,
      "loss": 0.1457,
      "step": 7719
    },
    {
      "epoch": 5.216289286921257,
      "grad_norm": 2.347470283508301,
      "learning_rate": 2e-05,
      "loss": 0.1941,
      "step": 7720
    },
    {
      "epoch": 5.2169651909428865,
      "grad_norm": 2.356278896331787,
      "learning_rate": 2e-05,
      "loss": 0.1768,
      "step": 7721
    },
    {
      "epoch": 5.217641094964515,
      "grad_norm": 4.292178153991699,
      "learning_rate": 2e-05,
      "loss": 0.1655,
      "step": 7722
    },
    {
      "epoch": 5.218316998986144,
      "grad_norm": 1.9170118570327759,
      "learning_rate": 2e-05,
      "loss": 0.171,
      "step": 7723
    },
    {
      "epoch": 5.218992903007773,
      "grad_norm": 3.993256092071533,
      "learning_rate": 2e-05,
      "loss": 0.1704,
      "step": 7724
    },
    {
      "epoch": 5.219668807029402,
      "grad_norm": 3.2520413398742676,
      "learning_rate": 2e-05,
      "loss": 0.1651,
      "step": 7725
    },
    {
      "epoch": 5.22034471105103,
      "grad_norm": 3.4597980976104736,
      "learning_rate": 2e-05,
      "loss": 0.1496,
      "step": 7726
    },
    {
      "epoch": 5.22102061507266,
      "grad_norm": 4.235325336456299,
      "learning_rate": 2e-05,
      "loss": 0.183,
      "step": 7727
    },
    {
      "epoch": 5.221696519094289,
      "grad_norm": 4.302570819854736,
      "learning_rate": 2e-05,
      "loss": 0.1579,
      "step": 7728
    },
    {
      "epoch": 5.222372423115917,
      "grad_norm": 2.6642370223999023,
      "learning_rate": 2e-05,
      "loss": 0.1443,
      "step": 7729
    },
    {
      "epoch": 5.223048327137547,
      "grad_norm": 1.6116433143615723,
      "learning_rate": 2e-05,
      "loss": 0.1285,
      "step": 7730
    },
    {
      "epoch": 5.223724231159175,
      "grad_norm": 1.5634629726409912,
      "learning_rate": 2e-05,
      "loss": 0.1287,
      "step": 7731
    },
    {
      "epoch": 5.2244001351808045,
      "grad_norm": 4.59535026550293,
      "learning_rate": 2e-05,
      "loss": 0.1887,
      "step": 7732
    },
    {
      "epoch": 5.225076039202433,
      "grad_norm": 5.815956115722656,
      "learning_rate": 2e-05,
      "loss": 0.2417,
      "step": 7733
    },
    {
      "epoch": 5.225751943224062,
      "grad_norm": 3.283252477645874,
      "learning_rate": 2e-05,
      "loss": 0.1961,
      "step": 7734
    },
    {
      "epoch": 5.2264278472456915,
      "grad_norm": 1.7255275249481201,
      "learning_rate": 2e-05,
      "loss": 0.1843,
      "step": 7735
    },
    {
      "epoch": 5.22710375126732,
      "grad_norm": 3.513934373855591,
      "learning_rate": 2e-05,
      "loss": 0.1391,
      "step": 7736
    },
    {
      "epoch": 5.227779655288949,
      "grad_norm": 5.174921989440918,
      "learning_rate": 2e-05,
      "loss": 0.1427,
      "step": 7737
    },
    {
      "epoch": 5.228455559310578,
      "grad_norm": 2.1740567684173584,
      "learning_rate": 2e-05,
      "loss": 0.1025,
      "step": 7738
    },
    {
      "epoch": 5.229131463332207,
      "grad_norm": 4.982848167419434,
      "learning_rate": 2e-05,
      "loss": 0.1306,
      "step": 7739
    },
    {
      "epoch": 5.2298073673538354,
      "grad_norm": 2.4152607917785645,
      "learning_rate": 2e-05,
      "loss": 0.1337,
      "step": 7740
    },
    {
      "epoch": 5.230483271375465,
      "grad_norm": 2.987527370452881,
      "learning_rate": 2e-05,
      "loss": 0.1595,
      "step": 7741
    },
    {
      "epoch": 5.231159175397093,
      "grad_norm": 4.380283832550049,
      "learning_rate": 2e-05,
      "loss": 0.161,
      "step": 7742
    },
    {
      "epoch": 5.2318350794187225,
      "grad_norm": 4.0562849044799805,
      "learning_rate": 2e-05,
      "loss": 0.1552,
      "step": 7743
    },
    {
      "epoch": 5.232510983440352,
      "grad_norm": 2.0913281440734863,
      "learning_rate": 2e-05,
      "loss": 0.1274,
      "step": 7744
    },
    {
      "epoch": 5.23318688746198,
      "grad_norm": 3.852877140045166,
      "learning_rate": 2e-05,
      "loss": 0.1126,
      "step": 7745
    },
    {
      "epoch": 5.2338627914836096,
      "grad_norm": 3.46061635017395,
      "learning_rate": 2e-05,
      "loss": 0.1595,
      "step": 7746
    },
    {
      "epoch": 5.234538695505238,
      "grad_norm": 4.369128227233887,
      "learning_rate": 2e-05,
      "loss": 0.2126,
      "step": 7747
    },
    {
      "epoch": 5.235214599526867,
      "grad_norm": 2.5409908294677734,
      "learning_rate": 2e-05,
      "loss": 0.1552,
      "step": 7748
    },
    {
      "epoch": 5.235890503548496,
      "grad_norm": 3.379042625427246,
      "learning_rate": 2e-05,
      "loss": 0.0765,
      "step": 7749
    },
    {
      "epoch": 5.236566407570125,
      "grad_norm": 4.122787952423096,
      "learning_rate": 2e-05,
      "loss": 0.1827,
      "step": 7750
    },
    {
      "epoch": 5.236566407570125,
      "eval_accuracy": 0.7740709459459459,
      "eval_loss": 0.224008709192276,
      "eval_runtime": 46.808,
      "eval_samples_per_second": 505.897,
      "eval_steps_per_second": 15.809,
      "step": 7750
    },
    {
      "epoch": 5.237242311591754,
      "grad_norm": 3.340651035308838,
      "learning_rate": 2e-05,
      "loss": 0.138,
      "step": 7751
    },
    {
      "epoch": 5.237918215613383,
      "grad_norm": 3.1696698665618896,
      "learning_rate": 2e-05,
      "loss": 0.188,
      "step": 7752
    },
    {
      "epoch": 5.238594119635012,
      "grad_norm": 4.270773887634277,
      "learning_rate": 2e-05,
      "loss": 0.1779,
      "step": 7753
    },
    {
      "epoch": 5.2392700236566405,
      "grad_norm": 3.622224807739258,
      "learning_rate": 2e-05,
      "loss": 0.1563,
      "step": 7754
    },
    {
      "epoch": 5.23994592767827,
      "grad_norm": 2.0196430683135986,
      "learning_rate": 2e-05,
      "loss": 0.1464,
      "step": 7755
    },
    {
      "epoch": 5.240621831699898,
      "grad_norm": 2.2345926761627197,
      "learning_rate": 2e-05,
      "loss": 0.1729,
      "step": 7756
    },
    {
      "epoch": 5.241297735721528,
      "grad_norm": 2.204047203063965,
      "learning_rate": 2e-05,
      "loss": 0.1165,
      "step": 7757
    },
    {
      "epoch": 5.241973639743157,
      "grad_norm": 3.9162979125976562,
      "learning_rate": 2e-05,
      "loss": 0.1562,
      "step": 7758
    },
    {
      "epoch": 5.242649543764785,
      "grad_norm": 2.490898609161377,
      "learning_rate": 2e-05,
      "loss": 0.0861,
      "step": 7759
    },
    {
      "epoch": 5.243325447786415,
      "grad_norm": 3.179051399230957,
      "learning_rate": 2e-05,
      "loss": 0.1214,
      "step": 7760
    },
    {
      "epoch": 5.244001351808043,
      "grad_norm": 5.54720401763916,
      "learning_rate": 2e-05,
      "loss": 0.1197,
      "step": 7761
    },
    {
      "epoch": 5.244677255829672,
      "grad_norm": 2.661472797393799,
      "learning_rate": 2e-05,
      "loss": 0.1674,
      "step": 7762
    },
    {
      "epoch": 5.245353159851301,
      "grad_norm": 2.5141713619232178,
      "learning_rate": 2e-05,
      "loss": 0.1164,
      "step": 7763
    },
    {
      "epoch": 5.24602906387293,
      "grad_norm": 2.9590749740600586,
      "learning_rate": 2e-05,
      "loss": 0.1258,
      "step": 7764
    },
    {
      "epoch": 5.2467049678945585,
      "grad_norm": 3.6938023567199707,
      "learning_rate": 2e-05,
      "loss": 0.0857,
      "step": 7765
    },
    {
      "epoch": 5.247380871916188,
      "grad_norm": 4.803613662719727,
      "learning_rate": 2e-05,
      "loss": 0.1761,
      "step": 7766
    },
    {
      "epoch": 5.248056775937817,
      "grad_norm": 6.768067359924316,
      "learning_rate": 2e-05,
      "loss": 0.1291,
      "step": 7767
    },
    {
      "epoch": 5.248732679959446,
      "grad_norm": 1.8104239702224731,
      "learning_rate": 2e-05,
      "loss": 0.1378,
      "step": 7768
    },
    {
      "epoch": 5.249408583981075,
      "grad_norm": 4.549561023712158,
      "learning_rate": 2e-05,
      "loss": 0.1049,
      "step": 7769
    },
    {
      "epoch": 5.250084488002703,
      "grad_norm": 2.100612163543701,
      "learning_rate": 2e-05,
      "loss": 0.0873,
      "step": 7770
    },
    {
      "epoch": 5.250760392024333,
      "grad_norm": 2.343782424926758,
      "learning_rate": 2e-05,
      "loss": 0.1902,
      "step": 7771
    },
    {
      "epoch": 5.251436296045961,
      "grad_norm": 4.362542629241943,
      "learning_rate": 2e-05,
      "loss": 0.1631,
      "step": 7772
    },
    {
      "epoch": 5.25211220006759,
      "grad_norm": 4.185332775115967,
      "learning_rate": 2e-05,
      "loss": 0.2068,
      "step": 7773
    },
    {
      "epoch": 5.25278810408922,
      "grad_norm": 11.091008186340332,
      "learning_rate": 2e-05,
      "loss": 0.1724,
      "step": 7774
    },
    {
      "epoch": 5.253464008110848,
      "grad_norm": 7.003860950469971,
      "learning_rate": 2e-05,
      "loss": 0.1597,
      "step": 7775
    },
    {
      "epoch": 5.254139912132477,
      "grad_norm": 5.205738544464111,
      "learning_rate": 2e-05,
      "loss": 0.1044,
      "step": 7776
    },
    {
      "epoch": 5.254815816154106,
      "grad_norm": 8.612655639648438,
      "learning_rate": 2e-05,
      "loss": 0.1848,
      "step": 7777
    },
    {
      "epoch": 5.255491720175735,
      "grad_norm": 2.4999399185180664,
      "learning_rate": 2e-05,
      "loss": 0.1673,
      "step": 7778
    },
    {
      "epoch": 5.256167624197364,
      "grad_norm": 3.071669340133667,
      "learning_rate": 2e-05,
      "loss": 0.1367,
      "step": 7779
    },
    {
      "epoch": 5.256843528218993,
      "grad_norm": 3.2939555644989014,
