{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.9898832684824903,
  "eval_steps": 21,
  "global_step": 1284,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0031128404669260703,
      "grad_norm": 14.75,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 2.064,
      "step": 1
    },
    {
      "epoch": 0.0031128404669260703,
      "eval_loss": 4.825625419616699,
      "eval_runtime": 42.977,
      "eval_samples_per_second": 114.201,
      "eval_steps_per_second": 11.425,
      "step": 1
    },
    {
      "epoch": 0.0062256809338521405,
      "grad_norm": 16.25,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 2.2079,
      "step": 2
    },
    {
      "epoch": 0.00933852140077821,
      "grad_norm": 15.625,
      "learning_rate": 3e-06,
      "loss": 2.099,
      "step": 3
    },
    {
      "epoch": 0.012451361867704281,
      "grad_norm": 15.4375,
      "learning_rate": 4.000000000000001e-06,
      "loss": 1.9552,
      "step": 4
    },
    {
      "epoch": 0.01556420233463035,
      "grad_norm": 13.0,
      "learning_rate": 5e-06,
      "loss": 2.0341,
      "step": 5
    },
    {
      "epoch": 0.01867704280155642,
      "grad_norm": 8.4375,
      "learning_rate": 6e-06,
      "loss": 1.8984,
      "step": 6
    },
    {
      "epoch": 0.02178988326848249,
      "grad_norm": 7.25,
      "learning_rate": 7e-06,
      "loss": 1.9003,
      "step": 7
    },
    {
      "epoch": 0.024902723735408562,
      "grad_norm": 5.4375,
      "learning_rate": 8.000000000000001e-06,
      "loss": 1.813,
      "step": 8
    },
    {
      "epoch": 0.02801556420233463,
      "grad_norm": 5.71875,
      "learning_rate": 9e-06,
      "loss": 1.8366,
      "step": 9
    },
    {
      "epoch": 0.0311284046692607,
      "grad_norm": 5.15625,
      "learning_rate": 1e-05,
      "loss": 1.6888,
      "step": 10
    },
    {
      "epoch": 0.034241245136186774,
      "grad_norm": 3.859375,
      "learning_rate": 9.999984798001758e-06,
      "loss": 1.6119,
      "step": 11
    },
    {
      "epoch": 0.03735408560311284,
      "grad_norm": 3.109375,
      "learning_rate": 9.99993919209947e-06,
      "loss": 1.5192,
      "step": 12
    },
    {
      "epoch": 0.04046692607003891,
      "grad_norm": 2.921875,
      "learning_rate": 9.99986318257046e-06,
      "loss": 1.5577,
      "step": 13
    },
    {
      "epoch": 0.04357976653696498,
      "grad_norm": 2.96875,
      "learning_rate": 9.999756769876921e-06,
      "loss": 1.5569,
      "step": 14
    },
    {
      "epoch": 0.04669260700389105,
      "grad_norm": 2.921875,
      "learning_rate": 9.99961995466593e-06,
      "loss": 1.5299,
      "step": 15
    },
    {
      "epoch": 0.049805447470817124,
      "grad_norm": 2.859375,
      "learning_rate": 9.999452737769438e-06,
      "loss": 1.4747,
      "step": 16
    },
    {
      "epoch": 0.05291828793774319,
      "grad_norm": 2.671875,
      "learning_rate": 9.999255120204248e-06,
      "loss": 1.4891,
      "step": 17
    },
    {
      "epoch": 0.05603112840466926,
      "grad_norm": 2.59375,
      "learning_rate": 9.99902710317204e-06,
      "loss": 1.457,
      "step": 18
    },
    {
      "epoch": 0.05914396887159533,
      "grad_norm": 2.5,
      "learning_rate": 9.998768688059334e-06,
      "loss": 1.518,
      "step": 19
    },
    {
      "epoch": 0.0622568093385214,
      "grad_norm": 2.34375,
      "learning_rate": 9.998479876437506e-06,
      "loss": 1.3843,
      "step": 20
    },
    {
      "epoch": 0.06536964980544747,
      "grad_norm": 2.59375,
      "learning_rate": 9.99816067006276e-06,
      "loss": 1.4511,
      "step": 21
    },
    {
      "epoch": 0.06536964980544747,
      "eval_loss": 3.4685511589050293,
      "eval_runtime": 238.3402,
      "eval_samples_per_second": 20.592,
      "eval_steps_per_second": 2.06,
      "step": 21
    },
    {
      "epoch": 0.06848249027237355,
      "grad_norm": 2.296875,
      "learning_rate": 9.997811070876122e-06,
      "loss": 1.4491,
      "step": 22
    },
    {
      "epoch": 0.0715953307392996,
      "grad_norm": 2.234375,
      "learning_rate": 9.99743108100344e-06,
      "loss": 1.4585,
      "step": 23
    },
    {
      "epoch": 0.07470817120622568,
      "grad_norm": 2.203125,
      "learning_rate": 9.997020702755353e-06,
      "loss": 1.4002,
      "step": 24
    },
    {
      "epoch": 0.07782101167315175,
      "grad_norm": 2.21875,
      "learning_rate": 9.996579938627288e-06,
      "loss": 1.4493,
      "step": 25
    },
    {
      "epoch": 0.08093385214007782,
      "grad_norm": 2.234375,
      "learning_rate": 9.996108791299447e-06,
      "loss": 1.4291,
      "step": 26
    },
    {
      "epoch": 0.0840466926070039,
      "grad_norm": 2.25,
      "learning_rate": 9.99560726363678e-06,
      "loss": 1.394,
      "step": 27
    },
    {
      "epoch": 0.08715953307392996,
      "grad_norm": 2.28125,
      "learning_rate": 9.995075358688975e-06,
      "loss": 1.4258,
      "step": 28
    },
    {
      "epoch": 0.09027237354085603,
      "grad_norm": 2.28125,
      "learning_rate": 9.99451307969044e-06,
      "loss": 1.3931,
      "step": 29
    },
    {
      "epoch": 0.0933852140077821,
      "grad_norm": 2.359375,
      "learning_rate": 9.993920430060282e-06,
      "loss": 1.4083,
      "step": 30
    },
    {
      "epoch": 0.09649805447470818,
      "grad_norm": 2.25,
      "learning_rate": 9.993297413402282e-06,
      "loss": 1.3743,
      "step": 31
    },
    {
      "epoch": 0.09961089494163425,
      "grad_norm": 2.171875,
      "learning_rate": 9.992644033504881e-06,
      "loss": 1.3644,
      "step": 32
    },
    {
      "epoch": 0.10272373540856031,
      "grad_norm": 2.125,
      "learning_rate": 9.991960294341151e-06,
      "loss": 1.4141,
      "step": 33
    },
    {
      "epoch": 0.10583657587548638,
      "grad_norm": 2.3125,
      "learning_rate": 9.991246200068773e-06,
      "loss": 1.4032,
      "step": 34
    },
    {
      "epoch": 0.10894941634241245,
      "grad_norm": 2.234375,
      "learning_rate": 9.990501755030009e-06,
      "loss": 1.3794,
      "step": 35
    },
    {
      "epoch": 0.11206225680933853,
      "grad_norm": 2.234375,
      "learning_rate": 9.989726963751683e-06,
      "loss": 1.3835,
      "step": 36
    },
    {
      "epoch": 0.11517509727626458,
      "grad_norm": 2.21875,
      "learning_rate": 9.988921830945141e-06,
      "loss": 1.3217,
      "step": 37
    },
    {
      "epoch": 0.11828793774319066,
      "grad_norm": 2.203125,
      "learning_rate": 9.98808636150624e-06,
      "loss": 1.3395,
      "step": 38
    },
    {
      "epoch": 0.12140077821011673,
      "grad_norm": 2.171875,
      "learning_rate": 9.987220560515296e-06,
      "loss": 1.4044,
      "step": 39
    },
    {
      "epoch": 0.1245136186770428,
      "grad_norm": 2.34375,
      "learning_rate": 9.986324433237073e-06,
      "loss": 1.443,
      "step": 40
    },
    {
      "epoch": 0.12762645914396886,
      "grad_norm": 2.140625,
      "learning_rate": 9.985397985120742e-06,
      "loss": 1.4289,
      "step": 41
    },
    {
      "epoch": 0.13073929961089495,
      "grad_norm": 2.15625,
      "learning_rate": 9.984441221799847e-06,
      "loss": 1.3398,
      "step": 42
    },
    {
      "epoch": 0.13073929961089495,
      "eval_loss": 3.445338010787964,
      "eval_runtime": 43.0264,
      "eval_samples_per_second": 114.07,
      "eval_steps_per_second": 11.412,
      "step": 42
    },
    {
      "epoch": 0.133852140077821,
      "grad_norm": 2.109375,
      "learning_rate": 9.983454149092277e-06,
      "loss": 1.3594,
      "step": 43
    },
    {
      "epoch": 0.1369649805447471,
      "grad_norm": 2.109375,
      "learning_rate": 9.982436773000217e-06,
      "loss": 1.3072,
      "step": 44
    },
    {
      "epoch": 0.14007782101167315,
      "grad_norm": 2.203125,
      "learning_rate": 9.981389099710132e-06,
      "loss": 1.3684,
      "step": 45
    },
    {
      "epoch": 0.1431906614785992,
      "grad_norm": 2.21875,
      "learning_rate": 9.980311135592712e-06,
      "loss": 1.3878,
      "step": 46
    },
    {
      "epoch": 0.1463035019455253,
      "grad_norm": 2.171875,
      "learning_rate": 9.979202887202839e-06,
      "loss": 1.3926,
      "step": 47
    },
    {
      "epoch": 0.14941634241245136,
      "grad_norm": 2.125,
      "learning_rate": 9.978064361279549e-06,
      "loss": 1.3231,
      "step": 48
    },
    {
      "epoch": 0.15252918287937745,
      "grad_norm": 2.078125,
      "learning_rate": 9.976895564745993e-06,
      "loss": 1.3403,
      "step": 49
    },
    {
      "epoch": 0.1556420233463035,
      "grad_norm": 2.0625,
      "learning_rate": 9.975696504709382e-06,
      "loss": 1.3065,
      "step": 50
    },
    {
      "epoch": 0.15875486381322956,
      "grad_norm": 2.15625,
      "learning_rate": 9.974467188460964e-06,
      "loss": 1.372,
      "step": 51
    },
    {
      "epoch": 0.16186770428015565,
      "grad_norm": 2.234375,
      "learning_rate": 9.973207623475964e-06,
      "loss": 1.384,
      "step": 52
    },
    {
      "epoch": 0.1649805447470817,
      "grad_norm": 2.296875,
      "learning_rate": 9.971917817413543e-06,
      "loss": 1.39,
      "step": 53
    },
    {
      "epoch": 0.1680933852140078,
      "grad_norm": 2.265625,
      "learning_rate": 9.970597778116753e-06,
      "loss": 1.3777,
      "step": 54
    },
    {
      "epoch": 0.17120622568093385,
      "grad_norm": 2.09375,
      "learning_rate": 9.969247513612488e-06,
      "loss": 1.3244,
      "step": 55
    },
    {
      "epoch": 0.1743190661478599,
      "grad_norm": 2.234375,
      "learning_rate": 9.967867032111435e-06,
      "loss": 1.3791,
      "step": 56
    },
    {
      "epoch": 0.177431906614786,
      "grad_norm": 2.125,
      "learning_rate": 9.966456342008026e-06,
      "loss": 1.3357,
      "step": 57
    },
    {
      "epoch": 0.18054474708171206,
      "grad_norm": 2.046875,
      "learning_rate": 9.965015451880383e-06,
      "loss": 1.347,
      "step": 58
    },
    {
      "epoch": 0.18365758754863815,
      "grad_norm": 1.96875,
      "learning_rate": 9.96354437049027e-06,
      "loss": 1.2783,
      "step": 59
    },
    {
      "epoch": 0.1867704280155642,
      "grad_norm": 2.078125,
      "learning_rate": 9.96204310678304e-06,
      "loss": 1.336,
      "step": 60
    },
    {
      "epoch": 0.18988326848249026,
      "grad_norm": 2.078125,
      "learning_rate": 9.960511669887574e-06,
      "loss": 1.3173,
      "step": 61
    },
    {
      "epoch": 0.19299610894941635,
      "grad_norm": 2.21875,
      "learning_rate": 9.95895006911623e-06,
      "loss": 1.3325,
      "step": 62
    },
    {
      "epoch": 0.1961089494163424,
      "grad_norm": 2.296875,
      "learning_rate": 9.957358313964796e-06,
      "loss": 1.3432,
      "step": 63
    },
    {
      "epoch": 0.1961089494163424,
      "eval_loss": 3.4063408374786377,
      "eval_runtime": 43.0394,
      "eval_samples_per_second": 114.035,
      "eval_steps_per_second": 11.408,
      "step": 63
    },
    {
      "epoch": 0.1992217898832685,
      "grad_norm": 2.15625,
      "learning_rate": 9.95573641411241e-06,
      "loss": 1.3829,
      "step": 64
    },
    {
      "epoch": 0.20233463035019456,
      "grad_norm": 2.3125,
      "learning_rate": 9.954084379421519e-06,
      "loss": 1.3878,
      "step": 65
    },
    {
      "epoch": 0.20544747081712061,
      "grad_norm": 2.125,
      "learning_rate": 9.952402219937817e-06,
      "loss": 1.3584,
      "step": 66
    },
    {
      "epoch": 0.2085603112840467,
      "grad_norm": 2.171875,
      "learning_rate": 9.950689945890177e-06,
      "loss": 1.364,
      "step": 67
    },
    {
      "epoch": 0.21167315175097276,
      "grad_norm": 2.15625,
      "learning_rate": 9.948947567690594e-06,
      "loss": 1.3538,
      "step": 68
    },
    {
      "epoch": 0.21478599221789882,
      "grad_norm": 2.0625,
      "learning_rate": 9.947175095934123e-06,
      "loss": 1.3933,
      "step": 69
    },
    {
      "epoch": 0.2178988326848249,
      "grad_norm": 2.109375,
      "learning_rate": 9.945372541398804e-06,
      "loss": 1.3466,
      "step": 70
    },
    {
      "epoch": 0.22101167315175096,
      "grad_norm": 2.234375,
      "learning_rate": 9.943539915045613e-06,
      "loss": 1.3053,
      "step": 71
    },
    {
      "epoch": 0.22412451361867705,
      "grad_norm": 2.078125,
      "learning_rate": 9.94167722801838e-06,
      "loss": 1.3427,
      "step": 72
    },
    {
      "epoch": 0.2272373540856031,
      "grad_norm": 2.125,
      "learning_rate": 9.939784491643734e-06,
      "loss": 1.282,
      "step": 73
    },
    {
      "epoch": 0.23035019455252917,
      "grad_norm": 2.078125,
      "learning_rate": 9.937861717431021e-06,
      "loss": 1.3111,
      "step": 74
    },
    {
      "epoch": 0.23346303501945526,
      "grad_norm": 2.234375,
      "learning_rate": 9.935908917072253e-06,
      "loss": 1.3097,
      "step": 75
    },
    {
      "epoch": 0.23657587548638132,
      "grad_norm": 2.109375,
      "learning_rate": 9.933926102442007e-06,
      "loss": 1.3106,
      "step": 76
    },
    {
      "epoch": 0.2396887159533074,
      "grad_norm": 2.046875,
      "learning_rate": 9.931913285597388e-06,
      "loss": 1.3336,
      "step": 77
    },
    {
      "epoch": 0.24280155642023346,
      "grad_norm": 2.140625,
      "learning_rate": 9.929870478777929e-06,
      "loss": 1.3764,
      "step": 78
    },
    {
      "epoch": 0.24591439688715952,
      "grad_norm": 2.078125,
      "learning_rate": 9.927797694405528e-06,
      "loss": 1.3375,
      "step": 79
    },
    {
      "epoch": 0.2490272373540856,
      "grad_norm": 2.203125,
      "learning_rate": 9.925694945084369e-06,
      "loss": 1.2792,
      "step": 80
    },
    {
      "epoch": 0.25214007782101167,
      "grad_norm": 2.15625,
      "learning_rate": 9.923562243600854e-06,
      "loss": 1.2541,
      "step": 81
    },
    {
      "epoch": 0.2552529182879377,
      "grad_norm": 2.171875,
      "learning_rate": 9.921399602923507e-06,
      "loss": 1.2923,
      "step": 82
    },
    {
      "epoch": 0.25836575875486384,
      "grad_norm": 2.125,
      "learning_rate": 9.919207036202914e-06,
      "loss": 1.3427,
      "step": 83
    },
    {
      "epoch": 0.2614785992217899,
      "grad_norm": 2.296875,
      "learning_rate": 9.916984556771634e-06,
      "loss": 1.3794,
      "step": 84
    },
    {
      "epoch": 0.2614785992217899,
      "eval_loss": 3.360220432281494,
      "eval_runtime": 43.0727,
      "eval_samples_per_second": 113.947,
      "eval_steps_per_second": 11.399,
      "step": 84
    },
    {
      "epoch": 0.26459143968871596,
      "grad_norm": 2.0625,
      "learning_rate": 9.914732178144118e-06,
      "loss": 1.3032,
      "step": 85
    },
    {
      "epoch": 0.267704280155642,
      "grad_norm": 2.078125,
      "learning_rate": 9.912449914016627e-06,
      "loss": 1.2801,
      "step": 86
    },
    {
      "epoch": 0.2708171206225681,
      "grad_norm": 2.140625,
      "learning_rate": 9.910137778267153e-06,
      "loss": 1.3511,
      "step": 87
    },
    {
      "epoch": 0.2739299610894942,
      "grad_norm": 2.140625,
      "learning_rate": 9.907795784955327e-06,
      "loss": 1.3418,
      "step": 88
    },
    {
      "epoch": 0.27704280155642025,
      "grad_norm": 2.078125,
      "learning_rate": 9.905423948322343e-06,
      "loss": 1.3268,
      "step": 89
    },
    {
      "epoch": 0.2801556420233463,
      "grad_norm": 2.1875,
      "learning_rate": 9.903022282790864e-06,
      "loss": 1.3076,
      "step": 90
    },
    {
      "epoch": 0.28326848249027237,
      "grad_norm": 2.140625,
      "learning_rate": 9.900590802964932e-06,
      "loss": 1.3174,
      "step": 91
    },
    {
      "epoch": 0.2863813229571984,
      "grad_norm": 2.03125,
      "learning_rate": 9.898129523629892e-06,
      "loss": 1.2773,
      "step": 92
    },
    {
      "epoch": 0.28949416342412454,
      "grad_norm": 2.296875,
      "learning_rate": 9.895638459752285e-06,
      "loss": 1.3194,
      "step": 93
    },
    {
      "epoch": 0.2926070038910506,
      "grad_norm": 2.0625,
      "learning_rate": 9.893117626479778e-06,
      "loss": 1.2994,
      "step": 94
    },
    {
      "epoch": 0.29571984435797666,
      "grad_norm": 2.140625,
      "learning_rate": 9.890567039141043e-06,
      "loss": 1.2786,
      "step": 95
    },
    {
      "epoch": 0.2988326848249027,
      "grad_norm": 2.15625,
      "learning_rate": 9.887986713245697e-06,
      "loss": 1.2812,
      "step": 96
    },
    {
      "epoch": 0.3019455252918288,
      "grad_norm": 2.09375,
      "learning_rate": 9.88537666448418e-06,
      "loss": 1.2752,
      "step": 97
    },
    {
      "epoch": 0.3050583657587549,
      "grad_norm": 2.21875,
      "learning_rate": 9.882736908727677e-06,
      "loss": 1.3149,
      "step": 98
    },
    {
      "epoch": 0.30817120622568095,
      "grad_norm": 2.203125,
      "learning_rate": 9.880067462028013e-06,
      "loss": 1.3206,
      "step": 99
    },
    {
      "epoch": 0.311284046692607,
      "grad_norm": 2.390625,
      "learning_rate": 9.877368340617554e-06,
      "loss": 1.3579,
      "step": 100
    },
    {
      "epoch": 0.31439688715953307,
      "grad_norm": 2.28125,
      "learning_rate": 9.874639560909118e-06,
      "loss": 1.2929,
      "step": 101
    },
    {
      "epoch": 0.3175097276264591,
      "grad_norm": 2.078125,
      "learning_rate": 9.87188113949587e-06,
      "loss": 1.3153,
      "step": 102
    },
    {
      "epoch": 0.32062256809338524,
      "grad_norm": 2.171875,
      "learning_rate": 9.86909309315121e-06,
      "loss": 1.3541,
      "step": 103
    },
    {
      "epoch": 0.3237354085603113,
      "grad_norm": 2.21875,
      "learning_rate": 9.866275438828694e-06,
      "loss": 1.3195,
      "step": 104
    },
    {
      "epoch": 0.32684824902723736,
      "grad_norm": 2.109375,
      "learning_rate": 9.863428193661908e-06,
      "loss": 1.2749,
      "step": 105
    },
    {
      "epoch": 0.32684824902723736,
      "eval_loss": 3.3013370037078857,
      "eval_runtime": 42.9791,
      "eval_samples_per_second": 114.195,
      "eval_steps_per_second": 11.424,
      "step": 105
    },
    {
      "epoch": 0.3299610894941634,
      "grad_norm": 2.078125,
      "learning_rate": 9.860551374964382e-06,
      "loss": 1.3183,
      "step": 106
    },
    {
      "epoch": 0.3330739299610895,
      "grad_norm": 2.125,
      "learning_rate": 9.857645000229473e-06,
      "loss": 1.312,
      "step": 107
    },
    {
      "epoch": 0.3361867704280156,
      "grad_norm": 2.125,
      "learning_rate": 9.854709087130261e-06,
      "loss": 1.3561,
      "step": 108
    },
    {
      "epoch": 0.33929961089494165,
      "grad_norm": 2.171875,
      "learning_rate": 9.851743653519445e-06,
      "loss": 1.2718,
      "step": 109
    },
    {
      "epoch": 0.3424124513618677,
      "grad_norm": 2.046875,
      "learning_rate": 9.848748717429231e-06,
      "loss": 1.3323,
      "step": 110
    },
    {
      "epoch": 0.34552529182879377,
      "grad_norm": 2.078125,
      "learning_rate": 9.845724297071223e-06,
      "loss": 1.3103,
      "step": 111
    },
    {
      "epoch": 0.3486381322957198,
      "grad_norm": 2.0625,
      "learning_rate": 9.842670410836317e-06,
      "loss": 1.3613,
      "step": 112
    },
    {
      "epoch": 0.35175097276264594,
      "grad_norm": 2.140625,
      "learning_rate": 9.839587077294581e-06,
      "loss": 1.2745,
      "step": 113
    },
    {
      "epoch": 0.354863813229572,
      "grad_norm": 2.03125,
      "learning_rate": 9.836474315195148e-06,
      "loss": 1.324,
      "step": 114
    },
    {
      "epoch": 0.35797665369649806,
      "grad_norm": 2.359375,
      "learning_rate": 9.833332143466099e-06,
      "loss": 1.3685,
      "step": 115
    },
    {
      "epoch": 0.3610894941634241,
      "grad_norm": 2.015625,
      "learning_rate": 9.830160581214349e-06,
      "loss": 1.2881,
      "step": 116
    },
    {
      "epoch": 0.3642023346303502,
      "grad_norm": 2.078125,
      "learning_rate": 9.826959647725531e-06,
      "loss": 1.3306,
      "step": 117
    },
    {
      "epoch": 0.3673151750972763,
      "grad_norm": 2.234375,
      "learning_rate": 9.823729362463884e-06,
      "loss": 1.3497,
      "step": 118
    },
    {
      "epoch": 0.37042801556420235,
      "grad_norm": 2.0625,
      "learning_rate": 9.820469745072117e-06,
      "loss": 1.3467,
      "step": 119
    },
    {
      "epoch": 0.3735408560311284,
      "grad_norm": 2.109375,
      "learning_rate": 9.817180815371315e-06,
      "loss": 1.2927,
      "step": 120
    },
    {
      "epoch": 0.37665369649805447,
      "grad_norm": 2.15625,
      "learning_rate": 9.813862593360795e-06,
      "loss": 1.3044,
      "step": 121
    },
    {
      "epoch": 0.37976653696498053,
      "grad_norm": 2.140625,
      "learning_rate": 9.810515099218004e-06,
      "loss": 1.3237,
      "step": 122
    },
    {
      "epoch": 0.38287937743190664,
      "grad_norm": 2.046875,
      "learning_rate": 9.807138353298378e-06,
      "loss": 1.3343,
      "step": 123
    },
    {
      "epoch": 0.3859922178988327,
      "grad_norm": 2.0625,
      "learning_rate": 9.803732376135233e-06,
      "loss": 1.3105,
      "step": 124
    },
    {
      "epoch": 0.38910505836575876,
      "grad_norm": 2.203125,
      "learning_rate": 9.80029718843963e-06,
      "loss": 1.3085,
      "step": 125
    },
    {
      "epoch": 0.3922178988326848,
      "grad_norm": 2.0625,
      "learning_rate": 9.796832811100261e-06,
      "loss": 1.3018,
      "step": 126
    },
    {
      "epoch": 0.3922178988326848,
      "eval_loss": 3.3111178874969482,
      "eval_runtime": 42.869,
      "eval_samples_per_second": 114.488,
      "eval_steps_per_second": 11.454,
      "step": 126
    },
    {
      "epoch": 0.3953307392996109,
      "grad_norm": 2.09375,
      "learning_rate": 9.793339265183303e-06,
      "loss": 1.2746,
      "step": 127
    },
    {
      "epoch": 0.398443579766537,
      "grad_norm": 2.203125,
      "learning_rate": 9.789816571932314e-06,
      "loss": 1.3269,
      "step": 128
    },
    {
      "epoch": 0.40155642023346305,
      "grad_norm": 2.109375,
      "learning_rate": 9.78626475276808e-06,
      "loss": 1.2881,
      "step": 129
    },
    {
      "epoch": 0.4046692607003891,
      "grad_norm": 2.0625,
      "learning_rate": 9.782683829288502e-06,
      "loss": 1.3172,
      "step": 130
    },
    {
      "epoch": 0.40778210116731517,
      "grad_norm": 2.03125,
      "learning_rate": 9.779073823268457e-06,
      "loss": 1.2327,
      "step": 131
    },
    {
      "epoch": 0.41089494163424123,
      "grad_norm": 2.125,
      "learning_rate": 9.775434756659667e-06,
      "loss": 1.3401,
      "step": 132
    },
    {
      "epoch": 0.4140077821011673,
      "grad_norm": 2.15625,
      "learning_rate": 9.771766651590566e-06,
      "loss": 1.2837,
      "step": 133
    },
    {
      "epoch": 0.4171206225680934,
      "grad_norm": 2.078125,
      "learning_rate": 9.768069530366164e-06,
      "loss": 1.3225,
      "step": 134
    },
    {
      "epoch": 0.42023346303501946,
      "grad_norm": 2.15625,
      "learning_rate": 9.764343415467914e-06,
      "loss": 1.29,
      "step": 135
    },
    {
      "epoch": 0.4233463035019455,
      "grad_norm": 2.0625,
      "learning_rate": 9.76058832955357e-06,
      "loss": 1.2578,
      "step": 136
    },
    {
      "epoch": 0.4264591439688716,
      "grad_norm": 2.203125,
      "learning_rate": 9.756804295457061e-06,
      "loss": 1.3049,
      "step": 137
    },
    {
      "epoch": 0.42957198443579764,
      "grad_norm": 1.9375,
      "learning_rate": 9.752991336188336e-06,
      "loss": 1.2803,
      "step": 138
    },
    {
      "epoch": 0.43268482490272375,
      "grad_norm": 2.109375,
      "learning_rate": 9.749149474933236e-06,
      "loss": 1.2986,
      "step": 139
    },
    {
      "epoch": 0.4357976653696498,
      "grad_norm": 2.34375,
      "learning_rate": 9.745278735053345e-06,
      "loss": 1.3462,
      "step": 140
    },
    {
      "epoch": 0.43891050583657587,
      "grad_norm": 2.28125,
      "learning_rate": 9.741379140085858e-06,
      "loss": 1.3486,
      "step": 141
    },
    {
      "epoch": 0.44202334630350193,
      "grad_norm": 2.125,
      "learning_rate": 9.737450713743429e-06,
      "loss": 1.2483,
      "step": 142
    },
    {
      "epoch": 0.445136186770428,
      "grad_norm": 2.0625,
      "learning_rate": 9.733493479914031e-06,
      "loss": 1.262,
      "step": 143
    },
    {
      "epoch": 0.4482490272373541,
      "grad_norm": 2.171875,
      "learning_rate": 9.729507462660808e-06,
      "loss": 1.2834,
      "step": 144
    },
    {
      "epoch": 0.45136186770428016,
      "grad_norm": 2.125,
      "learning_rate": 9.72549268622193e-06,
      "loss": 1.3382,
      "step": 145
    },
    {
      "epoch": 0.4544747081712062,
      "grad_norm": 2.078125,
      "learning_rate": 9.721449175010446e-06,
      "loss": 1.2732,
      "step": 146
    },
    {
      "epoch": 0.4575875486381323,
      "grad_norm": 2.15625,
      "learning_rate": 9.71737695361414e-06,
      "loss": 1.2893,
      "step": 147
    },
    {
      "epoch": 0.4575875486381323,
      "eval_loss": 3.2904868125915527,
      "eval_runtime": 42.9732,
      "eval_samples_per_second": 114.211,
      "eval_steps_per_second": 11.426,
      "step": 147
    },
    {
      "epoch": 0.46070038910505834,
      "grad_norm": 2.078125,
      "learning_rate": 9.71327604679537e-06,
      "loss": 1.2534,
      "step": 148
    },
    {
      "epoch": 0.46381322957198445,
      "grad_norm": 2.171875,
      "learning_rate": 9.709146479490927e-06,
      "loss": 1.2854,
      "step": 149
    },
    {
      "epoch": 0.4669260700389105,
      "grad_norm": 2.125,
      "learning_rate": 9.704988276811883e-06,
      "loss": 1.2836,
      "step": 150
    },
    {
      "epoch": 0.47003891050583657,
      "grad_norm": 2.09375,
      "learning_rate": 9.700801464043432e-06,
      "loss": 1.2975,
      "step": 151
    },
    {
      "epoch": 0.47315175097276263,
      "grad_norm": 2.15625,
      "learning_rate": 9.696586066644743e-06,
      "loss": 1.3118,
      "step": 152
    },
    {
      "epoch": 0.4762645914396887,
      "grad_norm": 2.125,
      "learning_rate": 9.692342110248802e-06,
      "loss": 1.3219,
      "step": 153
    },
    {
      "epoch": 0.4793774319066148,
      "grad_norm": 2.15625,
      "learning_rate": 9.688069620662255e-06,
      "loss": 1.2997,
      "step": 154
    },
    {
      "epoch": 0.48249027237354086,
      "grad_norm": 2.171875,
      "learning_rate": 9.683768623865256e-06,
      "loss": 1.3123,
      "step": 155
    },
    {
      "epoch": 0.4856031128404669,
      "grad_norm": 2.0,
      "learning_rate": 9.6794391460113e-06,
      "loss": 1.3152,
      "step": 156
    },
    {
      "epoch": 0.488715953307393,
      "grad_norm": 2.078125,
      "learning_rate": 9.675081213427076e-06,
      "loss": 1.271,
      "step": 157
    },
    {
      "epoch": 0.49182879377431904,
      "grad_norm": 2.109375,
      "learning_rate": 9.670694852612293e-06,
      "loss": 1.2867,
      "step": 158
    },
    {
      "epoch": 0.49494163424124515,
      "grad_norm": 2.109375,
      "learning_rate": 9.666280090239535e-06,
      "loss": 1.2872,
      "step": 159
    },
    {
      "epoch": 0.4980544747081712,
      "grad_norm": 2.171875,
      "learning_rate": 9.661836953154085e-06,
      "loss": 1.2621,
      "step": 160
    },
    {
      "epoch": 0.5011673151750973,
      "grad_norm": 2.265625,
