{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9904153354632586,
  "eval_steps": 500,
  "global_step": 468,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.006389776357827476,
      "grad_norm": 5.939312845608103,
      "learning_rate": 4.2553191489361704e-07,
      "loss": 0.8883,
      "step": 1
    },
    {
      "epoch": 0.012779552715654952,
      "grad_norm": 5.6459943274568385,
      "learning_rate": 8.510638297872341e-07,
      "loss": 0.8365,
      "step": 2
    },
    {
      "epoch": 0.019169329073482427,
      "grad_norm": 6.290489665408837,
      "learning_rate": 1.276595744680851e-06,
      "loss": 0.8864,
      "step": 3
    },
    {
      "epoch": 0.025559105431309903,
      "grad_norm": 5.8544636560570735,
      "learning_rate": 1.7021276595744682e-06,
      "loss": 0.879,
      "step": 4
    },
    {
      "epoch": 0.03194888178913738,
      "grad_norm": 5.478170251266386,
      "learning_rate": 2.1276595744680853e-06,
      "loss": 0.8245,
      "step": 5
    },
    {
      "epoch": 0.038338658146964855,
      "grad_norm": 4.688336713722404,
      "learning_rate": 2.553191489361702e-06,
      "loss": 0.849,
      "step": 6
    },
    {
      "epoch": 0.04472843450479233,
      "grad_norm": 4.278233500376253,
      "learning_rate": 2.978723404255319e-06,
      "loss": 0.8151,
      "step": 7
    },
    {
      "epoch": 0.051118210862619806,
      "grad_norm": 2.367029075074496,
      "learning_rate": 3.4042553191489363e-06,
      "loss": 0.7566,
      "step": 8
    },
    {
      "epoch": 0.05750798722044728,
      "grad_norm": 2.0707237905380484,
      "learning_rate": 3.8297872340425535e-06,
      "loss": 0.7551,
      "step": 9
    },
    {
      "epoch": 0.06389776357827476,
      "grad_norm": 1.7866725538131587,
      "learning_rate": 4.255319148936171e-06,
      "loss": 0.7457,
      "step": 10
    },
    {
      "epoch": 0.07028753993610223,
      "grad_norm": 3.404477977029493,
      "learning_rate": 4.680851063829788e-06,
      "loss": 0.699,
      "step": 11
    },
    {
      "epoch": 0.07667731629392971,
      "grad_norm": 3.8759087344215093,
      "learning_rate": 5.106382978723404e-06,
      "loss": 0.8106,
      "step": 12
    },
    {
      "epoch": 0.08306709265175719,
      "grad_norm": 4.206484510948959,
      "learning_rate": 5.531914893617022e-06,
      "loss": 0.7639,
      "step": 13
    },
    {
      "epoch": 0.08945686900958466,
      "grad_norm": 3.597855342218176,
      "learning_rate": 5.957446808510638e-06,
      "loss": 0.7143,
      "step": 14
    },
    {
      "epoch": 0.09584664536741214,
      "grad_norm": 3.493223809169224,
      "learning_rate": 6.382978723404256e-06,
      "loss": 0.7495,
      "step": 15
    },
    {
      "epoch": 0.10223642172523961,
      "grad_norm": 2.7730036124678588,
      "learning_rate": 6.808510638297873e-06,
      "loss": 0.7079,
      "step": 16
    },
    {
      "epoch": 0.10862619808306709,
      "grad_norm": 1.99917862715565,
      "learning_rate": 7.234042553191491e-06,
      "loss": 0.6637,
      "step": 17
    },
    {
      "epoch": 0.11501597444089456,
      "grad_norm": 1.7234540912576959,
      "learning_rate": 7.659574468085107e-06,
      "loss": 0.6563,
      "step": 18
    },
    {
      "epoch": 0.12140575079872204,
      "grad_norm": 1.9632219810386513,
      "learning_rate": 8.085106382978723e-06,
      "loss": 0.6753,
      "step": 19
    },
    {
      "epoch": 0.12779552715654952,
      "grad_norm": 2.1614567368781157,
      "learning_rate": 8.510638297872341e-06,
      "loss": 0.6063,
      "step": 20
    },
    {
      "epoch": 0.134185303514377,
      "grad_norm": 1.932507047332365,
      "learning_rate": 8.936170212765958e-06,
      "loss": 0.629,
      "step": 21
    },
    {
      "epoch": 0.14057507987220447,
      "grad_norm": 1.5481507427216847,
      "learning_rate": 9.361702127659576e-06,
      "loss": 0.6444,
      "step": 22
    },
    {
      "epoch": 0.14696485623003194,
      "grad_norm": 1.2177329752055572,
      "learning_rate": 9.787234042553192e-06,
      "loss": 0.641,
      "step": 23
    },
    {
      "epoch": 0.15335463258785942,
      "grad_norm": 1.4206725631658237,
      "learning_rate": 1.0212765957446808e-05,
      "loss": 0.6496,
      "step": 24
    },
    {
      "epoch": 0.1597444089456869,
      "grad_norm": 1.4667519530084758,
      "learning_rate": 1.0638297872340426e-05,
      "loss": 0.6304,
      "step": 25
    },
    {
      "epoch": 0.16613418530351437,
      "grad_norm": 1.2452708214307595,
      "learning_rate": 1.1063829787234044e-05,
      "loss": 0.6413,
      "step": 26
    },
    {
      "epoch": 0.17252396166134185,
      "grad_norm": 0.7759167145812482,
      "learning_rate": 1.1489361702127662e-05,
      "loss": 0.5927,
      "step": 27
    },
    {
      "epoch": 0.17891373801916932,
      "grad_norm": 1.0321850071310563,
      "learning_rate": 1.1914893617021277e-05,
      "loss": 0.5898,
      "step": 28
    },
    {
      "epoch": 0.1853035143769968,
      "grad_norm": 1.0266703283162124,
      "learning_rate": 1.2340425531914895e-05,
      "loss": 0.5895,
      "step": 29
    },
    {
      "epoch": 0.19169329073482427,
      "grad_norm": 0.9674925840183656,
      "learning_rate": 1.2765957446808513e-05,
      "loss": 0.5853,
      "step": 30
    },
    {
      "epoch": 0.19808306709265175,
      "grad_norm": 0.9805374530460067,
      "learning_rate": 1.3191489361702127e-05,
      "loss": 0.5756,
      "step": 31
    },
    {
      "epoch": 0.20447284345047922,
      "grad_norm": 0.9258462587522315,
      "learning_rate": 1.3617021276595745e-05,
      "loss": 0.585,
      "step": 32
    },
    {
      "epoch": 0.2108626198083067,
      "grad_norm": 0.8925757909283792,
      "learning_rate": 1.4042553191489363e-05,
      "loss": 0.584,
      "step": 33
    },
    {
      "epoch": 0.21725239616613418,
      "grad_norm": 0.7738950177471798,
      "learning_rate": 1.4468085106382981e-05,
      "loss": 0.553,
      "step": 34
    },
    {
      "epoch": 0.22364217252396165,
      "grad_norm": 0.823063478968403,
      "learning_rate": 1.4893617021276596e-05,
      "loss": 0.605,
      "step": 35
    },
    {
      "epoch": 0.23003194888178913,
      "grad_norm": 0.7769591687125452,
      "learning_rate": 1.5319148936170214e-05,
      "loss": 0.5398,
      "step": 36
    },
    {
      "epoch": 0.2364217252396166,
      "grad_norm": 0.724085749329525,
      "learning_rate": 1.5744680851063832e-05,
      "loss": 0.5541,
      "step": 37
    },
    {
      "epoch": 0.24281150159744408,
      "grad_norm": 0.7937699708520545,
      "learning_rate": 1.6170212765957446e-05,
      "loss": 0.5516,
      "step": 38
    },
    {
      "epoch": 0.24920127795527156,
      "grad_norm": 0.838915436660047,
      "learning_rate": 1.6595744680851064e-05,
      "loss": 0.5335,
      "step": 39
    },
    {
      "epoch": 0.25559105431309903,
      "grad_norm": 0.7447841608522099,
      "learning_rate": 1.7021276595744682e-05,
      "loss": 0.5414,
      "step": 40
    },
    {
      "epoch": 0.26198083067092653,
      "grad_norm": 0.8714110715774589,
      "learning_rate": 1.74468085106383e-05,
      "loss": 0.556,
      "step": 41
    },
    {
      "epoch": 0.268370607028754,
      "grad_norm": 0.7411083091257032,
      "learning_rate": 1.7872340425531915e-05,
      "loss": 0.5113,
      "step": 42
    },
    {
      "epoch": 0.2747603833865815,
      "grad_norm": 0.7680242229988357,
      "learning_rate": 1.8297872340425533e-05,
      "loss": 0.5684,
      "step": 43
    },
    {
      "epoch": 0.28115015974440893,
      "grad_norm": 0.7374046719973165,
      "learning_rate": 1.872340425531915e-05,
      "loss": 0.577,
      "step": 44
    },
    {
      "epoch": 0.28753993610223644,
      "grad_norm": 0.7709945486266545,
      "learning_rate": 1.914893617021277e-05,
      "loss": 0.5251,
      "step": 45
    },
    {
      "epoch": 0.2939297124600639,
      "grad_norm": 0.7420014515431359,
      "learning_rate": 1.9574468085106384e-05,
      "loss": 0.5358,
      "step": 46
    },
    {
      "epoch": 0.3003194888178914,
      "grad_norm": 0.7321704500165183,
      "learning_rate": 2e-05,
      "loss": 0.56,
      "step": 47
    },
    {
      "epoch": 0.30670926517571884,
      "grad_norm": 0.7855274459068936,
      "learning_rate": 1.9999721578003894e-05,
      "loss": 0.5427,
      "step": 48
    },
    {
      "epoch": 0.31309904153354634,
      "grad_norm": 0.7731901448013743,
      "learning_rate": 1.9998886327519337e-05,
      "loss": 0.5517,
      "step": 49
    },
    {
      "epoch": 0.3194888178913738,
      "grad_norm": 0.7327331638939495,
      "learning_rate": 1.999749429505675e-05,
      "loss": 0.5468,
      "step": 50
    },
    {
      "epoch": 0.3258785942492013,
      "grad_norm": 0.7419176272941636,
      "learning_rate": 1.9995545558130624e-05,
      "loss": 0.548,
      "step": 51
    },
    {
      "epoch": 0.33226837060702874,
      "grad_norm": 0.7873113572745489,
      "learning_rate": 1.9993040225255205e-05,
      "loss": 0.5403,
      "step": 52
    },
    {
      "epoch": 0.33865814696485624,
      "grad_norm": 0.6658693818077654,
      "learning_rate": 1.998997843593845e-05,
      "loss": 0.5054,
      "step": 53
    },
    {
      "epoch": 0.3450479233226837,
      "grad_norm": 0.7409323699660156,
      "learning_rate": 1.9986360360674252e-05,
      "loss": 0.5555,
      "step": 54
    },
    {
      "epoch": 0.3514376996805112,
      "grad_norm": 0.7094024540386773,
      "learning_rate": 1.9982186200932964e-05,
      "loss": 0.535,
      "step": 55
    },
    {
      "epoch": 0.35782747603833864,
      "grad_norm": 0.9538171828161622,
      "learning_rate": 1.9977456189150164e-05,
      "loss": 0.5788,
      "step": 56
    },
    {
      "epoch": 0.36421725239616615,
      "grad_norm": 0.7395956073936416,
      "learning_rate": 1.9972170588713715e-05,
      "loss": 0.5836,
      "step": 57
    },
    {
      "epoch": 0.3706070287539936,
      "grad_norm": 0.8649724098491106,
      "learning_rate": 1.9966329693949098e-05,
      "loss": 0.4918,
      "step": 58
    },
    {
      "epoch": 0.3769968051118211,
