{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.6026200873362445,
  "eval_steps": 500,
  "global_step": 276,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.002183406113537118,
      "grad_norm": 1.2640951871871948,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 1.8106,
      "step": 1
    },
    {
      "epoch": 0.004366812227074236,
      "grad_norm": 0.9510509967803955,
      "learning_rate": 4.800000000000001e-06,
      "loss": 1.7795,
      "step": 2
    },
    {
      "epoch": 0.006550218340611353,
      "grad_norm": 0.9909348487854004,
      "learning_rate": 7.2e-06,
      "loss": 1.7826,
      "step": 3
    },
    {
      "epoch": 0.008733624454148471,
      "grad_norm": 0.45971497893333435,
      "learning_rate": 9.600000000000001e-06,
      "loss": 1.8758,
      "step": 4
    },
    {
      "epoch": 0.010917030567685589,
      "grad_norm": 0.42884814739227295,
      "learning_rate": 1.2e-05,
      "loss": 1.7021,
      "step": 5
    },
    {
      "epoch": 0.013100436681222707,
      "grad_norm": 0.5919619798660278,
      "learning_rate": 1.44e-05,
      "loss": 1.8473,
      "step": 6
    },
    {
      "epoch": 0.015283842794759825,
      "grad_norm": 0.6785497665405273,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 1.7757,
      "step": 7
    },
    {
      "epoch": 0.017467248908296942,
      "grad_norm": 4.099381446838379,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 1.7825,
      "step": 8
    },
    {
      "epoch": 0.019650655021834062,
      "grad_norm": 2.532731294631958,
      "learning_rate": 2.16e-05,
      "loss": 1.7632,
      "step": 9
    },
    {
      "epoch": 0.021834061135371178,
      "grad_norm": 1.6794261932373047,
      "learning_rate": 2.4e-05,
      "loss": 1.5177,
      "step": 10
    },
    {
      "epoch": 0.024017467248908297,
      "grad_norm": 106.32669067382812,
      "learning_rate": 2.64e-05,
      "loss": 1.9722,
      "step": 11
    },
    {
      "epoch": 0.026200873362445413,
      "grad_norm": 86.48753356933594,
      "learning_rate": 2.88e-05,
      "loss": 2.0703,
      "step": 12
    },
    {
      "epoch": 0.028384279475982533,
      "grad_norm": 47.49798583984375,
      "learning_rate": 3.12e-05,
      "loss": 1.9955,
      "step": 13
    },
    {
      "epoch": 0.03056768558951965,
      "grad_norm": 22.287212371826172,
      "learning_rate": 3.3600000000000004e-05,
      "loss": 1.9062,
      "step": 14
    },
    {
      "epoch": 0.03275109170305677,
      "grad_norm": 4.17438268661499,
      "learning_rate": 3.6e-05,
      "loss": 1.9168,
      "step": 15
    },
    {
      "epoch": 0.034934497816593885,
      "grad_norm": 1.0019221305847168,
      "learning_rate": 3.8400000000000005e-05,
      "loss": 1.689,
      "step": 16
    },
    {
      "epoch": 0.03711790393013101,
      "grad_norm": 1.1589058637619019,
      "learning_rate": 4.08e-05,
      "loss": 1.8809,
      "step": 17
    },
    {
      "epoch": 0.039301310043668124,
      "grad_norm": 2.0959179401397705,
      "learning_rate": 4.32e-05,
      "loss": 1.6561,
      "step": 18
    },
    {
      "epoch": 0.04148471615720524,
      "grad_norm": 0.6226203441619873,
      "learning_rate": 4.5600000000000004e-05,
      "loss": 1.8873,
      "step": 19
    },
    {
      "epoch": 0.043668122270742356,
      "grad_norm": 1.2480379343032837,
      "learning_rate": 4.8e-05,
      "loss": 1.5221,
      "step": 20
    },
    {
      "epoch": 0.04585152838427948,
      "grad_norm": 0.5519077777862549,
      "learning_rate": 5.04e-05,
      "loss": 1.7395,
      "step": 21
    },
    {
      "epoch": 0.048034934497816595,
      "grad_norm": 1.2694586515426636,
      "learning_rate": 5.28e-05,
      "loss": 1.7318,
      "step": 22
    },
    {
      "epoch": 0.05021834061135371,
      "grad_norm": 1.5784529447555542,
      "learning_rate": 5.520000000000001e-05,
      "loss": 1.5382,
      "step": 23
    },
    {
      "epoch": 0.05240174672489083,
      "grad_norm": 1.124130129814148,
      "learning_rate": 5.76e-05,
      "loss": 1.7795,
      "step": 24
    },
    {
      "epoch": 0.05458515283842795,
      "grad_norm": 0.5427664518356323,
      "learning_rate": 6e-05,
      "loss": 1.8873,
      "step": 25
    },
    {
      "epoch": 0.056768558951965066,
      "grad_norm": 0.4415256977081299,
      "learning_rate": 5.999981351876082e-05,
      "loss": 1.74,
      "step": 26
    },
    {
      "epoch": 0.05895196506550218,
      "grad_norm": 3.6468303203582764,
      "learning_rate": 5.9999254077361634e-05,
      "loss": 1.7995,
      "step": 27
    },
    {
      "epoch": 0.0611353711790393,
      "grad_norm": 1.0301928520202637,
      "learning_rate": 5.9998321682757455e-05,
      "loss": 1.7174,
      "step": 28
    },
    {
      "epoch": 0.06331877729257641,
      "grad_norm": 59.005828857421875,
      "learning_rate": 5.9997016346539894e-05,
      "loss": 2.0753,
      "step": 29
    },
    {
      "epoch": 0.06550218340611354,
      "grad_norm": 1.1177105903625488,
      "learning_rate": 5.9995338084937e-05,
      "loss": 1.6031,
      "step": 30
    },
    {
      "epoch": 0.06768558951965066,
      "grad_norm": 0.49058693647384644,
      "learning_rate": 5.9993286918813056e-05,
      "loss": 1.8835,
      "step": 31
    },
    {
      "epoch": 0.06986899563318777,
      "grad_norm": 3.5991365909576416,
      "learning_rate": 5.999086287366833e-05,
      "loss": 1.7765,
      "step": 32
    },
    {
      "epoch": 0.07205240174672489,
      "grad_norm": 0.6412646770477295,
      "learning_rate": 5.998806597963876e-05,
      "loss": 1.6924,
      "step": 33
    },
    {
      "epoch": 0.07423580786026202,
      "grad_norm": 0.7408072352409363,
      "learning_rate": 5.998489627149556e-05,
      "loss": 1.7422,
      "step": 34
    },
    {
      "epoch": 0.07641921397379912,
      "grad_norm": 0.7268134951591492,
      "learning_rate": 5.998135378864479e-05,
      "loss": 1.8242,
      "step": 35
    },
    {
