{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 375,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.008,
      "grad_norm": 14.56950855255127,
      "learning_rate": 2.6315789473684213e-07,
      "loss": 0.8339,
      "step": 1
    },
    {
      "epoch": 0.016,
      "grad_norm": 13.73933219909668,
      "learning_rate": 5.263157894736843e-07,
      "loss": 0.8732,
      "step": 2
    },
    {
      "epoch": 0.024,
      "grad_norm": 14.081198692321777,
      "learning_rate": 7.894736842105263e-07,
      "loss": 0.894,
      "step": 3
    },
    {
      "epoch": 0.032,
      "grad_norm": 14.026392936706543,
      "learning_rate": 1.0526315789473685e-06,
      "loss": 0.8509,
      "step": 4
    },
    {
      "epoch": 0.04,
      "grad_norm": 14.380779266357422,
      "learning_rate": 1.3157894736842106e-06,
      "loss": 0.8362,
      "step": 5
    },
    {
      "epoch": 0.048,
      "grad_norm": 12.984580993652344,
      "learning_rate": 1.5789473684210526e-06,
      "loss": 0.8653,
      "step": 6
    },
    {
      "epoch": 0.056,
      "grad_norm": 10.433991432189941,
      "learning_rate": 1.8421052631578948e-06,
      "loss": 0.8321,
      "step": 7
    },
    {
      "epoch": 0.064,
      "grad_norm": 9.493083000183105,
      "learning_rate": 2.105263157894737e-06,
      "loss": 0.8143,
      "step": 8
    },
    {
      "epoch": 0.072,
      "grad_norm": 7.640753746032715,
      "learning_rate": 2.368421052631579e-06,
      "loss": 0.7905,
      "step": 9
    },
    {
      "epoch": 0.08,
      "grad_norm": 4.019260883331299,
      "learning_rate": 2.631578947368421e-06,
      "loss": 0.7129,
      "step": 10
    },
    {
      "epoch": 0.088,
      "grad_norm": 3.2670819759368896,
      "learning_rate": 2.8947368421052634e-06,
      "loss": 0.7227,
      "step": 11
    },
    {
      "epoch": 0.096,
      "grad_norm": 2.992112398147583,
      "learning_rate": 3.157894736842105e-06,
      "loss": 0.6972,
      "step": 12
    },
    {
      "epoch": 0.104,
      "grad_norm": 2.052677869796753,
      "learning_rate": 3.421052631578948e-06,
      "loss": 0.678,
      "step": 13
    },
    {
      "epoch": 0.112,
      "grad_norm": 1.920199990272522,
      "learning_rate": 3.6842105263157896e-06,
      "loss": 0.667,
      "step": 14
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.706879734992981,
      "learning_rate": 3.947368421052632e-06,
      "loss": 0.6192,
      "step": 15
    },
    {
      "epoch": 0.128,
      "grad_norm": 1.6659389734268188,
      "learning_rate": 4.210526315789474e-06,
      "loss": 0.6118,
      "step": 16
    },
    {
      "epoch": 0.136,
      "grad_norm": 1.3653850555419922,
      "learning_rate": 4.473684210526316e-06,
      "loss": 0.5681,
      "step": 17
    },
    {
      "epoch": 0.144,
      "grad_norm": 1.4679189920425415,
      "learning_rate": 4.736842105263158e-06,
      "loss": 0.607,
      "step": 18
    },
    {
      "epoch": 0.152,
      "grad_norm": 1.2237054109573364,
      "learning_rate": 5e-06,
      "loss": 0.5273,
      "step": 19
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.9686553478240967,
      "learning_rate": 5.263157894736842e-06,
      "loss": 0.5578,
      "step": 20
    },
    {
      "epoch": 0.168,
      "grad_norm": 0.7895737290382385,
      "learning_rate": 5.526315789473685e-06,
      "loss": 0.5794,
      "step": 21
    },
    {
      "epoch": 0.176,
      "grad_norm": 0.8087520599365234,
      "learning_rate": 5.789473684210527e-06,
      "loss": 0.5229,
      "step": 22
    },
    {
      "epoch": 0.184,
      "grad_norm": 0.7652890086174011,
      "learning_rate": 6.0526315789473685e-06,
      "loss": 0.5292,
      "step": 23
    },
    {
      "epoch": 0.192,
      "grad_norm": 0.9132710695266724,
      "learning_rate": 6.31578947368421e-06,
      "loss": 0.5271,
      "step": 24
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.8860102295875549,
      "learning_rate": 6.578947368421054e-06,
      "loss": 0.5155,
      "step": 25
    },
    {
      "epoch": 0.208,
      "grad_norm": 0.8902170062065125,
      "learning_rate": 6.842105263157896e-06,
      "loss": 0.5287,
      "step": 26
    },
    {
      "epoch": 0.216,
      "grad_norm": 0.8824247121810913,
      "learning_rate": 7.1052631578947375e-06,
      "loss": 0.4791,
      "step": 27
    },
    {
      "epoch": 0.224,
      "grad_norm": 1.2683454751968384,
      "learning_rate": 7.368421052631579e-06,
      "loss": 0.5329,
      "step": 28
    },
    {
      "epoch": 0.232,
      "grad_norm": 0.7248997092247009,
      "learning_rate": 7.631578947368423e-06,
      "loss": 0.5294,
      "step": 29
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.7209534645080566,
      "learning_rate": 7.894736842105265e-06,
      "loss": 0.4842,
      "step": 30
    },
    {
      "epoch": 0.248,
      "grad_norm": 0.7555510401725769,
      "learning_rate": 8.157894736842106e-06,
      "loss": 0.5348,
      "step": 31
    },
    {
      "epoch": 0.256,
      "grad_norm": 0.6169366240501404,
      "learning_rate": 8.421052631578948e-06,
      "loss": 0.5382,
      "step": 32
    },
    {
      "epoch": 0.264,
      "grad_norm": 0.6152195930480957,
      "learning_rate": 8.68421052631579e-06,
      "loss": 0.5039,
      "step": 33
    },
    {
      "epoch": 0.272,
      "grad_norm": 0.631288468837738,
      "learning_rate": 8.947368421052632e-06,
      "loss": 0.5037,
      "step": 34
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.5909512042999268,
      "learning_rate": 9.210526315789474e-06,
      "loss": 0.5017,
      "step": 35
    },
    {
      "epoch": 0.288,
      "grad_norm": 0.5489953756332397,
      "learning_rate": 9.473684210526315e-06,
      "loss": 0.4469,
      "step": 36
    },
    {
      "epoch": 0.296,
      "grad_norm": 0.565674901008606,
      "learning_rate": 9.736842105263159e-06,
      "loss": 0.478,
      "step": 37
    },
    {
      "epoch": 0.304,
      "grad_norm": 0.680474579334259,
      "learning_rate": 1e-05,
      "loss": 0.5392,
      "step": 38
    },
    {
      "epoch": 0.312,
      "grad_norm": 0.607657253742218,
      "learning_rate": 9.99978274148479e-06,
      "loss": 0.5022,
      "step": 39
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.5720808506011963,
      "learning_rate": 9.999130984819662e-06,
      "loss": 0.5057,
      "step": 40
    },
    {
      "epoch": 0.328,
      "grad_norm": 0.5648075938224792,
      "learning_rate": 9.998044786644492e-06,
      "loss": 0.4718,
      "step": 41
    },
    {
      "epoch": 0.336,
      "grad_norm": 0.5617280006408691,
      "learning_rate": 9.9965242413536e-06,
      "loss": 0.5049,
      "step": 42
    },
    {
      "epoch": 0.344,
      "grad_norm": 0.6486408114433289,
      "learning_rate": 9.994569481087552e-06,
      "loss": 0.5203,
      "step": 43
    },
    {
      "epoch": 0.352,
      "grad_norm": 0.5352612733840942,
      "learning_rate": 9.992180675721671e-06,
      "loss": 0.4967,
      "step": 44
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.5438513159751892,
      "learning_rate": 9.989358032851283e-06,
      "loss": 0.5188,
      "step": 45
    },
    {
      "epoch": 0.368,
      "grad_norm": 0.5454742908477783,
      "learning_rate": 9.986101797773667e-06,
      "loss": 0.4864,
      "step": 46
    },
    {
      "epoch": 0.376,
      "grad_norm": 0.536598265171051,
      "learning_rate": 9.98241225346674e-06,
      "loss": 0.4883,
      "step": 47
    },
    {
      "epoch": 0.384,
      "grad_norm": 0.4659676253795624,
      "learning_rate": 9.978289720564471e-06,