      "learning_rate": 2e-05,
      "loss": 0.0944,
      "step": 7780
    },
    {
      "epoch": 5.257519432240622,
      "grad_norm": 1.5121705532073975,
      "learning_rate": 2e-05,
      "loss": 0.1031,
      "step": 7781
    },
    {
      "epoch": 5.258195336262251,
      "grad_norm": 3.35766339302063,
      "learning_rate": 2e-05,
      "loss": 0.2041,
      "step": 7782
    },
    {
      "epoch": 5.25887124028388,
      "grad_norm": 2.3838374614715576,
      "learning_rate": 2e-05,
      "loss": 0.1304,
      "step": 7783
    },
    {
      "epoch": 5.259547144305508,
      "grad_norm": 4.535953521728516,
      "learning_rate": 2e-05,
      "loss": 0.1863,
      "step": 7784
    },
    {
      "epoch": 5.260223048327138,
      "grad_norm": 5.58568000793457,
      "learning_rate": 2e-05,
      "loss": 0.1873,
      "step": 7785
    },
    {
      "epoch": 5.260898952348766,
      "grad_norm": 3.9569578170776367,
      "learning_rate": 2e-05,
      "loss": 0.102,
      "step": 7786
    },
    {
      "epoch": 5.2615748563703955,
      "grad_norm": 1.4250510931015015,
      "learning_rate": 2e-05,
      "loss": 0.1065,
      "step": 7787
    },
    {
      "epoch": 5.262250760392025,
      "grad_norm": 3.4730563163757324,
      "learning_rate": 2e-05,
      "loss": 0.1118,
      "step": 7788
    },
    {
      "epoch": 5.262926664413653,
      "grad_norm": 4.936408042907715,
      "learning_rate": 2e-05,
      "loss": 0.1906,
      "step": 7789
    },
    {
      "epoch": 5.2636025684352825,
      "grad_norm": 2.333921432495117,
      "learning_rate": 2e-05,
      "loss": 0.1326,
      "step": 7790
    },
    {
      "epoch": 5.264278472456911,
      "grad_norm": 2.2163355350494385,
      "learning_rate": 2e-05,
      "loss": 0.1246,
      "step": 7791
    },
    {
      "epoch": 5.26495437647854,
      "grad_norm": 1.5227880477905273,
      "learning_rate": 2e-05,
      "loss": 0.1179,
      "step": 7792
    },
    {
      "epoch": 5.265630280500169,
      "grad_norm": 3.9934394359588623,
      "learning_rate": 2e-05,
      "loss": 0.1496,
      "step": 7793
    },
    {
      "epoch": 5.266306184521798,
      "grad_norm": 5.392880916595459,
      "learning_rate": 2e-05,
      "loss": 0.1634,
      "step": 7794
    },
    {
      "epoch": 5.266982088543426,
      "grad_norm": 7.169835090637207,
      "learning_rate": 2e-05,
      "loss": 0.1451,
      "step": 7795
    },
    {
      "epoch": 5.267657992565056,
      "grad_norm": 1.6798864603042603,
      "learning_rate": 2e-05,
      "loss": 0.1687,
      "step": 7796
    },
    {
      "epoch": 5.268333896586685,
      "grad_norm": 2.9034018516540527,
      "learning_rate": 2e-05,
      "loss": 0.1542,
      "step": 7797
    },
    {
      "epoch": 5.2690098006083135,
      "grad_norm": 2.421116828918457,
      "learning_rate": 2e-05,
      "loss": 0.2033,
      "step": 7798
    },
    {
      "epoch": 5.269685704629943,
      "grad_norm": 2.206718683242798,
      "learning_rate": 2e-05,
      "loss": 0.1464,
      "step": 7799
    },
    {
      "epoch": 5.270361608651571,
      "grad_norm": 5.3825201988220215,
      "learning_rate": 2e-05,
      "loss": 0.1656,
      "step": 7800
    },
    {
      "epoch": 5.270361608651571,
      "eval_accuracy": 0.7547297297297297,
      "eval_loss": 0.23314695060253143,
      "eval_runtime": 45.1163,
      "eval_samples_per_second": 524.865,
      "eval_steps_per_second": 16.402,
      "step": 7800
    },
    {
      "epoch": 5.2710375126732005,
      "grad_norm": 7.410736560821533,
      "learning_rate": 2e-05,
      "loss": 0.1796,
      "step": 7801
    },
    {
      "epoch": 5.271713416694829,
      "grad_norm": 3.097135066986084,
      "learning_rate": 2e-05,
      "loss": 0.1414,
      "step": 7802
    },
    {
      "epoch": 5.272389320716458,
      "grad_norm": 3.3226327896118164,
      "learning_rate": 2e-05,
      "loss": 0.1306,
      "step": 7803
    },
    {
      "epoch": 5.273065224738088,
      "grad_norm": 2.2362101078033447,
      "learning_rate": 2e-05,
      "loss": 0.1726,
      "step": 7804
    },
    {
      "epoch": 5.273741128759716,
      "grad_norm": 3.5470709800720215,
      "learning_rate": 2e-05,
      "loss": 0.0985,
      "step": 7805
    },
    {
      "epoch": 5.274417032781345,
      "grad_norm": 2.6714799404144287,
      "learning_rate": 2e-05,
      "loss": 0.1509,
      "step": 7806
    },
    {
      "epoch": 5.275092936802974,
      "grad_norm": 4.122203350067139,
      "learning_rate": 2e-05,
      "loss": 0.1617,
      "step": 7807
    },
    {
      "epoch": 5.275768840824603,
      "grad_norm": 1.896407127380371,
      "learning_rate": 2e-05,
      "loss": 0.1628,
      "step": 7808
    },
    {
      "epoch": 5.2764447448462315,
      "grad_norm": 1.9419585466384888,
      "learning_rate": 2e-05,
      "loss": 0.163,
      "step": 7809
    },
    {
      "epoch": 5.277120648867861,
      "grad_norm": 2.418752431869507,
      "learning_rate": 2e-05,
      "loss": 0.1255,
      "step": 7810
    },
    {
      "epoch": 5.277796552889489,
      "grad_norm": 3.6051857471466064,
      "learning_rate": 2e-05,
      "loss": 0.2276,
      "step": 7811
    },
    {
      "epoch": 5.2784724569111185,
      "grad_norm": 6.553763389587402,
      "learning_rate": 2e-05,
      "loss": 0.1376,
      "step": 7812
    },
    {
      "epoch": 5.279148360932748,
      "grad_norm": 2.0427801609039307,
      "learning_rate": 2e-05,
      "loss": 0.1201,
      "step": 7813
    },
    {
      "epoch": 5.279824264954376,
      "grad_norm": 2.534533977508545,
      "learning_rate": 2e-05,
      "loss": 0.1783,
      "step": 7814
    },
    {
      "epoch": 5.280500168976006,
      "grad_norm": 4.691914081573486,
      "learning_rate": 2e-05,
      "loss": 0.154,
      "step": 7815
    },
    {
      "epoch": 5.281176072997634,
      "grad_norm": 3.283292531967163,
      "learning_rate": 2e-05,
      "loss": 0.1653,
      "step": 7816
    },
    {
      "epoch": 5.281851977019263,
      "grad_norm": 4.686107635498047,
      "learning_rate": 2e-05,
      "loss": 0.164,
      "step": 7817
    },
    {
      "epoch": 5.282527881040892,
      "grad_norm": 2.838021993637085,
      "learning_rate": 2e-05,
      "loss": 0.1395,
      "step": 7818
    },
    {
      "epoch": 5.283203785062521,
      "grad_norm": 2.384554624557495,
      "learning_rate": 2e-05,
      "loss": 0.1583,
      "step": 7819
    },
    {
      "epoch": 5.28387968908415,
      "grad_norm": 4.725687026977539,
      "learning_rate": 2e-05,
      "loss": 0.1468,
      "step": 7820
    },
    {
      "epoch": 5.284555593105779,
      "grad_norm": 4.343028545379639,
      "learning_rate": 2e-05,
      "loss": 0.1521,
      "step": 7821
    },
    {
      "epoch": 5.285231497127408,
      "grad_norm": 2.9689042568206787,
      "learning_rate": 2e-05,
      "loss": 0.2214,
      "step": 7822
    },
    {
      "epoch": 5.285907401149037,
      "grad_norm": 2.6109619140625,
      "learning_rate": 2e-05,
      "loss": 0.1135,
      "step": 7823
    },
    {
      "epoch": 5.286583305170666,
      "grad_norm": 2.253833770751953,
      "learning_rate": 2e-05,
      "loss": 0.1832,
      "step": 7824
    },
    {
      "epoch": 5.287259209192294,
      "grad_norm": 1.9873716831207275,
      "learning_rate": 2e-05,
      "loss": 0.1663,
      "step": 7825
    },
    {
      "epoch": 5.287935113213924,
      "grad_norm": 4.689773082733154,
      "learning_rate": 2e-05,
      "loss": 0.1023,
      "step": 7826
    },
    {
      "epoch": 5.288611017235553,
      "grad_norm": 2.6034796237945557,
      "learning_rate": 2e-05,
      "loss": 0.2113,
      "step": 7827
    },
    {
      "epoch": 5.289286921257181,
      "grad_norm": 2.9998409748077393,
      "learning_rate": 2e-05,
      "loss": 0.166,
      "step": 7828
    },
    {
      "epoch": 5.289962825278811,
      "grad_norm": 1.9468022584915161,
      "learning_rate": 2e-05,
      "loss": 0.149,
      "step": 7829
    },
    {
      "epoch": 5.290638729300439,
      "grad_norm": 2.0202958583831787,
      "learning_rate": 2e-05,
      "loss": 0.192,
      "step": 7830
    },
    {
      "epoch": 5.291314633322068,
      "grad_norm": 2.3394505977630615,
      "learning_rate": 2e-05,
      "loss": 0.0858,
      "step": 7831
    },
    {
      "epoch": 5.291990537343697,
      "grad_norm": 3.1548824310302734,
      "learning_rate": 2e-05,
      "loss": 0.1429,
      "step": 7832
    },
    {
      "epoch": 5.292666441365326,
      "grad_norm": 1.5719163417816162,
      "learning_rate": 2e-05,
      "loss": 0.1688,
      "step": 7833
    },
    {
      "epoch": 5.2933423453869555,
      "grad_norm": 2.0011210441589355,
      "learning_rate": 2e-05,
      "loss": 0.1172,
      "step": 7834
    },
    {
      "epoch": 5.294018249408584,
      "grad_norm": 4.25998592376709,
      "learning_rate": 2e-05,
      "loss": 0.1236,
      "step": 7835
    },
    {
      "epoch": 5.294694153430213,
      "grad_norm": 4.332117080688477,
      "learning_rate": 2e-05,
      "loss": 0.1213,
      "step": 7836
    },
    {
      "epoch": 5.295370057451842,
      "grad_norm": 2.200256586074829,
      "learning_rate": 2e-05,
      "loss": 0.2003,
      "step": 7837
    },
    {
      "epoch": 5.296045961473471,
      "grad_norm": 4.093990325927734,
      "learning_rate": 2e-05,
      "loss": 0.1542,
      "step": 7838
    },
    {
      "epoch": 5.296721865495099,
      "grad_norm": 4.966525554656982,
      "learning_rate": 2e-05,
      "loss": 0.1819,
      "step": 7839
    },
    {
      "epoch": 5.297397769516729,
      "grad_norm": 6.960391044616699,
      "learning_rate": 2e-05,
      "loss": 0.1803,
      "step": 7840
    },
    {
      "epoch": 5.298073673538358,
