{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 438,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00684931506849315,
      "grad_norm": 5.92502498626709,
      "learning_rate": 2.2727272727272729e-07,
      "loss": 0.8592,
      "step": 1
    },
    {
      "epoch": 0.0136986301369863,
      "grad_norm": 5.975358009338379,
      "learning_rate": 4.5454545454545457e-07,
      "loss": 0.8441,
      "step": 2
    },
    {
      "epoch": 0.02054794520547945,
      "grad_norm": 6.038137435913086,
      "learning_rate": 6.818181818181818e-07,
      "loss": 0.8847,
      "step": 3
    },
    {
      "epoch": 0.0273972602739726,
      "grad_norm": 6.023561000823975,
      "learning_rate": 9.090909090909091e-07,
      "loss": 0.8802,
      "step": 4
    },
    {
      "epoch": 0.03424657534246575,
      "grad_norm": 5.958974361419678,
      "learning_rate": 1.1363636363636364e-06,
      "loss": 0.9015,
      "step": 5
    },
    {
      "epoch": 0.0410958904109589,
      "grad_norm": 5.522011756896973,
      "learning_rate": 1.3636363636363636e-06,
      "loss": 0.8771,
      "step": 6
    },
    {
      "epoch": 0.04794520547945205,
      "grad_norm": 5.2923994064331055,
      "learning_rate": 1.590909090909091e-06,
      "loss": 0.8668,
      "step": 7
    },
    {
      "epoch": 0.0547945205479452,
      "grad_norm": 4.468110084533691,
      "learning_rate": 1.8181818181818183e-06,
      "loss": 0.8269,
      "step": 8
    },
    {
      "epoch": 0.06164383561643835,
      "grad_norm": 3.9916787147521973,
      "learning_rate": 2.0454545454545457e-06,
      "loss": 0.7879,
      "step": 9
    },
    {
      "epoch": 0.0684931506849315,
      "grad_norm": 2.3367724418640137,
      "learning_rate": 2.2727272727272728e-06,
      "loss": 0.7631,
      "step": 10
    },
    {
      "epoch": 0.07534246575342465,
      "grad_norm": 2.0509274005889893,
      "learning_rate": 2.5e-06,
      "loss": 0.7555,
      "step": 11
    },
    {
      "epoch": 0.0821917808219178,
      "grad_norm": 1.9280574321746826,
      "learning_rate": 2.7272727272727272e-06,
      "loss": 0.7413,
      "step": 12
    },
    {
      "epoch": 0.08904109589041095,
      "grad_norm": 1.8343186378479004,
      "learning_rate": 2.954545454545455e-06,
      "loss": 0.7588,
      "step": 13
    },
    {
      "epoch": 0.0958904109589041,
      "grad_norm": 3.5027122497558594,
      "learning_rate": 3.181818181818182e-06,
      "loss": 0.7552,
      "step": 14
    },
    {
      "epoch": 0.10273972602739725,
      "grad_norm": 3.766814947128296,
      "learning_rate": 3.409090909090909e-06,
      "loss": 0.7653,
      "step": 15
    },
    {
      "epoch": 0.1095890410958904,
      "grad_norm": 3.9283227920532227,
      "learning_rate": 3.6363636363636366e-06,
      "loss": 0.7895,
      "step": 16
    },
    {
      "epoch": 0.11643835616438356,
      "grad_norm": 3.853430986404419,
      "learning_rate": 3.863636363636364e-06,
      "loss": 0.7636,
      "step": 17
    },
    {
      "epoch": 0.1232876712328767,
      "grad_norm": 3.0980968475341797,
      "learning_rate": 4.0909090909090915e-06,
      "loss": 0.7197,
      "step": 18
    },
    {
      "epoch": 0.13013698630136986,
      "grad_norm": 2.2356159687042236,
      "learning_rate": 4.3181818181818185e-06,
      "loss": 0.6567,
      "step": 19
    },
    {
      "epoch": 0.136986301369863,
      "grad_norm": 2.1344797611236572,
      "learning_rate": 4.5454545454545455e-06,
      "loss": 0.7121,
      "step": 20
    },
    {
      "epoch": 0.14383561643835616,
      "grad_norm": 1.604323148727417,
      "learning_rate": 4.772727272727273e-06,
      "loss": 0.7,
      "step": 21
    },
    {
      "epoch": 0.1506849315068493,
      "grad_norm": 1.3561691045761108,
      "learning_rate": 5e-06,
      "loss": 0.6897,
      "step": 22
    },
    {
      "epoch": 0.15753424657534246,
      "grad_norm": 1.3645248413085938,
      "learning_rate": 5.2272727272727274e-06,
      "loss": 0.668,
      "step": 23
    },
    {
      "epoch": 0.1643835616438356,
      "grad_norm": 1.4382814168930054,
      "learning_rate": 5.4545454545454545e-06,
      "loss": 0.6585,
      "step": 24
    },
    {
      "epoch": 0.17123287671232876,
      "grad_norm": 1.3407988548278809,
      "learning_rate": 5.681818181818183e-06,
      "loss": 0.6639,
      "step": 25
    },
    {
      "epoch": 0.1780821917808219,
      "grad_norm": 1.243182897567749,
      "learning_rate": 5.90909090909091e-06,
      "loss": 0.6519,
      "step": 26
    },
    {
      "epoch": 0.18493150684931506,
      "grad_norm": 0.9758120775222778,
      "learning_rate": 6.136363636363637e-06,
      "loss": 0.6133,
      "step": 27
    },
    {
      "epoch": 0.1917808219178082,
      "grad_norm": 0.9279192686080933,
      "learning_rate": 6.363636363636364e-06,
      "loss": 0.6404,
      "step": 28
    },
    {
      "epoch": 0.19863013698630136,
      "grad_norm": 1.1322438716888428,
      "learning_rate": 6.590909090909091e-06,
      "loss": 0.6385,
      "step": 29
    },
    {
      "epoch": 0.2054794520547945,
      "grad_norm": 1.0254402160644531,
      "learning_rate": 6.818181818181818e-06,
      "loss": 0.6447,
      "step": 30
    },
    {
      "epoch": 0.21232876712328766,
      "grad_norm": 0.7659736275672913,
      "learning_rate": 7.045454545454546e-06,
      "loss": 0.6244,
      "step": 31
    },
    {
      "epoch": 0.2191780821917808,
      "grad_norm": 0.8017905354499817,
      "learning_rate": 7.272727272727273e-06,
      "loss": 0.6137,
      "step": 32
    },
    {
      "epoch": 0.22602739726027396,
      "grad_norm": 0.8775509595870972,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.6211,
      "step": 33
    },
    {
      "epoch": 0.2328767123287671,
      "grad_norm": 0.8729256987571716,
      "learning_rate": 7.727272727272727e-06,
      "loss": 0.6034,
      "step": 34
    },
    {
      "epoch": 0.23972602739726026,
      "grad_norm": 0.6592076420783997,
      "learning_rate": 7.954545454545455e-06,
      "loss": 0.6151,
      "step": 35
    },
    {
      "epoch": 0.2465753424657534,
      "grad_norm": 0.5613800883293152,
      "learning_rate": 8.181818181818183e-06,
      "loss": 0.5855,
      "step": 36
    },
    {
      "epoch": 0.2534246575342466,
      "grad_norm": 0.6701193451881409,
      "learning_rate": 8.40909090909091e-06,
      "loss": 0.6001,
      "step": 37
    },
    {
      "epoch": 0.2602739726027397,
      "grad_norm": 0.708198070526123,
      "learning_rate": 8.636363636363637e-06,
      "loss": 0.5876,
      "step": 38
    },
    {
      "epoch": 0.2671232876712329,
      "grad_norm": 0.5600025653839111,
      "learning_rate": 8.863636363636365e-06,
      "loss": 0.5883,
      "step": 39
    },
    {
      "epoch": 0.273972602739726,
      "grad_norm": 0.5653926134109497,
      "learning_rate": 9.090909090909091e-06,
      "loss": 0.5614,
      "step": 40
    },
    {
      "epoch": 0.2808219178082192,
      "grad_norm": 0.7606906890869141,
      "learning_rate": 9.318181818181819e-06,
      "loss": 0.6156,
      "step": 41
    },
    {
      "epoch": 0.2876712328767123,
      "grad_norm": 0.5844655632972717,
      "learning_rate": 9.545454545454547e-06,
      "loss": 0.586,
      "step": 42
    },
    {
      "epoch": 0.2945205479452055,
      "grad_norm": 0.49894917011260986,
      "learning_rate": 9.772727272727273e-06,
      "loss": 0.5747,
      "step": 43
    },
    {
      "epoch": 0.3013698630136986,
      "grad_norm": 0.5528267621994019,
      "learning_rate": 1e-05,
      "loss": 0.5419,
      "step": 44
    },
    {
      "epoch": 0.3082191780821918,
      "grad_norm": 0.6437212228775024,
      "learning_rate": 9.999841055681184e-06,
      "loss": 0.568,
      "step": 45
    },
    {
      "epoch": 0.3150684931506849,
      "grad_norm": 0.5419601798057556,
      "learning_rate": 9.999364232830053e-06,
      "loss": 0.5715,
      "step": 46
    },
    {
      "epoch": 0.3219178082191781,
      "grad_norm": 0.4963567554950714,
      "learning_rate": 9.99856956176192e-06,
      "loss": 0.5492,
      "step": 47
    },
    {
      "epoch": 0.3287671232876712,
      "grad_norm": 0.7037913203239441,
      "learning_rate": 9.997457093000165e-06,
      "loss": 0.5849,
      "step": 48
    },
    {
      "epoch": 0.3356164383561644,
      "grad_norm": 0.7691269516944885,
      "learning_rate": 9.996026897273024e-06,
      "loss": 0.585,
      "step": 49
    },
    {
      "epoch": 0.3424657534246575,
      "grad_norm": 0.4761511981487274,
      "learning_rate": 9.994279065509094e-06,
      "loss": 0.5813,
      "step": 50
    },
    {
      "epoch": 0.3493150684931507,
      "grad_norm": 0.60807204246521,
      "learning_rate": 9.992213708831542e-06,
      "loss": 0.5576,
      "step": 51
    },
    {
      "epoch": 0.3561643835616438,
      "grad_norm": 0.7713252902030945,
      "learning_rate": 9.989830958551058e-06,
      "loss": 0.5776,
      "step": 52
    },
    {
      "epoch": 0.363013698630137,
      "grad_norm": 0.5777241587638855,
      "learning_rate": 9.987130966157486e-06,
      "loss": 0.5973,
      "step": 53
    },
    {
      "epoch": 0.3698630136986301,
      "grad_norm": 0.4666934907436371,
      "learning_rate": 9.984113903310206e-06,
      "loss": 0.5383,
      "step": 54
    },
    {
      "epoch": 0.3767123287671233,
      "grad_norm": 0.6798987984657288,
      "learning_rate": 9.98077996182722e-06,
      "loss": 0.5371,
      "step": 55
    },
    {