      "loss": 0.5121,
      "step": 48
    },
    {
      "epoch": 0.392,
      "grad_norm": 0.5836896300315857,
      "learning_rate": 9.97373455732901e-06,
      "loss": 0.4975,
      "step": 49
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.5226072669029236,
      "learning_rate": 9.968747159619556e-06,
      "loss": 0.4944,
      "step": 50
    },
    {
      "epoch": 0.408,
      "grad_norm": 0.5444321632385254,
      "learning_rate": 9.963327960857962e-06,
      "loss": 0.5209,
      "step": 51
    },
    {
      "epoch": 0.416,
      "grad_norm": 0.5670180320739746,
      "learning_rate": 9.957477431991053e-06,
      "loss": 0.514,
      "step": 52
    },
    {
      "epoch": 0.424,
      "grad_norm": 0.511341392993927,
      "learning_rate": 9.95119608144972e-06,
      "loss": 0.526,
      "step": 53
    },
    {
      "epoch": 0.432,
      "grad_norm": 0.5060410499572754,
      "learning_rate": 9.944484455104716e-06,
      "loss": 0.4806,
      "step": 54
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.487128347158432,
      "learning_rate": 9.937343136219234e-06,
      "loss": 0.5273,
      "step": 55
    },
    {
      "epoch": 0.448,
      "grad_norm": 0.5598118305206299,
      "learning_rate": 9.929772745398207e-06,
      "loss": 0.5745,
      "step": 56
    },
    {
      "epoch": 0.456,
      "grad_norm": 0.5080751180648804,
      "learning_rate": 9.921773940534382e-06,
      "loss": 0.5082,
      "step": 57
    },
    {
      "epoch": 0.464,
      "grad_norm": 0.5492854118347168,
      "learning_rate": 9.913347416751148e-06,
      "loss": 0.5244,
      "step": 58
    },
    {
      "epoch": 0.472,
      "grad_norm": 0.5565740466117859,
      "learning_rate": 9.904493906342124e-06,
      "loss": 0.5117,
      "step": 59
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.5090964436531067,
      "learning_rate": 9.895214178707516e-06,
      "loss": 0.4942,
      "step": 60
    },
    {
      "epoch": 0.488,
      "grad_norm": 0.5151892304420471,
      "learning_rate": 9.885509040287267e-06,
      "loss": 0.5165,
      "step": 61
    },
    {
      "epoch": 0.496,
      "grad_norm": 0.5298195481300354,
      "learning_rate": 9.875379334490962e-06,
      "loss": 0.5218,
      "step": 62
    },
    {
      "epoch": 0.504,
      "grad_norm": 0.5165826082229614,
      "learning_rate": 9.864825941624538e-06,
      "loss": 0.4986,
      "step": 63
    },
    {
      "epoch": 0.512,
      "grad_norm": 0.506293773651123,
      "learning_rate": 9.853849778813777e-06,
      "loss": 0.5127,
      "step": 64
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.5140017867088318,
      "learning_rate": 9.842451799924616e-06,
      "loss": 0.5158,
      "step": 65
    },
    {
      "epoch": 0.528,
      "grad_norm": 0.5674154162406921,
      "learning_rate": 9.830632995480243e-06,
      "loss": 0.5006,
      "step": 66
    },
    {
      "epoch": 0.536,
      "grad_norm": 0.5346920490264893,
      "learning_rate": 9.818394392575018e-06,
      "loss": 0.5126,
      "step": 67
    },
    {
      "epoch": 0.544,
      "grad_norm": 0.5248146653175354,
      "learning_rate": 9.805737054785223e-06,
      "loss": 0.5394,
      "step": 68
    },
    {
      "epoch": 0.552,
      "grad_norm": 0.5473419427871704,
      "learning_rate": 9.792662082076618e-06,
      "loss": 0.5224,
      "step": 69
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.5517472624778748,
      "learning_rate": 9.779170610708872e-06,
      "loss": 0.5125,
      "step": 70
    },
    {
      "epoch": 0.568,
      "grad_norm": 0.4769006669521332,
      "learning_rate": 9.765263813136796e-06,
      "loss": 0.5081,
      "step": 71
    },
    {
      "epoch": 0.576,
      "grad_norm": 0.5075954794883728,
      "learning_rate": 9.750942897908468e-06,
      "loss": 0.4958,
      "step": 72
    },
    {
      "epoch": 0.584,
      "grad_norm": 0.5006887316703796,
      "learning_rate": 9.736209109560201e-06,
      "loss": 0.5279,
      "step": 73
    },
    {
      "epoch": 0.592,
      "grad_norm": 0.5326476693153381,
      "learning_rate": 9.721063728508384e-06,
      "loss": 0.5136,
      "step": 74
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.5512110590934753,
      "learning_rate": 9.705508070938219e-06,
      "loss": 0.474,
      "step": 75
    },
    {
      "epoch": 0.608,
      "grad_norm": 0.5133175253868103,
      "learning_rate": 9.689543488689332e-06,
      "loss": 0.4994,
      "step": 76
    },
    {
      "epoch": 0.616,
      "grad_norm": 0.5255376696586609,
      "learning_rate": 9.673171369138297e-06,
      "loss": 0.5183,
      "step": 77
    },
    {
      "epoch": 0.624,
      "grad_norm": 0.5707758665084839,
      "learning_rate": 9.656393135078067e-06,
      "loss": 0.493,
      "step": 78
    },
    {
      "epoch": 0.632,
      "grad_norm": 0.5201893448829651,
      "learning_rate": 9.639210244594335e-06,
      "loss": 0.4635,
      "step": 79
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.5058737397193909,
      "learning_rate": 9.621624190938802e-06,
      "loss": 0.5312,
      "step": 80
    },
    {
      "epoch": 0.648,
      "grad_norm": 0.43761327862739563,
      "learning_rate": 9.603636502399436e-06,
      "loss": 0.4524,
      "step": 81
    },
    {
      "epoch": 0.656,
      "grad_norm": 0.5335831642150879,
      "learning_rate": 9.585248742167638e-06,
      "loss": 0.4648,
      "step": 82
    },
    {
      "epoch": 0.664,
      "grad_norm": 0.5373964309692383,
      "learning_rate": 9.566462508202403e-06,
      "loss": 0.4911,
      "step": 83
    },
    {
      "epoch": 0.672,
      "grad_norm": 0.5256152153015137,
      "learning_rate": 9.547279433091446e-06,
      "loss": 0.4968,
      "step": 84
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.5257714986801147,
      "learning_rate": 9.527701183909336e-06,
      "loss": 0.4879,
      "step": 85
    },
    {
      "epoch": 0.688,
      "grad_norm": 0.5462167859077454,
      "learning_rate": 9.507729462072615e-06,
      "loss": 0.4577,
      "step": 86
    },
    {
      "epoch": 0.696,
      "grad_norm": 0.5356786847114563,
      "learning_rate": 9.48736600319193e-06,
      "loss": 0.4922,
      "step": 87
    },
    {
      "epoch": 0.704,
      "grad_norm": 0.5588122010231018,
      "learning_rate": 9.466612576921223e-06,
      "loss": 0.4915,
      "step": 88
    },
    {
      "epoch": 0.712,
      "grad_norm": 0.5467613935470581,
      "learning_rate": 9.445470986803922e-06,
      "loss": 0.4711,
      "step": 89
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.5490957498550415,
      "learning_rate": 9.423943070116219e-06,
      "loss": 0.5108,
      "step": 90
    },
    {
      "epoch": 0.728,
      "grad_norm": 0.4839731752872467,
      "learning_rate": 9.402030697707398e-06,
      "loss": 0.5128,
      "step": 91
    },
    {
      "epoch": 0.736,
      "grad_norm": 0.5173898935317993,
      "learning_rate": 9.37973577383726e-06,
      "loss": 0.4823,
      "step": 92
    },
    {
      "epoch": 0.744,
      "grad_norm": 0.563199520111084,
      "learning_rate": 9.357060236010626e-06,
      "loss": 0.4985,
      "step": 93
    },
    {
      "epoch": 0.752,
      "grad_norm": 0.513721227645874,
      "learning_rate": 9.334006054808966e-06,
      "loss": 0.4484,
      "step": 94
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.5133054256439209,
      "learning_rate": 9.310575233719155e-06,
      "loss": 0.5047,
      "step": 95
    },
    {
      "epoch": 0.768,
      "grad_norm": 0.5787802338600159,
