diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,28193 @@
+{
+  "best_metric": 0.18977239727973938,
+  "best_model_checkpoint": "task_base/checkpoint-4000",
+  "epoch": 20.0,
+  "eval_steps": 500,
+  "global_step": 4000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.005,
+      "grad_norm": 98.6826400756836,
+      "learning_rate": 9.9975e-06,
+      "loss": 26.7875,
+      "step": 1
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 130.16192626953125,
+      "learning_rate": 9.995000000000002e-06,
+      "loss": 29.5502,
+      "step": 2
+    },
+    {
+      "epoch": 0.015,
+      "grad_norm": 128.5795440673828,
+      "learning_rate": 9.992500000000001e-06,
+      "loss": 28.2612,
+      "step": 3
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 129.13333129882812,
+      "learning_rate": 9.990000000000001e-06,
+      "loss": 29.185,
+      "step": 4
+    },
+    {
+      "epoch": 0.025,
+      "grad_norm": 80.05451965332031,
+      "learning_rate": 9.9875e-06,
+      "loss": 28.0726,
+      "step": 5
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 113.97787475585938,
+      "learning_rate": 9.985000000000002e-06,
+      "loss": 28.509,
+      "step": 6
+    },
+    {
+      "epoch": 0.035,
+      "grad_norm": 129.45773315429688,
+      "learning_rate": 9.9825e-06,
+      "loss": 30.8329,
+      "step": 7
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 76.6866226196289,
+      "learning_rate": 9.980000000000001e-06,
+      "loss": 26.9813,
+      "step": 8
+    },
+    {
+      "epoch": 0.045,
+      "grad_norm": 76.29522705078125,
+      "learning_rate": 9.9775e-06,
+      "loss": 25.1497,
+      "step": 9
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 133.18101501464844,
+      "learning_rate": 9.975000000000002e-06,
+      "loss": 29.4541,
+      "step": 10
+    },
+    {
+      "epoch": 0.055,
+      "grad_norm": 135.04808044433594,
+      "learning_rate": 9.9725e-06,
+      "loss": 30.4396,
+      "step": 11
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 100.15875244140625,
+      "learning_rate": 9.970000000000001e-06,
+      "loss": 27.681,
+      "step": 12
+    },
+    {
+      "epoch": 0.065,
+      "grad_norm": 124.39775848388672,
+      "learning_rate": 9.9675e-06,
+      "loss": 30.8462,
+      "step": 13
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 99.69007110595703,
+      "learning_rate": 9.965000000000002e-06,
+      "loss": 27.6581,
+      "step": 14
+    },
+    {
+      "epoch": 0.075,
+      "grad_norm": 80.10343170166016,
+      "learning_rate": 9.9625e-06,
+      "loss": 26.3459,
+      "step": 15
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 101.93919372558594,
+      "learning_rate": 9.960000000000001e-06,
+      "loss": 25.0478,
+      "step": 16
+    },
+    {
+      "epoch": 0.085,
+      "grad_norm": 87.31743621826172,
+      "learning_rate": 9.9575e-06,
+      "loss": 24.6051,
+      "step": 17
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 169.87258911132812,
+      "learning_rate": 9.955000000000002e-06,
+      "loss": 33.0518,
+      "step": 18
+    },
+    {
+      "epoch": 0.095,
+      "grad_norm": 128.26705932617188,
+      "learning_rate": 9.9525e-06,
+      "loss": 30.3029,
+      "step": 19
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 121.78145599365234,
+      "learning_rate": 9.950000000000001e-06,
+      "loss": 28.4835,
+      "step": 20
+    },
+    {
+      "epoch": 0.105,
+      "grad_norm": 92.1102066040039,
+      "learning_rate": 9.947500000000001e-06,
+      "loss": 27.2581,
+      "step": 21
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 94.41499328613281,
+      "learning_rate": 9.945e-06,
+      "loss": 26.9037,
+      "step": 22
+    },
+    {
+      "epoch": 0.115,
+      "grad_norm": 71.15790557861328,
+      "learning_rate": 9.9425e-06,
+      "loss": 25.2449,
+      "step": 23
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 90.54814910888672,
+      "learning_rate": 9.940000000000001e-06,
+      "loss": 24.2899,
+      "step": 24
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 158.75157165527344,
+      "learning_rate": 9.937500000000001e-06,
+      "loss": 29.8552,
+      "step": 25
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 142.3026580810547,
+      "learning_rate": 9.935e-06,
+      "loss": 32.3428,
+      "step": 26
+    },
+    {
+      "epoch": 0.135,
+      "grad_norm": 99.76361846923828,
+      "learning_rate": 9.9325e-06,
+      "loss": 22.78,
+      "step": 27
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 84.86548614501953,
+      "learning_rate": 9.930000000000001e-06,
+      "loss": 23.5209,
+      "step": 28
+    },
+    {
+      "epoch": 0.145,
+      "grad_norm": 128.35513305664062,
+      "learning_rate": 9.927500000000001e-06,
+      "loss": 27.2072,
+      "step": 29
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 97.47346496582031,
+      "learning_rate": 9.925e-06,
+      "loss": 26.5399,
+      "step": 30
+    },
+    {
+      "epoch": 0.155,
+      "grad_norm": 210.4054718017578,
+      "learning_rate": 9.9225e-06,
+      "loss": 28.0418,
+      "step": 31
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 127.63544464111328,
+      "learning_rate": 9.920000000000002e-06,
+      "loss": 25.0774,
+      "step": 32
+    },
+    {
+      "epoch": 0.165,
+      "grad_norm": 65.77703094482422,
+      "learning_rate": 9.917500000000001e-06,
+      "loss": 22.8484,
+      "step": 33
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 75.45140075683594,
+      "learning_rate": 9.915e-06,
+      "loss": 23.0948,
+      "step": 34
+    },
+    {
+      "epoch": 0.175,
+      "grad_norm": 89.79125213623047,
+      "learning_rate": 9.9125e-06,
+      "loss": 26.0596,
+      "step": 35
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 76.27912139892578,
+      "learning_rate": 9.91e-06,
+      "loss": 23.8138,
+      "step": 36
+    },
+    {
+      "epoch": 0.185,
+      "grad_norm": 62.73405838012695,
+      "learning_rate": 9.907500000000001e-06,
+      "loss": 20.6296,
+      "step": 37
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 45.03425216674805,
+      "learning_rate": 9.905000000000001e-06,
+      "loss": 15.0878,
+      "step": 38
+    },
+    {
+      "epoch": 0.195,
+      "grad_norm": 88.85426330566406,
+      "learning_rate": 9.9025e-06,
+      "loss": 24.3421,
+      "step": 39
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 76.42606353759766,
+      "learning_rate": 9.9e-06,
+      "loss": 24.4541,
+      "step": 40
+    },
+    {
+      "epoch": 0.205,
+      "grad_norm": 71.12284088134766,
+      "learning_rate": 9.897500000000001e-06,
+      "loss": 23.9476,
+      "step": 41
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 72.09452819824219,
+      "learning_rate": 9.895000000000001e-06,
+      "loss": 22.3881,
+      "step": 42
+    },
+    {
+      "epoch": 0.215,
+      "grad_norm": 82.90807342529297,
+      "learning_rate": 9.8925e-06,
+      "loss": 22.1626,
+      "step": 43
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 106.95568084716797,
+      "learning_rate": 9.89e-06,
+      "loss": 27.6259,
+      "step": 44
+    },
+    {
+      "epoch": 0.225,
+      "grad_norm": 93.54569244384766,
+      "learning_rate": 9.887500000000001e-06,
+      "loss": 25.3901,
+      "step": 45
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 80.322509765625,
+      "learning_rate": 9.885000000000001e-06,
+      "loss": 21.0109,
+      "step": 46
+    },
+    {
+      "epoch": 0.235,
+      "grad_norm": 101.02912139892578,
+      "learning_rate": 9.8825e-06,
+      "loss": 26.5691,
+      "step": 47
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 137.3873748779297,
+      "learning_rate": 9.88e-06,
+      "loss": 27.7835,
+      "step": 48
+    },
+    {
+      "epoch": 0.245,
+      "grad_norm": 115.74402618408203,
+      "learning_rate": 9.877500000000002e-06,
+      "loss": 25.5815,
+      "step": 49
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 114.11056518554688,
+      "learning_rate": 9.875000000000001e-06,
+      "loss": 27.1426,
+      "step": 50
+    },
+    {
+      "epoch": 0.255,
+      "grad_norm": 94.43346405029297,
+      "learning_rate": 9.8725e-06,
+      "loss": 23.5383,
+      "step": 51
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 108.61998748779297,
+      "learning_rate": 9.87e-06,
+      "loss": 23.55,
+      "step": 52
+    },
+    {
+      "epoch": 0.265,
+      "grad_norm": 99.12030792236328,
+      "learning_rate": 9.867500000000002e-06,
+      "loss": 25.2893,
+      "step": 53
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 87.00041961669922,
+      "learning_rate": 9.865000000000001e-06,
+      "loss": 26.3382,
+      "step": 54
+    },
+    {
+      "epoch": 0.275,
+      "grad_norm": 71.67118072509766,
+      "learning_rate": 9.862500000000001e-06,
+      "loss": 20.7062,
+      "step": 55
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 68.72937774658203,
+      "learning_rate": 9.86e-06,
+      "loss": 21.0595,
+      "step": 56
+    },
+    {
+      "epoch": 0.285,
+      "grad_norm": 84.18948364257812,
+      "learning_rate": 9.857500000000002e-06,
+      "loss": 22.0963,
+      "step": 57
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 102.69549560546875,
+      "learning_rate": 9.855000000000001e-06,
+      "loss": 25.2563,
+      "step": 58
+    },
+    {
+      "epoch": 0.295,
+      "grad_norm": 67.48748779296875,
+      "learning_rate": 9.852500000000001e-06,
+      "loss": 19.4465,
+      "step": 59
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 117.39988708496094,
+      "learning_rate": 9.85e-06,
+      "loss": 25.3728,
+      "step": 60
+    },
+    {
+      "epoch": 0.305,
+      "grad_norm": 96.82878875732422,
+      "learning_rate": 9.8475e-06,
+      "loss": 24.8773,
+      "step": 61
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 48.93696212768555,
+      "learning_rate": 9.845000000000001e-06,
+      "loss": 16.9369,
+      "step": 62
+    },
+    {
+      "epoch": 0.315,
+      "grad_norm": 140.5491180419922,
+      "learning_rate": 9.842500000000001e-06,
+      "loss": 29.7644,
+      "step": 63
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 91.80207824707031,
+      "learning_rate": 9.84e-06,
+      "loss": 21.6525,
+      "step": 64
+    },
+    {
+      "epoch": 0.325,
+      "grad_norm": 101.0759048461914,
+      "learning_rate": 9.8375e-06,
+      "loss": 25.469,
+      "step": 65
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 70.44091033935547,
+      "learning_rate": 9.835000000000002e-06,
+      "loss": 23.0819,
+      "step": 66
+    },
+    {
+      "epoch": 0.335,
+      "grad_norm": 84.1961441040039,
+      "learning_rate": 9.832500000000001e-06,
+      "loss": 20.3565,
+      "step": 67
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 77.37662506103516,
+      "learning_rate": 9.83e-06,
+      "loss": 20.2491,
+      "step": 68
+    },
+    {
+      "epoch": 0.345,
+      "grad_norm": 52.62104797363281,
+      "learning_rate": 9.8275e-06,
+      "loss": 17.9186,
+      "step": 69
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 76.30178833007812,
+      "learning_rate": 9.825000000000002e-06,
+      "loss": 20.3917,
+      "step": 70
+    },
+    {
+      "epoch": 0.355,
+      "grad_norm": 69.18999481201172,
+      "learning_rate": 9.8225e-06,
+      "loss": 20.6128,
+      "step": 71
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 60.806060791015625,
+      "learning_rate": 9.820000000000001e-06,
+      "loss": 17.7748,
+      "step": 72
+    },
+    {
+      "epoch": 0.365,
+      "grad_norm": 104.89543914794922,
+      "learning_rate": 9.8175e-06,
+      "loss": 25.6384,
+      "step": 73
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 81.19650268554688,
+      "learning_rate": 9.815000000000002e-06,
+      "loss": 21.5553,
+      "step": 74
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 87.89532470703125,
+      "learning_rate": 9.8125e-06,
+      "loss": 23.4621,
+      "step": 75
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 60.43477249145508,
+      "learning_rate": 9.810000000000001e-06,
+      "loss": 16.6604,
+      "step": 76
+    },
+    {
+      "epoch": 0.385,
+      "grad_norm": 86.52584075927734,
+      "learning_rate": 9.8075e-06,
+      "loss": 22.5716,
+      "step": 77
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 83.47530364990234,
+      "learning_rate": 9.805000000000002e-06,
+      "loss": 23.1142,
+      "step": 78
+    },
+    {
+      "epoch": 0.395,
+      "grad_norm": 129.93313598632812,
+      "learning_rate": 9.8025e-06,
+      "loss": 28.2888,
+      "step": 79