      "epoch": 0.07860262008733625,
      "grad_norm": 0.5574382543563843,
      "learning_rate": 5.9977438575126904e-05,
      "loss": 1.5462,
      "step": 36
    },
    {
      "epoch": 0.08078602620087336,
      "grad_norm": 0.4807906448841095,
      "learning_rate": 5.9973150679616154e-05,
      "loss": 1.7738,
      "step": 37
    },
    {
      "epoch": 0.08296943231441048,
      "grad_norm": 0.6777800917625427,
      "learning_rate": 5.9968490155420015e-05,
      "loss": 1.5089,
      "step": 38
    },
    {
      "epoch": 0.0851528384279476,
      "grad_norm": 0.517384946346283,
      "learning_rate": 5.99634570604785e-05,
      "loss": 1.5988,
      "step": 39
    },
    {
      "epoch": 0.08733624454148471,
      "grad_norm": 0.6687184572219849,
      "learning_rate": 5.995805145736348e-05,
      "loss": 1.6098,
      "step": 40
    },
    {
      "epoch": 0.08951965065502183,
      "grad_norm": 0.695644736289978,
      "learning_rate": 5.9952273413277836e-05,
      "loss": 1.8774,
      "step": 41
    },
    {
      "epoch": 0.09170305676855896,
      "grad_norm": 0.49494823813438416,
      "learning_rate": 5.9946123000054706e-05,
      "loss": 1.8832,
      "step": 42
    },
    {
      "epoch": 0.09388646288209607,
      "grad_norm": 0.6464128494262695,
      "learning_rate": 5.993960029415653e-05,
      "loss": 1.6651,
      "step": 43
    },
    {
      "epoch": 0.09606986899563319,
      "grad_norm": 0.5558153986930847,
      "learning_rate": 5.993270537667413e-05,
      "loss": 1.4509,
      "step": 44
    },
    {
      "epoch": 0.0982532751091703,
      "grad_norm": 0.7220055460929871,
      "learning_rate": 5.992543833332568e-05,
      "loss": 1.713,
      "step": 45
    },
    {
      "epoch": 0.10043668122270742,
      "grad_norm": 0.5923091769218445,
      "learning_rate": 5.991779925445568e-05,
      "loss": 1.852,
      "step": 46
    },
    {
      "epoch": 0.10262008733624454,
      "grad_norm": 0.9899210929870605,
      "learning_rate": 5.990978823503376e-05,
      "loss": 1.9421,
      "step": 47
    },
    {
      "epoch": 0.10480349344978165,
      "grad_norm": 0.49990367889404297,
      "learning_rate": 5.990140537465361e-05,
      "loss": 1.6797,
      "step": 48
    },
    {
      "epoch": 0.10698689956331878,
      "grad_norm": 0.7551549673080444,
      "learning_rate": 5.9892650777531624e-05,
      "loss": 1.5336,
      "step": 49
    },
    {
      "epoch": 0.1091703056768559,
      "grad_norm": 2.107231378555298,
      "learning_rate": 5.9883524552505676e-05,
      "loss": 1.8295,
      "step": 50
    },
    {
      "epoch": 0.11135371179039301,
      "grad_norm": 0.5479142069816589,
      "learning_rate": 5.987402681303376e-05,
      "loss": 1.8246,
      "step": 51
    },
    {
      "epoch": 0.11353711790393013,
      "grad_norm": 0.565462589263916,
      "learning_rate": 5.986415767719254e-05,
      "loss": 1.8011,
      "step": 52
    },
    {
      "epoch": 0.11572052401746726,
      "grad_norm": 0.6389402747154236,
      "learning_rate": 5.985391726767594e-05,
      "loss": 1.6677,
      "step": 53
    },
    {
      "epoch": 0.11790393013100436,
      "grad_norm": 0.9275972247123718,
      "learning_rate": 5.9843305711793584e-05,
      "loss": 1.758,
      "step": 54
    },
    {
      "epoch": 0.12008733624454149,
      "grad_norm": 0.5152640342712402,
      "learning_rate": 5.98323231414692e-05,
      "loss": 1.5863,
      "step": 55
    },
    {
      "epoch": 0.1222707423580786,
      "grad_norm": 0.6816053986549377,
      "learning_rate": 5.982096969323902e-05,
      "loss": 1.6378,
      "step": 56
    },
    {
      "epoch": 0.12445414847161572,
      "grad_norm": 0.7598839402198792,
      "learning_rate": 5.980924550825004e-05,
      "loss": 1.664,
      "step": 57
    },
    {
      "epoch": 0.12663755458515283,
      "grad_norm": 0.5732179880142212,
      "learning_rate": 5.979715073225829e-05,
      "loss": 1.7683,
      "step": 58
    },
    {
      "epoch": 0.12882096069868995,
      "grad_norm": 1.1897765398025513,
      "learning_rate": 5.978468551562704e-05,
      "loss": 1.6833,
      "step": 59
    },
    {
      "epoch": 0.13100436681222707,
      "grad_norm": 0.6417302489280701,
      "learning_rate": 5.977185001332489e-05,
      "loss": 1.7572,
      "step": 60
    },
    {
      "epoch": 0.1331877729257642,
      "grad_norm": 0.6772924661636353,
      "learning_rate": 5.9758644384923865e-05,
      "loss": 1.7905,
      "step": 61
    },
    {
      "epoch": 0.13537117903930132,
      "grad_norm": 0.7241927981376648,
      "learning_rate": 5.974506879459742e-05,
      "loss": 1.7925,
      "step": 62
    },
    {
      "epoch": 0.13755458515283842,
      "grad_norm": 0.6865760684013367,
      "learning_rate": 5.9731123411118416e-05,
      "loss": 1.7268,
      "step": 63
    },
    {
      "epoch": 0.13973799126637554,
      "grad_norm": 0.5667518377304077,
      "learning_rate": 5.971680840785702e-05,
      "loss": 1.5103,
      "step": 64
    },
    {
      "epoch": 0.14192139737991266,
      "grad_norm": 0.727475643157959,
      "learning_rate": 5.970212396277854e-05,
      "loss": 1.6056,
      "step": 65
    },
    {
      "epoch": 0.14410480349344978,
      "grad_norm": 0.5900119543075562,
      "learning_rate": 5.9687070258441195e-05,
      "loss": 1.6795,
      "step": 66
    },
    {
      "epoch": 0.1462882096069869,
      "grad_norm": 0.5586450695991516,
      "learning_rate": 5.967164748199389e-05,
      "loss": 1.4165,
      "step": 67
    },
    {
      "epoch": 0.14847161572052403,
      "grad_norm": 0.7147237658500671,
      "learning_rate": 5.9655855825173855e-05,
      "loss": 1.7175,
      "step": 68
    },
    {
      "epoch": 0.15065502183406113,
      "grad_norm": 0.5687724947929382,
      "learning_rate": 5.963969548430428e-05,
      "loss": 1.7112,
      "step": 69
    },
    {
      "epoch": 0.15283842794759825,
      "grad_norm": 0.9042346477508545,
      "learning_rate": 5.962316666029183e-05,