      "learning_rate": 9.28676980895935e-06,
      "loss": 0.5064,
      "step": 96
    },
    {
      "epoch": 0.776,
      "grad_norm": 0.4748121500015259,
      "learning_rate": 9.262591849302049e-06,
      "loss": 0.4604,
      "step": 97
    },
    {
      "epoch": 0.784,
      "grad_norm": 0.556131899356842,
      "learning_rate": 9.238043455894294e-06,
      "loss": 0.5295,
      "step": 98
    },
    {
      "epoch": 0.792,
      "grad_norm": 0.5900002717971802,
      "learning_rate": 9.213126762075088e-06,
      "loss": 0.4977,
      "step": 99
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.5788969397544861,
      "learning_rate": 9.187843933189994e-06,
      "loss": 0.4732,
      "step": 100
    },
    {
      "epoch": 0.808,
      "grad_norm": 0.5433998107910156,
      "learning_rate": 9.162197166402957e-06,
      "loss": 0.5175,
      "step": 101
    },
    {
      "epoch": 0.816,
      "grad_norm": 0.5408159494400024,
      "learning_rate": 9.136188690505363e-06,
      "loss": 0.5087,
      "step": 102
    },
    {
      "epoch": 0.824,
      "grad_norm": 0.5552597045898438,
      "learning_rate": 9.109820765722357e-06,
      "loss": 0.4835,
      "step": 103
    },
    {
      "epoch": 0.832,
      "grad_norm": 0.5368766188621521,
      "learning_rate": 9.083095683516414e-06,
      "loss": 0.5349,
      "step": 104
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.46206405758857727,
      "learning_rate": 9.056015766388205e-06,
      "loss": 0.4567,
      "step": 105
    },
    {
      "epoch": 0.848,
      "grad_norm": 0.5309408903121948,
      "learning_rate": 9.028583367674767e-06,
      "loss": 0.5166,
      "step": 106
    },
    {
      "epoch": 0.856,
      "grad_norm": 0.528049886226654,
      "learning_rate": 9.00080087134498e-06,
      "loss": 0.4526,
      "step": 107
    },
    {
      "epoch": 0.864,
      "grad_norm": 0.5379208326339722,
      "learning_rate": 8.972670691792409e-06,
      "loss": 0.4893,
      "step": 108
    },
    {
      "epoch": 0.872,
      "grad_norm": 0.5074729919433594,
      "learning_rate": 8.944195273625472e-06,
      "loss": 0.4924,
      "step": 109
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.5556723475456238,
      "learning_rate": 8.915377091454992e-06,
      "loss": 0.4981,
      "step": 110
    },
    {
      "epoch": 0.888,
      "grad_norm": 0.5345364212989807,
      "learning_rate": 8.886218649679162e-06,
      "loss": 0.5134,
      "step": 111
    },
    {
      "epoch": 0.896,
      "grad_norm": 0.511635959148407,
      "learning_rate": 8.856722482265886e-06,
      "loss": 0.461,
      "step": 112
    },
    {
      "epoch": 0.904,
      "grad_norm": 0.49792084097862244,
      "learning_rate": 8.826891152532579e-06,
      "loss": 0.505,
      "step": 113
    },
    {
      "epoch": 0.912,
      "grad_norm": 0.4829058051109314,
      "learning_rate": 8.796727252923403e-06,
      "loss": 0.4716,
      "step": 114
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.5941436886787415,
      "learning_rate": 8.766233404783975e-06,
      "loss": 0.5296,
      "step": 115
    },
    {
      "epoch": 0.928,
      "grad_norm": 0.5134578943252563,
      "learning_rate": 8.735412258133562e-06,
      "loss": 0.5045,
      "step": 116
    },
    {
      "epoch": 0.936,
      "grad_norm": 0.5325246453285217,
      "learning_rate": 8.704266491434787e-06,
      "loss": 0.472,
      "step": 117
    },
    {
      "epoch": 0.944,
      "grad_norm": 0.538865327835083,
      "learning_rate": 8.672798811360863e-06,
      "loss": 0.5026,
      "step": 118
    },
    {
      "epoch": 0.952,
      "grad_norm": 0.48347458243370056,
      "learning_rate": 8.641011952560372e-06,
      "loss": 0.484,
      "step": 119
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.5306034684181213,
      "learning_rate": 8.608908677419606e-06,
      "loss": 0.4997,
      "step": 120
    },
    {
      "epoch": 0.968,
      "grad_norm": 0.5399216413497925,
      "learning_rate": 8.576491775822527e-06,
      "loss": 0.4804,
      "step": 121
    },
    {
      "epoch": 0.976,
      "grad_norm": 0.5183812975883484,
      "learning_rate": 8.543764064908295e-06,
      "loss": 0.4887,
      "step": 122
    },
    {
      "epoch": 0.984,
      "grad_norm": 0.4991128444671631,
      "learning_rate": 8.510728388826464e-06,
      "loss": 0.4453,
      "step": 123
    },
    {
      "epoch": 0.992,
      "grad_norm": 0.5225016474723816,
      "learning_rate": 8.477387618489808e-06,
      "loss": 0.5123,
      "step": 124
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.46465304493904114,
      "learning_rate": 8.443744651324828e-06,
      "loss": 0.4633,
      "step": 125
    },
    {
      "epoch": 1.008,
      "grad_norm": 0.5113005638122559,
      "learning_rate": 8.409802411019962e-06,
      "loss": 0.427,
      "step": 126
    },
    {
      "epoch": 1.016,
      "grad_norm": 0.5328509211540222,
      "learning_rate": 8.375563847271506e-06,
      "loss": 0.4324,
      "step": 127
    },
    {
      "epoch": 1.024,
      "grad_norm": 0.519523024559021,
      "learning_rate": 8.341031935527267e-06,
      "loss": 0.4387,
      "step": 128
    },
    {
      "epoch": 1.032,
      "grad_norm": 0.506823718547821,
      "learning_rate": 8.306209676727994e-06,
      "loss": 0.4469,
      "step": 129
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.4721229374408722,
      "learning_rate": 8.271100097046585e-06,
      "loss": 0.4226,
      "step": 130
    },
    {
      "epoch": 1.048,
      "grad_norm": 0.4656391739845276,
      "learning_rate": 8.235706247625098e-06,
      "loss": 0.3729,
      "step": 131
    },
    {
      "epoch": 1.056,
      "grad_norm": 0.4901474118232727,
      "learning_rate": 8.200031204309604e-06,
      "loss": 0.4627,
      "step": 132
    },
    {
      "epoch": 1.064,
      "grad_norm": 0.606958270072937,
      "learning_rate": 8.16407806738288e-06,
      "loss": 0.4347,
      "step": 133
    },
    {
      "epoch": 1.072,
      "grad_norm": 0.4573584496974945,
      "learning_rate": 8.127849961294984e-06,
      "loss": 0.402,
      "step": 134
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.4766193926334381,
      "learning_rate": 8.091350034391732e-06,
      "loss": 0.4203,
      "step": 135
    },
    {
      "epoch": 1.088,
      "grad_norm": 0.5167376399040222,
      "learning_rate": 8.05458145864109e-06,
      "loss": 0.4521,
      "step": 136
    },
    {
      "epoch": 1.096,
      "grad_norm": 0.52271968126297,
      "learning_rate": 8.017547429357532e-06,
      "loss": 0.4321,
      "step": 137
    },
    {
      "epoch": 1.104,
      "grad_norm": 0.49505582451820374,
      "learning_rate": 7.980251164924342e-06,
      "loss": 0.4591,
      "step": 138
    },
    {
      "epoch": 1.112,
      "grad_norm": 0.6967505216598511,
      "learning_rate": 7.94269590651393e-06,
      "loss": 0.4322,
      "step": 139
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.46629568934440613,
      "learning_rate": 7.904884917806174e-06,
      "loss": 0.4612,
      "step": 140
    },
    {
      "epoch": 1.1280000000000001,
      "grad_norm": 0.49921801686286926,
      "learning_rate": 7.866821484704777e-06,
      "loss": 0.4353,
      "step": 141
    },
    {
      "epoch": 1.1360000000000001,
      "grad_norm": 0.5447902083396912,
      "learning_rate": 7.828508915051724e-06,
      "loss": 0.4323,
      "step": 142
    },
    {
      "epoch": 1.144,
      "grad_norm": 0.5017483830451965,
      "learning_rate": 7.789950538339813e-06,
      "loss": 0.4352,
      "step": 143
    },
    {