      "epoch": 0.3835616438356164,
      "grad_norm": 0.6998324394226074,
      "learning_rate": 9.977129353672951e-06,
      "loss": 0.5708,
      "step": 56
    },
    {
      "epoch": 0.3904109589041096,
      "grad_norm": 0.46906980872154236,
      "learning_rate": 9.973162310944769e-06,
      "loss": 0.5777,
      "step": 57
    },
    {
      "epoch": 0.3972602739726027,
      "grad_norm": 0.595076858997345,
      "learning_rate": 9.968879085858234e-06,
      "loss": 0.5366,
      "step": 58
    },
    {
      "epoch": 0.4041095890410959,
      "grad_norm": 0.6636199355125427,
      "learning_rate": 9.964279950731066e-06,
      "loss": 0.5822,
      "step": 59
    },
    {
      "epoch": 0.410958904109589,
      "grad_norm": 0.4243064224720001,
      "learning_rate": 9.959365197965824e-06,
      "loss": 0.5814,
      "step": 60
    },
    {
      "epoch": 0.4178082191780822,
      "grad_norm": 0.49584805965423584,
      "learning_rate": 9.954135140031322e-06,
      "loss": 0.5594,
      "step": 61
    },
    {
      "epoch": 0.4246575342465753,
      "grad_norm": 0.6090409755706787,
      "learning_rate": 9.948590109442755e-06,
      "loss": 0.5516,
      "step": 62
    },
    {
      "epoch": 0.4315068493150685,
      "grad_norm": 0.5092521905899048,
      "learning_rate": 9.942730458740568e-06,
      "loss": 0.5335,
      "step": 63
    },
    {
      "epoch": 0.4383561643835616,
      "grad_norm": 0.4809435307979584,
      "learning_rate": 9.936556560468037e-06,
      "loss": 0.5421,
      "step": 64
    },
    {
      "epoch": 0.4452054794520548,
      "grad_norm": 0.545276939868927,
      "learning_rate": 9.930068807147585e-06,
      "loss": 0.5375,
      "step": 65
    },
    {
      "epoch": 0.4520547945205479,
      "grad_norm": 0.4461447596549988,
      "learning_rate": 9.923267611255824e-06,
      "loss": 0.5605,
      "step": 66
    },
    {
      "epoch": 0.4589041095890411,
      "grad_norm": 0.4646362364292145,
      "learning_rate": 9.916153405197333e-06,
      "loss": 0.5229,
      "step": 67
    },
    {
      "epoch": 0.4657534246575342,
      "grad_norm": 0.40728962421417236,
      "learning_rate": 9.908726641277167e-06,
      "loss": 0.5364,
      "step": 68
    },
    {
      "epoch": 0.4726027397260274,
      "grad_norm": 0.4487316906452179,
      "learning_rate": 9.9009877916721e-06,
      "loss": 0.5351,
      "step": 69
    },
    {
      "epoch": 0.4794520547945205,
      "grad_norm": 0.4371180832386017,
      "learning_rate": 9.8929373484006e-06,
      "loss": 0.543,
      "step": 70
    },
    {
      "epoch": 0.4863013698630137,
      "grad_norm": 0.40577659010887146,
      "learning_rate": 9.884575823291561e-06,
      "loss": 0.5959,
      "step": 71
    },
    {
      "epoch": 0.4931506849315068,
      "grad_norm": 0.45803892612457275,
      "learning_rate": 9.875903747951742e-06,
      "loss": 0.569,
      "step": 72
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.41057565808296204,
      "learning_rate": 9.866921673731991e-06,
      "loss": 0.5376,
      "step": 73
    },
    {
      "epoch": 0.5068493150684932,
      "grad_norm": 0.40417274832725525,
      "learning_rate": 9.857630171692175e-06,
      "loss": 0.5356,
      "step": 74
    },
    {
      "epoch": 0.5136986301369864,
      "grad_norm": 0.4480303227901459,
      "learning_rate": 9.848029832564875e-06,
      "loss": 0.5151,
      "step": 75
    },
    {
      "epoch": 0.5205479452054794,
      "grad_norm": 0.4397015869617462,
      "learning_rate": 9.83812126671784e-06,
      "loss": 0.5274,
      "step": 76
    },
    {
      "epoch": 0.5273972602739726,
      "grad_norm": 0.4616395831108093,
      "learning_rate": 9.827905104115167e-06,
      "loss": 0.5365,
      "step": 77
    },
    {
      "epoch": 0.5342465753424658,
      "grad_norm": 0.40124353766441345,
      "learning_rate": 9.81738199427726e-06,
      "loss": 0.5463,
      "step": 78
    },
    {
      "epoch": 0.541095890410959,
      "grad_norm": 0.4343346059322357,
      "learning_rate": 9.80655260623953e-06,
      "loss": 0.5346,
      "step": 79
    },
    {
      "epoch": 0.547945205479452,
      "grad_norm": 0.44946178793907166,
      "learning_rate": 9.795417628509857e-06,
      "loss": 0.5553,
      "step": 80
    },
    {
      "epoch": 0.5547945205479452,
      "grad_norm": 0.416344553232193,
      "learning_rate": 9.783977769024821e-06,
      "loss": 0.5467,
      "step": 81
    },
    {
      "epoch": 0.5616438356164384,
      "grad_norm": 0.41620340943336487,
      "learning_rate": 9.772233755104695e-06,
      "loss": 0.5365,
      "step": 82
    },
    {
      "epoch": 0.5684931506849316,
      "grad_norm": 0.43192237615585327,
      "learning_rate": 9.76018633340719e-06,
      "loss": 0.5366,
      "step": 83
    },
    {
      "epoch": 0.5753424657534246,
      "grad_norm": 0.4866390526294708,
      "learning_rate": 9.747836269880005e-06,
      "loss": 0.5549,
      "step": 84
    },
    {
      "epoch": 0.5821917808219178,
      "grad_norm": 0.5592001676559448,
      "learning_rate": 9.73518434971211e-06,
      "loss": 0.5605,
      "step": 85
    },
    {
      "epoch": 0.589041095890411,
      "grad_norm": 0.4245613217353821,
      "learning_rate": 9.722231377283841e-06,
      "loss": 0.5304,
      "step": 86
    },
    {
      "epoch": 0.5958904109589042,
      "grad_norm": 0.434048056602478,
      "learning_rate": 9.70897817611575e-06,
      "loss": 0.5392,
      "step": 87
    },
    {
      "epoch": 0.6027397260273972,
      "grad_norm": 0.36968135833740234,
      "learning_rate": 9.695425588816248e-06,
      "loss": 0.536,
      "step": 88
    },
    {
      "epoch": 0.6095890410958904,
      "grad_norm": 0.4373226463794708,
      "learning_rate": 9.681574477028039e-06,
      "loss": 0.531,
      "step": 89
    },
    {
      "epoch": 0.6164383561643836,
      "grad_norm": 0.3857197165489197,
      "learning_rate": 9.667425721373333e-06,
      "loss": 0.5239,
      "step": 90
    },
    {
      "epoch": 0.6232876712328768,
      "grad_norm": 0.4224616587162018,
      "learning_rate": 9.65298022139786e-06,
      "loss": 0.5462,
      "step": 91
    },
    {
      "epoch": 0.6301369863013698,
      "grad_norm": 0.4477905035018921,
      "learning_rate": 9.638238895513687e-06,
      "loss": 0.496,
      "step": 92
    },
    {
      "epoch": 0.636986301369863,
      "grad_norm": 0.38324639201164246,
      "learning_rate": 9.623202680940811e-06,
      "loss": 0.5521,
      "step": 93
    },
    {
      "epoch": 0.6438356164383562,
      "grad_norm": 0.5522112250328064,
      "learning_rate": 9.607872533647584e-06,
      "loss": 0.5576,
      "step": 94
    },
    {
      "epoch": 0.6506849315068494,
      "grad_norm": 0.45175299048423767,
      "learning_rate": 9.592249428289935e-06,
      "loss": 0.5461,
      "step": 95
    },
    {
      "epoch": 0.6575342465753424,
      "grad_norm": 0.5615484714508057,
      "learning_rate": 9.5763343581494e-06,
      "loss": 0.5065,
      "step": 96
    },
    {
      "epoch": 0.6643835616438356,
      "grad_norm": 0.46789291501045227,
      "learning_rate": 9.560128335069971e-06,
      "loss": 0.5649,
      "step": 97
    },
    {
      "epoch": 0.6712328767123288,
      "grad_norm": 0.45602020621299744,
      "learning_rate": 9.543632389393767e-06,
      "loss": 0.5235,
      "step": 98
    },
    {
      "epoch": 0.678082191780822,
      "grad_norm": 0.5114532113075256,
      "learning_rate": 9.526847569895529e-06,
      "loss": 0.5454,
      "step": 99
    },
    {
      "epoch": 0.684931506849315,
      "grad_norm": 0.39697399735450745,
      "learning_rate": 9.50977494371594e-06,
      "loss": 0.5445,
      "step": 100
    },
    {
      "epoch": 0.6917808219178082,
      "grad_norm": 0.47998112440109253,
      "learning_rate": 9.49241559629377e-06,
      "loss": 0.5241,
      "step": 101
    },
    {
      "epoch": 0.6986301369863014,
      "grad_norm": 0.4278751313686371,
      "learning_rate": 9.474770631296882e-06,
      "loss": 0.512,
      "step": 102
    },
    {
      "epoch": 0.7054794520547946,
      "grad_norm": 0.5100277662277222,
      "learning_rate": 9.456841170552054e-06,
      "loss": 0.5387,
      "step": 103
    },
    {
      "epoch": 0.7123287671232876,
      "grad_norm": 0.41954633593559265,
      "learning_rate": 9.438628353973654e-06,
      "loss": 0.5454,
      "step": 104
    },
    {
      "epoch": 0.7191780821917808,
      "grad_norm": 0.42173242568969727,
      "learning_rate": 9.420133339491171e-06,
      "loss": 0.4957,
      "step": 105
    },
    {
      "epoch": 0.726027397260274,
      "grad_norm": 0.5260700583457947,
      "learning_rate": 9.4013573029756e-06,
      "loss": 0.5745,
      "step": 106
    },
    {
      "epoch": 0.7328767123287672,
      "grad_norm": 0.3974344730377197,
      "learning_rate": 9.382301438164673e-06,
      "loss": 0.5289,
      "step": 107
    },
    {
      "epoch": 0.7397260273972602,
      "grad_norm": 0.4914883077144623,
      "learning_rate": 9.36296695658697e-06,
      "loss": 0.5337,
      "step": 108
    },
    {
      "epoch": 0.7465753424657534,
      "grad_norm": 0.454052597284317,
      "learning_rate": 9.343355087484893e-06,
      "loss": 0.5377,
      "step": 109
    },
    {
      "epoch": 0.7534246575342466,
      "grad_norm": 0.42050087451934814,
      "learning_rate": 9.323467077736513e-06,
      "loss": 0.5216,
      "step": 110
    },
    {
      "epoch": 0.7602739726027398,
      "grad_norm": 0.506266713142395,