      "grad_norm": 4.510560035705566,
      "learning_rate": 2e-05,
      "loss": 0.1043,
      "step": 7841
    },
    {
      "epoch": 5.298749577559986,
      "grad_norm": 4.054383277893066,
      "learning_rate": 2e-05,
      "loss": 0.2101,
      "step": 7842
    },
    {
      "epoch": 5.299425481581616,
      "grad_norm": 2.3288893699645996,
      "learning_rate": 2e-05,
      "loss": 0.1264,
      "step": 7843
    },
    {
      "epoch": 5.300101385603244,
      "grad_norm": 3.783109426498413,
      "learning_rate": 2e-05,
      "loss": 0.1367,
      "step": 7844
    },
    {
      "epoch": 5.3007772896248735,
      "grad_norm": 2.581752061843872,
      "learning_rate": 2e-05,
      "loss": 0.1409,
      "step": 7845
    },
    {
      "epoch": 5.301453193646502,
      "grad_norm": 2.727297782897949,
      "learning_rate": 2e-05,
      "loss": 0.1804,
      "step": 7846
    },
    {
      "epoch": 5.302129097668131,
      "grad_norm": 2.9843616485595703,
      "learning_rate": 2e-05,
      "loss": 0.0807,
      "step": 7847
    },
    {
      "epoch": 5.30280500168976,
      "grad_norm": 5.421690464019775,
      "learning_rate": 2e-05,
      "loss": 0.1731,
      "step": 7848
    },
    {
      "epoch": 5.303480905711389,
      "grad_norm": 2.693195104598999,
      "learning_rate": 2e-05,
      "loss": 0.1964,
      "step": 7849
    },
    {
      "epoch": 5.304156809733018,
      "grad_norm": 2.47967267036438,
      "learning_rate": 2e-05,
      "loss": 0.1655,
      "step": 7850
    },
    {
      "epoch": 5.304156809733018,
      "eval_accuracy": 0.7727618243243243,
      "eval_loss": 0.24182885885238647,
      "eval_runtime": 47.7928,
      "eval_samples_per_second": 495.472,
      "eval_steps_per_second": 15.484,
      "step": 7850
    },
    {
      "epoch": 5.304832713754647,
      "grad_norm": 3.034242630004883,
      "learning_rate": 2e-05,
      "loss": 0.2324,
      "step": 7851
    },
    {
      "epoch": 5.305508617776276,
      "grad_norm": 2.6198153495788574,
      "learning_rate": 2e-05,
      "loss": 0.1958,
      "step": 7852
    },
    {
      "epoch": 5.3061845217979045,
      "grad_norm": 3.8410730361938477,
      "learning_rate": 2e-05,
      "loss": 0.122,
      "step": 7853
    },
    {
      "epoch": 5.306860425819534,
      "grad_norm": 1.7651058435440063,
      "learning_rate": 2e-05,
      "loss": 0.197,
      "step": 7854
    },
    {
      "epoch": 5.307536329841162,
      "grad_norm": 2.5943515300750732,
      "learning_rate": 2e-05,
      "loss": 0.1558,
      "step": 7855
    },
    {
      "epoch": 5.3082122338627915,
      "grad_norm": 4.948856830596924,
      "learning_rate": 2e-05,
      "loss": 0.1743,
      "step": 7856
    },
    {
      "epoch": 5.308888137884421,
      "grad_norm": 3.422135829925537,
      "learning_rate": 2e-05,
      "loss": 0.1548,
      "step": 7857
    },
    {
      "epoch": 5.309564041906049,
      "grad_norm": 1.5855540037155151,
      "learning_rate": 2e-05,
      "loss": 0.125,
      "step": 7858
    },
    {
      "epoch": 5.310239945927679,
      "grad_norm": 7.950244426727295,
      "learning_rate": 2e-05,
      "loss": 0.272,
      "step": 7859
    },
    {
      "epoch": 5.310915849949307,
      "grad_norm": 2.2730164527893066,
      "learning_rate": 2e-05,
      "loss": 0.1074,
      "step": 7860
    },
    {
      "epoch": 5.311591753970936,
      "grad_norm": 2.676892042160034,
      "learning_rate": 2e-05,
      "loss": 0.1476,
      "step": 7861
    },
    {
      "epoch": 5.312267657992565,
      "grad_norm": 1.875929594039917,
      "learning_rate": 2e-05,
      "loss": 0.1299,
      "step": 7862
    },
    {
      "epoch": 5.312943562014194,
      "grad_norm": 2.1491076946258545,
      "learning_rate": 2e-05,
      "loss": 0.1073,
      "step": 7863
    },
    {
      "epoch": 5.3136194660358225,
      "grad_norm": 3.7750210762023926,
      "learning_rate": 2e-05,
      "loss": 0.1671,
      "step": 7864
    },
    {
      "epoch": 5.314295370057452,
      "grad_norm": 3.729395866394043,
      "learning_rate": 2e-05,
      "loss": 0.1579,
      "step": 7865
    },
    {
      "epoch": 5.314971274079081,
      "grad_norm": 3.5847854614257812,
      "learning_rate": 2e-05,
      "loss": 0.1503,
      "step": 7866
    },
    {
      "epoch": 5.3156471781007095,
      "grad_norm": 2.267773151397705,
      "learning_rate": 2e-05,
      "loss": 0.2095,
      "step": 7867
    },
    {
      "epoch": 5.316323082122339,
      "grad_norm": 3.2098798751831055,
      "learning_rate": 2e-05,
      "loss": 0.1456,
      "step": 7868
    },
    {
      "epoch": 5.316998986143967,
      "grad_norm": 3.2272284030914307,
      "learning_rate": 2e-05,
      "loss": 0.1856,
      "step": 7869
    },
    {
      "epoch": 5.317674890165597,
      "grad_norm": 2.059460401535034,
      "learning_rate": 2e-05,
      "loss": 0.1505,
      "step": 7870
    },
    {
      "epoch": 5.318350794187225,
      "grad_norm": 8.380192756652832,
      "learning_rate": 2e-05,
      "loss": 0.156,
      "step": 7871
    },
    {
      "epoch": 5.319026698208854,
      "grad_norm": 3.4495134353637695,
      "learning_rate": 2e-05,
      "loss": 0.1396,
      "step": 7872
    },
    {
      "epoch": 5.319702602230484,
      "grad_norm": 1.5988106727600098,
      "learning_rate": 2e-05,
      "loss": 0.1427,
      "step": 7873
    },
    {
      "epoch": 5.320378506252112,
      "grad_norm": 2.1003403663635254,
      "learning_rate": 2e-05,
      "loss": 0.1311,
      "step": 7874
    },
    {
      "epoch": 5.321054410273741,
      "grad_norm": 2.8812108039855957,
      "learning_rate": 2e-05,
      "loss": 0.1832,
      "step": 7875
    },
    {
      "epoch": 5.32173031429537,
      "grad_norm": 5.099373817443848,
      "learning_rate": 2e-05,
      "loss": 0.1033,
      "step": 7876
    },
    {
      "epoch": 5.322406218316999,
      "grad_norm": 3.029226303100586,
      "learning_rate": 2e-05,
      "loss": 0.182,
      "step": 7877
    },
    {
      "epoch": 5.3230821223386275,
      "grad_norm": 2.542929172515869,
      "learning_rate": 2e-05,
      "loss": 0.1562,
      "step": 7878
    },
    {
      "epoch": 5.323758026360257,
      "grad_norm": 4.844904899597168,
      "learning_rate": 2e-05,
      "loss": 0.1645,
      "step": 7879
    },
    {
      "epoch": 5.324433930381886,
      "grad_norm": 5.973598480224609,
      "learning_rate": 2e-05,
      "loss": 0.1686,
      "step": 7880
    },
    {
      "epoch": 5.325109834403515,
      "grad_norm": 3.499695062637329,
      "learning_rate": 2e-05,
      "loss": 0.1586,
      "step": 7881
    },
    {
      "epoch": 5.325785738425144,
      "grad_norm": 2.0030157566070557,
      "learning_rate": 2e-05,
      "loss": 0.1319,
      "step": 7882
    },
    {
      "epoch": 5.326461642446772,
      "grad_norm": 2.4791159629821777,
      "learning_rate": 2e-05,
      "loss": 0.0968,
      "step": 7883
    },
    {
      "epoch": 5.327137546468402,
      "grad_norm": 2.602193832397461,
      "learning_rate": 2e-05,
      "loss": 0.1593,
      "step": 7884
    },
    {
      "epoch": 5.32781345049003,
      "grad_norm": 2.7834293842315674,
      "learning_rate": 2e-05,
      "loss": 0.1473,
      "step": 7885
    },
    {
      "epoch": 5.328489354511659,
      "grad_norm": 3.411999464035034,
      "learning_rate": 2e-05,
      "loss": 0.1916,
      "step": 7886
    },
    {
      "epoch": 5.329165258533289,
      "grad_norm": 2.5570743083953857,
      "learning_rate": 2e-05,
      "loss": 0.1569,
      "step": 7887
    },
    {
      "epoch": 5.329841162554917,
      "grad_norm": 2.2579636573791504,
      "learning_rate": 2e-05,
      "loss": 0.1765,
      "step": 7888
    },
    {
      "epoch": 5.3305170665765464,
      "grad_norm": 3.2423651218414307,
      "learning_rate": 2e-05,
      "loss": 0.1222,
      "step": 7889
    },
    {
      "epoch": 5.331192970598175,
      "grad_norm": 2.248751640319824,
      "learning_rate": 2e-05,
      "loss": 0.1376,
      "step": 7890
    },
    {
      "epoch": 5.331868874619804,
      "grad_norm": 4.842864990234375,
      "learning_rate": 2e-05,
      "loss": 0.1001,
      "step": 7891
    },
    {
      "epoch": 5.332544778641433,
      "grad_norm": 4.0180768966674805,
      "learning_rate": 2e-05,
      "loss": 0.1392,
      "step": 7892
    },
    {
      "epoch": 5.333220682663062,
      "grad_norm": 2.389572858810425,
      "learning_rate": 2e-05,
      "loss": 0.1646,
      "step": 7893
    },
    {
      "epoch": 5.333896586684691,
      "grad_norm": 3.979931592941284,
      "learning_rate": 2e-05,
      "loss": 0.2314,
      "step": 7894
    },
    {
      "epoch": 5.33457249070632,
      "grad_norm": 6.642794609069824,
      "learning_rate": 2e-05,
      "loss": 0.1193,
      "step": 7895
    },
    {
      "epoch": 5.335248394727949,
      "grad_norm": 3.402167320251465,
      "learning_rate": 2e-05,
      "loss": 0.12,
      "step": 7896
    },
    {
      "epoch": 5.335924298749577,
      "grad_norm": 3.138007640838623,
      "learning_rate": 2e-05,
      "loss": 0.1912,
      "step": 7897
    },
    {
      "epoch": 5.336600202771207,
      "grad_norm": 5.7851386070251465,
      "learning_rate": 2e-05,
      "loss": 0.1581,
      "step": 7898
    },
    {
      "epoch": 5.337276106792835,
      "grad_norm": 6.598158836364746,
      "learning_rate": 2e-05,
      "loss": 0.1871,
      "step": 7899
    },
    {
      "epoch": 5.3379520108144645,
      "grad_norm": 6.073897361755371,
      "learning_rate": 2e-05,
      "loss": 0.1334,
      "step": 7900
    },
    {
      "epoch": 5.3379520108144645,
      "eval_accuracy": 0.7560810810810811,
      "eval_loss": 0.2409779578447342,
      "eval_runtime": 45.8945,