      "grad_norm": 0.8838321465136332,
      "learning_rate": 1.9959933830103034e-05,
      "loss": 0.514,
      "step": 59
    },
    {
      "epoch": 0.38338658146964855,
      "grad_norm": 0.7993190296372936,
      "learning_rate": 1.9952983353325358e-05,
      "loss": 0.5191,
      "step": 60
    },
    {
      "epoch": 0.38977635782747605,
      "grad_norm": 1.0003320683454586,
      "learning_rate": 1.9945478650649192e-05,
      "loss": 0.5395,
      "step": 61
    },
    {
      "epoch": 0.3961661341853035,
      "grad_norm": 0.8174378430731664,
      "learning_rate": 1.9937420139969397e-05,
      "loss": 0.5428,
      "step": 62
    },
    {
      "epoch": 0.402555910543131,
      "grad_norm": 0.7497946237416347,
      "learning_rate": 1.9928808270019297e-05,
      "loss": 0.5826,
      "step": 63
    },
    {
      "epoch": 0.40894568690095845,
      "grad_norm": 0.8640015891418493,
      "learning_rate": 1.9919643520345698e-05,
      "loss": 0.5343,
      "step": 64
    },
    {
      "epoch": 0.41533546325878595,
      "grad_norm": 1.0033449675940591,
      "learning_rate": 1.990992640128218e-05,
      "loss": 0.5623,
      "step": 65
    },
    {
      "epoch": 0.4217252396166134,
      "grad_norm": 0.7213392023439573,
      "learning_rate": 1.989965745392068e-05,
      "loss": 0.5314,
      "step": 66
    },
    {
      "epoch": 0.4281150159744409,
      "grad_norm": 0.8863830620435368,
      "learning_rate": 1.988883725008136e-05,
      "loss": 0.5458,
      "step": 67
    },
    {
      "epoch": 0.43450479233226835,
      "grad_norm": 0.7362416113857194,
      "learning_rate": 1.9877466392280773e-05,
      "loss": 0.5428,
      "step": 68
    },
    {
      "epoch": 0.44089456869009586,
      "grad_norm": 0.9788737478764125,
      "learning_rate": 1.9865545513698304e-05,
      "loss": 0.5412,
      "step": 69
    },
    {
      "epoch": 0.4472843450479233,
      "grad_norm": 0.8021931238319318,
      "learning_rate": 1.9853075278140913e-05,
      "loss": 0.5522,
      "step": 70
    },
    {
      "epoch": 0.4536741214057508,
      "grad_norm": 0.85887825826516,
      "learning_rate": 1.984005638000618e-05,
      "loss": 0.5702,
      "step": 71
    },
    {
      "epoch": 0.46006389776357826,
      "grad_norm": 0.7815409304110528,
      "learning_rate": 1.9826489544243623e-05,
      "loss": 0.5482,
      "step": 72
    },
    {
      "epoch": 0.46645367412140576,
      "grad_norm": 0.7632274039193099,
      "learning_rate": 1.981237552631434e-05,
      "loss": 0.5387,
      "step": 73
    },
    {
      "epoch": 0.4728434504792332,
      "grad_norm": 0.7141397681245097,
      "learning_rate": 1.9797715112148937e-05,
      "loss": 0.5546,
      "step": 74
    },
    {
      "epoch": 0.4792332268370607,
      "grad_norm": 0.9775621975797985,
      "learning_rate": 1.9782509118103773e-05,
      "loss": 0.5308,
      "step": 75
    },
    {
      "epoch": 0.48562300319488816,
      "grad_norm": 0.7330287655613912,
      "learning_rate": 1.9766758390915494e-05,
      "loss": 0.5431,
      "step": 76
    },
    {
      "epoch": 0.49201277955271566,
      "grad_norm": 0.8510881793812443,
      "learning_rate": 1.9750463807653873e-05,
      "loss": 0.5187,
      "step": 77
    },
    {
      "epoch": 0.4984025559105431,
      "grad_norm": 0.8486211359598791,
      "learning_rate": 1.9733626275673e-05,
      "loss": 0.55,
      "step": 78
    },
    {
      "epoch": 0.5047923322683706,
      "grad_norm": 0.7806116343240348,
      "learning_rate": 1.9716246732560715e-05,
      "loss": 0.4813,
      "step": 79
    },
    {
      "epoch": 0.5111821086261981,
      "grad_norm": 0.6353268171693698,
      "learning_rate": 1.9698326146086446e-05,
      "loss": 0.5078,
      "step": 80
    },
    {
      "epoch": 0.5175718849840255,
      "grad_norm": 0.8766273238149929,
      "learning_rate": 1.967986551414728e-05,
      "loss": 0.544,
      "step": 81
    },
    {
      "epoch": 0.5239616613418531,
      "grad_norm": 0.7124929308186037,
      "learning_rate": 1.9660865864712413e-05,
      "loss": 0.5547,
      "step": 82
    },
    {
      "epoch": 0.5303514376996805,
      "grad_norm": 0.7985052392042088,
      "learning_rate": 1.9641328255765916e-05,
      "loss": 0.5215,
      "step": 83
    },
    {
      "epoch": 0.536741214057508,
      "grad_norm": 0.6406165534662566,
      "learning_rate": 1.96212537752478e-05,
      "loss": 0.4978,
      "step": 84
    },
    {
      "epoch": 0.5431309904153354,
      "grad_norm": 0.7351154516097609,
      "learning_rate": 1.9600643540993453e-05,
      "loss": 0.5498,
      "step": 85
    },
    {
      "epoch": 0.549520766773163,
      "grad_norm": 0.7023736701907295,
      "learning_rate": 1.9579498700671386e-05,
      "loss": 0.5271,
      "step": 86
    },
    {
      "epoch": 0.5559105431309904,
      "grad_norm": 0.7585882232459317,
      "learning_rate": 1.9557820431719333e-05,
      "loss": 0.5419,
      "step": 87
    },
    {
      "epoch": 0.5623003194888179,
      "grad_norm": 0.7609703576418915,
      "learning_rate": 1.9535609941278676e-05,
      "loss": 0.5115,
      "step": 88
    },
    {
      "epoch": 0.5686900958466453,
      "grad_norm": 0.8355505315864735,
      "learning_rate": 1.9512868466127232e-05,
      "loss": 0.5214,
      "step": 89
    },
    {
      "epoch": 0.5750798722044729,
      "grad_norm": 0.8985531652526637,
      "learning_rate": 1.9489597272610377e-05,
      "loss": 0.5217,
      "step": 90
    },
    {
      "epoch": 0.5814696485623003,
      "grad_norm": 0.8296453033353365,
      "learning_rate": 1.9465797656570546e-05,
      "loss": 0.5235,
      "step": 91
    },
    {
      "epoch": 0.5878594249201278,
      "grad_norm": 0.8400305646818227,
      "learning_rate": 1.944147094327506e-05,
      "loss": 0.5028,
      "step": 92
    },
    {
      "epoch": 0.5942492012779552,
      "grad_norm": 0.8027178786911765,
      "learning_rate": 1.9416618487342333e-05,
      "loss": 0.4932,
      "step": 93
    },
    {
      "epoch": 0.6006389776357828,
      "grad_norm": 0.831774187088593,
      "learning_rate": 1.9391241672666438e-05,
      "loss": 0.5015,
      "step": 94
    },
    {
      "epoch": 0.6070287539936102,
      "grad_norm": 0.6880013029770987,
      "learning_rate": 1.936534191234006e-05,
      "loss": 0.4851,
      "step": 95
    },
    {
      "epoch": 0.6134185303514377,
      "grad_norm": 0.7874317799707009,
      "learning_rate": 1.9338920648575798e-05,
      "loss": 0.4993,
      "step": 96
    },
    {
      "epoch": 0.6198083067092651,
      "grad_norm": 0.8086603460274078,
      "learning_rate": 1.9311979352625837e-05,
      "loss": 0.5472,
      "step": 97
    },
    {
      "epoch": 0.6261980830670927,
      "grad_norm": 0.7330826956678399,
      "learning_rate": 1.928451952470007e-05,
      "loss": 0.5267,
      "step": 98
    },
    {
      "epoch": 0.6325878594249201,
      "grad_norm": 0.6774581525725395,
      "learning_rate": 1.9256542693882505e-05,
      "loss": 0.4964,
      "step": 99
    },
    {
      "epoch": 0.6389776357827476,
      "grad_norm": 0.699096603267482,
      "learning_rate": 1.922805041804617e-05,
      "loss": 0.5051,
      "step": 100
    },
    {
      "epoch": 0.645367412140575,
      "grad_norm": 0.7625712048870313,
      "learning_rate": 1.919904428376632e-05,
      "loss": 0.5301,
      "step": 101
    },
    {
      "epoch": 0.6517571884984026,
      "grad_norm": 0.7442243954348883,
      "learning_rate": 1.916952590623212e-05,
      "loss": 0.5114,
      "step": 102
    },
    {
      "epoch": 0.65814696485623,
      "grad_norm": 0.6319971486530133,
      "learning_rate": 1.9139496929156685e-05,
      "loss": 0.5106,
      "step": 103
    },
    {
      "epoch": 0.6645367412140575,
      "grad_norm": 0.6634321632909197,
      "learning_rate": 1.910895902468557e-05,
      "loss": 0.5211,
      "step": 104
    },
    {
      "epoch": 0.670926517571885,
      "grad_norm": 0.5959863886093508,
      "learning_rate": 1.907791389330363e-05,
      "loss": 0.5124,
      "step": 105
    },
    {
      "epoch": 0.6773162939297125,
      "grad_norm": 0.7995435481771501,
      "learning_rate": 1.904636326374036e-05,
      "loss": 0.52,
      "step": 106
    },
    {
      "epoch": 0.6837060702875399,
      "grad_norm": 0.7383352863257084,
      "learning_rate": 1.9014308892873612e-05,
      "loss": 0.5516,
      "step": 107
    },
    {
      "epoch": 0.6900958466453674,
      "grad_norm": 0.8315686682034763,
      "learning_rate": 1.8981752565631767e-05,
      "loss": 0.4891,
      "step": 108
    },
    {
      "epoch": 0.6964856230031949,
      "grad_norm": 0.6508951908046772,
      "learning_rate": 1.8948696094894354e-05,
      "loss": 0.5591,
      "step": 109
    },
    {
      "epoch": 0.7028753993610224,
      "grad_norm": 0.8009701429333234,
      "learning_rate": 1.8915141321391083e-05,
      "loss": 0.5305,
      "step": 110
    },
    {
      "epoch": 0.7092651757188498,
      "grad_norm": 0.7308231155353043,
      "learning_rate": 1.8881090113599353e-05,
      "loss": 0.5191,
      "step": 111
    },
    {
      "epoch": 0.7156549520766773,
      "grad_norm": 0.7808253591246626,
      "learning_rate": 1.8846544367640218e-05,
      "loss": 0.4906,
      "step": 112
    },
    {
      "epoch": 0.7220447284345048,
      "grad_norm": 0.8760149164716509,
      "learning_rate": 1.881150600717279e-05,
      "loss": 0.532,
      "step": 113
    },
    {
      "epoch": 0.7284345047923323,
      "grad_norm": 0.6776840820965584,
      "learning_rate": 1.8775976983287117e-05,
      "loss": 0.521,
      "step": 114
    },
    {
      "epoch": 0.7348242811501597,
      "grad_norm": 0.8898200759371685,
      "learning_rate": 1.873995927439555e-05,
      "loss": 0.5251,
      "step": 115
    },
    {
      "epoch": 0.7412140575079872,
      "grad_norm": 0.7499195664294598,
      "learning_rate": 1.8703454886122568e-05,
      "loss": 0.5348,
      "step": 116
    },
    {
      "epoch": 0.7476038338658147,
      "grad_norm": 0.8505865320259214,
      "learning_rate": 1.86664658511931e-05,
      "loss": 0.5054,
      "step": 117
    },
    {
      "epoch": 0.7539936102236422,