      "learning_rate": 9.303304191776291e-06,
      "loss": 0.502,
      "step": 111
    },
    {
      "epoch": 0.7671232876712328,
      "grad_norm": 0.44245612621307373,
      "learning_rate": 9.282867711514703e-06,
      "loss": 0.548,
      "step": 112
    },
    {
      "epoch": 0.773972602739726,
      "grad_norm": 0.4844461679458618,
      "learning_rate": 9.262158936256717e-06,
      "loss": 0.5564,
      "step": 113
    },
    {
      "epoch": 0.7808219178082192,
      "grad_norm": 0.44854727387428284,
      "learning_rate": 9.241179182619207e-06,
      "loss": 0.495,
      "step": 114
    },
    {
      "epoch": 0.7876712328767124,
      "grad_norm": 0.44603100419044495,
      "learning_rate": 9.219929784447232e-06,
      "loss": 0.5371,
      "step": 115
    },
    {
      "epoch": 0.7945205479452054,
      "grad_norm": 0.4621168076992035,
      "learning_rate": 9.19841209272924e-06,
      "loss": 0.5636,
      "step": 116
    },
    {
      "epoch": 0.8013698630136986,
      "grad_norm": 0.36872074007987976,
      "learning_rate": 9.176627475511171e-06,
      "loss": 0.5362,
      "step": 117
    },
    {
      "epoch": 0.8082191780821918,
      "grad_norm": 0.442421019077301,
      "learning_rate": 9.154577317809483e-06,
      "loss": 0.5426,
      "step": 118
    },
    {
      "epoch": 0.815068493150685,
      "grad_norm": 0.47269773483276367,
      "learning_rate": 9.132263021523096e-06,
      "loss": 0.5525,
      "step": 119
    },
    {
      "epoch": 0.821917808219178,
      "grad_norm": 0.39111027121543884,
      "learning_rate": 9.109686005344258e-06,
      "loss": 0.5411,
      "step": 120
    },
    {
      "epoch": 0.8287671232876712,
      "grad_norm": 0.4781222641468048,
      "learning_rate": 9.086847704668352e-06,
      "loss": 0.5368,
      "step": 121
    },
    {
      "epoch": 0.8356164383561644,
      "grad_norm": 0.49406808614730835,
      "learning_rate": 9.063749571502633e-06,
      "loss": 0.5375,
      "step": 122
    },
    {
      "epoch": 0.8424657534246576,
      "grad_norm": 0.4259958565235138,
      "learning_rate": 9.040393074373921e-06,
      "loss": 0.5504,
      "step": 123
    },
    {
      "epoch": 0.8493150684931506,
      "grad_norm": 0.41925352811813354,
      "learning_rate": 9.016779698235227e-06,
      "loss": 0.5184,
      "step": 124
    },
    {
      "epoch": 0.8561643835616438,
      "grad_norm": 0.5044152736663818,
      "learning_rate": 8.992910944371343e-06,
      "loss": 0.5405,
      "step": 125
    },
    {
      "epoch": 0.863013698630137,
      "grad_norm": 0.4105503261089325,
      "learning_rate": 8.9687883303034e-06,
      "loss": 0.5234,
      "step": 126
    },
    {
      "epoch": 0.8698630136986302,
      "grad_norm": 0.46981051564216614,
      "learning_rate": 8.94441338969238e-06,
      "loss": 0.5467,
      "step": 127
    },
    {
      "epoch": 0.8767123287671232,
      "grad_norm": 0.4437629282474518,
      "learning_rate": 8.919787672241619e-06,
      "loss": 0.5259,
      "step": 128
    },
    {
      "epoch": 0.8835616438356164,
      "grad_norm": 0.41010263562202454,
      "learning_rate": 8.894912743598269e-06,
      "loss": 0.557,
      "step": 129
    },
    {
      "epoch": 0.8904109589041096,
      "grad_norm": 0.4265572130680084,
      "learning_rate": 8.869790185253766e-06,
      "loss": 0.4936,
      "step": 130
    },
    {
      "epoch": 0.8972602739726028,
      "grad_norm": 0.516581118106842,
      "learning_rate": 8.84442159444328e-06,
      "loss": 0.5336,
      "step": 131
    },
    {
      "epoch": 0.9041095890410958,
      "grad_norm": 0.4106625020503998,
      "learning_rate": 8.818808584044163e-06,
      "loss": 0.5374,
      "step": 132
    },
    {
      "epoch": 0.910958904109589,
      "grad_norm": 0.397277295589447,
      "learning_rate": 8.792952782473415e-06,
      "loss": 0.5485,
      "step": 133
    },
    {
      "epoch": 0.9178082191780822,
      "grad_norm": 0.40992820262908936,
      "learning_rate": 8.76685583358414e-06,
      "loss": 0.5257,
      "step": 134
    },
    {
      "epoch": 0.9246575342465754,
      "grad_norm": 0.4277738332748413,
      "learning_rate": 8.740519396561045e-06,
      "loss": 0.5102,
      "step": 135
    },
    {
      "epoch": 0.9315068493150684,
      "grad_norm": 0.40470704436302185,
      "learning_rate": 8.713945145814948e-06,
      "loss": 0.5169,
      "step": 136
    },
    {
      "epoch": 0.9383561643835616,
      "grad_norm": 0.4239768385887146,
      "learning_rate": 8.68713477087632e-06,
      "loss": 0.5474,
      "step": 137
    },
    {
      "epoch": 0.9452054794520548,
      "grad_norm": 0.4196901023387909,
      "learning_rate": 8.660089976287875e-06,
      "loss": 0.511,
      "step": 138
    },
    {
      "epoch": 0.952054794520548,
      "grad_norm": 0.48978888988494873,
      "learning_rate": 8.632812481496195e-06,
      "loss": 0.5108,
      "step": 139
    },
    {
      "epoch": 0.958904109589041,
      "grad_norm": 0.4350135922431946,
      "learning_rate": 8.60530402074241e-06,
      "loss": 0.5004,
      "step": 140
    },
    {
      "epoch": 0.9657534246575342,
      "grad_norm": 0.4894917607307434,
      "learning_rate": 8.577566342951944e-06,
      "loss": 0.5342,
      "step": 141
    },
    {
      "epoch": 0.9726027397260274,
      "grad_norm": 0.4722926616668701,
      "learning_rate": 8.549601211623316e-06,
      "loss": 0.5374,
      "step": 142
    },
    {
      "epoch": 0.9794520547945206,
      "grad_norm": 0.40020692348480225,
      "learning_rate": 8.521410404716029e-06,
      "loss": 0.5241,
      "step": 143
    },
    {
      "epoch": 0.9863013698630136,
      "grad_norm": 0.4775395691394806,
      "learning_rate": 8.492995714537519e-06,
      "loss": 0.5149,
      "step": 144
    },
    {
      "epoch": 0.9931506849315068,
      "grad_norm": 0.4094436764717102,
      "learning_rate": 8.46435894762922e-06,
      "loss": 0.5025,
      "step": 145
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.3945600688457489,
      "learning_rate": 8.43550192465169e-06,
      "loss": 0.5152,
      "step": 146
    },
    {
      "epoch": 1.0068493150684932,
      "grad_norm": 0.48859384655952454,
      "learning_rate": 8.406426480268881e-06,
      "loss": 0.4662,
      "step": 147
    },
    {
      "epoch": 1.0136986301369864,
      "grad_norm": 0.42010384798049927,
      "learning_rate": 8.377134463031468e-06,
      "loss": 0.4996,
      "step": 148
    },
    {
      "epoch": 1.0205479452054795,
      "grad_norm": 0.6074568033218384,
      "learning_rate": 8.347627735259344e-06,
      "loss": 0.4944,
      "step": 149
    },
    {
      "epoch": 1.0273972602739727,
      "grad_norm": 0.40178409218788147,
      "learning_rate": 8.317908172923207e-06,
      "loss": 0.5116,
      "step": 150
    },
    {
      "epoch": 1.0342465753424657,
      "grad_norm": 0.4420071244239807,
      "learning_rate": 8.287977665525292e-06,
      "loss": 0.4948,
      "step": 151
    },
    {
      "epoch": 1.0410958904109588,
      "grad_norm": 0.4358409643173218,
      "learning_rate": 8.257838115979244e-06,
      "loss": 0.4929,
      "step": 152
    },
    {
      "epoch": 1.047945205479452,
      "grad_norm": 0.4392963945865631,
      "learning_rate": 8.227491440489134e-06,
      "loss": 0.4923,
      "step": 153
    },
    {
      "epoch": 1.0547945205479452,
      "grad_norm": 0.44299256801605225,
      "learning_rate": 8.196939568427624e-06,
      "loss": 0.4921,
      "step": 154
    },
    {
      "epoch": 1.0616438356164384,
      "grad_norm": 0.41032472252845764,
      "learning_rate": 8.166184442213314e-06,
      "loss": 0.4864,
      "step": 155
    },
    {
      "epoch": 1.0684931506849316,
      "grad_norm": 0.4872773289680481,
      "learning_rate": 8.135228017187238e-06,
      "loss": 0.4849,
      "step": 156
    },
    {
      "epoch": 1.0753424657534247,
      "grad_norm": 0.4646698236465454,
      "learning_rate": 8.10407226148855e-06,
      "loss": 0.4813,
      "step": 157
    },
    {
      "epoch": 1.0821917808219177,
      "grad_norm": 0.42592811584472656,
      "learning_rate": 8.0727191559294e-06,
      "loss": 0.4919,
      "step": 158
    },
    {
      "epoch": 1.0890410958904109,
      "grad_norm": 0.39574742317199707,
      "learning_rate": 8.041170693868985e-06,
      "loss": 0.4844,
      "step": 159
    },
    {
      "epoch": 1.095890410958904,
      "grad_norm": 0.4392130970954895,
      "learning_rate": 8.009428881086836e-06,
      "loss": 0.4963,
      "step": 160
    },
    {
      "epoch": 1.1027397260273972,
      "grad_norm": 0.47602227330207825,
      "learning_rate": 7.977495735655271e-06,
      "loss": 0.5001,
      "step": 161
    },
    {
      "epoch": 1.1095890410958904,
      "grad_norm": 0.39845725893974304,
      "learning_rate": 7.945373287811116e-06,
      "loss": 0.4943,
      "step": 162
    },
    {
      "epoch": 1.1164383561643836,
      "grad_norm": 0.41520991921424866,
      "learning_rate": 7.913063579826601e-06,
      "loss": 0.5106,
      "step": 163
    },
    {
      "epoch": 1.1232876712328768,
      "grad_norm": 0.3922985792160034,
      "learning_rate": 7.880568665879542e-06,
      "loss": 0.4906,
      "step": 164
    },
    {
      "epoch": 1.13013698630137,
      "grad_norm": 0.46458759903907776,
      "learning_rate": 7.847890611922721e-06,
      "loss": 0.4927,
      "step": 165
    },
    {
      "epoch": 1.1369863013698631,
      "grad_norm": 0.37780556082725525,