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 98.76376342773438,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 23.4802,
+      "step": 80
+    },
+    {
+      "epoch": 0.405,
+      "grad_norm": 91.87368774414062,
+      "learning_rate": 9.7975e-06,
+      "loss": 20.8709,
+      "step": 81
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 78.7389144897461,
+      "learning_rate": 9.795000000000002e-06,
+      "loss": 19.2707,
+      "step": 82
+    },
+    {
+      "epoch": 0.415,
+      "grad_norm": 73.52212524414062,
+      "learning_rate": 9.7925e-06,
+      "loss": 20.4787,
+      "step": 83
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 62.861488342285156,
+      "learning_rate": 9.790000000000001e-06,
+      "loss": 18.7466,
+      "step": 84
+    },
+    {
+      "epoch": 0.425,
+      "grad_norm": 99.57554626464844,
+      "learning_rate": 9.7875e-06,
+      "loss": 23.2627,
+      "step": 85
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 83.78347778320312,
+      "learning_rate": 9.785e-06,
+      "loss": 20.8797,
+      "step": 86
+    },
+    {
+      "epoch": 0.435,
+      "grad_norm": 57.15800857543945,
+      "learning_rate": 9.7825e-06,
+      "loss": 14.9965,
+      "step": 87
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 78.029052734375,
+      "learning_rate": 9.780000000000001e-06,
+      "loss": 16.8036,
+      "step": 88
+    },
+    {
+      "epoch": 0.445,
+      "grad_norm": 61.482418060302734,
+      "learning_rate": 9.7775e-06,
+      "loss": 16.3136,
+      "step": 89
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 66.77391052246094,
+      "learning_rate": 9.775e-06,
+      "loss": 18.7183,
+      "step": 90
+    },
+    {
+      "epoch": 0.455,
+      "grad_norm": 67.68733215332031,
+      "learning_rate": 9.7725e-06,
+      "loss": 19.0895,
+      "step": 91
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 55.95758056640625,
+      "learning_rate": 9.770000000000001e-06,
+      "loss": 15.0203,
+      "step": 92
+    },
+    {
+      "epoch": 0.465,
+      "grad_norm": 69.09195709228516,
+      "learning_rate": 9.767500000000001e-06,
+      "loss": 16.5279,
+      "step": 93
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 48.976558685302734,
+      "learning_rate": 9.765e-06,
+      "loss": 14.8395,
+      "step": 94
+    },
+    {
+      "epoch": 0.475,
+      "grad_norm": 90.89331817626953,
+      "learning_rate": 9.7625e-06,
+      "loss": 19.8422,
+      "step": 95
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 85.39977264404297,
+      "learning_rate": 9.760000000000001e-06,
+      "loss": 20.1714,
+      "step": 96
+    },
+    {
+      "epoch": 0.485,
+      "grad_norm": 104.15357971191406,
+      "learning_rate": 9.757500000000001e-06,
+      "loss": 22.6641,
+      "step": 97
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 152.04747009277344,
+      "learning_rate": 9.755e-06,
+      "loss": 17.8898,
+      "step": 98
+    },
+    {
+      "epoch": 0.495,
+      "grad_norm": 64.95588684082031,
+      "learning_rate": 9.7525e-06,
+      "loss": 17.6483,
+      "step": 99
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 108.30925750732422,
+      "learning_rate": 9.75e-06,
+      "loss": 21.5085,
+      "step": 100
+    },
+    {
+      "epoch": 0.505,
+      "grad_norm": 69.2565689086914,
+      "learning_rate": 9.747500000000001e-06,
+      "loss": 16.7347,
+      "step": 101
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 81.48815155029297,
+      "learning_rate": 9.745e-06,
+      "loss": 16.962,
+      "step": 102
+    },
+    {
+      "epoch": 0.515,
+      "grad_norm": 84.01866149902344,
+      "learning_rate": 9.7425e-06,
+      "loss": 18.084,
+      "step": 103
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 73.81437683105469,
+      "learning_rate": 9.74e-06,
+      "loss": 16.6248,
+      "step": 104
+    },
+    {
+      "epoch": 0.525,
+      "grad_norm": 45.05405044555664,
+      "learning_rate": 9.737500000000001e-06,
+      "loss": 14.0475,
+      "step": 105
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 67.52337646484375,
+      "learning_rate": 9.735e-06,
+      "loss": 17.407,
+      "step": 106
+    },
+    {
+      "epoch": 0.535,
+      "grad_norm": 65.25855255126953,
+      "learning_rate": 9.7325e-06,
+      "loss": 16.1082,
+      "step": 107
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 51.94172286987305,
+      "learning_rate": 9.73e-06,
+      "loss": 13.5911,
+      "step": 108
+    },
+    {
+      "epoch": 0.545,
+      "grad_norm": 52.56631088256836,
+      "learning_rate": 9.727500000000001e-06,
+      "loss": 13.8112,
+      "step": 109
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 68.89740753173828,
+      "learning_rate": 9.725000000000001e-06,
+      "loss": 15.7815,
+      "step": 110
+    },
+    {
+      "epoch": 0.555,
+      "grad_norm": 51.8447380065918,
+      "learning_rate": 9.7225e-06,
+      "loss": 14.6614,
+      "step": 111
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 56.773799896240234,
+      "learning_rate": 9.72e-06,
+      "loss": 14.1768,
+      "step": 112
+    },
+    {
+      "epoch": 0.565,
+      "grad_norm": 56.056732177734375,
+      "learning_rate": 9.717500000000001e-06,
+      "loss": 14.2544,
+      "step": 113
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 86.14506530761719,
+      "learning_rate": 9.715000000000001e-06,
+      "loss": 19.0712,
+      "step": 114
+    },
+    {
+      "epoch": 0.575,
+      "grad_norm": 60.531982421875,
+      "learning_rate": 9.7125e-06,
+      "loss": 13.9446,
+      "step": 115
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 81.2743148803711,
+      "learning_rate": 9.71e-06,
+      "loss": 17.7113,
+      "step": 116
+    },
+    {
+      "epoch": 0.585,
+      "grad_norm": 65.71122741699219,
+      "learning_rate": 9.707500000000002e-06,
+      "loss": 17.3789,
+      "step": 117
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 73.02108764648438,
+      "learning_rate": 9.705000000000001e-06,
+      "loss": 17.0654,
+      "step": 118
+    },
+    {
+      "epoch": 0.595,
+      "grad_norm": 45.88516616821289,
+      "learning_rate": 9.7025e-06,
+      "loss": 13.668,
+      "step": 119
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 71.4030990600586,
+      "learning_rate": 9.7e-06,
+      "loss": 13.1978,
+      "step": 120
+    },
+    {
+      "epoch": 0.605,
+      "grad_norm": 60.879302978515625,
+      "learning_rate": 9.697500000000002e-06,
+      "loss": 15.14,
+      "step": 121
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 64.47191619873047,
+      "learning_rate": 9.695000000000001e-06,
+      "loss": 13.4599,
+      "step": 122
+    },
+    {
+      "epoch": 0.615,
+      "grad_norm": 64.46489715576172,
+      "learning_rate": 9.6925e-06,
+      "loss": 15.1128,
+      "step": 123
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 75.43682861328125,
+      "learning_rate": 9.69e-06,
+      "loss": 16.5048,
+      "step": 124
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 54.55022430419922,
+      "learning_rate": 9.6875e-06,
+      "loss": 12.699,
+      "step": 125
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 65.59871673583984,
+      "learning_rate": 9.685000000000001e-06,
+      "loss": 14.7196,
+      "step": 126
+    },
+    {
+      "epoch": 0.635,
+      "grad_norm": 45.54060745239258,
+      "learning_rate": 9.682500000000001e-06,
+      "loss": 10.5676,
+      "step": 127
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 85.62071990966797,
+      "learning_rate": 9.68e-06,
+      "loss": 15.2813,
+      "step": 128
+    },
+    {
+      "epoch": 0.645,
+      "grad_norm": 75.14244842529297,
+      "learning_rate": 9.6775e-06,
+      "loss": 15.7681,
+      "step": 129
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 56.10523223876953,
+      "learning_rate": 9.675000000000001e-06,
+      "loss": 12.3535,
+      "step": 130
+    },
+    {
+      "epoch": 0.655,
+      "grad_norm": 97.01177978515625,
+      "learning_rate": 9.672500000000001e-06,
+      "loss": 16.4449,
+      "step": 131
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 62.3574104309082,
+      "learning_rate": 9.67e-06,
+      "loss": 12.2822,
+      "step": 132
+    },
+    {
+      "epoch": 0.665,
+      "grad_norm": 88.70935821533203,
+      "learning_rate": 9.6675e-06,
+      "loss": 16.6384,
+      "step": 133
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 90.1426773071289,
+      "learning_rate": 9.665000000000001e-06,
+      "loss": 15.6102,
+      "step": 134
+    },
+    {
+      "epoch": 0.675,
+      "grad_norm": 76.1562728881836,
+      "learning_rate": 9.662500000000001e-06,
+      "loss": 15.6129,
+      "step": 135
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 53.55042266845703,
+      "learning_rate": 9.66e-06,
+      "loss": 12.8199,
+      "step": 136
+    },
+    {
+      "epoch": 0.685,
+      "grad_norm": 51.21839141845703,
+      "learning_rate": 9.6575e-06,
+      "loss": 11.6335,
+      "step": 137
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 76.6912612915039,
+      "learning_rate": 9.655000000000002e-06,
+      "loss": 13.5561,
+      "step": 138
+    },
+    {
+      "epoch": 0.695,
+      "grad_norm": 60.01485061645508,
+      "learning_rate": 9.652500000000001e-06,
+      "loss": 10.2351,
+      "step": 139
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 106.69459533691406,
+      "learning_rate": 9.65e-06,
+      "loss": 13.3251,
+      "step": 140
+    },
+    {
+      "epoch": 0.705,
+      "grad_norm": 73.65582275390625,
+      "learning_rate": 9.6475e-06,
+      "loss": 12.7267,
+      "step": 141
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 77.97052001953125,
+      "learning_rate": 9.645000000000002e-06,
+      "loss": 12.6486,
+      "step": 142
+    },
+    {
+      "epoch": 0.715,
+      "grad_norm": 79.05912780761719,
+      "learning_rate": 9.642500000000001e-06,
+      "loss": 17.0956,
+      "step": 143
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 62.184776306152344,
+      "learning_rate": 9.640000000000001e-06,
+      "loss": 12.7541,
+      "step": 144
+    },
+    {
+      "epoch": 0.725,
+      "grad_norm": 72.90044403076172,
+      "learning_rate": 9.6375e-06,
+      "loss": 13.0718,
+      "step": 145
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 56.604209899902344,
+      "learning_rate": 9.635000000000002e-06,
+      "loss": 9.7636,
+      "step": 146
+    },
+    {
+      "epoch": 0.735,
+      "grad_norm": 70.15077209472656,
+      "learning_rate": 9.632500000000001e-06,
+      "loss": 12.739,
+      "step": 147
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 80.87680053710938,
+      "learning_rate": 9.630000000000001e-06,
+      "loss": 12.102,
+      "step": 148
+    },
+    {
+      "epoch": 0.745,
+      "grad_norm": 72.05039978027344,
+      "learning_rate": 9.6275e-06,
+      "loss": 12.7588,
+      "step": 149
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 77.3372573852539,
+      "learning_rate": 9.625e-06,
+      "loss": 11.6244,
+      "step": 150
+    },
+    {
+      "epoch": 0.755,
+      "grad_norm": 83.22505187988281,
+      "learning_rate": 9.622500000000001e-06,
+      "loss": 12.8399,
+      "step": 151
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 50.26594161987305,
+      "learning_rate": 9.620000000000001e-06,
+      "loss": 11.0361,
+      "step": 152
+    },
+    {
+      "epoch": 0.765,
+      "grad_norm": 60.52434158325195,
+      "learning_rate": 9.6175e-06,
+      "loss": 10.189,
+      "step": 153
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 43.105796813964844,
+      "learning_rate": 9.615e-06,
+      "loss": 9.3596,
+      "step": 154
+    },
+    {
+      "epoch": 0.775,
+      "grad_norm": 52.157371520996094,
+      "learning_rate": 9.612500000000002e-06,
+      "loss": 9.8226,
+      "step": 155
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 79.5045394897461,
+      "learning_rate": 9.610000000000001e-06,
+      "loss": 13.821,
+      "step": 156
+    },
+    {
+      "epoch": 0.785,
+      "grad_norm": 33.26095962524414,
+      "learning_rate": 9.6075e-06,
+      "loss": 7.971,
+      "step": 157
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 34.945560455322266,
+      "learning_rate": 9.605e-06,
+      "loss": 7.9237,
+      "step": 158
+    },
+    {
+      "epoch": 0.795,
+      "grad_norm": 51.239471435546875,
+      "learning_rate": 9.602500000000002e-06,
+      "loss": 9.0766,
+      "step": 159
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 60.95249938964844,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 11.288,
+      "step": 160
+    },
+    {