      "loss": 1.9019,
      "step": 70
    },
    {
      "epoch": 0.15502183406113537,
      "grad_norm": 0.5679914355278015,
      "learning_rate": 5.960626955862424e-05,
      "loss": 1.6929,
      "step": 71
    },
    {
      "epoch": 0.1572052401746725,
      "grad_norm": 0.8124454617500305,
      "learning_rate": 5.958900438936767e-05,
      "loss": 1.9199,
      "step": 72
    },
    {
      "epoch": 0.15938864628820962,
      "grad_norm": 0.4910396635532379,
      "learning_rate": 5.9571371367164115e-05,
      "loss": 1.6558,
      "step": 73
    },
    {
      "epoch": 0.1615720524017467,
      "grad_norm": 0.5993661880493164,
      "learning_rate": 5.955337071122878e-05,
      "loss": 1.5878,
      "step": 74
    },
    {
      "epoch": 0.16375545851528384,
      "grad_norm": 2.5788443088531494,
      "learning_rate": 5.953500264534729e-05,
      "loss": 1.7942,
      "step": 75
    },
    {
      "epoch": 0.16593886462882096,
      "grad_norm": 0.7565484046936035,
      "learning_rate": 5.951626739787298e-05,
      "loss": 1.609,
      "step": 76
    },
    {
      "epoch": 0.16812227074235808,
      "grad_norm": 0.5368123054504395,
      "learning_rate": 5.9497165201723974e-05,
      "loss": 1.7037,
      "step": 77
    },
    {
      "epoch": 0.1703056768558952,
      "grad_norm": 0.6212489008903503,
      "learning_rate": 5.947769629438036e-05,
      "loss": 1.8458,
      "step": 78
    },
    {
      "epoch": 0.17248908296943233,
      "grad_norm": 0.6993361115455627,
      "learning_rate": 5.94578609178812e-05,
      "loss": 1.9143,
      "step": 79
    },
    {
      "epoch": 0.17467248908296942,
      "grad_norm": 0.4691102206707001,
      "learning_rate": 5.9437659318821544e-05,
      "loss": 1.5512,
      "step": 80
    },
    {
      "epoch": 0.17685589519650655,
      "grad_norm": 0.6164081692695618,
      "learning_rate": 5.9417091748349326e-05,
      "loss": 1.748,
      "step": 81
    },
    {
      "epoch": 0.17903930131004367,
      "grad_norm": 2.626962900161743,
      "learning_rate": 5.939615846216229e-05,
      "loss": 1.7222,
      "step": 82
    },
    {
      "epoch": 0.1812227074235808,
      "grad_norm": 0.7723209261894226,
      "learning_rate": 5.937485972050478e-05,
      "loss": 1.6113,
      "step": 83
    },
    {
      "epoch": 0.18340611353711792,
      "grad_norm": 0.5663449764251709,
      "learning_rate": 5.9353195788164495e-05,
      "loss": 1.9,
      "step": 84
    },
    {
      "epoch": 0.185589519650655,
      "grad_norm": 0.6012480854988098,
      "learning_rate": 5.933116693446926e-05,
      "loss": 1.7153,
      "step": 85
    },
    {
      "epoch": 0.18777292576419213,
      "grad_norm": 0.7445963025093079,
      "learning_rate": 5.930877343328357e-05,
      "loss": 1.8219,
      "step": 86
    },
    {
      "epoch": 0.18995633187772926,
      "grad_norm": 0.5544320344924927,
      "learning_rate": 5.9286015563005326e-05,
      "loss": 1.603,
      "step": 87
    },
    {
      "epoch": 0.19213973799126638,
      "grad_norm": 0.8514794111251831,
      "learning_rate": 5.926289360656222e-05,
      "loss": 1.5782,
      "step": 88
    },
    {
      "epoch": 0.1943231441048035,
      "grad_norm": 0.638144850730896,
      "learning_rate": 5.923940785140832e-05,
      "loss": 1.5709,
      "step": 89
    },
    {
      "epoch": 0.1965065502183406,
      "grad_norm": 0.6473650336265564,
      "learning_rate": 5.9215558589520485e-05,
      "loss": 1.6066,
      "step": 90
    },
    {
      "epoch": 0.19868995633187772,
      "grad_norm": 0.7627925872802734,
      "learning_rate": 5.919134611739472e-05,
      "loss": 1.6075,
      "step": 91
    },
    {
      "epoch": 0.20087336244541484,
      "grad_norm": 0.623173713684082,
      "learning_rate": 5.9166770736042455e-05,
      "loss": 1.8107,
      "step": 92
    },
    {
      "epoch": 0.20305676855895197,
      "grad_norm": 0.5980973243713379,
      "learning_rate": 5.914183275098688e-05,
      "loss": 1.5835,
      "step": 93
    },
    {
      "epoch": 0.2052401746724891,
      "grad_norm": 0.5484164953231812,
      "learning_rate": 5.911653247225908e-05,
      "loss": 1.5582,
      "step": 94
    },
    {
      "epoch": 0.2074235807860262,
      "grad_norm": 0.6916869282722473,
      "learning_rate": 5.9090870214394215e-05,
      "loss": 1.8005,
      "step": 95
    },
    {
      "epoch": 0.2096069868995633,
      "grad_norm": 5.73091459274292,
      "learning_rate": 5.906484629642758e-05,
      "loss": 1.8641,
      "step": 96
    },
    {
      "epoch": 0.21179039301310043,
      "grad_norm": 14.361653327941895,
      "learning_rate": 5.903846104189069e-05,
      "loss": 1.8477,
      "step": 97
    },
    {
      "epoch": 0.21397379912663755,
      "grad_norm": 139.7395782470703,
      "learning_rate": 5.9011714778807196e-05,
      "loss": 2.3126,
      "step": 98
    },
    {
      "epoch": 0.21615720524017468,
      "grad_norm": 59.63438415527344,
      "learning_rate": 5.898460783968886e-05,
      "loss": 2.2169,
      "step": 99
    },
    {
      "epoch": 0.2183406113537118,
      "grad_norm": 13.395294189453125,
      "learning_rate": 5.895714056153138e-05,
      "loss": 1.7024,
      "step": 100
    },
    {
      "epoch": 0.2205240174672489,
      "grad_norm": 2.569697141647339,
      "learning_rate": 5.8929313285810235e-05,
      "loss": 1.8547,
      "step": 101
    },
    {
      "epoch": 0.22270742358078602,
      "grad_norm": 2.247088670730591,
      "learning_rate": 5.8901126358476414e-05,
      "loss": 1.71,
      "step": 102
    },
    {
      "epoch": 0.22489082969432314,
      "grad_norm": 7.35801362991333,
      "learning_rate": 5.887258012995212e-05,
      "loss": 1.7473,
      "step": 103
    },
    {
      "epoch": 0.22707423580786026,
      "grad_norm": 16.6738338470459,
      "learning_rate": 5.884367495512643e-05,
      "loss": 1.9232,
      "step": 104
    },
    {
      "epoch": 0.2292576419213974,
      "grad_norm": 1.1531955003738403,