      "eval_samples_per_second": 515.966,
      "eval_steps_per_second": 16.124,
      "step": 7900
    },
    {
      "epoch": 5.338627914836093,
      "grad_norm": 2.6977810859680176,
      "learning_rate": 2e-05,
      "loss": 0.1911,
      "step": 7901
    },
    {
      "epoch": 5.339303818857722,
      "grad_norm": 2.9232304096221924,
      "learning_rate": 2e-05,
      "loss": 0.1747,
      "step": 7902
    },
    {
      "epoch": 5.3399797228793515,
      "grad_norm": 3.8402035236358643,
      "learning_rate": 2e-05,
      "loss": 0.129,
      "step": 7903
    },
    {
      "epoch": 5.34065562690098,
      "grad_norm": 2.4316093921661377,
      "learning_rate": 2e-05,
      "loss": 0.1457,
      "step": 7904
    },
    {
      "epoch": 5.341331530922609,
      "grad_norm": 5.564116477966309,
      "learning_rate": 2e-05,
      "loss": 0.207,
      "step": 7905
    },
    {
      "epoch": 5.342007434944238,
      "grad_norm": 11.324166297912598,
      "learning_rate": 2e-05,
      "loss": 0.1649,
      "step": 7906
    },
    {
      "epoch": 5.342683338965867,
      "grad_norm": 2.9032819271087646,
      "learning_rate": 2e-05,
      "loss": 0.1046,
      "step": 7907
    },
    {
      "epoch": 5.343359242987495,
      "grad_norm": 1.9037877321243286,
      "learning_rate": 2e-05,
      "loss": 0.1679,
      "step": 7908
    },
    {
      "epoch": 5.344035147009125,
      "grad_norm": 3.3165245056152344,
      "learning_rate": 2e-05,
      "loss": 0.1346,
      "step": 7909
    },
    {
      "epoch": 5.344711051030753,
      "grad_norm": 1.5227092504501343,
      "learning_rate": 2e-05,
      "loss": 0.0974,
      "step": 7910
    },
    {
      "epoch": 5.3453869550523825,
      "grad_norm": 1.7548339366912842,
      "learning_rate": 2e-05,
      "loss": 0.1682,
      "step": 7911
    },
    {
      "epoch": 5.346062859074012,
      "grad_norm": 2.4545738697052,
      "learning_rate": 2e-05,
      "loss": 0.1471,
      "step": 7912
    },
    {
      "epoch": 5.34673876309564,
      "grad_norm": 6.880537986755371,
      "learning_rate": 2e-05,
      "loss": 0.2092,
      "step": 7913
    },
    {
      "epoch": 5.3474146671172695,
      "grad_norm": 6.384661674499512,
      "learning_rate": 2e-05,
      "loss": 0.2079,
      "step": 7914
    },
    {
      "epoch": 5.348090571138898,
      "grad_norm": 3.8031864166259766,
      "learning_rate": 2e-05,
      "loss": 0.1613,
      "step": 7915
    },
    {
      "epoch": 5.348766475160527,
      "grad_norm": 2.1729094982147217,
      "learning_rate": 2e-05,
      "loss": 0.1422,
      "step": 7916
    },
    {
      "epoch": 5.349442379182156,
      "grad_norm": 1.7727444171905518,
      "learning_rate": 2e-05,
      "loss": 0.1618,
      "step": 7917
    },
    {
      "epoch": 5.350118283203785,
      "grad_norm": 1.8491606712341309,
      "learning_rate": 2e-05,
      "loss": 0.1449,
      "step": 7918
    },
    {
      "epoch": 5.350794187225414,
      "grad_norm": 2.3346617221832275,
      "learning_rate": 2e-05,
      "loss": 0.0805,
      "step": 7919
    },
    {
      "epoch": 5.351470091247043,
      "grad_norm": 2.487428665161133,
      "learning_rate": 2e-05,
      "loss": 0.1508,
      "step": 7920
    },
    {
      "epoch": 5.352145995268672,
      "grad_norm": 3.2748818397521973,
      "learning_rate": 2e-05,
      "loss": 0.1503,
      "step": 7921
    },
    {
      "epoch": 5.3528218992903005,
      "grad_norm": 1.7182923555374146,
      "learning_rate": 2e-05,
      "loss": 0.0794,
      "step": 7922
    },
    {
      "epoch": 5.35349780331193,
      "grad_norm": 1.9498655796051025,
      "learning_rate": 2e-05,
      "loss": 0.1327,
      "step": 7923
    },
    {
      "epoch": 5.354173707333558,
      "grad_norm": 3.2763071060180664,
      "learning_rate": 2e-05,
      "loss": 0.2031,
      "step": 7924
    },
    {
      "epoch": 5.3548496113551876,
      "grad_norm": 3.296825408935547,
      "learning_rate": 2e-05,
      "loss": 0.1409,
      "step": 7925
    },
    {
      "epoch": 5.355525515376817,
      "grad_norm": 2.5777413845062256,
      "learning_rate": 2e-05,
      "loss": 0.1813,
      "step": 7926
    },
    {
      "epoch": 5.356201419398445,
      "grad_norm": 7.068456649780273,
      "learning_rate": 2e-05,
      "loss": 0.1867,
      "step": 7927
    },
    {
      "epoch": 5.356877323420075,
      "grad_norm": 2.9255924224853516,
      "learning_rate": 2e-05,
      "loss": 0.1713,
      "step": 7928
    },
    {
      "epoch": 5.357553227441703,
      "grad_norm": 2.626345157623291,
      "learning_rate": 2e-05,
      "loss": 0.1743,
      "step": 7929
    },
    {
      "epoch": 5.358229131463332,
      "grad_norm": 2.2415947914123535,
      "learning_rate": 2e-05,
      "loss": 0.1711,
      "step": 7930
    },
    {
      "epoch": 5.358905035484961,
      "grad_norm": 3.7084856033325195,
      "learning_rate": 2e-05,
      "loss": 0.2095,
      "step": 7931
    },
    {
      "epoch": 5.35958093950659,
      "grad_norm": 5.402217864990234,
      "learning_rate": 2e-05,
      "loss": 0.1214,
      "step": 7932
    },
    {
      "epoch": 5.360256843528219,
      "grad_norm": 3.6127915382385254,
      "learning_rate": 2e-05,
      "loss": 0.125,
      "step": 7933
    },
    {
      "epoch": 5.360932747549848,
      "grad_norm": 2.725952625274658,
      "learning_rate": 2e-05,
      "loss": 0.168,
      "step": 7934
    },
    {
      "epoch": 5.361608651571477,
      "grad_norm": 2.5320851802825928,
      "learning_rate": 2e-05,
      "loss": 0.1502,
      "step": 7935
    },
    {
      "epoch": 5.362284555593106,
      "grad_norm": 2.4890198707580566,
      "learning_rate": 2e-05,
      "loss": 0.1615,
      "step": 7936
    },
    {
      "epoch": 5.362960459614735,
      "grad_norm": 2.2450287342071533,
      "learning_rate": 2e-05,
      "loss": 0.1215,
      "step": 7937
    },
    {
      "epoch": 5.363636363636363,
      "grad_norm": 8.87985610961914,
      "learning_rate": 2e-05,
      "loss": 0.1296,
      "step": 7938
    },
    {
      "epoch": 5.364312267657993,
      "grad_norm": 1.8532203435897827,
      "learning_rate": 2e-05,
      "loss": 0.1429,
      "step": 7939
    },
    {
      "epoch": 5.364988171679622,
      "grad_norm": 3.942521810531616,
      "learning_rate": 2e-05,
      "loss": 0.1135,
      "step": 7940
    },
    {
      "epoch": 5.36566407570125,
      "grad_norm": 2.2157390117645264,
      "learning_rate": 2e-05,
      "loss": 0.1822,
      "step": 7941
    },
    {
      "epoch": 5.36633997972288,
      "grad_norm": 3.398202419281006,
      "learning_rate": 2e-05,
      "loss": 0.1327,
      "step": 7942
    },
    {
      "epoch": 5.367015883744508,
      "grad_norm": 4.862818717956543,
      "learning_rate": 2e-05,
      "loss": 0.1651,
      "step": 7943
    },
    {
      "epoch": 5.367691787766137,
      "grad_norm": 2.831875801086426,
      "learning_rate": 2e-05,
      "loss": 0.1576,
      "step": 7944
    },
    {
      "epoch": 5.368367691787766,
      "grad_norm": 3.4283573627471924,
      "learning_rate": 2e-05,
      "loss": 0.1679,
      "step": 7945
    },
    {
      "epoch": 5.369043595809395,
      "grad_norm": 4.099649429321289,
      "learning_rate": 2e-05,
      "loss": 0.1365,
      "step": 7946
    },
    {
      "epoch": 5.369719499831024,
      "grad_norm": 5.207469463348389,
      "learning_rate": 2e-05,
      "loss": 0.1547,
      "step": 7947
    },
    {
      "epoch": 5.370395403852653,
      "grad_norm": 1.7626183032989502,
      "learning_rate": 2e-05,
      "loss": 0.1215,
      "step": 7948
    },
    {
      "epoch": 5.371071307874282,
      "grad_norm": 2.457165241241455,
      "learning_rate": 2e-05,
      "loss": 0.1383,
      "step": 7949
    },
    {
      "epoch": 5.371747211895911,
      "grad_norm": 3.6659369468688965,
      "learning_rate": 2e-05,
      "loss": 0.1613,
      "step": 7950
    },
    {
      "epoch": 5.371747211895911,
      "eval_accuracy": 0.783910472972973,
      "eval_loss": 0.24081507325172424,
      "eval_runtime": 51.538,
      "eval_samples_per_second": 459.467,
      "eval_steps_per_second": 14.358,
      "step": 7950
    },
    {
      "epoch": 5.37242311591754,
      "grad_norm": 6.179384708404541,
      "learning_rate": 2e-05,
      "loss": 0.1356,
      "step": 7951
    },
    {
      "epoch": 5.373099019939168,
      "grad_norm": 2.5349457263946533,
      "learning_rate": 2e-05,
      "loss": 0.1674,
      "step": 7952
    },
    {
      "epoch": 5.373774923960798,
      "grad_norm": 2.7111783027648926,
      "learning_rate": 2e-05,
      "loss": 0.1332,
      "step": 7953
    },
    {
      "epoch": 5.374450827982426,
      "grad_norm": 2.004240036010742,
      "learning_rate": 2e-05,
      "loss": 0.1782,
      "step": 7954
    },
    {
      "epoch": 5.375126732004055,
      "grad_norm": 2.1309237480163574,
      "learning_rate": 2e-05,
      "loss": 0.1706,
      "step": 7955
    },
    {
      "epoch": 5.375802636025685,
      "grad_norm": 4.5736870765686035,
      "learning_rate": 2e-05,
      "loss": 0.1586,
      "step": 7956
    },
    {
      "epoch": 5.376478540047313,
      "grad_norm": 3.2083311080932617,
      "learning_rate": 2e-05,
      "loss": 0.1714,
      "step": 7957
    },
    {
      "epoch": 5.3771544440689425,
      "grad_norm": 2.5687193870544434,
      "learning_rate": 2e-05,
      "loss": 0.1347,
      "step": 7958
    },
    {
      "epoch": 5.377830348090571,
      "grad_norm": 2.330441474914551,
      "learning_rate": 2e-05,
      "loss": 0.1814,
      "step": 7959
    },
    {
      "epoch": 5.3785062521122,
      "grad_norm": 4.437220096588135,
      "learning_rate": 2e-05,
      "loss": 0.1677,
      "step": 7960
    },