+      "epoch": 0.805,
+      "grad_norm": 48.51112747192383,
+      "learning_rate": 9.597500000000001e-06,
+      "loss": 8.1339,
+      "step": 161
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 38.652732849121094,
+      "learning_rate": 9.595e-06,
+      "loss": 9.2921,
+      "step": 162
+    },
+    {
+      "epoch": 0.815,
+      "grad_norm": 91.49453735351562,
+      "learning_rate": 9.592500000000002e-06,
+      "loss": 13.1772,
+      "step": 163
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 34.79182815551758,
+      "learning_rate": 9.59e-06,
+      "loss": 7.7615,
+      "step": 164
+    },
+    {
+      "epoch": 0.825,
+      "grad_norm": 43.71617889404297,
+      "learning_rate": 9.587500000000001e-06,
+      "loss": 9.2388,
+      "step": 165
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 75.4392318725586,
+      "learning_rate": 9.585e-06,
+      "loss": 12.5248,
+      "step": 166
+    },
+    {
+      "epoch": 0.835,
+      "grad_norm": 50.73188018798828,
+      "learning_rate": 9.582500000000002e-06,
+      "loss": 11.2405,
+      "step": 167
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 60.70325469970703,
+      "learning_rate": 9.58e-06,
+      "loss": 10.031,
+      "step": 168
+    },
+    {
+      "epoch": 0.845,
+      "grad_norm": 52.77183532714844,
+      "learning_rate": 9.577500000000001e-06,
+      "loss": 8.8415,
+      "step": 169
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 59.920921325683594,
+      "learning_rate": 9.575e-06,
+      "loss": 11.9102,
+      "step": 170
+    },
+    {
+      "epoch": 0.855,
+      "grad_norm": 56.02485656738281,
+      "learning_rate": 9.572500000000002e-06,
+      "loss": 10.2426,
+      "step": 171
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 48.06016159057617,
+      "learning_rate": 9.57e-06,
+      "loss": 8.535,
+      "step": 172
+    },
+    {
+      "epoch": 0.865,
+      "grad_norm": 66.89437103271484,
+      "learning_rate": 9.567500000000001e-06,
+      "loss": 10.1777,
+      "step": 173
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 73.92242431640625,
+      "learning_rate": 9.565e-06,
+      "loss": 10.38,
+      "step": 174
+    },
+    {
+      "epoch": 0.875,
+      "grad_norm": 60.38460159301758,
+      "learning_rate": 9.562500000000002e-06,
+      "loss": 8.9321,
+      "step": 175
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 33.883548736572266,
+      "learning_rate": 9.56e-06,
+      "loss": 7.9813,
+      "step": 176
+    },
+    {
+      "epoch": 0.885,
+      "grad_norm": 43.302982330322266,
+      "learning_rate": 9.557500000000001e-06,
+      "loss": 7.996,
+      "step": 177
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 30.48736572265625,
+      "learning_rate": 9.555e-06,
+      "loss": 6.9998,
+      "step": 178
+    },
+    {
+      "epoch": 0.895,
+      "grad_norm": 43.6906852722168,
+      "learning_rate": 9.5525e-06,
+      "loss": 7.7499,
+      "step": 179
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 33.401893615722656,
+      "learning_rate": 9.55e-06,
+      "loss": 7.3932,
+      "step": 180
+    },
+    {
+      "epoch": 0.905,
+      "grad_norm": 26.68390655517578,
+      "learning_rate": 9.547500000000001e-06,
+      "loss": 6.3025,
+      "step": 181
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 18.920473098754883,
+      "learning_rate": 9.545000000000001e-06,
+      "loss": 6.0356,
+      "step": 182
+    },
+    {
+      "epoch": 0.915,
+      "grad_norm": 36.93348693847656,
+      "learning_rate": 9.5425e-06,
+      "loss": 6.9846,
+      "step": 183
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 20.760738372802734,
+      "learning_rate": 9.54e-06,
+      "loss": 5.7417,
+      "step": 184
+    },
+    {
+      "epoch": 0.925,
+      "grad_norm": 54.30912780761719,
+      "learning_rate": 9.537500000000001e-06,
+      "loss": 8.7363,
+      "step": 185
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 44.32710647583008,
+      "learning_rate": 9.535000000000001e-06,
+      "loss": 6.9432,
+      "step": 186
+    },
+    {
+      "epoch": 0.935,
+      "grad_norm": 33.204708099365234,
+      "learning_rate": 9.5325e-06,
+      "loss": 7.6664,
+      "step": 187
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 39.57059860229492,
+      "learning_rate": 9.53e-06,
+      "loss": 7.9256,
+      "step": 188
+    },
+    {
+      "epoch": 0.945,
+      "grad_norm": 34.433956146240234,
+      "learning_rate": 9.5275e-06,
+      "loss": 7.7483,
+      "step": 189
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 59.56363296508789,
+      "learning_rate": 9.525000000000001e-06,
+      "loss": 9.6471,
+      "step": 190
+    },
+    {
+      "epoch": 0.955,
+      "grad_norm": 20.172027587890625,
+      "learning_rate": 9.5225e-06,
+      "loss": 6.3757,
+      "step": 191
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 55.61946105957031,
+      "learning_rate": 9.52e-06,
+      "loss": 8.6142,
+      "step": 192
+    },
+    {
+      "epoch": 0.965,
+      "grad_norm": 17.15169906616211,
+      "learning_rate": 9.5175e-06,
+      "loss": 6.3336,
+      "step": 193
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 70.50774383544922,
+      "learning_rate": 9.515000000000001e-06,
+      "loss": 7.7348,
+      "step": 194
+    },
+    {
+      "epoch": 0.975,
+      "grad_norm": 22.035673141479492,
+      "learning_rate": 9.5125e-06,
+      "loss": 6.8391,
+      "step": 195
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 54.45932388305664,
+      "learning_rate": 9.51e-06,
+      "loss": 8.642,
+      "step": 196
+    },
+    {
+      "epoch": 0.985,
+      "grad_norm": 18.351449966430664,
+      "learning_rate": 9.5075e-06,
+      "loss": 6.6387,
+      "step": 197
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 45.91608810424805,
+      "learning_rate": 9.505000000000001e-06,
+      "loss": 9.2273,
+      "step": 198
+    },
+    {
+      "epoch": 0.995,
+      "grad_norm": 49.50335693359375,
+      "learning_rate": 9.502500000000001e-06,
+      "loss": 8.2617,
+      "step": 199
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 68.06169891357422,
+      "learning_rate": 9.5e-06,
+      "loss": 9.5812,
+      "step": 200
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 5.654143810272217,
+      "eval_runtime": 16.4561,
+      "eval_samples_per_second": 48.493,
+      "eval_steps_per_second": 6.077,
+      "step": 200
+    },
+    {
+      "epoch": 1.005,
+      "grad_norm": 19.128555297851562,
+      "learning_rate": 9.4975e-06,
+      "loss": 6.5611,
+      "step": 201
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 36.564815521240234,
+      "learning_rate": 9.495000000000001e-06,
+      "loss": 7.1963,
+      "step": 202
+    },
+    {
+      "epoch": 1.015,
+      "grad_norm": 22.483034133911133,
+      "learning_rate": 9.492500000000001e-06,
+      "loss": 6.7288,
+      "step": 203
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 30.927268981933594,
+      "learning_rate": 9.49e-06,
+      "loss": 6.5283,
+      "step": 204
+    },
+    {
+      "epoch": 1.025,
+      "grad_norm": 48.2264404296875,
+      "learning_rate": 9.4875e-06,
+      "loss": 7.7132,
+      "step": 205
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 51.05891418457031,
+      "learning_rate": 9.485000000000002e-06,
+      "loss": 7.5766,
+      "step": 206
+    },
+    {
+      "epoch": 1.035,
+      "grad_norm": 32.528873443603516,
+      "learning_rate": 9.482500000000001e-06,
+      "loss": 6.7818,
+      "step": 207
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 25.026874542236328,
+      "learning_rate": 9.48e-06,
+      "loss": 6.5215,
+      "step": 208
+    },
+    {
+      "epoch": 1.045,
+      "grad_norm": 39.68418502807617,
+      "learning_rate": 9.4775e-06,
+      "loss": 6.3149,
+      "step": 209
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 19.05411720275879,
+      "learning_rate": 9.475000000000002e-06,
+      "loss": 5.6285,
+      "step": 210
+    },
+    {
+      "epoch": 1.055,
+      "grad_norm": 72.75760650634766,
+      "learning_rate": 9.472500000000001e-06,
+      "loss": 9.9226,
+      "step": 211
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 36.22922134399414,
+      "learning_rate": 9.47e-06,
+      "loss": 6.9132,
+      "step": 212
+    },
+    {
+      "epoch": 1.065,
+      "grad_norm": 21.323957443237305,
+      "learning_rate": 9.4675e-06,
+      "loss": 6.1066,
+      "step": 213
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 48.98505401611328,
+      "learning_rate": 9.465e-06,
+      "loss": 7.1241,
+      "step": 214
+    },
+    {
+      "epoch": 1.075,
+      "grad_norm": 14.870978355407715,
+      "learning_rate": 9.462500000000001e-06,
+      "loss": 5.5579,
+      "step": 215
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 16.360605239868164,
+      "learning_rate": 9.460000000000001e-06,
+      "loss": 5.5027,
+      "step": 216
+    },
+    {
+      "epoch": 1.085,
+      "grad_norm": 14.37453842163086,
+      "learning_rate": 9.4575e-06,
+      "loss": 5.5242,
+      "step": 217
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 19.441381454467773,
+      "learning_rate": 9.455e-06,
+      "loss": 6.493,
+      "step": 218
+    },
+    {
+      "epoch": 1.095,
+      "grad_norm": 34.427207946777344,
+      "learning_rate": 9.452500000000001e-06,
+      "loss": 7.8062,
+      "step": 219
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 32.38112258911133,
+      "learning_rate": 9.450000000000001e-06,
+      "loss": 6.688,
+      "step": 220
+    },
+    {
+      "epoch": 1.105,
+      "grad_norm": 22.37595558166504,
+      "learning_rate": 9.4475e-06,
+      "loss": 6.2928,
+      "step": 221
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 31.03255844116211,
+      "learning_rate": 9.445e-06,
+      "loss": 6.4515,
+      "step": 222
+    },
+    {
+      "epoch": 1.115,
+      "grad_norm": 25.20647621154785,
+      "learning_rate": 9.442500000000001e-06,
+      "loss": 6.0165,
+      "step": 223
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 24.565000534057617,
+      "learning_rate": 9.440000000000001e-06,
+      "loss": 6.601,
+      "step": 224
+    },
+    {
+      "epoch": 1.125,
+      "grad_norm": 28.457658767700195,
+      "learning_rate": 9.4375e-06,
+      "loss": 5.9095,
+      "step": 225
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 59.42367935180664,
+      "learning_rate": 9.435e-06,
+      "loss": 8.9338,
+      "step": 226
+    },
+    {
+      "epoch": 1.135,
+      "grad_norm": 66.78691101074219,
+      "learning_rate": 9.432500000000002e-06,
+      "loss": 8.2992,
+      "step": 227
+    },
+    {
+      "epoch": 1.1400000000000001,
+      "grad_norm": 11.681739807128906,
+      "learning_rate": 9.43e-06,
+      "loss": 5.5204,
+      "step": 228
+    },
+    {
+      "epoch": 1.145,
+      "grad_norm": 12.233701705932617,
+      "learning_rate": 9.4275e-06,
+      "loss": 5.834,
+      "step": 229
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 30.185115814208984,
+      "learning_rate": 9.425e-06,
+      "loss": 6.296,
+      "step": 230
+    },
+    {
+      "epoch": 1.155,
+      "grad_norm": 26.54530143737793,
+      "learning_rate": 9.422500000000002e-06,
+      "loss": 5.1923,
+      "step": 231
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 96.65982055664062,
+      "learning_rate": 9.42e-06,
+      "loss": 6.2701,
+      "step": 232
+    },
+    {
+      "epoch": 1.165,
+      "grad_norm": 17.669124603271484,
+      "learning_rate": 9.417500000000001e-06,
+      "loss": 5.1329,
+      "step": 233
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 14.36751651763916,
+      "learning_rate": 9.415e-06,
+      "loss": 5.8423,
+      "step": 234
+    },
+    {
+      "epoch": 1.175,
+      "grad_norm": 13.57278823852539,
+      "learning_rate": 9.412500000000002e-06,
+      "loss": 5.3048,
+      "step": 235
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 27.2868709564209,
+      "learning_rate": 9.41e-06,
+      "loss": 6.5665,
+      "step": 236
+    },
+    {
+      "epoch": 1.185,
+      "grad_norm": 12.037609100341797,
+      "learning_rate": 9.407500000000001e-06,
+      "loss": 4.8999,
+      "step": 237
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 47.080013275146484,
+      "learning_rate": 9.405e-06,
+      "loss": 7.4874,
+      "step": 238
+    },
+    {
+      "epoch": 1.195,
+      "grad_norm": 12.558409690856934,
+      "learning_rate": 9.402500000000002e-06,
+      "loss": 5.2037,
+      "step": 239
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 22.80580711364746,
+      "learning_rate": 9.4e-06,
+      "loss": 5.6469,