      "grad_norm": 0.6522705814194051,
      "learning_rate": 1.862899422931934e-05,
      "loss": 0.4996,
      "step": 118
    },
    {
      "epoch": 0.7603833865814696,
      "grad_norm": 0.6577321085300217,
      "learning_rate": 1.859104210708604e-05,
      "loss": 0.5106,
      "step": 119
    },
    {
      "epoch": 0.7667731629392971,
      "grad_norm": 0.6237691862811225,
      "learning_rate": 1.855261159783432e-05,
      "loss": 0.4952,
      "step": 120
    },
    {
      "epoch": 0.7731629392971247,
      "grad_norm": 0.7454463301186427,
      "learning_rate": 1.8513704841543997e-05,
      "loss": 0.5026,
      "step": 121
    },
    {
      "epoch": 0.7795527156549521,
      "grad_norm": 0.7934145741008819,
      "learning_rate": 1.847432400471443e-05,
      "loss": 0.549,
      "step": 122
    },
    {
      "epoch": 0.7859424920127795,
      "grad_norm": 0.6802649460778195,
      "learning_rate": 1.8434471280243854e-05,
      "loss": 0.506,
      "step": 123
    },
    {
      "epoch": 0.792332268370607,
      "grad_norm": 0.7304620814404005,
      "learning_rate": 1.8394148887307286e-05,
      "loss": 0.5383,
      "step": 124
    },
    {
      "epoch": 0.7987220447284346,
      "grad_norm": 0.6957782218041328,
      "learning_rate": 1.8353359071232954e-05,
      "loss": 0.5019,
      "step": 125
    },
    {
      "epoch": 0.805111821086262,
      "grad_norm": 0.6826820301754135,
      "learning_rate": 1.8312104103377266e-05,
      "loss": 0.4937,
      "step": 126
    },
    {
      "epoch": 0.8115015974440895,
      "grad_norm": 0.598906813674708,
      "learning_rate": 1.827038628099831e-05,
      "loss": 0.488,
      "step": 127
    },
    {
      "epoch": 0.8178913738019169,
      "grad_norm": 0.6633449549765212,
      "learning_rate": 1.822820792712797e-05,
      "loss": 0.4971,
      "step": 128
    },
    {
      "epoch": 0.8242811501597445,
      "grad_norm": 0.6710913601976871,
      "learning_rate": 1.8185571390442542e-05,
      "loss": 0.4836,
      "step": 129
    },
    {
      "epoch": 0.8306709265175719,
      "grad_norm": 0.5762592167913351,
      "learning_rate": 1.8142479045131956e-05,
      "loss": 0.5192,
      "step": 130
    },
    {
      "epoch": 0.8370607028753994,
      "grad_norm": 0.6738150719804582,
      "learning_rate": 1.809893329076757e-05,
      "loss": 0.5228,
      "step": 131
    },
    {
      "epoch": 0.8434504792332268,
      "grad_norm": 0.6500141385276138,
      "learning_rate": 1.8054936552168548e-05,
      "loss": 0.5332,
      "step": 132
    },
    {
      "epoch": 0.8498402555910544,
      "grad_norm": 0.5886713445337487,
      "learning_rate": 1.801049127926686e-05,
      "loss": 0.4787,
      "step": 133
    },
    {
      "epoch": 0.8562300319488818,
      "grad_norm": 0.7649082888316344,
      "learning_rate": 1.7965599946970814e-05,
      "loss": 0.4914,
      "step": 134
    },
    {
      "epoch": 0.8626198083067093,
      "grad_norm": 0.8014082101971098,
      "learning_rate": 1.7920265055027285e-05,
      "loss": 0.5171,
      "step": 135
    },
    {
      "epoch": 0.8690095846645367,
      "grad_norm": 0.7706425976997076,
      "learning_rate": 1.7874489127882496e-05,
      "loss": 0.5267,
      "step": 136
    },
    {
      "epoch": 0.8753993610223643,
      "grad_norm": 0.9916919676402002,
      "learning_rate": 1.7828274714541445e-05,
      "loss": 0.5225,
      "step": 137
    },
    {
      "epoch": 0.8817891373801917,
      "grad_norm": 0.7510268200588113,
      "learning_rate": 1.7781624388425974e-05,
      "loss": 0.5056,
      "step": 138
    },
    {
      "epoch": 0.8881789137380192,
      "grad_norm": 1.0622441121039243,
      "learning_rate": 1.773454074723147e-05,
      "loss": 0.4773,
      "step": 139
    },
    {
      "epoch": 0.8945686900958466,
      "grad_norm": 0.8430739649427611,
      "learning_rate": 1.76870264127822e-05,
      "loss": 0.5384,
      "step": 140
    },
    {
      "epoch": 0.9009584664536742,
      "grad_norm": 0.7217579935144164,
      "learning_rate": 1.763908403088534e-05,
      "loss": 0.5108,
      "step": 141
    },
    {
      "epoch": 0.9073482428115016,
      "grad_norm": 1.0377962316213807,
      "learning_rate": 1.759071627118362e-05,
      "loss": 0.4961,
      "step": 142
    },
    {
      "epoch": 0.9137380191693291,
      "grad_norm": 0.6423592379266523,
      "learning_rate": 1.754192582700668e-05,
      "loss": 0.5061,
      "step": 143
    },
    {
      "epoch": 0.9201277955271565,
      "grad_norm": 0.7302473302785821,
      "learning_rate": 1.7492715415221087e-05,
      "loss": 0.5074,
      "step": 144
    },
    {
      "epoch": 0.9265175718849841,
      "grad_norm": 0.709358335589859,
      "learning_rate": 1.7443087776079068e-05,
      "loss": 0.4867,
      "step": 145
    },
    {
      "epoch": 0.9329073482428115,
      "grad_norm": 0.6708691396098521,
      "learning_rate": 1.739304567306588e-05,
      "loss": 0.4955,
      "step": 146
    },
    {
      "epoch": 0.939297124600639,
      "grad_norm": 0.7224681216999947,
      "learning_rate": 1.7342591892745978e-05,
      "loss": 0.5063,
      "step": 147
    },
    {
      "epoch": 0.9456869009584664,
      "grad_norm": 0.692060343253805,
      "learning_rate": 1.72917292446078e-05,
      "loss": 0.5395,
      "step": 148
    },
    {
      "epoch": 0.952076677316294,
      "grad_norm": 0.6503466417528021,
      "learning_rate": 1.7240460560907345e-05,
      "loss": 0.502,
      "step": 149
    },
    {
      "epoch": 0.9584664536741214,
      "grad_norm": 0.7852299377720712,
      "learning_rate": 1.7188788696510477e-05,
      "loss": 0.4982,
      "step": 150
    },
    {
      "epoch": 0.9648562300319489,
      "grad_norm": 0.6487201669982385,
      "learning_rate": 1.7136716528733916e-05,
      "loss": 0.491,
      "step": 151
    },
    {
      "epoch": 0.9712460063897763,
      "grad_norm": 0.7162643954228396,
      "learning_rate": 1.7084246957185036e-05,
      "loss": 0.4715,
      "step": 152
    },
    {
      "epoch": 0.9776357827476039,
      "grad_norm": 0.6882308824929978,
      "learning_rate": 1.703138290360041e-05,
      "loss": 0.4884,
      "step": 153
    },
    {
      "epoch": 0.9840255591054313,
      "grad_norm": 0.6558983320574183,
      "learning_rate": 1.6978127311683103e-05,
      "loss": 0.5053,
      "step": 154
    },
    {
      "epoch": 0.9904153354632588,
      "grad_norm": 0.7793396179424056,
      "learning_rate": 1.6924483146938756e-05,
      "loss": 0.4891,
      "step": 155
    },
    {
      "epoch": 0.9968051118210862,
      "grad_norm": 0.6408866926804856,
      "learning_rate": 1.6870453396510456e-05,
      "loss": 0.5253,
      "step": 156
    },
    {
      "epoch": 1.0031948881789137,
      "grad_norm": 1.219315523607184,
      "learning_rate": 1.681604106901239e-05,
      "loss": 0.7727,
      "step": 157
    },
    {
      "epoch": 1.0095846645367412,
      "grad_norm": 0.7363761093026954,
      "learning_rate": 1.676124919436233e-05,
      "loss": 0.4369,
      "step": 158
    },
    {
      "epoch": 1.0159744408945688,
      "grad_norm": 0.6104584236502094,
      "learning_rate": 1.6706080823612897e-05,
      "loss": 0.4074,
      "step": 159
    },
    {
      "epoch": 1.0223642172523961,
      "grad_norm": 0.6688279966321403,
      "learning_rate": 1.665053902878167e-05,
      "loss": 0.3875,
      "step": 160
    },
    {
      "epoch": 1.0287539936102237,
      "grad_norm": 0.7714442502018793,
      "learning_rate": 1.659462690268013e-05,
      "loss": 0.4693,
      "step": 161
    },
    {
      "epoch": 1.035143769968051,
      "grad_norm": 0.6581337601062037,
      "learning_rate": 1.6538347558741424e-05,
      "loss": 0.4088,
      "step": 162
    },
    {
      "epoch": 1.0415335463258786,
      "grad_norm": 0.8907132608840105,
      "learning_rate": 1.6481704130847013e-05,
      "loss": 0.4421,
      "step": 163
    },
    {
      "epoch": 1.0479233226837061,
      "grad_norm": 0.7281123005456749,
      "learning_rate": 1.642469977315214e-05,
      "loss": 0.4466,
      "step": 164
    },
    {
      "epoch": 1.0543130990415335,
      "grad_norm": 0.8713456380104283,
      "learning_rate": 1.6367337659910223e-05,
      "loss": 0.4205,
      "step": 165
    },
    {
      "epoch": 1.060702875399361,
      "grad_norm": 0.8221934081375099,
      "learning_rate": 1.6309620985296075e-05,
      "loss": 0.4811,
      "step": 166
    },
    {
      "epoch": 1.0670926517571886,
      "grad_norm": 0.6414871893647255,
      "learning_rate": 1.625155296322805e-05,
      "loss": 0.3994,
      "step": 167
    },
    {
      "epoch": 1.073482428115016,
      "grad_norm": 0.6020141646779518,
      "learning_rate": 1.6193136827189067e-05,
      "loss": 0.3544,
      "step": 168
    },
    {
      "epoch": 1.0798722044728435,
      "grad_norm": 0.6620341470772692,
      "learning_rate": 1.6134375830046566e-05,
      "loss": 0.3865,
      "step": 169
    },
    {
      "epoch": 1.0862619808306708,
      "grad_norm": 0.7529504403287007,
      "learning_rate": 1.607527324387137e-05,
      "loss": 0.5138,
      "step": 170
    },
    {
      "epoch": 1.0926517571884984,
      "grad_norm": 0.5746068027715405,
      "learning_rate": 1.6015832359755483e-05,
      "loss": 0.4053,
      "step": 171
    },
    {
      "epoch": 1.099041533546326,
      "grad_norm": 0.6423681462846987,
      "learning_rate": 1.5956056487628832e-05,
      "loss": 0.4878,
      "step": 172
    },
    {
      "epoch": 1.1054313099041533,
      "grad_norm": 0.6192258722203541,
      "learning_rate": 1.5895948956074937e-05,
      "loss": 0.4188,
      "step": 173
    },
    {
      "epoch": 1.1118210862619808,
      "grad_norm": 0.6163929815962566,
      "learning_rate": 1.5835513112145583e-05,
      "loss": 0.3854,
      "step": 174
    },
    {
      "epoch": 1.1182108626198084,
      "grad_norm": 0.5697582130826278,
      "learning_rate": 1.5774752321174428e-05,
      "loss": 0.4073,
      "step": 175
    },
    {
      "epoch": 1.1246006389776357,
      "grad_norm": 0.6223679982213549,
      "learning_rate": 1.571366996658962e-05,
      "loss": 0.4224,
      "step": 176
    },
    {