      "learning_rate": 7.81503149555255e-06,
      "loss": 0.4639,
      "step": 166
    },
    {
      "epoch": 1.143835616438356,
      "grad_norm": 0.3818896412849426,
      "learning_rate": 7.781993405876974e-06,
      "loss": 0.4851,
      "step": 167
    },
    {
      "epoch": 1.1506849315068493,
      "grad_norm": 0.5063320398330688,
      "learning_rate": 7.748778443382658e-06,
      "loss": 0.4807,
      "step": 168
    },
    {
      "epoch": 1.1575342465753424,
      "grad_norm": 0.4002656936645508,
      "learning_rate": 7.715388719801437e-06,
      "loss": 0.4678,
      "step": 169
    },
    {
      "epoch": 1.1643835616438356,
      "grad_norm": 0.41656002402305603,
      "learning_rate": 7.68182635797606e-06,
      "loss": 0.4762,
      "step": 170
    },
    {
      "epoch": 1.1712328767123288,
      "grad_norm": 0.42518362402915955,
      "learning_rate": 7.648093491725224e-06,
      "loss": 0.5075,
      "step": 171
    },
    {
      "epoch": 1.178082191780822,
      "grad_norm": 0.3999499976634979,
      "learning_rate": 7.6141922657079045e-06,
      "loss": 0.4519,
      "step": 172
    },
    {
      "epoch": 1.1849315068493151,
      "grad_norm": 0.40690121054649353,
      "learning_rate": 7.580124835287013e-06,
      "loss": 0.4662,
      "step": 173
    },
    {
      "epoch": 1.191780821917808,
      "grad_norm": 0.4337243139743805,
      "learning_rate": 7.545893366392358e-06,
      "loss": 0.4894,
      "step": 174
    },
    {
      "epoch": 1.1986301369863013,
      "grad_norm": 0.40752294659614563,
      "learning_rate": 7.511500035382943e-06,
      "loss": 0.4788,
      "step": 175
    },
    {
      "epoch": 1.2054794520547945,
      "grad_norm": 0.4354107975959778,
      "learning_rate": 7.476947028908595e-06,
      "loss": 0.5209,
      "step": 176
    },
    {
      "epoch": 1.2123287671232876,
      "grad_norm": 0.37461692094802856,
      "learning_rate": 7.442236543770945e-06,
      "loss": 0.4733,
      "step": 177
    },
    {
      "epoch": 1.2191780821917808,
      "grad_norm": 0.3634370267391205,
      "learning_rate": 7.407370786783757e-06,
      "loss": 0.4707,
      "step": 178
    },
    {
      "epoch": 1.226027397260274,
      "grad_norm": 0.41078001260757446,
      "learning_rate": 7.372351974632634e-06,
      "loss": 0.4897,
      "step": 179
    },
    {
      "epoch": 1.2328767123287672,
      "grad_norm": 0.3920040428638458,
      "learning_rate": 7.33718233373407e-06,
      "loss": 0.4995,
      "step": 180
    },
    {
      "epoch": 1.2397260273972603,
      "grad_norm": 0.400307297706604,
      "learning_rate": 7.3018641000939115e-06,
      "loss": 0.5126,
      "step": 181
    },
    {
      "epoch": 1.2465753424657535,
      "grad_norm": 0.390546590089798,
      "learning_rate": 7.266399519165193e-06,
      "loss": 0.4777,
      "step": 182
    },
    {
      "epoch": 1.2534246575342465,
      "grad_norm": 0.3763103187084198,
      "learning_rate": 7.2307908457053786e-06,
      "loss": 0.4714,
      "step": 183
    },
    {
      "epoch": 1.2602739726027397,
      "grad_norm": 0.4261626899242401,
      "learning_rate": 7.195040343633006e-06,
      "loss": 0.4964,
      "step": 184
    },
    {
      "epoch": 1.2671232876712328,
      "grad_norm": 0.38287729024887085,
      "learning_rate": 7.159150285883757e-06,
      "loss": 0.5032,
      "step": 185
    },
    {
      "epoch": 1.273972602739726,
      "grad_norm": 0.34323829412460327,
      "learning_rate": 7.123122954265942e-06,
      "loss": 0.4533,
      "step": 186
    },
    {
      "epoch": 1.2808219178082192,
      "grad_norm": 0.3638916313648224,
      "learning_rate": 7.086960639315437e-06,
      "loss": 0.4923,
      "step": 187
    },
    {
      "epoch": 1.2876712328767124,
      "grad_norm": 0.41420572996139526,
      "learning_rate": 7.050665640150045e-06,
      "loss": 0.4993,
      "step": 188
    },
    {
      "epoch": 1.2945205479452055,
      "grad_norm": 0.3671460449695587,
      "learning_rate": 7.0142402643233346e-06,
      "loss": 0.4941,
      "step": 189
    },
    {
      "epoch": 1.3013698630136985,
      "grad_norm": 0.4007507860660553,
      "learning_rate": 6.977686827677926e-06,
      "loss": 0.5012,
      "step": 190
    },
    {
      "epoch": 1.308219178082192,
      "grad_norm": 0.4122563302516937,
      "learning_rate": 6.941007654198254e-06,
      "loss": 0.4887,
      "step": 191
    },
    {
      "epoch": 1.3150684931506849,
      "grad_norm": 0.4061591327190399,
      "learning_rate": 6.904205075862816e-06,
      "loss": 0.455,
      "step": 192
    },
    {
      "epoch": 1.321917808219178,
      "grad_norm": 0.38111311197280884,
      "learning_rate": 6.867281432495911e-06,
      "loss": 0.4815,
      "step": 193
    },
    {
      "epoch": 1.3287671232876712,
      "grad_norm": 0.41904938220977783,
      "learning_rate": 6.830239071618874e-06,
      "loss": 0.473,
      "step": 194
    },
    {
      "epoch": 1.3356164383561644,
      "grad_norm": 0.35697585344314575,
      "learning_rate": 6.793080348300834e-06,
      "loss": 0.5002,
      "step": 195
    },
    {
      "epoch": 1.3424657534246576,
      "grad_norm": 0.3706590533256531,
      "learning_rate": 6.755807625008974e-06,
      "loss": 0.4928,
      "step": 196
    },
    {
      "epoch": 1.3493150684931507,
      "grad_norm": 0.37768539786338806,
      "learning_rate": 6.718423271458343e-06,
      "loss": 0.4932,
      "step": 197
    },
    {
      "epoch": 1.356164383561644,
      "grad_norm": 0.3710521161556244,
      "learning_rate": 6.680929664461184e-06,
      "loss": 0.4774,
      "step": 198
    },
    {
      "epoch": 1.3630136986301369,
      "grad_norm": 0.36909735202789307,
      "learning_rate": 6.643329187775827e-06,
      "loss": 0.45,
      "step": 199
    },
    {
      "epoch": 1.36986301369863,
      "grad_norm": 0.3874911069869995,
      "learning_rate": 6.6056242319551315e-06,
      "loss": 0.4698,
      "step": 200
    },
    {
      "epoch": 1.3767123287671232,
      "grad_norm": 0.39832133054733276,
      "learning_rate": 6.567817194194508e-06,
      "loss": 0.4626,
      "step": 201
    },
    {
      "epoch": 1.3835616438356164,
      "grad_norm": 0.4187299311161041,
      "learning_rate": 6.529910478179499e-06,
      "loss": 0.4854,
      "step": 202
    },
    {
      "epoch": 1.3904109589041096,
      "grad_norm": 0.3984171450138092,
      "learning_rate": 6.491906493932968e-06,
      "loss": 0.4633,
      "step": 203
    },
    {
      "epoch": 1.3972602739726028,
      "grad_norm": 0.4085666835308075,
      "learning_rate": 6.45380765766187e-06,
      "loss": 0.477,
      "step": 204
    },
    {
      "epoch": 1.404109589041096,
      "grad_norm": 0.4559183716773987,
      "learning_rate": 6.415616391603639e-06,
      "loss": 0.478,
      "step": 205
    },
    {
      "epoch": 1.410958904109589,
      "grad_norm": 0.4116540551185608,
      "learning_rate": 6.377335123872177e-06,
      "loss": 0.4867,
      "step": 206
    },
    {
      "epoch": 1.4178082191780823,
      "grad_norm": 0.34374645352363586,
      "learning_rate": 6.338966288303499e-06,
      "loss": 0.4915,
      "step": 207
    },
    {
      "epoch": 1.4246575342465753,
      "grad_norm": 0.45456692576408386,
      "learning_rate": 6.300512324300975e-06,
      "loss": 0.4754,
      "step": 208
    },
    {
      "epoch": 1.4315068493150684,
      "grad_norm": 0.3950651288032532,
      "learning_rate": 6.261975676680252e-06,
      "loss": 0.4742,
      "step": 209
    },
    {
      "epoch": 1.4383561643835616,
      "grad_norm": 0.3701169192790985,
      "learning_rate": 6.223358795513812e-06,
      "loss": 0.4917,
      "step": 210
    },
    {
      "epoch": 1.4452054794520548,
      "grad_norm": 0.45799094438552856,
      "learning_rate": 6.184664135975202e-06,
      "loss": 0.4665,
      "step": 211
    },
    {
      "epoch": 1.452054794520548,
      "grad_norm": 0.40342605113983154,
      "learning_rate": 6.145894158182945e-06,
      "loss": 0.4827,
      "step": 212
    },
    {
      "epoch": 1.4589041095890412,
      "grad_norm": 0.3885434567928314,
      "learning_rate": 6.107051327044124e-06,
      "loss": 0.4705,
      "step": 213
    },
    {
      "epoch": 1.4657534246575343,
      "grad_norm": 0.41012731194496155,
      "learning_rate": 6.0681381120976745e-06,
      "loss": 0.4676,
      "step": 214
    },
    {
      "epoch": 1.4726027397260273,
      "grad_norm": 0.40364816784858704,
      "learning_rate": 6.029156987357373e-06,
      "loss": 0.4675,
      "step": 215
    },
    {
      "epoch": 1.4794520547945205,
      "grad_norm": 0.40886637568473816,
      "learning_rate": 5.990110431154549e-06,
      "loss": 0.4718,
      "step": 216
    },
    {
      "epoch": 1.4863013698630136,
      "grad_norm": 0.35799840092658997,
      "learning_rate": 5.951000925980509e-06,
      "loss": 0.5037,
      "step": 217
    },
    {
      "epoch": 1.4931506849315068,
      "grad_norm": 0.39343416690826416,
      "learning_rate": 5.9118309583287205e-06,
      "loss": 0.4883,
      "step": 218
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.4305885434150696,
      "learning_rate": 5.872603018536713e-06,
      "loss": 0.5251,
      "step": 219
    },
    {
      "epoch": 1.5068493150684932,
      "grad_norm": 0.3737599551677704,
      "learning_rate": 5.8333196006277536e-06,
      "loss": 0.4737,
      "step": 220
    },
    {
      "epoch": 1.5136986301369864,
      "grad_norm": 0.3764914572238922,