      "learning_rate": 5.881441119335086e-05,
      "loss": 1.9778,
      "step": 105
    },
    {
      "epoch": 0.2314410480349345,
      "grad_norm": 1.6605702638626099,
      "learning_rate": 5.878478920843492e-05,
      "loss": 1.7249,
      "step": 106
    },
    {
      "epoch": 0.2336244541484716,
      "grad_norm": 1.4826135635375977,
      "learning_rate": 5.875480936864158e-05,
      "loss": 1.974,
      "step": 107
    },
    {
      "epoch": 0.23580786026200873,
      "grad_norm": 1.0353524684906006,
      "learning_rate": 5.872447204668266e-05,
      "loss": 1.7089,
      "step": 108
    },
    {
      "epoch": 0.23799126637554585,
      "grad_norm": 3.460448980331421,
      "learning_rate": 5.869377761971428e-05,
      "loss": 1.6767,
      "step": 109
    },
    {
      "epoch": 0.24017467248908297,
      "grad_norm": 0.7396925091743469,
      "learning_rate": 5.866272646933208e-05,
      "loss": 1.8262,
      "step": 110
    },
    {
      "epoch": 0.2423580786026201,
      "grad_norm": 0.9432264566421509,
      "learning_rate": 5.863131898156653e-05,
      "loss": 1.7891,
      "step": 111
    },
    {
      "epoch": 0.2445414847161572,
      "grad_norm": 1.2679530382156372,
      "learning_rate": 5.859955554687811e-05,
      "loss": 1.6454,
      "step": 112
    },
    {
      "epoch": 0.24672489082969432,
      "grad_norm": 0.9549881815910339,
      "learning_rate": 5.856743656015246e-05,
      "loss": 1.7474,
      "step": 113
    },
    {
      "epoch": 0.24890829694323144,
      "grad_norm": 0.991295337677002,
      "learning_rate": 5.8534962420695485e-05,
      "loss": 1.7915,
      "step": 114
    },
    {
      "epoch": 0.25109170305676853,
      "grad_norm": 1.199068546295166,
      "learning_rate": 5.850213353222836e-05,
      "loss": 1.8153,
      "step": 115
    },
    {
      "epoch": 0.25327510917030566,
      "grad_norm": 1.3128461837768555,
      "learning_rate": 5.846895030288255e-05,
      "loss": 1.7813,
      "step": 116
    },
    {
      "epoch": 0.2554585152838428,
      "grad_norm": 0.8711860775947571,
      "learning_rate": 5.84354131451947e-05,
      "loss": 1.9726,
      "step": 117
    },
    {
      "epoch": 0.2576419213973799,
      "grad_norm": 0.8892795443534851,
      "learning_rate": 5.840152247610151e-05,
      "loss": 1.6705,
      "step": 118
    },
    {
      "epoch": 0.259825327510917,
      "grad_norm": 1.0668408870697021,
      "learning_rate": 5.83672787169346e-05,
      "loss": 1.6662,
      "step": 119
    },
    {
      "epoch": 0.26200873362445415,
      "grad_norm": 0.8376436233520508,
      "learning_rate": 5.83326822934152e-05,
      "loss": 1.7861,
      "step": 120
    },
    {
      "epoch": 0.26419213973799127,
      "grad_norm": 1.3424566984176636,
      "learning_rate": 5.829773363564891e-05,
      "loss": 1.8182,
      "step": 121
    },
    {
      "epoch": 0.2663755458515284,
      "grad_norm": 1.0533416271209717,
      "learning_rate": 5.8262433178120326e-05,
      "loss": 1.9233,
      "step": 122
    },
    {
      "epoch": 0.2685589519650655,
      "grad_norm": 0.8255435824394226,
      "learning_rate": 5.822678135968765e-05,
      "loss": 1.6769,
      "step": 123
    },
    {
      "epoch": 0.27074235807860264,
      "grad_norm": 0.7774696350097656,
      "learning_rate": 5.819077862357725e-05,
      "loss": 1.6095,
      "step": 124
    },
    {
      "epoch": 0.27292576419213976,
      "grad_norm": 0.8042762875556946,
      "learning_rate": 5.81544254173781e-05,
      "loss": 1.8795,
      "step": 125
    },
    {
      "epoch": 0.27510917030567683,
      "grad_norm": 0.7848981618881226,
      "learning_rate": 5.811772219303627e-05,
      "loss": 1.7101,
      "step": 126
    },
    {
      "epoch": 0.27729257641921395,
      "grad_norm": 0.7639428377151489,
      "learning_rate": 5.8080669406849274e-05,
      "loss": 1.8064,
      "step": 127
    },
    {
      "epoch": 0.2794759825327511,
      "grad_norm": 0.9971274137496948,
      "learning_rate": 5.804326751946041e-05,
      "loss": 1.5637,
      "step": 128
    },
    {
      "epoch": 0.2816593886462882,
      "grad_norm": 1.073482632637024,
      "learning_rate": 5.8005516995853034e-05,
      "loss": 1.7532,
      "step": 129
    },
    {
      "epoch": 0.2838427947598253,
      "grad_norm": 0.8909122347831726,
      "learning_rate": 5.796741830534477e-05,
      "loss": 1.703,
      "step": 130
    },
    {
      "epoch": 0.28602620087336245,
      "grad_norm": 1.0015509128570557,
      "learning_rate": 5.79289719215817e-05,
      "loss": 1.6742,
      "step": 131
    },
    {
      "epoch": 0.28820960698689957,
      "grad_norm": 0.8277704119682312,
      "learning_rate": 5.789017832253243e-05,
      "loss": 1.8545,
      "step": 132
    },
    {
      "epoch": 0.2903930131004367,
      "grad_norm": 2.0752744674682617,
      "learning_rate": 5.785103799048218e-05,
      "loss": 1.594,
      "step": 133
    },
    {
      "epoch": 0.2925764192139738,
      "grad_norm": 0.880749523639679,
      "learning_rate": 5.781155141202681e-05,
      "loss": 1.7113,
      "step": 134
    },
    {
      "epoch": 0.29475982532751094,
      "grad_norm": 1.5237421989440918,
      "learning_rate": 5.777171907806672e-05,
      "loss": 1.8,
      "step": 135
    },
    {
      "epoch": 0.29694323144104806,
      "grad_norm": 0.7517329454421997,
      "learning_rate": 5.7731541483800764e-05,
      "loss": 1.8649,
      "step": 136
    },
    {
      "epoch": 0.29912663755458513,
      "grad_norm": 0.9960501194000244,
      "learning_rate": 5.7691019128720127e-05,
      "loss": 1.8795,
      "step": 137
    },
    {
      "epoch": 0.30131004366812225,
      "grad_norm": 2.075016498565674,
      "learning_rate": 5.7650152516602075e-05,
      "loss": 1.7784,
      "step": 138
    },
    {
      "epoch": 0.3034934497816594,
      "grad_norm": 0.8353100419044495,
      "learning_rate": 5.7608942155503694e-05,
      "loss": 1.6494,
      "step": 139