      "epoch": 1.1309904153354633,
      "grad_norm": 0.6320866995001497,
      "learning_rate": 1.5652269449725375e-05,
      "loss": 0.4248,
      "step": 177
    },
    {
      "epoch": 1.1373801916932909,
      "grad_norm": 0.5297458143137096,
      "learning_rate": 1.5590554189632585e-05,
      "loss": 0.3748,
      "step": 178
    },
    {
      "epoch": 1.1437699680511182,
      "grad_norm": 0.6358623803264093,
      "learning_rate": 1.552852762288843e-05,
      "loss": 0.4747,
      "step": 179
    },
    {
      "epoch": 1.1501597444089458,
      "grad_norm": 0.6278673693951984,
      "learning_rate": 1.5466193203405017e-05,
      "loss": 0.4036,
      "step": 180
    },
    {
      "epoch": 1.156549520766773,
      "grad_norm": 0.6151559404690766,
      "learning_rate": 1.540355440223704e-05,
      "loss": 0.4124,
      "step": 181
    },
    {
      "epoch": 1.1629392971246006,
      "grad_norm": 0.5891168526046718,
      "learning_rate": 1.534061470738852e-05,
      "loss": 0.4637,
      "step": 182
    },
    {
      "epoch": 1.1693290734824282,
      "grad_norm": 0.6184989510741438,
      "learning_rate": 1.527737762361855e-05,
      "loss": 0.4161,
      "step": 183
    },
    {
      "epoch": 1.1757188498402555,
      "grad_norm": 0.6419175983144096,
      "learning_rate": 1.5213846672246139e-05,
      "loss": 0.4533,
      "step": 184
    },
    {
      "epoch": 1.182108626198083,
      "grad_norm": 0.5120065057835093,
      "learning_rate": 1.5150025390954153e-05,
      "loss": 0.375,
      "step": 185
    },
    {
      "epoch": 1.1884984025559104,
      "grad_norm": 0.698299656514935,
      "learning_rate": 1.5085917333592299e-05,
      "loss": 0.4193,
      "step": 186
    },
    {
      "epoch": 1.194888178913738,
      "grad_norm": 0.6532253196374253,
      "learning_rate": 1.5021526069979232e-05,
      "loss": 0.4596,
      "step": 187
    },
    {
      "epoch": 1.2012779552715656,
      "grad_norm": 0.6028992706390495,
      "learning_rate": 1.4956855185703787e-05,
      "loss": 0.4207,
      "step": 188
    },
    {
      "epoch": 1.207667731629393,
      "grad_norm": 0.6185875205322425,
      "learning_rate": 1.48919082819253e-05,
      "loss": 0.4243,
      "step": 189
    },
    {
      "epoch": 1.2140575079872205,
      "grad_norm": 0.6043057260440954,
      "learning_rate": 1.4826688975173085e-05,
      "loss": 0.437,
      "step": 190
    },
    {
      "epoch": 1.220447284345048,
      "grad_norm": 0.6821203725825057,
      "learning_rate": 1.4761200897145063e-05,
      "loss": 0.5037,
      "step": 191
    },
    {
      "epoch": 1.2268370607028753,
      "grad_norm": 0.6008298374460067,
      "learning_rate": 1.4695447694505512e-05,
      "loss": 0.3661,
      "step": 192
    },
    {
      "epoch": 1.233226837060703,
      "grad_norm": 0.7199503911011816,
      "learning_rate": 1.4629433028682014e-05,
      "loss": 0.5043,
      "step": 193
    },
    {
      "epoch": 1.2396166134185305,
      "grad_norm": 0.5180806806501965,
      "learning_rate": 1.456316057566158e-05,
      "loss": 0.3819,
      "step": 194
    },
    {
      "epoch": 1.2460063897763578,
      "grad_norm": 0.6928136290488852,
      "learning_rate": 1.4496634025785938e-05,
      "loss": 0.4263,
      "step": 195
    },
    {
      "epoch": 1.2523961661341854,
      "grad_norm": 0.5665537937081578,
      "learning_rate": 1.4429857083546054e-05,
      "loss": 0.4844,
      "step": 196
    },
    {
      "epoch": 1.2587859424920127,
      "grad_norm": 0.5563376996661965,
      "learning_rate": 1.4362833467375839e-05,
      "loss": 0.4156,
      "step": 197
    },
    {
      "epoch": 1.2651757188498403,
      "grad_norm": 0.5377783142222045,
      "learning_rate": 1.429556690944509e-05,
      "loss": 0.4025,
      "step": 198
    },
    {
      "epoch": 1.2715654952076676,
      "grad_norm": 0.5413290657480134,
      "learning_rate": 1.4228061155451671e-05,
      "loss": 0.4066,
      "step": 199
    },
    {
      "epoch": 1.2779552715654952,
      "grad_norm": 0.5854137012981868,
      "learning_rate": 1.4160319964412943e-05,
      "loss": 0.4631,
      "step": 200
    },
    {
      "epoch": 1.2843450479233227,
      "grad_norm": 0.5549643478979397,
      "learning_rate": 1.4092347108456425e-05,
      "loss": 0.4203,
      "step": 201
    },
    {
      "epoch": 1.29073482428115,
      "grad_norm": 0.6090412261860937,
      "learning_rate": 1.402414637260977e-05,
      "loss": 0.4818,
      "step": 202
    },
    {
      "epoch": 1.2971246006389776,
      "grad_norm": 0.5907663028777652,
      "learning_rate": 1.3955721554589979e-05,
      "loss": 0.4253,
      "step": 203
    },
    {
      "epoch": 1.3035143769968052,
      "grad_norm": 0.6715293474231077,
      "learning_rate": 1.388707646459193e-05,
      "loss": 0.418,
      "step": 204
    },
    {
      "epoch": 1.3099041533546325,
      "grad_norm": 0.6297284099203906,
      "learning_rate": 1.3818214925076226e-05,
      "loss": 0.4147,
      "step": 205
    },
    {
      "epoch": 1.31629392971246,
      "grad_norm": 0.5667392659435537,
      "learning_rate": 1.3749140770556322e-05,
      "loss": 0.4311,
      "step": 206
    },
    {
      "epoch": 1.3226837060702876,
      "grad_norm": 0.6322500298227558,
      "learning_rate": 1.367985784738501e-05,
      "loss": 0.4203,
      "step": 207
    },
    {
      "epoch": 1.329073482428115,
      "grad_norm": 0.6001231519749656,
      "learning_rate": 1.361037001354025e-05,
      "loss": 0.4476,
      "step": 208
    },
    {
      "epoch": 1.3354632587859425,
      "grad_norm": 0.5824816964719087,
      "learning_rate": 1.3540681138410317e-05,
      "loss": 0.3966,
      "step": 209
    },
    {
      "epoch": 1.34185303514377,
      "grad_norm": 0.6931574895039062,
      "learning_rate": 1.3470795102578358e-05,
      "loss": 0.5033,
      "step": 210
    },
    {
      "epoch": 1.3482428115015974,
      "grad_norm": 0.5416798455387101,
      "learning_rate": 1.3400715797606293e-05,
      "loss": 0.4008,
      "step": 211
    },
    {
      "epoch": 1.354632587859425,
      "grad_norm": 0.6524420733651487,
      "learning_rate": 1.3330447125818115e-05,
      "loss": 0.4436,
      "step": 212
    },
    {
      "epoch": 1.3610223642172525,
      "grad_norm": 0.6109732592249346,
      "learning_rate": 1.3259993000082599e-05,
      "loss": 0.4139,
      "step": 213
    },
    {
      "epoch": 1.3674121405750799,
      "grad_norm": 0.6089156914272807,
      "learning_rate": 1.3189357343595405e-05,
      "loss": 0.4287,
      "step": 214
    },
    {
      "epoch": 1.3738019169329074,
      "grad_norm": 0.5100478253771064,
      "learning_rate": 1.3118544089660635e-05,
      "loss": 0.3773,
      "step": 215
    },
    {
      "epoch": 1.3801916932907348,
      "grad_norm": 0.665636694875197,
      "learning_rate": 1.3047557181471784e-05,
      "loss": 0.3922,
      "step": 216
    },
    {
      "epoch": 1.3865814696485623,
      "grad_norm": 0.6421315985345839,
      "learning_rate": 1.2976400571892189e-05,
      "loss": 0.4432,
      "step": 217
    },
    {
      "epoch": 1.3929712460063897,
      "grad_norm": 0.5630313538184247,
      "learning_rate": 1.2905078223234907e-05,
      "loss": 0.4191,
      "step": 218
    },
    {
      "epoch": 1.3993610223642172,
      "grad_norm": 0.6745179299688657,
      "learning_rate": 1.2833594107042078e-05,
      "loss": 0.4333,
      "step": 219
    },
    {
      "epoch": 1.4057507987220448,
      "grad_norm": 0.5454374317298349,
      "learning_rate": 1.2761952203863759e-05,
      "loss": 0.4277,
      "step": 220
    },
    {
      "epoch": 1.4121405750798721,
      "grad_norm": 0.5789898470817271,
      "learning_rate": 1.2690156503036288e-05,
      "loss": 0.4036,
      "step": 221
    },
    {
      "epoch": 1.4185303514376997,
      "grad_norm": 0.5949332555607629,
      "learning_rate": 1.2618211002460135e-05,
      "loss": 0.4785,
      "step": 222
    },
    {
      "epoch": 1.4249201277955272,
      "grad_norm": 0.5587473061257787,
      "learning_rate": 1.2546119708377273e-05,
      "loss": 0.4109,
      "step": 223
    },
    {
      "epoch": 1.4313099041533546,
      "grad_norm": 0.5092986860891325,
      "learning_rate": 1.2473886635148109e-05,
      "loss": 0.4375,
      "step": 224
    },
    {
      "epoch": 1.4376996805111821,
      "grad_norm": 0.49324268916930714,
      "learning_rate": 1.2401515805027924e-05,
      "loss": 0.4339,
      "step": 225
    },
    {
      "epoch": 1.4440894568690097,
      "grad_norm": 0.521973180886398,
      "learning_rate": 1.2329011247942915e-05,
      "loss": 0.3929,
      "step": 226
    },
    {
      "epoch": 1.450479233226837,
      "grad_norm": 0.49449177011483986,
      "learning_rate": 1.2256377001265785e-05,
      "loss": 0.3688,
      "step": 227
    },
    {
      "epoch": 1.4568690095846646,
      "grad_norm": 0.688803254610615,
      "learning_rate": 1.2183617109590925e-05,
      "loss": 0.5054,
      "step": 228
    },
    {
      "epoch": 1.4632587859424921,
      "grad_norm": 0.5222053229962778,
      "learning_rate": 1.2110735624509184e-05,
      "loss": 0.3946,
      "step": 229
    },
    {
      "epoch": 1.4696485623003195,
      "grad_norm": 0.5032950143354372,
      "learning_rate": 1.2037736604382279e-05,
      "loss": 0.4158,
      "step": 230
    },
    {
      "epoch": 1.476038338658147,
      "grad_norm": 0.5739740428574489,
      "learning_rate": 1.1964624114116784e-05,
      "loss": 0.3794,
      "step": 231
    },
    {
      "epoch": 1.4824281150159744,
      "grad_norm": 0.6710884805647371,
      "learning_rate": 1.1891402224937805e-05,
      "loss": 0.466,
      "step": 232
    },
    {
      "epoch": 1.488817891373802,
      "grad_norm": 0.4582879331386117,
      "learning_rate": 1.1818075014162243e-05,
      "loss": 0.3585,
      "step": 233
    },
    {
      "epoch": 1.4952076677316293,
      "grad_norm": 0.5448073298830566,
      "learning_rate": 1.1744646564971777e-05,
      "loss": 0.4238,
      "step": 234
    },
    {
      "epoch": 1.5015974440894568,
      "grad_norm": 0.5760693930458392,
      "learning_rate": 1.1671120966185486e-05,
      "loss": 0.4324,