      "learning_rate": 5.793983202152283e-06,
      "loss": 0.508,
      "step": 221
    },
    {
      "epoch": 1.5205479452054793,
      "grad_norm": 0.416117787361145,
      "learning_rate": 5.754596324029125e-06,
      "loss": 0.4693,
      "step": 222
    },
    {
      "epoch": 1.5273972602739727,
      "grad_norm": 0.3617672026157379,
      "learning_rate": 5.715161470386485e-06,
      "loss": 0.4683,
      "step": 223
    },
    {
      "epoch": 1.5342465753424657,
      "grad_norm": 0.3951134979724884,
      "learning_rate": 5.675681148402743e-06,
      "loss": 0.4884,
      "step": 224
    },
    {
      "epoch": 1.541095890410959,
      "grad_norm": 0.3958381414413452,
      "learning_rate": 5.636157868147054e-06,
      "loss": 0.4914,
      "step": 225
    },
    {
      "epoch": 1.547945205479452,
      "grad_norm": 0.38841086626052856,
      "learning_rate": 5.596594142419759e-06,
      "loss": 0.4879,
      "step": 226
    },
    {
      "epoch": 1.5547945205479452,
      "grad_norm": 0.3935975730419159,
      "learning_rate": 5.556992486592634e-06,
      "loss": 0.4929,
      "step": 227
    },
    {
      "epoch": 1.5616438356164384,
      "grad_norm": 0.4356209337711334,
      "learning_rate": 5.517355418448961e-06,
      "loss": 0.485,
      "step": 228
    },
    {
      "epoch": 1.5684931506849316,
      "grad_norm": 0.36613333225250244,
      "learning_rate": 5.47768545802346e-06,
      "loss": 0.4623,
      "step": 229
    },
    {
      "epoch": 1.5753424657534247,
      "grad_norm": 0.37810492515563965,
      "learning_rate": 5.437985127442065e-06,
      "loss": 0.478,
      "step": 230
    },
    {
      "epoch": 1.5821917808219177,
      "grad_norm": 0.41265738010406494,
      "learning_rate": 5.398256950761578e-06,
      "loss": 0.508,
      "step": 231
    },
    {
      "epoch": 1.589041095890411,
      "grad_norm": 0.3820304870605469,
      "learning_rate": 5.3585034538091885e-06,
      "loss": 0.468,
      "step": 232
    },
    {
      "epoch": 1.595890410958904,
      "grad_norm": 0.36171847581863403,
      "learning_rate": 5.318727164021896e-06,
      "loss": 0.474,
      "step": 233
    },
    {
      "epoch": 1.6027397260273972,
      "grad_norm": 0.4011186361312866,
      "learning_rate": 5.278930610285813e-06,
      "loss": 0.4931,
      "step": 234
    },
    {
      "epoch": 1.6095890410958904,
      "grad_norm": 0.3681487441062927,
      "learning_rate": 5.239116322775392e-06,
      "loss": 0.464,
      "step": 235
    },
    {
      "epoch": 1.6164383561643836,
      "grad_norm": 0.3996686339378357,
      "learning_rate": 5.199286832792553e-06,
      "loss": 0.5076,
      "step": 236
    },
    {
      "epoch": 1.6232876712328768,
      "grad_norm": 0.40732523798942566,
      "learning_rate": 5.159444672605759e-06,
      "loss": 0.4675,
      "step": 237
    },
    {
      "epoch": 1.6301369863013697,
      "grad_norm": 0.4019952714443207,
      "learning_rate": 5.119592375289015e-06,
      "loss": 0.4653,
      "step": 238
    },
    {
      "epoch": 1.6369863013698631,
      "grad_norm": 0.4392927587032318,
      "learning_rate": 5.079732474560821e-06,
      "loss": 0.4886,
      "step": 239
    },
    {
      "epoch": 1.643835616438356,
      "grad_norm": 0.42995181679725647,
      "learning_rate": 5.039867504623084e-06,
      "loss": 0.5046,
      "step": 240
    },
    {
      "epoch": 1.6506849315068495,
      "grad_norm": 0.36546292901039124,
      "learning_rate": 5e-06,
      "loss": 0.4801,
      "step": 241
    },
    {
      "epoch": 1.6575342465753424,
      "grad_norm": 0.4144984781742096,
      "learning_rate": 4.960132495376919e-06,
      "loss": 0.4831,
      "step": 242
    },
    {
      "epoch": 1.6643835616438356,
      "grad_norm": 0.38707658648490906,
      "learning_rate": 4.92026752543918e-06,
      "loss": 0.4937,
      "step": 243
    },
    {
      "epoch": 1.6712328767123288,
      "grad_norm": 0.37487655878067017,
      "learning_rate": 4.880407624710986e-06,
      "loss": 0.4648,
      "step": 244
    },
    {
      "epoch": 1.678082191780822,
      "grad_norm": 0.38767164945602417,
      "learning_rate": 4.8405553273942415e-06,
      "loss": 0.4995,
      "step": 245
    },
    {
      "epoch": 1.6849315068493151,
      "grad_norm": 0.401931494474411,
      "learning_rate": 4.800713167207449e-06,
      "loss": 0.4724,
      "step": 246
    },
    {
      "epoch": 1.691780821917808,
      "grad_norm": 0.35886698961257935,
      "learning_rate": 4.760883677224609e-06,
      "loss": 0.4681,
      "step": 247
    },
    {
      "epoch": 1.6986301369863015,
      "grad_norm": 0.3592085540294647,
      "learning_rate": 4.721069389714188e-06,
      "loss": 0.485,
      "step": 248
    },
    {
      "epoch": 1.7054794520547945,
      "grad_norm": 0.3601101040840149,
      "learning_rate": 4.6812728359781064e-06,
      "loss": 0.4881,
      "step": 249
    },
    {
      "epoch": 1.7123287671232876,
      "grad_norm": 0.37123242020606995,
      "learning_rate": 4.641496546190813e-06,
      "loss": 0.4868,
      "step": 250
    },
    {
      "epoch": 1.7191780821917808,
      "grad_norm": 0.3683170974254608,
      "learning_rate": 4.601743049238425e-06,
      "loss": 0.4827,
      "step": 251
    },
    {
      "epoch": 1.726027397260274,
      "grad_norm": 0.3582712411880493,
      "learning_rate": 4.562014872557936e-06,
      "loss": 0.4646,
      "step": 252
    },
    {
      "epoch": 1.7328767123287672,
      "grad_norm": 0.369663268327713,
      "learning_rate": 4.522314541976541e-06,
      "loss": 0.4802,
      "step": 253
    },
    {
      "epoch": 1.7397260273972601,
      "grad_norm": 0.3558022677898407,
      "learning_rate": 4.48264458155104e-06,
      "loss": 0.4901,
      "step": 254
    },
    {
      "epoch": 1.7465753424657535,
      "grad_norm": 0.3494178354740143,
      "learning_rate": 4.443007513407368e-06,
      "loss": 0.4881,
      "step": 255
    },
    {
      "epoch": 1.7534246575342465,
      "grad_norm": 0.3548129200935364,
      "learning_rate": 4.403405857580243e-06,
      "loss": 0.5079,
      "step": 256
    },
    {
      "epoch": 1.7602739726027399,
      "grad_norm": 0.3643389940261841,
      "learning_rate": 4.363842131852948e-06,
      "loss": 0.4889,
      "step": 257
    },
    {
      "epoch": 1.7671232876712328,
      "grad_norm": 0.3427479863166809,
      "learning_rate": 4.3243188515972575e-06,
      "loss": 0.4791,
      "step": 258
    },
    {
      "epoch": 1.773972602739726,
      "grad_norm": 0.38511818647384644,
      "learning_rate": 4.2848385296135165e-06,
      "loss": 0.473,
      "step": 259
    },
    {
      "epoch": 1.7808219178082192,
      "grad_norm": 0.37743064761161804,
      "learning_rate": 4.245403675970877e-06,
      "loss": 0.5084,
      "step": 260
    },
    {
      "epoch": 1.7876712328767124,
      "grad_norm": 0.33601808547973633,
      "learning_rate": 4.206016797847718e-06,
      "loss": 0.4822,
      "step": 261
    },
    {
      "epoch": 1.7945205479452055,
      "grad_norm": 0.3497495949268341,
      "learning_rate": 4.166680399372248e-06,
      "loss": 0.4936,
      "step": 262
    },
    {
      "epoch": 1.8013698630136985,
      "grad_norm": 0.32923176884651184,
      "learning_rate": 4.127396981463289e-06,
      "loss": 0.4987,
      "step": 263
    },
    {
      "epoch": 1.808219178082192,
      "grad_norm": 0.37325793504714966,
      "learning_rate": 4.08816904167128e-06,
      "loss": 0.4708,
      "step": 264
    },
    {
      "epoch": 1.8150684931506849,
      "grad_norm": 0.3285587728023529,
      "learning_rate": 4.048999074019493e-06,
      "loss": 0.4612,
      "step": 265
    },
    {
      "epoch": 1.821917808219178,
      "grad_norm": 0.34880074858665466,
      "learning_rate": 4.009889568845453e-06,
      "loss": 0.4805,
      "step": 266
    },
    {
      "epoch": 1.8287671232876712,
      "grad_norm": 0.366190105676651,
      "learning_rate": 3.9708430126426286e-06,
      "loss": 0.478,
      "step": 267
    },
    {
      "epoch": 1.8356164383561644,
      "grad_norm": 0.37551647424697876,
      "learning_rate": 3.9318618879023255e-06,
      "loss": 0.4897,
      "step": 268
    },
    {
      "epoch": 1.8424657534246576,
      "grad_norm": 0.3399874269962311,
      "learning_rate": 3.8929486729558775e-06,
      "loss": 0.4558,
      "step": 269
    },
    {
      "epoch": 1.8493150684931505,
      "grad_norm": 0.37752845883369446,
      "learning_rate": 3.854105841817056e-06,
      "loss": 0.476,
      "step": 270
    },
    {
      "epoch": 1.856164383561644,
      "grad_norm": 0.38380375504493713,
      "learning_rate": 3.8153358640248e-06,
      "loss": 0.4845,
      "step": 271
    },
    {
      "epoch": 1.8630136986301369,
      "grad_norm": 0.35891273617744446,
      "learning_rate": 3.776641204486191e-06,
      "loss": 0.4911,
      "step": 272
    },
    {
      "epoch": 1.8698630136986303,
      "grad_norm": 0.34820806980133057,
      "learning_rate": 3.738024323319749e-06,
      "loss": 0.5001,
      "step": 273
    },
    {
      "epoch": 1.8767123287671232,
      "grad_norm": 0.381046861410141,
      "learning_rate": 3.699487675699027e-06,
      "loss": 0.4655,
      "step": 274
    },
    {
      "epoch": 1.8835616438356164,
      "grad_norm": 0.3240637183189392,
      "learning_rate": 3.661033711696501e-06,
      "loss": 0.4779,
      "step": 275
    },
    {
      "epoch": 1.8904109589041096,
      "grad_norm": 0.32344701886177063,