    {
      "epoch": 5.379182156133829,
      "grad_norm": 3.4925131797790527,
      "learning_rate": 2e-05,
      "loss": 0.1364,
      "step": 7961
    },
    {
      "epoch": 5.379858060155458,
      "grad_norm": 2.174591302871704,
      "learning_rate": 2e-05,
      "loss": 0.165,
      "step": 7962
    },
    {
      "epoch": 5.380533964177086,
      "grad_norm": 2.7302117347717285,
      "learning_rate": 2e-05,
      "loss": 0.2341,
      "step": 7963
    },
    {
      "epoch": 5.381209868198716,
      "grad_norm": 2.231147289276123,
      "learning_rate": 2e-05,
      "loss": 0.1794,
      "step": 7964
    },
    {
      "epoch": 5.381885772220345,
      "grad_norm": 5.23698091506958,
      "learning_rate": 2e-05,
      "loss": 0.1575,
      "step": 7965
    },
    {
      "epoch": 5.3825616762419735,
      "grad_norm": 5.360477924346924,
      "learning_rate": 2e-05,
      "loss": 0.1446,
      "step": 7966
    },
    {
      "epoch": 5.383237580263603,
      "grad_norm": 1.9946458339691162,
      "learning_rate": 2e-05,
      "loss": 0.1616,
      "step": 7967
    },
    {
      "epoch": 5.383913484285231,
      "grad_norm": 2.8778746128082275,
      "learning_rate": 2e-05,
      "loss": 0.1476,
      "step": 7968
    },
    {
      "epoch": 5.3845893883068605,
      "grad_norm": 7.915513038635254,
      "learning_rate": 2e-05,
      "loss": 0.1577,
      "step": 7969
    },
    {
      "epoch": 5.385265292328489,
      "grad_norm": 5.779706001281738,
      "learning_rate": 2e-05,
      "loss": 0.1522,
      "step": 7970
    },
    {
      "epoch": 5.385941196350118,
      "grad_norm": 2.7936716079711914,
      "learning_rate": 2e-05,
      "loss": 0.113,
      "step": 7971
    },
    {
      "epoch": 5.386617100371748,
      "grad_norm": 1.3860760927200317,
      "learning_rate": 2e-05,
      "loss": 0.1457,
      "step": 7972
    },
    {
      "epoch": 5.387293004393376,
      "grad_norm": 2.9187939167022705,
      "learning_rate": 2e-05,
      "loss": 0.1755,
      "step": 7973
    },
    {
      "epoch": 5.387968908415005,
      "grad_norm": 2.273254871368408,
      "learning_rate": 2e-05,
      "loss": 0.2016,
      "step": 7974
    },
    {
      "epoch": 5.388644812436634,
      "grad_norm": 2.1389544010162354,
      "learning_rate": 2e-05,
      "loss": 0.2106,
      "step": 7975
    },
    {
      "epoch": 5.389320716458263,
      "grad_norm": 4.08931827545166,
      "learning_rate": 2e-05,
      "loss": 0.1083,
      "step": 7976
    },
    {
      "epoch": 5.3899966204798915,
      "grad_norm": 4.216609477996826,
      "learning_rate": 2e-05,
      "loss": 0.1744,
      "step": 7977
    },
    {
      "epoch": 5.390672524501521,
      "grad_norm": 3.7249157428741455,
      "learning_rate": 2e-05,
      "loss": 0.1613,
      "step": 7978
    },
    {
      "epoch": 5.39134842852315,
      "grad_norm": 2.9566690921783447,
      "learning_rate": 2e-05,
      "loss": 0.1711,
      "step": 7979
    },
    {
      "epoch": 5.3920243325447785,
      "grad_norm": 2.347580909729004,
      "learning_rate": 2e-05,
      "loss": 0.19,
      "step": 7980
    },
    {
      "epoch": 5.392700236566408,
      "grad_norm": 4.401803493499756,
      "learning_rate": 2e-05,
      "loss": 0.1678,
      "step": 7981
    },
    {
      "epoch": 5.393376140588036,
      "grad_norm": 3.6637418270111084,
      "learning_rate": 2e-05,
      "loss": 0.1352,
      "step": 7982
    },
    {
      "epoch": 5.394052044609666,
      "grad_norm": 1.9445013999938965,
      "learning_rate": 2e-05,
      "loss": 0.1105,
      "step": 7983
    },
    {
      "epoch": 5.394727948631294,
      "grad_norm": 3.784882068634033,
      "learning_rate": 2e-05,
      "loss": 0.1553,
      "step": 7984
    },
    {
      "epoch": 5.395403852652923,
      "grad_norm": 2.4891490936279297,
      "learning_rate": 2e-05,
      "loss": 0.1351,
      "step": 7985
    },
    {
      "epoch": 5.396079756674553,
      "grad_norm": 2.1436052322387695,
      "learning_rate": 2e-05,
      "loss": 0.1414,
      "step": 7986
    },
    {
      "epoch": 5.396755660696181,
      "grad_norm": 3.144345283508301,
      "learning_rate": 2e-05,
      "loss": 0.1914,
      "step": 7987
    },
    {
      "epoch": 5.39743156471781,
      "grad_norm": 3.9820144176483154,
      "learning_rate": 2e-05,
      "loss": 0.167,
      "step": 7988
    },
    {
      "epoch": 5.398107468739439,
      "grad_norm": 1.758431077003479,
      "learning_rate": 2e-05,
      "loss": 0.17,
      "step": 7989
    },
    {
      "epoch": 5.398783372761068,
      "grad_norm": 4.134430408477783,
      "learning_rate": 2e-05,
      "loss": 0.114,
      "step": 7990
    },
    {
      "epoch": 5.3994592767826965,
      "grad_norm": 1.3700863122940063,
      "learning_rate": 2e-05,
      "loss": 0.1487,
      "step": 7991
    },
    {
      "epoch": 5.400135180804326,
      "grad_norm": 1.8225936889648438,
      "learning_rate": 2e-05,
      "loss": 0.11,
      "step": 7992
    },
    {
      "epoch": 5.400811084825955,
      "grad_norm": 5.02482795715332,
      "learning_rate": 2e-05,
      "loss": 0.1476,
      "step": 7993
    },
    {
      "epoch": 5.401486988847584,
      "grad_norm": 7.1898016929626465,
      "learning_rate": 2e-05,
      "loss": 0.1696,
      "step": 7994
    },
    {
      "epoch": 5.402162892869213,
      "grad_norm": 5.825398921966553,
      "learning_rate": 2e-05,
      "loss": 0.326,
      "step": 7995
    },
    {
      "epoch": 5.402838796890841,
      "grad_norm": 3.065668821334839,
      "learning_rate": 2e-05,
      "loss": 0.1688,
      "step": 7996
    },
    {
      "epoch": 5.403514700912471,
      "grad_norm": 3.2382516860961914,
      "learning_rate": 2e-05,
      "loss": 0.1315,
      "step": 7997
    },
    {
      "epoch": 5.404190604934099,
      "grad_norm": 4.7119975090026855,
      "learning_rate": 2e-05,
      "loss": 0.1639,
      "step": 7998
    },
    {
      "epoch": 5.404866508955728,
      "grad_norm": 6.367834091186523,
      "learning_rate": 2e-05,
      "loss": 0.1386,
      "step": 7999
    },
    {
      "epoch": 5.405542412977357,
      "grad_norm": 5.836191177368164,
      "learning_rate": 2e-05,
      "loss": 0.156,
      "step": 8000
    },
    {
      "epoch": 5.405542412977357,
      "eval_accuracy": 0.7590371621621622,
      "eval_loss": 0.23090939223766327,
      "eval_runtime": 46.9846,
      "eval_samples_per_second": 503.995,
      "eval_steps_per_second": 15.75,
      "step": 8000
    },
    {
      "epoch": 5.406218316998986,
      "grad_norm": 3.404524803161621,
      "learning_rate": 2e-05,
      "loss": 0.169,
      "step": 8001
    },
    {
      "epoch": 5.4068942210206155,
      "grad_norm": 3.062309741973877,
      "learning_rate": 2e-05,
      "loss": 0.2084,
      "step": 8002
    },
    {
      "epoch": 5.407570125042244,
      "grad_norm": 5.214565277099609,
      "learning_rate": 2e-05,
      "loss": 0.2009,
      "step": 8003
    },
    {
      "epoch": 5.408246029063873,
      "grad_norm": 3.271944046020508,
      "learning_rate": 2e-05,
      "loss": 0.1466,
      "step": 8004
    },
    {
      "epoch": 5.408921933085502,
      "grad_norm": 1.4363462924957275,
      "learning_rate": 2e-05,
      "loss": 0.1337,
      "step": 8005
    },
    {
      "epoch": 5.409597837107131,
      "grad_norm": 2.7465384006500244,
      "learning_rate": 2e-05,
      "loss": 0.1468,
      "step": 8006
    },
    {
      "epoch": 5.410273741128759,
      "grad_norm": 3.8201546669006348,
      "learning_rate": 2e-05,
      "loss": 0.1673,
      "step": 8007
    },
    {
      "epoch": 5.410949645150389,
      "grad_norm": 4.509533405303955,
      "learning_rate": 2e-05,
      "loss": 0.1998,
      "step": 8008
    },
    {
      "epoch": 5.411625549172017,
      "grad_norm": 4.022464752197266,
      "learning_rate": 2e-05,
      "loss": 0.1576,
      "step": 8009
    },
    {
      "epoch": 5.412301453193646,
      "grad_norm": 1.9652345180511475,
      "learning_rate": 2e-05,
      "loss": 0.1812,
      "step": 8010
    },
    {
      "epoch": 5.412977357215276,
      "grad_norm": 4.745934009552002,
      "learning_rate": 2e-05,
      "loss": 0.1278,
      "step": 8011
    },
    {
      "epoch": 5.413653261236904,
      "grad_norm": 2.337362051010132,
      "learning_rate": 2e-05,
      "loss": 0.1694,
      "step": 8012
    },
    {
      "epoch": 5.4143291652585335,
      "grad_norm": 5.684236526489258,
      "learning_rate": 2e-05,
      "loss": 0.1643,
      "step": 8013
    },
    {
      "epoch": 5.415005069280162,
      "grad_norm": 2.9609451293945312,
      "learning_rate": 2e-05,
      "loss": 0.1257,
      "step": 8014
    },
    {
      "epoch": 5.415680973301791,
      "grad_norm": 2.584446668624878,
      "learning_rate": 2e-05,
      "loss": 0.1617,
      "step": 8015
    },
    {
      "epoch": 5.41635687732342,
      "grad_norm": 3.9783143997192383,
      "learning_rate": 2e-05,
      "loss": 0.0964,
      "step": 8016
    },
    {
      "epoch": 5.417032781345049,
      "grad_norm": 2.3925812244415283,
      "learning_rate": 2e-05,
      "loss": 0.2364,
      "step": 8017
    },
    {
      "epoch": 5.417708685366678,
      "grad_norm": 3.717879295349121,
      "learning_rate": 2e-05,
      "loss": 0.1347,
      "step": 8018
    },
    {
      "epoch": 5.418384589388307,
      "grad_norm": 1.9484272003173828,
      "learning_rate": 2e-05,
      "loss": 0.1433,
      "step": 8019
    },
    {
      "epoch": 5.419060493409936,
      "grad_norm": 3.597914934158325,
      "learning_rate": 2e-05,
      "loss": 0.2066,
      "step": 8020
    },
    {
      "epoch": 5.419736397431564,
      "grad_norm": 7.674152374267578,
      "learning_rate": 2e-05,
      "loss": 0.1063,