      "step": 235
    },
    {
      "epoch": 1.5079872204472844,
      "grad_norm": 0.49993740420542576,
      "learning_rate": 1.159750231203217e-05,
      "loss": 0.4078,
      "step": 236
    },
    {
      "epoch": 1.5143769968051117,
      "grad_norm": 0.563168928495261,
      "learning_rate": 1.1523794701922351e-05,
      "loss": 0.4351,
      "step": 237
    },
    {
      "epoch": 1.5207667731629393,
      "grad_norm": 0.5313983639628916,
      "learning_rate": 1.145000224022002e-05,
      "loss": 0.3571,
      "step": 238
    },
    {
      "epoch": 1.5271565495207668,
      "grad_norm": 0.501524364831029,
      "learning_rate": 1.1376129036014073e-05,
      "loss": 0.4304,
      "step": 239
    },
    {
      "epoch": 1.5335463258785942,
      "grad_norm": 0.44709921144798975,
      "learning_rate": 1.1302179202889505e-05,
      "loss": 0.3948,
      "step": 240
    },
    {
      "epoch": 1.5399361022364217,
      "grad_norm": 0.6024736966395187,
      "learning_rate": 1.1228156858698344e-05,
      "loss": 0.5055,
      "step": 241
    },
    {
      "epoch": 1.5463258785942493,
      "grad_norm": 0.5039190418971187,
      "learning_rate": 1.1154066125330358e-05,
      "loss": 0.3853,
      "step": 242
    },
    {
      "epoch": 1.5527156549520766,
      "grad_norm": 0.5082416278400913,
      "learning_rate": 1.107991112848352e-05,
      "loss": 0.365,
      "step": 243
    },
    {
      "epoch": 1.5591054313099042,
      "grad_norm": 0.571424017024242,
      "learning_rate": 1.100569599743428e-05,
      "loss": 0.4157,
      "step": 244
    },
    {
      "epoch": 1.5654952076677318,
      "grad_norm": 0.5185521716853748,
      "learning_rate": 1.0931424864807624e-05,
      "loss": 0.4278,
      "step": 245
    },
    {
      "epoch": 1.571884984025559,
      "grad_norm": 0.48800280720450917,
      "learning_rate": 1.0857101866346953e-05,
      "loss": 0.3713,
      "step": 246
    },
    {
      "epoch": 1.5782747603833864,
      "grad_norm": 0.6095490026665179,
      "learning_rate": 1.0782731140683786e-05,
      "loss": 0.4746,
      "step": 247
    },
    {
      "epoch": 1.5846645367412142,
      "grad_norm": 0.44928616025284457,
      "learning_rate": 1.0708316829107295e-05,
      "loss": 0.3991,
      "step": 248
    },
    {
      "epoch": 1.5910543130990416,
      "grad_norm": 0.527002704565493,
      "learning_rate": 1.0633863075333713e-05,
      "loss": 0.4499,
      "step": 249
    },
    {
      "epoch": 1.5974440894568689,
      "grad_norm": 0.5073694564198865,
      "learning_rate": 1.0559374025275597e-05,
      "loss": 0.4451,
      "step": 250
    },
    {
      "epoch": 1.6038338658146964,
      "grad_norm": 0.43991601364268823,
      "learning_rate": 1.0484853826810942e-05,
      "loss": 0.3585,
      "step": 251
    },
    {
      "epoch": 1.610223642172524,
      "grad_norm": 0.5793726292991741,
      "learning_rate": 1.0410306629552231e-05,
      "loss": 0.4504,
      "step": 252
    },
    {
      "epoch": 1.6166134185303513,
      "grad_norm": 0.5109980782065752,
      "learning_rate": 1.0335736584615357e-05,
      "loss": 0.4323,
      "step": 253
    },
    {
      "epoch": 1.623003194888179,
      "grad_norm": 0.5541438519921948,
      "learning_rate": 1.0261147844388472e-05,
      "loss": 0.4108,
      "step": 254
    },
    {
      "epoch": 1.6293929712460065,
      "grad_norm": 0.49521914364363934,
      "learning_rate": 1.0186544562300766e-05,
      "loss": 0.3646,
      "step": 255
    },
    {
      "epoch": 1.6357827476038338,
      "grad_norm": 0.5712895311131073,
      "learning_rate": 1.011193089259118e-05,
      "loss": 0.4436,
      "step": 256
    },
    {
      "epoch": 1.6421725239616614,
      "grad_norm": 0.4759908770222787,
      "learning_rate": 1.0037310990077083e-05,
      "loss": 0.4114,
      "step": 257
    },
    {
      "epoch": 1.648562300319489,
      "grad_norm": 0.5916396180237442,
      "learning_rate": 9.962689009922918e-06,
      "loss": 0.4351,
      "step": 258
    },
    {
      "epoch": 1.6549520766773163,
      "grad_norm": 0.5121135747386429,
      "learning_rate": 9.888069107408824e-06,
      "loss": 0.4054,
      "step": 259
    },
    {
      "epoch": 1.6613418530351438,
      "grad_norm": 0.5138320946770819,
      "learning_rate": 9.813455437699238e-06,
      "loss": 0.3825,
      "step": 260
    },
    {
      "epoch": 1.6677316293929714,
      "grad_norm": 0.6049983944001385,
      "learning_rate": 9.738852155611531e-06,
      "loss": 0.4499,
      "step": 261
    },
    {
      "epoch": 1.6741214057507987,
      "grad_norm": 0.5783031442949135,
      "learning_rate": 9.664263415384644e-06,
      "loss": 0.4481,
      "step": 262
    },
    {
      "epoch": 1.680511182108626,
      "grad_norm": 0.5079212196836366,
      "learning_rate": 9.589693370447769e-06,
      "loss": 0.4215,
      "step": 263
    },
    {
      "epoch": 1.6869009584664538,
      "grad_norm": 0.5362072676088528,
      "learning_rate": 9.515146173189058e-06,
      "loss": 0.382,
      "step": 264
    },
    {
      "epoch": 1.6932907348242812,
      "grad_norm": 0.5939862572487508,
      "learning_rate": 9.440625974724408e-06,
      "loss": 0.4359,
      "step": 265
    },
    {
      "epoch": 1.6996805111821085,
      "grad_norm": 0.509251006459368,
      "learning_rate": 9.366136924666289e-06,
      "loss": 0.4187,
      "step": 266
    },
    {
      "epoch": 1.706070287539936,
      "grad_norm": 0.5360988896841928,
      "learning_rate": 9.291683170892712e-06,
      "loss": 0.4397,
      "step": 267
    },
    {
      "epoch": 1.7124600638977636,
      "grad_norm": 0.4969550304271485,
      "learning_rate": 9.217268859316219e-06,
      "loss": 0.4203,
      "step": 268
    },
    {
      "epoch": 1.718849840255591,
      "grad_norm": 0.5191197891519107,
      "learning_rate": 9.142898133653049e-06,
      "loss": 0.3609,
      "step": 269
    },
    {
      "epoch": 1.7252396166134185,
      "grad_norm": 0.4714057411588144,
      "learning_rate": 9.068575135192377e-06,
      "loss": 0.4465,
      "step": 270
    },
    {
      "epoch": 1.731629392971246,
      "grad_norm": 0.4652015886422851,
      "learning_rate": 8.994304002565723e-06,
      "loss": 0.4304,
      "step": 271
    },
    {
      "epoch": 1.7380191693290734,
      "grad_norm": 0.5542538380015932,
      "learning_rate": 8.920088871516482e-06,
      "loss": 0.4401,
      "step": 272
    },
    {
      "epoch": 1.744408945686901,
      "grad_norm": 0.5308379584880095,
      "learning_rate": 8.845933874669645e-06,
      "loss": 0.4408,
      "step": 273
    },
    {
      "epoch": 1.7507987220447285,
      "grad_norm": 0.5930183810947137,
      "learning_rate": 8.771843141301659e-06,
      "loss": 0.4715,
      "step": 274
    },
    {
      "epoch": 1.7571884984025559,
      "grad_norm": 0.5544444716982725,
      "learning_rate": 8.697820797110499e-06,
      "loss": 0.4054,
      "step": 275
    },
    {
      "epoch": 1.7635782747603834,
      "grad_norm": 0.5332926586542652,
      "learning_rate": 8.62387096398593e-06,
      "loss": 0.4075,
      "step": 276
    },
    {
      "epoch": 1.769968051118211,
      "grad_norm": 0.46783125273604637,
      "learning_rate": 8.549997759779981e-06,
      "loss": 0.383,
      "step": 277
    },
    {
      "epoch": 1.7763578274760383,
      "grad_norm": 0.5521530233312871,
      "learning_rate": 8.47620529807765e-06,
      "loss": 0.4533,
      "step": 278
    },
    {
      "epoch": 1.7827476038338657,
      "grad_norm": 0.5736693711263415,
      "learning_rate": 8.402497687967837e-06,
      "loss": 0.4041,
      "step": 279
    },
    {
      "epoch": 1.7891373801916934,
      "grad_norm": 0.5178503013621817,
      "learning_rate": 8.328879033814516e-06,
      "loss": 0.397,
      "step": 280
    },
    {
      "epoch": 1.7955271565495208,
      "grad_norm": 0.5005614658444079,
      "learning_rate": 8.255353435028228e-06,
      "loss": 0.432,
      "step": 281
    },
    {
      "epoch": 1.8019169329073481,
      "grad_norm": 0.466589550480172,
      "learning_rate": 8.181924985837762e-06,
      "loss": 0.3921,
      "step": 282
    },
    {
      "epoch": 1.8083067092651757,
      "grad_norm": 0.492210299786068,
      "learning_rate": 8.108597775062198e-06,
      "loss": 0.3952,
      "step": 283
    },
    {
      "epoch": 1.8146964856230032,
      "grad_norm": 0.5488180113006943,
      "learning_rate": 8.035375885883217e-06,
      "loss": 0.4721,
      "step": 284
    },
    {
      "epoch": 1.8210862619808306,
      "grad_norm": 0.46872330736771156,
      "learning_rate": 7.962263395617724e-06,
      "loss": 0.4027,
      "step": 285
    },
    {
      "epoch": 1.8274760383386581,
      "grad_norm": 0.5921155853332334,
      "learning_rate": 7.88926437549082e-06,
      "loss": 0.4784,
      "step": 286
    },
    {
      "epoch": 1.8338658146964857,
      "grad_norm": 0.44822680013304916,
      "learning_rate": 7.81638289040908e-06,
      "loss": 0.3745,
      "step": 287
    },
    {
      "epoch": 1.840255591054313,
      "grad_norm": 0.5070334055575236,
      "learning_rate": 7.743622998734217e-06,
      "loss": 0.4493,
      "step": 288
    },
    {
      "epoch": 1.8466453674121406,
      "grad_norm": 0.45829357905039364,
      "learning_rate": 7.670988752057088e-06,
      "loss": 0.3906,
      "step": 289
    },
    {
      "epoch": 1.8530351437699681,
      "grad_norm": 0.4401957393457272,
      "learning_rate": 7.598484194972076e-06,
      "loss": 0.4351,
      "step": 290
    },
    {
      "epoch": 1.8594249201277955,
      "grad_norm": 0.45173410773764394,
      "learning_rate": 7.526113364851891e-06,
      "loss": 0.404,
      "step": 291
    },
    {
      "epoch": 1.865814696485623,
      "grad_norm": 0.48332838338770834,
      "learning_rate": 7.453880291622726e-06,
      "loss": 0.39,
      "step": 292
    },
    {
      "epoch": 1.8722044728434506,
      "grad_norm": 0.49775533365807295,
      "learning_rate": 7.381788997539869e-06,
      "loss": 0.4263,
      "step": 293
    },
    {
      "epoch": 1.878594249201278,
      "grad_norm": 0.45957391883705206,
      "learning_rate": 7.309843496963715e-06,