      "learning_rate": 9.657365468373767e-06,
      "loss": 1.3111,
      "step": 161
    },
    {
      "epoch": 0.5042801556420233,
      "grad_norm": 2.125,
      "learning_rate": 9.652865663088783e-06,
      "loss": 1.2811,
      "step": 162
    },
    {
      "epoch": 0.5073929961089494,
      "grad_norm": 2.171875,
      "learning_rate": 9.648337564661544e-06,
      "loss": 1.3022,
      "step": 163
    },
    {
      "epoch": 0.5105058365758754,
      "grad_norm": 2.109375,
      "learning_rate": 9.643781200626512e-06,
      "loss": 1.2897,
      "step": 164
    },
    {
      "epoch": 0.5136186770428015,
      "grad_norm": 2.171875,
      "learning_rate": 9.639196598690019e-06,
      "loss": 1.3004,
      "step": 165
    },
    {
      "epoch": 0.5167315175097277,
      "grad_norm": 2.109375,
      "learning_rate": 9.63458378673011e-06,
      "loss": 1.325,
      "step": 166
    },
    {
      "epoch": 0.5198443579766537,
      "grad_norm": 2.203125,
      "learning_rate": 9.629942792796368e-06,
      "loss": 1.2936,
      "step": 167
    },
    {
      "epoch": 0.5229571984435798,
      "grad_norm": 2.265625,
      "learning_rate": 9.625273645109746e-06,
      "loss": 1.2704,
      "step": 168
    },
    {
      "epoch": 0.5229571984435798,
      "eval_loss": 3.3199856281280518,
      "eval_runtime": 43.0255,
      "eval_samples_per_second": 114.072,
      "eval_steps_per_second": 11.412,
      "step": 168
    },
    {
      "epoch": 0.5260700389105059,
      "grad_norm": 2.078125,
      "learning_rate": 9.620576372062395e-06,
      "loss": 1.3304,
      "step": 169
    },
    {
      "epoch": 0.5291828793774319,
      "grad_norm": 2.171875,
      "learning_rate": 9.61585100221749e-06,
      "loss": 1.3155,
      "step": 170
    },
    {
      "epoch": 0.532295719844358,
      "grad_norm": 2.1875,
      "learning_rate": 9.611097564309054e-06,
      "loss": 1.3404,
      "step": 171
    },
    {
      "epoch": 0.535408560311284,
      "grad_norm": 2.203125,
      "learning_rate": 9.60631608724179e-06,
      "loss": 1.3219,
      "step": 172
    },
    {
      "epoch": 0.5385214007782101,
      "grad_norm": 2.125,
      "learning_rate": 9.601506600090899e-06,
      "loss": 1.3454,
      "step": 173
    },
    {
      "epoch": 0.5416342412451361,
      "grad_norm": 2.125,
      "learning_rate": 9.596669132101913e-06,
      "loss": 1.3173,
      "step": 174
    },
    {
      "epoch": 0.5447470817120622,
      "grad_norm": 2.21875,
      "learning_rate": 9.591803712690498e-06,
      "loss": 1.3023,
      "step": 175
    },
    {
      "epoch": 0.5478599221789884,
      "grad_norm": 2.109375,
      "learning_rate": 9.586910371442294e-06,
      "loss": 1.2795,
      "step": 176
    },
    {
      "epoch": 0.5509727626459144,
      "grad_norm": 2.078125,
      "learning_rate": 9.58198913811273e-06,
      "loss": 1.2241,
      "step": 177
    },
    {
      "epoch": 0.5540856031128405,
      "grad_norm": 2.046875,
      "learning_rate": 9.577040042626832e-06,
      "loss": 1.3081,
      "step": 178
    },
    {
      "epoch": 0.5571984435797666,
      "grad_norm": 2.15625,
      "learning_rate": 9.572063115079063e-06,
      "loss": 1.3233,
      "step": 179
    },
    {
      "epoch": 0.5603112840466926,
      "grad_norm": 2.09375,
      "learning_rate": 9.567058385733118e-06,
      "loss": 1.2702,
      "step": 180
    },
    {
      "epoch": 0.5634241245136187,
      "grad_norm": 2.109375,
      "learning_rate": 9.56202588502175e-06,
      "loss": 1.3005,
      "step": 181
    },
    {
      "epoch": 0.5665369649805447,
      "grad_norm": 2.078125,
      "learning_rate": 9.55696564354659e-06,
      "loss": 1.3114,
      "step": 182
    },
    {
      "epoch": 0.5696498054474708,
      "grad_norm": 2.125,
      "learning_rate": 9.551877692077946e-06,
      "loss": 1.3246,
      "step": 183
    },
    {
      "epoch": 0.5727626459143969,
      "grad_norm": 2.046875,
      "learning_rate": 9.546762061554632e-06,
      "loss": 1.2964,
      "step": 184
    },
    {
      "epoch": 0.5758754863813229,
      "grad_norm": 2.03125,
      "learning_rate": 9.54161878308377e-06,
      "loss": 1.2711,
      "step": 185
    },
    {
      "epoch": 0.5789883268482491,
      "grad_norm": 2.15625,
      "learning_rate": 9.536447887940606e-06,
      "loss": 1.3415,
      "step": 186
    },
    {
      "epoch": 0.5821011673151751,
      "grad_norm": 2.3125,
      "learning_rate": 9.531249407568312e-06,
      "loss": 1.3254,
      "step": 187
    },
    {
      "epoch": 0.5852140077821012,
      "grad_norm": 2.1875,
      "learning_rate": 9.526023373577808e-06,
      "loss": 1.3008,
      "step": 188
    },
    {
      "epoch": 0.5883268482490273,
      "grad_norm": 2.015625,
      "learning_rate": 9.520769817747554e-06,
      "loss": 1.2826,
      "step": 189
    },
    {
      "epoch": 0.5883268482490273,
      "eval_loss": 3.284386396408081,
      "eval_runtime": 42.9458,
      "eval_samples_per_second": 114.284,
      "eval_steps_per_second": 11.433,
      "step": 189
    },
    {
      "epoch": 0.5914396887159533,
      "grad_norm": 2.15625,
      "learning_rate": 9.515488772023371e-06,
      "loss": 1.2494,
      "step": 190
    },
    {
      "epoch": 0.5945525291828794,
      "grad_norm": 2.109375,
      "learning_rate": 9.510180268518236e-06,
      "loss": 1.2436,
      "step": 191
    },
    {
      "epoch": 0.5976653696498054,
      "grad_norm": 2.109375,
      "learning_rate": 9.504844339512096e-06,
      "loss": 1.2947,
      "step": 192
    },
    {
      "epoch": 0.6007782101167315,
      "grad_norm": 2.171875,
      "learning_rate": 9.499481017451665e-06,
      "loss": 1.3407,
      "step": 193
    },
    {
      "epoch": 0.6038910505836576,
      "grad_norm": 2.109375,
      "learning_rate": 9.494090334950221e-06,
      "loss": 1.2728,
      "step": 194
    },
    {
      "epoch": 0.6070038910505836,
      "grad_norm": 2.265625,
      "learning_rate": 9.488672324787431e-06,
      "loss": 1.2765,
      "step": 195
    },
    {
      "epoch": 0.6101167315175098,
      "grad_norm": 2.203125,
      "learning_rate": 9.483227019909123e-06,
      "loss": 1.3156,
      "step": 196
    },
    {
      "epoch": 0.6132295719844358,
      "grad_norm": 2.109375,
      "learning_rate": 9.477754453427104e-06,
      "loss": 1.2856,
      "step": 197
    },
    {
      "epoch": 0.6163424124513619,
      "grad_norm": 2.0625,
      "learning_rate": 9.472254658618952e-06,
      "loss": 1.2391,
      "step": 198
    },
    {
      "epoch": 0.619455252918288,
      "grad_norm": 2.0625,
      "learning_rate": 9.466727668927817e-06,
      "loss": 1.2687,
      "step": 199
    },
    {
      "epoch": 0.622568093385214,
      "grad_norm": 1.9921875,
      "learning_rate": 9.46117351796221e-06,
      "loss": 1.2733,
      "step": 200
    },
    {
      "epoch": 0.6256809338521401,
      "grad_norm": 2.078125,
      "learning_rate": 9.455592239495814e-06,
      "loss": 1.2881,
      "step": 201
    },
    {
      "epoch": 0.6287937743190661,
      "grad_norm": 2.0,
      "learning_rate": 9.449983867467259e-06,
      "loss": 1.2382,
      "step": 202
    },
    {
      "epoch": 0.6319066147859922,
      "grad_norm": 2.140625,
      "learning_rate": 9.444348435979932e-06,
      "loss": 1.2663,
      "step": 203
    },
    {
      "epoch": 0.6350194552529183,
      "grad_norm": 2.203125,
      "learning_rate": 9.438685979301757e-06,
      "loss": 1.253,
      "step": 204
    },
    {
      "epoch": 0.6381322957198443,
      "grad_norm": 2.03125,
      "learning_rate": 9.432996531865001e-06,
      "loss": 1.3063,
      "step": 205
    },
    {
      "epoch": 0.6412451361867705,
      "grad_norm": 2.0625,
      "learning_rate": 9.427280128266049e-06,
      "loss": 1.2458,
      "step": 206
    },
    {
      "epoch": 0.6443579766536965,
      "grad_norm": 2.109375,
      "learning_rate": 9.421536803265208e-06,
      "loss": 1.2767,
      "step": 207
    },
    {
      "epoch": 0.6474708171206226,
      "grad_norm": 2.203125,
      "learning_rate": 9.415766591786481e-06,
      "loss": 1.2608,
      "step": 208
    },
    {
      "epoch": 0.6505836575875487,
      "grad_norm": 2.125,
      "learning_rate": 9.409969528917366e-06,
      "loss": 1.3592,
      "step": 209
    },
    {
      "epoch": 0.6536964980544747,
      "grad_norm": 2.125,
      "learning_rate": 9.40414564990864e-06,
      "loss": 1.3037,
      "step": 210
    },
    {
      "epoch": 0.6536964980544747,
      "eval_loss": 3.2836391925811768,
      "eval_runtime": 43.7873,
      "eval_samples_per_second": 112.087,
      "eval_steps_per_second": 11.213,
      "step": 210
    },
    {
      "epoch": 0.6568093385214008,
      "grad_norm": 2.21875,
      "learning_rate": 9.39829499017414e-06,
      "loss": 1.2922,
      "step": 211
    },
    {
      "epoch": 0.6599221789883268,
      "grad_norm": 2.09375,
      "learning_rate": 9.39241758529056e-06,
      "loss": 1.2364,
      "step": 212
    },
    {
      "epoch": 0.6630350194552529,
      "grad_norm": 2.09375,
      "learning_rate": 9.38651347099721e-06,
      "loss": 1.3096,
      "step": 213
    },
    {
      "epoch": 0.666147859922179,
      "grad_norm": 2.03125,
      "learning_rate": 9.380582683195832e-06,
      "loss": 1.2232,
      "step": 214
    },
    {
      "epoch": 0.669260700389105,
      "grad_norm": 2.15625,
      "learning_rate": 9.374625257950354e-06,
      "loss": 1.2558,
      "step": 215
    },
    {
      "epoch": 0.6723735408560312,
      "grad_norm": 2.328125,
      "learning_rate": 9.368641231486682e-06,
      "loss": 1.3062,
      "step": 216
    },
    {
      "epoch": 0.6754863813229572,
      "grad_norm": 2.125,
      "learning_rate": 9.36263064019248e-06,
      "loss": 1.3384,
      "step": 217
    },
    {
      "epoch": 0.6785992217898833,
      "grad_norm": 2.15625,
      "learning_rate": 9.356593520616948e-06,
      "loss": 1.2555,
      "step": 218
    },
    {
      "epoch": 0.6817120622568094,
      "grad_norm": 2.03125,
      "learning_rate": 9.350529909470598e-06,
      "loss": 1.246,
      "step": 219
    },
    {
      "epoch": 0.6848249027237354,
      "grad_norm": 1.9921875,
      "learning_rate": 9.344439843625034e-06,
      "loss": 1.2964,
      "step": 220
    },
    {
      "epoch": 0.6879377431906615,
      "grad_norm": 2.03125,
      "learning_rate": 9.338323360112722e-06,
      "loss": 1.2676,
      "step": 221
    },
    {
      "epoch": 0.6910505836575875,
      "grad_norm": 2.21875,
      "learning_rate": 9.332180496126771e-06,
      "loss": 1.2853,
      "step": 222
    },
    {
      "epoch": 0.6941634241245136,
      "grad_norm": 2.171875,
      "learning_rate": 9.326011289020706e-06,
      "loss": 1.2989,
      "step": 223
    },
    {
      "epoch": 0.6972762645914397,
      "grad_norm": 2.171875,
      "learning_rate": 9.319815776308235e-06,
      "loss": 1.3034,
      "step": 224
    },
    {
      "epoch": 0.7003891050583657,
      "grad_norm": 2.09375,
      "learning_rate": 9.313593995663029e-06,
      "loss": 1.3238,
      "step": 225
    },
    {
      "epoch": 0.7035019455252919,
      "grad_norm": 2.03125,
      "learning_rate": 9.307345984918484e-06,
      "loss": 1.3038,
      "step": 226
    },
    {
      "epoch": 0.7066147859922179,
      "grad_norm": 2.125,
      "learning_rate": 9.301071782067504e-06,
      "loss": 1.2933,
      "step": 227
    },
    {
      "epoch": 0.709727626459144,
      "grad_norm": 2.140625,
      "learning_rate": 9.294771425262254e-06,
      "loss": 1.3431,
      "step": 228
    },
    {
      "epoch": 0.7128404669260701,
      "grad_norm": 2.1875,
      "learning_rate": 9.288444952813941e-06,
      "loss": 1.2981,
      "step": 229
    },
    {
      "epoch": 0.7159533073929961,
      "grad_norm": 2.203125,
      "learning_rate": 9.282092403192574e-06,
      "loss": 1.2719,
      "step": 230
    },
    {
      "epoch": 0.7190661478599222,
      "grad_norm": 2.078125,
      "learning_rate": 9.275713815026732e-06,
      "loss": 1.3189,
      "step": 231
    },
    {
      "epoch": 0.7190661478599222,
      "eval_loss": 3.2407894134521484,
      "eval_runtime": 42.9656,
      "eval_samples_per_second": 114.231,
      "eval_steps_per_second": 11.428,
      "step": 231
    },
    {
      "epoch": 0.7221789883268482,
      "grad_norm": 2.0625,
      "learning_rate": 9.269309227103329e-06,
      "loss": 1.2471,
      "step": 232
    },
    {
      "epoch": 0.7252918287937743,
      "grad_norm": 2.046875,
      "learning_rate": 9.262878678367377e-06,
      "loss": 1.2623,
      "step": 233
    },
    {
      "epoch": 0.7284046692607004,
      "grad_norm": 2.046875,
      "learning_rate": 9.256422207921757e-06,
      "loss": 1.2689,
      "step": 234
    },
    {
      "epoch": 0.7315175097276264,
      "grad_norm": 2.140625,
      "learning_rate": 9.249939855026966e-06,
      "loss": 1.25,
      "step": 235
    },
    {
      "epoch": 0.7346303501945526,
      "grad_norm": 2.171875,
      "learning_rate": 9.243431659100892e-06,
      "loss": 1.3142,
      "step": 236
    },
    {
      "epoch": 0.7377431906614786,
      "grad_norm": 2.1875,
      "learning_rate": 9.236897659718568e-06,
      "loss": 1.288,
      "step": 237
    },
    {
      "epoch": 0.7408560311284047,
      "grad_norm": 2.078125,
      "learning_rate": 9.230337896611934e-06,
      "loss": 1.2822,
      "step": 238
    },
    {
      "epoch": 0.7439688715953308,
      "grad_norm": 1.984375,
      "learning_rate": 9.223752409669592e-06,
      "loss": 1.2812,
      "step": 239
    },
    {
      "epoch": 0.7470817120622568,
      "grad_norm": 2.078125,
      "learning_rate": 9.217141238936567e-06,
      "loss": 1.3169,
      "step": 240
    },
    {
      "epoch": 0.7501945525291829,
      "grad_norm": 2.1875,
      "learning_rate": 9.21050442461406e-06,
      "loss": 1.2667,
      "step": 241
    },
    {
      "epoch": 0.7533073929961089,
      "grad_norm": 2.046875,
      "learning_rate": 9.203842007059208e-06,
      "loss": 1.2509,
      "step": 242
    },
    {
      "epoch": 0.756420233463035,
      "grad_norm": 2.03125,
      "learning_rate": 9.197154026784833e-06,
      "loss": 1.2419,
      "step": 243
    },
    {
      "epoch": 0.7595330739299611,
      "grad_norm": 2.21875,
      "learning_rate": 9.190440524459203e-06,
      "loss": 1.2279,
      "step": 244
    },
    {
      "epoch": 0.7626459143968871,
      "grad_norm": 1.9609375,
      "learning_rate": 9.18370154090578e-06,
      "loss": 1.2637,
      "step": 245
    },
    {
      "epoch": 0.7657587548638133,
      "grad_norm": 2.140625,
      "learning_rate": 9.176937117102967e-06,
      "loss": 1.2851,
      "step": 246
    },
    {
      "epoch": 0.7688715953307393,
      "grad_norm": 2.09375,
      "learning_rate": 9.170147294183866e-06,
      "loss": 1.2998,
      "step": 247
    },
    {
      "epoch": 0.7719844357976654,
      "grad_norm": 2.046875,
      "learning_rate": 9.163332113436031e-06,
      "loss": 1.2498,
      "step": 248
    },
    {
      "epoch": 0.7750972762645915,
      "grad_norm": 2.125,
      "learning_rate": 9.15649161630121e-06,
      "loss": 1.3203,
      "step": 249
    },
    {
      "epoch": 0.7782101167315175,
      "grad_norm": 2.09375,
      "learning_rate": 9.149625844375088e-06,
      "loss": 1.269,
      "step": 250
    },
    {
      "epoch": 0.7813229571984436,
      "grad_norm": 2.15625,
      "learning_rate": 9.14273483940705e-06,
      "loss": 1.2679,
      "step": 251
    },
    {
      "epoch": 0.7844357976653696,
      "grad_norm": 2.03125,
      "learning_rate": 9.135818643299911e-06,
      "loss": 1.2845,
      "step": 252
    },
    {
      "epoch": 0.7844357976653696,
      "eval_loss": 3.2150230407714844,
      "eval_runtime": 43.6603,
      "eval_samples_per_second": 112.413,
      "eval_steps_per_second": 11.246,
      "step": 252
    },
    {
      "epoch": 0.7875486381322957,
      "grad_norm": 2.203125,
      "learning_rate": 9.128877298109675e-06,
      "loss": 1.3251,
      "step": 253
    },
    {
      "epoch": 0.7906614785992218,
      "grad_norm": 2.046875,
      "learning_rate": 9.121910846045267e-06,
      "loss": 1.2985,
      "step": 254
    },
    {
      "epoch": 0.7937743190661478,
      "grad_norm": 2.15625,
      "learning_rate": 9.114919329468283e-06,
      "loss": 1.2773,
      "step": 255
    },
    {
      "epoch": 0.796887159533074,
      "grad_norm": 2.09375,
      "learning_rate": 9.107902790892734e-06,
      "loss": 1.3517,
      "step": 256
    },
    {
      "epoch": 0.8,
      "grad_norm": 2.0625,
      "learning_rate": 9.10086127298478e-06,
      "loss": 1.2972,
      "step": 257
    },
    {
      "epoch": 0.8031128404669261,
      "grad_norm": 2.015625,
      "learning_rate": 9.093794818562482e-06,
      "loss": 1.2717,
      "step": 258
    },
    {
      "epoch": 0.8062256809338522,
      "grad_norm": 2.140625,
      "learning_rate": 9.086703470595528e-06,
      "loss": 1.2771,
      "step": 259
    },
    {
      "epoch": 0.8093385214007782,
      "grad_norm": 1.984375,
      "learning_rate": 9.079587272204983e-06,
      "loss": 1.2546,
      "step": 260
    },
    {
      "epoch": 0.8124513618677043,
      "grad_norm": 2.0625,
      "learning_rate": 9.072446266663021e-06,
      "loss": 1.2385,
      "step": 261
    },
    {
      "epoch": 0.8155642023346303,
      "grad_norm": 2.15625,
      "learning_rate": 9.065280497392663e-06,
      "loss": 1.2407,
      "step": 262
    },
    {
      "epoch": 0.8186770428015564,
      "grad_norm": 2.15625,
      "learning_rate": 9.058090007967515e-06,
      "loss": 1.3042,
      "step": 263
    },
    {
      "epoch": 0.8217898832684825,
      "grad_norm": 2.09375,
      "learning_rate": 9.050874842111498e-06,
      "loss": 1.2714,
      "step": 264
    },
    {
      "epoch": 0.8249027237354085,
      "grad_norm": 2.078125,
      "learning_rate": 9.04363504369859e-06,
      "loss": 1.2733,
      "step": 265
    },
    {
      "epoch": 0.8280155642023346,
      "grad_norm": 2.03125,
      "learning_rate": 9.03637065675255e-06,
      "loss": 1.2357,
      "step": 266
    },
    {
      "epoch": 0.8311284046692607,
      "grad_norm": 2.0625,
      "learning_rate": 9.02908172544666e-06,
      "loss": 1.2595,
      "step": 267
    },
    {
      "epoch": 0.8342412451361868,
      "grad_norm": 2.0625,
      "learning_rate": 9.021768294103444e-06,
      "loss": 1.3116,
      "step": 268
    },
    {
      "epoch": 0.8373540856031129,
      "grad_norm": 2.03125,
      "learning_rate": 9.014430407194413e-06,
      "loss": 1.2891,
      "step": 269
    },
    {
      "epoch": 0.8404669260700389,
      "grad_norm": 2.109375,
      "learning_rate": 9.007068109339783e-06,
      "loss": 1.2659,
      "step": 270
    },
    {
      "epoch": 0.843579766536965,
      "grad_norm": 2.03125,
      "learning_rate": 8.999681445308213e-06,
      "loss": 1.3067,
      "step": 271
    },
    {
      "epoch": 0.846692607003891,
      "grad_norm": 2.078125,
      "learning_rate": 8.99227046001652e-06,
      "loss": 1.3106,
      "step": 272
    },
    {
      "epoch": 0.8498054474708171,
      "grad_norm": 2.28125,
      "learning_rate": 8.98483519852942e-06,
      "loss": 1.2808,
      "step": 273
    },
    {
      "epoch": 0.8498054474708171,
      "eval_loss": 3.245403289794922,
      "eval_runtime": 246.1743,
      "eval_samples_per_second": 19.937,
      "eval_steps_per_second": 1.995,
      "step": 273
    },
    {
      "epoch": 0.8529182879377432,
      "grad_norm": 2.109375,
      "learning_rate": 8.977375706059248e-06,
      "loss": 1.3074,
      "step": 274
    },
    {
      "epoch": 0.8560311284046692,
      "grad_norm": 2.171875,
      "learning_rate": 8.969892027965677e-06,
      "loss": 1.2758,
      "step": 275
    },
    {
      "epoch": 0.8591439688715953,
      "grad_norm": 2.078125,
      "learning_rate": 8.962384209755453e-06,
      "loss": 1.2666,
      "step": 276
    },
    {
      "epoch": 0.8622568093385214,
      "grad_norm": 2.125,
      "learning_rate": 8.954852297082112e-06,
      "loss": 1.288,
      "step": 277
    },
    {
      "epoch": 0.8653696498054475,
      "grad_norm": 2.125,
      "learning_rate": 8.947296335745704e-06,
      "loss": 1.2702,
      "step": 278
    },
    {
      "epoch": 0.8684824902723736,
      "grad_norm": 2.15625,
      "learning_rate": 8.93971637169251e-06,
      "loss": 1.2791,
      "step": 279
    },
    {
      "epoch": 0.8715953307392996,
      "grad_norm": 2.296875,
      "learning_rate": 8.932112451014773e-06,
      "loss": 1.2467,
      "step": 280
    },
    {
      "epoch": 0.8747081712062257,
      "grad_norm": 2.203125,
      "learning_rate": 8.92448461995041e-06,
      "loss": 1.3387,
      "step": 281
    },
    {
      "epoch": 0.8778210116731517,
      "grad_norm": 2.09375,
      "learning_rate": 8.916832924882727e-06,
      "loss": 1.2826,
      "step": 282
    },
    {
      "epoch": 0.8809338521400778,
      "grad_norm": 2.078125,
      "learning_rate": 8.90915741234015e-06,
      "loss": 1.2473,
      "step": 283
    },
    {
      "epoch": 0.8840466926070039,
      "grad_norm": 2.078125,
      "learning_rate": 8.901458128995926e-06,
      "loss": 1.3222,
      "step": 284
    },
    {
      "epoch": 0.8871595330739299,
      "grad_norm": 2.109375,
      "learning_rate": 8.893735121667855e-06,
      "loss": 1.2884,
      "step": 285
    },
    {
      "epoch": 0.890272373540856,
      "grad_norm": 2.125,
      "learning_rate": 8.885988437317993e-06,
      "loss": 1.2405,
      "step": 286
    },
    {
      "epoch": 0.8933852140077821,
      "grad_norm": 2.15625,
      "learning_rate": 8.878218123052373e-06,
      "loss": 1.3235,
      "step": 287
    },
    {
      "epoch": 0.8964980544747082,
      "grad_norm": 2.171875,
      "learning_rate": 8.870424226120715e-06,
      "loss": 1.2807,
      "step": 288
    },
    {
      "epoch": 0.8996108949416343,
      "grad_norm": 2.09375,
      "learning_rate": 8.862606793916147e-06,
      "loss": 1.2793,
      "step": 289
    },
    {
      "epoch": 0.9027237354085603,
      "grad_norm": 2.140625,
      "learning_rate": 8.854765873974898e-06,
      "loss": 1.2713,
      "step": 290
    },
    {
      "epoch": 0.9058365758754864,
      "grad_norm": 2.03125,
      "learning_rate": 8.846901513976034e-06,
      "loss": 1.2734,
      "step": 291
    },
    {
      "epoch": 0.9089494163424124,
      "grad_norm": 2.0625,
      "learning_rate": 8.83901376174115e-06,
      "loss": 1.2414,
      "step": 292
    },
    {
      "epoch": 0.9120622568093385,
      "grad_norm": 2.109375,
      "learning_rate": 8.831102665234079e-06,
      "loss": 1.2459,
      "step": 293
    },
    {
      "epoch": 0.9151750972762646,
      "grad_norm": 2.078125,
      "learning_rate": 8.823168272560616e-06,
      "loss": 1.2534,
      "step": 294
    },
    {
      "epoch": 0.9151750972762646,
      "eval_loss": 3.2486953735351562,
      "eval_runtime": 42.8446,
      "eval_samples_per_second": 114.554,
      "eval_steps_per_second": 11.46,
      "step": 294
    },
    {
      "epoch": 0.9182879377431906,
      "grad_norm": 1.984375,
      "learning_rate": 8.81521063196821e-06,
      "loss": 1.248,
      "step": 295
    },
    {
      "epoch": 0.9214007782101167,
      "grad_norm": 2.109375,
      "learning_rate": 8.807229791845673e-06,
      "loss": 1.2233,
      "step": 296
    },
    {
      "epoch": 0.9245136186770428,
      "grad_norm": 2.109375,
      "learning_rate": 8.799225800722895e-06,
      "loss": 1.2105,
      "step": 297
    },
    {
      "epoch": 0.9276264591439689,
      "grad_norm": 2.046875,
      "learning_rate": 8.791198707270538e-06,
      "loss": 1.2302,
      "step": 298
    },
    {
      "epoch": 0.930739299610895,
      "grad_norm": 2.140625,
      "learning_rate": 8.783148560299749e-06,
      "loss": 1.27,
      "step": 299
    },
    {
      "epoch": 0.933852140077821,
      "grad_norm": 2.21875,
      "learning_rate": 8.775075408761851e-06,
      "loss": 1.2852,
      "step": 300
    },
    {
      "epoch": 0.9369649805447471,
      "grad_norm": 2.125,
      "learning_rate": 8.766979301748063e-06,
      "loss": 1.3053,
      "step": 301
    },
    {
      "epoch": 0.9400778210116731,
      "grad_norm": 2.140625,
      "learning_rate": 8.758860288489184e-06,
      "loss": 1.2874,
      "step": 302
    },
    {
      "epoch": 0.9431906614785992,
      "grad_norm": 2.09375,
      "learning_rate": 8.750718418355307e-06,
      "loss": 1.3175,
      "step": 303
    },
    {
      "epoch": 0.9463035019455253,
      "grad_norm": 2.015625,
      "learning_rate": 8.742553740855507e-06,
      "loss": 1.2873,
      "step": 304
    },
    {
      "epoch": 0.9494163424124513,
      "grad_norm": 2.109375,
      "learning_rate": 8.734366305637549e-06,
      "loss": 1.2571,
      "step": 305
    },
    {
      "epoch": 0.9525291828793774,
      "grad_norm": 2.03125,
      "learning_rate": 8.726156162487586e-06,
      "loss": 1.2534,
      "step": 306
    },
    {
      "epoch": 0.9556420233463035,
      "grad_norm": 2.09375,
      "learning_rate": 8.71792336132985e-06,
      "loss": 1.3129,
      "step": 307
    },
    {
      "epoch": 0.9587548638132296,
      "grad_norm": 2.046875,
      "learning_rate": 8.709667952226352e-06,
      "loss": 1.3245,
      "step": 308
    },
    {
      "epoch": 0.9618677042801557,
      "grad_norm": 2.15625,
      "learning_rate": 8.701389985376578e-06,
      "loss": 1.2881,
      "step": 309
    },
    {
      "epoch": 0.9649805447470817,
      "grad_norm": 2.046875,
      "learning_rate": 8.693089511117183e-06,
      "loss": 1.2398,
      "step": 310
    },
    {
      "epoch": 0.9680933852140078,
      "grad_norm": 2.09375,
      "learning_rate": 8.684766579921684e-06,
      "loss": 1.2676,
      "step": 311
    },
    {
      "epoch": 0.9712062256809338,
      "grad_norm": 2.0625,
      "learning_rate": 8.676421242400156e-06,
      "loss": 1.2232,
      "step": 312
    },
    {
      "epoch": 0.9743190661478599,
      "grad_norm": 2.0625,
      "learning_rate": 8.668053549298922e-06,
      "loss": 1.2429,
      "step": 313
    },
    {
      "epoch": 0.977431906614786,
      "grad_norm": 2.015625,
      "learning_rate": 8.659663551500247e-06,
      "loss": 1.2742,
      "step": 314
    },
    {
      "epoch": 0.980544747081712,
      "grad_norm": 2.125,
      "learning_rate": 8.651251300022017e-06,
      "loss": 1.2902,
      "step": 315
    },
    {
      "epoch": 0.980544747081712,
      "eval_loss": 3.2123305797576904,
      "eval_runtime": 42.9008,
      "eval_samples_per_second": 114.403,
      "eval_steps_per_second": 11.445,
      "step": 315
    },
    {
      "epoch": 0.9836575875486381,
      "grad_norm": 2.0,
      "learning_rate": 8.642816846017448e-06,
      "loss": 1.2077,
      "step": 316
    },
    {
      "epoch": 0.9867704280155642,
      "grad_norm": 2.015625,
      "learning_rate": 8.634360240774764e-06,
      "loss": 1.2571,
      "step": 317
    },
    {
      "epoch": 0.9898832684824903,
      "grad_norm": 2.109375,
      "learning_rate": 8.625881535716883e-06,
      "loss": 1.2814,
      "step": 318
    },
    {
      "epoch": 0.9929961089494164,
      "grad_norm": 2.140625,
      "learning_rate": 8.617380782401108e-06,
      "loss": 1.2944,
      "step": 319
    },
    {
      "epoch": 0.9961089494163424,
      "grad_norm": 2.0625,
      "learning_rate": 8.608858032518815e-06,
      "loss": 1.2779,
      "step": 320
    },
    {
      "epoch": 0.9992217898832685,
      "grad_norm": 1.921875,
      "learning_rate": 8.600313337895134e-06,
      "loss": 1.2236,
      "step": 321