+      "step": 240
+    },
+    {
+      "epoch": 1.205,
+      "grad_norm": 13.897841453552246,
+      "learning_rate": 9.397500000000001e-06,
+      "loss": 5.3935,
+      "step": 241
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 22.320159912109375,
+      "learning_rate": 9.395e-06,
+      "loss": 6.2718,
+      "step": 242
+    },
+    {
+      "epoch": 1.215,
+      "grad_norm": 16.342809677124023,
+      "learning_rate": 9.3925e-06,
+      "loss": 5.8632,
+      "step": 243
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 21.0856876373291,
+      "learning_rate": 9.39e-06,
+      "loss": 5.7345,
+      "step": 244
+    },
+    {
+      "epoch": 1.225,
+      "grad_norm": 15.833319664001465,
+      "learning_rate": 9.387500000000001e-06,
+      "loss": 5.8321,
+      "step": 245
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 14.124794960021973,
+      "learning_rate": 9.385e-06,
+      "loss": 4.9309,
+      "step": 246
+    },
+    {
+      "epoch": 1.2349999999999999,
+      "grad_norm": 48.34260177612305,
+      "learning_rate": 9.3825e-06,
+      "loss": 5.2509,
+      "step": 247
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 16.606124877929688,
+      "learning_rate": 9.38e-06,
+      "loss": 6.1208,
+      "step": 248
+    },
+    {
+      "epoch": 1.245,
+      "grad_norm": 13.457324028015137,
+      "learning_rate": 9.377500000000001e-06,
+      "loss": 5.362,
+      "step": 249
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 28.144323348999023,
+      "learning_rate": 9.375000000000001e-06,
+      "loss": 5.8093,
+      "step": 250
+    },
+    {
+      "epoch": 1.255,
+      "grad_norm": 19.901933670043945,
+      "learning_rate": 9.3725e-06,
+      "loss": 5.1549,
+      "step": 251
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 14.436508178710938,
+      "learning_rate": 9.370000000000002e-06,
+      "loss": 5.4795,
+      "step": 252
+    },
+    {
+      "epoch": 1.2650000000000001,
+      "grad_norm": 17.70148277282715,
+      "learning_rate": 9.3675e-06,
+      "loss": 5.2439,
+      "step": 253
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 15.188850402832031,
+      "learning_rate": 9.365000000000001e-06,
+      "loss": 5.4887,
+      "step": 254
+    },
+    {
+      "epoch": 1.275,
+      "grad_norm": 23.32349967956543,
+      "learning_rate": 9.3625e-06,
+      "loss": 5.5844,
+      "step": 255
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 14.434393882751465,
+      "learning_rate": 9.360000000000002e-06,
+      "loss": 5.1482,
+      "step": 256
+    },
+    {
+      "epoch": 1.285,
+      "grad_norm": 20.46137809753418,
+      "learning_rate": 9.3575e-06,
+      "loss": 5.2333,
+      "step": 257
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 28.507661819458008,
+      "learning_rate": 9.355000000000001e-06,
+      "loss": 5.7363,
+      "step": 258
+    },
+    {
+      "epoch": 1.295,
+      "grad_norm": 16.82868194580078,
+      "learning_rate": 9.3525e-06,
+      "loss": 5.6211,
+      "step": 259
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 17.33863067626953,
+      "learning_rate": 9.350000000000002e-06,
+      "loss": 5.4096,
+      "step": 260
+    },
+    {
+      "epoch": 1.305,
+      "grad_norm": 13.24040699005127,
+      "learning_rate": 9.3475e-06,
+      "loss": 5.4104,
+      "step": 261
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 17.547054290771484,
+      "learning_rate": 9.345000000000001e-06,
+      "loss": 4.9409,
+      "step": 262
+    },
+    {
+      "epoch": 1.315,
+      "grad_norm": 15.64814567565918,
+      "learning_rate": 9.3425e-06,
+      "loss": 4.8756,
+      "step": 263
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 24.33405303955078,
+      "learning_rate": 9.340000000000002e-06,
+      "loss": 5.0255,
+      "step": 264
+    },
+    {
+      "epoch": 1.325,
+      "grad_norm": 12.722175598144531,
+      "learning_rate": 9.3375e-06,
+      "loss": 4.4596,
+      "step": 265
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 16.860002517700195,
+      "learning_rate": 9.335000000000001e-06,
+      "loss": 5.3056,
+      "step": 266
+    },
+    {
+      "epoch": 1.335,
+      "grad_norm": 21.540050506591797,
+      "learning_rate": 9.332500000000001e-06,
+      "loss": 5.067,
+      "step": 267
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 17.87770652770996,
+      "learning_rate": 9.33e-06,
+      "loss": 4.8456,
+      "step": 268
+    },
+    {
+      "epoch": 1.345,
+      "grad_norm": 16.975229263305664,
+      "learning_rate": 9.3275e-06,
+      "loss": 5.165,
+      "step": 269
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 29.36518669128418,
+      "learning_rate": 9.325000000000001e-06,
+      "loss": 4.9931,
+      "step": 270
+    },
+    {
+      "epoch": 1.355,
+      "grad_norm": 22.000598907470703,
+      "learning_rate": 9.322500000000001e-06,
+      "loss": 6.3005,
+      "step": 271
+    },
+    {
+      "epoch": 1.3599999999999999,
+      "grad_norm": 15.200225830078125,
+      "learning_rate": 9.32e-06,
+      "loss": 4.6263,
+      "step": 272
+    },
+    {
+      "epoch": 1.365,
+      "grad_norm": 26.238624572753906,
+      "learning_rate": 9.3175e-06,
+      "loss": 5.5159,
+      "step": 273
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 20.63860321044922,
+      "learning_rate": 9.315000000000001e-06,
+      "loss": 5.3313,
+      "step": 274
+    },
+    {
+      "epoch": 1.375,
+      "grad_norm": 19.636701583862305,
+      "learning_rate": 9.312500000000001e-06,
+      "loss": 5.1795,
+      "step": 275
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 45.85452651977539,
+      "learning_rate": 9.31e-06,
+      "loss": 5.9442,
+      "step": 276
+    },
+    {
+      "epoch": 1.385,
+      "grad_norm": 37.55158233642578,
+      "learning_rate": 9.3075e-06,
+      "loss": 6.0706,
+      "step": 277
+    },
+    {
+      "epoch": 1.3900000000000001,
+      "grad_norm": 21.770265579223633,
+      "learning_rate": 9.305000000000002e-06,
+      "loss": 5.4709,
+      "step": 278
+    },
+    {
+      "epoch": 1.395,
+      "grad_norm": 15.628860473632812,
+      "learning_rate": 9.302500000000001e-06,
+      "loss": 4.6683,
+      "step": 279
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 18.33157730102539,
+      "learning_rate": 9.3e-06,
+      "loss": 3.8461,
+      "step": 280
+    },
+    {
+      "epoch": 1.405,
+      "grad_norm": 19.2154541015625,
+      "learning_rate": 9.2975e-06,
+      "loss": 4.0097,
+      "step": 281
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 12.176109313964844,
+      "learning_rate": 9.295e-06,
+      "loss": 5.104,
+      "step": 282
+    },
+    {
+      "epoch": 1.415,
+      "grad_norm": 13.833988189697266,
+      "learning_rate": 9.292500000000001e-06,
+      "loss": 3.9899,
+      "step": 283
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 21.36445426940918,
+      "learning_rate": 9.29e-06,
+      "loss": 5.057,
+      "step": 284
+    },
+    {
+      "epoch": 1.425,
+      "grad_norm": 158.6640625,
+      "learning_rate": 9.2875e-06,
+      "loss": 4.7766,
+      "step": 285
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 18.744028091430664,
+      "learning_rate": 9.285e-06,
+      "loss": 5.0241,
+      "step": 286
+    },
+    {
+      "epoch": 1.435,
+      "grad_norm": 35.88058853149414,
+      "learning_rate": 9.282500000000001e-06,
+      "loss": 5.8649,
+      "step": 287
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 15.70468521118164,
+      "learning_rate": 9.280000000000001e-06,
+      "loss": 4.0588,
+      "step": 288
+    },
+    {
+      "epoch": 1.445,
+      "grad_norm": 15.735753059387207,
+      "learning_rate": 9.2775e-06,
+      "loss": 4.1133,
+      "step": 289
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 26.22873306274414,
+      "learning_rate": 9.275e-06,
+      "loss": 4.7342,
+      "step": 290
+    },
+    {
+      "epoch": 1.455,
+      "grad_norm": 17.944917678833008,
+      "learning_rate": 9.272500000000001e-06,
+      "loss": 3.9767,
+      "step": 291
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 23.74828338623047,
+      "learning_rate": 9.270000000000001e-06,
+      "loss": 3.7015,
+      "step": 292
+    },
+    {
+      "epoch": 1.465,
+      "grad_norm": 34.3891487121582,
+      "learning_rate": 9.2675e-06,
+      "loss": 4.5864,
+      "step": 293
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 15.60699462890625,
+      "learning_rate": 9.265e-06,
+      "loss": 5.2254,
+      "step": 294
+    },
+    {
+      "epoch": 1.475,
+      "grad_norm": 15.06813907623291,
+      "learning_rate": 9.262500000000002e-06,
+      "loss": 4.1259,
+      "step": 295
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 13.980265617370605,
+      "learning_rate": 9.260000000000001e-06,
+      "loss": 4.4235,
+      "step": 296
+    },
+    {
+      "epoch": 1.4849999999999999,
+      "grad_norm": 12.710746765136719,
+      "learning_rate": 9.2575e-06,
+      "loss": 4.4286,
+      "step": 297
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 15.565831184387207,
+      "learning_rate": 9.255e-06,
+      "loss": 3.9422,
+      "step": 298
+    },
+    {
+      "epoch": 1.495,
+      "grad_norm": 13.09053897857666,
+      "learning_rate": 9.252500000000002e-06,
+      "loss": 3.9304,
+      "step": 299
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 12.519309043884277,
+      "learning_rate": 9.250000000000001e-06,
+      "loss": 4.4601,
+      "step": 300
+    },
+    {
+      "epoch": 1.505,
+      "grad_norm": 39.41765594482422,
+      "learning_rate": 9.2475e-06,
+      "loss": 5.2401,
+      "step": 301
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 16.737485885620117,
+      "learning_rate": 9.245e-06,
+      "loss": 3.5006,
+      "step": 302
+    },
+    {
+      "epoch": 1.5150000000000001,
+      "grad_norm": 19.48272132873535,
+      "learning_rate": 9.242500000000002e-06,
+      "loss": 4.401,
+      "step": 303
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 21.063579559326172,
+      "learning_rate": 9.240000000000001e-06,
+      "loss": 4.2256,
+      "step": 304
+    },
+    {
+      "epoch": 1.525,
+      "grad_norm": 20.35517120361328,
+      "learning_rate": 9.237500000000001e-06,
+      "loss": 4.3057,
+      "step": 305
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 15.529792785644531,
+      "learning_rate": 9.235e-06,
+      "loss": 3.7291,
+      "step": 306
+    },
+    {
+      "epoch": 1.5350000000000001,
+      "grad_norm": 27.14385223388672,
+      "learning_rate": 9.2325e-06,
+      "loss": 4.345,
+      "step": 307
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 27.56285285949707,
+      "learning_rate": 9.230000000000001e-06,
+      "loss": 4.3392,
+      "step": 308
+    },
+    {
+      "epoch": 1.545,
+      "grad_norm": 24.108661651611328,
+      "learning_rate": 9.227500000000001e-06,
+      "loss": 4.7867,
+      "step": 309
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 15.940447807312012,
+      "learning_rate": 9.225e-06,
+      "loss": 4.9862,
+      "step": 310
+    },
+    {
+      "epoch": 1.5550000000000002,
+      "grad_norm": 13.561243057250977,
+      "learning_rate": 9.2225e-06,
+      "loss": 4.2635,
+      "step": 311
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 25.27726936340332,
+      "learning_rate": 9.220000000000002e-06,
+      "loss": 4.4889,
+      "step": 312
+    },
+    {
+      "epoch": 1.565,
+      "grad_norm": 12.268288612365723,
+      "learning_rate": 9.217500000000001e-06,
+      "loss": 3.9313,
+      "step": 313
+    },
+    {
+      "epoch": 1.5699999999999998,
+      "grad_norm": 16.627775192260742,
+      "learning_rate": 9.215e-06,
+      "loss": 4.374,
+      "step": 314
+    },
+    {
+      "epoch": 1.575,
+      "grad_norm": 13.112637519836426,
+      "learning_rate": 9.2125e-06,
+      "loss": 3.63,
+      "step": 315
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 20.270721435546875,
+      "learning_rate": 9.210000000000002e-06,
+      "loss": 4.6569,
+      "step": 316
+    },
+    {
+      "epoch": 1.585,
+      "grad_norm": 14.087224006652832,
+      "learning_rate": 9.2075e-06,
+      "loss": 4.1085,
+      "step": 317
+    },
+    {
+      "epoch": 1.5899999999999999,
+      "grad_norm": 16.007585525512695,
+      "learning_rate": 9.205e-06,
+      "loss": 3.8162,
+      "step": 318
+    },
+    {
+      "epoch": 1.595,
+      "grad_norm": 15.254408836364746,
+      "learning_rate": 9.2025e-06,
+      "loss": 4.218,
+      "step": 319
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 13.776213645935059,
+      "learning_rate": 9.200000000000002e-06,
+      "loss": 4.0463,