      "loss": 0.362,
      "step": 294
    },
    {
      "epoch": 1.8849840255591053,
      "grad_norm": 0.5317384668186387,
      "learning_rate": 7.238047796136247e-06,
      "loss": 0.4593,
      "step": 295
    },
    {
      "epoch": 1.891373801916933,
      "grad_norm": 0.4766410704227857,
      "learning_rate": 7.166405892957926e-06,
      "loss": 0.3991,
      "step": 296
    },
    {
      "epoch": 1.8977635782747604,
      "grad_norm": 0.4468475803781737,
      "learning_rate": 7.094921776765095e-06,
      "loss": 0.4128,
      "step": 297
    },
    {
      "epoch": 1.9041533546325877,
      "grad_norm": 0.5191666964746032,
      "learning_rate": 7.023599428107815e-06,
      "loss": 0.457,
      "step": 298
    },
    {
      "epoch": 1.9105431309904153,
      "grad_norm": 0.44496722949584133,
      "learning_rate": 6.95244281852822e-06,
      "loss": 0.3908,
      "step": 299
    },
    {
      "epoch": 1.9169329073482428,
      "grad_norm": 0.4827518065178973,
      "learning_rate": 6.881455910339369e-06,
      "loss": 0.4515,
      "step": 300
    },
    {
      "epoch": 1.9233226837060702,
      "grad_norm": 0.5066972710057507,
      "learning_rate": 6.8106426564045965e-06,
      "loss": 0.4084,
      "step": 301
    },
    {
      "epoch": 1.9297124600638977,
      "grad_norm": 0.468517851202735,
      "learning_rate": 6.740006999917406e-06,
      "loss": 0.4292,
      "step": 302
    },
    {
      "epoch": 1.9361022364217253,
      "grad_norm": 0.5620319742163961,
      "learning_rate": 6.669552874181888e-06,
      "loss": 0.4285,
      "step": 303
    },
    {
      "epoch": 1.9424920127795526,
      "grad_norm": 0.46212229948639855,
      "learning_rate": 6.599284202393709e-06,
      "loss": 0.4009,
      "step": 304
    },
    {
      "epoch": 1.9488817891373802,
      "grad_norm": 0.4618831752703247,
      "learning_rate": 6.529204897421644e-06,
      "loss": 0.471,
      "step": 305
    },
    {
      "epoch": 1.9552715654952078,
      "grad_norm": 0.4453523115078921,
      "learning_rate": 6.4593188615896855e-06,
      "loss": 0.4149,
      "step": 306
    },
    {
      "epoch": 1.961661341853035,
      "grad_norm": 0.5565537423113835,
      "learning_rate": 6.389629986459756e-06,
      "loss": 0.4595,
      "step": 307
    },
    {
      "epoch": 1.9680511182108626,
      "grad_norm": 0.4817525299898893,
      "learning_rate": 6.3201421526149945e-06,
      "loss": 0.351,
      "step": 308
    },
    {
      "epoch": 1.9744408945686902,
      "grad_norm": 0.49529548437523874,
      "learning_rate": 6.250859229443684e-06,
      "loss": 0.4219,
      "step": 309
    },
    {
      "epoch": 1.9808306709265175,
      "grad_norm": 0.4677244453368462,
      "learning_rate": 6.181785074923778e-06,
      "loss": 0.4204,
      "step": 310
    },
    {
      "epoch": 1.9872204472843449,
      "grad_norm": 0.5533811435739016,
      "learning_rate": 6.112923535408074e-06,
      "loss": 0.4005,
      "step": 311
    },
    {
      "epoch": 1.9936102236421727,
      "grad_norm": 0.6829408529085472,
      "learning_rate": 6.044278445410025e-06,
      "loss": 0.457,
      "step": 312
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.7425708478665342,
      "learning_rate": 5.975853627390233e-06,
      "loss": 0.5428,
      "step": 313
    },
    {
      "epoch": 2.0063897763578273,
      "grad_norm": 0.6185014849673437,
      "learning_rate": 5.907652891543576e-06,
      "loss": 0.3446,
      "step": 314
    },
    {
      "epoch": 2.012779552715655,
      "grad_norm": 0.550249298005834,
      "learning_rate": 5.839680035587061e-06,
      "loss": 0.3567,
      "step": 315
    },
    {
      "epoch": 2.0191693290734825,
      "grad_norm": 0.6406389855268264,
      "learning_rate": 5.771938844548331e-06,
      "loss": 0.3439,
      "step": 316
    },
    {
      "epoch": 2.02555910543131,
      "grad_norm": 0.7498793828860592,
      "learning_rate": 5.704433090554912e-06,
      "loss": 0.3525,
      "step": 317
    },
    {
      "epoch": 2.0319488817891376,
      "grad_norm": 0.6058522485872244,
      "learning_rate": 5.6371665326241635e-06,
      "loss": 0.3493,
      "step": 318
    },
    {
      "epoch": 2.038338658146965,
      "grad_norm": 0.4684273692024481,
      "learning_rate": 5.570142916453944e-06,
      "loss": 0.3529,
      "step": 319
    },
    {
      "epoch": 2.0447284345047922,
      "grad_norm": 0.5736798457088337,
      "learning_rate": 5.503365974214059e-06,
      "loss": 0.3601,
      "step": 320
    },
    {
      "epoch": 2.0511182108626196,
      "grad_norm": 0.6933501821794038,
      "learning_rate": 5.436839424338426e-06,
      "loss": 0.3495,
      "step": 321
    },
    {
      "epoch": 2.0575079872204474,
      "grad_norm": 0.5220030865474038,
      "learning_rate": 5.37056697131799e-06,
      "loss": 0.3313,
      "step": 322
    },
    {
      "epoch": 2.0638977635782747,
      "grad_norm": 0.48932445764444477,
      "learning_rate": 5.304552305494493e-06,
      "loss": 0.3429,
      "step": 323
    },
    {
      "epoch": 2.070287539936102,
      "grad_norm": 0.5367328609042773,
      "learning_rate": 5.238799102854941e-06,
      "loss": 0.3631,
      "step": 324
    },
    {
      "epoch": 2.07667731629393,
      "grad_norm": 0.4753482664474989,
      "learning_rate": 5.173311024826916e-06,
      "loss": 0.3164,
      "step": 325
    },
    {
      "epoch": 2.083067092651757,
      "grad_norm": 0.5857488280714297,
      "learning_rate": 5.108091718074706e-06,
      "loss": 0.3684,
      "step": 326
    },
    {
      "epoch": 2.0894568690095845,
      "grad_norm": 0.5577030340771072,
      "learning_rate": 5.043144814296214e-06,
      "loss": 0.3393,
      "step": 327
    },
    {
      "epoch": 2.0958466453674123,
      "grad_norm": 0.5001282088257198,
      "learning_rate": 4.9784739300207675e-06,
      "loss": 0.3289,
      "step": 328
    },
    {
      "epoch": 2.1022364217252396,
      "grad_norm": 0.49155760192395415,
      "learning_rate": 4.914082666407705e-06,
      "loss": 0.3126,
      "step": 329
    },
    {
      "epoch": 2.108626198083067,
      "grad_norm": 0.4721435423436379,
      "learning_rate": 4.849974609045849e-06,
      "loss": 0.3412,
      "step": 330
    },
    {
      "epoch": 2.1150159744408947,
      "grad_norm": 0.4887160051279591,
      "learning_rate": 4.786153327753865e-06,
      "loss": 0.344,
      "step": 331
    },
    {
      "epoch": 2.121405750798722,
      "grad_norm": 0.4325584192349116,
      "learning_rate": 4.722622376381455e-06,
      "loss": 0.3316,
      "step": 332
    },
    {
      "epoch": 2.1277955271565494,
      "grad_norm": 0.4294996718579918,
      "learning_rate": 4.659385292611479e-06,
      "loss": 0.3339,
      "step": 333
    },
    {
      "epoch": 2.134185303514377,
      "grad_norm": 0.41116978589443015,
      "learning_rate": 4.59644559776296e-06,
      "loss": 0.3325,
      "step": 334
    },
    {
      "epoch": 2.1405750798722045,
      "grad_norm": 0.4411913227153377,
      "learning_rate": 4.533806796594989e-06,
      "loss": 0.342,
      "step": 335
    },
    {
      "epoch": 2.146964856230032,
      "grad_norm": 0.44530187962960605,
      "learning_rate": 4.471472377111574e-06,
      "loss": 0.364,
      "step": 336
    },
    {
      "epoch": 2.1533546325878596,
      "grad_norm": 0.4681980949823063,
      "learning_rate": 4.409445810367421e-06,
      "loss": 0.3749,
      "step": 337
    },
    {
      "epoch": 2.159744408945687,
      "grad_norm": 0.4082359456108807,
      "learning_rate": 4.347730550274628e-06,
      "loss": 0.3586,
      "step": 338
    },
    {
      "epoch": 2.1661341853035143,
      "grad_norm": 0.41583681737409806,
      "learning_rate": 4.286330033410384e-06,
      "loss": 0.3047,
      "step": 339
    },
    {
      "epoch": 2.1725239616613417,
      "grad_norm": 0.4540111795301057,
      "learning_rate": 4.2252476788255735e-06,
      "loss": 0.3254,
      "step": 340
    },
    {
      "epoch": 2.1789137380191694,
      "grad_norm": 0.45433063555043707,
      "learning_rate": 4.164486887854424e-06,
      "loss": 0.3459,
      "step": 341
    },
    {
      "epoch": 2.1853035143769968,
      "grad_norm": 0.43892759194821435,
      "learning_rate": 4.104051043925068e-06,
      "loss": 0.354,
      "step": 342
    },
    {
      "epoch": 2.191693290734824,
      "grad_norm": 0.43626351327017776,
      "learning_rate": 4.043943512371171e-06,
      "loss": 0.3484,
      "step": 343
    },
    {
      "epoch": 2.198083067092652,
      "grad_norm": 0.41866320314755173,
      "learning_rate": 3.984167640244518e-06,
      "loss": 0.3595,
      "step": 344
    },
    {
      "epoch": 2.2044728434504792,
      "grad_norm": 0.43171890711936267,
      "learning_rate": 3.924726756128632e-06,
      "loss": 0.3683,
      "step": 345
    },
    {
      "epoch": 2.2108626198083066,
      "grad_norm": 0.39657734240201464,
      "learning_rate": 3.8656241699534396e-06,
      "loss": 0.3319,
      "step": 346
    },
    {
      "epoch": 2.2172523961661343,
      "grad_norm": 0.4390716578058822,
      "learning_rate": 3.8068631728109364e-06,
      "loss": 0.3718,
      "step": 347
    },
    {
      "epoch": 2.2236421725239617,
      "grad_norm": 0.43526087538819297,
      "learning_rate": 3.7484470367719493e-06,
      "loss": 0.3586,
      "step": 348
    },
    {
      "epoch": 2.230031948881789,
      "grad_norm": 0.41150940940323066,
      "learning_rate": 3.6903790147039286e-06,
      "loss": 0.3219,
      "step": 349
    },
    {
      "epoch": 2.236421725239617,
      "grad_norm": 0.4434399201886998,
      "learning_rate": 3.6326623400897797e-06,
      "loss": 0.3496,
      "step": 350
    },
    {
      "epoch": 2.242811501597444,
      "grad_norm": 0.4226836887457401,
      "learning_rate": 3.575300226847863e-06,
      "loss": 0.3369,
      "step": 351
    },
    {
      "epoch": 2.2492012779552715,
      "grad_norm": 0.449011167578069,
      "learning_rate": 3.5182958691529945e-06,
      "loss": 0.3456,
      "step": 352
    },
    {
      "epoch": 2.255591054313099,
      "grad_norm": 0.40337825402760535,
      "learning_rate": 3.4616524412585797e-06,