    },
    {
      "epoch": 0.3056768558951965,
      "grad_norm": 1.3495029211044312,
      "learning_rate": 5.756738855775561e-05,
      "loss": 1.6013,
      "step": 140
    },
    {
      "epoch": 0.3078602620087336,
      "grad_norm": 1.128967046737671,
      "learning_rate": 5.7525492239955586e-05,
      "loss": 1.5049,
      "step": 141
    },
    {
      "epoch": 0.31004366812227074,
      "grad_norm": 1.7347663640975952,
      "learning_rate": 5.7483253722962086e-05,
      "loss": 1.6187,
      "step": 142
    },
    {
      "epoch": 0.31222707423580787,
      "grad_norm": 0.8418498635292053,
      "learning_rate": 5.744067353188787e-05,
      "loss": 1.679,
      "step": 143
    },
    {
      "epoch": 0.314410480349345,
      "grad_norm": 0.8698752522468567,
      "learning_rate": 5.739775219609336e-05,
      "loss": 1.8017,
      "step": 144
    },
    {
      "epoch": 0.3165938864628821,
      "grad_norm": 0.7123454809188843,
      "learning_rate": 5.735449024918017e-05,
      "loss": 1.7967,
      "step": 145
    },
    {
      "epoch": 0.31877729257641924,
      "grad_norm": 0.8182825446128845,
      "learning_rate": 5.73108882289844e-05,
      "loss": 1.7494,
      "step": 146
    },
    {
      "epoch": 0.32096069868995636,
      "grad_norm": 0.7949255704879761,
      "learning_rate": 5.726694667756996e-05,
      "loss": 1.6424,
      "step": 147
    },
    {
      "epoch": 0.3231441048034934,
      "grad_norm": 0.7940742373466492,
      "learning_rate": 5.7222666141221836e-05,
      "loss": 1.5466,
      "step": 148
    },
    {
      "epoch": 0.32532751091703055,
      "grad_norm": 0.8936315178871155,
      "learning_rate": 5.7178047170439335e-05,
      "loss": 1.7015,
      "step": 149
    },
    {
      "epoch": 0.32751091703056767,
      "grad_norm": 0.7524909377098083,
      "learning_rate": 5.713309031992917e-05,
      "loss": 1.7601,
      "step": 150
    },
    {
      "epoch": 0.3296943231441048,
      "grad_norm": 0.85063636302948,
      "learning_rate": 5.7087796148598644e-05,
      "loss": 1.7546,
      "step": 151
    },
    {
      "epoch": 0.3318777292576419,
      "grad_norm": 0.8636225461959839,
      "learning_rate": 5.704216521954861e-05,
      "loss": 1.6332,
      "step": 152
    },
    {
      "epoch": 0.33406113537117904,
      "grad_norm": 1.071815013885498,
      "learning_rate": 5.699619810006655e-05,
      "loss": 1.958,
      "step": 153
    },
    {
      "epoch": 0.33624454148471616,
      "grad_norm": 0.8274393081665039,
      "learning_rate": 5.694989536161951e-05,
      "loss": 1.6706,
      "step": 154
    },
    {
      "epoch": 0.3384279475982533,
      "grad_norm": 1.9788514375686646,
      "learning_rate": 5.6903257579846945e-05,
      "loss": 1.7573,
      "step": 155
    },
    {
      "epoch": 0.3406113537117904,
      "grad_norm": 0.7105967402458191,
      "learning_rate": 5.6856285334553625e-05,
      "loss": 1.6248,
      "step": 156
    },
    {
      "epoch": 0.34279475982532753,
      "grad_norm": 2.3344361782073975,
      "learning_rate": 5.680897920970237e-05,
      "loss": 1.7104,
      "step": 157
    },
    {
      "epoch": 0.34497816593886466,
      "grad_norm": 0.7904492020606995,
      "learning_rate": 5.676133979340683e-05,
      "loss": 1.8896,
      "step": 158
    },
    {
      "epoch": 0.3471615720524017,
      "grad_norm": 0.8370620012283325,
      "learning_rate": 5.6713367677924176e-05,
      "loss": 2.0439,
      "step": 159
    },
    {
      "epoch": 0.34934497816593885,
      "grad_norm": 0.8244124054908752,
      "learning_rate": 5.666506345964771e-05,
      "loss": 1.7606,
      "step": 160
    },
    {
      "epoch": 0.35152838427947597,
      "grad_norm": 1.2729883193969727,
      "learning_rate": 5.6616427739099445e-05,
      "loss": 1.8514,
      "step": 161
    },
    {
      "epoch": 0.3537117903930131,
      "grad_norm": 2.0001423358917236,
      "learning_rate": 5.65674611209227e-05,
      "loss": 1.7505,
      "step": 162
    },
    {
      "epoch": 0.3558951965065502,
      "grad_norm": 0.9735695123672485,
      "learning_rate": 5.651816421387451e-05,
      "loss": 1.8041,
      "step": 163
    },
    {
      "epoch": 0.35807860262008734,
      "grad_norm": 1.3495986461639404,
      "learning_rate": 5.646853763081809e-05,
      "loss": 1.897,
      "step": 164
    },
    {
      "epoch": 0.36026200873362446,
      "grad_norm": 1.0789763927459717,
      "learning_rate": 5.641858198871523e-05,
      "loss": 1.7141,
      "step": 165
    },
    {
      "epoch": 0.3624454148471616,
      "grad_norm": 1.5300121307373047,
      "learning_rate": 5.636829790861859e-05,
      "loss": 1.601,
      "step": 166
    },
    {
      "epoch": 0.3646288209606987,
      "grad_norm": 1.3608547449111938,
      "learning_rate": 5.631768601566401e-05,
      "loss": 1.8016,
      "step": 167
    },
    {
      "epoch": 0.36681222707423583,
      "grad_norm": 2.203794240951538,
      "learning_rate": 5.6266746939062725e-05,
      "loss": 1.5859,
      "step": 168
    },
    {
      "epoch": 0.36899563318777295,
      "grad_norm": 0.9623550772666931,
      "learning_rate": 5.621548131209355e-05,
      "loss": 1.7289,
      "step": 169
    },
    {
      "epoch": 0.37117903930131,
      "grad_norm": 2.7508177757263184,
      "learning_rate": 5.616388977209499e-05,
      "loss": 1.8764,
      "step": 170
    },
    {
      "epoch": 0.37336244541484714,
      "grad_norm": 4.835301876068115,
      "learning_rate": 5.611197296045733e-05,
      "loss": 1.6522,
      "step": 171
    },
    {
      "epoch": 0.37554585152838427,
      "grad_norm": 0.8887910842895508,
      "learning_rate": 5.605973152261466e-05,
      "loss": 1.7881,
      "step": 172
    },
    {
      "epoch": 0.3777292576419214,
      "grad_norm": 1.3616653680801392,
      "learning_rate": 5.600716610803685e-05,
      "loss": 1.7412,
      "step": 173
    },
    {
      "epoch": 0.3799126637554585,
      "grad_norm": 0.735529363155365,