      "step": 8021
    },
    {
      "epoch": 5.420412301453194,
      "grad_norm": 2.419720411300659,
      "learning_rate": 2e-05,
      "loss": 0.1557,
      "step": 8022
    },
    {
      "epoch": 5.421088205474822,
      "grad_norm": 4.6300554275512695,
      "learning_rate": 2e-05,
      "loss": 0.1523,
      "step": 8023
    },
    {
      "epoch": 5.4217641094964515,
      "grad_norm": 4.82394552230835,
      "learning_rate": 2e-05,
      "loss": 0.1369,
      "step": 8024
    },
    {
      "epoch": 5.422440013518081,
      "grad_norm": 1.8317391872406006,
      "learning_rate": 2e-05,
      "loss": 0.1014,
      "step": 8025
    },
    {
      "epoch": 5.423115917539709,
      "grad_norm": 2.216949939727783,
      "learning_rate": 2e-05,
      "loss": 0.1044,
      "step": 8026
    },
    {
      "epoch": 5.4237918215613385,
      "grad_norm": 2.2950382232666016,
      "learning_rate": 2e-05,
      "loss": 0.1653,
      "step": 8027
    },
    {
      "epoch": 5.424467725582967,
      "grad_norm": 5.666065216064453,
      "learning_rate": 2e-05,
      "loss": 0.1535,
      "step": 8028
    },
    {
      "epoch": 5.425143629604596,
      "grad_norm": 4.567964553833008,
      "learning_rate": 2e-05,
      "loss": 0.1395,
      "step": 8029
    },
    {
      "epoch": 5.425819533626225,
      "grad_norm": 8.939783096313477,
      "learning_rate": 2e-05,
      "loss": 0.298,
      "step": 8030
    },
    {
      "epoch": 5.426495437647854,
      "grad_norm": 9.608133316040039,
      "learning_rate": 2e-05,
      "loss": 0.1908,
      "step": 8031
    },
    {
      "epoch": 5.427171341669483,
      "grad_norm": 2.508772611618042,
      "learning_rate": 2e-05,
      "loss": 0.1765,
      "step": 8032
    },
    {
      "epoch": 5.427847245691112,
      "grad_norm": 6.4339213371276855,
      "learning_rate": 2e-05,
      "loss": 0.0802,
      "step": 8033
    },
    {
      "epoch": 5.428523149712741,
      "grad_norm": 2.171945095062256,
      "learning_rate": 2e-05,
      "loss": 0.1831,
      "step": 8034
    },
    {
      "epoch": 5.4291990537343695,
      "grad_norm": 5.7636213302612305,
      "learning_rate": 2e-05,
      "loss": 0.1318,
      "step": 8035
    },
    {
      "epoch": 5.429874957755999,
      "grad_norm": 3.754774808883667,
      "learning_rate": 2e-05,
      "loss": 0.181,
      "step": 8036
    },
    {
      "epoch": 5.430550861777627,
      "grad_norm": 9.709126472473145,
      "learning_rate": 2e-05,
      "loss": 0.2092,
      "step": 8037
    },
    {
      "epoch": 5.431226765799257,
      "grad_norm": 6.297460556030273,
      "learning_rate": 2e-05,
      "loss": 0.1754,
      "step": 8038
    },
    {
      "epoch": 5.431902669820886,
      "grad_norm": 5.596184730529785,
      "learning_rate": 2e-05,
      "loss": 0.1176,
      "step": 8039
    },
    {
      "epoch": 5.432578573842514,
      "grad_norm": 5.357330322265625,
      "learning_rate": 2e-05,
      "loss": 0.1541,
      "step": 8040
    },
    {
      "epoch": 5.433254477864144,
      "grad_norm": 6.889560699462891,
      "learning_rate": 2e-05,
      "loss": 0.196,
      "step": 8041
    },
    {
      "epoch": 5.433930381885772,
      "grad_norm": 4.768102169036865,
      "learning_rate": 2e-05,
      "loss": 0.1618,
      "step": 8042
    },
    {
      "epoch": 5.434606285907401,
      "grad_norm": 2.7197234630584717,
      "learning_rate": 2e-05,
      "loss": 0.1927,
      "step": 8043
    },
    {
      "epoch": 5.43528218992903,
      "grad_norm": 4.664916515350342,
      "learning_rate": 2e-05,
      "loss": 0.1699,
      "step": 8044
    },
    {
      "epoch": 5.435958093950659,
      "grad_norm": 2.961902379989624,
      "learning_rate": 2e-05,
      "loss": 0.152,
      "step": 8045
    },
    {
      "epoch": 5.4366339979722875,
      "grad_norm": 3.636505365371704,
      "learning_rate": 2e-05,
      "loss": 0.1628,
      "step": 8046
    },
    {
      "epoch": 5.437309901993917,
      "grad_norm": 4.048971176147461,
      "learning_rate": 2e-05,
      "loss": 0.1298,
      "step": 8047
    },
    {
      "epoch": 5.437985806015546,
      "grad_norm": 2.1962311267852783,
      "learning_rate": 2e-05,
      "loss": 0.1861,
      "step": 8048
    },
    {
      "epoch": 5.438661710037175,
      "grad_norm": 2.238931894302368,
      "learning_rate": 2e-05,
      "loss": 0.1237,
      "step": 8049
    },
    {
      "epoch": 5.439337614058804,
      "grad_norm": 1.9443804025650024,
      "learning_rate": 2e-05,
      "loss": 0.1275,
      "step": 8050
    },
    {
      "epoch": 5.439337614058804,
      "eval_accuracy": 0.754856418918919,
      "eval_loss": 0.22382204234600067,
      "eval_runtime": 48.2558,
      "eval_samples_per_second": 490.718,
      "eval_steps_per_second": 15.335,
      "step": 8050
    },
    {
      "epoch": 5.440013518080432,
      "grad_norm": 2.6425788402557373,
      "learning_rate": 2e-05,
      "loss": 0.1079,
      "step": 8051
    },
    {
      "epoch": 5.440689422102062,
      "grad_norm": 6.247486591339111,
      "learning_rate": 2e-05,
      "loss": 0.1698,
      "step": 8052
    },
    {
      "epoch": 5.44136532612369,
      "grad_norm": 2.954197406768799,
      "learning_rate": 2e-05,
      "loss": 0.1611,
      "step": 8053
    },
    {
      "epoch": 5.442041230145319,
      "grad_norm": 1.4814592599868774,
      "learning_rate": 2e-05,
      "loss": 0.1201,
      "step": 8054
    },
    {
      "epoch": 5.442717134166949,
      "grad_norm": 2.2507503032684326,
      "learning_rate": 2e-05,
      "loss": 0.1705,
      "step": 8055
    },
    {
      "epoch": 5.443393038188577,
      "grad_norm": 3.2818100452423096,
      "learning_rate": 2e-05,
      "loss": 0.1342,
      "step": 8056
    },
    {
      "epoch": 5.444068942210206,
      "grad_norm": 2.25887393951416,
      "learning_rate": 2e-05,
      "loss": 0.1142,
      "step": 8057
    },
    {
      "epoch": 5.444744846231835,
      "grad_norm": 2.8252665996551514,
      "learning_rate": 2e-05,
      "loss": 0.2132,
      "step": 8058
    },
    {
      "epoch": 5.445420750253464,
      "grad_norm": 6.699522972106934,
      "learning_rate": 2e-05,
      "loss": 0.1343,
      "step": 8059
    },
    {
      "epoch": 5.446096654275093,
      "grad_norm": 2.644026517868042,
      "learning_rate": 2e-05,
      "loss": 0.0978,
      "step": 8060
    },
    {
      "epoch": 5.446772558296722,
      "grad_norm": 4.866253852844238,
      "learning_rate": 2e-05,
      "loss": 0.1775,
      "step": 8061
    },
    {
      "epoch": 5.44744846231835,
      "grad_norm": 4.144374370574951,
      "learning_rate": 2e-05,
      "loss": 0.1441,
      "step": 8062
    },
    {
      "epoch": 5.44812436633998,
      "grad_norm": 3.4824368953704834,
      "learning_rate": 2e-05,
      "loss": 0.1096,
      "step": 8063
    },
    {
      "epoch": 5.448800270361609,
      "grad_norm": 2.6276326179504395,
      "learning_rate": 2e-05,
      "loss": 0.1411,
      "step": 8064
    },
    {
      "epoch": 5.449476174383237,
      "grad_norm": 2.491126298904419,
      "learning_rate": 2e-05,
      "loss": 0.2254,
      "step": 8065
    },
    {
      "epoch": 5.450152078404867,
      "grad_norm": 5.0437211990356445,
      "learning_rate": 2e-05,
      "loss": 0.1588,
      "step": 8066
    },
    {
      "epoch": 5.450827982426495,
      "grad_norm": 3.8109912872314453,
      "learning_rate": 2e-05,
      "loss": 0.1172,
      "step": 8067
    },
    {
      "epoch": 5.4515038864481244,
      "grad_norm": 4.161117076873779,
      "learning_rate": 2e-05,
      "loss": 0.219,
      "step": 8068
    },
    {
      "epoch": 5.452179790469753,
      "grad_norm": 5.982975959777832,
      "learning_rate": 2e-05,
      "loss": 0.1536,
      "step": 8069
    },
    {
      "epoch": 5.452855694491382,
      "grad_norm": 1.975697636604309,
      "learning_rate": 2e-05,
      "loss": 0.1151,
      "step": 8070
    },
    {
      "epoch": 5.4535315985130115,
      "grad_norm": 5.402202129364014,
      "learning_rate": 2e-05,
      "loss": 0.1993,
      "step": 8071
    },
    {
      "epoch": 5.45420750253464,
      "grad_norm": 4.223846435546875,
      "learning_rate": 2e-05,
      "loss": 0.148,
      "step": 8072
    },
    {
      "epoch": 5.454883406556269,
      "grad_norm": 1.8662627935409546,
      "learning_rate": 2e-05,
      "loss": 0.1104,
      "step": 8073
    },
    {
      "epoch": 5.455559310577898,
      "grad_norm": 3.8122551441192627,
      "learning_rate": 2e-05,
      "loss": 0.0814,
      "step": 8074
    },
    {
      "epoch": 5.456235214599527,
      "grad_norm": 2.6086301803588867,
      "learning_rate": 2e-05,
      "loss": 0.142,
      "step": 8075
    },
    {
      "epoch": 5.456911118621155,
      "grad_norm": 2.8411920070648193,
      "learning_rate": 2e-05,
      "loss": 0.1177,
      "step": 8076
    },
    {
      "epoch": 5.457587022642785,
      "grad_norm": 2.8973851203918457,
      "learning_rate": 2e-05,
      "loss": 0.1975,
      "step": 8077
    },
    {
      "epoch": 5.458262926664414,
      "grad_norm": 4.383665561676025,
      "learning_rate": 2e-05,
      "loss": 0.1105,
      "step": 8078
    },
    {
      "epoch": 5.4589388306860425,
      "grad_norm": 6.952496528625488,
      "learning_rate": 2e-05,
      "loss": 0.1598,
      "step": 8079
    },
    {
      "epoch": 5.459614734707672,
      "grad_norm": 3.20971417427063,
      "learning_rate": 2e-05,
      "loss": 0.1095,
      "step": 8080
    },
    {
      "epoch": 5.4602906387293,
      "grad_norm": 5.113437175750732,
      "learning_rate": 2e-05,
      "loss": 0.1623,
      "step": 8081
    },
    {
      "epoch": 5.4609665427509295,
      "grad_norm": 2.0057482719421387,