      "loss": 0.3569,
      "step": 353
    },
    {
      "epoch": 2.2619808306709266,
      "grad_norm": 0.4417946658839781,
      "learning_rate": 3.405373097319875e-06,
      "loss": 0.2965,
      "step": 354
    },
    {
      "epoch": 2.268370607028754,
      "grad_norm": 0.4226400522886019,
      "learning_rate": 3.3494609712183323e-06,
      "loss": 0.3143,
      "step": 355
    },
    {
      "epoch": 2.2747603833865817,
      "grad_norm": 0.42990286288068097,
      "learning_rate": 3.2939191763871047e-06,
      "loss": 0.325,
      "step": 356
    },
    {
      "epoch": 2.281150159744409,
      "grad_norm": 0.48668783932063814,
      "learning_rate": 3.2387508056376726e-06,
      "loss": 0.3393,
      "step": 357
    },
    {
      "epoch": 2.2875399361022364,
      "grad_norm": 0.45993768274586794,
      "learning_rate": 3.183958930987612e-06,
      "loss": 0.3271,
      "step": 358
    },
    {
      "epoch": 2.2939297124600637,
      "grad_norm": 0.3696554039297162,
      "learning_rate": 3.1295466034895482e-06,
      "loss": 0.356,
      "step": 359
    },
    {
      "epoch": 2.3003194888178915,
      "grad_norm": 0.41320713175521795,
      "learning_rate": 3.0755168530612444e-06,
      "loss": 0.321,
      "step": 360
    },
    {
      "epoch": 2.306709265175719,
      "grad_norm": 0.42742117582586886,
      "learning_rate": 3.021872688316896e-06,
      "loss": 0.3161,
      "step": 361
    },
    {
      "epoch": 2.313099041533546,
      "grad_norm": 0.41256761998572333,
      "learning_rate": 2.968617096399592e-06,
      "loss": 0.3095,
      "step": 362
    },
    {
      "epoch": 2.319488817891374,
      "grad_norm": 0.418828546634445,
      "learning_rate": 2.9157530428149683e-06,
      "loss": 0.3434,
      "step": 363
    },
    {
      "epoch": 2.3258785942492013,
      "grad_norm": 0.40161025260603606,
      "learning_rate": 2.8632834712660882e-06,
      "loss": 0.3203,
      "step": 364
    },
    {
      "epoch": 2.3322683706070286,
      "grad_norm": 0.4182135730630846,
      "learning_rate": 2.8112113034895273e-06,
      "loss": 0.3135,
      "step": 365
    },
    {
      "epoch": 2.3386581469648564,
      "grad_norm": 0.40613861239190874,
      "learning_rate": 2.7595394390926557e-06,
      "loss": 0.3643,
      "step": 366
    },
    {
      "epoch": 2.3450479233226837,
      "grad_norm": 0.41302826534326115,
      "learning_rate": 2.708270755392207e-06,
      "loss": 0.3424,
      "step": 367
    },
    {
      "epoch": 2.351437699680511,
      "grad_norm": 0.3930172226452873,
      "learning_rate": 2.657408107254027e-06,
      "loss": 0.3429,
      "step": 368
    },
    {
      "epoch": 2.357827476038339,
      "grad_norm": 0.41867946397201244,
      "learning_rate": 2.60695432693412e-06,
      "loss": 0.3657,
      "step": 369
    },
    {
      "epoch": 2.364217252396166,
      "grad_norm": 0.3952956485859315,
      "learning_rate": 2.5569122239209366e-06,
      "loss": 0.2999,
      "step": 370
    },
    {
      "epoch": 2.3706070287539935,
      "grad_norm": 0.38547686501477857,
      "learning_rate": 2.507284584778913e-06,
      "loss": 0.3229,
      "step": 371
    },
    {
      "epoch": 2.376996805111821,
      "grad_norm": 0.4071024941451391,
      "learning_rate": 2.4580741729933246e-06,
      "loss": 0.3398,
      "step": 372
    },
    {
      "epoch": 2.3833865814696487,
      "grad_norm": 0.3975504941590724,
      "learning_rate": 2.4092837288163807e-06,
      "loss": 0.3246,
      "step": 373
    },
    {
      "epoch": 2.389776357827476,
      "grad_norm": 0.4007814100724813,
      "learning_rate": 2.3609159691146577e-06,
      "loss": 0.3834,
      "step": 374
    },
    {
      "epoch": 2.3961661341853033,
      "grad_norm": 0.4132465009460835,
      "learning_rate": 2.312973587217798e-06,
      "loss": 0.3552,
      "step": 375
    },
    {
      "epoch": 2.402555910543131,
      "grad_norm": 0.38136466401080477,
      "learning_rate": 2.2654592527685305e-06,
      "loss": 0.3394,
      "step": 376
    },
    {
      "epoch": 2.4089456869009584,
      "grad_norm": 0.38378352088164336,
      "learning_rate": 2.2183756115740274e-06,
      "loss": 0.3417,
      "step": 377
    },
    {
      "epoch": 2.415335463258786,
      "grad_norm": 0.42705803790306746,
      "learning_rate": 2.171725285458559e-06,
      "loss": 0.3657,
      "step": 378
    },
    {
      "epoch": 2.4217252396166136,
      "grad_norm": 0.3893534985584009,
      "learning_rate": 2.1255108721175066e-06,
      "loss": 0.3439,
      "step": 379
    },
    {
      "epoch": 2.428115015974441,
      "grad_norm": 0.443399810426828,
      "learning_rate": 2.079734944972717e-06,
      "loss": 0.3216,
      "step": 380
    },
    {
      "epoch": 2.4345047923322682,
      "grad_norm": 0.39874922783339656,
      "learning_rate": 2.0344000530291875e-06,
      "loss": 0.3434,
      "step": 381
    },
    {
      "epoch": 2.440894568690096,
      "grad_norm": 0.3771297738742276,
      "learning_rate": 1.9895087207331422e-06,
      "loss": 0.3409,
      "step": 382
    },
    {
      "epoch": 2.4472843450479234,
      "grad_norm": 0.40027082321568275,
      "learning_rate": 1.945063447831452e-06,
      "loss": 0.3265,
      "step": 383
    },
    {
      "epoch": 2.4536741214057507,
      "grad_norm": 0.3737581626959035,
      "learning_rate": 1.9010667092324342e-06,
      "loss": 0.34,
      "step": 384
    },
    {
      "epoch": 2.460063897763578,
      "grad_norm": 0.3841031529537794,
      "learning_rate": 1.8575209548680472e-06,
      "loss": 0.3503,
      "step": 385
    },
    {
      "epoch": 2.466453674121406,
      "grad_norm": 0.4038855857442773,
      "learning_rate": 1.814428609557458e-06,
      "loss": 0.3201,
      "step": 386
    },
    {
      "epoch": 2.472843450479233,
      "grad_norm": 0.40823279792933154,
      "learning_rate": 1.7717920728720284e-06,
      "loss": 0.3219,
      "step": 387
    },
    {
      "epoch": 2.479233226837061,
      "grad_norm": 0.45571229187170187,
      "learning_rate": 1.7296137190016916e-06,
      "loss": 0.3646,
      "step": 388
    },
    {
      "epoch": 2.4856230031948883,
      "grad_norm": 0.3992759518176683,
      "learning_rate": 1.6878958966227366e-06,
      "loss": 0.3514,
      "step": 389
    },
    {
      "epoch": 2.4920127795527156,
      "grad_norm": 0.44997582744977993,
      "learning_rate": 1.646640928767047e-06,
      "loss": 0.3676,
      "step": 390
    },
    {
      "epoch": 2.498402555910543,
      "grad_norm": 0.4036812441443117,
      "learning_rate": 1.6058511126927178e-06,
      "loss": 0.3572,
      "step": 391
    },
    {
      "epoch": 2.5047923322683707,
      "grad_norm": 0.38861999088805715,
      "learning_rate": 1.5655287197561497e-06,
      "loss": 0.3176,
      "step": 392
    },
    {
      "epoch": 2.511182108626198,
      "grad_norm": 0.3958987299112533,
      "learning_rate": 1.5256759952855737e-06,
      "loss": 0.3411,
      "step": 393
    },
    {
      "epoch": 2.5175718849840254,
      "grad_norm": 0.3958900572864119,
      "learning_rate": 1.4862951584560037e-06,
      "loss": 0.3391,
      "step": 394
    },
    {
      "epoch": 2.523961661341853,
      "grad_norm": 0.3918198479937304,
      "learning_rate": 1.447388402165686e-06,
      "loss": 0.3178,
      "step": 395
    },
    {
      "epoch": 2.5303514376996805,
      "grad_norm": 0.40756583772176336,
      "learning_rate": 1.4089578929139635e-06,
      "loss": 0.3264,
      "step": 396
    },
    {
      "epoch": 2.536741214057508,
      "grad_norm": 0.40960058637131896,
      "learning_rate": 1.371005770680659e-06,
      "loss": 0.3385,
      "step": 397
    },
    {
      "epoch": 2.543130990415335,
      "grad_norm": 0.40903897126869604,
      "learning_rate": 1.3335341488068997e-06,
      "loss": 0.3805,
      "step": 398
    },
    {
      "epoch": 2.549520766773163,
      "grad_norm": 0.39992597114439177,
      "learning_rate": 1.2965451138774343e-06,
      "loss": 0.3196,
      "step": 399
    },
    {
      "epoch": 2.5559105431309903,
      "grad_norm": 0.3894861786588525,
      "learning_rate": 1.2600407256044544e-06,
      "loss": 0.354,
      "step": 400
    },
    {
      "epoch": 2.562300319488818,
      "grad_norm": 0.3773260960595135,
      "learning_rate": 1.2240230167128863e-06,
      "loss": 0.3408,
      "step": 401
    },
    {
      "epoch": 2.5686900958466454,
      "grad_norm": 0.3916021706010166,
      "learning_rate": 1.188493992827211e-06,
      "loss": 0.3141,
      "step": 402
    },
    {
      "epoch": 2.5750798722044728,
      "grad_norm": 0.3906836314398566,
      "learning_rate": 1.1534556323597824e-06,
      "loss": 0.2924,
      "step": 403
    },
    {
      "epoch": 2.5814696485623,
      "grad_norm": 0.4038749879309955,
      "learning_rate": 1.1189098864006488e-06,
      "loss": 0.3365,
      "step": 404
    },
    {
      "epoch": 2.587859424920128,
      "grad_norm": 0.40279549296487077,
      "learning_rate": 1.084858678608922e-06,
      "loss": 0.3239,
      "step": 405
    },
    {
      "epoch": 2.594249201277955,
      "grad_norm": 0.36365947820301786,
      "learning_rate": 1.0513039051056507e-06,
      "loss": 0.3326,
      "step": 406
    },
    {
      "epoch": 2.600638977635783,
      "grad_norm": 0.3696286967249599,
      "learning_rate": 1.0182474343682346e-06,
      "loss": 0.3292,
      "step": 407
    },
    {
      "epoch": 2.6070287539936103,
      "grad_norm": 0.4175231953214467,
      "learning_rate": 9.85691107126392e-07,
      "loss": 0.3744,
      "step": 408
    },
    {
      "epoch": 2.6134185303514377,
      "grad_norm": 0.39195370451842787,
      "learning_rate": 9.536367362596422e-07,
      "loss": 0.3695,
      "step": 409
    },
    {
      "epoch": 2.619808306709265,
      "grad_norm": 0.37605332595306007,
      "learning_rate": 9.220861066963715e-07,
      "loss": 0.3418,
      "step": 410
    },
    {
      "epoch": 2.626198083067093,
      "grad_norm": 0.3995977702745196,
      "learning_rate": 8.910409753144344e-07,
      "loss": 0.3549,
      "step": 411
    },
    {
      "epoch": 2.63258785942492,
      "grad_norm": 0.3834229012458761,