+      "step": 320
+    },
+    {
+      "epoch": 1.605,
+      "grad_norm": 11.78616714477539,
+      "learning_rate": 9.1975e-06,
+      "loss": 3.9821,
+      "step": 321
+    },
+    {
+      "epoch": 1.6099999999999999,
+      "grad_norm": 15.8296537399292,
+      "learning_rate": 9.195000000000001e-06,
+      "loss": 3.3531,
+      "step": 322
+    },
+    {
+      "epoch": 1.615,
+      "grad_norm": 18.58026885986328,
+      "learning_rate": 9.1925e-06,
+      "loss": 4.4408,
+      "step": 323
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 50.75275802612305,
+      "learning_rate": 9.190000000000002e-06,
+      "loss": 4.4751,
+      "step": 324
+    },
+    {
+      "epoch": 1.625,
+      "grad_norm": 19.88287925720215,
+      "learning_rate": 9.1875e-06,
+      "loss": 3.8357,
+      "step": 325
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 17.656288146972656,
+      "learning_rate": 9.185000000000001e-06,
+      "loss": 4.3007,
+      "step": 326
+    },
+    {
+      "epoch": 1.635,
+      "grad_norm": 13.043316841125488,
+      "learning_rate": 9.1825e-06,
+      "loss": 3.6775,
+      "step": 327
+    },
+    {
+      "epoch": 1.6400000000000001,
+      "grad_norm": 12.428067207336426,
+      "learning_rate": 9.180000000000002e-06,
+      "loss": 3.3341,
+      "step": 328
+    },
+    {
+      "epoch": 1.645,
+      "grad_norm": 13.9157133102417,
+      "learning_rate": 9.1775e-06,
+      "loss": 3.4281,
+      "step": 329
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 13.140881538391113,
+      "learning_rate": 9.175000000000001e-06,
+      "loss": 3.8547,
+      "step": 330
+    },
+    {
+      "epoch": 1.655,
+      "grad_norm": 18.8797607421875,
+      "learning_rate": 9.1725e-06,
+      "loss": 4.0548,
+      "step": 331
+    },
+    {
+      "epoch": 1.6600000000000001,
+      "grad_norm": 259.4732666015625,
+      "learning_rate": 9.17e-06,
+      "loss": 3.6696,
+      "step": 332
+    },
+    {
+      "epoch": 1.665,
+      "grad_norm": 16.65726089477539,
+      "learning_rate": 9.1675e-06,
+      "loss": 4.3721,
+      "step": 333
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 17.92304039001465,
+      "learning_rate": 9.165000000000001e-06,
+      "loss": 4.3179,
+      "step": 334
+    },
+    {
+      "epoch": 1.675,
+      "grad_norm": 18.12198257446289,
+      "learning_rate": 9.1625e-06,
+      "loss": 3.7976,
+      "step": 335
+    },
+    {
+      "epoch": 1.6800000000000002,
+      "grad_norm": 18.153013229370117,
+      "learning_rate": 9.16e-06,
+      "loss": 3.6183,
+      "step": 336
+    },
+    {
+      "epoch": 1.685,
+      "grad_norm": 17.950115203857422,
+      "learning_rate": 9.1575e-06,
+      "loss": 3.1439,
+      "step": 337
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 17.969717025756836,
+      "learning_rate": 9.155000000000001e-06,
+      "loss": 3.9333,
+      "step": 338
+    },
+    {
+      "epoch": 1.6949999999999998,
+      "grad_norm": 26.744415283203125,
+      "learning_rate": 9.152500000000001e-06,
+      "loss": 3.3695,
+      "step": 339
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 21.588607788085938,
+      "learning_rate": 9.15e-06,
+      "loss": 4.485,
+      "step": 340
+    },
+    {
+      "epoch": 1.705,
+      "grad_norm": 16.382080078125,
+      "learning_rate": 9.1475e-06,
+      "loss": 3.3703,
+      "step": 341
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 14.781033515930176,
+      "learning_rate": 9.145000000000001e-06,
+      "loss": 3.3075,
+      "step": 342
+    },
+    {
+      "epoch": 1.7149999999999999,
+      "grad_norm": 9.9727783203125,
+      "learning_rate": 9.142500000000001e-06,
+      "loss": 3.2767,
+      "step": 343
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 20.854019165039062,
+      "learning_rate": 9.14e-06,
+      "loss": 4.526,
+      "step": 344
+    },
+    {
+      "epoch": 1.725,
+      "grad_norm": 13.6721830368042,
+      "learning_rate": 9.1375e-06,
+      "loss": 3.7636,
+      "step": 345
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 15.286890983581543,
+      "learning_rate": 9.135e-06,
+      "loss": 3.6536,
+      "step": 346
+    },
+    {
+      "epoch": 1.7349999999999999,
+      "grad_norm": 12.589783668518066,
+      "learning_rate": 9.132500000000001e-06,
+      "loss": 4.2612,
+      "step": 347
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 14.62938117980957,
+      "learning_rate": 9.13e-06,
+      "loss": 3.6396,
+      "step": 348
+    },
+    {
+      "epoch": 1.745,
+      "grad_norm": 17.03478240966797,
+      "learning_rate": 9.1275e-06,
+      "loss": 3.6213,
+      "step": 349
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 13.38320541381836,
+      "learning_rate": 9.125e-06,
+      "loss": 3.8689,
+      "step": 350
+    },
+    {
+      "epoch": 1.755,
+      "grad_norm": 18.37547492980957,
+      "learning_rate": 9.122500000000001e-06,
+      "loss": 3.686,
+      "step": 351
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 17.98255157470703,
+      "learning_rate": 9.12e-06,
+      "loss": 3.9075,
+      "step": 352
+    },
+    {
+      "epoch": 1.7650000000000001,
+      "grad_norm": 19.452011108398438,
+      "learning_rate": 9.1175e-06,
+      "loss": 3.7084,
+      "step": 353
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 38.63498306274414,
+      "learning_rate": 9.115e-06,
+      "loss": 3.9272,
+      "step": 354
+    },
+    {
+      "epoch": 1.775,
+      "grad_norm": 11.674823760986328,
+      "learning_rate": 9.112500000000001e-06,
+      "loss": 3.5812,
+      "step": 355
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 15.235624313354492,
+      "learning_rate": 9.110000000000001e-06,
+      "loss": 2.854,
+      "step": 356
+    },
+    {
+      "epoch": 1.7850000000000001,
+      "grad_norm": 22.987409591674805,
+      "learning_rate": 9.1075e-06,
+      "loss": 4.5732,
+      "step": 357
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 13.051648139953613,
+      "learning_rate": 9.105e-06,
+      "loss": 3.3024,
+      "step": 358
+    },
+    {
+      "epoch": 1.795,
+      "grad_norm": 15.852628707885742,
+      "learning_rate": 9.102500000000001e-06,
+      "loss": 3.9023,
+      "step": 359
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 13.126157760620117,
+      "learning_rate": 9.100000000000001e-06,
+      "loss": 3.2484,
+      "step": 360
+    },
+    {
+      "epoch": 1.8050000000000002,
+      "grad_norm": 12.919014930725098,
+      "learning_rate": 9.0975e-06,
+      "loss": 3.1793,
+      "step": 361
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 13.069762229919434,
+      "learning_rate": 9.095e-06,
+      "loss": 3.5063,
+      "step": 362
+    },
+    {
+      "epoch": 1.815,
+      "grad_norm": 12.379800796508789,
+      "learning_rate": 9.092500000000001e-06,
+      "loss": 3.5545,
+      "step": 363
+    },
+    {
+      "epoch": 1.8199999999999998,
+      "grad_norm": 14.911227226257324,
+      "learning_rate": 9.090000000000001e-06,
+      "loss": 3.2847,
+      "step": 364
+    },
+    {
+      "epoch": 1.825,
+      "grad_norm": 11.489347457885742,
+      "learning_rate": 9.0875e-06,
+      "loss": 3.1312,
+      "step": 365
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 13.805174827575684,
+      "learning_rate": 9.085e-06,
+      "loss": 3.0411,
+      "step": 366
+    },
+    {
+      "epoch": 1.835,
+      "grad_norm": 13.91395378112793,
+      "learning_rate": 9.082500000000002e-06,
+      "loss": 3.588,
+      "step": 367
+    },
+    {
+      "epoch": 1.8399999999999999,
+      "grad_norm": 16.66775894165039,
+      "learning_rate": 9.080000000000001e-06,
+      "loss": 2.9282,
+      "step": 368
+    },
+    {
+      "epoch": 1.845,
+      "grad_norm": 13.521129608154297,
+      "learning_rate": 9.0775e-06,
+      "loss": 3.735,
+      "step": 369
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 16.441198348999023,
+      "learning_rate": 9.075e-06,
+      "loss": 3.2501,
+      "step": 370
+    },
+    {
+      "epoch": 1.855,
+      "grad_norm": 16.147235870361328,
+      "learning_rate": 9.0725e-06,
+      "loss": 3.2706,
+      "step": 371
+    },
+    {
+      "epoch": 1.8599999999999999,
+      "grad_norm": 20.167339324951172,
+      "learning_rate": 9.070000000000001e-06,
+      "loss": 4.2765,
+      "step": 372
+    },
+    {
+      "epoch": 1.865,
+      "grad_norm": 19.808313369750977,
+      "learning_rate": 9.067500000000001e-06,
+      "loss": 3.457,
+      "step": 373
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 17.18585777282715,
+      "learning_rate": 9.065e-06,
+      "loss": 3.5391,
+      "step": 374
+    },
+    {
+      "epoch": 1.875,
+      "grad_norm": 12.90639877319336,
+      "learning_rate": 9.0625e-06,
+      "loss": 3.316,
+      "step": 375
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 16.349651336669922,
+      "learning_rate": 9.060000000000001e-06,
+      "loss": 2.8231,
+      "step": 376
+    },
+    {
+      "epoch": 1.885,
+      "grad_norm": 12.448009490966797,
+      "learning_rate": 9.057500000000001e-06,
+      "loss": 2.8123,
+      "step": 377
+    },
+    {
+      "epoch": 1.8900000000000001,
+      "grad_norm": 17.73921775817871,
+      "learning_rate": 9.055e-06,
+      "loss": 3.5571,
+      "step": 378
+    },
+    {
+      "epoch": 1.895,
+      "grad_norm": 22.70398712158203,
+      "learning_rate": 9.0525e-06,
+      "loss": 4.43,
+      "step": 379
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 235.1514434814453,
+      "learning_rate": 9.050000000000001e-06,
+      "loss": 4.6211,
+      "step": 380
+    },
+    {
+      "epoch": 1.905,
+      "grad_norm": 18.66480255126953,
+      "learning_rate": 9.047500000000001e-06,
+      "loss": 3.0494,
+      "step": 381
+    },
+    {
+      "epoch": 1.9100000000000001,
+      "grad_norm": 16.810522079467773,
+      "learning_rate": 9.045e-06,
+      "loss": 3.6914,
+      "step": 382
+    },
+    {
+      "epoch": 1.915,
+      "grad_norm": 19.71428871154785,
+      "learning_rate": 9.0425e-06,
+      "loss": 3.2481,
+      "step": 383
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 14.474647521972656,
+      "learning_rate": 9.040000000000002e-06,
+      "loss": 2.828,
+      "step": 384
+    },
+    {
+      "epoch": 1.925,
+      "grad_norm": 16.400922775268555,
+      "learning_rate": 9.037500000000001e-06,
+      "loss": 3.4751,
+      "step": 385
+    },
+    {
+      "epoch": 1.9300000000000002,
+      "grad_norm": 18.586286544799805,
+      "learning_rate": 9.035e-06,
+      "loss": 3.2448,
+      "step": 386
+    },
+    {
+      "epoch": 1.935,
+      "grad_norm": 18.10755729675293,
+      "learning_rate": 9.0325e-06,
+      "loss": 3.827,
+      "step": 387
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 14.871902465820312,
+      "learning_rate": 9.030000000000002e-06,
+      "loss": 2.9371,
+      "step": 388
+    },
+    {
+      "epoch": 1.9449999999999998,
+      "grad_norm": 16.365745544433594,
+      "learning_rate": 9.027500000000001e-06,
+      "loss": 2.7089,
+      "step": 389
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 16.51506233215332,
+      "learning_rate": 9.025e-06,
+      "loss": 3.0837,
+      "step": 390
+    },
+    {
+      "epoch": 1.955,
+      "grad_norm": 42.9727897644043,
+      "learning_rate": 9.0225e-06,
+      "loss": 4.2563,
+      "step": 391
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 23.49498176574707,
+      "learning_rate": 9.020000000000002e-06,
+      "loss": 3.8626,
+      "step": 392
+    },
+    {
+      "epoch": 1.9649999999999999,
+      "grad_norm": 14.26905632019043,
+      "learning_rate": 9.017500000000001e-06,
+      "loss": 2.6303,
+      "step": 393
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 30.369155883789062,
+      "learning_rate": 9.015000000000001e-06,
+      "loss": 3.1674,
+      "step": 394
+    },
+    {
+      "epoch": 1.975,
+      "grad_norm": 10.437726974487305,
+      "learning_rate": 9.0125e-06,
+      "loss": 3.1162,
+      "step": 395
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 16.37222671508789,
+      "learning_rate": 9.01e-06,
+      "loss": 2.8724,
+      "step": 396
+    },
+    {
+      "epoch": 1.9849999999999999,
+      "grad_norm": 14.053140640258789,
+      "learning_rate": 9.007500000000001e-06,
+      "loss": 2.86,
+      "step": 397
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 15.641444206237793,
+      "learning_rate": 9.005000000000001e-06,
+      "loss": 3.8803,
+      "step": 398
+    },
+    {
+      "epoch": 1.995,
+      "grad_norm": 13.041229248046875,
+      "learning_rate": 9.0025e-06,
+      "loss": 3.2471,
+      "step": 399