    },
    {
      "epoch": 1.0,
      "grad_norm": 4.4375,
      "learning_rate": 8.591746750488639e-06,
      "loss": 1.3442,
      "step": 322
    },
    {
      "epoch": 1.003112840466926,
      "grad_norm": 2.0,
      "learning_rate": 8.583158322391027e-06,
      "loss": 1.2512,
      "step": 323
    },
    {
      "epoch": 1.0062256809338521,
      "grad_norm": 1.953125,
      "learning_rate": 8.57454810582681e-06,
      "loss": 1.1895,
      "step": 324
    },
    {
      "epoch": 1.0093385214007782,
      "grad_norm": 2.046875,
      "learning_rate": 8.565916153152982e-06,
      "loss": 1.1798,
      "step": 325
    },
    {
      "epoch": 1.0124513618677042,
      "grad_norm": 2.0625,
      "learning_rate": 8.557262516858718e-06,
      "loss": 1.26,
      "step": 326
    },
    {
      "epoch": 1.0155642023346303,
      "grad_norm": 2.109375,
      "learning_rate": 8.54858724956504e-06,
      "loss": 1.2325,
      "step": 327
    },
    {
      "epoch": 1.0186770428015564,
      "grad_norm": 2.015625,
      "learning_rate": 8.539890404024511e-06,
      "loss": 1.1924,
      "step": 328
    },
    {
      "epoch": 1.0217898832684824,
      "grad_norm": 2.140625,
      "learning_rate": 8.5311720331209e-06,
      "loss": 1.2143,
      "step": 329
    },
    {
      "epoch": 1.0249027237354085,
      "grad_norm": 1.984375,
      "learning_rate": 8.522432189868875e-06,
      "loss": 1.1977,
      "step": 330
    },
    {
      "epoch": 1.0280155642023345,
      "grad_norm": 2.109375,
      "learning_rate": 8.513670927413664e-06,
      "loss": 1.2362,
      "step": 331
    },
    {
      "epoch": 1.0311284046692606,
      "grad_norm": 1.9609375,
      "learning_rate": 8.504888299030748e-06,
      "loss": 1.2298,
      "step": 332
    },
    {
      "epoch": 1.0342412451361869,
      "grad_norm": 2.09375,
      "learning_rate": 8.496084358125528e-06,
      "loss": 1.1911,
      "step": 333
    },
    {
      "epoch": 1.037354085603113,
      "grad_norm": 2.109375,
      "learning_rate": 8.487259158232999e-06,
      "loss": 1.2295,
      "step": 334
    },
    {
      "epoch": 1.040466926070039,
      "grad_norm": 2.09375,
      "learning_rate": 8.478412753017433e-06,
      "loss": 1.2631,
      "step": 335
    },
    {
      "epoch": 1.043579766536965,
      "grad_norm": 2.0,
      "learning_rate": 8.469545196272042e-06,
      "loss": 1.1715,
      "step": 336
    },
    {
      "epoch": 1.043579766536965,
      "eval_loss": 3.2468361854553223,
      "eval_runtime": 43.5607,
      "eval_samples_per_second": 112.67,
      "eval_steps_per_second": 11.272,
      "step": 336
    },
    {
      "epoch": 1.046692607003891,
      "grad_norm": 2.0625,
      "learning_rate": 8.460656541918664e-06,
      "loss": 1.1867,
      "step": 337
    },
    {
      "epoch": 1.0498054474708172,
      "grad_norm": 2.15625,
      "learning_rate": 8.451746844007415e-06,
      "loss": 1.1942,
      "step": 338
    },
    {
      "epoch": 1.0529182879377432,
      "grad_norm": 2.09375,
      "learning_rate": 8.442816156716386e-06,
      "loss": 1.2204,
      "step": 339
    },
    {
      "epoch": 1.0560311284046693,
      "grad_norm": 2.125,
      "learning_rate": 8.433864534351291e-06,
      "loss": 1.1683,
      "step": 340
    },
    {
      "epoch": 1.0591439688715953,
      "grad_norm": 2.03125,
      "learning_rate": 8.424892031345148e-06,
      "loss": 1.1914,
      "step": 341
    },
    {
      "epoch": 1.0622568093385214,
      "grad_norm": 2.0625,
      "learning_rate": 8.41589870225795e-06,
      "loss": 1.2484,
      "step": 342
    },
    {
      "epoch": 1.0653696498054475,
      "grad_norm": 2.109375,
      "learning_rate": 8.406884601776325e-06,
      "loss": 1.2643,
      "step": 343
    },
    {
      "epoch": 1.0684824902723735,
      "grad_norm": 2.0,
      "learning_rate": 8.397849784713206e-06,
      "loss": 1.2014,
      "step": 344
    },
    {
      "epoch": 1.0715953307392996,
      "grad_norm": 2.125,
      "learning_rate": 8.388794306007507e-06,
      "loss": 1.2362,
      "step": 345
    },
    {
      "epoch": 1.0747081712062256,
      "grad_norm": 2.15625,
      "learning_rate": 8.379718220723772e-06,
      "loss": 1.1732,
      "step": 346
    },
    {
      "epoch": 1.0778210116731517,
      "grad_norm": 2.109375,
      "learning_rate": 8.370621584051858e-06,
      "loss": 1.2718,
      "step": 347
    },
    {
      "epoch": 1.0809338521400778,
      "grad_norm": 2.0625,
      "learning_rate": 8.361504451306585e-06,
      "loss": 1.2143,
      "step": 348
    },
    {
      "epoch": 1.0840466926070038,
      "grad_norm": 2.046875,
      "learning_rate": 8.352366877927407e-06,
      "loss": 1.2278,
      "step": 349
    },
    {
      "epoch": 1.0871595330739299,
      "grad_norm": 2.125,
      "learning_rate": 8.343208919478075e-06,
      "loss": 1.2641,
      "step": 350
    },
    {
      "epoch": 1.090272373540856,
      "grad_norm": 2.09375,
      "learning_rate": 8.334030631646295e-06,
      "loss": 1.3161,
      "step": 351
    },
    {
      "epoch": 1.0933852140077822,
      "grad_norm": 2.03125,
      "learning_rate": 8.324832070243393e-06,
      "loss": 1.1935,
      "step": 352
    },
    {
      "epoch": 1.0964980544747083,
      "grad_norm": 2.078125,
      "learning_rate": 8.315613291203977e-06,
      "loss": 1.2077,
      "step": 353
    },
    {
      "epoch": 1.0996108949416343,
      "grad_norm": 2.265625,
      "learning_rate": 8.30637435058559e-06,
      "loss": 1.223,
      "step": 354
    },
    {
      "epoch": 1.1027237354085604,
      "grad_norm": 2.125,
      "learning_rate": 8.297115304568375e-06,
      "loss": 1.2927,
      "step": 355
    },
    {
      "epoch": 1.1058365758754864,
      "grad_norm": 2.03125,
      "learning_rate": 8.287836209454736e-06,
      "loss": 1.2412,
      "step": 356
    },
    {
      "epoch": 1.1089494163424125,
      "grad_norm": 2.171875,
      "learning_rate": 8.278537121668985e-06,
      "loss": 1.2641,
      "step": 357
    },
    {
      "epoch": 1.1089494163424125,
      "eval_loss": 3.286686420440674,
      "eval_runtime": 42.981,
      "eval_samples_per_second": 114.19,
      "eval_steps_per_second": 11.424,
      "step": 357
    },
    {
      "epoch": 1.1120622568093386,
      "grad_norm": 1.96875,
      "learning_rate": 8.269218097757007e-06,
      "loss": 1.2338,
      "step": 358
    },
    {
      "epoch": 1.1151750972762646,
      "grad_norm": 2.3125,
      "learning_rate": 8.259879194385923e-06,
      "loss": 1.2401,
      "step": 359
    },
    {
      "epoch": 1.1182879377431907,
      "grad_norm": 2.03125,
      "learning_rate": 8.250520468343722e-06,
      "loss": 1.2099,
      "step": 360
    },
    {
      "epoch": 1.1214007782101167,
      "grad_norm": 2.203125,
      "learning_rate": 8.241141976538944e-06,
      "loss": 1.2651,
      "step": 361
    },
    {
      "epoch": 1.1245136186770428,
      "grad_norm": 2.125,
      "learning_rate": 8.231743776000312e-06,
      "loss": 1.2307,
      "step": 362
    },
    {
      "epoch": 1.1276264591439689,
      "grad_norm": 2.078125,
      "learning_rate": 8.222325923876399e-06,
      "loss": 1.1858,
      "step": 363
    },
    {
      "epoch": 1.130739299610895,
      "grad_norm": 2.125,
      "learning_rate": 8.212888477435272e-06,
      "loss": 1.2304,
      "step": 364
    },
    {
      "epoch": 1.133852140077821,
      "grad_norm": 2.0625,
      "learning_rate": 8.203431494064152e-06,
      "loss": 1.2025,
      "step": 365
    },
    {
      "epoch": 1.136964980544747,
      "grad_norm": 2.265625,
      "learning_rate": 8.193955031269055e-06,
      "loss": 1.1993,
      "step": 366
    },
    {
      "epoch": 1.140077821011673,
      "grad_norm": 2.015625,
      "learning_rate": 8.184459146674447e-06,
      "loss": 1.24,
      "step": 367
    },
    {
      "epoch": 1.1431906614785992,
      "grad_norm": 1.9921875,
      "learning_rate": 8.1749438980229e-06,
      "loss": 1.1573,
      "step": 368
    },
    {
      "epoch": 1.1463035019455252,
      "grad_norm": 2.0625,
      "learning_rate": 8.165409343174729e-06,
      "loss": 1.1965,
      "step": 369
    },
    {
      "epoch": 1.1494163424124513,
      "grad_norm": 2.09375,
      "learning_rate": 8.15585554010765e-06,
      "loss": 1.2264,
      "step": 370
    },
    {
      "epoch": 1.1525291828793773,
      "grad_norm": 1.96875,
      "learning_rate": 8.14628254691642e-06,
      "loss": 1.1693,
      "step": 371
    },
    {
      "epoch": 1.1556420233463034,
      "grad_norm": 2.28125,
      "learning_rate": 8.136690421812491e-06,
      "loss": 1.2597,
      "step": 372
    },
    {
      "epoch": 1.1587548638132295,
      "grad_norm": 2.109375,
      "learning_rate": 8.12707922312365e-06,
      "loss": 1.212,
      "step": 373
    },
    {
      "epoch": 1.1618677042801557,
      "grad_norm": 2.09375,
      "learning_rate": 8.117449009293668e-06,
      "loss": 1.2231,
      "step": 374
    },
    {
      "epoch": 1.1649805447470818,
      "grad_norm": 2.21875,
      "learning_rate": 8.107799838881942e-06,
      "loss": 1.235,
      "step": 375
    },
    {
      "epoch": 1.1680933852140079,
      "grad_norm": 2.0625,
      "learning_rate": 8.09813177056314e-06,
      "loss": 1.2185,
      "step": 376
    },
    {
      "epoch": 1.171206225680934,
      "grad_norm": 1.96875,
      "learning_rate": 8.088444863126845e-06,
      "loss": 1.1977,
      "step": 377
    },
    {
      "epoch": 1.17431906614786,
      "grad_norm": 2.109375,
      "learning_rate": 8.078739175477198e-06,
      "loss": 1.1966,
      "step": 378
    },
    {
      "epoch": 1.17431906614786,
      "eval_loss": 3.2749640941619873,
      "eval_runtime": 42.9779,
      "eval_samples_per_second": 114.198,
      "eval_steps_per_second": 11.424,
      "step": 378
    },
    {
      "epoch": 1.177431906614786,
      "grad_norm": 2.046875,
      "learning_rate": 8.069014766632536e-06,
      "loss": 1.2058,
      "step": 379
    },
    {
      "epoch": 1.180544747081712,
      "grad_norm": 2.125,
      "learning_rate": 8.059271695725038e-06,
      "loss": 1.1528,
      "step": 380
    },
    {
      "epoch": 1.1836575875486381,
      "grad_norm": 1.984375,
      "learning_rate": 8.049510022000365e-06,
      "loss": 1.1938,
      "step": 381
    },
    {
      "epoch": 1.1867704280155642,
      "grad_norm": 2.09375,
      "learning_rate": 8.039729804817294e-06,
      "loss": 1.1424,
      "step": 382
    },
    {
      "epoch": 1.1898832684824903,
      "grad_norm": 2.0625,
      "learning_rate": 8.029931103647362e-06,
      "loss": 1.2536,
      "step": 383
    },
    {
      "epoch": 1.1929961089494163,
      "grad_norm": 2.109375,
      "learning_rate": 8.020113978074504e-06,
      "loss": 1.1442,
      "step": 384
    },
    {
      "epoch": 1.1961089494163424,
      "grad_norm": 2.203125,
      "learning_rate": 8.010278487794692e-06,
      "loss": 1.2111,
      "step": 385
    },
    {
      "epoch": 1.1992217898832684,
      "grad_norm": 2.125,
      "learning_rate": 8.000424692615568e-06,
      "loss": 1.2681,
      "step": 386
    },
    {
      "epoch": 1.2023346303501945,
      "grad_norm": 2.125,
      "learning_rate": 7.99055265245608e-06,
      "loss": 1.2749,
      "step": 387
    },
    {
      "epoch": 1.2054474708171206,
      "grad_norm": 2.09375,
      "learning_rate": 7.980662427346127e-06,
      "loss": 1.2126,
      "step": 388
    },
    {
      "epoch": 1.2085603112840466,
      "grad_norm": 2.15625,
      "learning_rate": 7.970754077426181e-06,
      "loss": 1.2009,
      "step": 389
    },
    {
      "epoch": 1.2116731517509727,
      "grad_norm": 2.109375,
      "learning_rate": 7.96082766294693e-06,
      "loss": 1.2169,
      "step": 390
    },
    {
      "epoch": 1.2147859922178987,
      "grad_norm": 2.15625,
      "learning_rate": 7.950883244268907e-06,
      "loss": 1.1784,
      "step": 391
    },
    {
      "epoch": 1.217898832684825,
      "grad_norm": 2.125,
      "learning_rate": 7.940920881862126e-06,
      "loss": 1.2622,
      "step": 392
    },
    {
      "epoch": 1.221011673151751,
      "grad_norm": 2.15625,
      "learning_rate": 7.930940636305716e-06,
      "loss": 1.2552,
      "step": 393
    },
    {
      "epoch": 1.2241245136186771,
      "grad_norm": 2.046875,
      "learning_rate": 7.920942568287542e-06,
      "loss": 1.1769,
      "step": 394
    },
    {
      "epoch": 1.2272373540856032,
      "grad_norm": 2.1875,
      "learning_rate": 7.910926738603855e-06,
      "loss": 1.2153,
      "step": 395
    },
    {
      "epoch": 1.2303501945525293,
      "grad_norm": 2.09375,
      "learning_rate": 7.900893208158899e-06,
      "loss": 1.1812,
      "step": 396
    },
    {
      "epoch": 1.2334630350194553,
      "grad_norm": 2.046875,
      "learning_rate": 7.890842037964566e-06,
      "loss": 1.2756,
      "step": 397
    },
    {
      "epoch": 1.2365758754863814,
      "grad_norm": 2.15625,
      "learning_rate": 7.880773289139998e-06,
      "loss": 1.247,
      "step": 398
    },
    {
      "epoch": 1.2396887159533074,
      "grad_norm": 2.046875,
      "learning_rate": 7.870687022911238e-06,
      "loss": 1.2555,
      "step": 399
    },
    {
      "epoch": 1.2396887159533074,
      "eval_loss": 3.2569496631622314,
      "eval_runtime": 42.989,
      "eval_samples_per_second": 114.169,
      "eval_steps_per_second": 11.422,
      "step": 399
    },
    {
      "epoch": 1.2428015564202335,
      "grad_norm": 2.03125,
      "learning_rate": 7.860583300610849e-06,
      "loss": 1.2343,
      "step": 400
    },
    {
      "epoch": 1.2459143968871595,
      "grad_norm": 1.9921875,
      "learning_rate": 7.850462183677536e-06,
      "loss": 1.1612,
      "step": 401
    },
    {
      "epoch": 1.2490272373540856,
      "grad_norm": 2.109375,
      "learning_rate": 7.84032373365578e-06,
      "loss": 1.2072,
      "step": 402
    },
    {
      "epoch": 1.2521400778210117,
      "grad_norm": 2.125,
      "learning_rate": 7.830168012195461e-06,
      "loss": 1.1671,
      "step": 403
    },
    {
      "epoch": 1.2552529182879377,
      "grad_norm": 2.09375,
      "learning_rate": 7.819995081051483e-06,
      "loss": 1.2154,
      "step": 404
    },
    {
      "epoch": 1.2583657587548638,
      "grad_norm": 2.0625,
      "learning_rate": 7.8098050020834e-06,
      "loss": 1.2939,
      "step": 405
    },
    {
      "epoch": 1.2614785992217898,
      "grad_norm": 2.0625,
      "learning_rate": 7.799597837255033e-06,
      "loss": 1.2114,
      "step": 406
    },
    {
      "epoch": 1.264591439688716,
      "grad_norm": 2.046875,
      "learning_rate": 7.789373648634107e-06,
      "loss": 1.1948,
      "step": 407
    },
    {
      "epoch": 1.267704280155642,
      "grad_norm": 2.03125,
      "learning_rate": 7.779132498391857e-06,
      "loss": 1.265,
      "step": 408
    },
    {
      "epoch": 1.270817120622568,
      "grad_norm": 2.0625,
      "learning_rate": 7.768874448802665e-06,
      "loss": 1.2301,
      "step": 409
    },
    {
      "epoch": 1.273929961089494,
      "grad_norm": 2.1875,
      "learning_rate": 7.758599562243673e-06,
      "loss": 1.2844,
      "step": 410
    },
    {
      "epoch": 1.2770428015564201,
      "grad_norm": 2.171875,
      "learning_rate": 7.748307901194401e-06,
      "loss": 1.2818,
      "step": 411
    },
    {
      "epoch": 1.2801556420233462,
      "grad_norm": 2.03125,
      "learning_rate": 7.737999528236377e-06,
      "loss": 1.2296,
      "step": 412
    },
    {
      "epoch": 1.2832684824902723,
      "grad_norm": 2.125,
      "learning_rate": 7.727674506052744e-06,
      "loss": 1.188,
      "step": 413
    },
    {
      "epoch": 1.2863813229571983,
      "grad_norm": 2.140625,
      "learning_rate": 7.717332897427893e-06,
      "loss": 1.2312,
      "step": 414
    },
    {
      "epoch": 1.2894941634241246,
      "grad_norm": 2.1875,
      "learning_rate": 7.706974765247071e-06,
      "loss": 1.1914,
      "step": 415
    },
    {
      "epoch": 1.2926070038910507,
      "grad_norm": 2.234375,
      "learning_rate": 7.696600172495997e-06,
      "loss": 1.1891,
      "step": 416
    },
    {
      "epoch": 1.2957198443579767,
      "grad_norm": 2.171875,
      "learning_rate": 7.68620918226049e-06,
      "loss": 1.2064,
      "step": 417
    },
    {
      "epoch": 1.2988326848249028,
      "grad_norm": 2.078125,
      "learning_rate": 7.675801857726077e-06,
      "loss": 1.2027,
      "step": 418
    },
    {
      "epoch": 1.3019455252918288,
      "grad_norm": 2.0625,
      "learning_rate": 7.665378262177608e-06,
      "loss": 1.2334,
      "step": 419
    },
    {
      "epoch": 1.305058365758755,
      "grad_norm": 2.078125,
      "learning_rate": 7.654938458998874e-06,
      "loss": 1.2654,
      "step": 420
    },
    {
      "epoch": 1.305058365758755,
      "eval_loss": 3.2143044471740723,
      "eval_runtime": 43.0867,
      "eval_samples_per_second": 113.91,
      "eval_steps_per_second": 11.396,
      "step": 420
    },
    {
      "epoch": 1.308171206225681,
      "grad_norm": 2.046875,
      "learning_rate": 7.644482511672229e-06,
      "loss": 1.2291,
      "step": 421
    },
    {
      "epoch": 1.311284046692607,
      "grad_norm": 2.15625,
      "learning_rate": 7.634010483778185e-06,
      "loss": 1.2156,
      "step": 422
    },
    {
      "epoch": 1.314396887159533,
      "grad_norm": 2.03125,
      "learning_rate": 7.62352243899504e-06,
      "loss": 1.168,
      "step": 423
    },
    {
      "epoch": 1.3175097276264591,
      "grad_norm": 2.09375,
      "learning_rate": 7.613018441098494e-06,
      "loss": 1.2032,
      "step": 424
    },
    {
      "epoch": 1.3206225680933852,
      "grad_norm": 2.078125,
      "learning_rate": 7.602498553961248e-06,
      "loss": 1.2162,
      "step": 425
    },
    {
      "epoch": 1.3237354085603112,
      "grad_norm": 2.203125,
      "learning_rate": 7.591962841552627e-06,
      "loss": 1.218,
      "step": 426
    },
    {
      "epoch": 1.3268482490272373,
      "grad_norm": 2.15625,
      "learning_rate": 7.581411367938178e-06,
      "loss": 1.2663,
      "step": 427
    },
    {
      "epoch": 1.3299610894941634,
      "grad_norm": 2.171875,
      "learning_rate": 7.570844197279299e-06,
      "loss": 1.2426,
      "step": 428
    },
    {
      "epoch": 1.3330739299610894,
      "grad_norm": 2.171875,
      "learning_rate": 7.560261393832831e-06,
      "loss": 1.2026,
      "step": 429
    },
    {
      "epoch": 1.3361867704280157,
      "grad_norm": 2.171875,
      "learning_rate": 7.5496630219506805e-06,
      "loss": 1.2685,
      "step": 430
    },
    {
      "epoch": 1.3392996108949418,
      "grad_norm": 2.046875,
      "learning_rate": 7.539049146079418e-06,
      "loss": 1.2207,
      "step": 431
    },
    {
      "epoch": 1.3424124513618678,
      "grad_norm": 2.078125,
      "learning_rate": 7.528419830759892e-06,
      "loss": 1.2472,
      "step": 432
    },
    {
      "epoch": 1.3455252918287939,
      "grad_norm": 2.09375,
      "learning_rate": 7.517775140626838e-06,
      "loss": 1.2088,
      "step": 433
    },
    {
      "epoch": 1.34863813229572,
      "grad_norm": 2.046875,
      "learning_rate": 7.507115140408477e-06,
      "loss": 1.1932,
      "step": 434
    },
    {
      "epoch": 1.351750972762646,
      "grad_norm": 2.078125,
      "learning_rate": 7.496439894926135e-06,
      "loss": 1.1883,
      "step": 435
    },
    {
      "epoch": 1.354863813229572,
      "grad_norm": 2.046875,
      "learning_rate": 7.485749469093833e-06,
      "loss": 1.1911,
      "step": 436
    },
    {
      "epoch": 1.3579766536964981,
      "grad_norm": 2.125,
      "learning_rate": 7.475043927917908e-06,
      "loss": 1.2517,
      "step": 437
    },
    {
      "epoch": 1.3610894941634242,
      "grad_norm": 1.9765625,
      "learning_rate": 7.464323336496606e-06,
      "loss": 1.1531,
      "step": 438
    },
    {
      "epoch": 1.3642023346303502,
      "grad_norm": 2.09375,
      "learning_rate": 7.453587760019691e-06,
      "loss": 1.1705,
      "step": 439
    },
    {
      "epoch": 1.3673151750972763,
      "grad_norm": 2.078125,
      "learning_rate": 7.442837263768049e-06,
      "loss": 1.2622,
      "step": 440
    },
    {
      "epoch": 1.3704280155642024,
      "grad_norm": 2.09375,
      "learning_rate": 7.432071913113292e-06,
      "loss": 1.2229,
      "step": 441
    },
    {
      "epoch": 1.3704280155642024,
      "eval_loss": 3.245072603225708,
      "eval_runtime": 42.9891,
      "eval_samples_per_second": 114.168,
      "eval_steps_per_second": 11.421,
      "step": 441
    },
    {
      "epoch": 1.3735408560311284,
      "grad_norm": 2.078125,
      "learning_rate": 7.421291773517354e-06,
      "loss": 1.2302,
      "step": 442
    },
    {
      "epoch": 1.3766536964980545,
      "grad_norm": 2.078125,
      "learning_rate": 7.410496910532103e-06,
      "loss": 1.2408,
      "step": 443
    },
    {
      "epoch": 1.3797665369649805,
      "grad_norm": 2.15625,
      "learning_rate": 7.399687389798933e-06,
      "loss": 1.212,
      "step": 444
    },
    {
      "epoch": 1.3828793774319066,
      "grad_norm": 2.1875,
      "learning_rate": 7.388863277048368e-06,
      "loss": 1.1861,
      "step": 445
    },
    {
      "epoch": 1.3859922178988326,
      "grad_norm": 2.09375,
      "learning_rate": 7.37802463809967e-06,
      "loss": 1.2117,
      "step": 446
    },
    {
      "epoch": 1.3891050583657587,
      "grad_norm": 2.015625,
      "learning_rate": 7.367171538860422e-06,
      "loss": 1.2034,
      "step": 447
    },
    {
      "epoch": 1.3922178988326848,
      "grad_norm": 2.0625,
      "learning_rate": 7.356304045326147e-06,
      "loss": 1.1683,
      "step": 448
    },
    {
      "epoch": 1.3953307392996108,
      "grad_norm": 2.0625,
      "learning_rate": 7.345422223579888e-06,
      "loss": 1.2002,
      "step": 449
    },
    {
      "epoch": 1.3984435797665369,
      "grad_norm": 2.09375,
      "learning_rate": 7.33452613979182e-06,
      "loss": 1.2097,
      "step": 450
    },
    {
      "epoch": 1.401556420233463,
      "grad_norm": 1.984375,
      "learning_rate": 7.323615860218844e-06,
      "loss": 1.174,
      "step": 451
    },
    {
      "epoch": 1.404669260700389,
      "grad_norm": 2.09375,
      "learning_rate": 7.312691451204178e-06,
      "loss": 1.188,
      "step": 452
    },
    {
      "epoch": 1.407782101167315,
      "grad_norm": 2.140625,
      "learning_rate": 7.3017529791769595e-06,
      "loss": 1.281,
      "step": 453
    },
    {
      "epoch": 1.4108949416342411,
      "grad_norm": 2.109375,
      "learning_rate": 7.290800510651845e-06,
      "loss": 1.1868,
      "step": 454
    },
    {
      "epoch": 1.4140077821011672,
      "grad_norm": 2.0625,
      "learning_rate": 7.279834112228597e-06,
      "loss": 1.2264,
      "step": 455
    },
    {
      "epoch": 1.4171206225680935,
      "grad_norm": 2.09375,
      "learning_rate": 7.268853850591678e-06,
      "loss": 1.2228,
      "step": 456
    },
    {
      "epoch": 1.4202334630350195,
      "grad_norm": 2.09375,
      "learning_rate": 7.2578597925098624e-06,
      "loss": 1.2144,
      "step": 457
    },
    {
      "epoch": 1.4233463035019456,
      "grad_norm": 2.046875,
      "learning_rate": 7.246852004835807e-06,
      "loss": 1.1469,
      "step": 458
    },
    {
      "epoch": 1.4264591439688716,
      "grad_norm": 2.0625,
      "learning_rate": 7.235830554505661e-06,
      "loss": 1.2039,
      "step": 459
    },
    {
      "epoch": 1.4295719844357977,
      "grad_norm": 2.09375,
      "learning_rate": 7.22479550853865e-06,
      "loss": 1.2336,
      "step": 460
    },
    {
      "epoch": 1.4326848249027238,
      "grad_norm": 2.1875,
      "learning_rate": 7.213746934036674e-06,
      "loss": 1.2167,
      "step": 461
    },
    {
      "epoch": 1.4357976653696498,
      "grad_norm": 2.203125,
      "learning_rate": 7.202684898183898e-06,
      "loss": 1.2142,
      "step": 462
    },
    {
      "epoch": 1.4357976653696498,
      "eval_loss": 3.242511034011841,
      "eval_runtime": 42.9464,
      "eval_samples_per_second": 114.282,
      "eval_steps_per_second": 11.433,
      "step": 462
    },
    {
      "epoch": 1.4389105058365759,
      "grad_norm": 2.078125,
      "learning_rate": 7.1916094682463425e-06,
      "loss": 1.2513,
      "step": 463
    },
    {
      "epoch": 1.442023346303502,
      "grad_norm": 2.078125,
      "learning_rate": 7.1805207115714715e-06,
      "loss": 1.1959,
      "step": 464
    },
    {
      "epoch": 1.445136186770428,
      "grad_norm": 2.0,
      "learning_rate": 7.169418695587791e-06,
      "loss": 1.2024,
      "step": 465
    },
    {
      "epoch": 1.448249027237354,
      "grad_norm": 2.109375,
      "learning_rate": 7.1583034878044325e-06,
      "loss": 1.2115,
      "step": 466
    },
    {
      "epoch": 1.45136186770428,
      "grad_norm": 2.15625,
      "learning_rate": 7.147175155810741e-06,
      "loss": 1.2186,
      "step": 467
    },
    {
      "epoch": 1.4544747081712062,
      "grad_norm": 1.9921875,
      "learning_rate": 7.136033767275872e-06,
      "loss": 1.196,
      "step": 468
    },
    {
      "epoch": 1.4575875486381322,
      "grad_norm": 2.078125,
      "learning_rate": 7.1248793899483735e-06,
      "loss": 1.2852,
      "step": 469
    },
    {
      "epoch": 1.4607003891050583,
      "grad_norm": 2.109375,
      "learning_rate": 7.113712091655773e-06,
      "loss": 1.2415,
      "step": 470
    },
    {
      "epoch": 1.4638132295719846,
      "grad_norm": 2.0,
      "learning_rate": 7.1025319403041725e-06,
      "loss": 1.2232,
      "step": 471
    },
    {
      "epoch": 1.4669260700389106,
      "grad_norm": 2.09375,
      "learning_rate": 7.091339003877826e-06,
      "loss": 1.195,
      "step": 472
    },
    {
      "epoch": 1.4700389105058367,
      "grad_norm": 2.15625,
      "learning_rate": 7.080133350438736e-06,
      "loss": 1.2175,
      "step": 473
    },
    {
      "epoch": 1.4731517509727627,
      "grad_norm": 2.140625,
      "learning_rate": 7.068915048126231e-06,
      "loss": 1.2311,
      "step": 474
    },
    {
      "epoch": 1.4762645914396888,
      "grad_norm": 1.96875,
      "learning_rate": 7.057684165156556e-06,
      "loss": 1.1713,
      "step": 475
    },
    {
      "epoch": 1.4793774319066149,
      "grad_norm": 2.078125,
      "learning_rate": 7.0464407698224546e-06,
      "loss": 1.1632,
      "step": 476
    },
    {
      "epoch": 1.482490272373541,
      "grad_norm": 2.0625,
      "learning_rate": 7.03518493049276e-06,
      "loss": 1.2709,
      "step": 477
    },
    {
      "epoch": 1.485603112840467,
      "grad_norm": 2.140625,
      "learning_rate": 7.023916715611969e-06,
      "loss": 1.2537,
      "step": 478
    },
    {
      "epoch": 1.488715953307393,
      "grad_norm": 2.109375,
      "learning_rate": 7.012636193699838e-06,
      "loss": 1.2134,
      "step": 479
    },
    {
      "epoch": 1.491828793774319,
      "grad_norm": 2.1875,
      "learning_rate": 7.001343433350956e-06,
      "loss": 1.2369,
      "step": 480
    },
    {
      "epoch": 1.4949416342412452,
      "grad_norm": 2.078125,
      "learning_rate": 6.9900385032343296e-06,
      "loss": 1.1922,
      "step": 481
    },
    {
      "epoch": 1.4980544747081712,
      "grad_norm": 2.03125,
      "learning_rate": 6.978721472092973e-06,
      "loss": 1.1372,
      "step": 482
    },
    {
      "epoch": 1.5011673151750973,
      "grad_norm": 2.15625,
      "learning_rate": 6.967392408743479e-06,
      "loss": 1.2211,
      "step": 483
    },
    {