      "epoch": 1.152,
      "grad_norm": 0.49647992849349976,
      "learning_rate": 7.751149705423313e-06,
      "loss": 0.3971,
      "step": 144
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.5001079440116882,
      "learning_rate": 7.712109788226763e-06,
      "loss": 0.4566,
      "step": 145
    },
    {
      "epoch": 1.168,
      "grad_norm": 0.4786428213119507,
      "learning_rate": 7.672834179451943e-06,
      "loss": 0.446,
      "step": 146
    },
    {
      "epoch": 1.176,
      "grad_norm": 0.5244361758232117,
      "learning_rate": 7.633326292283028e-06,
      "loss": 0.4759,
      "step": 147
    },
    {
      "epoch": 1.184,
      "grad_norm": 0.4763104319572449,
      "learning_rate": 7.593589560089984e-06,
      "loss": 0.4239,
      "step": 148
    },
    {
      "epoch": 1.192,
      "grad_norm": 0.4974442422389984,
      "learning_rate": 7.553627436130183e-06,
      "loss": 0.4053,
      "step": 149
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.5250166654586792,
      "learning_rate": 7.513443393248312e-06,
      "loss": 0.4521,
      "step": 150
    },
    {
      "epoch": 1.208,
      "grad_norm": 0.4822094440460205,
      "learning_rate": 7.473040923574567e-06,
      "loss": 0.3883,
      "step": 151
    },
    {
      "epoch": 1.216,
      "grad_norm": 0.4862370491027832,
      "learning_rate": 7.432423538221179e-06,
      "loss": 0.4322,
      "step": 152
    },
    {
      "epoch": 1.224,
      "grad_norm": 0.482036828994751,
      "learning_rate": 7.391594766977277e-06,
      "loss": 0.4105,
      "step": 153
    },
    {
      "epoch": 1.232,
      "grad_norm": 0.5110259056091309,
      "learning_rate": 7.350558158002154e-06,
      "loss": 0.4103,
      "step": 154
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.5026330351829529,
      "learning_rate": 7.3093172775169e-06,
      "loss": 0.4312,
      "step": 155
    },
    {
      "epoch": 1.248,
      "grad_norm": 0.4886966943740845,
      "learning_rate": 7.2678757094945e-06,
      "loss": 0.4325,
      "step": 156
    },
    {
      "epoch": 1.256,
      "grad_norm": 0.4985504150390625,
      "learning_rate": 7.226237055348369e-06,
      "loss": 0.4232,
      "step": 157
    },
    {
      "epoch": 1.264,
      "grad_norm": 0.47912245988845825,
      "learning_rate": 7.184404933619377e-06,
      "loss": 0.4359,
      "step": 158
    },
    {
      "epoch": 1.272,
      "grad_norm": 0.48068177700042725,
      "learning_rate": 7.142382979661386e-06,
      "loss": 0.4158,
      "step": 159
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.4839794933795929,
      "learning_rate": 7.100174845325327e-06,
      "loss": 0.4035,
      "step": 160
    },
    {
      "epoch": 1.288,
      "grad_norm": 0.5554186105728149,
      "learning_rate": 7.057784198641835e-06,
      "loss": 0.3862,
      "step": 161
    },
    {
      "epoch": 1.296,
      "grad_norm": 0.5477203726768494,
      "learning_rate": 7.015214723502496e-06,
      "loss": 0.4347,
      "step": 162
    },
    {
      "epoch": 1.304,
      "grad_norm": 0.48494377732276917,
      "learning_rate": 6.972470119339692e-06,
      "loss": 0.4183,
      "step": 163
    },
    {
      "epoch": 1.312,
      "grad_norm": 0.4876761734485626,
      "learning_rate": 6.929554100805118e-06,
      "loss": 0.4459,
      "step": 164
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.4854036867618561,
      "learning_rate": 6.886470397446958e-06,
      "loss": 0.4249,
      "step": 165
    },
    {
      "epoch": 1.328,
      "grad_norm": 0.4709065556526184,
      "learning_rate": 6.843222753385785e-06,
      "loss": 0.4376,
      "step": 166
    },
    {
      "epoch": 1.336,
      "grad_norm": 0.4834536910057068,
      "learning_rate": 6.799814926989171e-06,
      "loss": 0.4017,
      "step": 167
    },
    {
      "epoch": 1.3439999999999999,
      "grad_norm": 0.4833991527557373,
      "learning_rate": 6.756250690545079e-06,
      "loss": 0.4145,
      "step": 168
    },
    {
      "epoch": 1.3519999999999999,
      "grad_norm": 0.5171043872833252,
      "learning_rate": 6.712533829934042e-06,
      "loss": 0.4373,
      "step": 169
    },
    {
      "epoch": 1.3599999999999999,
      "grad_norm": 0.5456111431121826,
      "learning_rate": 6.6686681443001485e-06,
      "loss": 0.4272,
      "step": 170
    },
    {
      "epoch": 1.3679999999999999,
      "grad_norm": 0.5275494456291199,
      "learning_rate": 6.62465744572089e-06,
      "loss": 0.4628,
      "step": 171
    },
    {
      "epoch": 1.376,
      "grad_norm": 0.46399417519569397,
      "learning_rate": 6.580505558875878e-06,
      "loss": 0.481,
      "step": 172
    },
    {
      "epoch": 1.384,
      "grad_norm": 0.4880594313144684,
      "learning_rate": 6.536216320714466e-06,
      "loss": 0.4285,
      "step": 173
    },
    {
      "epoch": 1.392,
      "grad_norm": 0.48990365862846375,
      "learning_rate": 6.491793580122301e-06,
      "loss": 0.3997,
      "step": 174
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.49995189905166626,
      "learning_rate": 6.447241197586847e-06,
      "loss": 0.4332,
      "step": 175
    },
    {
      "epoch": 1.408,
      "grad_norm": 0.4638576805591583,
      "learning_rate": 6.402563044861899e-06,
      "loss": 0.4236,
      "step": 176
    },
    {
      "epoch": 1.416,
      "grad_norm": 0.49920183420181274,
      "learning_rate": 6.357763004631104e-06,
      "loss": 0.4349,
      "step": 177
    },
    {
      "epoch": 1.424,
      "grad_norm": 0.5193583965301514,
      "learning_rate": 6.312844970170551e-06,
      "loss": 0.4089,
      "step": 178
    },
    {
      "epoch": 1.432,
      "grad_norm": 0.47679874300956726,
      "learning_rate": 6.267812845010431e-06,
      "loss": 0.4298,
      "step": 179
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.49884992837905884,
      "learning_rate": 6.2226705425958e-06,
      "loss": 0.4517,
      "step": 180
    },
    {
      "epoch": 1.448,
      "grad_norm": 0.48136106133461,
      "learning_rate": 6.177421985946499e-06,
      "loss": 0.4393,
      "step": 181
    },
    {
      "epoch": 1.456,
      "grad_norm": 0.5779682397842407,
      "learning_rate": 6.132071107316221e-06,
      "loss": 0.4225,
      "step": 182
    },
    {
      "epoch": 1.464,
      "grad_norm": 0.5509202480316162,
      "learning_rate": 6.0866218478507875e-06,
      "loss": 0.4579,
      "step": 183
    },
    {
      "epoch": 1.472,
      "grad_norm": 0.5101506114006042,
      "learning_rate": 6.041078157245649e-06,
      "loss": 0.4036,
      "step": 184
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.623524010181427,
      "learning_rate": 5.995443993402647e-06,
      "loss": 0.451,
      "step": 185
    },
    {
      "epoch": 1.488,
      "grad_norm": 0.4368286728858948,
      "learning_rate": 5.949723322086053e-06,
      "loss": 0.3881,
      "step": 186
    },
    {
      "epoch": 1.496,
      "grad_norm": 0.4305906295776367,
      "learning_rate": 5.9039201165779315e-06,
      "loss": 0.4036,
      "step": 187
    },
    {
      "epoch": 1.504,
      "grad_norm": 0.4749639630317688,
      "learning_rate": 5.858038357332851e-06,
      "loss": 0.4048,
      "step": 188
    },
    {
      "epoch": 1.512,
      "grad_norm": 0.48530155420303345,
      "learning_rate": 5.812082031631966e-06,
      "loss": 0.4464,
      "step": 189
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.555038571357727,
      "learning_rate": 5.766055133236513e-06,
      "loss": 0.4485,
      "step": 190
    },
    {
      "epoch": 1.528,