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 29.292346954345703,
+      "learning_rate": 9e-06,
+      "loss": 3.8949,
+      "step": 400
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 2.678745746612549,
+      "eval_runtime": 15.5316,
+      "eval_samples_per_second": 51.379,
+      "eval_steps_per_second": 6.438,
+      "step": 400
+    },
+    {
+      "epoch": 2.005,
+      "grad_norm": 14.086385726928711,
+      "learning_rate": 8.997500000000002e-06,
+      "loss": 2.7675,
+      "step": 401
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 22.927457809448242,
+      "learning_rate": 8.995000000000001e-06,
+      "loss": 4.6297,
+      "step": 402
+    },
+    {
+      "epoch": 2.015,
+      "grad_norm": 20.272979736328125,
+      "learning_rate": 8.9925e-06,
+      "loss": 3.2192,
+      "step": 403
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 13.211341857910156,
+      "learning_rate": 8.99e-06,
+      "loss": 3.0902,
+      "step": 404
+    },
+    {
+      "epoch": 2.025,
+      "grad_norm": 10.61660099029541,
+      "learning_rate": 8.987500000000002e-06,
+      "loss": 2.637,
+      "step": 405
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 17.614643096923828,
+      "learning_rate": 8.985000000000001e-06,
+      "loss": 2.9662,
+      "step": 406
+    },
+    {
+      "epoch": 2.035,
+      "grad_norm": 17.1745548248291,
+      "learning_rate": 8.9825e-06,
+      "loss": 3.3596,
+      "step": 407
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 13.972363471984863,
+      "learning_rate": 8.98e-06,
+      "loss": 3.8103,
+      "step": 408
+    },
+    {
+      "epoch": 2.045,
+      "grad_norm": 25.1292781829834,
+      "learning_rate": 8.977500000000002e-06,
+      "loss": 2.9338,
+      "step": 409
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 14.076967239379883,
+      "learning_rate": 8.975e-06,
+      "loss": 2.5592,
+      "step": 410
+    },
+    {
+      "epoch": 2.055,
+      "grad_norm": 12.190113067626953,
+      "learning_rate": 8.972500000000001e-06,
+      "loss": 2.6047,
+      "step": 411
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 19.26861572265625,
+      "learning_rate": 8.97e-06,
+      "loss": 3.4299,
+      "step": 412
+    },
+    {
+      "epoch": 2.065,
+      "grad_norm": 12.272963523864746,
+      "learning_rate": 8.967500000000002e-06,
+      "loss": 2.9492,
+      "step": 413
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 29.705860137939453,
+      "learning_rate": 8.965e-06,
+      "loss": 3.4982,
+      "step": 414
+    },
+    {
+      "epoch": 2.075,
+      "grad_norm": 13.402477264404297,
+      "learning_rate": 8.962500000000001e-06,
+      "loss": 2.9006,
+      "step": 415
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 81.86488342285156,
+      "learning_rate": 8.96e-06,
+      "loss": 2.7961,
+      "step": 416
+    },
+    {
+      "epoch": 2.085,
+      "grad_norm": 15.22852611541748,
+      "learning_rate": 8.957500000000002e-06,
+      "loss": 2.9847,
+      "step": 417
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 20.04500389099121,
+      "learning_rate": 8.955e-06,
+      "loss": 3.5712,
+      "step": 418
+    },
+    {
+      "epoch": 2.095,
+      "grad_norm": 16.6635684967041,
+      "learning_rate": 8.952500000000001e-06,
+      "loss": 2.7727,
+      "step": 419
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 12.978435516357422,
+      "learning_rate": 8.95e-06,
+      "loss": 2.5574,
+      "step": 420
+    },
+    {
+      "epoch": 2.105,
+      "grad_norm": 13.691923141479492,
+      "learning_rate": 8.947500000000002e-06,
+      "loss": 3.0499,
+      "step": 421
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 15.54564094543457,
+      "learning_rate": 8.945e-06,
+      "loss": 2.6433,
+      "step": 422
+    },
+    {
+      "epoch": 2.115,
+      "grad_norm": 11.948934555053711,
+      "learning_rate": 8.942500000000001e-06,
+      "loss": 2.53,
+      "step": 423
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 13.564254760742188,
+      "learning_rate": 8.94e-06,
+      "loss": 3.2759,
+      "step": 424
+    },
+    {
+      "epoch": 2.125,
+      "grad_norm": 11.439271926879883,
+      "learning_rate": 8.9375e-06,
+      "loss": 2.1494,
+      "step": 425
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 27.548776626586914,
+      "learning_rate": 8.935e-06,
+      "loss": 3.4425,
+      "step": 426
+    },
+    {
+      "epoch": 2.135,
+      "grad_norm": 13.43251895904541,
+      "learning_rate": 8.932500000000001e-06,
+      "loss": 3.3022,
+      "step": 427
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 14.409812927246094,
+      "learning_rate": 8.930000000000001e-06,
+      "loss": 2.7803,
+      "step": 428
+    },
+    {
+      "epoch": 2.145,
+      "grad_norm": 14.661177635192871,
+      "learning_rate": 8.9275e-06,
+      "loss": 2.9182,
+      "step": 429
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 12.333430290222168,
+      "learning_rate": 8.925e-06,
+      "loss": 3.2206,
+      "step": 430
+    },
+    {
+      "epoch": 2.155,
+      "grad_norm": 11.485803604125977,
+      "learning_rate": 8.922500000000001e-06,
+      "loss": 2.9742,
+      "step": 431
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 12.47711181640625,
+      "learning_rate": 8.920000000000001e-06,
+      "loss": 2.9922,
+      "step": 432
+    },
+    {
+      "epoch": 2.165,
+      "grad_norm": 15.269696235656738,
+      "learning_rate": 8.9175e-06,
+      "loss": 2.8085,
+      "step": 433
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 11.993585586547852,
+      "learning_rate": 8.915e-06,
+      "loss": 3.2323,
+      "step": 434
+    },
+    {
+      "epoch": 2.175,
+      "grad_norm": 16.55668830871582,
+      "learning_rate": 8.9125e-06,
+      "loss": 3.4861,
+      "step": 435
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 26.93693733215332,
+      "learning_rate": 8.910000000000001e-06,
+      "loss": 2.9854,
+      "step": 436
+    },
+    {
+      "epoch": 2.185,
+      "grad_norm": 11.806517601013184,
+      "learning_rate": 8.9075e-06,
+      "loss": 2.803,
+      "step": 437
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 19.67263412475586,
+      "learning_rate": 8.905e-06,
+      "loss": 3.1462,
+      "step": 438
+    },
+    {
+      "epoch": 2.195,
+      "grad_norm": 15.999809265136719,
+      "learning_rate": 8.9025e-06,
+      "loss": 2.5959,
+      "step": 439
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 17.06156349182129,
+      "learning_rate": 8.900000000000001e-06,
+      "loss": 2.722,
+      "step": 440
+    },
+    {
+      "epoch": 2.205,
+      "grad_norm": 9.786123275756836,
+      "learning_rate": 8.8975e-06,
+      "loss": 2.3772,
+      "step": 441
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 14.28829574584961,
+      "learning_rate": 8.895e-06,
+      "loss": 3.1507,
+      "step": 442
+    },
+    {
+      "epoch": 2.215,
+      "grad_norm": 15.327752113342285,
+      "learning_rate": 8.8925e-06,
+      "loss": 3.1637,
+      "step": 443
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 14.561240196228027,
+      "learning_rate": 8.890000000000001e-06,
+      "loss": 2.8827,
+      "step": 444
+    },
+    {
+      "epoch": 2.225,
+      "grad_norm": 15.798303604125977,
+      "learning_rate": 8.887500000000001e-06,
+      "loss": 3.1251,
+      "step": 445
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 16.426897048950195,
+      "learning_rate": 8.885e-06,
+      "loss": 2.8098,
+      "step": 446
+    },
+    {
+      "epoch": 2.235,
+      "grad_norm": 13.455865859985352,
+      "learning_rate": 8.8825e-06,
+      "loss": 3.3504,
+      "step": 447
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 15.910579681396484,
+      "learning_rate": 8.880000000000001e-06,
+      "loss": 3.4012,
+      "step": 448
+    },
+    {
+      "epoch": 2.245,
+      "grad_norm": 12.965435981750488,
+      "learning_rate": 8.877500000000001e-06,
+      "loss": 3.0774,
+      "step": 449
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 16.696971893310547,
+      "learning_rate": 8.875e-06,
+      "loss": 3.0263,
+      "step": 450
+    },
+    {
+      "epoch": 2.255,
+      "grad_norm": 14.485050201416016,
+      "learning_rate": 8.8725e-06,
+      "loss": 2.7659,
+      "step": 451
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 22.13248634338379,
+      "learning_rate": 8.870000000000001e-06,
+      "loss": 3.4155,
+      "step": 452
+    },
+    {
+      "epoch": 2.265,
+      "grad_norm": 60.460723876953125,
+      "learning_rate": 8.867500000000001e-06,
+      "loss": 3.9045,
+      "step": 453
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 12.62005615234375,
+      "learning_rate": 8.865e-06,
+      "loss": 2.7981,
+      "step": 454
+    },
+    {
+      "epoch": 2.275,
+      "grad_norm": 15.277429580688477,
+      "learning_rate": 8.8625e-06,
+      "loss": 3.3721,
+      "step": 455
+    },
+    {
+      "epoch": 2.2800000000000002,
+      "grad_norm": 14.150629997253418,
+      "learning_rate": 8.860000000000002e-06,
+      "loss": 2.8419,
+      "step": 456
+    },
+    {
+      "epoch": 2.285,
+      "grad_norm": 14.799385070800781,
+      "learning_rate": 8.857500000000001e-06,
+      "loss": 2.7765,
+      "step": 457
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 37.28092575073242,
+      "learning_rate": 8.855e-06,
+      "loss": 3.2277,
+      "step": 458
+    },
+    {
+      "epoch": 2.295,
+      "grad_norm": 19.254230499267578,
+      "learning_rate": 8.8525e-06,
+      "loss": 2.523,
+      "step": 459
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 28.826946258544922,
+      "learning_rate": 8.85e-06,
+      "loss": 3.545,
+      "step": 460
+    },
+    {
+      "epoch": 2.305,
+      "grad_norm": 44.31504821777344,
+      "learning_rate": 8.847500000000001e-06,
+      "loss": 3.1874,
+      "step": 461
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 10.268237113952637,
+      "learning_rate": 8.845000000000001e-06,
+      "loss": 2.4315,
+      "step": 462
+    },
+    {
+      "epoch": 2.315,
+      "grad_norm": 14.920913696289062,
+      "learning_rate": 8.8425e-06,
+      "loss": 3.0079,
+      "step": 463
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 15.898146629333496,
+      "learning_rate": 8.84e-06,
+      "loss": 3.0055,
+      "step": 464
+    },
+    {
+      "epoch": 2.325,
+      "grad_norm": 18.439987182617188,
+      "learning_rate": 8.837500000000001e-06,
+      "loss": 3.1911,
+      "step": 465
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 13.268722534179688,
+      "learning_rate": 8.835000000000001e-06,
+      "loss": 2.5,
+      "step": 466
+    },
+    {
+      "epoch": 2.335,
+      "grad_norm": 13.253763198852539,
+      "learning_rate": 8.8325e-06,
+      "loss": 2.676,
+      "step": 467
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 18.609159469604492,
+      "learning_rate": 8.83e-06,
+      "loss": 2.7842,
+      "step": 468
+    },
+    {
+      "epoch": 2.3449999999999998,
+      "grad_norm": 13.124605178833008,
+      "learning_rate": 8.827500000000001e-06,
+      "loss": 2.6416,
+      "step": 469
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 15.474005699157715,
+      "learning_rate": 8.825000000000001e-06,
+      "loss": 2.7274,
+      "step": 470
+    },
+    {
+      "epoch": 2.355,
+      "grad_norm": 14.222945213317871,
+      "learning_rate": 8.8225e-06,
+      "loss": 2.5155,
+      "step": 471
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 17.53885841369629,
+      "learning_rate": 8.82e-06,
+      "loss": 2.7828,
+      "step": 472
+    },
+    {
+      "epoch": 2.365,
+      "grad_norm": 13.242814064025879,
+      "learning_rate": 8.817500000000002e-06,
+      "loss": 2.7482,
+      "step": 473
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 17.79203224182129,
+      "learning_rate": 8.815e-06,
+      "loss": 2.9659,
+      "step": 474
+    },
+    {
+      "epoch": 2.375,
+      "grad_norm": 15.509217262268066,
+      "learning_rate": 8.8125e-06,
+      "loss": 2.8296,
+      "step": 475
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 15.005112648010254,
+      "learning_rate": 8.81e-06,
+      "loss": 2.8999,
+      "step": 476
+    },
+    {
+      "epoch": 2.385,
+      "grad_norm": 23.550128936767578,
+      "learning_rate": 8.807500000000002e-06,
+      "loss": 2.2158,
+      "step": 477
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 15.104799270629883,
+      "learning_rate": 8.805e-06,
+      "loss": 2.6378,
+      "step": 478
+    },
+    {
+      "epoch": 2.395,
+      "grad_norm": 21.498380661010742,
+      "learning_rate": 8.802500000000001e-06,