      "epoch": 1.5011673151750973,
      "eval_loss": 3.246913433074951,
      "eval_runtime": 42.9553,
      "eval_samples_per_second": 114.258,
      "eval_steps_per_second": 11.43,
      "step": 483
    },
    {
      "epoch": 1.5042801556420233,
      "grad_norm": 2.140625,
      "learning_rate": 6.956051382075608e-06,
      "loss": 1.2201,
      "step": 484
    },
    {
      "epoch": 1.5073929961089494,
      "grad_norm": 2.1875,
      "learning_rate": 6.94469846105187e-06,
      "loss": 1.2123,
      "step": 485
    },
    {
      "epoch": 1.5105058365758754,
      "grad_norm": 1.953125,
      "learning_rate": 6.933333714707094e-06,
      "loss": 1.2233,
      "step": 486
    },
    {
      "epoch": 1.5136186770428015,
      "grad_norm": 2.0,
      "learning_rate": 6.921957212148027e-06,
      "loss": 1.1983,
      "step": 487
    },
    {
      "epoch": 1.5167315175097276,
      "grad_norm": 2.046875,
      "learning_rate": 6.910569022552896e-06,
      "loss": 1.2192,
      "step": 488
    },
    {
      "epoch": 1.5198443579766536,
      "grad_norm": 1.9765625,
      "learning_rate": 6.8991692151709945e-06,
      "loss": 1.211,
      "step": 489
    },
    {
      "epoch": 1.5229571984435797,
      "grad_norm": 2.109375,
      "learning_rate": 6.887757859322265e-06,
      "loss": 1.1689,
      "step": 490
    },
    {
      "epoch": 1.5260700389105057,
      "grad_norm": 2.109375,
      "learning_rate": 6.876335024396872e-06,
      "loss": 1.2438,
      "step": 491
    },
    {
      "epoch": 1.5291828793774318,
      "grad_norm": 2.0625,
      "learning_rate": 6.86490077985478e-06,
      "loss": 1.16,
      "step": 492
    },
    {
      "epoch": 1.5322957198443579,
      "grad_norm": 2.203125,
      "learning_rate": 6.8534551952253395e-06,
      "loss": 1.1946,
      "step": 493
    },
    {
      "epoch": 1.535408560311284,
      "grad_norm": 2.171875,
      "learning_rate": 6.841998340106848e-06,
      "loss": 1.2187,
      "step": 494
    },
    {
      "epoch": 1.53852140077821,
      "grad_norm": 2.0,
      "learning_rate": 6.830530284166146e-06,
      "loss": 1.1818,
      "step": 495
    },
    {
      "epoch": 1.541634241245136,
      "grad_norm": 2.234375,
      "learning_rate": 6.819051097138179e-06,
      "loss": 1.2058,
      "step": 496
    },
    {
      "epoch": 1.544747081712062,
      "grad_norm": 2.046875,
      "learning_rate": 6.8075608488255786e-06,
      "loss": 1.2351,
      "step": 497
    },
    {
      "epoch": 1.5478599221789884,
      "grad_norm": 2.171875,
      "learning_rate": 6.79605960909824e-06,
      "loss": 1.2329,
      "step": 498
    },
    {
      "epoch": 1.5509727626459144,
      "grad_norm": 2.171875,
      "learning_rate": 6.784547447892892e-06,
      "loss": 1.175,
      "step": 499
    },
    {
      "epoch": 1.5540856031128405,
      "grad_norm": 2.15625,
      "learning_rate": 6.773024435212678e-06,
      "loss": 1.223,
      "step": 500
    },
    {
      "epoch": 1.5571984435797666,
      "grad_norm": 2.109375,
      "learning_rate": 6.761490641126725e-06,
      "loss": 1.2402,
      "step": 501
    },
    {
      "epoch": 1.5603112840466926,
      "grad_norm": 2.015625,
      "learning_rate": 6.7499461357697206e-06,
      "loss": 1.1994,
      "step": 502
    },
    {
      "epoch": 1.5634241245136187,
      "grad_norm": 2.09375,
      "learning_rate": 6.738390989341483e-06,
      "loss": 1.2562,
      "step": 503
    },
    {
      "epoch": 1.5665369649805447,
      "grad_norm": 2.09375,
      "learning_rate": 6.726825272106539e-06,
      "loss": 1.2439,
      "step": 504
    },
    {
      "epoch": 1.5665369649805447,
      "eval_loss": 3.2792906761169434,
      "eval_runtime": 43.1283,
      "eval_samples_per_second": 113.8,
      "eval_steps_per_second": 11.385,
      "step": 504
    },
    {
      "epoch": 1.5696498054474708,
      "grad_norm": 2.0625,
      "learning_rate": 6.715249054393694e-06,
      "loss": 1.2385,
      "step": 505
    },
    {
      "epoch": 1.5727626459143969,
      "grad_norm": 2.125,
      "learning_rate": 6.703662406595606e-06,
      "loss": 1.2241,
      "step": 506
    },
    {
      "epoch": 1.575875486381323,
      "grad_norm": 2.1875,
      "learning_rate": 6.692065399168352e-06,
      "loss": 1.2576,
      "step": 507
    },
    {
      "epoch": 1.5789883268482492,
      "grad_norm": 2.078125,
      "learning_rate": 6.68045810263101e-06,
      "loss": 1.2125,
      "step": 508
    },
    {
      "epoch": 1.5821011673151752,
      "grad_norm": 2.078125,
      "learning_rate": 6.668840587565219e-06,
      "loss": 1.1819,
      "step": 509
    },
    {
      "epoch": 1.5852140077821013,
      "grad_norm": 2.0625,
      "learning_rate": 6.657212924614754e-06,
      "loss": 1.2201,
      "step": 510
    },
    {
      "epoch": 1.5883268482490274,
      "grad_norm": 2.140625,
      "learning_rate": 6.6455751844851035e-06,
      "loss": 1.2827,
      "step": 511
    },
    {
      "epoch": 1.5914396887159534,
      "grad_norm": 2.125,
      "learning_rate": 6.6339274379430285e-06,
      "loss": 1.2311,
      "step": 512
    },
    {
      "epoch": 1.5945525291828795,
      "grad_norm": 2.25,
      "learning_rate": 6.622269755816136e-06,
      "loss": 1.245,
      "step": 513
    },
    {
      "epoch": 1.5976653696498055,
      "grad_norm": 2.0625,
      "learning_rate": 6.6106022089924535e-06,
      "loss": 1.2014,
      "step": 514
    },
    {
      "epoch": 1.6007782101167316,
      "grad_norm": 2.046875,
      "learning_rate": 6.598924868419992e-06,
      "loss": 1.1873,
      "step": 515
    },
    {
      "epoch": 1.6038910505836577,
      "grad_norm": 1.96875,
      "learning_rate": 6.587237805106312e-06,
      "loss": 1.2058,
      "step": 516
    },
    {
      "epoch": 1.6070038910505837,
      "grad_norm": 2.140625,
      "learning_rate": 6.575541090118105e-06,
      "loss": 1.2358,
      "step": 517
    },
    {
      "epoch": 1.6101167315175098,
      "grad_norm": 2.125,
      "learning_rate": 6.563834794580742e-06,
      "loss": 1.2446,
      "step": 518
    },
    {
      "epoch": 1.6132295719844358,
      "grad_norm": 2.109375,
      "learning_rate": 6.5521189896778615e-06,
      "loss": 1.1952,
      "step": 519
    },
    {
      "epoch": 1.616342412451362,
      "grad_norm": 2.046875,
      "learning_rate": 6.5403937466509195e-06,
      "loss": 1.1753,
      "step": 520
    },
    {
      "epoch": 1.619455252918288,
      "grad_norm": 2.109375,
      "learning_rate": 6.5286591367987655e-06,
      "loss": 1.1946,
      "step": 521
    },
    {
      "epoch": 1.622568093385214,
      "grad_norm": 2.09375,
      "learning_rate": 6.516915231477207e-06,
      "loss": 1.2307,
      "step": 522
    },
    {
      "epoch": 1.62568093385214,
      "grad_norm": 2.015625,
      "learning_rate": 6.505162102098574e-06,
      "loss": 1.2299,
      "step": 523
    },
    {
      "epoch": 1.6287937743190661,
      "grad_norm": 2.15625,
      "learning_rate": 6.493399820131291e-06,
      "loss": 1.2615,
      "step": 524
    },
    {
      "epoch": 1.6319066147859922,
      "grad_norm": 2.140625,
      "learning_rate": 6.48162845709943e-06,
      "loss": 1.2578,
      "step": 525
    },
    {
      "epoch": 1.6319066147859922,
      "eval_loss": 3.2751851081848145,
      "eval_runtime": 242.1819,
      "eval_samples_per_second": 20.266,
      "eval_steps_per_second": 2.027,
      "step": 525
    },
    {
      "epoch": 1.6350194552529183,
      "grad_norm": 2.203125,
      "learning_rate": 6.469848084582288e-06,
      "loss": 1.2736,
      "step": 526
    },
    {
      "epoch": 1.6381322957198443,
      "grad_norm": 2.15625,
      "learning_rate": 6.458058774213949e-06,
      "loss": 1.1787,
      "step": 527
    },
    {
      "epoch": 1.6412451361867704,
      "grad_norm": 2.109375,
      "learning_rate": 6.4462605976828395e-06,
      "loss": 1.2024,
      "step": 528
    },
    {
      "epoch": 1.6443579766536964,
      "grad_norm": 2.25,
      "learning_rate": 6.434453626731303e-06,
      "loss": 1.2447,
      "step": 529
    },
    {
      "epoch": 1.6474708171206225,
      "grad_norm": 2.203125,
      "learning_rate": 6.4226379331551625e-06,
      "loss": 1.2083,
      "step": 530
    },
    {
      "epoch": 1.6505836575875485,
      "grad_norm": 2.234375,
      "learning_rate": 6.410813588803279e-06,
      "loss": 1.1872,
      "step": 531
    },
    {
      "epoch": 1.6536964980544746,
      "grad_norm": 2.078125,
      "learning_rate": 6.398980665577115e-06,
      "loss": 1.2469,
      "step": 532
    },
    {
      "epoch": 1.6568093385214007,
      "grad_norm": 1.9453125,
      "learning_rate": 6.387139235430304e-06,
      "loss": 1.2334,
      "step": 533
    },
    {
      "epoch": 1.6599221789883267,
      "grad_norm": 1.9453125,
      "learning_rate": 6.375289370368205e-06,
      "loss": 1.2429,
      "step": 534
    },
    {
      "epoch": 1.6630350194552528,
      "grad_norm": 2.125,
      "learning_rate": 6.363431142447469e-06,
      "loss": 1.2982,
      "step": 535
    },
    {
      "epoch": 1.6661478599221788,
      "grad_norm": 1.9921875,
      "learning_rate": 6.351564623775602e-06,
      "loss": 1.1676,
      "step": 536
    },
    {
      "epoch": 1.669260700389105,
      "grad_norm": 2.15625,
      "learning_rate": 6.339689886510519e-06,
      "loss": 1.2115,
      "step": 537
    },
    {
      "epoch": 1.6723735408560312,
      "grad_norm": 2.078125,
      "learning_rate": 6.327807002860117e-06,
      "loss": 1.2076,
      "step": 538
    },
    {
      "epoch": 1.6754863813229572,
      "grad_norm": 2.03125,
      "learning_rate": 6.3159160450818255e-06,
      "loss": 1.1962,
      "step": 539
    },
    {
      "epoch": 1.6785992217898833,
      "grad_norm": 2.015625,
      "learning_rate": 6.304017085482171e-06,
      "loss": 1.1989,
      "step": 540
    },
    {
      "epoch": 1.6817120622568094,
      "grad_norm": 2.25,
      "learning_rate": 6.292110196416341e-06,
      "loss": 1.2069,
      "step": 541
    },
    {
      "epoch": 1.6848249027237354,
      "grad_norm": 2.078125,
      "learning_rate": 6.280195450287736e-06,
      "loss": 1.208,
      "step": 542
    },
    {
      "epoch": 1.6879377431906615,
      "grad_norm": 2.03125,
      "learning_rate": 6.268272919547537e-06,
      "loss": 1.2107,
      "step": 543
    },
    {
      "epoch": 1.6910505836575875,
      "grad_norm": 2.140625,
      "learning_rate": 6.256342676694261e-06,
      "loss": 1.2123,
      "step": 544
    },
    {
      "epoch": 1.6941634241245136,
      "grad_norm": 2.109375,
      "learning_rate": 6.24440479427332e-06,
      "loss": 1.2055,
      "step": 545
    },
    {
      "epoch": 1.6972762645914397,
      "grad_norm": 2.28125,
      "learning_rate": 6.23245934487658e-06,
      "loss": 1.2614,
      "step": 546
    },
    {
      "epoch": 1.6972762645914397,
      "eval_loss": 3.263235569000244,
      "eval_runtime": 43.0547,
      "eval_samples_per_second": 113.994,
      "eval_steps_per_second": 11.404,
      "step": 546
    },
    {
      "epoch": 1.7003891050583657,
      "grad_norm": 2.140625,
      "learning_rate": 6.220506401141923e-06,
      "loss": 1.2194,
      "step": 547
    },
    {
      "epoch": 1.703501945525292,
      "grad_norm": 2.28125,
      "learning_rate": 6.208546035752798e-06,
      "loss": 1.2122,
      "step": 548
    },
    {
      "epoch": 1.706614785992218,
      "grad_norm": 2.09375,
      "learning_rate": 6.1965783214377895e-06,
      "loss": 1.2646,
      "step": 549
    },
    {
      "epoch": 1.7097276264591441,
      "grad_norm": 2.21875,
      "learning_rate": 6.184603330970166e-06,
      "loss": 1.1832,
      "step": 550
    },
    {
      "epoch": 1.7128404669260702,
      "grad_norm": 2.140625,
      "learning_rate": 6.172621137167438e-06,
      "loss": 1.255,
      "step": 551
    },
    {
      "epoch": 1.7159533073929962,
      "grad_norm": 2.0625,
      "learning_rate": 6.160631812890925e-06,
      "loss": 1.2262,
      "step": 552
    },
    {
      "epoch": 1.7190661478599223,
      "grad_norm": 1.9921875,
      "learning_rate": 6.1486354310453e-06,
      "loss": 1.1897,
      "step": 553
    },
    {
      "epoch": 1.7221789883268483,
      "grad_norm": 2.046875,
      "learning_rate": 6.136632064578153e-06,
      "loss": 1.2697,
      "step": 554
    },
    {
      "epoch": 1.7252918287937744,
      "grad_norm": 2.015625,
      "learning_rate": 6.1246217864795475e-06,
      "loss": 1.1557,
      "step": 555
    },
    {
      "epoch": 1.7284046692607005,
      "grad_norm": 2.140625,
      "learning_rate": 6.112604669781572e-06,
      "loss": 1.2349,
      "step": 556
    },
    {
      "epoch": 1.7315175097276265,
      "grad_norm": 2.03125,
      "learning_rate": 6.1005807875579046e-06,
      "loss": 1.2456,
      "step": 557
    },
    {
      "epoch": 1.7346303501945526,
      "grad_norm": 2.046875,
      "learning_rate": 6.088550212923356e-06,
      "loss": 1.1561,
      "step": 558
    },
    {
      "epoch": 1.7377431906614786,
      "grad_norm": 2.046875,
      "learning_rate": 6.076513019033438e-06,
      "loss": 1.1835,
      "step": 559
    },
    {
      "epoch": 1.7408560311284047,
      "grad_norm": 2.109375,
      "learning_rate": 6.064469279083912e-06,
      "loss": 1.2335,
      "step": 560
    },
    {
      "epoch": 1.7439688715953308,
      "grad_norm": 1.984375,
      "learning_rate": 6.052419066310342e-06,
      "loss": 1.1958,
      "step": 561
    },
    {
      "epoch": 1.7470817120622568,
      "grad_norm": 2.015625,
      "learning_rate": 6.040362453987651e-06,
      "loss": 1.2303,
      "step": 562
    },
    {
      "epoch": 1.7501945525291829,
      "grad_norm": 2.171875,
      "learning_rate": 6.028299515429683e-06,
      "loss": 1.2403,
      "step": 563
    },
    {
      "epoch": 1.753307392996109,
      "grad_norm": 2.09375,
      "learning_rate": 6.016230323988745e-06,
      "loss": 1.1959,
      "step": 564
    },
    {
      "epoch": 1.756420233463035,
      "grad_norm": 2.21875,
      "learning_rate": 6.004154953055166e-06,
      "loss": 1.2325,
      "step": 565
    },
    {
      "epoch": 1.759533073929961,
      "grad_norm": 1.984375,
      "learning_rate": 5.992073476056854e-06,
      "loss": 1.1498,
      "step": 566
    },
    {
      "epoch": 1.7626459143968871,
      "grad_norm": 2.03125,
      "learning_rate": 5.9799859664588475e-06,
      "loss": 1.2385,
      "step": 567
    },
    {
      "epoch": 1.7626459143968871,
      "eval_loss": 3.2315096855163574,
      "eval_runtime": 43.1354,
      "eval_samples_per_second": 113.781,
      "eval_steps_per_second": 11.383,
      "step": 567
    },
    {
      "epoch": 1.7657587548638132,
      "grad_norm": 2.015625,
      "learning_rate": 5.967892497762866e-06,
      "loss": 1.2011,
      "step": 568
    },
    {
      "epoch": 1.7688715953307392,
      "grad_norm": 2.0625,
      "learning_rate": 5.955793143506863e-06,
      "loss": 1.1966,
      "step": 569
    },
    {
      "epoch": 1.7719844357976653,
      "grad_norm": 2.125,
      "learning_rate": 5.943687977264584e-06,
      "loss": 1.2229,
      "step": 570
    },
    {
      "epoch": 1.7750972762645914,
      "grad_norm": 2.0,
      "learning_rate": 5.931577072645117e-06,
      "loss": 1.1918,
      "step": 571
    },
    {
      "epoch": 1.7782101167315174,
      "grad_norm": 2.21875,
      "learning_rate": 5.919460503292443e-06,
      "loss": 1.2622,
      "step": 572
    },
    {
      "epoch": 1.7813229571984435,
      "grad_norm": 2.109375,
      "learning_rate": 5.907338342884984e-06,
      "loss": 1.2479,
      "step": 573
    },
    {
      "epoch": 1.7844357976653695,
      "grad_norm": 2.0625,
      "learning_rate": 5.8952106651351695e-06,
      "loss": 1.2034,
      "step": 574
    },
    {
      "epoch": 1.7875486381322956,
      "grad_norm": 2.0,
      "learning_rate": 5.883077543788972e-06,
      "loss": 1.1843,
      "step": 575
    },
    {
      "epoch": 1.7906614785992216,
      "grad_norm": 2.015625,
      "learning_rate": 5.870939052625465e-06,
      "loss": 1.2454,
      "step": 576
    },
    {
      "epoch": 1.7937743190661477,
      "grad_norm": 2.125,
      "learning_rate": 5.858795265456382e-06,
      "loss": 1.2067,
      "step": 577
    },
    {
      "epoch": 1.796887159533074,
      "grad_norm": 2.0625,
      "learning_rate": 5.84664625612565e-06,
      "loss": 1.2409,
      "step": 578
    },
    {
      "epoch": 1.8,
      "grad_norm": 2.0,
      "learning_rate": 5.834492098508959e-06,
      "loss": 1.1994,
      "step": 579
    },
    {
      "epoch": 1.803112840466926,
      "grad_norm": 2.171875,
      "learning_rate": 5.822332866513304e-06,
      "loss": 1.1593,
      "step": 580
    },
    {
      "epoch": 1.8062256809338522,
      "grad_norm": 2.046875,
      "learning_rate": 5.8101686340765305e-06,
      "loss": 1.2573,
      "step": 581
    },
    {
      "epoch": 1.8093385214007782,
      "grad_norm": 2.109375,
      "learning_rate": 5.797999475166897e-06,
      "loss": 1.2329,
      "step": 582
    },
    {
      "epoch": 1.8124513618677043,
      "grad_norm": 2.078125,
      "learning_rate": 5.785825463782616e-06,
      "loss": 1.1768,
      "step": 583
    },
    {
      "epoch": 1.8155642023346303,
      "grad_norm": 2.046875,
      "learning_rate": 5.773646673951406e-06,
      "loss": 1.1488,
      "step": 584
    },
    {
      "epoch": 1.8186770428015564,
      "grad_norm": 2.09375,
      "learning_rate": 5.761463179730047e-06,
      "loss": 1.275,
      "step": 585
    },
    {
      "epoch": 1.8217898832684825,
      "grad_norm": 2.125,
      "learning_rate": 5.749275055203918e-06,
      "loss": 1.2906,
      "step": 586
    },
    {
      "epoch": 1.8249027237354085,
      "grad_norm": 2.328125,
      "learning_rate": 5.7370823744865615e-06,
      "loss": 1.223,
      "step": 587
    },
    {
      "epoch": 1.8280155642023346,
      "grad_norm": 2.03125,
      "learning_rate": 5.7248852117192215e-06,
      "loss": 1.1596,
      "step": 588
    },
    {
      "epoch": 1.8280155642023346,
      "eval_loss": 3.2958641052246094,
      "eval_runtime": 43.1257,
      "eval_samples_per_second": 113.807,
      "eval_steps_per_second": 11.385,
      "step": 588
    },
    {
      "epoch": 1.8311284046692609,
      "grad_norm": 2.125,
      "learning_rate": 5.712683641070393e-06,
      "loss": 1.2058,
      "step": 589
    },
    {
      "epoch": 1.834241245136187,
      "grad_norm": 2.09375,
      "learning_rate": 5.700477736735383e-06,
      "loss": 1.1994,
      "step": 590
    },
    {
      "epoch": 1.837354085603113,
      "grad_norm": 2.0625,
      "learning_rate": 5.688267572935843e-06,
      "loss": 1.1652,
      "step": 591
    },
    {
      "epoch": 1.840466926070039,
      "grad_norm": 2.0625,
      "learning_rate": 5.676053223919328e-06,
      "loss": 1.2258,
      "step": 592
    },
    {
      "epoch": 1.843579766536965,
      "grad_norm": 2.046875,
      "learning_rate": 5.663834763958845e-06,
      "loss": 1.2107,
      "step": 593
    },
    {
      "epoch": 1.8466926070038912,
      "grad_norm": 2.09375,
      "learning_rate": 5.651612267352396e-06,
      "loss": 1.2265,
      "step": 594
    },
    {
      "epoch": 1.8498054474708172,
      "grad_norm": 2.15625,
      "learning_rate": 5.6393858084225305e-06,
      "loss": 1.2716,
      "step": 595
    },
    {
      "epoch": 1.8529182879377433,
      "grad_norm": 2.078125,
      "learning_rate": 5.627155461515891e-06,
      "loss": 1.1541,
      "step": 596
    },
    {
      "epoch": 1.8560311284046693,
      "grad_norm": 2.0625,
      "learning_rate": 5.61492130100276e-06,
      "loss": 1.1998,
      "step": 597
    },
    {
      "epoch": 1.8591439688715954,
      "grad_norm": 2.09375,
      "learning_rate": 5.6026834012766155e-06,
      "loss": 1.2547,
      "step": 598
    },
    {
      "epoch": 1.8622568093385214,
      "grad_norm": 2.171875,
      "learning_rate": 5.590441836753669e-06,
      "loss": 1.2091,
      "step": 599
    },
    {
      "epoch": 1.8653696498054475,
      "grad_norm": 2.140625,
      "learning_rate": 5.5781966818724155e-06,
      "loss": 1.1906,
      "step": 600
    },
    {
      "epoch": 1.8684824902723736,
      "grad_norm": 2.0625,
      "learning_rate": 5.565948011093185e-06,
      "loss": 1.1979,
      "step": 601
    },
    {
      "epoch": 1.8715953307392996,
      "grad_norm": 2.078125,
      "learning_rate": 5.553695898897686e-06,
      "loss": 1.2465,
      "step": 602
    },
    {
      "epoch": 1.8747081712062257,
      "grad_norm": 2.078125,
      "learning_rate": 5.5414404197885565e-06,
      "loss": 1.2629,
      "step": 603
    },
    {
      "epoch": 1.8778210116731517,
      "grad_norm": 2.109375,
      "learning_rate": 5.529181648288902e-06,
      "loss": 1.2341,
      "step": 604
    },
    {
      "epoch": 1.8809338521400778,
      "grad_norm": 2.09375,
      "learning_rate": 5.51691965894185e-06,
      "loss": 1.1666,
      "step": 605
    },
    {
      "epoch": 1.8840466926070039,
      "grad_norm": 2.125,
      "learning_rate": 5.5046545263101026e-06,
      "loss": 1.2525,
      "step": 606
    },
    {
      "epoch": 1.88715953307393,
      "grad_norm": 2.09375,
      "learning_rate": 5.492386324975466e-06,
      "loss": 1.1903,
      "step": 607
    },
    {
      "epoch": 1.890272373540856,
      "grad_norm": 2.015625,
      "learning_rate": 5.480115129538409e-06,
      "loss": 1.2323,
      "step": 608
    },
    {
      "epoch": 1.893385214007782,
      "grad_norm": 2.15625,
      "learning_rate": 5.467841014617611e-06,
      "loss": 1.2503,
      "step": 609
    },
    {
      "epoch": 1.893385214007782,
      "eval_loss": 3.2654647827148438,
      "eval_runtime": 43.0024,
      "eval_samples_per_second": 114.133,
      "eval_steps_per_second": 11.418,
      "step": 609
    },
    {
      "epoch": 1.896498054474708,
      "grad_norm": 2.046875,
      "learning_rate": 5.455564054849501e-06,
      "loss": 1.21,
      "step": 610
    },
    {
      "epoch": 1.8996108949416342,
      "grad_norm": 2.109375,
      "learning_rate": 5.443284324887804e-06,
      "loss": 1.2073,
      "step": 611
    },
    {
      "epoch": 1.9027237354085602,
      "grad_norm": 1.984375,
      "learning_rate": 5.4310018994030974e-06,
      "loss": 1.2073,
      "step": 612
    },
    {
      "epoch": 1.9058365758754863,
      "grad_norm": 2.0625,
      "learning_rate": 5.418716853082344e-06,
      "loss": 1.1849,
      "step": 613
    },
    {
      "epoch": 1.9089494163424123,
      "grad_norm": 2.0625,
      "learning_rate": 5.406429260628444e-06,
      "loss": 1.2102,
      "step": 614
    },
    {
      "epoch": 1.9120622568093384,
      "grad_norm": 2.125,
      "learning_rate": 5.394139196759783e-06,
      "loss": 1.2716,
      "step": 615
    },
    {
      "epoch": 1.9151750972762644,
      "grad_norm": 2.140625,
      "learning_rate": 5.381846736209769e-06,
      "loss": 1.1827,
      "step": 616
    },
    {
      "epoch": 1.9182879377431905,
      "grad_norm": 2.125,
      "learning_rate": 5.369551953726392e-06,
      "loss": 1.2514,
      "step": 617
    },
    {
      "epoch": 1.9214007782101166,
      "grad_norm": 2.046875,
      "learning_rate": 5.357254924071755e-06,
      "loss": 1.1653,
      "step": 618
    },
    {
      "epoch": 1.9245136186770428,
      "grad_norm": 2.109375,
      "learning_rate": 5.3449557220216245e-06,
      "loss": 1.2268,
      "step": 619
    },
    {
      "epoch": 1.927626459143969,
      "grad_norm": 2.109375,
      "learning_rate": 5.332654422364984e-06,
      "loss": 1.1909,
      "step": 620
    },
    {
      "epoch": 1.930739299610895,
      "grad_norm": 2.109375,
      "learning_rate": 5.320351099903565e-06,
      "loss": 1.2197,
      "step": 621
    },
    {
      "epoch": 1.933852140077821,
      "grad_norm": 2.09375,
      "learning_rate": 5.308045829451403e-06,
      "loss": 1.2061,
      "step": 622
    },
    {
      "epoch": 1.936964980544747,
      "grad_norm": 2.09375,
      "learning_rate": 5.295738685834379e-06,
      "loss": 1.2213,
      "step": 623
    },
    {
      "epoch": 1.9400778210116731,
      "grad_norm": 2.03125,
      "learning_rate": 5.28342974388976e-06,
      "loss": 1.2409,
      "step": 624
    },
    {
      "epoch": 1.9431906614785992,
      "grad_norm": 2.078125,
      "learning_rate": 5.271119078465755e-06,
      "loss": 1.2393,
      "step": 625
    },
    {
      "epoch": 1.9463035019455253,
      "grad_norm": 2.03125,
      "learning_rate": 5.258806764421048e-06,
      "loss": 1.2064,
      "step": 626
    },
    {
      "epoch": 1.9494163424124513,
      "grad_norm": 2.125,
      "learning_rate": 5.2464928766243485e-06,
      "loss": 1.2455,
      "step": 627
    },
    {
      "epoch": 1.9525291828793774,
      "grad_norm": 2.03125,
      "learning_rate": 5.23417748995394e-06,
      "loss": 1.1626,
      "step": 628
    },
    {
      "epoch": 1.9556420233463037,
      "grad_norm": 2.21875,
      "learning_rate": 5.221860679297215e-06,
      "loss": 1.2355,
      "step": 629
    },
    {
      "epoch": 1.9587548638132297,
      "grad_norm": 2.171875,
      "learning_rate": 5.2095425195502254e-06,
      "loss": 1.2989,
      "step": 630
    },
    {
      "epoch": 1.9587548638132297,
      "eval_loss": 3.300218105316162,
      "eval_runtime": 43.0432,
      "eval_samples_per_second": 114.025,
      "eval_steps_per_second": 11.407,
      "step": 630
    },
    {
      "epoch": 1.9618677042801558,
      "grad_norm": 2.03125,
      "learning_rate": 5.1972230856172335e-06,
      "loss": 1.1382,
      "step": 631
    },
    {
      "epoch": 1.9649805447470818,
      "grad_norm": 2.046875,
      "learning_rate": 5.184902452410238e-06,
      "loss": 1.2134,
      "step": 632
    },
    {
      "epoch": 1.968093385214008,
      "grad_norm": 2.046875,
      "learning_rate": 5.172580694848541e-06,
      "loss": 1.2543,
      "step": 633
    },
    {
      "epoch": 1.971206225680934,
      "grad_norm": 2.109375,
      "learning_rate": 5.160257887858278e-06,
      "loss": 1.2156,
      "step": 634
    },
    {
      "epoch": 1.97431906614786,
      "grad_norm": 2.140625,
      "learning_rate": 5.14793410637196e-06,
      "loss": 1.1832,
      "step": 635
    },
    {
      "epoch": 1.977431906614786,
      "grad_norm": 2.09375,
      "learning_rate": 5.135609425328033e-06,
      "loss": 1.2077,
      "step": 636
    },
    {
      "epoch": 1.9805447470817121,
      "grad_norm": 2.09375,
      "learning_rate": 5.123283919670408e-06,
      "loss": 1.1961,
      "step": 637
    },
    {
      "epoch": 1.9836575875486382,
      "grad_norm": 2.046875,
      "learning_rate": 5.110957664348011e-06,
      "loss": 1.1559,
      "step": 638
    },
    {
      "epoch": 1.9867704280155642,
      "grad_norm": 2.03125,
      "learning_rate": 5.098630734314327e-06,
      "loss": 1.2404,
      "step": 639
    },
    {
      "epoch": 1.9898832684824903,
      "grad_norm": 2.046875,
      "learning_rate": 5.0863032045269435e-06,
      "loss": 1.196,
      "step": 640
    },
    {
      "epoch": 1.9929961089494164,
      "grad_norm": 2.140625,
      "learning_rate": 5.0739751499470925e-06,
      "loss": 1.1919,
      "step": 641
    },
    {
      "epoch": 1.9961089494163424,
      "grad_norm": 2.046875,
      "learning_rate": 5.0616466455392056e-06,
      "loss": 1.2101,
      "step": 642
    },
    {
      "epoch": 1.9992217898832685,
      "grad_norm": 2.125,
      "learning_rate": 5.049317766270436e-06,
      "loss": 1.2414,
      "step": 643
    },
    {
      "epoch": 2.0,
      "grad_norm": 4.34375,
      "learning_rate": 5.03698858711023e-06,