      "grad_norm": 0.4892346262931824,
      "learning_rate": 5.7199616620407325e-06,
      "loss": 0.4522,
      "step": 191
    },
    {
      "epoch": 1.536,
      "grad_norm": 0.5365411639213562,
      "learning_rate": 5.673805623724272e-06,
      "loss": 0.4752,
      "step": 192
    },
    {
      "epoch": 1.544,
      "grad_norm": 0.5186147689819336,
      "learning_rate": 5.627591029404072e-06,
      "loss": 0.4021,
      "step": 193
    },
    {
      "epoch": 1.552,
      "grad_norm": 0.48625391721725464,
      "learning_rate": 5.581321895285787e-06,
      "loss": 0.4396,
      "step": 194
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.5071629285812378,
      "learning_rate": 5.535002242314772e-06,
      "loss": 0.4227,
      "step": 195
    },
    {
      "epoch": 1.568,
      "grad_norm": 0.5189520120620728,
      "learning_rate": 5.488636095826636e-06,
      "loss": 0.4512,
      "step": 196
    },
    {
      "epoch": 1.576,
      "grad_norm": 0.5435962677001953,
      "learning_rate": 5.4422274851974356e-06,
      "loss": 0.4767,
      "step": 197
    },
    {
      "epoch": 1.584,
      "grad_norm": 0.4914834797382355,
      "learning_rate": 5.395780443493508e-06,
      "loss": 0.4035,
      "step": 198
    },
    {
      "epoch": 1.592,
      "grad_norm": 0.5276004672050476,
      "learning_rate": 5.34929900712098e-06,
      "loss": 0.4329,
      "step": 199
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.4546476900577545,
      "learning_rate": 5.302787215474992e-06,
      "loss": 0.3965,
      "step": 200
    },
    {
      "epoch": 1.608,
      "grad_norm": 0.48372510075569153,
      "learning_rate": 5.256249110588659e-06,
      "loss": 0.4272,
      "step": 201
    },
    {
      "epoch": 1.616,
      "grad_norm": 0.49213626980781555,
      "learning_rate": 5.209688736781811e-06,
      "loss": 0.4409,
      "step": 202
    },
    {
      "epoch": 1.624,
      "grad_norm": 0.4945046603679657,
      "learning_rate": 5.163110140309518e-06,
      "loss": 0.4419,
      "step": 203
    },
    {
      "epoch": 1.6320000000000001,
      "grad_norm": 0.4468039274215698,
      "learning_rate": 5.116517369010467e-06,
      "loss": 0.398,
      "step": 204
    },
    {
      "epoch": 1.6400000000000001,
      "grad_norm": 0.4727765917778015,
      "learning_rate": 5.069914471955179e-06,
      "loss": 0.4229,
      "step": 205
    },
    {
      "epoch": 1.6480000000000001,
      "grad_norm": 0.5005367398262024,
      "learning_rate": 5.023305499094145e-06,
      "loss": 0.4153,
      "step": 206
    },
    {
      "epoch": 1.6560000000000001,
      "grad_norm": 0.4818647801876068,
      "learning_rate": 4.976694500905858e-06,
      "loss": 0.443,
      "step": 207
    },
    {
      "epoch": 1.6640000000000001,
      "grad_norm": 0.47024303674697876,
      "learning_rate": 4.930085528044823e-06,
      "loss": 0.4541,
      "step": 208
    },
    {
      "epoch": 1.6720000000000002,
      "grad_norm": 0.47553378343582153,
      "learning_rate": 4.883482630989536e-06,
      "loss": 0.4069,
      "step": 209
    },
    {
      "epoch": 1.6800000000000002,
      "grad_norm": 0.44939181208610535,
      "learning_rate": 4.8368898596904834e-06,
      "loss": 0.3964,
      "step": 210
    },
    {
      "epoch": 1.688,
      "grad_norm": 0.46793460845947266,
      "learning_rate": 4.790311263218191e-06,
      "loss": 0.4339,
      "step": 211
    },
    {
      "epoch": 1.696,
      "grad_norm": 0.45191749930381775,
      "learning_rate": 4.743750889411342e-06,
      "loss": 0.426,
      "step": 212
    },
    {
      "epoch": 1.704,
      "grad_norm": 0.45308157801628113,
      "learning_rate": 4.697212784525009e-06,
      "loss": 0.4338,
      "step": 213
    },
    {
      "epoch": 1.712,
      "grad_norm": 0.4543735384941101,
      "learning_rate": 4.65070099287902e-06,
      "loss": 0.4083,
      "step": 214
    },
    {
      "epoch": 1.72,
      "grad_norm": 0.49145424365997314,
      "learning_rate": 4.604219556506492e-06,
      "loss": 0.4658,
      "step": 215
    },
    {
      "epoch": 1.728,
      "grad_norm": 0.4785975515842438,
      "learning_rate": 4.557772514802564e-06,
      "loss": 0.4248,
      "step": 216
    },
    {
      "epoch": 1.736,
      "grad_norm": 0.5019366145133972,
      "learning_rate": 4.511363904173366e-06,
      "loss": 0.4482,
      "step": 217
    },
    {
      "epoch": 1.744,
      "grad_norm": 0.4578736126422882,
      "learning_rate": 4.46499775768523e-06,
      "loss": 0.4391,
      "step": 218
    },
    {
      "epoch": 1.752,
      "grad_norm": 0.46869972348213196,
      "learning_rate": 4.418678104714214e-06,
      "loss": 0.4087,
      "step": 219
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.480589359998703,
      "learning_rate": 4.372408970595931e-06,
      "loss": 0.418,
      "step": 220
    },
    {
      "epoch": 1.768,
      "grad_norm": 0.5114148855209351,
      "learning_rate": 4.326194376275729e-06,
      "loss": 0.4009,
      "step": 221
    },
    {
      "epoch": 1.776,
      "grad_norm": 0.4714803099632263,
      "learning_rate": 4.280038337959268e-06,
      "loss": 0.4119,
      "step": 222
    },
    {
      "epoch": 1.784,
      "grad_norm": 0.4820714592933655,
      "learning_rate": 4.2339448667634885e-06,
      "loss": 0.4364,
      "step": 223
    },
    {
      "epoch": 1.792,
      "grad_norm": 0.45575451850891113,
      "learning_rate": 4.187917968368036e-06,
      "loss": 0.3943,
      "step": 224
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.4869477152824402,
      "learning_rate": 4.141961642667152e-06,
      "loss": 0.4447,
      "step": 225
    },
    {
      "epoch": 1.808,
      "grad_norm": 0.4546818435192108,
      "learning_rate": 4.09607988342207e-06,
      "loss": 0.4397,
      "step": 226
    },
    {
      "epoch": 1.8159999999999998,
      "grad_norm": 0.5653818249702454,
      "learning_rate": 4.0502766779139485e-06,
      "loss": 0.3881,
      "step": 227
    },
    {
      "epoch": 1.8239999999999998,
      "grad_norm": 0.484779953956604,
      "learning_rate": 4.0045560065973535e-06,
      "loss": 0.4673,
      "step": 228
    },
    {
      "epoch": 1.8319999999999999,
      "grad_norm": 0.48289933800697327,
      "learning_rate": 3.958921842754351e-06,
      "loss": 0.4156,
      "step": 229
    },
    {
      "epoch": 1.8399999999999999,
      "grad_norm": 0.43402060866355896,
      "learning_rate": 3.913378152149214e-06,
      "loss": 0.4044,
      "step": 230
    },
    {
      "epoch": 1.8479999999999999,
      "grad_norm": 0.47271519899368286,
      "learning_rate": 3.86792889268378e-06,
      "loss": 0.4387,
      "step": 231
    },
    {
      "epoch": 1.8559999999999999,
      "grad_norm": 0.5104248523712158,
      "learning_rate": 3.8225780140535025e-06,
      "loss": 0.3976,
      "step": 232
    },
    {
      "epoch": 1.8639999999999999,
      "grad_norm": 0.44085070490837097,
      "learning_rate": 3.777329457404202e-06,
      "loss": 0.4062,
      "step": 233
    },
    {
      "epoch": 1.8719999999999999,
      "grad_norm": 0.5098981857299805,
      "learning_rate": 3.7321871549895715e-06,
      "loss": 0.4458,
      "step": 234
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.42704570293426514,
      "learning_rate": 3.68715502982945e-06,
      "loss": 0.3947,
      "step": 235
    },
    {
      "epoch": 1.888,
      "grad_norm": 0.46332859992980957,
      "learning_rate": 3.6422369953688973e-06,
      "loss": 0.4294,
      "step": 236
    },
    {
      "epoch": 1.896,
      "grad_norm": 0.4884559214115143,