      "learning_rate": 5.595427737022149e-05,
      "loss": 1.6339,
      "step": 174
    },
    {
      "epoch": 0.38209606986899564,
      "grad_norm": 2.6323893070220947,
      "learning_rate": 5.5901065966685726e-05,
      "loss": 1.7204,
      "step": 175
    },
    {
      "epoch": 0.38427947598253276,
      "grad_norm": 1.0813663005828857,
      "learning_rate": 5.584753255895813e-05,
      "loss": 1.6998,
      "step": 176
    },
    {
      "epoch": 0.3864628820960699,
      "grad_norm": 0.8331813216209412,
      "learning_rate": 5.5793677812570434e-05,
      "loss": 2.0123,
      "step": 177
    },
    {
      "epoch": 0.388646288209607,
      "grad_norm": 1.8793230056762695,
      "learning_rate": 5.573950239704931e-05,
      "loss": 1.7991,
      "step": 178
    },
    {
      "epoch": 0.39082969432314413,
      "grad_norm": 0.7234326004981995,
      "learning_rate": 5.568500698590799e-05,
      "loss": 1.6716,
      "step": 179
    },
    {
      "epoch": 0.3930131004366812,
      "grad_norm": 1.0117108821868896,
      "learning_rate": 5.563019225663794e-05,
      "loss": 1.6616,
      "step": 180
    },
    {
      "epoch": 0.3951965065502183,
      "grad_norm": 0.8955837488174438,
      "learning_rate": 5.5575058890700383e-05,
      "loss": 1.6803,
      "step": 181
    },
    {
      "epoch": 0.39737991266375544,
      "grad_norm": 0.9817379117012024,
      "learning_rate": 5.5519607573517906e-05,
      "loss": 1.7891,
      "step": 182
    },
    {
      "epoch": 0.39956331877729256,
      "grad_norm": 1.2921302318572998,
      "learning_rate": 5.5463838994465835e-05,
      "loss": 1.8692,
      "step": 183
    },
    {
      "epoch": 0.4017467248908297,
      "grad_norm": 1.0093421936035156,
      "learning_rate": 5.5407753846863764e-05,
      "loss": 1.5303,
      "step": 184
    },
    {
      "epoch": 0.4039301310043668,
      "grad_norm": 2.0379128456115723,
      "learning_rate": 5.535135282796689e-05,
      "loss": 1.6398,
      "step": 185
    },
    {
      "epoch": 0.40611353711790393,
      "grad_norm": 1.897610068321228,
      "learning_rate": 5.5294636638957335e-05,
      "loss": 1.9405,
      "step": 186
    },
    {
      "epoch": 0.40829694323144106,
      "grad_norm": 1.1107155084609985,
      "learning_rate": 5.5237605984935435e-05,
      "loss": 1.9509,
      "step": 187
    },
    {
      "epoch": 0.4104803493449782,
      "grad_norm": 1.278809905052185,
      "learning_rate": 5.518026157491101e-05,
      "loss": 1.7298,
      "step": 188
    },
    {
      "epoch": 0.4126637554585153,
      "grad_norm": 1.1539089679718018,
      "learning_rate": 5.512260412179449e-05,
      "loss": 1.9932,
      "step": 189
    },
    {
      "epoch": 0.4148471615720524,
      "grad_norm": 0.9997171759605408,
      "learning_rate": 5.50646343423881e-05,
      "loss": 1.6493,
      "step": 190
    },
    {
      "epoch": 0.4170305676855895,
      "grad_norm": 0.9165805578231812,
      "learning_rate": 5.5006352957376916e-05,
      "loss": 1.9506,
      "step": 191
    },
    {
      "epoch": 0.4192139737991266,
      "grad_norm": 0.9080667495727539,
      "learning_rate": 5.494776069131995e-05,
      "loss": 1.6588,
      "step": 192
    },
    {
      "epoch": 0.42139737991266374,
      "grad_norm": 1.0236437320709229,
      "learning_rate": 5.488885827264108e-05,
      "loss": 1.8508,
      "step": 193
    },
    {
      "epoch": 0.42358078602620086,
      "grad_norm": 0.7350343465805054,
      "learning_rate": 5.482964643362005e-05,
      "loss": 1.7548,
      "step": 194
    },
    {
      "epoch": 0.425764192139738,
      "grad_norm": 1.1321355104446411,
      "learning_rate": 5.4770125910383335e-05,
      "loss": 1.7822,
      "step": 195
    },
    {
      "epoch": 0.4279475982532751,
      "grad_norm": 1.1031216382980347,
      "learning_rate": 5.4710297442894984e-05,
      "loss": 1.7718,
      "step": 196
    },
    {
      "epoch": 0.43013100436681223,
      "grad_norm": 1.2498975992202759,
      "learning_rate": 5.4650161774947456e-05,
      "loss": 1.6593,
      "step": 197
    },
    {
      "epoch": 0.43231441048034935,
      "grad_norm": 1.3090567588806152,
      "learning_rate": 5.4589719654152346e-05,
      "loss": 1.8223,
      "step": 198
    },
    {
      "epoch": 0.4344978165938865,
      "grad_norm": 0.9731519222259521,
      "learning_rate": 5.4528971831931095e-05,
      "loss": 1.7375,
      "step": 199
    },
    {
      "epoch": 0.4366812227074236,
      "grad_norm": 2.541288375854492,
      "learning_rate": 5.446791906350563e-05,
      "loss": 1.8305,
      "step": 200
    },
    {
      "epoch": 0.4388646288209607,
      "grad_norm": 0.7675471305847168,
      "learning_rate": 5.4406562107889024e-05,
      "loss": 1.7076,
      "step": 201
    },
    {
      "epoch": 0.4410480349344978,
      "grad_norm": 0.8400501608848572,
      "learning_rate": 5.434490172787602e-05,
      "loss": 1.7853,
      "step": 202
    },
    {
      "epoch": 0.4432314410480349,
      "grad_norm": 1.086655855178833,
      "learning_rate": 5.428293869003356e-05,
      "loss": 1.9688,
      "step": 203
    },
    {
      "epoch": 0.44541484716157204,
      "grad_norm": 1.0826491117477417,
      "learning_rate": 5.4220673764691235e-05,
      "loss": 1.7438,
      "step": 204
    },
    {
      "epoch": 0.44759825327510916,
      "grad_norm": 1.3664004802703857,
      "learning_rate": 5.415810772593176e-05,
      "loss": 1.7104,
      "step": 205
    },
    {
      "epoch": 0.4497816593886463,
      "grad_norm": 2.6102070808410645,
      "learning_rate": 5.409524135158128e-05,
      "loss": 1.6784,
      "step": 206
    },
    {
      "epoch": 0.4519650655021834,
      "grad_norm": 1.0826854705810547,
      "learning_rate": 5.403207542319976e-05,
      "loss": 1.6317,
      "step": 207
    },
    {
      "epoch": 0.45414847161572053,
      "grad_norm": 2.249889612197876,
      "learning_rate": 5.396861072607123e-05,
      "loss": 1.6128,
      "step": 208
    },