      "loss": 1.1381,
      "step": 644
    },
    {
      "epoch": 2.003112840466926,
      "grad_norm": 2.171875,
      "learning_rate": 5.024659183029849e-06,
      "loss": 1.2396,
      "step": 645
    },
    {
      "epoch": 2.006225680933852,
      "grad_norm": 1.984375,
      "learning_rate": 5.012329629001924e-06,
      "loss": 1.1929,
      "step": 646
    },
    {
      "epoch": 2.009338521400778,
      "grad_norm": 2.203125,
      "learning_rate": 5e-06,
      "loss": 1.1997,
      "step": 647
    },
    {
      "epoch": 2.0124513618677042,
      "grad_norm": 2.0,
      "learning_rate": 4.987670370998078e-06,
      "loss": 1.203,
      "step": 648
    },
    {
      "epoch": 2.0155642023346303,
      "grad_norm": 2.015625,
      "learning_rate": 4.975340816970153e-06,
      "loss": 1.2107,
      "step": 649
    },
    {
      "epoch": 2.0186770428015564,
      "grad_norm": 1.9453125,
      "learning_rate": 4.9630114128897734e-06,
      "loss": 1.1766,
      "step": 650
    },
    {
      "epoch": 2.0217898832684824,
      "grad_norm": 2.0625,
      "learning_rate": 4.950682233729566e-06,
      "loss": 1.1709,
      "step": 651
    },
    {
      "epoch": 2.0217898832684824,
      "eval_loss": 3.283766984939575,
      "eval_runtime": 43.4423,
      "eval_samples_per_second": 112.977,
      "eval_steps_per_second": 11.302,
      "step": 651
    },
    {
      "epoch": 2.0249027237354085,
      "grad_norm": 2.09375,
      "learning_rate": 4.938353354460798e-06,
      "loss": 1.1733,
      "step": 652
    },
    {
      "epoch": 2.0280155642023345,
      "grad_norm": 2.0,
      "learning_rate": 4.9260248500529074e-06,
      "loss": 1.1782,
      "step": 653
    },
    {
      "epoch": 2.0311284046692606,
      "grad_norm": 2.078125,
      "learning_rate": 4.913696795473058e-06,
      "loss": 1.2085,
      "step": 654
    },
    {
      "epoch": 2.0342412451361866,
      "grad_norm": 2.0625,
      "learning_rate": 4.901369265685673e-06,
      "loss": 1.1772,
      "step": 655
    },
    {
      "epoch": 2.0373540856031127,
      "grad_norm": 1.953125,
      "learning_rate": 4.88904233565199e-06,
      "loss": 1.1123,
      "step": 656
    },
    {
      "epoch": 2.0404669260700388,
      "grad_norm": 2.1875,
      "learning_rate": 4.8767160803295925e-06,
      "loss": 1.186,
      "step": 657
    },
    {
      "epoch": 2.043579766536965,
      "grad_norm": 2.015625,
      "learning_rate": 4.8643905746719685e-06,
      "loss": 1.1662,
      "step": 658
    },
    {
      "epoch": 2.046692607003891,
      "grad_norm": 2.234375,
      "learning_rate": 4.852065893628041e-06,
      "loss": 1.2002,
      "step": 659
    },
    {
      "epoch": 2.049805447470817,
      "grad_norm": 2.109375,
      "learning_rate": 4.839742112141725e-06,
      "loss": 1.1697,
      "step": 660
    },
    {
      "epoch": 2.052918287937743,
      "grad_norm": 2.109375,
      "learning_rate": 4.827419305151461e-06,
      "loss": 1.2458,
      "step": 661
    },
    {
      "epoch": 2.056031128404669,
      "grad_norm": 2.078125,
      "learning_rate": 4.815097547589763e-06,
      "loss": 1.1816,
      "step": 662
    },
    {
      "epoch": 2.059143968871595,
      "grad_norm": 2.109375,
      "learning_rate": 4.80277691438277e-06,
      "loss": 1.1545,
      "step": 663
    },
    {
      "epoch": 2.062256809338521,
      "grad_norm": 2.21875,
      "learning_rate": 4.790457480449775e-06,
      "loss": 1.133,
      "step": 664
    },
    {
      "epoch": 2.0653696498054477,
      "grad_norm": 2.140625,
      "learning_rate": 4.778139320702787e-06,
      "loss": 1.1207,
      "step": 665
    },
    {
      "epoch": 2.0684824902723737,
      "grad_norm": 2.09375,
      "learning_rate": 4.765822510046061e-06,
      "loss": 1.1869,
      "step": 666
    },
    {
      "epoch": 2.0715953307393,
      "grad_norm": 2.0625,
      "learning_rate": 4.753507123375652e-06,
      "loss": 1.1695,
      "step": 667
    },
    {
      "epoch": 2.074708171206226,
      "grad_norm": 2.046875,
      "learning_rate": 4.741193235578953e-06,
      "loss": 1.1401,
      "step": 668
    },
    {
      "epoch": 2.077821011673152,
      "grad_norm": 2.15625,
      "learning_rate": 4.728880921534247e-06,
      "loss": 1.2059,
      "step": 669
    },
    {
      "epoch": 2.080933852140078,
      "grad_norm": 2.03125,
      "learning_rate": 4.716570256110242e-06,
      "loss": 1.1971,
      "step": 670
    },
    {
      "epoch": 2.084046692607004,
      "grad_norm": 2.1875,
      "learning_rate": 4.704261314165623e-06,
      "loss": 1.1764,
      "step": 671
    },
    {
      "epoch": 2.08715953307393,
      "grad_norm": 2.09375,
      "learning_rate": 4.691954170548599e-06,
      "loss": 1.2425,
      "step": 672
    },
    {
      "epoch": 2.08715953307393,
      "eval_loss": 3.2765259742736816,
      "eval_runtime": 43.0614,
      "eval_samples_per_second": 113.977,
      "eval_steps_per_second": 11.402,
      "step": 672
    },
    {
      "epoch": 2.090272373540856,
      "grad_norm": 2.015625,
      "learning_rate": 4.679648900096436e-06,
      "loss": 1.1348,
      "step": 673
    },
    {
      "epoch": 2.093385214007782,
      "grad_norm": 2.171875,
      "learning_rate": 4.667345577635018e-06,
      "loss": 1.2403,
      "step": 674
    },
    {
      "epoch": 2.0964980544747083,
      "grad_norm": 2.25,
      "learning_rate": 4.6550442779783755e-06,
      "loss": 1.1199,
      "step": 675
    },
    {
      "epoch": 2.0996108949416343,
      "grad_norm": 2.109375,
      "learning_rate": 4.642745075928246e-06,
      "loss": 1.2197,
      "step": 676
    },
    {
      "epoch": 2.1027237354085604,
      "grad_norm": 2.046875,
      "learning_rate": 4.6304480462736075e-06,
      "loss": 1.239,
      "step": 677
    },
    {
      "epoch": 2.1058365758754864,
      "grad_norm": 1.984375,
      "learning_rate": 4.618153263790232e-06,
      "loss": 1.1698,
      "step": 678
    },
    {
      "epoch": 2.1089494163424125,
      "grad_norm": 2.078125,
      "learning_rate": 4.605860803240217e-06,
      "loss": 1.1116,
      "step": 679
    },
    {
      "epoch": 2.1120622568093386,
      "grad_norm": 2.046875,
      "learning_rate": 4.593570739371557e-06,
      "loss": 1.188,
      "step": 680
    },
    {
      "epoch": 2.1151750972762646,
      "grad_norm": 2.03125,
      "learning_rate": 4.581283146917658e-06,
      "loss": 1.1811,
      "step": 681
    },
    {
      "epoch": 2.1182879377431907,
      "grad_norm": 2.15625,
      "learning_rate": 4.568998100596903e-06,
      "loss": 1.1754,
      "step": 682
    },
    {
      "epoch": 2.1214007782101167,
      "grad_norm": 2.171875,
      "learning_rate": 4.556715675112197e-06,
      "loss": 1.1726,
      "step": 683
    },
    {
      "epoch": 2.124513618677043,
      "grad_norm": 2.0,
      "learning_rate": 4.5444359451505e-06,
      "loss": 1.1446,
      "step": 684
    },
    {
      "epoch": 2.127626459143969,
      "grad_norm": 2.109375,
      "learning_rate": 4.532158985382391e-06,
      "loss": 1.2274,
      "step": 685
    },
    {
      "epoch": 2.130739299610895,
      "grad_norm": 2.046875,
      "learning_rate": 4.5198848704615915e-06,
      "loss": 1.1853,
      "step": 686
    },
    {
      "epoch": 2.133852140077821,
      "grad_norm": 2.09375,
      "learning_rate": 4.507613675024536e-06,
      "loss": 1.1777,
      "step": 687
    },
    {
      "epoch": 2.136964980544747,
      "grad_norm": 2.1875,
      "learning_rate": 4.495345473689897e-06,
      "loss": 1.1696,
      "step": 688
    },
    {
      "epoch": 2.140077821011673,
      "grad_norm": 2.0625,
      "learning_rate": 4.4830803410581506e-06,
      "loss": 1.188,
      "step": 689
    },
    {
      "epoch": 2.143190661478599,
      "grad_norm": 2.15625,
      "learning_rate": 4.470818351711098e-06,
      "loss": 1.1923,
      "step": 690
    },
    {
      "epoch": 2.146303501945525,
      "grad_norm": 2.140625,
      "learning_rate": 4.458559580211445e-06,
      "loss": 1.1903,
      "step": 691
    },
    {
      "epoch": 2.1494163424124513,
      "grad_norm": 2.125,
      "learning_rate": 4.4463041011023144e-06,
      "loss": 1.1476,
      "step": 692
    },
    {
      "epoch": 2.1525291828793773,
      "grad_norm": 1.984375,
      "learning_rate": 4.434051988906816e-06,
      "loss": 1.1725,
      "step": 693
    },
    {
      "epoch": 2.1525291828793773,
      "eval_loss": 3.2997207641601562,
      "eval_runtime": 42.9302,
      "eval_samples_per_second": 114.325,
      "eval_steps_per_second": 11.437,
      "step": 693
    },
    {
      "epoch": 2.1556420233463034,
      "grad_norm": 2.078125,
      "learning_rate": 4.421803318127587e-06,
      "loss": 1.1651,
      "step": 694
    },
    {
      "epoch": 2.1587548638132295,
      "grad_norm": 2.03125,
      "learning_rate": 4.409558163246332e-06,
      "loss": 1.2687,
      "step": 695
    },
    {
      "epoch": 2.1618677042801555,
      "grad_norm": 2.1875,
      "learning_rate": 4.397316598723385e-06,
      "loss": 1.2086,
      "step": 696
    },
    {
      "epoch": 2.1649805447470816,
      "grad_norm": 2.25,
      "learning_rate": 4.385078698997241e-06,
      "loss": 1.1491,
      "step": 697
    },
    {
      "epoch": 2.1680933852140076,
      "grad_norm": 2.0625,
      "learning_rate": 4.372844538484111e-06,
      "loss": 1.1862,
      "step": 698
    },
    {
      "epoch": 2.1712062256809337,
      "grad_norm": 2.109375,
      "learning_rate": 4.3606141915774695e-06,
      "loss": 1.1655,
      "step": 699
    },
    {
      "epoch": 2.1743190661478597,
      "grad_norm": 1.96875,
      "learning_rate": 4.3483877326476055e-06,
      "loss": 1.102,
      "step": 700
    },
    {
      "epoch": 2.177431906614786,
      "grad_norm": 2.078125,
      "learning_rate": 4.336165236041157e-06,
      "loss": 1.1863,
      "step": 701
    },
    {
      "epoch": 2.180544747081712,
      "grad_norm": 2.046875,
      "learning_rate": 4.323946776080674e-06,
      "loss": 1.1347,
      "step": 702
    },
    {
      "epoch": 2.1836575875486384,
      "grad_norm": 2.0625,
      "learning_rate": 4.31173242706416e-06,
      "loss": 1.172,
      "step": 703
    },
    {
      "epoch": 2.1867704280155644,
      "grad_norm": 2.0625,
      "learning_rate": 4.299522263264619e-06,
      "loss": 1.152,
      "step": 704
    },
    {
      "epoch": 2.1898832684824905,
      "grad_norm": 2.171875,
      "learning_rate": 4.28731635892961e-06,
      "loss": 1.2071,
      "step": 705
    },
    {
      "epoch": 2.1929961089494165,
      "grad_norm": 1.96875,
      "learning_rate": 4.275114788280779e-06,
      "loss": 1.0892,
      "step": 706
    },
    {
      "epoch": 2.1961089494163426,
      "grad_norm": 2.09375,
      "learning_rate": 4.262917625513439e-06,
      "loss": 1.1899,
      "step": 707
    },
    {
      "epoch": 2.1992217898832687,
      "grad_norm": 2.015625,
      "learning_rate": 4.250724944796082e-06,
      "loss": 1.1914,
      "step": 708
    },
    {
      "epoch": 2.2023346303501947,
      "grad_norm": 2.078125,
      "learning_rate": 4.238536820269955e-06,
      "loss": 1.1188,
      "step": 709
    },
    {
      "epoch": 2.205447470817121,
      "grad_norm": 1.96875,
      "learning_rate": 4.226353326048594e-06,
      "loss": 1.1404,
      "step": 710
    },
    {
      "epoch": 2.208560311284047,
      "grad_norm": 2.078125,
      "learning_rate": 4.214174536217385e-06,
      "loss": 1.236,
      "step": 711
    },
    {
      "epoch": 2.211673151750973,
      "grad_norm": 2.09375,
      "learning_rate": 4.2020005248331056e-06,
      "loss": 1.1953,
      "step": 712
    },
    {
      "epoch": 2.214785992217899,
      "grad_norm": 2.046875,
      "learning_rate": 4.189831365923471e-06,
      "loss": 1.2732,
      "step": 713
    },
    {
      "epoch": 2.217898832684825,
      "grad_norm": 2.25,
      "learning_rate": 4.177667133486698e-06,
      "loss": 1.2112,
      "step": 714
    },
    {
      "epoch": 2.217898832684825,
      "eval_loss": 3.287874698638916,
      "eval_runtime": 42.942,
      "eval_samples_per_second": 114.294,
      "eval_steps_per_second": 11.434,
      "step": 714
    },
    {
      "epoch": 2.221011673151751,
      "grad_norm": 2.0625,
      "learning_rate": 4.165507901491042e-06,
      "loss": 1.2023,
      "step": 715
    },
    {
      "epoch": 2.224124513618677,
      "grad_norm": 2.140625,
      "learning_rate": 4.153353743874353e-06,
      "loss": 1.1439,
      "step": 716
    },
    {
      "epoch": 2.227237354085603,
      "grad_norm": 2.125,
      "learning_rate": 4.14120473454362e-06,
      "loss": 1.2107,
      "step": 717
    },
    {
      "epoch": 2.2303501945525293,
      "grad_norm": 2.125,
      "learning_rate": 4.129060947374537e-06,
      "loss": 1.1555,
      "step": 718
    },
    {
      "epoch": 2.2334630350194553,
      "grad_norm": 2.046875,
      "learning_rate": 4.11692245621103e-06,
      "loss": 1.165,
      "step": 719
    },
    {
      "epoch": 2.2365758754863814,
      "grad_norm": 2.0625,
      "learning_rate": 4.104789334864831e-06,
      "loss": 1.2151,
      "step": 720
    },
    {
      "epoch": 2.2396887159533074,
      "grad_norm": 2.21875,
      "learning_rate": 4.092661657115016e-06,
      "loss": 1.2173,
      "step": 721
    },
    {
      "epoch": 2.2428015564202335,
      "grad_norm": 2.125,
      "learning_rate": 4.080539496707559e-06,
      "loss": 1.1734,
      "step": 722
    },
    {
      "epoch": 2.2459143968871595,
      "grad_norm": 2.046875,
      "learning_rate": 4.0684229273548845e-06,
      "loss": 1.1991,
      "step": 723
    },
    {
      "epoch": 2.2490272373540856,
      "grad_norm": 2.0625,
      "learning_rate": 4.056312022735417e-06,
      "loss": 1.1924,
      "step": 724
    },
    {
      "epoch": 2.2521400778210117,
      "grad_norm": 1.9921875,
      "learning_rate": 4.04420685649314e-06,
      "loss": 1.1325,
      "step": 725
    },
    {
      "epoch": 2.2552529182879377,
      "grad_norm": 2.125,
      "learning_rate": 4.0321075022371356e-06,
      "loss": 1.1987,
      "step": 726
    },
    {
      "epoch": 2.258365758754864,
      "grad_norm": 2.0625,
      "learning_rate": 4.020014033541153e-06,
      "loss": 1.1968,
      "step": 727
    },
    {
      "epoch": 2.26147859922179,
      "grad_norm": 2.0,
      "learning_rate": 4.007926523943145e-06,
      "loss": 1.1738,
      "step": 728
    },
    {
      "epoch": 2.264591439688716,
      "grad_norm": 2.078125,
      "learning_rate": 3.995845046944836e-06,
      "loss": 1.2113,
      "step": 729
    },
    {
      "epoch": 2.267704280155642,
      "grad_norm": 2.125,
      "learning_rate": 3.983769676011257e-06,
      "loss": 1.1855,
      "step": 730
    },
    {
      "epoch": 2.270817120622568,
      "grad_norm": 1.9609375,
      "learning_rate": 3.9717004845703175e-06,
      "loss": 1.1826,
      "step": 731
    },
    {
      "epoch": 2.273929961089494,
      "grad_norm": 2.078125,
      "learning_rate": 3.959637546012349e-06,
      "loss": 1.112,
      "step": 732
    },
    {
      "epoch": 2.27704280155642,
      "grad_norm": 2.046875,
      "learning_rate": 3.94758093368966e-06,
      "loss": 1.1504,
      "step": 733
    },
    {
      "epoch": 2.280155642023346,
      "grad_norm": 2.171875,
      "learning_rate": 3.93553072091609e-06,
      "loss": 1.1842,
      "step": 734
    },
    {
      "epoch": 2.2832684824902723,
      "grad_norm": 1.984375,
      "learning_rate": 3.923486980966562e-06,
      "loss": 1.1627,
      "step": 735
    },
    {
      "epoch": 2.2832684824902723,
      "eval_loss": 3.3009390830993652,
      "eval_runtime": 43.0229,
      "eval_samples_per_second": 114.079,
      "eval_steps_per_second": 11.413,
      "step": 735
    },
    {
      "epoch": 2.2863813229571983,
      "grad_norm": 1.9296875,
      "learning_rate": 3.911449787076645e-06,
      "loss": 1.1457,
      "step": 736
    },
    {
      "epoch": 2.2894941634241244,
      "grad_norm": 2.109375,
      "learning_rate": 3.899419212442097e-06,
      "loss": 1.1701,
      "step": 737
    },
    {
      "epoch": 2.2926070038910504,
      "grad_norm": 2.109375,
      "learning_rate": 3.887395330218429e-06,
      "loss": 1.174,
      "step": 738
    },
    {
      "epoch": 2.2957198443579765,
      "grad_norm": 2.015625,
      "learning_rate": 3.875378213520453e-06,
      "loss": 1.1557,
      "step": 739
    },
    {
      "epoch": 2.2988326848249026,
      "grad_norm": 2.03125,
      "learning_rate": 3.863367935421848e-06,
      "loss": 1.1732,
      "step": 740
    },
    {
      "epoch": 2.3019455252918286,
      "grad_norm": 2.109375,
      "learning_rate": 3.8513645689547005e-06,
      "loss": 1.2185,
      "step": 741
    },
    {
      "epoch": 2.3050583657587547,
      "grad_norm": 2.0625,
      "learning_rate": 3.839368187109076e-06,
      "loss": 1.1686,
      "step": 742
    },
    {
      "epoch": 2.3081712062256807,
      "grad_norm": 2.015625,
      "learning_rate": 3.827378862832562e-06,
      "loss": 1.1905,
      "step": 743
    },
    {
      "epoch": 2.311284046692607,
      "grad_norm": 1.9375,
      "learning_rate": 3.815396669029836e-06,
      "loss": 1.1317,
      "step": 744
    },
    {
      "epoch": 2.314396887159533,
      "grad_norm": 2.046875,
      "learning_rate": 3.803421678562213e-06,
      "loss": 1.2107,
      "step": 745
    },
    {
      "epoch": 2.317509727626459,
      "grad_norm": 2.09375,
      "learning_rate": 3.7914539642472035e-06,
      "loss": 1.1715,
      "step": 746
    },
    {
      "epoch": 2.3206225680933854,
      "grad_norm": 2.0,
      "learning_rate": 3.77949359885808e-06,
      "loss": 1.1503,
      "step": 747
    },
    {
      "epoch": 2.3237354085603115,
      "grad_norm": 2.109375,
      "learning_rate": 3.7675406551234213e-06,
      "loss": 1.2384,
      "step": 748
    },
    {
      "epoch": 2.3268482490272375,
      "grad_norm": 2.109375,
      "learning_rate": 3.755595205726682e-06,
      "loss": 1.2157,
      "step": 749
    },
    {
      "epoch": 2.3299610894941636,
      "grad_norm": 2.046875,
      "learning_rate": 3.7436573233057392e-06,
      "loss": 1.1955,
      "step": 750
    },
    {
      "epoch": 2.3330739299610896,
      "grad_norm": 1.9453125,
      "learning_rate": 3.731727080452464e-06,
      "loss": 1.1629,
      "step": 751
    },
    {
      "epoch": 2.3361867704280157,
      "grad_norm": 2.15625,
      "learning_rate": 3.7198045497122647e-06,
      "loss": 1.2091,
      "step": 752
    },
    {
      "epoch": 2.3392996108949418,
      "grad_norm": 2.140625,
      "learning_rate": 3.707889803583661e-06,
      "loss": 1.2029,
      "step": 753
    },
    {
      "epoch": 2.342412451361868,
      "grad_norm": 2.109375,
      "learning_rate": 3.6959829145178293e-06,
      "loss": 1.1842,
      "step": 754
    },
    {
      "epoch": 2.345525291828794,
      "grad_norm": 2.125,
      "learning_rate": 3.684083954918176e-06,
      "loss": 1.2187,
      "step": 755
    },
    {
      "epoch": 2.34863813229572,
      "grad_norm": 2.109375,
      "learning_rate": 3.672192997139885e-06,
      "loss": 1.2418,
      "step": 756
    },
    {
      "epoch": 2.34863813229572,
      "eval_loss": 3.2847952842712402,
      "eval_runtime": 42.9189,
      "eval_samples_per_second": 114.355,
      "eval_steps_per_second": 11.44,
      "step": 756
    },
    {
      "epoch": 2.351750972762646,
      "grad_norm": 2.078125,
      "learning_rate": 3.6603101134894826e-06,
      "loss": 1.182,
      "step": 757
    },
    {
      "epoch": 2.354863813229572,
      "grad_norm": 2.015625,
      "learning_rate": 3.6484353762244007e-06,
      "loss": 1.1894,
      "step": 758
    },
    {
      "epoch": 2.357976653696498,
      "grad_norm": 2.015625,
      "learning_rate": 3.6365688575525315e-06,
      "loss": 1.1311,
      "step": 759
    },
    {
      "epoch": 2.361089494163424,
      "grad_norm": 2.15625,
      "learning_rate": 3.6247106296317965e-06,
      "loss": 1.2006,
      "step": 760
    },
    {
      "epoch": 2.3642023346303502,
      "grad_norm": 2.015625,
      "learning_rate": 3.6128607645696967e-06,
      "loss": 1.145,
      "step": 761
    },
    {
      "epoch": 2.3673151750972763,
      "grad_norm": 2.03125,
      "learning_rate": 3.601019334422887e-06,
      "loss": 1.2284,
      "step": 762
    },
    {
      "epoch": 2.3704280155642024,
      "grad_norm": 1.9921875,
      "learning_rate": 3.5891864111967224e-06,
      "loss": 1.1433,
      "step": 763
    },
    {
      "epoch": 2.3735408560311284,
      "grad_norm": 2.0625,
      "learning_rate": 3.5773620668448384e-06,
      "loss": 1.1372,
      "step": 764
    },
    {
      "epoch": 2.3766536964980545,
      "grad_norm": 2.140625,
      "learning_rate": 3.5655463732686975e-06,
      "loss": 1.1784,
      "step": 765
    },
    {
      "epoch": 2.3797665369649805,
      "grad_norm": 2.234375,
      "learning_rate": 3.553739402317162e-06,
      "loss": 1.2172,
      "step": 766
    },
    {
      "epoch": 2.3828793774319066,
      "grad_norm": 2.15625,
      "learning_rate": 3.5419412257860536e-06,
      "loss": 1.2109,
      "step": 767
    },
    {
      "epoch": 2.3859922178988326,
      "grad_norm": 2.234375,
      "learning_rate": 3.530151915417712e-06,
      "loss": 1.1934,
      "step": 768
    },
    {
      "epoch": 2.3891050583657587,
      "grad_norm": 2.09375,
      "learning_rate": 3.5183715429005717e-06,
      "loss": 1.1769,
      "step": 769
    },
    {
      "epoch": 2.3922178988326848,
      "grad_norm": 2.21875,
      "learning_rate": 3.5066001798687106e-06,
      "loss": 1.2034,
      "step": 770
    },
    {
      "epoch": 2.395330739299611,
      "grad_norm": 2.1875,
      "learning_rate": 3.4948378979014266e-06,
      "loss": 1.1938,
      "step": 771
    },
    {
      "epoch": 2.398443579766537,
      "grad_norm": 2.171875,
      "learning_rate": 3.483084768522794e-06,
      "loss": 1.1838,
      "step": 772
    },
    {
      "epoch": 2.401556420233463,
      "grad_norm": 2.15625,
      "learning_rate": 3.471340863201237e-06,
      "loss": 1.1273,
      "step": 773
    },
    {
      "epoch": 2.404669260700389,
      "grad_norm": 2.0625,
      "learning_rate": 3.4596062533490813e-06,
      "loss": 1.1895,
      "step": 774
    },
    {
      "epoch": 2.407782101167315,
      "grad_norm": 2.0625,
      "learning_rate": 3.4478810103221393e-06,
      "loss": 1.1818,
      "step": 775
    },
    {
      "epoch": 2.410894941634241,
      "grad_norm": 2.046875,
      "learning_rate": 3.4361652054192577e-06,
      "loss": 1.2297,
      "step": 776
    },
    {
      "epoch": 2.414007782101167,
      "grad_norm": 2.0625,
      "learning_rate": 3.424458909881897e-06,
      "loss": 1.1648,
      "step": 777
    },
    {
      "epoch": 2.414007782101167,
      "eval_loss": 3.3048324584960938,
      "eval_runtime": 243.7491,
      "eval_samples_per_second": 20.135,
      "eval_steps_per_second": 2.014,
      "step": 777
    },
    {
      "epoch": 2.4171206225680932,
      "grad_norm": 2.0625,
      "learning_rate": 3.41276219489369e-06,
      "loss": 1.2064,
      "step": 778
    },
    {
      "epoch": 2.4202334630350193,
      "grad_norm": 2.046875,
      "learning_rate": 3.4010751315800105e-06,
      "loss": 1.1375,
      "step": 779
    },
    {
      "epoch": 2.4233463035019454,
      "grad_norm": 2.125,
      "learning_rate": 3.389397791007548e-06,
      "loss": 1.1842,
      "step": 780
    },
    {
      "epoch": 2.4264591439688714,
      "grad_norm": 2.15625,
      "learning_rate": 3.3777302441838645e-06,
      "loss": 1.2123,
      "step": 781
    },
    {
      "epoch": 2.4295719844357975,
      "grad_norm": 2.03125,
      "learning_rate": 3.3660725620569736e-06,
      "loss": 1.1457,
      "step": 782
    },
    {
      "epoch": 2.432684824902724,
      "grad_norm": 2.15625,
      "learning_rate": 3.354424815514897e-06,
      "loss": 1.1927,
      "step": 783
    },
    {
      "epoch": 2.43579766536965,
      "grad_norm": 2.03125,
      "learning_rate": 3.3427870753852478e-06,
      "loss": 1.1588,
      "step": 784
    },
    {
      "epoch": 2.438910505836576,
      "grad_norm": 2.078125,
      "learning_rate": 3.3311594124347826e-06,
      "loss": 1.1572,
      "step": 785
    },
    {
      "epoch": 2.442023346303502,
      "grad_norm": 2.078125,
      "learning_rate": 3.3195418973689913e-06,
      "loss": 1.2321,
      "step": 786
    },
    {
      "epoch": 2.445136186770428,
      "grad_norm": 2.125,
      "learning_rate": 3.307934600831648e-06,
      "loss": 1.2187,
      "step": 787
    },
    {
      "epoch": 2.4482490272373543,
      "grad_norm": 2.03125,
      "learning_rate": 3.296337593404395e-06,
      "loss": 1.1736,
      "step": 788
    },
    {
      "epoch": 2.4513618677042803,
      "grad_norm": 2.171875,
      "learning_rate": 3.2847509456063076e-06,
      "loss": 1.1947,
      "step": 789
    },
    {
      "epoch": 2.4544747081712064,
      "grad_norm": 2.15625,
      "learning_rate": 3.273174727893463e-06,
      "loss": 1.2165,
      "step": 790
    },
    {
      "epoch": 2.4575875486381324,
      "grad_norm": 2.015625,
      "learning_rate": 3.2616090106585195e-06,
      "loss": 1.1586,
      "step": 791
    },
    {
      "epoch": 2.4607003891050585,
      "grad_norm": 2.046875,
      "learning_rate": 3.250053864230281e-06,
      "loss": 1.1948,
      "step": 792
    },
    {
      "epoch": 2.4638132295719846,
      "grad_norm": 2.046875,
      "learning_rate": 3.2385093588732756e-06,
      "loss": 1.1544,
      "step": 793
    },
    {
      "epoch": 2.4669260700389106,
      "grad_norm": 2.03125,
      "learning_rate": 3.226975564787322e-06,
      "loss": 1.1431,
      "step": 794
    },
    {
      "epoch": 2.4700389105058367,
      "grad_norm": 2.15625,
      "learning_rate": 3.215452552107109e-06,
      "loss": 1.1509,
      "step": 795
    },
    {
      "epoch": 2.4731517509727627,
      "grad_norm": 2.0625,
      "learning_rate": 3.2039403909017603e-06,
      "loss": 1.2013,
      "step": 796
    },
    {
      "epoch": 2.476264591439689,
      "grad_norm": 2.140625,
      "learning_rate": 3.1924391511744223e-06,
      "loss": 1.1019,
      "step": 797
    },
    {
      "epoch": 2.479377431906615,
      "grad_norm": 2.078125,
      "learning_rate": 3.180948902861822e-06,
      "loss": 1.2011,
      "step": 798
    },
    {
      "epoch": 2.479377431906615,
      "eval_loss": 3.2985522747039795,
      "eval_runtime": 42.9344,
      "eval_samples_per_second": 114.314,
      "eval_steps_per_second": 11.436,
      "step": 798
    },
    {
      "epoch": 2.482490272373541,
      "grad_norm": 2.125,
      "learning_rate": 3.169469715833855e-06,
      "loss": 1.1646,
      "step": 799
    },
    {
      "epoch": 2.485603112840467,
      "grad_norm": 2.03125,
      "learning_rate": 3.1580016598931546e-06,
      "loss": 1.1283,
      "step": 800
    },
    {
      "epoch": 2.488715953307393,
      "grad_norm": 2.09375,
      "learning_rate": 3.1465448047746626e-06,
      "loss": 1.1744,
      "step": 801
    },
    {
      "epoch": 2.491828793774319,
      "grad_norm": 2.109375,
      "learning_rate": 3.1350992201452212e-06,
      "loss": 1.1623,
      "step": 802
    },
    {
      "epoch": 2.494941634241245,
      "grad_norm": 2.046875,
      "learning_rate": 3.12366497560313e-06,
      "loss": 1.1813,
      "step": 803
    },
    {
      "epoch": 2.498054474708171,
      "grad_norm": 2.140625,
      "learning_rate": 3.1122421406777365e-06,
      "loss": 1.2278,
      "step": 804
    },
    {
      "epoch": 2.5011673151750973,
      "grad_norm": 2.109375,