      "learning_rate": 3.6226648761278238e-06,
      "loss": 0.4519,
      "step": 276
    },
    {
      "epoch": 1.8972602739726028,
      "grad_norm": 0.3502044677734375,
      "learning_rate": 3.5843836083963625e-06,
      "loss": 0.4843,
      "step": 277
    },
    {
      "epoch": 1.904109589041096,
      "grad_norm": 0.31665265560150146,
      "learning_rate": 3.5461923423381313e-06,
      "loss": 0.4679,
      "step": 278
    },
    {
      "epoch": 1.910958904109589,
      "grad_norm": 0.35308602452278137,
      "learning_rate": 3.5080935060670345e-06,
      "loss": 0.4798,
      "step": 279
    },
    {
      "epoch": 1.9178082191780823,
      "grad_norm": 0.32667192816734314,
      "learning_rate": 3.4700895218205026e-06,
      "loss": 0.4814,
      "step": 280
    },
    {
      "epoch": 1.9246575342465753,
      "grad_norm": 0.34751468896865845,
      "learning_rate": 3.432182805805495e-06,
      "loss": 0.4553,
      "step": 281
    },
    {
      "epoch": 1.9315068493150684,
      "grad_norm": 0.32501402497291565,
      "learning_rate": 3.3943757680448697e-06,
      "loss": 0.4725,
      "step": 282
    },
    {
      "epoch": 1.9383561643835616,
      "grad_norm": 0.3369636535644531,
      "learning_rate": 3.3566708122241753e-06,
      "loss": 0.4462,
      "step": 283
    },
    {
      "epoch": 1.9452054794520548,
      "grad_norm": 0.3620050549507141,
      "learning_rate": 3.3190703355388165e-06,
      "loss": 0.4758,
      "step": 284
    },
    {
      "epoch": 1.952054794520548,
      "grad_norm": 0.33129626512527466,
      "learning_rate": 3.2815767285416576e-06,
      "loss": 0.492,
      "step": 285
    },
    {
      "epoch": 1.958904109589041,
      "grad_norm": 0.3341040015220642,
      "learning_rate": 3.244192374991027e-06,
      "loss": 0.4611,
      "step": 286
    },
    {
      "epoch": 1.9657534246575343,
      "grad_norm": 0.3279862701892853,
      "learning_rate": 3.2069196516991685e-06,
      "loss": 0.4802,
      "step": 287
    },
    {
      "epoch": 1.9726027397260273,
      "grad_norm": 0.3518829643726349,
      "learning_rate": 3.169760928381127e-06,
      "loss": 0.4795,
      "step": 288
    },
    {
      "epoch": 1.9794520547945207,
      "grad_norm": 0.3467582166194916,
      "learning_rate": 3.1327185675040907e-06,
      "loss": 0.4993,
      "step": 289
    },
    {
      "epoch": 1.9863013698630136,
      "grad_norm": 0.34387102723121643,
      "learning_rate": 3.0957949241371845e-06,
      "loss": 0.4846,
      "step": 290
    },
    {
      "epoch": 1.9931506849315068,
      "grad_norm": 0.3198384642601013,
      "learning_rate": 3.0589923458017467e-06,
      "loss": 0.4681,
      "step": 291
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.3447113633155823,
      "learning_rate": 3.0223131723220756e-06,
      "loss": 0.4804,
      "step": 292
    },
    {
      "epoch": 2.006849315068493,
      "grad_norm": 0.3846930265426636,
      "learning_rate": 2.9857597356766675e-06,
      "loss": 0.4599,
      "step": 293
    },
    {
      "epoch": 2.0136986301369864,
      "grad_norm": 0.3663610517978668,
      "learning_rate": 2.949334359849957e-06,
      "loss": 0.4611,
      "step": 294
    },
    {
      "epoch": 2.0205479452054793,
      "grad_norm": 0.34175989031791687,
      "learning_rate": 2.913039360684565e-06,
      "loss": 0.4806,
      "step": 295
    },
    {
      "epoch": 2.0273972602739727,
      "grad_norm": 0.29779112339019775,
      "learning_rate": 2.876877045734058e-06,
      "loss": 0.4257,
      "step": 296
    },
    {
      "epoch": 2.0342465753424657,
      "grad_norm": 0.32330426573753357,
      "learning_rate": 2.840849714116244e-06,
      "loss": 0.4458,
      "step": 297
    },
    {
      "epoch": 2.041095890410959,
      "grad_norm": 0.3471793234348297,
      "learning_rate": 2.8049596563669936e-06,
      "loss": 0.4457,
      "step": 298
    },
    {
      "epoch": 2.047945205479452,
      "grad_norm": 0.36945298314094543,
      "learning_rate": 2.769209154294623e-06,
      "loss": 0.4343,
      "step": 299
    },
    {
      "epoch": 2.0547945205479454,
      "grad_norm": 0.354952335357666,
      "learning_rate": 2.7336004808348094e-06,
      "loss": 0.4463,
      "step": 300
    },
    {
      "epoch": 2.0616438356164384,
      "grad_norm": 0.3860822916030884,
      "learning_rate": 2.69813589990609e-06,
      "loss": 0.4459,
      "step": 301
    },
    {
      "epoch": 2.0684931506849313,
      "grad_norm": 0.3559304475784302,
      "learning_rate": 2.662817666265932e-06,
      "loss": 0.4594,
      "step": 302
    },
    {
      "epoch": 2.0753424657534247,
      "grad_norm": 0.3339487314224243,
      "learning_rate": 2.6276480253673663e-06,
      "loss": 0.4612,
      "step": 303
    },
    {
      "epoch": 2.0821917808219177,
      "grad_norm": 0.404588520526886,
      "learning_rate": 2.5926292132162432e-06,
      "loss": 0.4606,
      "step": 304
    },
    {
      "epoch": 2.089041095890411,
      "grad_norm": 0.3664696216583252,
      "learning_rate": 2.5577634562290567e-06,
      "loss": 0.4416,
      "step": 305
    },
    {
      "epoch": 2.095890410958904,
      "grad_norm": 0.35626354813575745,
      "learning_rate": 2.5230529710914074e-06,
      "loss": 0.47,
      "step": 306
    },
    {
      "epoch": 2.1027397260273974,
      "grad_norm": 0.32117801904678345,
      "learning_rate": 2.48849996461706e-06,
      "loss": 0.4475,
      "step": 307
    },
    {
      "epoch": 2.1095890410958904,
      "grad_norm": 0.34588146209716797,
      "learning_rate": 2.4541066336076434e-06,
      "loss": 0.4436,
      "step": 308
    },
    {
      "epoch": 2.1164383561643834,
      "grad_norm": 0.31559157371520996,
      "learning_rate": 2.4198751647129896e-06,
      "loss": 0.458,
      "step": 309
    },
    {
      "epoch": 2.1232876712328768,
      "grad_norm": 0.3515860140323639,
      "learning_rate": 2.385807734292097e-06,
      "loss": 0.4564,
      "step": 310
    },
    {
      "epoch": 2.1301369863013697,
      "grad_norm": 0.35864606499671936,
      "learning_rate": 2.3519065082747777e-06,
      "loss": 0.4636,
      "step": 311
    },
    {
      "epoch": 2.136986301369863,
      "grad_norm": 0.34572356939315796,
      "learning_rate": 2.318173642023939e-06,
      "loss": 0.4431,
      "step": 312
    },
    {
      "epoch": 2.143835616438356,
      "grad_norm": 0.33984869718551636,
      "learning_rate": 2.284611280198563e-06,
      "loss": 0.4658,
      "step": 313
    },
    {
      "epoch": 2.1506849315068495,
      "grad_norm": 0.3217160105705261,
      "learning_rate": 2.251221556617344e-06,
      "loss": 0.4493,
      "step": 314
    },
    {
      "epoch": 2.1575342465753424,
      "grad_norm": 0.3655526340007782,
      "learning_rate": 2.218006594123028e-06,
      "loss": 0.4517,
      "step": 315
    },
    {
      "epoch": 2.1643835616438354,
      "grad_norm": 0.35365915298461914,
      "learning_rate": 2.184968504447453e-06,
      "loss": 0.4542,
      "step": 316
    },
    {
      "epoch": 2.171232876712329,
      "grad_norm": 0.30709370970726013,
      "learning_rate": 2.15210938807728e-06,
      "loss": 0.4265,
      "step": 317
    },
    {
      "epoch": 2.1780821917808217,
      "grad_norm": 0.31368520855903625,
      "learning_rate": 2.11943133412046e-06,
      "loss": 0.4563,
      "step": 318
    },
    {
      "epoch": 2.184931506849315,
      "grad_norm": 0.3410292863845825,
      "learning_rate": 2.086936420173399e-06,
      "loss": 0.4515,
      "step": 319
    },
    {
      "epoch": 2.191780821917808,
      "grad_norm": 0.3358498215675354,
      "learning_rate": 2.0546267121888863e-06,
      "loss": 0.4197,
      "step": 320
    },
    {
      "epoch": 2.1986301369863015,
      "grad_norm": 0.3185282051563263,
      "learning_rate": 2.0225042643447283e-06,
      "loss": 0.4613,
      "step": 321
    },
    {
      "epoch": 2.2054794520547945,
      "grad_norm": 0.3557981252670288,
      "learning_rate": 1.990571118913166e-06,
      "loss": 0.4824,
      "step": 322
    },
    {
      "epoch": 2.212328767123288,
      "grad_norm": 0.31085366010665894,
      "learning_rate": 1.9588293061310165e-06,
      "loss": 0.4413,
      "step": 323
    },
    {
      "epoch": 2.219178082191781,
      "grad_norm": 0.3326927125453949,
      "learning_rate": 1.9272808440706024e-06,
      "loss": 0.453,
      "step": 324
    },
    {
      "epoch": 2.2260273972602738,
      "grad_norm": 0.32726767659187317,
      "learning_rate": 1.8959277385114516e-06,
      "loss": 0.454,
      "step": 325
    },
    {
      "epoch": 2.232876712328767,
      "grad_norm": 0.3480210602283478,
      "learning_rate": 1.864771982812763e-06,
      "loss": 0.4168,
      "step": 326
    },
    {
      "epoch": 2.23972602739726,
      "grad_norm": 0.311626672744751,
      "learning_rate": 1.8338155577866873e-06,
      "loss": 0.4566,
      "step": 327
    },
    {
      "epoch": 2.2465753424657535,
      "grad_norm": 0.2938528060913086,
      "learning_rate": 1.8030604315723765e-06,
      "loss": 0.4372,
      "step": 328
    },
    {
      "epoch": 2.2534246575342465,
      "grad_norm": 0.2981064021587372,
      "learning_rate": 1.7725085595108682e-06,
      "loss": 0.4396,
      "step": 329
    },
    {
      "epoch": 2.26027397260274,
      "grad_norm": 0.3302954435348511,
      "learning_rate": 1.7421618840207576e-06,
      "loss": 0.4634,
      "step": 330
    },
    {
      "epoch": 2.267123287671233,
      "grad_norm": 0.31649407744407654,