      "learning_rate": 2e-05,
      "loss": 0.2012,
      "step": 8082
    },
    {
      "epoch": 5.461642446772558,
      "grad_norm": 5.132128715515137,
      "learning_rate": 2e-05,
      "loss": 0.1392,
      "step": 8083
    },
    {
      "epoch": 5.462318350794187,
      "grad_norm": 5.90605354309082,
      "learning_rate": 2e-05,
      "loss": 0.1604,
      "step": 8084
    },
    {
      "epoch": 5.462994254815817,
      "grad_norm": 5.0869140625,
      "learning_rate": 2e-05,
      "loss": 0.1796,
      "step": 8085
    },
    {
      "epoch": 5.463670158837445,
      "grad_norm": 2.3975448608398438,
      "learning_rate": 2e-05,
      "loss": 0.1319,
      "step": 8086
    },
    {
      "epoch": 5.464346062859074,
      "grad_norm": 1.5773835182189941,
      "learning_rate": 2e-05,
      "loss": 0.1273,
      "step": 8087
    },
    {
      "epoch": 5.465021966880703,
      "grad_norm": 3.670591115951538,
      "learning_rate": 2e-05,
      "loss": 0.202,
      "step": 8088
    },
    {
      "epoch": 5.465697870902332,
      "grad_norm": 3.0327939987182617,
      "learning_rate": 2e-05,
      "loss": 0.2207,
      "step": 8089
    },
    {
      "epoch": 5.4663737749239605,
      "grad_norm": 2.5611302852630615,
      "learning_rate": 2e-05,
      "loss": 0.1471,
      "step": 8090
    },
    {
      "epoch": 5.46704967894559,
      "grad_norm": 5.4189133644104,
      "learning_rate": 2e-05,
      "loss": 0.1144,
      "step": 8091
    },
    {
      "epoch": 5.467725582967219,
      "grad_norm": 3.3233118057250977,
      "learning_rate": 2e-05,
      "loss": 0.1629,
      "step": 8092
    },
    {
      "epoch": 5.4684014869888475,
      "grad_norm": 3.724411964416504,
      "learning_rate": 2e-05,
      "loss": 0.1433,
      "step": 8093
    },
    {
      "epoch": 5.469077391010477,
      "grad_norm": 1.5861526727676392,
      "learning_rate": 2e-05,
      "loss": 0.143,
      "step": 8094
    },
    {
      "epoch": 5.469753295032105,
      "grad_norm": 3.4550628662109375,
      "learning_rate": 2e-05,
      "loss": 0.1872,
      "step": 8095
    },
    {
      "epoch": 5.470429199053735,
      "grad_norm": 1.9694149494171143,
      "learning_rate": 2e-05,
      "loss": 0.1606,
      "step": 8096
    },
    {
      "epoch": 5.471105103075363,
      "grad_norm": 4.59539270401001,
      "learning_rate": 2e-05,
      "loss": 0.1404,
      "step": 8097
    },
    {
      "epoch": 5.471781007096992,
      "grad_norm": 2.1944072246551514,
      "learning_rate": 2e-05,
      "loss": 0.1627,
      "step": 8098
    },
    {
      "epoch": 5.472456911118621,
      "grad_norm": 2.329324245452881,
      "learning_rate": 2e-05,
      "loss": 0.1969,
      "step": 8099
    },
    {
      "epoch": 5.47313281514025,
      "grad_norm": 1.9469135999679565,
      "learning_rate": 2e-05,
      "loss": 0.1616,
      "step": 8100
    },
    {
      "epoch": 5.47313281514025,
      "eval_accuracy": 0.7670608108108108,
      "eval_loss": 0.2248397022485733,
      "eval_runtime": 46.8654,
      "eval_samples_per_second": 505.277,
      "eval_steps_per_second": 15.79,
      "step": 8100
    },
    {
      "epoch": 5.473808719161879,
      "grad_norm": 1.6778180599212646,
      "learning_rate": 2e-05,
      "loss": 0.145,
      "step": 8101
    },
    {
      "epoch": 5.474484623183508,
      "grad_norm": 3.5693540573120117,
      "learning_rate": 2e-05,
      "loss": 0.1273,
      "step": 8102
    },
    {
      "epoch": 5.475160527205137,
      "grad_norm": 1.8195767402648926,
      "learning_rate": 2e-05,
      "loss": 0.1563,
      "step": 8103
    },
    {
      "epoch": 5.4758364312267656,
      "grad_norm": 7.075252056121826,
      "learning_rate": 2e-05,
      "loss": 0.1273,
      "step": 8104
    },
    {
      "epoch": 5.476512335248395,
      "grad_norm": 2.416256904602051,
      "learning_rate": 2e-05,
      "loss": 0.1408,
      "step": 8105
    },
    {
      "epoch": 5.477188239270023,
      "grad_norm": 3.0363404750823975,
      "learning_rate": 2e-05,
      "loss": 0.1406,
      "step": 8106
    },
    {
      "epoch": 5.477864143291653,
      "grad_norm": 4.125951766967773,
      "learning_rate": 2e-05,
      "loss": 0.1701,
      "step": 8107
    },
    {
      "epoch": 5.478540047313282,
      "grad_norm": 5.397008419036865,
      "learning_rate": 2e-05,
      "loss": 0.1436,
      "step": 8108
    },
    {
      "epoch": 5.47921595133491,
      "grad_norm": 2.9974348545074463,
      "learning_rate": 2e-05,
      "loss": 0.1496,
      "step": 8109
    },
    {
      "epoch": 5.47989185535654,
      "grad_norm": 6.323788642883301,
      "learning_rate": 2e-05,
      "loss": 0.1853,
      "step": 8110
    },
    {
      "epoch": 5.480567759378168,
      "grad_norm": 8.098483085632324,
      "learning_rate": 2e-05,
      "loss": 0.2775,
      "step": 8111
    },
    {
      "epoch": 5.481243663399797,
      "grad_norm": 4.499943256378174,
      "learning_rate": 2e-05,
      "loss": 0.2056,
      "step": 8112
    },
    {
      "epoch": 5.481919567421426,
      "grad_norm": 2.0933756828308105,
      "learning_rate": 2e-05,
      "loss": 0.236,
      "step": 8113
    },
    {
      "epoch": 5.482595471443055,
      "grad_norm": 7.813767910003662,
      "learning_rate": 2e-05,
      "loss": 0.1602,
      "step": 8114
    },
    {
      "epoch": 5.483271375464684,
      "grad_norm": 1.7120471000671387,
      "learning_rate": 2e-05,
      "loss": 0.1357,
      "step": 8115
    },
    {
      "epoch": 5.483947279486313,
      "grad_norm": 2.943671464920044,
      "learning_rate": 2e-05,
      "loss": 0.1986,
      "step": 8116
    },
    {
      "epoch": 5.484623183507942,
      "grad_norm": 6.154226779937744,
      "learning_rate": 2e-05,
      "loss": 0.1797,
      "step": 8117
    },
    {
      "epoch": 5.485299087529571,
      "grad_norm": 1.7254897356033325,
      "learning_rate": 2e-05,
      "loss": 0.1672,
      "step": 8118
    },
    {
      "epoch": 5.4859749915512,
      "grad_norm": 4.117249965667725,
      "learning_rate": 2e-05,
      "loss": 0.1283,
      "step": 8119
    },
    {
      "epoch": 5.486650895572828,
      "grad_norm": 6.098010540008545,
      "learning_rate": 2e-05,
      "loss": 0.1971,
      "step": 8120
    },
    {
      "epoch": 5.487326799594458,
      "grad_norm": 7.379197597503662,
      "learning_rate": 2e-05,
      "loss": 0.179,
      "step": 8121
    },
    {
      "epoch": 5.488002703616086,
      "grad_norm": 2.24263596534729,
      "learning_rate": 2e-05,
      "loss": 0.1442,
      "step": 8122
    },
    {
      "epoch": 5.488678607637715,
      "grad_norm": 6.069986820220947,
      "learning_rate": 2e-05,
      "loss": 0.1895,
      "step": 8123
    },
    {
      "epoch": 5.489354511659345,
      "grad_norm": 5.038661003112793,
      "learning_rate": 2e-05,
      "loss": 0.1769,
      "step": 8124
    },
    {
      "epoch": 5.490030415680973,
      "grad_norm": 4.5749831199646,
      "learning_rate": 2e-05,
      "loss": 0.12,
      "step": 8125
    },
    {
      "epoch": 5.4907063197026025,
      "grad_norm": 1.4540882110595703,
      "learning_rate": 2e-05,
      "loss": 0.1165,
      "step": 8126
    },
    {
      "epoch": 5.491382223724231,
      "grad_norm": 1.7803140878677368,
      "learning_rate": 2e-05,
      "loss": 0.1324,
      "step": 8127
    },
    {
      "epoch": 5.49205812774586,
      "grad_norm": 1.5550788640975952,
      "learning_rate": 2e-05,
      "loss": 0.1492,
      "step": 8128
    },
    {
      "epoch": 5.492734031767489,
      "grad_norm": 5.089910984039307,
      "learning_rate": 2e-05,
      "loss": 0.1287,
      "step": 8129
    },
    {
      "epoch": 5.493409935789118,
      "grad_norm": 1.9813536405563354,
      "learning_rate": 2e-05,
      "loss": 0.119,
      "step": 8130
    },
    {
      "epoch": 5.494085839810747,
      "grad_norm": 1.7066471576690674,
      "learning_rate": 2e-05,
      "loss": 0.1167,
      "step": 8131
    },
    {
      "epoch": 5.494761743832376,
      "grad_norm": 3.798050880432129,
      "learning_rate": 2e-05,
      "loss": 0.1472,
      "step": 8132
    },
    {
      "epoch": 5.495437647854005,
      "grad_norm": 3.552870512008667,
      "learning_rate": 2e-05,
      "loss": 0.1443,
      "step": 8133
    },
    {
      "epoch": 5.496113551875633,
      "grad_norm": 2.7776198387145996,
      "learning_rate": 2e-05,
      "loss": 0.1645,
      "step": 8134
    },
    {
      "epoch": 5.496789455897263,
      "grad_norm": 1.6635762453079224,
      "learning_rate": 2e-05,
      "loss": 0.185,
      "step": 8135
    },
    {
      "epoch": 5.497465359918891,
      "grad_norm": 5.148570537567139,
      "learning_rate": 2e-05,
      "loss": 0.214,
      "step": 8136
    },
    {
      "epoch": 5.4981412639405205,
      "grad_norm": 4.951303482055664,
      "learning_rate": 2e-05,
      "loss": 0.1518,
      "step": 8137
    },
    {
      "epoch": 5.49881716796215,
      "grad_norm": 3.5021257400512695,
      "learning_rate": 2e-05,
      "loss": 0.1409,
      "step": 8138
    },
    {
      "epoch": 5.499493071983778,
      "grad_norm": 4.066040992736816,
      "learning_rate": 2e-05,
      "loss": 0.2014,
      "step": 8139
    },
    {
      "epoch": 5.5001689760054076,
      "grad_norm": 7.118319988250732,
      "learning_rate": 2e-05,
      "loss": 0.1383,
      "step": 8140
    },
    {
      "epoch": 5.500844880027036,
      "grad_norm": 4.574049472808838,
      "learning_rate": 2e-05,
      "loss": 0.1563,
      "step": 8141
    },
    {
      "epoch": 5.501520784048665,
      "grad_norm": 3.2503786087036133,
      "learning_rate": 2e-05,
      "loss": 0.1707,
      "step": 8142
    },
    {
      "epoch": 5.502196688070294,
      "grad_norm": 3.8982605934143066,