      "learning_rate": 3.100830784829006e-06,
      "loss": 1.1908,
      "step": 805
    },
    {
      "epoch": 2.5042801556420233,
      "grad_norm": 2.3125,
      "learning_rate": 3.0894309774471064e-06,
      "loss": 1.2309,
      "step": 806
    },
    {
      "epoch": 2.5073929961089494,
      "grad_norm": 2.15625,
      "learning_rate": 3.0780427878519727e-06,
      "loss": 1.1926,
      "step": 807
    },
    {
      "epoch": 2.5105058365758754,
      "grad_norm": 2.140625,
      "learning_rate": 3.0666662852929063e-06,
      "loss": 1.1554,
      "step": 808
    },
    {
      "epoch": 2.5136186770428015,
      "grad_norm": 2.203125,
      "learning_rate": 3.0553015389481328e-06,
      "loss": 1.1879,
      "step": 809
    },
    {
      "epoch": 2.5167315175097276,
      "grad_norm": 2.109375,
      "learning_rate": 3.0439486179243927e-06,
      "loss": 1.1447,
      "step": 810
    },
    {
      "epoch": 2.5198443579766536,
      "grad_norm": 2.046875,
      "learning_rate": 3.0326075912565244e-06,
      "loss": 1.1789,
      "step": 811
    },
    {
      "epoch": 2.5229571984435797,
      "grad_norm": 2.109375,
      "learning_rate": 3.021278527907029e-06,
      "loss": 1.2125,
      "step": 812
    },
    {
      "epoch": 2.5260700389105057,
      "grad_norm": 2.03125,
      "learning_rate": 3.0099614967656725e-06,
      "loss": 1.1863,
      "step": 813
    },
    {
      "epoch": 2.529182879377432,
      "grad_norm": 2.078125,
      "learning_rate": 2.998656566649046e-06,
      "loss": 1.1595,
      "step": 814
    },
    {
      "epoch": 2.532295719844358,
      "grad_norm": 2.046875,
      "learning_rate": 2.9873638063001633e-06,
      "loss": 1.2422,
      "step": 815
    },
    {
      "epoch": 2.535408560311284,
      "grad_norm": 2.046875,
      "learning_rate": 2.976083284388031e-06,
      "loss": 1.2126,
      "step": 816
    },
    {
      "epoch": 2.53852140077821,
      "grad_norm": 2.09375,
      "learning_rate": 2.9648150695072413e-06,
      "loss": 1.1802,
      "step": 817
    },
    {
      "epoch": 2.541634241245136,
      "grad_norm": 1.984375,
      "learning_rate": 2.953559230177545e-06,
      "loss": 1.1175,
      "step": 818
    },
    {
      "epoch": 2.544747081712062,
      "grad_norm": 2.125,
      "learning_rate": 2.942315834843446e-06,
      "loss": 1.2197,
      "step": 819
    },
    {
      "epoch": 2.544747081712062,
      "eval_loss": 3.2813668251037598,
      "eval_runtime": 43.2418,
      "eval_samples_per_second": 113.501,
      "eval_steps_per_second": 11.355,
      "step": 819
    },
    {
      "epoch": 2.547859922178988,
      "grad_norm": 2.078125,
      "learning_rate": 2.9310849518737703e-06,
      "loss": 1.162,
      "step": 820
    },
    {
      "epoch": 2.550972762645914,
      "grad_norm": 2.078125,
      "learning_rate": 2.919866649561265e-06,
      "loss": 1.216,
      "step": 821
    },
    {
      "epoch": 2.5540856031128403,
      "grad_norm": 2.078125,
      "learning_rate": 2.9086609961221758e-06,
      "loss": 1.132,
      "step": 822
    },
    {
      "epoch": 2.5571984435797663,
      "grad_norm": 2.0625,
      "learning_rate": 2.897468059695829e-06,
      "loss": 1.2035,
      "step": 823
    },
    {
      "epoch": 2.5603112840466924,
      "grad_norm": 2.09375,
      "learning_rate": 2.8862879083442274e-06,
      "loss": 1.1952,
      "step": 824
    },
    {
      "epoch": 2.5634241245136185,
      "grad_norm": 2.03125,
      "learning_rate": 2.8751206100516273e-06,
      "loss": 1.1815,
      "step": 825
    },
    {
      "epoch": 2.5665369649805445,
      "grad_norm": 2.0625,
      "learning_rate": 2.8639662327241292e-06,
      "loss": 1.264,
      "step": 826
    },
    {
      "epoch": 2.5696498054474706,
      "grad_norm": 2.015625,
      "learning_rate": 2.8528248441892593e-06,
      "loss": 1.1632,
      "step": 827
    },
    {
      "epoch": 2.5727626459143966,
      "grad_norm": 2.171875,
      "learning_rate": 2.8416965121955696e-06,
      "loss": 1.1482,
      "step": 828
    },
    {
      "epoch": 2.5758754863813227,
      "grad_norm": 2.078125,
      "learning_rate": 2.83058130441221e-06,
      "loss": 1.1939,
      "step": 829
    },
    {
      "epoch": 2.578988326848249,
      "grad_norm": 2.015625,
      "learning_rate": 2.8194792884285293e-06,
      "loss": 1.1394,
      "step": 830
    },
    {
      "epoch": 2.5821011673151752,
      "grad_norm": 2.09375,
      "learning_rate": 2.8083905317536604e-06,
      "loss": 1.2135,
      "step": 831
    },
    {
      "epoch": 2.5852140077821013,
      "grad_norm": 2.21875,
      "learning_rate": 2.797315101816103e-06,
      "loss": 1.224,
      "step": 832
    },
    {
      "epoch": 2.5883268482490274,
      "grad_norm": 2.109375,
      "learning_rate": 2.786253065963329e-06,
      "loss": 1.2033,
      "step": 833
    },
    {
      "epoch": 2.5914396887159534,
      "grad_norm": 2.015625,
      "learning_rate": 2.775204491461353e-06,
      "loss": 1.2154,
      "step": 834
    },
    {
      "epoch": 2.5945525291828795,
      "grad_norm": 2.015625,
      "learning_rate": 2.764169445494341e-06,
      "loss": 1.1937,
      "step": 835
    },
    {
      "epoch": 2.5976653696498055,
      "grad_norm": 2.0625,
      "learning_rate": 2.7531479951641928e-06,
      "loss": 1.1827,
      "step": 836
    },
    {
      "epoch": 2.6007782101167316,
      "grad_norm": 2.078125,
      "learning_rate": 2.742140207490139e-06,
      "loss": 1.149,
      "step": 837
    },
    {
      "epoch": 2.6038910505836577,
      "grad_norm": 2.03125,
      "learning_rate": 2.7311461494083216e-06,
      "loss": 1.1063,
      "step": 838
    },
    {
      "epoch": 2.6070038910505837,
      "grad_norm": 2.1875,
      "learning_rate": 2.7201658877714067e-06,
      "loss": 1.1751,
      "step": 839
    },
    {
      "epoch": 2.61011673151751,
      "grad_norm": 2.140625,
      "learning_rate": 2.709199489348156e-06,
      "loss": 1.1962,
      "step": 840
    },
    {
      "epoch": 2.61011673151751,
      "eval_loss": 3.2692527770996094,
      "eval_runtime": 42.9318,
      "eval_samples_per_second": 114.321,
      "eval_steps_per_second": 11.437,
      "step": 840
    },
    {
      "epoch": 2.613229571984436,
      "grad_norm": 1.9921875,
      "learning_rate": 2.6982470208230405e-06,
      "loss": 1.2508,
      "step": 841
    },
    {
      "epoch": 2.616342412451362,
      "grad_norm": 2.09375,
      "learning_rate": 2.687308548795825e-06,
      "loss": 1.1406,
      "step": 842
    },
    {
      "epoch": 2.619455252918288,
      "grad_norm": 2.21875,
      "learning_rate": 2.6763841397811576e-06,
      "loss": 1.2393,
      "step": 843
    },
    {
      "epoch": 2.622568093385214,
      "grad_norm": 2.0625,
      "learning_rate": 2.665473860208182e-06,
      "loss": 1.1256,
      "step": 844
    },
    {
      "epoch": 2.62568093385214,
      "grad_norm": 2.125,
      "learning_rate": 2.654577776420114e-06,
      "loss": 1.2197,
      "step": 845
    },
    {
      "epoch": 2.628793774319066,
      "grad_norm": 2.078125,
      "learning_rate": 2.643695954673855e-06,
      "loss": 1.1413,
      "step": 846
    },
    {
      "epoch": 2.631906614785992,
      "grad_norm": 2.0,
      "learning_rate": 2.6328284611395773e-06,
      "loss": 1.1676,
      "step": 847
    },
    {
      "epoch": 2.6350194552529183,
      "grad_norm": 2.09375,
      "learning_rate": 2.621975361900332e-06,
      "loss": 1.1479,
      "step": 848
    },
    {
      "epoch": 2.6381322957198443,
      "grad_norm": 2.125,
      "learning_rate": 2.6111367229516317e-06,
      "loss": 1.2003,
      "step": 849
    },
    {
      "epoch": 2.6412451361867704,
      "grad_norm": 2.203125,
      "learning_rate": 2.6003126102010696e-06,
      "loss": 1.1806,
      "step": 850
    },
    {
      "epoch": 2.6443579766536964,
      "grad_norm": 2.078125,
      "learning_rate": 2.5895030894678985e-06,
      "loss": 1.232,
      "step": 851
    },
    {
      "epoch": 2.6474708171206225,
      "grad_norm": 2.140625,
      "learning_rate": 2.578708226482646e-06,
      "loss": 1.246,
      "step": 852
    },
    {
      "epoch": 2.6505836575875485,
      "grad_norm": 2.078125,
      "learning_rate": 2.5679280868867105e-06,
      "loss": 1.153,
      "step": 853
    },
    {
      "epoch": 2.6536964980544746,
      "grad_norm": 2.21875,
      "learning_rate": 2.557162736231952e-06,
      "loss": 1.2528,
      "step": 854
    },
    {
      "epoch": 2.6568093385214007,
      "grad_norm": 2.140625,
      "learning_rate": 2.5464122399803126e-06,
      "loss": 1.1769,
      "step": 855
    },
    {
      "epoch": 2.6599221789883267,
      "grad_norm": 2.0625,
      "learning_rate": 2.535676663503397e-06,
      "loss": 1.1673,
      "step": 856
    },
    {
      "epoch": 2.663035019455253,
      "grad_norm": 2.140625,
      "learning_rate": 2.524956072082093e-06,
      "loss": 1.1874,
      "step": 857
    },
    {
      "epoch": 2.666147859922179,
      "grad_norm": 2.078125,
      "learning_rate": 2.514250530906167e-06,
      "loss": 1.2046,
      "step": 858
    },
    {
      "epoch": 2.669260700389105,
      "grad_norm": 2.0625,
      "learning_rate": 2.503560105073867e-06,
      "loss": 1.2124,
      "step": 859
    },
    {
      "epoch": 2.6723735408560314,
      "grad_norm": 2.078125,
      "learning_rate": 2.4928848595915226e-06,
      "loss": 1.1695,
      "step": 860
    },
    {
      "epoch": 2.6754863813229575,
      "grad_norm": 2.0625,
      "learning_rate": 2.482224859373164e-06,
      "loss": 1.1669,
      "step": 861
    },
    {
      "epoch": 2.6754863813229575,
      "eval_loss": 3.2722625732421875,
      "eval_runtime": 42.9586,
      "eval_samples_per_second": 114.25,
      "eval_steps_per_second": 11.43,
      "step": 861
    },
    {
      "epoch": 2.6785992217898835,
      "grad_norm": 2.140625,
      "learning_rate": 2.4715801692401086e-06,
      "loss": 1.1623,
      "step": 862
    },
    {
      "epoch": 2.6817120622568096,
      "grad_norm": 2.046875,
      "learning_rate": 2.4609508539205827e-06,
      "loss": 1.1396,
      "step": 863
    },
    {
      "epoch": 2.6848249027237356,
      "grad_norm": 2.0625,
      "learning_rate": 2.450336978049322e-06,
      "loss": 1.1819,
      "step": 864
    },
    {
      "epoch": 2.6879377431906617,
      "grad_norm": 2.0625,
      "learning_rate": 2.43973860616717e-06,
      "loss": 1.1677,
      "step": 865
    },
    {
      "epoch": 2.6910505836575878,
      "grad_norm": 2.171875,
      "learning_rate": 2.429155802720704e-06,
      "loss": 1.242,
      "step": 866
    },
    {
      "epoch": 2.694163424124514,
      "grad_norm": 2.15625,
      "learning_rate": 2.418588632061824e-06,
      "loss": 1.2427,
      "step": 867
    },
    {
      "epoch": 2.69727626459144,
      "grad_norm": 2.03125,
      "learning_rate": 2.408037158447375e-06,
      "loss": 1.1534,
      "step": 868
    },
    {
      "epoch": 2.700389105058366,
      "grad_norm": 2.015625,
      "learning_rate": 2.397501446038751e-06,
      "loss": 1.166,
      "step": 869
    },
    {
      "epoch": 2.703501945525292,
      "grad_norm": 2.09375,
      "learning_rate": 2.3869815589015065e-06,
      "loss": 1.209,
      "step": 870
    },
    {
      "epoch": 2.706614785992218,
      "grad_norm": 2.046875,
      "learning_rate": 2.37647756100496e-06,
      "loss": 1.2229,
      "step": 871
    },
    {
      "epoch": 2.709727626459144,
      "grad_norm": 2.03125,
      "learning_rate": 2.3659895162218184e-06,
      "loss": 1.1537,
      "step": 872
    },
    {
      "epoch": 2.71284046692607,
      "grad_norm": 2.140625,
      "learning_rate": 2.355517488327772e-06,
      "loss": 1.1835,
      "step": 873
    },
    {
      "epoch": 2.7159533073929962,
      "grad_norm": 2.09375,
      "learning_rate": 2.345061541001125e-06,
      "loss": 1.2162,
      "step": 874
    },
    {
      "epoch": 2.7190661478599223,
      "grad_norm": 2.03125,
      "learning_rate": 2.334621737822394e-06,
      "loss": 1.2539,
      "step": 875
    },
    {
      "epoch": 2.7221789883268483,
      "grad_norm": 2.09375,
      "learning_rate": 2.324198142273924e-06,
      "loss": 1.2008,
      "step": 876
    },
    {
      "epoch": 2.7252918287937744,
      "grad_norm": 2.140625,
      "learning_rate": 2.3137908177395117e-06,
      "loss": 1.1509,
      "step": 877
    },
    {
      "epoch": 2.7284046692607005,
      "grad_norm": 2.0625,
      "learning_rate": 2.3033998275040047e-06,
      "loss": 1.1845,
      "step": 878
    },
    {
      "epoch": 2.7315175097276265,
      "grad_norm": 2.09375,
      "learning_rate": 2.2930252347529316e-06,
      "loss": 1.1148,
      "step": 879
    },
    {
      "epoch": 2.7346303501945526,
      "grad_norm": 2.09375,
      "learning_rate": 2.282667102572107e-06,
      "loss": 1.1351,
      "step": 880
    },
    {
      "epoch": 2.7377431906614786,
      "grad_norm": 2.125,
      "learning_rate": 2.272325493947257e-06,
      "loss": 1.1882,
      "step": 881
    },
    {
      "epoch": 2.7408560311284047,
      "grad_norm": 2.015625,
      "learning_rate": 2.2620004717636246e-06,
      "loss": 1.1433,
      "step": 882
    },
    {
      "epoch": 2.7408560311284047,
      "eval_loss": 3.2814431190490723,
      "eval_runtime": 42.9282,
      "eval_samples_per_second": 114.33,
      "eval_steps_per_second": 11.438,
      "step": 882
    },
    {
      "epoch": 2.7439688715953308,
      "grad_norm": 2.046875,
      "learning_rate": 2.2516920988056007e-06,
      "loss": 1.1944,
      "step": 883
    },
    {
      "epoch": 2.747081712062257,
      "grad_norm": 2.140625,
      "learning_rate": 2.241400437756328e-06,
      "loss": 1.243,
      "step": 884
    },
    {
      "epoch": 2.750194552529183,
      "grad_norm": 2.1875,
      "learning_rate": 2.2311255511973347e-06,
      "loss": 1.2016,
      "step": 885
    },
    {
      "epoch": 2.753307392996109,
      "grad_norm": 2.125,
      "learning_rate": 2.220867501608145e-06,
      "loss": 1.2263,
      "step": 886
    },
    {
      "epoch": 2.756420233463035,
      "grad_norm": 2.0625,
      "learning_rate": 2.2106263513658954e-06,
      "loss": 1.2324,
      "step": 887
    },
    {
      "epoch": 2.759533073929961,
      "grad_norm": 2.0,
      "learning_rate": 2.2004021627449703e-06,
      "loss": 1.1838,
      "step": 888
    },
    {
      "epoch": 2.762645914396887,
      "grad_norm": 2.125,
      "learning_rate": 2.1901949979166033e-06,
      "loss": 1.1914,
      "step": 889
    },
    {
      "epoch": 2.765758754863813,
      "grad_norm": 2.03125,
      "learning_rate": 2.1800049189485182e-06,
      "loss": 1.1784,
      "step": 890
    },
    {
      "epoch": 2.7688715953307392,
      "grad_norm": 2.03125,
      "learning_rate": 2.1698319878045395e-06,
      "loss": 1.1625,
      "step": 891
    },
    {
      "epoch": 2.7719844357976653,
      "grad_norm": 2.203125,
      "learning_rate": 2.159676266344222e-06,
      "loss": 1.2084,
      "step": 892
    },
    {
      "epoch": 2.7750972762645914,
      "grad_norm": 1.9609375,
      "learning_rate": 2.149537816322465e-06,
      "loss": 1.2009,
      "step": 893
    },
    {
      "epoch": 2.7782101167315174,
      "grad_norm": 2.078125,
      "learning_rate": 2.139416699389153e-06,
      "loss": 1.1779,
      "step": 894
    },
    {
      "epoch": 2.7813229571984435,
      "grad_norm": 2.1875,
      "learning_rate": 2.1293129770887628e-06,
      "loss": 1.1865,
      "step": 895
    },
    {
      "epoch": 2.7844357976653695,
      "grad_norm": 2.046875,
      "learning_rate": 2.119226710860003e-06,
      "loss": 1.2406,
      "step": 896
    },
    {
      "epoch": 2.7875486381322956,
      "grad_norm": 2.0625,
      "learning_rate": 2.1091579620354367e-06,
      "loss": 1.1759,
      "step": 897
    },
    {
      "epoch": 2.7906614785992216,
      "grad_norm": 2.125,
      "learning_rate": 2.0991067918411013e-06,
      "loss": 1.1891,
      "step": 898
    },
    {
      "epoch": 2.7937743190661477,
      "grad_norm": 2.03125,
      "learning_rate": 2.089073261396148e-06,
      "loss": 1.1728,
      "step": 899
    },
    {
      "epoch": 2.7968871595330738,
      "grad_norm": 2.171875,
      "learning_rate": 2.079057431712459e-06,
      "loss": 1.226,
      "step": 900
    },
    {
      "epoch": 2.8,
      "grad_norm": 2.078125,
      "learning_rate": 2.069059363694286e-06,
      "loss": 1.1735,
      "step": 901
    },
    {
      "epoch": 2.803112840466926,
      "grad_norm": 2.078125,
      "learning_rate": 2.059079118137874e-06,
      "loss": 1.1537,
      "step": 902
    },
    {
      "epoch": 2.806225680933852,
      "grad_norm": 2.140625,
      "learning_rate": 2.0491167557310948e-06,
      "loss": 1.2126,
      "step": 903
    },
    {
      "epoch": 2.806225680933852,
      "eval_loss": 3.3012092113494873,
      "eval_runtime": 42.9262,
      "eval_samples_per_second": 114.336,
      "eval_steps_per_second": 11.438,
      "step": 903
    },
    {
      "epoch": 2.809338521400778,
      "grad_norm": 2.140625,
      "learning_rate": 2.0391723370530707e-06,
      "loss": 1.175,
      "step": 904
    },
    {
      "epoch": 2.812451361867704,
      "grad_norm": 2.109375,
      "learning_rate": 2.0292459225738203e-06,
      "loss": 1.1985,
      "step": 905
    },
    {
      "epoch": 2.81556420233463,
      "grad_norm": 2.015625,
      "learning_rate": 2.0193375726538737e-06,
      "loss": 1.2118,
      "step": 906
    },
    {
      "epoch": 2.818677042801556,
      "grad_norm": 2.015625,
      "learning_rate": 2.00944734754392e-06,
      "loss": 1.1593,
      "step": 907
    },
    {
      "epoch": 2.8217898832684822,
      "grad_norm": 2.125,
      "learning_rate": 1.999575307384435e-06,
      "loss": 1.2239,
      "step": 908
    },
    {
      "epoch": 2.8249027237354083,
      "grad_norm": 2.125,
      "learning_rate": 1.9897215122053097e-06,
      "loss": 1.2362,
      "step": 909
    },
    {
      "epoch": 2.8280155642023344,
      "grad_norm": 2.125,
      "learning_rate": 1.979886021925499e-06,
      "loss": 1.2016,
      "step": 910
    },
    {
      "epoch": 2.831128404669261,
      "grad_norm": 2.140625,
      "learning_rate": 1.970068896352641e-06,
      "loss": 1.1758,
      "step": 911
    },
    {
      "epoch": 2.834241245136187,
      "grad_norm": 2.171875,
      "learning_rate": 1.960270195182708e-06,
      "loss": 1.1537,
      "step": 912
    },
    {
      "epoch": 2.837354085603113,
      "grad_norm": 2.078125,
      "learning_rate": 1.9504899779996354e-06,
      "loss": 1.2304,
      "step": 913
    },
    {
      "epoch": 2.840466926070039,
      "grad_norm": 2.140625,
      "learning_rate": 1.940728304274963e-06,
      "loss": 1.1528,
      "step": 914
    },
    {
      "epoch": 2.843579766536965,
      "grad_norm": 2.078125,
      "learning_rate": 1.9309852333674657e-06,
      "loss": 1.1887,
      "step": 915
    },
    {
      "epoch": 2.846692607003891,
      "grad_norm": 2.015625,
      "learning_rate": 1.9212608245228047e-06,
      "loss": 1.172,
      "step": 916
    },
    {
      "epoch": 2.849805447470817,
      "grad_norm": 2.09375,
      "learning_rate": 1.9115551368731566e-06,
      "loss": 1.2059,
      "step": 917
    },
    {
      "epoch": 2.8529182879377433,
      "grad_norm": 2.046875,
      "learning_rate": 1.9018682294368613e-06,
      "loss": 1.1897,
      "step": 918
    },
    {
      "epoch": 2.8560311284046693,
      "grad_norm": 2.140625,
      "learning_rate": 1.8922001611180602e-06,
      "loss": 1.1567,
      "step": 919
    },
    {
      "epoch": 2.8591439688715954,
      "grad_norm": 2.015625,
      "learning_rate": 1.8825509907063328e-06,
      "loss": 1.2059,
      "step": 920
    },
    {
      "epoch": 2.8622568093385214,
      "grad_norm": 2.09375,
      "learning_rate": 1.8729207768763514e-06,
      "loss": 1.2275,
      "step": 921
    },
    {
      "epoch": 2.8653696498054475,
      "grad_norm": 2.0625,
      "learning_rate": 1.8633095781875098e-06,
      "loss": 1.1889,
      "step": 922
    },
    {
      "epoch": 2.8684824902723736,
      "grad_norm": 2.125,
      "learning_rate": 1.8537174530835806e-06,
      "loss": 1.1889,
      "step": 923
    },
    {
      "epoch": 2.8715953307392996,
      "grad_norm": 2.09375,
      "learning_rate": 1.8441444598923508e-06,
      "loss": 1.1779,
      "step": 924
    },
    {
      "epoch": 2.8715953307392996,
      "eval_loss": 3.279463529586792,
      "eval_runtime": 42.9331,
      "eval_samples_per_second": 114.318,
      "eval_steps_per_second": 11.436,
      "step": 924
    },
    {
      "epoch": 2.8747081712062257,
      "grad_norm": 2.15625,
      "learning_rate": 1.8345906568252724e-06,
      "loss": 1.1725,
      "step": 925
    },
    {
      "epoch": 2.8778210116731517,
      "grad_norm": 2.078125,
      "learning_rate": 1.8250561019771012e-06,
      "loss": 1.1841,
      "step": 926
    },
    {
      "epoch": 2.880933852140078,
      "grad_norm": 2.109375,
      "learning_rate": 1.8155408533255553e-06,
      "loss": 1.157,
      "step": 927
    },
    {
      "epoch": 2.884046692607004,
      "grad_norm": 2.09375,
      "learning_rate": 1.8060449687309477e-06,
      "loss": 1.1782,
      "step": 928
    },
    {
      "epoch": 2.88715953307393,
      "grad_norm": 2.140625,
      "learning_rate": 1.7965685059358489e-06,
      "loss": 1.1945,
      "step": 929
    },
    {
      "epoch": 2.890272373540856,
      "grad_norm": 2.046875,
      "learning_rate": 1.7871115225647294e-06,
      "loss": 1.1762,
      "step": 930
    },
    {
      "epoch": 2.893385214007782,
      "grad_norm": 2.125,
      "learning_rate": 1.7776740761236034e-06,
      "loss": 1.1818,
      "step": 931
    },
    {
      "epoch": 2.896498054474708,
      "grad_norm": 2.125,
      "learning_rate": 1.7682562239996914e-06,
      "loss": 1.1458,
      "step": 932
    },
    {
      "epoch": 2.899610894941634,
      "grad_norm": 2.0625,
      "learning_rate": 1.7588580234610592e-06,
      "loss": 1.1681,
      "step": 933
    },
    {
      "epoch": 2.90272373540856,
      "grad_norm": 1.9921875,
      "learning_rate": 1.7494795316562791e-06,
      "loss": 1.1213,
      "step": 934
    },
    {
      "epoch": 2.9058365758754863,
      "grad_norm": 2.078125,
      "learning_rate": 1.7401208056140784e-06,
      "loss": 1.2343,
      "step": 935
    },
    {
      "epoch": 2.9089494163424123,
      "grad_norm": 2.015625,
      "learning_rate": 1.7307819022429928e-06,
      "loss": 1.1587,
      "step": 936
    },
    {
      "epoch": 2.9120622568093384,
      "grad_norm": 2.078125,
      "learning_rate": 1.7214628783310162e-06,
      "loss": 1.1476,
      "step": 937
    },
    {
      "epoch": 2.9151750972762644,
      "grad_norm": 2.125,
      "learning_rate": 1.7121637905452659e-06,
      "loss": 1.2009,
      "step": 938
    },
    {
      "epoch": 2.9182879377431905,
      "grad_norm": 2.171875,
      "learning_rate": 1.7028846954316252e-06,
      "loss": 1.1999,
      "step": 939
    },
    {
      "epoch": 2.9214007782101166,
      "grad_norm": 2.09375,
      "learning_rate": 1.693625649414411e-06,
      "loss": 1.1731,
      "step": 940
    },
    {
      "epoch": 2.924513618677043,
      "grad_norm": 2.015625,
      "learning_rate": 1.6843867087960252e-06,
      "loss": 1.1684,
      "step": 941
    },
    {
      "epoch": 2.927626459143969,
      "grad_norm": 2.046875,
      "learning_rate": 1.675167929756608e-06,
      "loss": 1.1816,
      "step": 942
    },
    {
      "epoch": 2.930739299610895,
      "grad_norm": 2.15625,
      "learning_rate": 1.6659693683537077e-06,
      "loss": 1.193,
      "step": 943
    },
    {
      "epoch": 2.9338521400778212,
      "grad_norm": 2.09375,
      "learning_rate": 1.6567910805219273e-06,
      "loss": 1.1551,
      "step": 944
    },
    {
      "epoch": 2.9369649805447473,
      "grad_norm": 2.03125,
      "learning_rate": 1.6476331220725943e-06,
      "loss": 1.1602,
      "step": 945
    },
    {
      "epoch": 2.9369649805447473,
      "eval_loss": 3.277841091156006,
      "eval_runtime": 42.9151,
      "eval_samples_per_second": 114.365,
      "eval_steps_per_second": 11.441,
      "step": 945
    },
    {
      "epoch": 2.9400778210116734,
      "grad_norm": 2.0625,
      "learning_rate": 1.6384955486934157e-06,
      "loss": 1.1917,
      "step": 946
    },
    {
      "epoch": 2.9431906614785994,
      "grad_norm": 1.984375,
      "learning_rate": 1.6293784159481434e-06,
      "loss": 1.1895,
      "step": 947
    },
    {
      "epoch": 2.9463035019455255,
      "grad_norm": 2.0625,
      "learning_rate": 1.6202817792762283e-06,
      "loss": 1.1747,
      "step": 948
    },
    {
      "epoch": 2.9494163424124515,
      "grad_norm": 2.046875,
      "learning_rate": 1.6112056939924957e-06,
      "loss": 1.1853,
      "step": 949
    },
    {
      "epoch": 2.9525291828793776,
      "grad_norm": 2.046875,
      "learning_rate": 1.6021502152867952e-06,
      "loss": 1.2185,
      "step": 950
    },
    {
      "epoch": 2.9556420233463037,
      "grad_norm": 1.953125,
      "learning_rate": 1.5931153982236768e-06,
      "loss": 1.1483,
      "step": 951
    },
    {
      "epoch": 2.9587548638132297,
      "grad_norm": 2.03125,
      "learning_rate": 1.5841012977420516e-06,
      "loss": 1.1556,
      "step": 952
    },
    {
      "epoch": 2.9618677042801558,
      "grad_norm": 2.046875,
      "learning_rate": 1.5751079686548527e-06,
      "loss": 1.1422,
      "step": 953
    },
    {
      "epoch": 2.964980544747082,
      "grad_norm": 2.21875,
      "learning_rate": 1.5661354656487115e-06,
      "loss": 1.1737,
      "step": 954
    },
    {
      "epoch": 2.968093385214008,
      "grad_norm": 2.140625,
      "learning_rate": 1.557183843283614e-06,
      "loss": 1.1485,
      "step": 955
    },
    {
      "epoch": 2.971206225680934,
      "grad_norm": 2.0625,
      "learning_rate": 1.548253155992585e-06,
      "loss": 1.1501,
      "step": 956
    },
    {
      "epoch": 2.97431906614786,
      "grad_norm": 2.0625,
      "learning_rate": 1.539343458081337e-06,
      "loss": 1.1971,
      "step": 957
    },
    {
      "epoch": 2.977431906614786,
      "grad_norm": 2.109375,
      "learning_rate": 1.5304548037279582e-06,
      "loss": 1.182,
      "step": 958
    },
    {
      "epoch": 2.980544747081712,
      "grad_norm": 2.0625,
      "learning_rate": 1.5215872469825682e-06,
      "loss": 1.1725,
      "step": 959
    },
    {
      "epoch": 2.983657587548638,
      "grad_norm": 2.09375,
      "learning_rate": 1.5127408417670031e-06,
      "loss": 1.1708,
      "step": 960
    },
    {
      "epoch": 2.9867704280155642,
      "grad_norm": 2.03125,
      "learning_rate": 1.5039156418744743e-06,
      "loss": 1.1779,
      "step": 961
    },
    {
      "epoch": 2.9898832684824903,
      "grad_norm": 2.0625,
      "learning_rate": 1.4951117009692528e-06,
      "loss": 1.186,
      "step": 962
    },
    {
      "epoch": 2.9929961089494164,
      "grad_norm": 2.09375,
      "learning_rate": 1.4863290725863383e-06,
      "loss": 1.1066,
      "step": 963
    },
    {
      "epoch": 2.9961089494163424,
      "grad_norm": 1.984375,
      "learning_rate": 1.477567810131127e-06,
      "loss": 1.242,
      "step": 964
    },
    {
      "epoch": 2.9992217898832685,
      "grad_norm": 2.03125,
      "learning_rate": 1.468827966879101e-06,
      "loss": 1.1625,
      "step": 965
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.09375,
      "learning_rate": 1.4601095959754897e-06,
      "loss": 1.1668,
      "step": 966
    },
    {