      "learning_rate": 1.71202233447471e-06,
      "loss": 0.4425,
      "step": 331
    },
    {
      "epoch": 2.2739726027397262,
      "grad_norm": 0.30021733045578003,
      "learning_rate": 1.682091827076796e-06,
      "loss": 0.4365,
      "step": 332
    },
    {
      "epoch": 2.280821917808219,
      "grad_norm": 0.30878108739852905,
      "learning_rate": 1.6523722647406575e-06,
      "loss": 0.435,
      "step": 333
    },
    {
      "epoch": 2.287671232876712,
      "grad_norm": 0.33258000016212463,
      "learning_rate": 1.6228655369685342e-06,
      "loss": 0.4679,
      "step": 334
    },
    {
      "epoch": 2.2945205479452055,
      "grad_norm": 0.3215084373950958,
      "learning_rate": 1.5935735197311204e-06,
      "loss": 0.4452,
      "step": 335
    },
    {
      "epoch": 2.3013698630136985,
      "grad_norm": 0.32681819796562195,
      "learning_rate": 1.5644980753483109e-06,
      "loss": 0.4886,
      "step": 336
    },
    {
      "epoch": 2.308219178082192,
      "grad_norm": 0.3088940382003784,
      "learning_rate": 1.5356410523707828e-06,
      "loss": 0.4603,
      "step": 337
    },
    {
      "epoch": 2.315068493150685,
      "grad_norm": 0.3307572901248932,
      "learning_rate": 1.5070042854624833e-06,
      "loss": 0.4558,
      "step": 338
    },
    {
      "epoch": 2.3219178082191783,
      "grad_norm": 0.31423696875572205,
      "learning_rate": 1.4785895952839735e-06,
      "loss": 0.4485,
      "step": 339
    },
    {
      "epoch": 2.328767123287671,
      "grad_norm": 0.32906076312065125,
      "learning_rate": 1.4503987883766857e-06,
      "loss": 0.4587,
      "step": 340
    },
    {
      "epoch": 2.3356164383561646,
      "grad_norm": 0.31784313917160034,
      "learning_rate": 1.4224336570480574e-06,
      "loss": 0.457,
      "step": 341
    },
    {
      "epoch": 2.3424657534246576,
      "grad_norm": 0.30230167508125305,
      "learning_rate": 1.3946959792575915e-06,
      "loss": 0.4805,
      "step": 342
    },
    {
      "epoch": 2.3493150684931505,
      "grad_norm": 0.33398470282554626,
      "learning_rate": 1.3671875185038064e-06,
      "loss": 0.4369,
      "step": 343
    },
    {
      "epoch": 2.356164383561644,
      "grad_norm": 0.35357654094696045,
      "learning_rate": 1.3399100237121266e-06,
      "loss": 0.4409,
      "step": 344
    },
    {
      "epoch": 2.363013698630137,
      "grad_norm": 0.344047486782074,
      "learning_rate": 1.312865229123681e-06,
      "loss": 0.4811,
      "step": 345
    },
    {
      "epoch": 2.3698630136986303,
      "grad_norm": 0.30871498584747314,
      "learning_rate": 1.2860548541850542e-06,
      "loss": 0.4625,
      "step": 346
    },
    {
      "epoch": 2.3767123287671232,
      "grad_norm": 0.3363014757633209,
      "learning_rate": 1.2594806034389556e-06,
      "loss": 0.4463,
      "step": 347
    },
    {
      "epoch": 2.383561643835616,
      "grad_norm": 0.3206148147583008,
      "learning_rate": 1.233144166415861e-06,
      "loss": 0.4366,
      "step": 348
    },
    {
      "epoch": 2.3904109589041096,
      "grad_norm": 0.3465731143951416,
      "learning_rate": 1.2070472175265857e-06,
      "loss": 0.4448,
      "step": 349
    },
    {
      "epoch": 2.3972602739726026,
      "grad_norm": 0.29774489998817444,
      "learning_rate": 1.1811914159558374e-06,
      "loss": 0.4441,
      "step": 350
    },
    {
      "epoch": 2.404109589041096,
      "grad_norm": 0.29331234097480774,
      "learning_rate": 1.155578405556722e-06,
      "loss": 0.4299,
      "step": 351
    },
    {
      "epoch": 2.410958904109589,
      "grad_norm": 0.3341335952281952,
      "learning_rate": 1.1302098147462348e-06,
      "loss": 0.4509,
      "step": 352
    },
    {
      "epoch": 2.4178082191780823,
      "grad_norm": 0.3187620937824249,
      "learning_rate": 1.1050872564017329e-06,
      "loss": 0.4656,
      "step": 353
    },
    {
      "epoch": 2.4246575342465753,
      "grad_norm": 0.33351418375968933,
      "learning_rate": 1.080212327758382e-06,
      "loss": 0.437,
      "step": 354
    },
    {
      "epoch": 2.4315068493150687,
      "grad_norm": 0.311325341463089,
      "learning_rate": 1.0555866103076212e-06,
      "loss": 0.4551,
      "step": 355
    },
    {
      "epoch": 2.4383561643835616,
      "grad_norm": 0.30476653575897217,
      "learning_rate": 1.0312116696966012e-06,
      "loss": 0.4347,
      "step": 356
    },
    {
      "epoch": 2.4452054794520546,
      "grad_norm": 0.3320573568344116,
      "learning_rate": 1.0070890556286578e-06,
      "loss": 0.4355,
      "step": 357
    },
    {
      "epoch": 2.452054794520548,
      "grad_norm": 0.2996487021446228,
      "learning_rate": 9.832203017647746e-07,
      "loss": 0.4531,
      "step": 358
    },
    {
      "epoch": 2.458904109589041,
      "grad_norm": 0.31164291501045227,
      "learning_rate": 9.596069256260792e-07,
      "loss": 0.4914,
      "step": 359
    },
    {
      "epoch": 2.4657534246575343,
      "grad_norm": 0.3011796176433563,
      "learning_rate": 9.362504284973683e-07,
      "loss": 0.425,
      "step": 360
    },
    {
      "epoch": 2.4726027397260273,
      "grad_norm": 0.29375994205474854,
      "learning_rate": 9.131522953316502e-07,
      "loss": 0.4695,
      "step": 361
    },
    {
      "epoch": 2.4794520547945207,
      "grad_norm": 0.30826085805892944,
      "learning_rate": 8.903139946557437e-07,
      "loss": 0.457,
      "step": 362
    },
    {
      "epoch": 2.4863013698630136,
      "grad_norm": 0.32142892479896545,
      "learning_rate": 8.677369784769041e-07,
      "loss": 0.4567,
      "step": 363
    },
    {
      "epoch": 2.493150684931507,
      "grad_norm": 0.2887185215950012,
      "learning_rate": 8.454226821905171e-07,
      "loss": 0.4601,
      "step": 364
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.33354097604751587,
      "learning_rate": 8.233725244888291e-07,
      "loss": 0.4549,
      "step": 365
    },
    {
      "epoch": 2.506849315068493,
      "grad_norm": 0.29360541701316833,
      "learning_rate": 8.015879072707611e-07,
      "loss": 0.4755,
      "step": 366
    },
    {
      "epoch": 2.5136986301369864,
      "grad_norm": 0.30357393622398376,
      "learning_rate": 7.800702155527695e-07,
      "loss": 0.437,
      "step": 367
    },
    {
      "epoch": 2.5205479452054793,
      "grad_norm": 0.2883985638618469,
      "learning_rate": 7.588208173807943e-07,
      "loss": 0.4511,
      "step": 368
    },
    {
      "epoch": 2.5273972602739727,
      "grad_norm": 0.31914854049682617,
      "learning_rate": 7.378410637432848e-07,
      "loss": 0.4499,
      "step": 369
    },
    {
      "epoch": 2.5342465753424657,
      "grad_norm": 0.30720996856689453,
      "learning_rate": 7.171322884852988e-07,
      "loss": 0.4815,
      "step": 370
    },
    {
      "epoch": 2.541095890410959,
      "grad_norm": 0.2876569926738739,
      "learning_rate": 6.966958082237096e-07,
      "loss": 0.4205,
      "step": 371
    },
    {
      "epoch": 2.547945205479452,
      "grad_norm": 0.3016420900821686,
      "learning_rate": 6.765329222634892e-07,
      "loss": 0.48,
      "step": 372
    },
    {
      "epoch": 2.5547945205479454,
      "grad_norm": 0.3060811161994934,
      "learning_rate": 6.566449125151086e-07,
      "loss": 0.4609,
      "step": 373
    },
    {
      "epoch": 2.5616438356164384,
      "grad_norm": 0.2952597141265869,
      "learning_rate": 6.370330434130317e-07,
      "loss": 0.4407,
      "step": 374
    },
    {
      "epoch": 2.5684931506849313,
      "grad_norm": 0.3057716488838196,
      "learning_rate": 6.176985618353282e-07,
      "loss": 0.4477,
      "step": 375
    },
    {
      "epoch": 2.5753424657534247,
      "grad_norm": 0.2792918086051941,
      "learning_rate": 5.986426970244009e-07,
      "loss": 0.4489,
      "step": 376
    },
    {
      "epoch": 2.5821917808219177,
      "grad_norm": 0.3092077672481537,
      "learning_rate": 5.798666605088293e-07,
      "loss": 0.4356,
      "step": 377
    },
    {
      "epoch": 2.589041095890411,
      "grad_norm": 0.27638155221939087,
      "learning_rate": 5.613716460263485e-07,
      "loss": 0.4553,
      "step": 378
    },
    {
      "epoch": 2.595890410958904,
      "grad_norm": 0.29136520624160767,
      "learning_rate": 5.431588294479479e-07,
      "loss": 0.4275,
      "step": 379
    },
    {
      "epoch": 2.602739726027397,
      "grad_norm": 0.2942065894603729,
      "learning_rate": 5.252293687031196e-07,
      "loss": 0.4159,
      "step": 380
    },
    {
      "epoch": 2.6095890410958904,
      "grad_norm": 0.3121497929096222,
      "learning_rate": 5.075844037062322e-07,
      "loss": 0.4503,
      "step": 381
    },
    {
      "epoch": 2.616438356164384,
      "grad_norm": 0.3024226129055023,
      "learning_rate": 4.902250562840622e-07,
      "loss": 0.4525,
      "step": 382
    },
    {
      "epoch": 2.6232876712328768,
      "grad_norm": 0.30542418360710144,
      "learning_rate": 4.7315243010447156e-07,
      "loss": 0.4335,
      "step": 383
    },
    {
      "epoch": 2.6301369863013697,
      "grad_norm": 0.3106946051120758,
      "learning_rate": 4.5636761060623314e-07,
      "loss": 0.4627,
      "step": 384
    },
    {
      "epoch": 2.636986301369863,
      "grad_norm": 0.294289231300354,
      "learning_rate": 4.398716649300311e-07,
      "loss": 0.4503,
      "step": 385
    },
    {
      "epoch": 2.643835616438356,