      "learning_rate": 3.5974369551381023e-06,
      "loss": 0.4498,
      "step": 237
    },
    {
      "epoch": 1.904,
      "grad_norm": 0.5290793180465698,
      "learning_rate": 3.5527588024131542e-06,
      "loss": 0.4312,
      "step": 238
    },
    {
      "epoch": 1.912,
      "grad_norm": 0.48744434118270874,
      "learning_rate": 3.5082064198777e-06,
      "loss": 0.4772,
      "step": 239
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.4615407884120941,
      "learning_rate": 3.463783679285535e-06,
      "loss": 0.4101,
      "step": 240
    },
    {
      "epoch": 1.928,
      "grad_norm": 0.49090084433555603,
      "learning_rate": 3.4194944411241213e-06,
      "loss": 0.4222,
      "step": 241
    },
    {
      "epoch": 1.936,
      "grad_norm": 0.4603979289531708,
      "learning_rate": 3.3753425542791106e-06,
      "loss": 0.4322,
      "step": 242
    },
    {
      "epoch": 1.944,
      "grad_norm": 0.4630926847457886,
      "learning_rate": 3.3313318556998523e-06,
      "loss": 0.4553,
      "step": 243
    },
    {
      "epoch": 1.952,
      "grad_norm": 0.5014007091522217,
      "learning_rate": 3.2874661700659586e-06,
      "loss": 0.4541,
      "step": 244
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.5217999815940857,
      "learning_rate": 3.2437493094549223e-06,
      "loss": 0.4164,
      "step": 245
    },
    {
      "epoch": 1.968,
      "grad_norm": 0.5072699785232544,
      "learning_rate": 3.200185073010831e-06,
      "loss": 0.4151,
      "step": 246
    },
    {
      "epoch": 1.976,
      "grad_norm": 0.4602278172969818,
      "learning_rate": 3.1567772466142156e-06,
      "loss": 0.3739,
      "step": 247
    },
    {
      "epoch": 1.984,
      "grad_norm": 0.5061410069465637,
      "learning_rate": 3.1135296025530426e-06,
      "loss": 0.4892,
      "step": 248
    },
    {
      "epoch": 1.992,
      "grad_norm": 0.5533497333526611,
      "learning_rate": 3.070445899194885e-06,
      "loss": 0.4132,
      "step": 249
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.52688068151474,
      "learning_rate": 3.0275298806603102e-06,
      "loss": 0.4463,
      "step": 250
    },
    {
      "epoch": 2.008,
      "grad_norm": 0.5126904249191284,
      "learning_rate": 2.984785276497507e-06,
      "loss": 0.3639,
      "step": 251
    },
    {
      "epoch": 2.016,
      "grad_norm": 0.5104655623435974,
      "learning_rate": 2.9422158013581658e-06,
      "loss": 0.3535,
      "step": 252
    },
    {
      "epoch": 2.024,
      "grad_norm": 0.5136793255805969,
      "learning_rate": 2.899825154674674e-06,
      "loss": 0.3835,
      "step": 253
    },
    {
      "epoch": 2.032,
      "grad_norm": 0.45347532629966736,
      "learning_rate": 2.8576170203386144e-06,
      "loss": 0.3412,
      "step": 254
    },
    {
      "epoch": 2.04,
      "grad_norm": 0.4745730757713318,
      "learning_rate": 2.8155950663806234e-06,
      "loss": 0.3863,
      "step": 255
    },
    {
      "epoch": 2.048,
      "grad_norm": 0.5067371129989624,
      "learning_rate": 2.7737629446516325e-06,
      "loss": 0.3514,
      "step": 256
    },
    {
      "epoch": 2.056,
      "grad_norm": 0.45462149381637573,
      "learning_rate": 2.732124290505501e-06,
      "loss": 0.3528,
      "step": 257
    },
    {
      "epoch": 2.064,
      "grad_norm": 0.46529221534729004,
      "learning_rate": 2.6906827224831024e-06,
      "loss": 0.3296,
      "step": 258
    },
    {
      "epoch": 2.072,
      "grad_norm": 0.5160210728645325,
      "learning_rate": 2.6494418419978485e-06,
      "loss": 0.3416,
      "step": 259
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.505672037601471,
      "learning_rate": 2.608405233022724e-06,
      "loss": 0.3793,
      "step": 260
    },
    {
      "epoch": 2.088,
      "grad_norm": 0.44298309087753296,
      "learning_rate": 2.5675764617788233e-06,
      "loss": 0.3298,
      "step": 261
    },
    {
      "epoch": 2.096,
      "grad_norm": 0.4680778682231903,
      "learning_rate": 2.526959076425434e-06,
      "loss": 0.3427,
      "step": 262
    },
    {
      "epoch": 2.104,
      "grad_norm": 0.4816036820411682,
      "learning_rate": 2.4865566067516896e-06,
      "loss": 0.3709,
      "step": 263
    },
    {
      "epoch": 2.112,
      "grad_norm": 0.47969356179237366,
      "learning_rate": 2.4463725638698182e-06,
      "loss": 0.3255,
      "step": 264
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.5166645646095276,
      "learning_rate": 2.406410439910017e-06,
      "loss": 0.3928,
      "step": 265
    },
    {
      "epoch": 2.128,
      "grad_norm": 0.523476243019104,
      "learning_rate": 2.366673707716973e-06,
      "loss": 0.3497,
      "step": 266
    },
    {
      "epoch": 2.136,
      "grad_norm": 0.5414285063743591,
      "learning_rate": 2.327165820548059e-06,
      "loss": 0.3653,
      "step": 267
    },
    {
      "epoch": 2.144,
      "grad_norm": 0.4763210713863373,
      "learning_rate": 2.287890211773238e-06,
      "loss": 0.3628,
      "step": 268
    },
    {
      "epoch": 2.152,
      "grad_norm": 0.5000659227371216,
      "learning_rate": 2.2488502945766893e-06,
      "loss": 0.3681,
      "step": 269
    },
    {
      "epoch": 2.16,
      "grad_norm": 0.48179784417152405,
      "learning_rate": 2.210049461660189e-06,
      "loss": 0.3542,
      "step": 270
    },
    {
      "epoch": 2.168,
      "grad_norm": 0.4680255651473999,
      "learning_rate": 2.1714910849482777e-06,
      "loss": 0.3618,
      "step": 271
    },
    {
      "epoch": 2.176,
      "grad_norm": 0.4781685173511505,
      "learning_rate": 2.1331785152952243e-06,
      "loss": 0.3533,
      "step": 272
    },
    {
      "epoch": 2.184,
      "grad_norm": 0.508620023727417,
      "learning_rate": 2.0951150821938278e-06,
      "loss": 0.3761,
      "step": 273
    },
    {
      "epoch": 2.192,
      "grad_norm": 0.48846498131752014,
      "learning_rate": 2.0573040934860717e-06,
      "loss": 0.3776,
      "step": 274
    },
    {
      "epoch": 2.2,
      "grad_norm": 0.48690298199653625,
      "learning_rate": 2.0197488350756618e-06,
      "loss": 0.326,
      "step": 275
    },
    {
      "epoch": 2.208,
      "grad_norm": 0.5058321356773376,
      "learning_rate": 1.98245257064247e-06,
      "loss": 0.3596,
      "step": 276
    },
    {
      "epoch": 2.216,
      "grad_norm": 0.5168304443359375,
      "learning_rate": 1.945418541358911e-06,
      "loss": 0.3443,
      "step": 277
    },
    {
      "epoch": 2.224,
      "grad_norm": 0.4696882367134094,
      "learning_rate": 1.9086499656082685e-06,
      "loss": 0.3469,
      "step": 278
    },
    {
      "epoch": 2.232,
      "grad_norm": 0.46734338998794556,
      "learning_rate": 1.872150038705015e-06,
      "loss": 0.3494,
      "step": 279
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.4970659017562866,
      "learning_rate": 1.835921932617119e-06,
      "loss": 0.3614,
      "step": 280
    },
    {
      "epoch": 2.248,
      "grad_norm": 0.4725131094455719,
      "learning_rate": 1.7999687956903955e-06,
      "loss": 0.3678,
      "step": 281
    },
    {
      "epoch": 2.2560000000000002,
      "grad_norm": 0.48716068267822266,
      "learning_rate": 1.7642937523749038e-06,
      "loss": 0.3514,
      "step": 282
    },
    {
      "epoch": 2.2640000000000002,
      "grad_norm": 0.4228149950504303,
      "learning_rate": 1.7288999029534177e-06,
      "loss": 0.3349,
      "step": 283
    },
    {
      "epoch": 2.2720000000000002,
      "grad_norm": 0.47141751646995544,