+      "loss": 2.3704,
+      "step": 479
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 18.903722763061523,
+      "learning_rate": 8.8e-06,
+      "loss": 2.4074,
+      "step": 480
+    },
+    {
+      "epoch": 2.4050000000000002,
+      "grad_norm": 39.530765533447266,
+      "learning_rate": 8.797500000000002e-06,
+      "loss": 3.7361,
+      "step": 481
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 13.20029354095459,
+      "learning_rate": 8.795e-06,
+      "loss": 2.5068,
+      "step": 482
+    },
+    {
+      "epoch": 2.415,
+      "grad_norm": 14.13471794128418,
+      "learning_rate": 8.792500000000001e-06,
+      "loss": 2.293,
+      "step": 483
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 12.765094757080078,
+      "learning_rate": 8.79e-06,
+      "loss": 2.7894,
+      "step": 484
+    },
+    {
+      "epoch": 2.425,
+      "grad_norm": 13.011454582214355,
+      "learning_rate": 8.787500000000002e-06,
+      "loss": 2.5509,
+      "step": 485
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 14.738533020019531,
+      "learning_rate": 8.785e-06,
+      "loss": 2.5786,
+      "step": 486
+    },
+    {
+      "epoch": 2.435,
+      "grad_norm": 15.5855073928833,
+      "learning_rate": 8.782500000000001e-06,
+      "loss": 2.6184,
+      "step": 487
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 17.30025863647461,
+      "learning_rate": 8.78e-06,
+      "loss": 2.6851,
+      "step": 488
+    },
+    {
+      "epoch": 2.445,
+      "grad_norm": 14.512161254882812,
+      "learning_rate": 8.7775e-06,
+      "loss": 2.2097,
+      "step": 489
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 21.018709182739258,
+      "learning_rate": 8.775e-06,
+      "loss": 2.6585,
+      "step": 490
+    },
+    {
+      "epoch": 2.455,
+      "grad_norm": 22.35019302368164,
+      "learning_rate": 8.772500000000001e-06,
+      "loss": 3.3053,
+      "step": 491
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 13.563121795654297,
+      "learning_rate": 8.77e-06,
+      "loss": 2.4763,
+      "step": 492
+    },
+    {
+      "epoch": 2.465,
+      "grad_norm": 17.950769424438477,
+      "learning_rate": 8.7675e-06,
+      "loss": 2.8311,
+      "step": 493
+    },
+    {
+      "epoch": 2.4699999999999998,
+      "grad_norm": 12.782695770263672,
+      "learning_rate": 8.765e-06,
+      "loss": 2.2945,
+      "step": 494
+    },
+    {
+      "epoch": 2.475,
+      "grad_norm": 19.259750366210938,
+      "learning_rate": 8.762500000000001e-06,
+      "loss": 3.0994,
+      "step": 495
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 16.258710861206055,
+      "learning_rate": 8.76e-06,
+      "loss": 2.84,
+      "step": 496
+    },
+    {
+      "epoch": 2.485,
+      "grad_norm": 13.990570068359375,
+      "learning_rate": 8.7575e-06,
+      "loss": 2.3911,
+      "step": 497
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 14.935282707214355,
+      "learning_rate": 8.755e-06,
+      "loss": 2.8683,
+      "step": 498
+    },
+    {
+      "epoch": 2.495,
+      "grad_norm": 20.888132095336914,
+      "learning_rate": 8.7525e-06,
+      "loss": 2.7134,
+      "step": 499
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1239.13525390625,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 3.0441,
+      "step": 500
+    },
+    {
+      "epoch": 2.505,
+      "grad_norm": 16.687185287475586,
+      "learning_rate": 8.7475e-06,
+      "loss": 2.707,
+      "step": 501
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 13.546875,
+      "learning_rate": 8.745000000000002e-06,
+      "loss": 2.2039,
+      "step": 502
+    },
+    {
+      "epoch": 2.515,
+      "grad_norm": 15.732977867126465,
+      "learning_rate": 8.7425e-06,
+      "loss": 2.2895,
+      "step": 503
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 15.949091911315918,
+      "learning_rate": 8.740000000000001e-06,
+      "loss": 2.3318,
+      "step": 504
+    },
+    {
+      "epoch": 2.525,
+      "grad_norm": 11.378392219543457,
+      "learning_rate": 8.7375e-06,
+      "loss": 2.2168,
+      "step": 505
+    },
+    {
+      "epoch": 2.5300000000000002,
+      "grad_norm": 10.679386138916016,
+      "learning_rate": 8.735000000000002e-06,
+      "loss": 2.1631,
+      "step": 506
+    },
+    {
+      "epoch": 2.535,
+      "grad_norm": 16.291479110717773,
+      "learning_rate": 8.7325e-06,
+      "loss": 2.7465,
+      "step": 507
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 13.583270072937012,
+      "learning_rate": 8.730000000000001e-06,
+      "loss": 2.7675,
+      "step": 508
+    },
+    {
+      "epoch": 2.545,
+      "grad_norm": 15.418150901794434,
+      "learning_rate": 8.7275e-06,
+      "loss": 2.3643,
+      "step": 509
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 11.673973083496094,
+      "learning_rate": 8.725000000000002e-06,
+      "loss": 3.1141,
+      "step": 510
+    },
+    {
+      "epoch": 2.555,
+      "grad_norm": 17.230979919433594,
+      "learning_rate": 8.7225e-06,
+      "loss": 2.8243,
+      "step": 511
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 29.459699630737305,
+      "learning_rate": 8.720000000000001e-06,
+      "loss": 3.2704,
+      "step": 512
+    },
+    {
+      "epoch": 2.565,
+      "grad_norm": 17.000028610229492,
+      "learning_rate": 8.7175e-06,
+      "loss": 2.6371,
+      "step": 513
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 17.9285888671875,
+      "learning_rate": 8.715e-06,
+      "loss": 3.0397,
+      "step": 514
+    },
+    {
+      "epoch": 2.575,
+      "grad_norm": 14.084884643554688,
+      "learning_rate": 8.7125e-06,
+      "loss": 1.9404,
+      "step": 515
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 18.782333374023438,
+      "learning_rate": 8.710000000000001e-06,
+      "loss": 2.3015,
+      "step": 516
+    },
+    {
+      "epoch": 2.585,
+      "grad_norm": 16.925233840942383,
+      "learning_rate": 8.707500000000001e-06,
+      "loss": 2.1995,
+      "step": 517
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 16.437618255615234,
+      "learning_rate": 8.705e-06,
+      "loss": 2.2906,
+      "step": 518
+    },
+    {
+      "epoch": 2.5949999999999998,
+      "grad_norm": 19.559993743896484,
+      "learning_rate": 8.7025e-06,
+      "loss": 2.2777,
+      "step": 519
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 15.771316528320312,
+      "learning_rate": 8.700000000000001e-06,
+      "loss": 2.1775,
+      "step": 520
+    },
+    {
+      "epoch": 2.605,
+      "grad_norm": 16.368480682373047,
+      "learning_rate": 8.697500000000001e-06,
+      "loss": 2.3869,
+      "step": 521
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 16.325727462768555,
+      "learning_rate": 8.695e-06,
+      "loss": 2.0706,
+      "step": 522
+    },
+    {
+      "epoch": 2.615,
+      "grad_norm": 15.276273727416992,
+      "learning_rate": 8.6925e-06,
+      "loss": 2.1412,
+      "step": 523
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 21.763044357299805,
+      "learning_rate": 8.690000000000002e-06,
+      "loss": 2.7799,
+      "step": 524
+    },
+    {
+      "epoch": 2.625,
+      "grad_norm": 17.660381317138672,
+      "learning_rate": 8.687500000000001e-06,
+      "loss": 2.8572,
+      "step": 525
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 14.32105827331543,
+      "learning_rate": 8.685e-06,
+      "loss": 1.9261,
+      "step": 526
+    },
+    {
+      "epoch": 2.635,
+      "grad_norm": 15.422761917114258,
+      "learning_rate": 8.6825e-06,
+      "loss": 2.2744,
+      "step": 527
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 20.47952651977539,
+      "learning_rate": 8.68e-06,
+      "loss": 2.2375,
+      "step": 528
+    },
+    {
+      "epoch": 2.645,
+      "grad_norm": 16.76172637939453,
+      "learning_rate": 8.677500000000001e-06,
+      "loss": 3.0991,
+      "step": 529
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 25.654781341552734,
+      "learning_rate": 8.675e-06,
+      "loss": 2.6155,
+      "step": 530
+    },
+    {
+      "epoch": 2.6550000000000002,
+      "grad_norm": 16.66886329650879,
+      "learning_rate": 8.6725e-06,
+      "loss": 2.1234,
+      "step": 531
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 14.76485538482666,
+      "learning_rate": 8.67e-06,
+      "loss": 2.3906,
+      "step": 532
+    },
+    {
+      "epoch": 2.665,
+      "grad_norm": 13.776233673095703,
+      "learning_rate": 8.667500000000001e-06,
+      "loss": 2.263,
+      "step": 533
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 13.37656021118164,
+      "learning_rate": 8.665000000000001e-06,
+      "loss": 2.4959,
+      "step": 534
+    },
+    {
+      "epoch": 2.675,
+      "grad_norm": 17.845413208007812,
+      "learning_rate": 8.6625e-06,
+      "loss": 2.4379,
+      "step": 535
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 22.331716537475586,
+      "learning_rate": 8.66e-06,
+      "loss": 2.5153,
+      "step": 536
+    },
+    {
+      "epoch": 2.685,
+      "grad_norm": 22.21134376525879,
+      "learning_rate": 8.657500000000001e-06,
+      "loss": 2.3762,
+      "step": 537
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 16.65618133544922,
+      "learning_rate": 8.655000000000001e-06,
+      "loss": 2.7025,
+      "step": 538
+    },
+    {
+      "epoch": 2.695,
+      "grad_norm": 16.927024841308594,
+      "learning_rate": 8.6525e-06,
+      "loss": 2.3765,
+      "step": 539
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 18.975791931152344,
+      "learning_rate": 8.65e-06,
+      "loss": 2.6658,
+      "step": 540
+    },
+    {
+      "epoch": 2.705,
+      "grad_norm": 15.92786979675293,
+      "learning_rate": 8.647500000000001e-06,
+      "loss": 2.2169,
+      "step": 541
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 17.268693923950195,
+      "learning_rate": 8.645000000000001e-06,
+      "loss": 2.8084,
+      "step": 542
+    },
+    {
+      "epoch": 2.715,
+      "grad_norm": 14.349884986877441,
+      "learning_rate": 8.6425e-06,
+      "loss": 2.0297,
+      "step": 543
+    },
+    {
+      "epoch": 2.7199999999999998,
+      "grad_norm": 17.469099044799805,
+      "learning_rate": 8.64e-06,
+      "loss": 2.9701,
+      "step": 544
+    },
+    {
+      "epoch": 2.725,
+      "grad_norm": 21.70449447631836,
+      "learning_rate": 8.637500000000002e-06,
+      "loss": 2.3512,
+      "step": 545
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 13.843066215515137,
+      "learning_rate": 8.635000000000001e-06,
+      "loss": 2.4677,
+      "step": 546
+    },
+    {
+      "epoch": 2.735,
+      "grad_norm": 14.652873992919922,
+      "learning_rate": 8.6325e-06,
+      "loss": 1.8917,
+      "step": 547
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 21.05759620666504,
+      "learning_rate": 8.63e-06,
+      "loss": 1.8814,
+      "step": 548
+    },
+    {
+      "epoch": 2.745,
+      "grad_norm": 13.386977195739746,
+      "learning_rate": 8.627500000000002e-06,
+      "loss": 1.7832,
+      "step": 549
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 14.645966529846191,
+      "learning_rate": 8.625000000000001e-06,
+      "loss": 2.1394,
+      "step": 550
+    },
+    {
+      "epoch": 2.755,
+      "grad_norm": 26.992563247680664,
+      "learning_rate": 8.622500000000001e-06,
+      "loss": 2.0789,
+      "step": 551
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 15.615961074829102,
+      "learning_rate": 8.62e-06,
+      "loss": 2.4726,
+      "step": 552
+    },
+    {
+      "epoch": 2.765,
+      "grad_norm": 18.06861686706543,
+      "learning_rate": 8.6175e-06,
+      "loss": 1.9526,
+      "step": 553
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 15.376697540283203,
+      "learning_rate": 8.615000000000001e-06,
+      "loss": 2.1394,
+      "step": 554
+    },
+    {
+      "epoch": 2.775,
+      "grad_norm": 16.77503776550293,
+      "learning_rate": 8.612500000000001e-06,
+      "loss": 2.0945,
+      "step": 555
+    },
+    {
+      "epoch": 2.7800000000000002,
+      "grad_norm": 18.850114822387695,
+      "learning_rate": 8.61e-06,
+      "loss": 2.3451,
+      "step": 556
+    },
+    {
+      "epoch": 2.785,
+      "grad_norm": 25.581178665161133,
+      "learning_rate": 8.6075e-06,
+      "loss": 2.1528,
+      "step": 557
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 16.494306564331055,
+      "learning_rate": 8.605000000000001e-06,
+      "loss": 2.5561,
+      "step": 558
+    },
+    {
+      "epoch": 2.795,
+      "grad_norm": 19.93292999267578,
+      "learning_rate": 8.602500000000001e-06,
+      "loss": 1.8065,
+      "step": 559