      "grad_norm": 0.29368045926094055,
      "learning_rate": 4.2366564185060134e-07,
      "loss": 0.4447,
      "step": 386
    },
    {
      "epoch": 2.6506849315068495,
      "grad_norm": 0.2810214161872864,
      "learning_rate": 4.077505717100666e-07,
      "loss": 0.4497,
      "step": 387
    },
    {
      "epoch": 2.6575342465753424,
      "grad_norm": 0.33020564913749695,
      "learning_rate": 3.921274663524183e-07,
      "loss": 0.4356,
      "step": 388
    },
    {
      "epoch": 2.6643835616438354,
      "grad_norm": 0.2780897319316864,
      "learning_rate": 3.767973190591906e-07,
      "loss": 0.3993,
      "step": 389
    },
    {
      "epoch": 2.671232876712329,
      "grad_norm": 0.3095431625843048,
      "learning_rate": 3.61761104486314e-07,
      "loss": 0.446,
      "step": 390
    },
    {
      "epoch": 2.678082191780822,
      "grad_norm": 0.3015943765640259,
      "learning_rate": 3.4701977860213956e-07,
      "loss": 0.4296,
      "step": 391
    },
    {
      "epoch": 2.684931506849315,
      "grad_norm": 0.2791678011417389,
      "learning_rate": 3.3257427862666894e-07,
      "loss": 0.442,
      "step": 392
    },
    {
      "epoch": 2.691780821917808,
      "grad_norm": 0.3039053678512573,
      "learning_rate": 3.184255229719624e-07,
      "loss": 0.4256,
      "step": 393
    },
    {
      "epoch": 2.6986301369863015,
      "grad_norm": 0.2986087203025818,
      "learning_rate": 3.045744111837529e-07,
      "loss": 0.4683,
      "step": 394
    },
    {
      "epoch": 2.7054794520547945,
      "grad_norm": 0.30011793971061707,
      "learning_rate": 2.9102182388425106e-07,
      "loss": 0.4459,
      "step": 395
    },
    {
      "epoch": 2.712328767123288,
      "grad_norm": 0.26771464943885803,
      "learning_rate": 2.777686227161591e-07,
      "loss": 0.4355,
      "step": 396
    },
    {
      "epoch": 2.719178082191781,
      "grad_norm": 0.28580793738365173,
      "learning_rate": 2.648156502878907e-07,
      "loss": 0.4607,
      "step": 397
    },
    {
      "epoch": 2.7260273972602738,
      "grad_norm": 0.30609655380249023,
      "learning_rate": 2.5216373011999697e-07,
      "loss": 0.4665,
      "step": 398
    },
    {
      "epoch": 2.732876712328767,
      "grad_norm": 0.30261853337287903,
      "learning_rate": 2.3981366659281135e-07,
      "loss": 0.4277,
      "step": 399
    },
    {
      "epoch": 2.73972602739726,
      "grad_norm": 0.2930530905723572,
      "learning_rate": 2.2776624489530664e-07,
      "loss": 0.4553,
      "step": 400
    },
    {
      "epoch": 2.7465753424657535,
      "grad_norm": 0.3095199465751648,
      "learning_rate": 2.1602223097517915e-07,
      "loss": 0.426,
      "step": 401
    },
    {
      "epoch": 2.7534246575342465,
      "grad_norm": 0.3025897741317749,
      "learning_rate": 2.0458237149014347e-07,
      "loss": 0.4385,
      "step": 402
    },
    {
      "epoch": 2.76027397260274,
      "grad_norm": 0.27555036544799805,
      "learning_rate": 1.9344739376047083e-07,
      "loss": 0.418,
      "step": 403
    },
    {
      "epoch": 2.767123287671233,
      "grad_norm": 0.27457016706466675,
      "learning_rate": 1.8261800572274001e-07,
      "loss": 0.4213,
      "step": 404
    },
    {
      "epoch": 2.7739726027397262,
      "grad_norm": 0.2816792130470276,
      "learning_rate": 1.7209489588483396e-07,
      "loss": 0.4562,
      "step": 405
    },
    {
      "epoch": 2.780821917808219,
      "grad_norm": 0.2918860614299774,
      "learning_rate": 1.6187873328216142e-07,
      "loss": 0.4405,
      "step": 406
    },
    {
      "epoch": 2.787671232876712,
      "grad_norm": 0.30940842628479004,
      "learning_rate": 1.519701674351265e-07,
      "loss": 0.4667,
      "step": 407
    },
    {
      "epoch": 2.7945205479452055,
      "grad_norm": 0.2933136224746704,
      "learning_rate": 1.4236982830782676e-07,
      "loss": 0.4426,
      "step": 408
    },
    {
      "epoch": 2.8013698630136985,
      "grad_norm": 0.2901608347892761,
      "learning_rate": 1.3307832626800966e-07,
      "loss": 0.4664,
      "step": 409
    },
    {
      "epoch": 2.808219178082192,
      "grad_norm": 0.2942031919956207,
      "learning_rate": 1.2409625204825802e-07,
      "loss": 0.4149,
      "step": 410
    },
    {
      "epoch": 2.815068493150685,
      "grad_norm": 0.29800334572792053,
      "learning_rate": 1.1542417670844075e-07,
      "loss": 0.4464,
      "step": 411
    },
    {
      "epoch": 2.821917808219178,
      "grad_norm": 0.27716100215911865,
      "learning_rate": 1.0706265159939944e-07,
      "loss": 0.4472,
      "step": 412
    },
    {
      "epoch": 2.828767123287671,
      "grad_norm": 0.2767964005470276,
      "learning_rate": 9.901220832790104e-08,
      "loss": 0.4535,
      "step": 413
    },
    {
      "epoch": 2.8356164383561646,
      "grad_norm": 0.2769363522529602,
      "learning_rate": 9.12733587228326e-08,
      "loss": 0.4219,
      "step": 414
    },
    {
      "epoch": 2.8424657534246576,
      "grad_norm": 0.29575496912002563,
      "learning_rate": 8.384659480266733e-08,
      "loss": 0.4549,
      "step": 415
    },
    {
      "epoch": 2.8493150684931505,
      "grad_norm": 0.2830677032470703,
      "learning_rate": 7.673238874417676e-08,
      "loss": 0.4469,
      "step": 416
    },
    {
      "epoch": 2.856164383561644,
      "grad_norm": 0.31160372495651245,
      "learning_rate": 6.9931192852416e-08,
      "loss": 0.4427,
      "step": 417
    },
    {
      "epoch": 2.863013698630137,
      "grad_norm": 0.28588008880615234,
      "learning_rate": 6.344343953196386e-08,
      "loss": 0.4769,
      "step": 418
    },
    {
      "epoch": 2.8698630136986303,
      "grad_norm": 0.2827588617801666,
      "learning_rate": 5.726954125943318e-08,
      "loss": 0.439,
      "step": 419
    },
    {
      "epoch": 2.8767123287671232,
      "grad_norm": 0.2822471857070923,
      "learning_rate": 5.1409890557246876e-08,
      "loss": 0.4169,
      "step": 420
    },
    {
      "epoch": 2.883561643835616,
      "grad_norm": 0.2951059937477112,
      "learning_rate": 4.586485996867951e-08,
      "loss": 0.4413,
      "step": 421
    },
    {
      "epoch": 2.8904109589041096,
      "grad_norm": 0.28772932291030884,
      "learning_rate": 4.063480203417625e-08,
      "loss": 0.4463,
      "step": 422
    },
    {
      "epoch": 2.897260273972603,
      "grad_norm": 0.2849860191345215,
      "learning_rate": 3.572004926893413e-08,
      "loss": 0.4184,
      "step": 423
    },
    {
      "epoch": 2.904109589041096,
      "grad_norm": 0.2915169298648834,
      "learning_rate": 3.1120914141766214e-08,
      "loss": 0.4391,
      "step": 424
    },
    {
      "epoch": 2.910958904109589,
      "grad_norm": 0.3080134987831116,
      "learning_rate": 2.683768905523243e-08,
      "loss": 0.4579,
      "step": 425
    },
    {
      "epoch": 2.9178082191780823,
      "grad_norm": 0.2908354103565216,
      "learning_rate": 2.287064632705005e-08,
      "loss": 0.456,
      "step": 426
    },
    {
      "epoch": 2.9246575342465753,
      "grad_norm": 0.29851260781288147,
      "learning_rate": 1.9220038172780843e-08,
      "loss": 0.4306,
      "step": 427
    },
    {
      "epoch": 2.9315068493150687,
      "grad_norm": 0.2807960510253906,
      "learning_rate": 1.588609668979446e-08,
      "loss": 0.4539,
      "step": 428
    },
    {
      "epoch": 2.9383561643835616,
      "grad_norm": 0.2773742079734802,
      "learning_rate": 1.286903384251581e-08,
      "loss": 0.4576,
      "step": 429
    },
    {
      "epoch": 2.9452054794520546,
      "grad_norm": 0.2927360534667969,
      "learning_rate": 1.016904144894304e-08,
      "loss": 0.4464,
      "step": 430
    },
    {
      "epoch": 2.952054794520548,
      "grad_norm": 0.28278273344039917,
      "learning_rate": 7.78629116845786e-09,
      "loss": 0.471,
      "step": 431
    },
    {
      "epoch": 2.958904109589041,
      "grad_norm": 0.2733604907989502,
      "learning_rate": 5.720934490907604e-09,
      "loss": 0.4357,
      "step": 432
    },
    {
      "epoch": 2.9657534246575343,
      "grad_norm": 0.2910250723361969,
      "learning_rate": 3.973102726976819e-09,
      "loss": 0.4451,
      "step": 433
    },
    {
      "epoch": 2.9726027397260273,
      "grad_norm": 0.29007551074028015,
      "learning_rate": 2.542906999836725e-09,
      "loss": 0.4503,
      "step": 434
    },
    {
      "epoch": 2.9794520547945207,
      "grad_norm": 0.2818272113800049,
      "learning_rate": 1.4304382380819771e-09,
      "loss": 0.4465,
      "step": 435
    },
    {
      "epoch": 2.9863013698630136,
      "grad_norm": 0.29641109704971313,
      "learning_rate": 6.357671699486201e-10,
      "loss": 0.4726,
      "step": 436
    },
    {
      "epoch": 2.993150684931507,
      "grad_norm": 0.3044302463531494,
      "learning_rate": 1.5894431881657845e-10,
      "loss": 0.4642,
      "step": 437
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.2847689986228943,
      "learning_rate": 0.0,
      "loss": 0.4255,
      "step": 438
    },
    {
      "epoch": 3.0,
      "step": 438,
      "total_flos": 666359318511616.0,
      "train_loss": 0.5066342979125237,
      "train_runtime": 9306.6951,
      "train_samples_per_second": 4.514,
      "train_steps_per_second": 0.047
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 438,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 666359318511616.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}