      "learning_rate": 1.6937903232720076e-06,
      "loss": 0.3655,
      "step": 284
    },
    {
      "epoch": 2.2800000000000002,
      "grad_norm": 0.44692301750183105,
      "learning_rate": 1.6589680644727347e-06,
      "loss": 0.3446,
      "step": 285
    },
    {
      "epoch": 2.288,
      "grad_norm": 0.46537840366363525,
      "learning_rate": 1.6244361527284953e-06,
      "loss": 0.3473,
      "step": 286
    },
    {
      "epoch": 2.296,
      "grad_norm": 0.4578072428703308,
      "learning_rate": 1.5901975889800387e-06,
      "loss": 0.3614,
      "step": 287
    },
    {
      "epoch": 2.304,
      "grad_norm": 0.4702107012271881,
      "learning_rate": 1.556255348675174e-06,
      "loss": 0.3417,
      "step": 288
    },
    {
      "epoch": 2.312,
      "grad_norm": 0.49962151050567627,
      "learning_rate": 1.522612381510195e-06,
      "loss": 0.348,
      "step": 289
    },
    {
      "epoch": 2.32,
      "grad_norm": 0.514380156993866,
      "learning_rate": 1.489271611173538e-06,
      "loss": 0.3587,
      "step": 290
    },
    {
      "epoch": 2.328,
      "grad_norm": 0.4479122459888458,
      "learning_rate": 1.4562359350917054e-06,
      "loss": 0.3479,
      "step": 291
    },
    {
      "epoch": 2.336,
      "grad_norm": 0.48699259757995605,
      "learning_rate": 1.423508224177474e-06,
      "loss": 0.3678,
      "step": 292
    },
    {
      "epoch": 2.344,
      "grad_norm": 0.46162307262420654,
      "learning_rate": 1.3910913225803946e-06,
      "loss": 0.3321,
      "step": 293
    },
    {
      "epoch": 2.352,
      "grad_norm": 0.4586437940597534,
      "learning_rate": 1.35898804743963e-06,
      "loss": 0.3748,
      "step": 294
    },
    {
      "epoch": 2.36,
      "grad_norm": 0.479818195104599,
      "learning_rate": 1.3272011886391368e-06,
      "loss": 0.3489,
      "step": 295
    },
    {
      "epoch": 2.368,
      "grad_norm": 0.44310837984085083,
      "learning_rate": 1.295733508565213e-06,
      "loss": 0.3427,
      "step": 296
    },
    {
      "epoch": 2.376,
      "grad_norm": 0.4900333881378174,
      "learning_rate": 1.2645877418664394e-06,
      "loss": 0.3484,
      "step": 297
    },
    {
      "epoch": 2.384,
      "grad_norm": 0.6010957360267639,
      "learning_rate": 1.2337665952160266e-06,
      "loss": 0.3486,
      "step": 298
    },
    {
      "epoch": 2.392,
      "grad_norm": 0.44970938563346863,
      "learning_rate": 1.2032727470765982e-06,
      "loss": 0.3345,
      "step": 299
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.46644240617752075,
      "learning_rate": 1.1731088474674235e-06,
      "loss": 0.3695,
      "step": 300
    },
    {
      "epoch": 2.408,
      "grad_norm": 0.4730045199394226,
      "learning_rate": 1.1432775177341165e-06,
      "loss": 0.3605,
      "step": 301
    },
    {
      "epoch": 2.416,
      "grad_norm": 0.4608900845050812,
      "learning_rate": 1.11378135032084e-06,
      "loss": 0.3328,
      "step": 302
    },
    {
      "epoch": 2.424,
      "grad_norm": 0.47910159826278687,
      "learning_rate": 1.08462290854501e-06,
      "loss": 0.3592,
      "step": 303
    },
    {
      "epoch": 2.432,
      "grad_norm": 0.5008491277694702,
      "learning_rate": 1.0558047263745297e-06,
      "loss": 0.3653,
      "step": 304
    },
    {
      "epoch": 2.44,
      "grad_norm": 0.45120254158973694,
      "learning_rate": 1.0273293082075914e-06,
      "loss": 0.3317,
      "step": 305
    },
    {
      "epoch": 2.448,
      "grad_norm": 0.44771674275398254,
      "learning_rate": 9.991991286550207e-07,
      "loss": 0.3417,
      "step": 306
    },
    {
      "epoch": 2.456,
      "grad_norm": 0.47719165682792664,
      "learning_rate": 9.71416632325235e-07,
      "loss": 0.3634,
      "step": 307
    },
    {
      "epoch": 2.464,
      "grad_norm": 0.5096871256828308,
      "learning_rate": 9.439842336117954e-07,
      "loss": 0.3428,
      "step": 308
    },
    {
      "epoch": 2.472,
      "grad_norm": 0.43439558148384094,
      "learning_rate": 9.169043164835867e-07,
      "loss": 0.3336,
      "step": 309
    },
    {
      "epoch": 2.48,
      "grad_norm": 0.4950801730155945,
      "learning_rate": 8.901792342776439e-07,
      "loss": 0.4171,
      "step": 310
    },
    {
      "epoch": 2.488,
      "grad_norm": 0.4548616409301758,
      "learning_rate": 8.638113094946382e-07,
      "loss": 0.3123,
      "step": 311
    },
    {
      "epoch": 2.496,
      "grad_norm": 0.5122199058532715,
      "learning_rate": 8.378028335970451e-07,
      "loss": 0.3784,
      "step": 312
    },
    {
      "epoch": 2.504,
      "grad_norm": 0.452438622713089,
      "learning_rate": 8.121560668100065e-07,
      "loss": 0.3631,
      "step": 313
    },
    {
      "epoch": 2.512,
      "grad_norm": 0.43378862738609314,
      "learning_rate": 7.868732379249122e-07,
      "loss": 0.3319,
      "step": 314
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.4786088466644287,
      "learning_rate": 7.619565441057075e-07,
      "loss": 0.3517,
      "step": 315
    },
    {
      "epoch": 2.528,
      "grad_norm": 0.5133551955223083,
      "learning_rate": 7.37408150697953e-07,
      "loss": 0.3702,
      "step": 316
    },
    {
      "epoch": 2.536,
      "grad_norm": 0.4935527741909027,
      "learning_rate": 7.132301910406503e-07,
      "loss": 0.3759,
      "step": 317
    },
    {
      "epoch": 2.544,
      "grad_norm": 0.4293310344219208,
      "learning_rate": 6.894247662808456e-07,
      "loss": 0.322,
      "step": 318
    },
    {
      "epoch": 2.552,
      "grad_norm": 0.46054044365882874,
      "learning_rate": 6.659939451910341e-07,
      "loss": 0.3343,
      "step": 319
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.5166534185409546,
      "learning_rate": 6.429397639893758e-07,
      "loss": 0.3813,
      "step": 320
    },
    {
      "epoch": 2.568,
      "grad_norm": 0.4581873118877411,
      "learning_rate": 6.202642261627411e-07,
      "loss": 0.368,
      "step": 321
    },
    {
      "epoch": 2.576,
      "grad_norm": 0.5063138008117676,
      "learning_rate": 5.979693022926025e-07,
      "loss": 0.4142,
      "step": 322
    },
    {
      "epoch": 2.584,
      "grad_norm": 0.47579020261764526,
      "learning_rate": 5.760569298837825e-07,
      "loss": 0.3717,
      "step": 323
    },
    {
      "epoch": 2.592,
      "grad_norm": 0.4639912545681,
      "learning_rate": 5.54529013196079e-07,
      "loss": 0.352,
      "step": 324
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.4613383710384369,
      "learning_rate": 5.333874230787772e-07,
      "loss": 0.3661,
      "step": 325
    },
    {
      "epoch": 2.608,
      "grad_norm": 0.42720070481300354,
      "learning_rate": 5.126339968080696e-07,
      "loss": 0.3616,
      "step": 326
    },
    {
      "epoch": 2.616,
      "grad_norm": 0.4943825304508209,
      "learning_rate": 4.922705379273862e-07,
      "loss": 0.3645,
      "step": 327
    },
    {
      "epoch": 2.624,
      "grad_norm": 0.4619237780570984,
      "learning_rate": 4.7229881609066387e-07,
      "loss": 0.3307,
      "step": 328
    },
    {
      "epoch": 2.632,
      "grad_norm": 0.47587689757347107,
      "learning_rate": 4.5272056690855494e-07,
      "loss": 0.3624,
      "step": 329
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.45437976717948914,
      "learning_rate": 4.335374917975982e-07,
      "loss": 0.3242,
      "step": 330
    },
    {
      "epoch": 2.648,
      "grad_norm": 0.5222877264022827,
      "learning_rate": 4.147512578323615e-07,
      "loss": 0.3808,