      "learning_rate": 2e-05,
      "loss": 0.224,
      "step": 8143
    },
    {
      "epoch": 5.502872592091923,
      "grad_norm": 5.437823295593262,
      "learning_rate": 2e-05,
      "loss": 0.188,
      "step": 8144
    },
    {
      "epoch": 5.503548496113552,
      "grad_norm": 4.843693256378174,
      "learning_rate": 2e-05,
      "loss": 0.2056,
      "step": 8145
    },
    {
      "epoch": 5.504224400135181,
      "grad_norm": 2.585991859436035,
      "learning_rate": 2e-05,
      "loss": 0.0846,
      "step": 8146
    },
    {
      "epoch": 5.50490030415681,
      "grad_norm": 6.999101638793945,
      "learning_rate": 2e-05,
      "loss": 0.2141,
      "step": 8147
    },
    {
      "epoch": 5.5055762081784385,
      "grad_norm": 3.0797572135925293,
      "learning_rate": 2e-05,
      "loss": 0.2172,
      "step": 8148
    },
    {
      "epoch": 5.506252112200068,
      "grad_norm": 3.8552968502044678,
      "learning_rate": 2e-05,
      "loss": 0.1977,
      "step": 8149
    },
    {
      "epoch": 5.506928016221696,
      "grad_norm": 6.724528789520264,
      "learning_rate": 2e-05,
      "loss": 0.1495,
      "step": 8150
    },
    {
      "epoch": 5.506928016221696,
      "eval_accuracy": 0.7847128378378379,
      "eval_loss": 0.228960320353508,
      "eval_runtime": 45.4781,
      "eval_samples_per_second": 520.69,
      "eval_steps_per_second": 16.272,
      "step": 8150
    },
    {
      "epoch": 5.507603920243326,
      "grad_norm": 5.112842559814453,
      "learning_rate": 2e-05,
      "loss": 0.1613,
      "step": 8151
    },
    {
      "epoch": 5.508279824264954,
      "grad_norm": 4.49193811416626,
      "learning_rate": 2e-05,
      "loss": 0.2238,
      "step": 8152
    },
    {
      "epoch": 5.508955728286583,
      "grad_norm": 2.868769407272339,
      "learning_rate": 2e-05,
      "loss": 0.2023,
      "step": 8153
    },
    {
      "epoch": 5.509631632308212,
      "grad_norm": 3.4136264324188232,
      "learning_rate": 2e-05,
      "loss": 0.1471,
      "step": 8154
    },
    {
      "epoch": 5.510307536329841,
      "grad_norm": 2.664824962615967,
      "learning_rate": 2e-05,
      "loss": 0.1754,
      "step": 8155
    },
    {
      "epoch": 5.51098344035147,
      "grad_norm": 7.76983118057251,
      "learning_rate": 2e-05,
      "loss": 0.1677,
      "step": 8156
    },
    {
      "epoch": 5.511659344373099,
      "grad_norm": 3.7269368171691895,
      "learning_rate": 2e-05,
      "loss": 0.1492,
      "step": 8157
    },
    {
      "epoch": 5.512335248394728,
      "grad_norm": 2.3503823280334473,
      "learning_rate": 2e-05,
      "loss": 0.1781,
      "step": 8158
    },
    {
      "epoch": 5.5130111524163565,
      "grad_norm": 6.973111152648926,
      "learning_rate": 2e-05,
      "loss": 0.1835,
      "step": 8159
    },
    {
      "epoch": 5.513687056437986,
      "grad_norm": 1.8280682563781738,
      "learning_rate": 2e-05,
      "loss": 0.1369,
      "step": 8160
    },
    {
      "epoch": 5.514362960459614,
      "grad_norm": 1.739029049873352,
      "learning_rate": 2e-05,
      "loss": 0.1322,
      "step": 8161
    },
    {
      "epoch": 5.515038864481244,
      "grad_norm": 2.764719009399414,
      "learning_rate": 2e-05,
      "loss": 0.1578,
      "step": 8162
    },
    {
      "epoch": 5.515714768502873,
      "grad_norm": 5.605714797973633,
      "learning_rate": 2e-05,
      "loss": 0.1143,
      "step": 8163
    },
    {
      "epoch": 5.516390672524501,
      "grad_norm": 1.3873389959335327,
      "learning_rate": 2e-05,
      "loss": 0.1487,
      "step": 8164
    },
    {
      "epoch": 5.517066576546131,
      "grad_norm": 3.4763031005859375,
      "learning_rate": 2e-05,
      "loss": 0.1648,
      "step": 8165
    },
    {
      "epoch": 5.517742480567759,
      "grad_norm": 2.8869946002960205,
      "learning_rate": 2e-05,
      "loss": 0.177,
      "step": 8166
    },
    {
      "epoch": 5.518418384589388,
      "grad_norm": 2.5075039863586426,
      "learning_rate": 2e-05,
      "loss": 0.1533,
      "step": 8167
    },
    {
      "epoch": 5.519094288611017,
      "grad_norm": 2.4915099143981934,
      "learning_rate": 2e-05,
      "loss": 0.1824,
      "step": 8168
    },
    {
      "epoch": 5.519770192632646,
      "grad_norm": 3.6291043758392334,
      "learning_rate": 2e-05,
      "loss": 0.1482,
      "step": 8169
    },
    {
      "epoch": 5.520446096654275,
      "grad_norm": 3.7394089698791504,
      "learning_rate": 2e-05,
      "loss": 0.1571,
      "step": 8170
    },
    {
      "epoch": 5.521122000675904,
      "grad_norm": 1.7682442665100098,
      "learning_rate": 2e-05,
      "loss": 0.1385,
      "step": 8171
    },
    {
      "epoch": 5.521797904697533,
      "grad_norm": 2.691680431365967,
      "learning_rate": 2e-05,
      "loss": 0.1938,
      "step": 8172
    },
    {
      "epoch": 5.522473808719162,
      "grad_norm": 6.185911655426025,
      "learning_rate": 2e-05,
      "loss": 0.1883,
      "step": 8173
    },
    {
      "epoch": 5.523149712740791,
      "grad_norm": 2.520591974258423,
      "learning_rate": 2e-05,
      "loss": 0.1653,
      "step": 8174
    },
    {
      "epoch": 5.523825616762419,
      "grad_norm": 3.3335742950439453,
      "learning_rate": 2e-05,
      "loss": 0.2127,
      "step": 8175
    },
    {
      "epoch": 5.524501520784049,
      "grad_norm": 3.0189685821533203,
      "learning_rate": 2e-05,
      "loss": 0.1697,
      "step": 8176
    },
    {
      "epoch": 5.525177424805678,
      "grad_norm": 6.585131645202637,
      "learning_rate": 2e-05,
      "loss": 0.1633,
      "step": 8177
    },
    {
      "epoch": 5.525853328827306,
      "grad_norm": 4.84722900390625,
      "learning_rate": 2e-05,
      "loss": 0.1181,
      "step": 8178
    },
    {
      "epoch": 5.526529232848936,
      "grad_norm": 5.942724704742432,
      "learning_rate": 2e-05,
      "loss": 0.1623,
      "step": 8179
    },
    {
      "epoch": 5.527205136870564,
      "grad_norm": 1.6563799381256104,
      "learning_rate": 2e-05,
      "loss": 0.1535,
      "step": 8180
    },
    {
      "epoch": 5.5278810408921935,
      "grad_norm": 4.105404853820801,
      "learning_rate": 2e-05,
      "loss": 0.1439,
      "step": 8181
    },
    {
      "epoch": 5.528556944913822,
      "grad_norm": 2.4570422172546387,
      "learning_rate": 2e-05,
      "loss": 0.121,
      "step": 8182
    },
    {
      "epoch": 5.529232848935451,
      "grad_norm": 5.813277721405029,
      "learning_rate": 2e-05,
      "loss": 0.1596,
      "step": 8183
    },
    {
      "epoch": 5.5299087529570805,
      "grad_norm": 2.833808183670044,
      "learning_rate": 2e-05,
      "loss": 0.0992,
      "step": 8184
    },
    {
      "epoch": 5.530584656978709,
      "grad_norm": 2.952531337738037,
      "learning_rate": 2e-05,
      "loss": 0.1266,
      "step": 8185
    },
    {
      "epoch": 5.531260561000338,
      "grad_norm": 1.9694437980651855,
      "learning_rate": 2e-05,
      "loss": 0.1232,
      "step": 8186
    },
    {
      "epoch": 5.531936465021967,
      "grad_norm": 1.9436503648757935,
      "learning_rate": 2e-05,
      "loss": 0.1532,
      "step": 8187
    },
    {
      "epoch": 5.532612369043596,
      "grad_norm": 2.5581021308898926,
      "learning_rate": 2e-05,
      "loss": 0.1982,
      "step": 8188
    },
    {
      "epoch": 5.533288273065224,
      "grad_norm": 2.634136915206909,
      "learning_rate": 2e-05,
      "loss": 0.1766,
      "step": 8189
    },
    {
      "epoch": 5.533964177086854,
      "grad_norm": 2.4210758209228516,
      "learning_rate": 2e-05,
      "loss": 0.1264,
      "step": 8190
    },
    {
      "epoch": 5.534640081108483,
      "grad_norm": 2.3380439281463623,
      "learning_rate": 2e-05,
      "loss": 0.1762,
      "step": 8191
    },
    {
      "epoch": 5.5353159851301115,
      "grad_norm": 3.963581085205078,
      "learning_rate": 2e-05,
      "loss": 0.131,
      "step": 8192
    },
    {
      "epoch": 5.535991889151741,
      "grad_norm": 3.746182918548584,
      "learning_rate": 2e-05,
      "loss": 0.2417,
      "step": 8193
    },
    {
      "epoch": 5.536667793173369,
      "grad_norm": 3.9044690132141113,
      "learning_rate": 2e-05,
      "loss": 0.1802,
      "step": 8194
    },
    {
      "epoch": 5.5373436971949985,
      "grad_norm": 4.878129959106445,
      "learning_rate": 2e-05,
      "loss": 0.1507,
      "step": 8195
    },
    {
      "epoch": 5.538019601216627,
      "grad_norm": 3.8750481605529785,
      "learning_rate": 2e-05,
      "loss": 0.1786,
      "step": 8196
    },
    {
      "epoch": 5.538695505238256,
      "grad_norm": 4.0778679847717285,
      "learning_rate": 2e-05,
      "loss": 0.1198,
      "step": 8197
    },
    {
      "epoch": 5.539371409259886,
      "grad_norm": 5.459138870239258,
      "learning_rate": 2e-05,
      "loss": 0.1518,
      "step": 8198
    },
    {
      "epoch": 5.540047313281514,
      "grad_norm": 3.302785634994507,
      "learning_rate": 2e-05,
      "loss": 0.2229,
      "step": 8199
    },
    {
      "epoch": 5.540723217303143,
      "grad_norm": 7.654257297515869,
      "learning_rate": 2e-05,
      "loss": 0.1213,
      "step": 8200
    },
    {
      "epoch": 5.540723217303143,
      "eval_accuracy": 0.7700591216216216,
      "eval_loss": 0.2180466502904892,
      "eval_runtime": 50.9147,
      "eval_samples_per_second": 465.092,
      "eval_steps_per_second": 14.534,
      "step": 8200
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 14800,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 1000.0,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.18198873651413e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}