    {
      "epoch": 0.45633187772925765,
      "grad_norm": 2.29790997505188,
      "learning_rate": 5.390484804919408e-05,
      "loss": 1.7648,
      "step": 209
    },
    {
      "epoch": 0.4585152838427948,
      "grad_norm": 7.300095081329346,
      "learning_rate": 5.3840788185271146e-05,
      "loss": 1.7678,
      "step": 210
    },
    {
      "epoch": 0.4606986899563319,
      "grad_norm": 1.031797170639038,
      "learning_rate": 5.377643193069995e-05,
      "loss": 1.835,
      "step": 211
    },
    {
      "epoch": 0.462882096069869,
      "grad_norm": 1.6482349634170532,
      "learning_rate": 5.371178008556278e-05,
      "loss": 1.9561,
      "step": 212
    },
    {
      "epoch": 0.4650655021834061,
      "grad_norm": 0.8721967339515686,
      "learning_rate": 5.3646833453616704e-05,
      "loss": 1.7253,
      "step": 213
    },
    {
      "epoch": 0.4672489082969432,
      "grad_norm": 0.9159957766532898,
      "learning_rate": 5.358159284228363e-05,
      "loss": 1.7271,
      "step": 214
    },
    {
      "epoch": 0.46943231441048033,
      "grad_norm": 0.9446601271629333,
      "learning_rate": 5.35160590626402e-05,
      "loss": 1.7168,
      "step": 215
    },
    {
      "epoch": 0.47161572052401746,
      "grad_norm": 1.9629846811294556,
      "learning_rate": 5.345023292940781e-05,
      "loss": 1.646,
      "step": 216
    },
    {
      "epoch": 0.4737991266375546,
      "grad_norm": 1.46189546585083,
      "learning_rate": 5.338411526094237e-05,
      "loss": 1.7028,
      "step": 217
    },
    {
      "epoch": 0.4759825327510917,
      "grad_norm": 1.596393346786499,
      "learning_rate": 5.331770687922419e-05,
      "loss": 1.8786,
      "step": 218
    },
    {
      "epoch": 0.4781659388646288,
      "grad_norm": 1.1326192617416382,
      "learning_rate": 5.325100860984777e-05,
      "loss": 1.8706,
      "step": 219
    },
    {
      "epoch": 0.48034934497816595,
      "grad_norm": 0.9226349592208862,
      "learning_rate": 5.318402128201149e-05,
      "loss": 1.7639,
      "step": 220
    },
    {
      "epoch": 0.48253275109170307,
      "grad_norm": 1.7078310251235962,
      "learning_rate": 5.311674572850737e-05,
      "loss": 1.8673,
      "step": 221
    },
    {
      "epoch": 0.4847161572052402,
      "grad_norm": 26.21409797668457,
      "learning_rate": 5.304918278571062e-05,
      "loss": 1.6653,
      "step": 222
    },
    {
      "epoch": 0.4868995633187773,
      "grad_norm": 1.0262807607650757,
      "learning_rate": 5.298133329356934e-05,
      "loss": 1.5772,
      "step": 223
    },
    {
      "epoch": 0.4890829694323144,
      "grad_norm": 1.1077808141708374,
      "learning_rate": 5.2913198095594026e-05,
      "loss": 1.7631,
      "step": 224
    },
    {
      "epoch": 0.4912663755458515,
      "grad_norm": 2.0438127517700195,
      "learning_rate": 5.284477803884708e-05,
      "loss": 1.8588,
      "step": 225
    },
    {
      "epoch": 0.49344978165938863,
      "grad_norm": 2.173848867416382,
      "learning_rate": 5.27760739739323e-05,
      "loss": 1.7636,
      "step": 226
    },
    {
      "epoch": 0.49563318777292575,
      "grad_norm": 2.206197500228882,
      "learning_rate": 5.27070867549843e-05,
      "loss": 1.7601,
      "step": 227
    },
    {
      "epoch": 0.4978165938864629,
      "grad_norm": 1.1620668172836304,
      "learning_rate": 5.263781723965789e-05,
      "loss": 1.9674,
      "step": 228
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.0804260969161987,
      "learning_rate": 5.256826628911739e-05,
      "loss": 1.7915,
      "step": 229
    },
    {
      "epoch": 0.5021834061135371,
      "grad_norm": 1.1664297580718994,
      "learning_rate": 5.2498434768025975e-05,
      "loss": 1.8489,
      "step": 230
    },
    {
      "epoch": 0.5043668122270742,
      "grad_norm": 2.807969093322754,
      "learning_rate": 5.242832354453489e-05,
      "loss": 1.6818,
      "step": 231
    },
    {
      "epoch": 0.5065502183406113,
      "grad_norm": 1.4173234701156616,
      "learning_rate": 5.2357933490272644e-05,
      "loss": 1.5504,
      "step": 232
    },
    {
      "epoch": 0.5087336244541485,
      "grad_norm": 0.9084534645080566,
      "learning_rate": 5.228726548033421e-05,
      "loss": 1.9287,
      "step": 233
    },
    {
      "epoch": 0.5109170305676856,
      "grad_norm": 1.155557632446289,
      "learning_rate": 5.221632039327014e-05,
      "loss": 1.9677,
      "step": 234
    },
    {
      "epoch": 0.5131004366812227,
      "grad_norm": 4.166019916534424,
      "learning_rate": 5.21450991110756e-05,
      "loss": 1.9809,
      "step": 235
    },
    {
      "epoch": 0.5152838427947598,
      "grad_norm": 1.8789658546447754,
      "learning_rate": 5.2073602519179454e-05,
      "loss": 1.9389,
      "step": 236
    },
    {
      "epoch": 0.517467248908297,
      "grad_norm": 1.3924376964569092,
      "learning_rate": 5.200183150643326e-05,
      "loss": 1.8056,
      "step": 237
    },
    {
      "epoch": 0.519650655021834,
      "grad_norm": 1.6234887838363647,
      "learning_rate": 5.192978696510015e-05,
      "loss": 1.9687,
      "step": 238
    },
    {
      "epoch": 0.5218340611353712,
      "grad_norm": 2.0562949180603027,
      "learning_rate": 5.1857469790843836e-05,
      "loss": 1.8923,
      "step": 239
    },
    {
      "epoch": 0.5240174672489083,
      "grad_norm": 4.884009838104248,
      "learning_rate": 5.178488088271738e-05,
      "loss": 1.8317,
      "step": 240
    },
    {
      "epoch": 0.5262008733624454,
      "grad_norm": 1.523373007774353,
      "learning_rate": 5.1712021143152105e-05,
      "loss": 1.6786,
      "step": 241
    },
    {
      "epoch": 0.5283842794759825,
      "grad_norm": 1.2392911911010742,
      "learning_rate": 5.163889147794631e-05,
      "loss": 1.6499,
      "step": 242
    },
    {
      "epoch": 0.5305676855895196,
      "grad_norm": 1.0265744924545288,
      "learning_rate": 5.1565492796254024e-05,