      "learning_rate": 8.605030708433149e-07,
      "loss": 0.3445,
      "step": 412
    },
    {
      "epoch": 2.6389776357827475,
      "grad_norm": 0.37644074285723944,
      "learning_rate": 8.304740937678835e-07,
      "loss": 0.3668,
      "step": 413
    },
    {
      "epoch": 2.6453674121405752,
      "grad_norm": 0.38149522660404983,
      "learning_rate": 8.009557162336823e-07,
      "loss": 0.3286,
      "step": 414
    },
    {
      "epoch": 2.6517571884984026,
      "grad_norm": 0.3965348662763524,
      "learning_rate": 7.719495819538325e-07,
      "loss": 0.3562,
      "step": 415
    },
    {
      "epoch": 2.65814696485623,
      "grad_norm": 0.3740820985387306,
      "learning_rate": 7.434573061174966e-07,
      "loss": 0.3171,
      "step": 416
    },
    {
      "epoch": 2.6645367412140573,
      "grad_norm": 0.37288133480679564,
      "learning_rate": 7.154804752999344e-07,
      "loss": 0.3263,
      "step": 417
    },
    {
      "epoch": 2.670926517571885,
      "grad_norm": 0.37023486110236786,
      "learning_rate": 6.880206473741646e-07,
      "loss": 0.368,
      "step": 418
    },
    {
      "epoch": 2.6773162939297124,
      "grad_norm": 0.3924590139873951,
      "learning_rate": 6.610793514242075e-07,
      "loss": 0.3348,
      "step": 419
    },
    {
      "epoch": 2.68370607028754,
      "grad_norm": 0.4026773365921225,
      "learning_rate": 6.346580876599395e-07,
      "loss": 0.3405,
      "step": 420
    },
    {
      "epoch": 2.6900958466453675,
      "grad_norm": 0.3855444929570631,
      "learning_rate": 6.08758327333564e-07,
      "loss": 0.3588,
      "step": 421
    },
    {
      "epoch": 2.696485623003195,
      "grad_norm": 0.4018291966488895,
      "learning_rate": 5.833815126576714e-07,
      "loss": 0.3259,
      "step": 422
    },
    {
      "epoch": 2.702875399361022,
      "grad_norm": 0.36389835567175366,
      "learning_rate": 5.585290567249424e-07,
      "loss": 0.3514,
      "step": 423
    },
    {
      "epoch": 2.70926517571885,
      "grad_norm": 0.40458156951380986,
      "learning_rate": 5.342023434294552e-07,
      "loss": 0.337,
      "step": 424
    },
    {
      "epoch": 2.7156549520766773,
      "grad_norm": 0.36807317654930927,
      "learning_rate": 5.104027273896239e-07,
      "loss": 0.3148,
      "step": 425
    },
    {
      "epoch": 2.722044728434505,
      "grad_norm": 0.3689483940462552,
      "learning_rate": 4.871315338727711e-07,
      "loss": 0.3181,
      "step": 426
    },
    {
      "epoch": 2.7284345047923324,
      "grad_norm": 0.3833457229816993,
      "learning_rate": 4.6439005872132457e-07,
      "loss": 0.3368,
      "step": 427
    },
    {
      "epoch": 2.7348242811501597,
      "grad_norm": 0.36604943775735366,
      "learning_rate": 4.421795682806662e-07,
      "loss": 0.3364,
      "step": 428
    },
    {
      "epoch": 2.741214057507987,
      "grad_norm": 0.3773811652537217,
      "learning_rate": 4.2050129932861394e-07,
      "loss": 0.3229,
      "step": 429
    },
    {
      "epoch": 2.747603833865815,
      "grad_norm": 0.3796116399398149,
      "learning_rate": 3.9935645900654906e-07,
      "loss": 0.3633,
      "step": 430
    },
    {
      "epoch": 2.753993610223642,
      "grad_norm": 0.37803460451421467,
      "learning_rate": 3.7874622475220336e-07,
      "loss": 0.3524,
      "step": 431
    },
    {
      "epoch": 2.7603833865814695,
      "grad_norm": 0.36675174735385174,
      "learning_rate": 3.58671744234087e-07,
      "loss": 0.365,
      "step": 432
    },
    {
      "epoch": 2.7667731629392973,
      "grad_norm": 0.3898919999844523,
      "learning_rate": 3.3913413528758877e-07,
      "loss": 0.3323,
      "step": 433
    },
    {
      "epoch": 2.7731629392971247,
      "grad_norm": 0.3936657666166714,
      "learning_rate": 3.2013448585272333e-07,
      "loss": 0.3337,
      "step": 434
    },
    {
      "epoch": 2.779552715654952,
      "grad_norm": 0.38046637386341803,
      "learning_rate": 3.016738539135566e-07,
      "loss": 0.344,
      "step": 435
    },
    {
      "epoch": 2.7859424920127793,
      "grad_norm": 0.3770525251793597,
      "learning_rate": 2.837532674392862e-07,
      "loss": 0.374,
      "step": 436
    },
    {
      "epoch": 2.792332268370607,
      "grad_norm": 0.3916319385244186,
      "learning_rate": 2.6637372432700483e-07,
      "loss": 0.3492,
      "step": 437
    },
    {
      "epoch": 2.7987220447284344,
      "grad_norm": 0.4083221537063404,
      "learning_rate": 2.4953619234612816e-07,
      "loss": 0.3056,
      "step": 438
    },
    {
      "epoch": 2.8051118210862622,
      "grad_norm": 0.36900391336674243,
      "learning_rate": 2.332416090845102e-07,
      "loss": 0.3071,
      "step": 439
    },
    {
      "epoch": 2.8115015974440896,
      "grad_norm": 0.4181996987699531,
      "learning_rate": 2.1749088189622848e-07,
      "loss": 0.3351,
      "step": 440
    },
    {
      "epoch": 2.817891373801917,
      "grad_norm": 0.3547619175700847,
      "learning_rate": 2.0228488785106636e-07,
      "loss": 0.3406,
      "step": 441
    },
    {
      "epoch": 2.8242811501597442,
      "grad_norm": 0.3977218103469981,
      "learning_rate": 1.8762447368566582e-07,
      "loss": 0.3505,
      "step": 442
    },
    {
      "epoch": 2.830670926517572,
      "grad_norm": 0.39003543295538634,
      "learning_rate": 1.7351045575638047e-07,
      "loss": 0.3753,
      "step": 443
    },
    {
      "epoch": 2.8370607028753994,
      "grad_norm": 0.37845868272626343,
      "learning_rate": 1.5994361999382202e-07,
      "loss": 0.3278,
      "step": 444
    },
    {
      "epoch": 2.8434504792332267,
      "grad_norm": 0.39842394374774287,
      "learning_rate": 1.4692472185908635e-07,
      "loss": 0.3682,
      "step": 445
    },
    {
      "epoch": 2.8498402555910545,
      "grad_norm": 0.3457049068963664,
      "learning_rate": 1.344544863016961e-07,
      "loss": 0.3438,
      "step": 446
    },
    {
      "epoch": 2.856230031948882,
      "grad_norm": 0.40463187861502226,
      "learning_rate": 1.225336077192274e-07,
      "loss": 0.3356,
      "step": 447
    },
    {
      "epoch": 2.862619808306709,
      "grad_norm": 0.5473639905587789,
      "learning_rate": 1.1116274991864073e-07,
      "loss": 0.3612,
      "step": 448
    },
    {
      "epoch": 2.8690095846645365,
      "grad_norm": 0.36738048359514414,
      "learning_rate": 1.003425460793217e-07,
      "loss": 0.3618,
      "step": 449
    },
    {
      "epoch": 2.8753993610223643,
      "grad_norm": 0.39397443170606444,
      "learning_rate": 9.00735987178214e-08,
      "loss": 0.3325,
      "step": 450
    },
    {
      "epoch": 2.8817891373801916,
      "grad_norm": 0.3641877860232395,
      "learning_rate": 8.035647965430215e-08,
      "loss": 0.3705,
      "step": 451
    },
    {
      "epoch": 2.8881789137380194,
      "grad_norm": 0.3766602850517281,
      "learning_rate": 7.119172998070412e-08,
      "loss": 0.3602,
      "step": 452
    },
    {
      "epoch": 2.8945686900958467,
      "grad_norm": 0.3504672529434601,
      "learning_rate": 6.25798600306049e-08,
      "loss": 0.345,
      "step": 453
    },
    {
      "epoch": 2.900958466453674,
      "grad_norm": 0.3689087024207003,
      "learning_rate": 5.4521349350808996e-08,
      "loss": 0.3711,
      "step": 454
    },
    {
      "epoch": 2.9073482428115014,
      "grad_norm": 0.3816320750179523,
      "learning_rate": 4.701664667464245e-08,
      "loss": 0.3646,
      "step": 455
    },
    {
      "epoch": 2.913738019169329,
      "grad_norm": 0.3679437431040902,
      "learning_rate": 4.006616989696621e-08,
      "loss": 0.3615,
      "step": 456
    },
    {
      "epoch": 2.9201277955271565,
      "grad_norm": 0.38571927082544183,
      "learning_rate": 3.367030605090249e-08,
      "loss": 0.3222,
      "step": 457
    },
    {
      "epoch": 2.9265175718849843,
      "grad_norm": 0.3971237004311223,
      "learning_rate": 2.7829411286287577e-08,
      "loss": 0.3485,
      "step": 458
    },
    {
      "epoch": 2.9329073482428116,
      "grad_norm": 0.3630064489830963,
      "learning_rate": 2.254381084983659e-08,
      "loss": 0.3065,
      "step": 459
    },
    {
      "epoch": 2.939297124600639,
      "grad_norm": 0.3754819920105193,
      "learning_rate": 1.781379906703573e-08,
      "loss": 0.36,
      "step": 460
    },
    {
      "epoch": 2.9456869009584663,
      "grad_norm": 0.37643335933636624,
      "learning_rate": 1.3639639325748744e-08,
      "loss": 0.3725,
      "step": 461
    },
    {
      "epoch": 2.952076677316294,
      "grad_norm": 0.397354621784988,
      "learning_rate": 1.0021564061554189e-08,
      "loss": 0.3378,
      "step": 462
    },
    {
      "epoch": 2.9584664536741214,
      "grad_norm": 0.37424179888551773,
      "learning_rate": 6.959774744796921e-09,
      "loss": 0.3399,
      "step": 463
    },
    {
      "epoch": 2.9648562300319488,
      "grad_norm": 0.37190360281608004,
      "learning_rate": 4.454441869377047e-09,
      "loss": 0.3219,
      "step": 464
    },
    {
      "epoch": 2.9712460063897765,
      "grad_norm": 0.3657736795626134,
      "learning_rate": 2.5057049432519744e-09,
      "loss": 0.3705,
      "step": 465
    },
    {
      "epoch": 2.977635782747604,
      "grad_norm": 0.35303669900268125,
      "learning_rate": 1.1136724806637411e-09,
      "loss": 0.3318,
      "step": 466
    },
    {
      "epoch": 2.984025559105431,
      "grad_norm": 0.36854160936154184,
      "learning_rate": 2.784219961060597e-10,
      "loss": 0.3518,
      "step": 467
    },
    {
      "epoch": 2.9904153354632586,
      "grad_norm": 0.3989983000303115,
      "learning_rate": 0.0,
      "loss": 0.3546,
      "step": 468
    },
    {
      "epoch": 2.9904153354632586,
      "step": 468,
      "total_flos": 3.310734730054861e+17,
      "train_loss": 0.4434524042229367,
      "train_runtime": 10205.5973,
      "train_samples_per_second": 2.939,
      "train_steps_per_second": 0.046
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 468,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.310734730054861e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}