      "epoch": 3.0,
      "eval_loss": 3.2712929248809814,
      "eval_runtime": 42.9361,
      "eval_samples_per_second": 114.309,
      "eval_steps_per_second": 11.436,
      "step": 966
    },
    {
      "epoch": 3.003112840466926,
      "grad_norm": 2.109375,
      "learning_rate": 1.4514127504349606e-06,
      "loss": 1.1526,
      "step": 967
    },
    {
      "epoch": 3.006225680933852,
      "grad_norm": 2.046875,
      "learning_rate": 1.4427374831412833e-06,
      "loss": 1.1423,
      "step": 968
    },
    {
      "epoch": 3.009338521400778,
      "grad_norm": 2.0625,
      "learning_rate": 1.4340838468470198e-06,
      "loss": 1.117,
      "step": 969
    },
    {
      "epoch": 3.0124513618677042,
      "grad_norm": 2.078125,
      "learning_rate": 1.4254518941731915e-06,
      "loss": 1.1568,
      "step": 970
    },
    {
      "epoch": 3.0155642023346303,
      "grad_norm": 2.15625,
      "learning_rate": 1.4168416776089722e-06,
      "loss": 1.192,
      "step": 971
    },
    {
      "epoch": 3.0186770428015564,
      "grad_norm": 2.03125,
      "learning_rate": 1.4082532495113627e-06,
      "loss": 1.1847,
      "step": 972
    },
    {
      "epoch": 3.0217898832684824,
      "grad_norm": 2.078125,
      "learning_rate": 1.3996866621048671e-06,
      "loss": 1.1979,
      "step": 973
    },
    {
      "epoch": 3.0249027237354085,
      "grad_norm": 2.09375,
      "learning_rate": 1.3911419674811871e-06,
      "loss": 1.1658,
      "step": 974
    },
    {
      "epoch": 3.0280155642023345,
      "grad_norm": 2.09375,
      "learning_rate": 1.3826192175988928e-06,
      "loss": 1.1972,
      "step": 975
    },
    {
      "epoch": 3.0311284046692606,
      "grad_norm": 2.0,
      "learning_rate": 1.374118464283119e-06,
      "loss": 1.2039,
      "step": 976
    },
    {
      "epoch": 3.0342412451361866,
      "grad_norm": 2.0625,
      "learning_rate": 1.3656397592252358e-06,
      "loss": 1.1566,
      "step": 977
    },
    {
      "epoch": 3.0373540856031127,
      "grad_norm": 2.09375,
      "learning_rate": 1.3571831539825526e-06,
      "loss": 1.1977,
      "step": 978
    },
    {
      "epoch": 3.0404669260700388,
      "grad_norm": 2.0,
      "learning_rate": 1.3487486999779843e-06,
      "loss": 1.168,
      "step": 979
    },
    {
      "epoch": 3.043579766536965,
      "grad_norm": 2.015625,
      "learning_rate": 1.340336448499756e-06,
      "loss": 1.2273,
      "step": 980
    },
    {
      "epoch": 3.046692607003891,
      "grad_norm": 2.09375,
      "learning_rate": 1.3319464507010776e-06,
      "loss": 1.217,
      "step": 981
    },
    {
      "epoch": 3.049805447470817,
      "grad_norm": 2.109375,
      "learning_rate": 1.3235787575998432e-06,
      "loss": 1.18,
      "step": 982
    },
    {
      "epoch": 3.052918287937743,
      "grad_norm": 2.09375,
      "learning_rate": 1.3152334200783167e-06,
      "loss": 1.22,
      "step": 983
    },
    {
      "epoch": 3.056031128404669,
      "grad_norm": 2.140625,
      "learning_rate": 1.3069104888828176e-06,
      "loss": 1.2424,
      "step": 984
    },
    {
      "epoch": 3.059143968871595,
      "grad_norm": 2.046875,
      "learning_rate": 1.298610014623423e-06,
      "loss": 1.1265,
      "step": 985
    },
    {
      "epoch": 3.062256809338521,
      "grad_norm": 2.03125,
      "learning_rate": 1.2903320477736486e-06,
      "loss": 1.1362,
      "step": 986
    },
    {
      "epoch": 3.0653696498054477,
      "grad_norm": 2.03125,
      "learning_rate": 1.28207663867015e-06,
      "loss": 1.1328,
      "step": 987
    },
    {
      "epoch": 3.0653696498054477,
      "eval_loss": 3.287999391555786,
      "eval_runtime": 43.0165,
      "eval_samples_per_second": 114.096,
      "eval_steps_per_second": 11.414,
      "step": 987
    },
    {
      "epoch": 3.0684824902723737,
      "grad_norm": 2.125,
      "learning_rate": 1.2738438375124129e-06,
      "loss": 1.1969,
      "step": 988
    },
    {
      "epoch": 3.0715953307393,
      "grad_norm": 1.9921875,
      "learning_rate": 1.2656336943624514e-06,
      "loss": 1.1719,
      "step": 989
    },
    {
      "epoch": 3.074708171206226,
      "grad_norm": 2.03125,
      "learning_rate": 1.257446259144494e-06,
      "loss": 1.1718,
      "step": 990
    },
    {
      "epoch": 3.077821011673152,
      "grad_norm": 2.015625,
      "learning_rate": 1.2492815816446952e-06,
      "loss": 1.1898,
      "step": 991
    },
    {
      "epoch": 3.080933852140078,
      "grad_norm": 2.15625,
      "learning_rate": 1.2411397115108164e-06,
      "loss": 1.2078,
      "step": 992
    },
    {
      "epoch": 3.084046692607004,
      "grad_norm": 1.984375,
      "learning_rate": 1.2330206982519377e-06,
      "loss": 1.1248,
      "step": 993
    },
    {
      "epoch": 3.08715953307393,
      "grad_norm": 2.109375,
      "learning_rate": 1.2249245912381502e-06,
      "loss": 1.1952,
      "step": 994
    },
    {
      "epoch": 3.090272373540856,
      "grad_norm": 2.015625,
      "learning_rate": 1.2168514397002528e-06,
      "loss": 1.1579,
      "step": 995
    },
    {
      "epoch": 3.093385214007782,
      "grad_norm": 2.15625,
      "learning_rate": 1.2088012927294635e-06,
      "loss": 1.1778,
      "step": 996
    },
    {
      "epoch": 3.0964980544747083,
      "grad_norm": 2.046875,
      "learning_rate": 1.2007741992771065e-06,
      "loss": 1.1836,
      "step": 997
    },
    {
      "epoch": 3.0996108949416343,
      "grad_norm": 2.15625,
      "learning_rate": 1.1927702081543279e-06,
      "loss": 1.1404,
      "step": 998
    },
    {
      "epoch": 3.1027237354085604,
      "grad_norm": 2.046875,
      "learning_rate": 1.1847893680317913e-06,
      "loss": 1.1981,
      "step": 999
    },
    {
      "epoch": 3.1058365758754864,
      "grad_norm": 2.0,
      "learning_rate": 1.1768317274393847e-06,
      "loss": 1.1896,
      "step": 1000
    },
    {
      "epoch": 3.1089494163424125,
      "grad_norm": 2.0625,
      "learning_rate": 1.168897334765921e-06,
      "loss": 1.1594,
      "step": 1001
    },
    {
      "epoch": 3.1120622568093386,
      "grad_norm": 1.9375,
      "learning_rate": 1.1609862382588527e-06,
      "loss": 1.1721,
      "step": 1002
    },
    {
      "epoch": 3.1151750972762646,
      "grad_norm": 2.0625,
      "learning_rate": 1.1530984860239664e-06,
      "loss": 1.2105,
      "step": 1003
    },
    {
      "epoch": 3.1182879377431907,
      "grad_norm": 1.953125,
      "learning_rate": 1.145234126025102e-06,
      "loss": 1.1141,
      "step": 1004
    },
    {
      "epoch": 3.1214007782101167,
      "grad_norm": 2.09375,
      "learning_rate": 1.1373932060838554e-06,
      "loss": 1.2397,
      "step": 1005
    },
    {
      "epoch": 3.124513618677043,
      "grad_norm": 2.109375,
      "learning_rate": 1.129575773879285e-06,
      "loss": 1.1664,
      "step": 1006
    },
    {
      "epoch": 3.127626459143969,
      "grad_norm": 2.0625,
      "learning_rate": 1.1217818769476297e-06,
      "loss": 1.1873,
      "step": 1007
    },
    {
      "epoch": 3.130739299610895,
      "grad_norm": 2.046875,
      "learning_rate": 1.1140115626820085e-06,
      "loss": 1.1431,
      "step": 1008
    },
    {
      "epoch": 3.130739299610895,
      "eval_loss": 3.2956442832946777,
      "eval_runtime": 42.9591,
      "eval_samples_per_second": 114.248,
      "eval_steps_per_second": 11.429,
      "step": 1008
    },
    {
      "epoch": 3.133852140077821,
      "grad_norm": 1.9765625,
      "learning_rate": 1.1062648783321462e-06,
      "loss": 1.1407,
      "step": 1009
    },
    {
      "epoch": 3.136964980544747,
      "grad_norm": 2.03125,
      "learning_rate": 1.0985418710040741e-06,
      "loss": 1.2096,
      "step": 1010
    },
    {
      "epoch": 3.140077821011673,
      "grad_norm": 2.03125,
      "learning_rate": 1.0908425876598512e-06,
      "loss": 1.1254,
      "step": 1011
    },
    {
      "epoch": 3.143190661478599,
      "grad_norm": 2.046875,
      "learning_rate": 1.0831670751172723e-06,
      "loss": 1.1761,
      "step": 1012
    },
    {
      "epoch": 3.146303501945525,
      "grad_norm": 2.1875,
      "learning_rate": 1.075515380049591e-06,
      "loss": 1.1313,
      "step": 1013
    },
    {
      "epoch": 3.1494163424124513,
      "grad_norm": 2.0625,
      "learning_rate": 1.0678875489852265e-06,
      "loss": 1.1635,
      "step": 1014
    },
    {
      "epoch": 3.1525291828793773,
      "grad_norm": 2.09375,
      "learning_rate": 1.0602836283074903e-06,
      "loss": 1.1565,
      "step": 1015
    },
    {
      "epoch": 3.1556420233463034,
      "grad_norm": 2.078125,
      "learning_rate": 1.0527036642542982e-06,
      "loss": 1.131,
      "step": 1016
    },
    {
      "epoch": 3.1587548638132295,
      "grad_norm": 2.03125,
      "learning_rate": 1.0451477029178891e-06,
      "loss": 1.1416,
      "step": 1017
    },
    {
      "epoch": 3.1618677042801555,
      "grad_norm": 1.9921875,
      "learning_rate": 1.037615790244549e-06,
      "loss": 1.175,
      "step": 1018
    },
    {
      "epoch": 3.1649805447470816,
      "grad_norm": 2.03125,
      "learning_rate": 1.030107972034325e-06,
      "loss": 1.1667,
      "step": 1019
    },
    {
      "epoch": 3.1680933852140076,
      "grad_norm": 2.140625,
      "learning_rate": 1.0226242939407532e-06,
      "loss": 1.193,
      "step": 1020
    },
    {
      "epoch": 3.1712062256809337,
      "grad_norm": 2.03125,
      "learning_rate": 1.0151648014705795e-06,
      "loss": 1.1768,
      "step": 1021
    },
    {
      "epoch": 3.1743190661478597,
      "grad_norm": 2.078125,
      "learning_rate": 1.0077295399834813e-06,
      "loss": 1.209,
      "step": 1022
    },
    {
      "epoch": 3.177431906614786,
      "grad_norm": 2.09375,
      "learning_rate": 1.0003185546917883e-06,
      "loss": 1.2198,
      "step": 1023
    },
    {
      "epoch": 3.180544747081712,
      "grad_norm": 2.109375,
      "learning_rate": 9.929318906602176e-07,
      "loss": 1.1712,
      "step": 1024
    },
    {
      "epoch": 3.1836575875486384,
      "grad_norm": 2.078125,
      "learning_rate": 9.85569592805588e-07,
      "loss": 1.1721,
      "step": 1025
    },
    {
      "epoch": 3.1867704280155644,
      "grad_norm": 1.96875,
      "learning_rate": 9.782317058965564e-07,
      "loss": 1.1319,
      "step": 1026
    },
    {
      "epoch": 3.1898832684824905,
      "grad_norm": 2.234375,
      "learning_rate": 9.70918274553342e-07,
      "loss": 1.2211,
      "step": 1027
    },
    {
      "epoch": 3.1929961089494165,
      "grad_norm": 2.03125,
      "learning_rate": 9.636293432474497e-07,
      "loss": 1.1954,
      "step": 1028
    },
    {
      "epoch": 3.1961089494163426,
      "grad_norm": 2.015625,
      "learning_rate": 9.56364956301411e-07,
      "loss": 1.1485,
      "step": 1029
    },
    {
      "epoch": 3.1961089494163426,
      "eval_loss": 3.274742841720581,
      "eval_runtime": 240.9744,
      "eval_samples_per_second": 20.367,
      "eval_steps_per_second": 2.038,
      "step": 1029
    },
    {
      "epoch": 3.1992217898832687,
      "grad_norm": 2.109375,
      "learning_rate": 9.491251578885024e-07,
      "loss": 1.1978,
      "step": 1030
    },
    {
      "epoch": 3.2023346303501947,
      "grad_norm": 1.9609375,
      "learning_rate": 9.419099920324859e-07,
      "loss": 1.1824,
      "step": 1031
    },
    {
      "epoch": 3.205447470817121,
      "grad_norm": 2.046875,
      "learning_rate": 9.347195026073369e-07,
      "loss": 1.1685,
      "step": 1032
    },
    {
      "epoch": 3.208560311284047,
      "grad_norm": 1.984375,
      "learning_rate": 9.275537333369805e-07,
      "loss": 1.1793,
      "step": 1033
    },
    {
      "epoch": 3.211673151750973,
      "grad_norm": 2.15625,
      "learning_rate": 9.204127277950181e-07,
      "loss": 1.1751,
      "step": 1034
    },
    {
      "epoch": 3.214785992217899,
      "grad_norm": 2.078125,
      "learning_rate": 9.132965294044738e-07,
      "loss": 1.2425,
      "step": 1035
    },
    {
      "epoch": 3.217898832684825,
      "grad_norm": 2.109375,
      "learning_rate": 9.062051814375195e-07,
      "loss": 1.1325,
      "step": 1036
    },
    {
      "epoch": 3.221011673151751,
      "grad_norm": 2.09375,
      "learning_rate": 8.991387270152202e-07,
      "loss": 1.1512,
      "step": 1037
    },
    {
      "epoch": 3.224124513618677,
      "grad_norm": 2.125,
      "learning_rate": 8.920972091072683e-07,
      "loss": 1.1377,
      "step": 1038
    },
    {
      "epoch": 3.227237354085603,
      "grad_norm": 2.078125,
      "learning_rate": 8.850806705317183e-07,
      "loss": 1.1729,
      "step": 1039
    },
    {
      "epoch": 3.2303501945525293,
      "grad_norm": 2.0,
      "learning_rate": 8.780891539547354e-07,
      "loss": 1.1681,
      "step": 1040
    },
    {
      "epoch": 3.2334630350194553,
      "grad_norm": 2.078125,
      "learning_rate": 8.711227018903263e-07,
      "loss": 1.175,
      "step": 1041
    },
    {
      "epoch": 3.2365758754863814,
      "grad_norm": 2.0625,
      "learning_rate": 8.64181356700089e-07,
      "loss": 1.2032,
      "step": 1042
    },
    {
      "epoch": 3.2396887159533074,
      "grad_norm": 2.015625,
      "learning_rate": 8.572651605929506e-07,
      "loss": 1.1943,
      "step": 1043
    },
    {
      "epoch": 3.2428015564202335,
      "grad_norm": 2.0625,
      "learning_rate": 8.503741556249123e-07,
      "loss": 1.1654,
      "step": 1044
    },
    {
      "epoch": 3.2459143968871595,
      "grad_norm": 2.03125,
      "learning_rate": 8.435083836987906e-07,
      "loss": 1.1497,
      "step": 1045
    },
    {
      "epoch": 3.2490272373540856,
      "grad_norm": 1.9765625,
      "learning_rate": 8.366678865639688e-07,
      "loss": 1.1297,
      "step": 1046
    },
    {
      "epoch": 3.2521400778210117,
      "grad_norm": 2.03125,
      "learning_rate": 8.298527058161343e-07,
      "loss": 1.1562,
      "step": 1047
    },
    {
      "epoch": 3.2552529182879377,
      "grad_norm": 2.046875,
      "learning_rate": 8.230628828970349e-07,
      "loss": 1.1892,
      "step": 1048
    },
    {
      "epoch": 3.258365758754864,
      "grad_norm": 2.109375,
      "learning_rate": 8.162984590942219e-07,
      "loss": 1.2183,
      "step": 1049
    },
    {
      "epoch": 3.26147859922179,
      "grad_norm": 2.078125,
      "learning_rate": 8.095594755407971e-07,
      "loss": 1.1921,
      "step": 1050
    },
    {
      "epoch": 3.26147859922179,
      "eval_loss": 3.2739920616149902,
      "eval_runtime": 43.0068,
      "eval_samples_per_second": 114.121,
      "eval_steps_per_second": 11.417,
      "step": 1050
    },
    {
      "epoch": 3.264591439688716,
      "grad_norm": 2.109375,
      "learning_rate": 8.028459732151684e-07,
      "loss": 1.2321,
      "step": 1051
    },
    {
      "epoch": 3.267704280155642,
      "grad_norm": 2.1875,
      "learning_rate": 7.961579929407942e-07,
      "loss": 1.2235,
      "step": 1052
    },
    {
      "epoch": 3.270817120622568,
      "grad_norm": 2.171875,
      "learning_rate": 7.894955753859412e-07,
      "loss": 1.165,
      "step": 1053
    },
    {
      "epoch": 3.273929961089494,
      "grad_norm": 2.203125,
      "learning_rate": 7.828587610634342e-07,
      "loss": 1.2115,
      "step": 1054
    },
    {
      "epoch": 3.27704280155642,
      "grad_norm": 2.171875,
      "learning_rate": 7.762475903304089e-07,
      "loss": 1.1842,
      "step": 1055
    },
    {
      "epoch": 3.280155642023346,
      "grad_norm": 2.078125,
      "learning_rate": 7.696621033880663e-07,
      "loss": 1.1326,
      "step": 1056
    },
    {
      "epoch": 3.2832684824902723,
      "grad_norm": 2.09375,
      "learning_rate": 7.63102340281433e-07,
      "loss": 1.2021,
      "step": 1057
    },
    {
      "epoch": 3.2863813229571983,
      "grad_norm": 2.078125,
      "learning_rate": 7.565683408991093e-07,
      "loss": 1.117,
      "step": 1058
    },
    {
      "epoch": 3.2894941634241244,
      "grad_norm": 2.046875,
      "learning_rate": 7.500601449730349e-07,
      "loss": 1.179,
      "step": 1059
    },
    {
      "epoch": 3.2926070038910504,
      "grad_norm": 2.125,
      "learning_rate": 7.435777920782444e-07,
      "loss": 1.1731,
      "step": 1060
    },
    {
      "epoch": 3.2957198443579765,
      "grad_norm": 2.109375,
      "learning_rate": 7.371213216326228e-07,
      "loss": 1.2202,
      "step": 1061
    },
    {
      "epoch": 3.2988326848249026,
      "grad_norm": 2.0625,
      "learning_rate": 7.306907728966733e-07,
      "loss": 1.1822,
      "step": 1062
    },
    {
      "epoch": 3.3019455252918286,
      "grad_norm": 2.015625,
      "learning_rate": 7.242861849732696e-07,
      "loss": 1.1967,
      "step": 1063
    },
    {
      "epoch": 3.3050583657587547,
      "grad_norm": 2.015625,
      "learning_rate": 7.179075968074267e-07,
      "loss": 1.1782,
      "step": 1064
    },
    {
      "epoch": 3.3081712062256807,
      "grad_norm": 2.15625,
      "learning_rate": 7.115550471860588e-07,
      "loss": 1.2184,
      "step": 1065
    },
    {
      "epoch": 3.311284046692607,
      "grad_norm": 2.078125,
      "learning_rate": 7.052285747377464e-07,
      "loss": 1.1973,
      "step": 1066
    },
    {
      "epoch": 3.314396887159533,
      "grad_norm": 2.109375,
      "learning_rate": 6.989282179324963e-07,
      "loss": 1.1788,
      "step": 1067
    },
    {
      "epoch": 3.317509727626459,
      "grad_norm": 2.140625,
      "learning_rate": 6.92654015081517e-07,
      "loss": 1.2031,
      "step": 1068
    },
    {
      "epoch": 3.3206225680933854,
      "grad_norm": 2.046875,
      "learning_rate": 6.86406004336973e-07,
      "loss": 1.1535,
      "step": 1069
    },
    {
      "epoch": 3.3237354085603115,
      "grad_norm": 2.109375,
      "learning_rate": 6.801842236917661e-07,
      "loss": 1.2004,
      "step": 1070
    },
    {
      "epoch": 3.3268482490272375,
      "grad_norm": 2.0625,
      "learning_rate": 6.739887109792958e-07,
      "loss": 1.1771,
      "step": 1071
    },
    {
      "epoch": 3.3268482490272375,
      "eval_loss": 3.2850749492645264,
      "eval_runtime": 43.054,
      "eval_samples_per_second": 113.996,
      "eval_steps_per_second": 11.404,
      "step": 1071
    },
    {
      "epoch": 3.3299610894941636,
      "grad_norm": 2.109375,
      "learning_rate": 6.678195038732294e-07,
      "loss": 1.2137,
      "step": 1072
    },
    {
      "epoch": 3.3330739299610896,
      "grad_norm": 2.046875,
      "learning_rate": 6.616766398872798e-07,
      "loss": 1.1614,
      "step": 1073
    },
    {
      "epoch": 3.3361867704280157,
      "grad_norm": 2.140625,
      "learning_rate": 6.555601563749675e-07,
      "loss": 1.1616,
      "step": 1074
    },
    {
      "epoch": 3.3392996108949418,
      "grad_norm": 2.109375,
      "learning_rate": 6.494700905294027e-07,
      "loss": 1.1915,
      "step": 1075
    },
    {
      "epoch": 3.342412451361868,
      "grad_norm": 2.109375,
      "learning_rate": 6.43406479383053e-07,
      "loss": 1.1942,
      "step": 1076
    },
    {
      "epoch": 3.345525291828794,
      "grad_norm": 2.0625,
      "learning_rate": 6.37369359807522e-07,
      "loss": 1.1904,
      "step": 1077
    },
    {
      "epoch": 3.34863813229572,
      "grad_norm": 2.0,
      "learning_rate": 6.313587685133199e-07,
      "loss": 1.1069,
      "step": 1078
    },
    {
      "epoch": 3.351750972762646,
      "grad_norm": 2.0625,
      "learning_rate": 6.253747420496481e-07,
      "loss": 1.2487,
      "step": 1079
    },
    {
      "epoch": 3.354863813229572,
      "grad_norm": 2.0,
      "learning_rate": 6.194173168041689e-07,
      "loss": 1.1303,
      "step": 1080
    },
    {
      "epoch": 3.357976653696498,
      "grad_norm": 2.15625,
      "learning_rate": 6.134865290027903e-07,
      "loss": 1.1703,
      "step": 1081
    },
    {
      "epoch": 3.361089494163424,
      "grad_norm": 2.171875,
      "learning_rate": 6.075824147094433e-07,
      "loss": 1.1606,
      "step": 1082
    },
    {
      "epoch": 3.3642023346303502,
      "grad_norm": 2.140625,
      "learning_rate": 6.017050098258604e-07,
      "loss": 1.1497,
      "step": 1083
    },
    {
      "epoch": 3.3673151750972763,
      "grad_norm": 2.09375,
      "learning_rate": 5.958543500913628e-07,
      "loss": 1.1738,
      "step": 1084
    },
    {
      "epoch": 3.3704280155642024,
      "grad_norm": 1.9609375,
      "learning_rate": 5.90030471082636e-07,
      "loss": 1.1763,
      "step": 1085
    },
    {
      "epoch": 3.3735408560311284,
      "grad_norm": 2.0625,
      "learning_rate": 5.842334082135209e-07,
      "loss": 1.1866,
      "step": 1086
    },
    {
      "epoch": 3.3766536964980545,
      "grad_norm": 2.171875,
      "learning_rate": 5.784631967347926e-07,
      "loss": 1.1936,
      "step": 1087
    },
    {
      "epoch": 3.3797665369649805,
      "grad_norm": 2.078125,
      "learning_rate": 5.727198717339511e-07,
      "loss": 1.1531,
      "step": 1088
    },
    {
      "epoch": 3.3828793774319066,
      "grad_norm": 2.0625,
      "learning_rate": 5.670034681349995e-07,
      "loss": 1.1828,
      "step": 1089
    },
    {
      "epoch": 3.3859922178988326,
      "grad_norm": 2.203125,
      "learning_rate": 5.613140206982443e-07,
      "loss": 1.1924,
      "step": 1090
    },
    {
      "epoch": 3.3891050583657587,
      "grad_norm": 2.140625,
      "learning_rate": 5.556515640200704e-07,
      "loss": 1.1463,
      "step": 1091
    },
    {
      "epoch": 3.3922178988326848,
      "grad_norm": 2.078125,
      "learning_rate": 5.500161325327419e-07,
      "loss": 1.1824,
      "step": 1092
    },
    {
      "epoch": 3.3922178988326848,
      "eval_loss": 3.2866530418395996,
      "eval_runtime": 42.9843,
      "eval_samples_per_second": 114.181,
      "eval_steps_per_second": 11.423,
      "step": 1092
    },
    {
      "epoch": 3.395330739299611,
      "grad_norm": 2.078125,
      "learning_rate": 5.444077605041875e-07,
      "loss": 1.2346,
      "step": 1093
    },
    {
      "epoch": 3.398443579766537,
      "grad_norm": 2.109375,
      "learning_rate": 5.388264820377898e-07,
      "loss": 1.1618,
      "step": 1094
    },
    {
      "epoch": 3.401556420233463,
      "grad_norm": 2.046875,
      "learning_rate": 5.332723310721855e-07,
      "loss": 1.2463,
      "step": 1095
    },
    {
      "epoch": 3.404669260700389,
      "grad_norm": 2.171875,
      "learning_rate": 5.277453413810491e-07,
      "loss": 1.1943,
      "step": 1096
    },
    {
      "epoch": 3.407782101167315,
      "grad_norm": 2.125,
      "learning_rate": 5.222455465728964e-07,
      "loss": 1.1848,
      "step": 1097
    },
    {
      "epoch": 3.410894941634241,
      "grad_norm": 2.078125,
      "learning_rate": 5.167729800908772e-07,
      "loss": 1.168,
      "step": 1098
    },
    {
      "epoch": 3.414007782101167,
      "grad_norm": 2.171875,
      "learning_rate": 5.113276752125701e-07,
      "loss": 1.1892,
      "step": 1099
    },
    {
      "epoch": 3.4171206225680932,
      "grad_norm": 2.03125,
      "learning_rate": 5.059096650497791e-07,
      "loss": 1.1267,
      "step": 1100
    },
    {
      "epoch": 3.4202334630350193,
      "grad_norm": 2.15625,
      "learning_rate": 5.00518982548338e-07,
      "loss": 1.1844,
      "step": 1101
    },
    {
      "epoch": 3.4233463035019454,
      "grad_norm": 2.0625,
      "learning_rate": 4.951556604879049e-07,
      "loss": 1.1633,
      "step": 1102
    },
    {
      "epoch": 3.4264591439688714,
      "grad_norm": 2.0625,
      "learning_rate": 4.898197314817638e-07,
      "loss": 1.1096,
      "step": 1103
    },
    {
      "epoch": 3.4295719844357975,
      "grad_norm": 2.09375,
      "learning_rate": 4.845112279766306e-07,
      "loss": 1.1706,
      "step": 1104
    },
    {
      "epoch": 3.432684824902724,
      "grad_norm": 2.203125,
      "learning_rate": 4.79230182252447e-07,
      "loss": 1.1753,
      "step": 1105
    },
    {
      "epoch": 3.43579766536965,
      "grad_norm": 2.0625,
      "learning_rate": 4.73976626422194e-07,
      "loss": 1.1905,
      "step": 1106
    },
    {
      "epoch": 3.438910505836576,
      "grad_norm": 2.15625,
      "learning_rate": 4.687505924316882e-07,
      "loss": 1.1849,
      "step": 1107
    },
    {
      "epoch": 3.442023346303502,
      "grad_norm": 2.109375,
      "learning_rate": 4.63552112059395e-07,
      "loss": 1.1271,
      "step": 1108
    },
    {
      "epoch": 3.445136186770428,
      "grad_norm": 1.9921875,
      "learning_rate": 4.5838121691622995e-07,
      "loss": 1.1679,
      "step": 1109
    },
    {
      "epoch": 3.4482490272373543,
      "grad_norm": 2.140625,
      "learning_rate": 4.532379384453689e-07,
      "loss": 1.2012,
      "step": 1110
    },
    {
      "epoch": 3.4513618677042803,
      "grad_norm": 2.046875,
      "learning_rate": 4.4812230792205537e-07,
      "loss": 1.1629,
      "step": 1111
    },
    {
      "epoch": 3.4544747081712064,
      "grad_norm": 2.09375,
      "learning_rate": 4.4303435645341233e-07,
      "loss": 1.214,
      "step": 1112
    },
    {
      "epoch": 3.4575875486381324,
      "grad_norm": 2.046875,
      "learning_rate": 4.3797411497825015e-07,
      "loss": 1.1816,
      "step": 1113
    },
    {
      "epoch": 3.4575875486381324,
      "eval_loss": 3.2928295135498047,
      "eval_runtime": 42.972,
      "eval_samples_per_second": 114.214,
      "eval_steps_per_second": 11.426,
      "step": 1113
    },
    {
      "epoch": 3.4607003891050585,
      "grad_norm": 2.078125,
      "learning_rate": 4.3294161426688297e-07,
      "loss": 1.2152,
      "step": 1114
    },
    {
      "epoch": 3.4638132295719846,
      "grad_norm": 2.0,
      "learning_rate": 4.279368849209381e-07,
      "loss": 1.1487,
      "step": 1115
    },
    {
      "epoch": 3.4669260700389106,
      "grad_norm": 2.03125,
      "learning_rate": 4.2295995737316854e-07,
      "loss": 1.1562,
      "step": 1116
    },
    {
      "epoch": 3.4700389105058367,
      "grad_norm": 2.046875,
      "learning_rate": 4.180108618872736e-07,
      "loss": 1.1669,
      "step": 1117
    },
    {
      "epoch": 3.4731517509727627,
      "grad_norm": 2.078125,
      "learning_rate": 4.130896285577074e-07,
      "loss": 1.192,
      "step": 1118
    },
    {
      "epoch": 3.476264591439689,
      "grad_norm": 2.0625,
      "learning_rate": 4.0819628730950414e-07,
      "loss": 1.1911,
      "step": 1119
    },
    {
      "epoch": 3.479377431906615,
      "grad_norm": 2.03125,
      "learning_rate": 4.033308678980885e-07,
      "loss": 1.1837,
      "step": 1120
    },
    {
      "epoch": 3.482490272373541,
      "grad_norm": 2.046875,
      "learning_rate": 3.984933999091012e-07,
      "loss": 1.1691,
      "step": 1121
    },
    {
      "epoch": 3.485603112840467,
      "grad_norm": 2.03125,
      "learning_rate": 3.9368391275821203e-07,
      "loss": 1.1544,
      "step": 1122
    },
    {
      "epoch": 3.488715953307393,
      "grad_norm": 2.125,
      "learning_rate": 3.8890243569094874e-07,
      "loss": 1.1699,
      "step": 1123
    },
    {
      "epoch": 3.491828793774319,
      "grad_norm": 2.015625,
      "learning_rate": 3.8414899778251246e-07,
      "loss": 1.1559,
      "step": 1124
    },
    {
      "epoch": 3.494941634241245,
      "grad_norm": 2.09375,
      "learning_rate": 3.7942362793760567e-07,
      "loss": 1.1697,
      "step": 1125
    },
    {
      "epoch": 3.498054474708171,
      "grad_norm": 2.0625,
      "learning_rate": 3.747263548902552e-07,
      "loss": 1.1151,