      "loss": 1.6074,
      "step": 243
    },
    {
      "epoch": 0.5327510917030568,
      "grad_norm": 10.814948081970215,
      "learning_rate": 5.149182601057372e-05,
      "loss": 1.7312,
      "step": 244
    },
    {
      "epoch": 0.5349344978165939,
      "grad_norm": 2.797271490097046,
      "learning_rate": 5.141789203673698e-05,
      "loss": 1.7403,
      "step": 245
    },
    {
      "epoch": 0.537117903930131,
      "grad_norm": 1.3630836009979248,
      "learning_rate": 5.134369179389705e-05,
      "loss": 2.029,
      "step": 246
    },
    {
      "epoch": 0.5393013100436681,
      "grad_norm": 1.0868947505950928,
      "learning_rate": 5.12692262045175e-05,
      "loss": 1.7049,
      "step": 247
    },
    {
      "epoch": 0.5414847161572053,
      "grad_norm": 1.1137341260910034,
      "learning_rate": 5.119449619436068e-05,
      "loss": 1.8015,
      "step": 248
    },
    {
      "epoch": 0.5436681222707423,
      "grad_norm": 2.2383477687835693,
      "learning_rate": 5.111950269247626e-05,
      "loss": 1.8359,
      "step": 249
    },
    {
      "epoch": 0.5458515283842795,
      "grad_norm": 1.1027755737304688,
      "learning_rate": 5.104424663118964e-05,
      "loss": 1.7755,
      "step": 250
    },
    {
      "epoch": 0.5480349344978166,
      "grad_norm": 2.326738119125366,
      "learning_rate": 5.096872894609039e-05,
      "loss": 1.8242,
      "step": 251
    },
    {
      "epoch": 0.5502183406113537,
      "grad_norm": 1.4068495035171509,
      "learning_rate": 5.0892950576020614e-05,
      "loss": 1.6378,
      "step": 252
    },
    {
      "epoch": 0.5524017467248908,
      "grad_norm": 1.416374683380127,
      "learning_rate": 5.081691246306327e-05,
      "loss": 1.7617,
      "step": 253
    },
    {
      "epoch": 0.5545851528384279,
      "grad_norm": 1.3435388803482056,
      "learning_rate": 5.074061555253046e-05,
      "loss": 1.6573,
      "step": 254
    },
    {
      "epoch": 0.5567685589519651,
      "grad_norm": 0.9839504957199097,
      "learning_rate": 5.0664060792951675e-05,
      "loss": 1.6569,
      "step": 255
    },
    {
      "epoch": 0.5589519650655022,
      "grad_norm": 0.856988787651062,
      "learning_rate": 5.0587249136062016e-05,
      "loss": 1.8644,
      "step": 256
    },
    {
      "epoch": 0.5611353711790393,
      "grad_norm": 1.7147274017333984,
      "learning_rate": 5.0510181536790336e-05,
      "loss": 1.8451,
      "step": 257
    },
    {
      "epoch": 0.5633187772925764,
      "grad_norm": 3.185845136642456,
      "learning_rate": 5.0432858953247404e-05,
      "loss": 1.5912,
      "step": 258
    },
    {
      "epoch": 0.5655021834061136,
      "grad_norm": 2.470088243484497,
      "learning_rate": 5.035528234671397e-05,
      "loss": 1.7339,
      "step": 259
    },
    {
      "epoch": 0.5676855895196506,
      "grad_norm": 1.6962615251541138,
      "learning_rate": 5.027745268162881e-05,
      "loss": 1.7884,
      "step": 260
    },
    {
      "epoch": 0.5698689956331878,
      "grad_norm": 0.8670432567596436,
      "learning_rate": 5.0199370925576744e-05,
      "loss": 1.9467,
      "step": 261
    },
    {
      "epoch": 0.5720524017467249,
      "grad_norm": 1.4319339990615845,
      "learning_rate": 5.0121038049276623e-05,
      "loss": 1.8036,
      "step": 262
    },
    {
      "epoch": 0.574235807860262,
      "grad_norm": 0.9973216652870178,
      "learning_rate": 5.0042455026569246e-05,
      "loss": 1.7747,
      "step": 263
    },
    {
      "epoch": 0.5764192139737991,
      "grad_norm": 4.020474910736084,
      "learning_rate": 4.996362283440522e-05,
      "loss": 1.7755,
      "step": 264
    },
    {
      "epoch": 0.5786026200873362,
      "grad_norm": 1.2609071731567383,
      "learning_rate": 4.988454245283289e-05,
      "loss": 1.6582,
      "step": 265
    },
    {
      "epoch": 0.5807860262008734,
      "grad_norm": 0.9482327103614807,
      "learning_rate": 4.9805214864986075e-05,
      "loss": 1.8606,
      "step": 266
    },
    {
      "epoch": 0.5829694323144105,
      "grad_norm": 1.0823765993118286,
      "learning_rate": 4.972564105707191e-05,
      "loss": 1.8036,
      "step": 267
    },
    {
      "epoch": 0.5851528384279476,
      "grad_norm": 1.4853042364120483,
      "learning_rate": 4.964582201835856e-05,
      "loss": 1.9104,
      "step": 268
    },
    {
      "epoch": 0.5873362445414847,
      "grad_norm": 6.773947238922119,
      "learning_rate": 4.9565758741162886e-05,
      "loss": 1.6319,
      "step": 269
    },
    {
      "epoch": 0.5895196506550219,
      "grad_norm": 1.9097248315811157,
      "learning_rate": 4.948545222083817e-05,
      "loss": 1.8555,
      "step": 270
    },
    {
      "epoch": 0.5917030567685589,
      "grad_norm": 2.1053848266601562,
      "learning_rate": 4.9404903455761724e-05,
      "loss": 1.6979,
      "step": 271
    },
    {
      "epoch": 0.5938864628820961,
      "grad_norm": 0.9235579967498779,
      "learning_rate": 4.9324113447322436e-05,
      "loss": 1.9056,
      "step": 272
    },
    {
      "epoch": 0.5960698689956332,
      "grad_norm": 1.186112880706787,
      "learning_rate": 4.924308319990837e-05,
      "loss": 1.5734,
      "step": 273
    },
    {
      "epoch": 0.5982532751091703,
      "grad_norm": 1.9173979759216309,
      "learning_rate": 4.916181372089424e-05,
      "loss": 1.6387,
      "step": 274
    },
    {
      "epoch": 0.6004366812227074,
      "grad_norm": 1.2616807222366333,
      "learning_rate": 4.908030602062895e-05,
      "loss": 1.8001,
      "step": 275
    },
    {
      "epoch": 0.6026200873362445,
      "grad_norm": 1.4946835041046143,
      "learning_rate": 4.899856111242294e-05,
      "loss": 1.6381,
      "step": 276
    }
  ],
  "logging_steps": 1,
  "max_steps": 916,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 46,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.4029508078057554e+18,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}