      "step": 331
    },
    {
      "epoch": 2.656,
      "grad_norm": 0.46543389558792114,
      "learning_rate": 3.9636349760056427e-07,
      "loss": 0.3389,
      "step": 332
    },
    {
      "epoch": 2.664,
      "grad_norm": 0.4624124765396118,
      "learning_rate": 3.783758090611983e-07,
      "loss": 0.3336,
      "step": 333
    },
    {
      "epoch": 2.672,
      "grad_norm": 0.45994681119918823,
      "learning_rate": 3.6078975540566716e-07,
      "loss": 0.3482,
      "step": 334
    },
    {
      "epoch": 2.68,
      "grad_norm": 0.45068469643592834,
      "learning_rate": 3.4360686492193263e-07,
      "loss": 0.3077,
      "step": 335
    },
    {
      "epoch": 2.6879999999999997,
      "grad_norm": 0.5126630663871765,
      "learning_rate": 3.268286308617041e-07,
      "loss": 0.3906,
      "step": 336
    },
    {
      "epoch": 2.6959999999999997,
      "grad_norm": 0.4749516546726227,
      "learning_rate": 3.104565113106689e-07,
      "loss": 0.342,
      "step": 337
    },
    {
      "epoch": 2.7039999999999997,
      "grad_norm": 0.43455156683921814,
      "learning_rate": 2.9449192906178205e-07,
      "loss": 0.3491,
      "step": 338
    },
    {
      "epoch": 2.7119999999999997,
      "grad_norm": 0.46082979440689087,
      "learning_rate": 2.789362714916172e-07,
      "loss": 0.3875,
      "step": 339
    },
    {
      "epoch": 2.7199999999999998,
      "grad_norm": 0.44106754660606384,
      "learning_rate": 2.6379089043980064e-07,
      "loss": 0.32,
      "step": 340
    },
    {
      "epoch": 2.7279999999999998,
      "grad_norm": 0.46814629435539246,
      "learning_rate": 2.4905710209153224e-07,
      "loss": 0.3451,
      "step": 341
    },
    {
      "epoch": 2.7359999999999998,
      "grad_norm": 0.5077950358390808,
      "learning_rate": 2.3473618686320477e-07,
      "loss": 0.3861,
      "step": 342
    },
    {
      "epoch": 2.7439999999999998,
      "grad_norm": 0.49315568804740906,
      "learning_rate": 2.208293892911284e-07,
      "loss": 0.3469,
      "step": 343
    },
    {
      "epoch": 2.752,
      "grad_norm": 0.4634096920490265,
      "learning_rate": 2.0733791792338197e-07,
      "loss": 0.3761,
      "step": 344
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.4556006193161011,
      "learning_rate": 1.9426294521477874e-07,
      "loss": 0.3504,
      "step": 345
    },
    {
      "epoch": 2.768,
      "grad_norm": 0.48851218819618225,
      "learning_rate": 1.8160560742498223e-07,
      "loss": 0.3652,
      "step": 346
    },
    {
      "epoch": 2.776,
      "grad_norm": 0.45232683420181274,
      "learning_rate": 1.6936700451975818e-07,
      "loss": 0.3259,
      "step": 347
    },
    {
      "epoch": 2.784,
      "grad_norm": 0.4734852910041809,
      "learning_rate": 1.5754820007538473e-07,
      "loss": 0.3555,
      "step": 348
    },
    {
      "epoch": 2.792,
      "grad_norm": 0.46094757318496704,
      "learning_rate": 1.461502211862237e-07,
      "loss": 0.3451,
      "step": 349
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.45597511529922485,
      "learning_rate": 1.3517405837546404e-07,
      "loss": 0.3481,
      "step": 350
    },
    {
      "epoch": 2.808,
      "grad_norm": 0.44162362813949585,
      "learning_rate": 1.2462066550903818e-07,
      "loss": 0.3324,
      "step": 351
    },
    {
      "epoch": 2.816,
      "grad_norm": 0.4328787326812744,
      "learning_rate": 1.1449095971273305e-07,
      "loss": 0.3304,
      "step": 352
    },
    {
      "epoch": 2.824,
      "grad_norm": 0.472141832113266,
      "learning_rate": 1.0478582129248516e-07,
      "loss": 0.3655,
      "step": 353
    },
    {
      "epoch": 2.832,
      "grad_norm": 0.4979706406593323,
      "learning_rate": 9.550609365787888e-08,
      "loss": 0.3815,
      "step": 354
    },
    {
      "epoch": 2.84,
      "grad_norm": 0.49750280380249023,
      "learning_rate": 8.66525832488535e-08,
      "loss": 0.3656,
      "step": 355
    },
    {
      "epoch": 2.848,
      "grad_norm": 0.46096086502075195,
      "learning_rate": 7.822605946561923e-08,
      "loss": 0.3437,
      "step": 356
    },
    {
      "epoch": 2.856,
      "grad_norm": 0.5005020499229431,
      "learning_rate": 7.022725460179459e-08,
      "loss": 0.3697,
      "step": 357
    },
    {
      "epoch": 2.864,
      "grad_norm": 0.4813578426837921,
      "learning_rate": 6.265686378076729e-08,
      "loss": 0.3595,
      "step": 358
    },
    {
      "epoch": 2.872,
      "grad_norm": 0.4750673472881317,
      "learning_rate": 5.5515544895284324e-08,
      "loss": 0.3706,
      "step": 359
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.48565149307250977,
      "learning_rate": 4.880391855028088e-08,
      "loss": 0.3524,
      "step": 360
    },
    {
      "epoch": 2.888,
      "grad_norm": 0.4671759307384491,
      "learning_rate": 4.252256800894694e-08,
      "loss": 0.3407,
      "step": 361
    },
    {
      "epoch": 2.896,
      "grad_norm": 0.4376312494277954,
      "learning_rate": 3.6672039142039426e-08,
      "loss": 0.3292,
      "step": 362
    },
    {
      "epoch": 2.904,
      "grad_norm": 0.4908435344696045,
      "learning_rate": 3.125284038044407e-08,
      "loss": 0.3429,
      "step": 363
    },
    {
      "epoch": 2.912,
      "grad_norm": 0.44629916548728943,
      "learning_rate": 2.6265442670991293e-08,
      "loss": 0.3428,
      "step": 364
    },
    {
      "epoch": 2.92,
      "grad_norm": 0.465819388628006,
      "learning_rate": 2.1710279435530058e-08,
      "loss": 0.3448,
      "step": 365
    },
    {
      "epoch": 2.928,
      "grad_norm": 0.4819665253162384,
      "learning_rate": 1.7587746533260786e-08,
      "loss": 0.3519,
      "step": 366
    },
    {
      "epoch": 2.936,
      "grad_norm": 0.4784250855445862,
      "learning_rate": 1.3898202226333424e-08,
      "loss": 0.3739,
      "step": 367
    },
    {
      "epoch": 2.944,
      "grad_norm": 0.495050847530365,
      "learning_rate": 1.0641967148716236e-08,
      "loss": 0.3628,
      "step": 368
    },
    {
      "epoch": 2.952,
      "grad_norm": 0.4823502004146576,
      "learning_rate": 7.819324278328099e-09,
      "loss": 0.3411,
      "step": 369
    },
    {
      "epoch": 2.96,
      "grad_norm": 0.41492488980293274,
      "learning_rate": 5.430518912448169e-09,
      "loss": 0.3326,
      "step": 370
    },
    {
      "epoch": 2.968,
      "grad_norm": 0.47112229466438293,
      "learning_rate": 3.4757586464001513e-09,
      "loss": 0.3696,
      "step": 371
    },
    {
      "epoch": 2.976,
      "grad_norm": 0.46793392300605774,
      "learning_rate": 1.9552133555084117e-09,
      "loss": 0.3661,
      "step": 372
    },
    {
      "epoch": 2.984,
      "grad_norm": 0.4385625720024109,
      "learning_rate": 8.690151803386615e-10,
      "loss": 0.3421,
      "step": 373
    },
    {
      "epoch": 2.992,
      "grad_norm": 0.43801286816596985,
      "learning_rate": 2.1725851521103847e-10,
      "loss": 0.3395,
      "step": 374
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.44443294405937195,
      "learning_rate": 0.0,
      "loss": 0.3399,
      "step": 375
    },
    {
      "epoch": 3.0,
      "step": 375,
      "total_flos": 223704105418752.0,
      "train_loss": 0.4401542440255483,
      "train_runtime": 59266.2263,
      "train_samples_per_second": 0.607,
      "train_steps_per_second": 0.006
    }
  ],
  "logging_steps": 1,
  "max_steps": 375,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 223704105418752.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}