      "step": 1126
    },
    {
      "epoch": 3.5011673151750973,
      "grad_norm": 2.125,
      "learning_rate": 3.700572072036335e-07,
      "loss": 1.172,
      "step": 1127
    },
    {
      "epoch": 3.5042801556420233,
      "grad_norm": 1.9609375,
      "learning_rate": 3.6541621326989183e-07,
      "loss": 1.1391,
      "step": 1128
    },
    {
      "epoch": 3.5073929961089494,
      "grad_norm": 2.125,
      "learning_rate": 3.6080340130998203e-07,
      "loss": 1.1785,
      "step": 1129
    },
    {
      "epoch": 3.5105058365758754,
      "grad_norm": 2.140625,
      "learning_rate": 3.5621879937348836e-07,
      "loss": 1.1435,
      "step": 1130
    },
    {
      "epoch": 3.5136186770428015,
      "grad_norm": 2.15625,
      "learning_rate": 3.516624353384551e-07,
      "loss": 1.1478,
      "step": 1131
    },
    {
      "epoch": 3.5167315175097276,
      "grad_norm": 2.0625,
      "learning_rate": 3.471343369112185e-07,
      "loss": 1.1333,
      "step": 1132
    },
    {
      "epoch": 3.5198443579766536,
      "grad_norm": 2.03125,
      "learning_rate": 3.426345316262347e-07,
      "loss": 1.1805,
      "step": 1133
    },
    {
      "epoch": 3.5229571984435797,
      "grad_norm": 2.015625,
      "learning_rate": 3.381630468459163e-07,
      "loss": 1.1706,
      "step": 1134
    },
    {
      "epoch": 3.5229571984435797,
      "eval_loss": 3.2802956104278564,
      "eval_runtime": 42.9404,
      "eval_samples_per_second": 114.298,
      "eval_steps_per_second": 11.434,
      "step": 1134
    },
    {
      "epoch": 3.5260700389105057,
      "grad_norm": 2.0625,
      "learning_rate": 3.337199097604654e-07,
      "loss": 1.137,
      "step": 1135
    },
    {
      "epoch": 3.529182879377432,
      "grad_norm": 2.109375,
      "learning_rate": 3.2930514738770736e-07,
      "loss": 1.1482,
      "step": 1136
    },
    {
      "epoch": 3.532295719844358,
      "grad_norm": 2.09375,
      "learning_rate": 3.2491878657292643e-07,
      "loss": 1.1435,
      "step": 1137
    },
    {
      "epoch": 3.535408560311284,
      "grad_norm": 2.03125,
      "learning_rate": 3.205608539887006e-07,
      "loss": 1.1791,
      "step": 1138
    },
    {
      "epoch": 3.53852140077821,
      "grad_norm": 2.0,
      "learning_rate": 3.162313761347452e-07,
      "loss": 1.1735,
      "step": 1139
    },
    {
      "epoch": 3.541634241245136,
      "grad_norm": 2.0,
      "learning_rate": 3.119303793377448e-07,
      "loss": 1.1814,
      "step": 1140
    },
    {
      "epoch": 3.544747081712062,
      "grad_norm": 2.015625,
      "learning_rate": 3.076578897511978e-07,
      "loss": 1.1861,
      "step": 1141
    },
    {
      "epoch": 3.547859922178988,
      "grad_norm": 2.09375,
      "learning_rate": 3.034139333552566e-07,
      "loss": 1.1609,
      "step": 1142
    },
    {
      "epoch": 3.550972762645914,
      "grad_norm": 2.078125,
      "learning_rate": 2.9919853595656856e-07,
      "loss": 1.1861,
      "step": 1143
    },
    {
      "epoch": 3.5540856031128403,
      "grad_norm": 2.0,
      "learning_rate": 2.9501172318811834e-07,
      "loss": 1.1108,
      "step": 1144
    },
    {
      "epoch": 3.5571984435797663,
      "grad_norm": 2.125,
      "learning_rate": 2.90853520509074e-07,
      "loss": 1.2099,
      "step": 1145
    },
    {
      "epoch": 3.5603112840466924,
      "grad_norm": 2.109375,
      "learning_rate": 2.867239532046312e-07,
      "loss": 1.232,
      "step": 1146
    },
    {
      "epoch": 3.5634241245136185,
      "grad_norm": 2.125,
      "learning_rate": 2.826230463858609e-07,
      "loss": 1.172,
      "step": 1147
    },
    {
      "epoch": 3.5665369649805445,
      "grad_norm": 2.109375,
      "learning_rate": 2.7855082498955465e-07,
      "loss": 1.1497,
      "step": 1148
    },
    {
      "epoch": 3.5696498054474706,
      "grad_norm": 2.015625,
      "learning_rate": 2.7450731377807185e-07,
      "loss": 1.1616,
      "step": 1149
    },
    {
      "epoch": 3.5727626459143966,
      "grad_norm": 2.0625,
      "learning_rate": 2.704925373391948e-07,
      "loss": 1.1562,
      "step": 1150
    },
    {
      "epoch": 3.5758754863813227,
      "grad_norm": 2.0625,
      "learning_rate": 2.6650652008597067e-07,
      "loss": 1.1515,
      "step": 1151
    },
    {
      "epoch": 3.578988326848249,
      "grad_norm": 1.984375,
      "learning_rate": 2.6254928625657203e-07,
      "loss": 1.1796,
      "step": 1152
    },
    {
      "epoch": 3.5821011673151752,
      "grad_norm": 2.09375,
      "learning_rate": 2.586208599141432e-07,
      "loss": 1.2615,
      "step": 1153
    },
    {
      "epoch": 3.5852140077821013,
      "grad_norm": 2.046875,
      "learning_rate": 2.547212649466568e-07,
      "loss": 1.1225,
      "step": 1154
    },
    {
      "epoch": 3.5883268482490274,
      "grad_norm": 2.15625,
      "learning_rate": 2.50850525066767e-07,
      "loss": 1.2012,
      "step": 1155
    },
    {
      "epoch": 3.5883268482490274,
      "eval_loss": 3.274099826812744,
      "eval_runtime": 42.951,
      "eval_samples_per_second": 114.27,
      "eval_steps_per_second": 11.432,
      "step": 1155
    },
    {
      "epoch": 3.5914396887159534,
      "grad_norm": 2.015625,
      "learning_rate": 2.470086638116653e-07,
      "loss": 1.1373,
      "step": 1156
    },
    {
      "epoch": 3.5945525291828795,
      "grad_norm": 2.078125,
      "learning_rate": 2.431957045429395e-07,
      "loss": 1.2289,
      "step": 1157
    },
    {
      "epoch": 3.5976653696498055,
      "grad_norm": 2.0625,
      "learning_rate": 2.394116704464294e-07,
      "loss": 1.1616,
      "step": 1158
    },
    {
      "epoch": 3.6007782101167316,
      "grad_norm": 2.015625,
      "learning_rate": 2.3565658453208794e-07,
      "loss": 1.1457,
      "step": 1159
    },
    {
      "epoch": 3.6038910505836577,
      "grad_norm": 2.15625,
      "learning_rate": 2.3193046963383702e-07,
      "loss": 1.1859,
      "step": 1160
    },
    {
      "epoch": 3.6070038910505837,
      "grad_norm": 2.078125,
      "learning_rate": 2.2823334840943568e-07,
      "loss": 1.1966,
      "step": 1161
    },
    {
      "epoch": 3.61011673151751,
      "grad_norm": 1.9765625,
      "learning_rate": 2.2456524334033337e-07,
      "loss": 1.1817,
      "step": 1162
    },
    {
      "epoch": 3.613229571984436,
      "grad_norm": 2.0625,
      "learning_rate": 2.2092617673154403e-07,
      "loss": 1.1853,
      "step": 1163
    },
    {
      "epoch": 3.616342412451362,
      "grad_norm": 1.9765625,
      "learning_rate": 2.1731617071149901e-07,
      "loss": 1.1332,
      "step": 1164
    },
    {
      "epoch": 3.619455252918288,
      "grad_norm": 2.109375,
      "learning_rate": 2.137352472319215e-07,
      "loss": 1.2159,
      "step": 1165
    },
    {
      "epoch": 3.622568093385214,
      "grad_norm": 1.9765625,
      "learning_rate": 2.1018342806768787e-07,
      "loss": 1.1265,
      "step": 1166
    },
    {
      "epoch": 3.62568093385214,
      "grad_norm": 2.109375,
      "learning_rate": 2.0666073481669714e-07,
      "loss": 1.1906,
      "step": 1167
    },
    {
      "epoch": 3.628793774319066,
      "grad_norm": 2.21875,
      "learning_rate": 2.0316718889974108e-07,
      "loss": 1.1619,
      "step": 1168
    },
    {
      "epoch": 3.631906614785992,
      "grad_norm": 2.0,
      "learning_rate": 1.9970281156037053e-07,
      "loss": 1.1665,
      "step": 1169
    },
    {
      "epoch": 3.6350194552529183,
      "grad_norm": 2.09375,
      "learning_rate": 1.9626762386476926e-07,
      "loss": 1.154,
      "step": 1170
    },
    {
      "epoch": 3.6381322957198443,
      "grad_norm": 1.9375,
      "learning_rate": 1.9286164670162366e-07,
      "loss": 1.1707,
      "step": 1171
    },
    {
      "epoch": 3.6412451361867704,
      "grad_norm": 2.0625,
      "learning_rate": 1.8948490078199767e-07,
      "loss": 1.1813,
      "step": 1172
    },
    {
      "epoch": 3.6443579766536964,
      "grad_norm": 2.109375,
      "learning_rate": 1.8613740663920466e-07,
      "loss": 1.1341,
      "step": 1173
    },
    {
      "epoch": 3.6474708171206225,
      "grad_norm": 2.125,
      "learning_rate": 1.8281918462868643e-07,
      "loss": 1.2165,
      "step": 1174
    },
    {
      "epoch": 3.6505836575875485,
      "grad_norm": 2.03125,
      "learning_rate": 1.7953025492788324e-07,
      "loss": 1.2061,
      "step": 1175
    },
    {
      "epoch": 3.6536964980544746,
      "grad_norm": 2.078125,
      "learning_rate": 1.762706375361184e-07,
      "loss": 1.1937,
      "step": 1176
    },
    {
      "epoch": 3.6536964980544746,
      "eval_loss": 3.2711799144744873,
      "eval_runtime": 42.9121,
      "eval_samples_per_second": 114.373,
      "eval_steps_per_second": 11.442,
      "step": 1176
    },
    {
      "epoch": 3.6568093385214007,
      "grad_norm": 2.0625,
      "learning_rate": 1.7304035227446946e-07,
      "loss": 1.1859,
      "step": 1177
    },
    {
      "epoch": 3.6599221789883267,
      "grad_norm": 2.125,
      "learning_rate": 1.6983941878565225e-07,
      "loss": 1.1621,
      "step": 1178
    },
    {
      "epoch": 3.663035019455253,
      "grad_norm": 2.015625,
      "learning_rate": 1.666678565339025e-07,
      "loss": 1.1477,
      "step": 1179
    },
    {
      "epoch": 3.666147859922179,
      "grad_norm": 2.078125,
      "learning_rate": 1.6352568480485277e-07,
      "loss": 1.1737,
      "step": 1180
    },
    {
      "epoch": 3.669260700389105,
      "grad_norm": 2.109375,
      "learning_rate": 1.6041292270541963e-07,
      "loss": 1.1712,
      "step": 1181
    },
    {
      "epoch": 3.6723735408560314,
      "grad_norm": 2.03125,
      "learning_rate": 1.5732958916368336e-07,
      "loss": 1.183,
      "step": 1182
    },
    {
      "epoch": 3.6754863813229575,
      "grad_norm": 1.9921875,
      "learning_rate": 1.542757029287778e-07,
      "loss": 1.1297,
      "step": 1183
    },
    {
      "epoch": 3.6785992217898835,
      "grad_norm": 2.171875,
      "learning_rate": 1.5125128257077072e-07,
      "loss": 1.1971,
      "step": 1184
    },
    {
      "epoch": 3.6817120622568096,
      "grad_norm": 2.046875,
      "learning_rate": 1.4825634648055643e-07,
      "loss": 1.1952,
      "step": 1185
    },
    {
      "epoch": 3.6848249027237356,
      "grad_norm": 2.125,
      "learning_rate": 1.4529091286973994e-07,
      "loss": 1.2041,
      "step": 1186
    },
    {
      "epoch": 3.6879377431906617,
      "grad_norm": 2.03125,
      "learning_rate": 1.4235499977052758e-07,
      "loss": 1.0966,
      "step": 1187
    },
    {
      "epoch": 3.6910505836575878,
      "grad_norm": 2.046875,
      "learning_rate": 1.3944862503561862e-07,
      "loss": 1.1306,
      "step": 1188
    },
    {
      "epoch": 3.694163424124514,
      "grad_norm": 2.078125,
      "learning_rate": 1.365718063380922e-07,
      "loss": 1.2135,
      "step": 1189
    },
    {
      "epoch": 3.69727626459144,
      "grad_norm": 2.0,
      "learning_rate": 1.3372456117130837e-07,
      "loss": 1.1751,
      "step": 1190
    },
    {
      "epoch": 3.700389105058366,
      "grad_norm": 2.09375,
      "learning_rate": 1.3090690684879115e-07,
      "loss": 1.1415,
      "step": 1191
    },
    {
      "epoch": 3.703501945525292,
      "grad_norm": 2.140625,
      "learning_rate": 1.2811886050413281e-07,
      "loss": 1.1645,
      "step": 1192
    },
    {
      "epoch": 3.706614785992218,
      "grad_norm": 2.125,
      "learning_rate": 1.253604390908819e-07,
      "loss": 1.2294,
      "step": 1193
    },
    {
      "epoch": 3.709727626459144,
      "grad_norm": 2.15625,
      "learning_rate": 1.226316593824467e-07,
      "loss": 1.1927,
      "step": 1194
    },
    {
      "epoch": 3.71284046692607,
      "grad_norm": 2.046875,
      "learning_rate": 1.1993253797198845e-07,
      "loss": 1.1526,
      "step": 1195
    },
    {
      "epoch": 3.7159533073929962,
      "grad_norm": 2.0625,
      "learning_rate": 1.1726309127232272e-07,
      "loss": 1.1144,
      "step": 1196
    },
    {
      "epoch": 3.7190661478599223,
      "grad_norm": 2.109375,
      "learning_rate": 1.1462333551581939e-07,
      "loss": 1.1837,
      "step": 1197
    },
    {
      "epoch": 3.7190661478599223,
      "eval_loss": 3.276212215423584,
      "eval_runtime": 42.9606,
      "eval_samples_per_second": 114.244,
      "eval_steps_per_second": 11.429,
      "step": 1197
    },
    {
      "epoch": 3.7221789883268483,
      "grad_norm": 2.203125,
      "learning_rate": 1.1201328675430334e-07,
      "loss": 1.1607,
      "step": 1198
    },
    {
      "epoch": 3.7252918287937744,
      "grad_norm": 2.171875,
      "learning_rate": 1.0943296085895728e-07,
      "loss": 1.1668,
      "step": 1199
    },
    {
      "epoch": 3.7284046692607005,
      "grad_norm": 2.109375,
      "learning_rate": 1.0688237352022346e-07,
      "loss": 1.242,
      "step": 1200
    },
    {
      "epoch": 3.7315175097276265,
      "grad_norm": 2.125,
      "learning_rate": 1.0436154024771439e-07,
      "loss": 1.2143,
      "step": 1201
    },
    {
      "epoch": 3.7346303501945526,
      "grad_norm": 2.109375,
      "learning_rate": 1.0187047637010949e-07,
      "loss": 1.1761,
      "step": 1202
    },
    {
      "epoch": 3.7377431906614786,
      "grad_norm": 2.15625,
      "learning_rate": 9.940919703506912e-08,
      "loss": 1.1917,
      "step": 1203
    },
    {
      "epoch": 3.7408560311284047,
      "grad_norm": 2.046875,
      "learning_rate": 9.697771720913796e-08,
      "loss": 1.1782,
      "step": 1204
    },
    {
      "epoch": 3.7439688715953308,
      "grad_norm": 2.21875,
      "learning_rate": 9.457605167765726e-08,
      "loss": 1.2115,
      "step": 1205
    },
    {
      "epoch": 3.747081712062257,
      "grad_norm": 2.109375,
      "learning_rate": 9.22042150446728e-08,
      "loss": 1.2066,
      "step": 1206
    },
    {
      "epoch": 3.750194552529183,
      "grad_norm": 2.125,
      "learning_rate": 8.986222173284876e-08,
      "loss": 1.1731,
      "step": 1207
    },
    {
      "epoch": 3.753307392996109,
      "grad_norm": 2.046875,
      "learning_rate": 8.755008598337389e-08,
      "loss": 1.1688,
      "step": 1208
    },
    {
      "epoch": 3.756420233463035,
      "grad_norm": 2.046875,
      "learning_rate": 8.526782185588389e-08,
      "loss": 1.1309,
      "step": 1209
    },
    {
      "epoch": 3.759533073929961,
      "grad_norm": 2.03125,
      "learning_rate": 8.301544322836752e-08,
      "loss": 1.1541,
      "step": 1210
    },
    {
      "epoch": 3.762645914396887,
      "grad_norm": 2.140625,
      "learning_rate": 8.079296379708668e-08,
      "loss": 1.1914,
      "step": 1211
    },
    {
      "epoch": 3.765758754863813,
      "grad_norm": 2.125,
      "learning_rate": 7.86003970764948e-08,
      "loss": 1.2207,
      "step": 1212
    },
    {
      "epoch": 3.7688715953307392,
      "grad_norm": 2.078125,
      "learning_rate": 7.643775639914752e-08,
      "loss": 1.2023,
      "step": 1213
    },
    {
      "epoch": 3.7719844357976653,
      "grad_norm": 2.21875,
      "learning_rate": 7.430505491563101e-08,
      "loss": 1.1829,
      "step": 1214
    },
    {
      "epoch": 3.7750972762645914,
      "grad_norm": 2.09375,
      "learning_rate": 7.220230559447372e-08,
      "loss": 1.1939,
      "step": 1215
    },
    {
      "epoch": 3.7782101167315174,
      "grad_norm": 2.09375,
      "learning_rate": 7.012952122207206e-08,
      "loss": 1.1673,
      "step": 1216
    },
    {
      "epoch": 3.7813229571984435,
      "grad_norm": 2.03125,
      "learning_rate": 6.808671440261261e-08,
      "loss": 1.1754,
      "step": 1217
    },
    {
      "epoch": 3.7844357976653695,
      "grad_norm": 2.125,
      "learning_rate": 6.607389755799331e-08,
      "loss": 1.1842,
      "step": 1218
    },
    {
      "epoch": 3.7844357976653695,
      "eval_loss": 3.2782089710235596,
      "eval_runtime": 43.7456,
      "eval_samples_per_second": 112.194,
      "eval_steps_per_second": 11.224,
      "step": 1218
    },
    {
      "epoch": 3.7875486381322956,
      "grad_norm": 2.03125,
      "learning_rate": 6.409108292774912e-08,
      "loss": 1.1551,
      "step": 1219
    },
    {
      "epoch": 3.7906614785992216,
      "grad_norm": 2.046875,
      "learning_rate": 6.213828256897869e-08,
      "loss": 1.1875,
      "step": 1220
    },
    {
      "epoch": 3.7937743190661477,
      "grad_norm": 2.125,
      "learning_rate": 6.021550835626777e-08,
      "loss": 1.2097,
      "step": 1221
    },
    {
      "epoch": 3.7968871595330738,
      "grad_norm": 2.0625,
      "learning_rate": 5.832277198162095e-08,
      "loss": 1.1573,
      "step": 1222
    },
    {
      "epoch": 3.8,
      "grad_norm": 2.21875,
      "learning_rate": 5.646008495438948e-08,
      "loss": 1.2244,
      "step": 1223
    },
    {
      "epoch": 3.803112840466926,
      "grad_norm": 2.0,
      "learning_rate": 5.462745860119745e-08,
      "loss": 1.1356,
      "step": 1224
    },
    {
      "epoch": 3.806225680933852,
      "grad_norm": 2.25,
      "learning_rate": 5.2824904065879034e-08,
      "loss": 1.2012,
      "step": 1225
    },
    {
      "epoch": 3.809338521400778,
      "grad_norm": 2.15625,
      "learning_rate": 5.105243230940582e-08,
      "loss": 1.2114,
      "step": 1226
    },
    {
      "epoch": 3.812451361867704,
      "grad_norm": 2.1875,
      "learning_rate": 4.931005410982348e-08,
      "loss": 1.1744,
      "step": 1227
    },
    {
      "epoch": 3.81556420233463,
      "grad_norm": 1.984375,
      "learning_rate": 4.759778006218407e-08,
      "loss": 1.1416,
      "step": 1228
    },
    {
      "epoch": 3.818677042801556,
      "grad_norm": 2.0625,
      "learning_rate": 4.59156205784822e-08,
      "loss": 1.1351,
      "step": 1229
    },
    {
      "epoch": 3.8217898832684822,
      "grad_norm": 2.03125,
      "learning_rate": 4.4263585887592275e-08,
      "loss": 1.1422,
      "step": 1230
    },
    {
      "epoch": 3.8249027237354083,
      "grad_norm": 2.09375,
      "learning_rate": 4.26416860352058e-08,
      "loss": 1.2102,
      "step": 1231
    },
    {
      "epoch": 3.8280155642023344,
      "grad_norm": 2.140625,
      "learning_rate": 4.104993088376974e-08,
      "loss": 1.2105,
      "step": 1232
    },
    {
      "epoch": 3.831128404669261,
      "grad_norm": 2.09375,
      "learning_rate": 3.94883301124277e-08,
      "loss": 1.1951,
      "step": 1233
    },
    {
      "epoch": 3.834241245136187,
      "grad_norm": 2.03125,
      "learning_rate": 3.795689321696105e-08,
      "loss": 1.194,
      "step": 1234
    },
    {
      "epoch": 3.837354085603113,
      "grad_norm": 2.09375,
      "learning_rate": 3.645562950973014e-08,
      "loss": 1.1633,
      "step": 1235
    },
    {
      "epoch": 3.840466926070039,
      "grad_norm": 2.21875,
      "learning_rate": 3.498454811961816e-08,
      "loss": 1.1361,
      "step": 1236
    },
    {
      "epoch": 3.843579766536965,
      "grad_norm": 2.09375,
      "learning_rate": 3.354365799197512e-08,
      "loss": 1.1964,
      "step": 1237
    },
    {
      "epoch": 3.846692607003891,
      "grad_norm": 2.078125,
      "learning_rate": 3.213296788856568e-08,
      "loss": 1.2301,
      "step": 1238
    },
    {
      "epoch": 3.849805447470817,
      "grad_norm": 2.09375,
      "learning_rate": 3.0752486387513046e-08,
      "loss": 1.1139,
      "step": 1239
    },
    {
      "epoch": 3.849805447470817,
      "eval_loss": 3.2763912677764893,
      "eval_runtime": 42.9699,
      "eval_samples_per_second": 114.22,
      "eval_steps_per_second": 11.427,
      "step": 1239
    },
    {
      "epoch": 3.8529182879377433,
      "grad_norm": 2.046875,
      "learning_rate": 2.940222188324793e-08,
      "loss": 1.1899,
      "step": 1240
    },
    {
      "epoch": 3.8560311284046693,
      "grad_norm": 2.0625,
      "learning_rate": 2.8082182586458006e-08,
      "loss": 1.1738,
      "step": 1241
    },
    {
      "epoch": 3.8591439688715954,
      "grad_norm": 2.015625,
      "learning_rate": 2.6792376524036878e-08,
      "loss": 1.195,
      "step": 1242
    },
    {
      "epoch": 3.8622568093385214,
      "grad_norm": 2.03125,
      "learning_rate": 2.5532811539036306e-08,
      "loss": 1.1924,
      "step": 1243
    },
    {
      "epoch": 3.8653696498054475,
      "grad_norm": 2.21875,
      "learning_rate": 2.4303495290618485e-08,
      "loss": 1.1191,
      "step": 1244
    },
    {
      "epoch": 3.8684824902723736,
      "grad_norm": 2.046875,
      "learning_rate": 2.3104435254008852e-08,
      "loss": 1.2247,
      "step": 1245
    },
    {
      "epoch": 3.8715953307392996,
      "grad_norm": 2.09375,
      "learning_rate": 2.1935638720451125e-08,
      "loss": 1.1613,
      "step": 1246
    },
    {
      "epoch": 3.8747081712062257,
      "grad_norm": 2.109375,
      "learning_rate": 2.0797112797161788e-08,
      "loss": 1.2406,
      "step": 1247
    },
    {
      "epoch": 3.8778210116731517,
      "grad_norm": 2.015625,
      "learning_rate": 1.9688864407288455e-08,
      "loss": 1.1949,
      "step": 1248
    },
    {
      "epoch": 3.880933852140078,
      "grad_norm": 2.109375,
      "learning_rate": 1.8610900289867673e-08,
      "loss": 1.0947,
      "step": 1249
    },
    {
      "epoch": 3.884046692607004,
      "grad_norm": 2.046875,
      "learning_rate": 1.7563226999782745e-08,
      "loss": 1.1669,
      "step": 1250
    },
    {
      "epoch": 3.88715953307393,
      "grad_norm": 2.125,
      "learning_rate": 1.6545850907724314e-08,
      "loss": 1.222,
      "step": 1251
    },
    {
      "epoch": 3.890272373540856,
      "grad_norm": 2.015625,
      "learning_rate": 1.555877820015317e-08,
      "loss": 1.1947,
      "step": 1252
    },
    {
      "epoch": 3.893385214007782,
      "grad_norm": 2.015625,
      "learning_rate": 1.4602014879258609e-08,
      "loss": 1.1431,
      "step": 1253
    },
    {
      "epoch": 3.896498054474708,
      "grad_norm": 2.078125,
      "learning_rate": 1.3675566762927916e-08,
      "loss": 1.2133,
      "step": 1254
    },
    {
      "epoch": 3.899610894941634,
      "grad_norm": 2.109375,
      "learning_rate": 1.2779439484705835e-08,
      "loss": 1.1487,
      "step": 1255
    },
    {
      "epoch": 3.90272373540856,
      "grad_norm": 2.125,
      "learning_rate": 1.1913638493762369e-08,
      "loss": 1.2437,
      "step": 1256
    },
    {
      "epoch": 3.9058365758754863,
      "grad_norm": 2.078125,
      "learning_rate": 1.1078169054858922e-08,
      "loss": 1.1619,
      "step": 1257
    },
    {
      "epoch": 3.9089494163424123,
      "grad_norm": 2.09375,
      "learning_rate": 1.0273036248318325e-08,
      "loss": 1.1291,
      "step": 1258
    },
    {
      "epoch": 3.9120622568093384,
      "grad_norm": 2.09375,
      "learning_rate": 9.49824496999152e-09,
      "loss": 1.1654,
      "step": 1259
    },
    {
      "epoch": 3.9151750972762644,
      "grad_norm": 2.109375,
      "learning_rate": 8.753799931228157e-09,
      "loss": 1.179,
      "step": 1260
    },
    {
      "epoch": 3.9151750972762644,
      "eval_loss": 3.2695014476776123,
      "eval_runtime": 42.9558,
      "eval_samples_per_second": 114.257,
      "eval_steps_per_second": 11.43,
      "step": 1260
    },
    {
      "epoch": 3.9182879377431905,
      "grad_norm": 2.109375,
      "learning_rate": 8.039705658849373e-09,
      "loss": 1.1706,
      "step": 1261
    },
    {
      "epoch": 3.9214007782101166,
      "grad_norm": 2.0625,
      "learning_rate": 7.355966495119493e-09,
      "loss": 1.1972,
      "step": 1262
    },
    {
      "epoch": 3.924513618677043,
      "grad_norm": 2.03125,
      "learning_rate": 6.702586597719385e-09,
      "loss": 1.1649,
      "step": 1263
    },
    {
      "epoch": 3.927626459143969,
      "grad_norm": 2.015625,
      "learning_rate": 6.079569939719809e-09,
      "loss": 1.1139,
      "step": 1264
    },
    {
      "epoch": 3.930739299610895,
      "grad_norm": 2.09375,
      "learning_rate": 5.4869203095608836e-09,
      "loss": 1.1733,
      "step": 1265
    },
    {
      "epoch": 3.9338521400778212,
      "grad_norm": 2.046875,
      "learning_rate": 4.924641311026546e-09,
      "loss": 1.1941,
      "step": 1266
    },
    {
      "epoch": 3.9369649805447473,
      "grad_norm": 2.09375,
      "learning_rate": 4.392736363221239e-09,
      "loss": 1.1806,
      "step": 1267
    },
    {
      "epoch": 3.9400778210116734,
      "grad_norm": 2.03125,
      "learning_rate": 3.891208700553262e-09,
      "loss": 1.1544,
      "step": 1268
    },
    {
      "epoch": 3.9431906614785994,
      "grad_norm": 2.078125,
      "learning_rate": 3.420061372711447e-09,
      "loss": 1.1528,
      "step": 1269
    },
    {
      "epoch": 3.9463035019455255,
      "grad_norm": 2.0625,
      "learning_rate": 2.9792972446479605e-09,
      "loss": 1.1665,
      "step": 1270
    },
    {
      "epoch": 3.9494163424124515,
      "grad_norm": 2.09375,
      "learning_rate": 2.568918996560532e-09,
      "loss": 1.1798,
      "step": 1271
    },
    {
      "epoch": 3.9525291828793776,
      "grad_norm": 2.140625,
      "learning_rate": 2.1889291238780254e-09,
      "loss": 1.1744,
      "step": 1272
    },
    {
      "epoch": 3.9556420233463037,
      "grad_norm": 2.15625,
      "learning_rate": 1.839329937241563e-09,
      "loss": 1.1989,
      "step": 1273
    },
    {
      "epoch": 3.9587548638132297,
      "grad_norm": 2.0,
      "learning_rate": 1.5201235624945355e-09,
      "loss": 1.1781,
      "step": 1274
    },
    {
      "epoch": 3.9618677042801558,
      "grad_norm": 2.078125,
      "learning_rate": 1.231311940665947e-09,
      "loss": 1.2166,
      "step": 1275
    },
    {
      "epoch": 3.964980544747082,
      "grad_norm": 2.234375,
      "learning_rate": 9.728968279620887e-10,
      "loss": 1.1727,
      "step": 1276
    },
    {
      "epoch": 3.968093385214008,
      "grad_norm": 1.984375,
      "learning_rate": 7.448797957526621e-10,
      "loss": 1.0861,
      "step": 1277
    },
    {
      "epoch": 3.971206225680934,
      "grad_norm": 1.984375,
      "learning_rate": 5.472622305646713e-10,
      "loss": 1.1443,
      "step": 1278
    },
    {
      "epoch": 3.97431906614786,
      "grad_norm": 2.328125,
      "learning_rate": 3.800453340691013e-10,
      "loss": 1.1482,
      "step": 1279
    },
    {
      "epoch": 3.977431906614786,
      "grad_norm": 2.078125,
      "learning_rate": 2.432301230798073e-10,
      "loss": 1.1517,
      "step": 1280
    },
    {
      "epoch": 3.980544747081712,
      "grad_norm": 2.0625,
      "learning_rate": 1.3681742954185783e-10,
      "loss": 1.1514,
      "step": 1281
    },
    {
      "epoch": 3.980544747081712,
      "eval_loss": 3.283127546310425,
      "eval_runtime": 42.8864,
      "eval_samples_per_second": 114.442,
      "eval_steps_per_second": 11.449,
      "step": 1281
    },
    {
      "epoch": 3.983657587548638,
      "grad_norm": 2.0625,
      "learning_rate": 6.080790052986896e-11,
      "loss": 1.1644,
      "step": 1282
    },
    {
      "epoch": 3.9867704280155642,
      "grad_norm": 2.046875,
      "learning_rate": 1.5201998242453387e-11,
      "loss": 1.1507,
      "step": 1283
    },
    {
      "epoch": 3.9898832684824903,
      "grad_norm": 2.03125,
      "learning_rate": 0.0,
      "loss": 1.0928,
      "step": 1284
    }
  ],
  "logging_steps": 1,
  "max_steps": 1284,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 81,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.7818123129784893e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}