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 15.045790672302246,
+      "learning_rate": 8.6e-06,
+      "loss": 1.9312,
+      "step": 560
+    },
+    {
+      "epoch": 2.805,
+      "grad_norm": 15.675711631774902,
+      "learning_rate": 8.5975e-06,
+      "loss": 2.2907,
+      "step": 561
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 22.94296646118164,
+      "learning_rate": 8.595000000000002e-06,
+      "loss": 2.6787,
+      "step": 562
+    },
+    {
+      "epoch": 2.815,
+      "grad_norm": 19.185422897338867,
+      "learning_rate": 8.5925e-06,
+      "loss": 1.9245,
+      "step": 563
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 62.77212142944336,
+      "learning_rate": 8.59e-06,
+      "loss": 3.3297,
+      "step": 564
+    },
+    {
+      "epoch": 2.825,
+      "grad_norm": 15.820260047912598,
+      "learning_rate": 8.5875e-06,
+      "loss": 2.0025,
+      "step": 565
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 13.180363655090332,
+      "learning_rate": 8.585000000000002e-06,
+      "loss": 2.358,
+      "step": 566
+    },
+    {
+      "epoch": 2.835,
+      "grad_norm": 13.250864028930664,
+      "learning_rate": 8.5825e-06,
+      "loss": 2.3346,
+      "step": 567
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 15.402334213256836,
+      "learning_rate": 8.580000000000001e-06,
+      "loss": 2.1041,
+      "step": 568
+    },
+    {
+      "epoch": 2.8449999999999998,
+      "grad_norm": 27.700410842895508,
+      "learning_rate": 8.5775e-06,
+      "loss": 1.9377,
+      "step": 569
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 14.532342910766602,
+      "learning_rate": 8.575000000000002e-06,
+      "loss": 2.1867,
+      "step": 570
+    },
+    {
+      "epoch": 2.855,
+      "grad_norm": 14.994359970092773,
+      "learning_rate": 8.5725e-06,
+      "loss": 1.3965,
+      "step": 571
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 102.76524353027344,
+      "learning_rate": 8.570000000000001e-06,
+      "loss": 2.881,
+      "step": 572
+    },
+    {
+      "epoch": 2.865,
+      "grad_norm": 17.63866424560547,
+      "learning_rate": 8.5675e-06,
+      "loss": 1.5081,
+      "step": 573
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 14.247568130493164,
+      "learning_rate": 8.565000000000002e-06,
+      "loss": 2.1059,
+      "step": 574
+    },
+    {
+      "epoch": 2.875,
+      "grad_norm": 18.873178482055664,
+      "learning_rate": 8.5625e-06,
+      "loss": 1.7967,
+      "step": 575
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 14.864493370056152,
+      "learning_rate": 8.560000000000001e-06,
+      "loss": 2.6559,
+      "step": 576
+    },
+    {
+      "epoch": 2.885,
+      "grad_norm": 14.47778606414795,
+      "learning_rate": 8.5575e-06,
+      "loss": 2.0776,
+      "step": 577
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 21.14068031311035,
+      "learning_rate": 8.555e-06,
+      "loss": 1.6381,
+      "step": 578
+    },
+    {
+      "epoch": 2.895,
+      "grad_norm": 20.787967681884766,
+      "learning_rate": 8.5525e-06,
+      "loss": 2.0916,
+      "step": 579
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 14.042333602905273,
+      "learning_rate": 8.550000000000001e-06,
+      "loss": 1.7869,
+      "step": 580
+    },
+    {
+      "epoch": 2.9050000000000002,
+      "grad_norm": 15.163957595825195,
+      "learning_rate": 8.5475e-06,
+      "loss": 2.1278,
+      "step": 581
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 23.296714782714844,
+      "learning_rate": 8.545e-06,
+      "loss": 2.4937,
+      "step": 582
+    },
+    {
+      "epoch": 2.915,
+      "grad_norm": 18.689794540405273,
+      "learning_rate": 8.5425e-06,
+      "loss": 1.8953,
+      "step": 583
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 17.24251365661621,
+      "learning_rate": 8.540000000000001e-06,
+      "loss": 1.7431,
+      "step": 584
+    },
+    {
+      "epoch": 2.925,
+      "grad_norm": 16.060791015625,
+      "learning_rate": 8.5375e-06,
+      "loss": 1.4524,
+      "step": 585
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 40.53474807739258,
+      "learning_rate": 8.535e-06,
+      "loss": 1.6281,
+      "step": 586
+    },
+    {
+      "epoch": 2.935,
+      "grad_norm": 16.636444091796875,
+      "learning_rate": 8.5325e-06,
+      "loss": 1.6733,
+      "step": 587
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 18.471477508544922,
+      "learning_rate": 8.530000000000001e-06,
+      "loss": 1.9083,
+      "step": 588
+    },
+    {
+      "epoch": 2.945,
+      "grad_norm": 16.971132278442383,
+      "learning_rate": 8.527500000000001e-06,
+      "loss": 1.6334,
+      "step": 589
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 15.937500953674316,
+      "learning_rate": 8.525e-06,
+      "loss": 2.0047,
+      "step": 590
+    },
+    {
+      "epoch": 2.955,
+      "grad_norm": 18.388408660888672,
+      "learning_rate": 8.5225e-06,
+      "loss": 1.4465,
+      "step": 591
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 15.089503288269043,
+      "learning_rate": 8.52e-06,
+      "loss": 2.0945,
+      "step": 592
+    },
+    {
+      "epoch": 2.965,
+      "grad_norm": 20.312637329101562,
+      "learning_rate": 8.517500000000001e-06,
+      "loss": 2.1393,
+      "step": 593
+    },
+    {
+      "epoch": 2.9699999999999998,
+      "grad_norm": 63.03879165649414,
+      "learning_rate": 8.515e-06,
+      "loss": 1.8488,
+      "step": 594
+    },
+    {
+      "epoch": 2.975,
+      "grad_norm": 16.02383804321289,
+      "learning_rate": 8.5125e-06,
+      "loss": 2.0682,
+      "step": 595
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 18.252872467041016,
+      "learning_rate": 8.51e-06,
+      "loss": 2.1854,
+      "step": 596
+    },
+    {
+      "epoch": 2.985,
+      "grad_norm": 17.822114944458008,
+      "learning_rate": 8.507500000000001e-06,
+      "loss": 2.3448,
+      "step": 597
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 15.47060775756836,
+      "learning_rate": 8.505e-06,
+      "loss": 1.7238,
+      "step": 598
+    },
+    {
+      "epoch": 2.995,
+      "grad_norm": 17.763710021972656,
+      "learning_rate": 8.5025e-06,
+      "loss": 1.5974,
+      "step": 599
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 20.52495765686035,
+      "learning_rate": 8.5e-06,
+      "loss": 1.611,
+      "step": 600
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 1.1276317834854126,
+      "eval_runtime": 15.0908,
+      "eval_samples_per_second": 52.88,
+      "eval_steps_per_second": 6.627,
+      "step": 600
+    },
+    {
+      "epoch": 3.005,
+      "grad_norm": 16.134418487548828,
+      "learning_rate": 8.497500000000001e-06,
+      "loss": 1.7627,
+      "step": 601
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 66.23824310302734,
+      "learning_rate": 8.495e-06,
+      "loss": 2.1128,
+      "step": 602
+    },
+    {
+      "epoch": 3.015,
+      "grad_norm": 23.679622650146484,
+      "learning_rate": 8.4925e-06,
+      "loss": 1.8941,
+      "step": 603
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 12.122626304626465,
+      "learning_rate": 8.49e-06,
+      "loss": 1.9734,
+      "step": 604
+    },
+    {
+      "epoch": 3.025,
+      "grad_norm": 36.0194091796875,
+      "learning_rate": 8.487500000000001e-06,
+      "loss": 1.8102,
+      "step": 605
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 15.40415096282959,
+      "learning_rate": 8.485000000000001e-06,
+      "loss": 2.0156,
+      "step": 606
+    },
+    {
+      "epoch": 3.035,
+      "grad_norm": 24.995080947875977,
+      "learning_rate": 8.4825e-06,
+      "loss": 1.6933,
+      "step": 607
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 20.26504898071289,
+      "learning_rate": 8.48e-06,
+      "loss": 1.7924,
+      "step": 608
+    },
+    {
+      "epoch": 3.045,
+      "grad_norm": 20.2662296295166,
+      "learning_rate": 8.477500000000001e-06,
+      "loss": 1.9866,
+      "step": 609
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 17.551681518554688,
+      "learning_rate": 8.475000000000001e-06,
+      "loss": 1.5987,
+      "step": 610
+    },
+    {
+      "epoch": 3.055,
+      "grad_norm": 22.877954483032227,
+      "learning_rate": 8.4725e-06,
+      "loss": 1.6017,
+      "step": 611
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 17.723478317260742,
+      "learning_rate": 8.47e-06,
+      "loss": 1.3522,
+      "step": 612
+    },
+    {
+      "epoch": 3.065,
+      "grad_norm": 22.660205841064453,
+      "learning_rate": 8.467500000000002e-06,
+      "loss": 1.7278,
+      "step": 613
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 16.1903018951416,
+      "learning_rate": 8.465000000000001e-06,
+      "loss": 1.9765,
+      "step": 614
+    },
+    {
+      "epoch": 3.075,
+      "grad_norm": 16.37009048461914,
+      "learning_rate": 8.4625e-06,
+      "loss": 2.0244,
+      "step": 615
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 14.325566291809082,
+      "learning_rate": 8.46e-06,
+      "loss": 1.2033,
+      "step": 616
+    },
+    {
+      "epoch": 3.085,
+      "grad_norm": 12.867656707763672,
+      "learning_rate": 8.4575e-06,
+      "loss": 1.6523,
+      "step": 617
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 18.460519790649414,
+      "learning_rate": 8.455000000000001e-06,
+      "loss": 1.5439,
+      "step": 618
+    },
+    {
+      "epoch": 3.095,
+      "grad_norm": 16.935596466064453,
+      "learning_rate": 8.4525e-06,
+      "loss": 1.876,
+      "step": 619
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 17.250337600708008,
+      "learning_rate": 8.45e-06,
+      "loss": 1.4761,
+      "step": 620
+    },
+    {
+      "epoch": 3.105,
+      "grad_norm": 15.343099594116211,
+      "learning_rate": 8.4475e-06,
+      "loss": 1.2432,
+      "step": 621
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 13.481271743774414,
+      "learning_rate": 8.445000000000001e-06,
+      "loss": 1.4933,
+      "step": 622
+    },
+    {
+      "epoch": 3.115,
+      "grad_norm": 15.92331314086914,
+      "learning_rate": 8.442500000000001e-06,
+      "loss": 1.4761,
+      "step": 623
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 15.47143840789795,
+      "learning_rate": 8.44e-06,
+      "loss": 1.7706,
+      "step": 624
+    },
+    {
+      "epoch": 3.125,
+      "grad_norm": 15.646211624145508,
+      "learning_rate": 8.4375e-06,
+      "loss": 1.5945,
+      "step": 625
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 16.18316650390625,
+      "learning_rate": 8.435000000000001e-06,
+      "loss": 1.5345,
+      "step": 626
+    },
+    {
+      "epoch": 3.135,
+      "grad_norm": 17.53881072998047,
+      "learning_rate": 8.432500000000001e-06,
+      "loss": 1.9626,
+      "step": 627
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 16.5322322845459,
+      "learning_rate": 8.43e-06,
+      "loss": 1.3358,
+      "step": 628
+    },
+    {
+      "epoch": 3.145,
+      "grad_norm": 19.785900115966797,
+      "learning_rate": 8.4275e-06,
+      "loss": 1.6812,
+      "step": 629
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 24.368757247924805,
+      "learning_rate": 8.425000000000001e-06,
+      "loss": 2.098,
+      "step": 630
+    },
+    {
+      "epoch": 3.155,
+      "grad_norm": 17.398273468017578,
+      "learning_rate": 8.422500000000001e-06,
+      "loss": 1.6573,
+      "step": 631
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 23.3609561920166,
+      "learning_rate": 8.42e-06,
+      "loss": 2.5114,
+      "step": 632
+    },
+    {
+      "epoch": 3.165,
+      "grad_norm": 18.4223575592041,
+      "learning_rate": 8.4175e-06,
+      "loss": 1.3063,
+      "step": 633
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 15.635534286499023,
+      "learning_rate": 8.415000000000002e-06,
+      "loss": 1.4639,
+      "step": 634
+    },
+    {
+      "epoch": 3.175,
+      "grad_norm": 12.268280982971191,
+      "learning_rate": 8.412500000000001e-06,
+      "loss": 1.8504,
+      "step": 635
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 11.06991195678711,
+      "learning_rate": 8.41e-06,
+      "loss": 1.2995,
+      "step": 636
+    },
+    {
+      "epoch": 3.185,
+      "grad_norm": 16.566307067871094,
+      "learning_rate": 8.4075e-06,
+      "loss": 1.27,
+      "step": 637
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 12.646732330322266,
+      "learning_rate": 8.405000000000002e-06,
+      "loss": 2.0293,
+      "step": 638
+    },
+    {
+      "epoch": 3.195,
+      "grad_norm": 12.561951637268066,
+      "learning_rate": 8.402500000000001e-06,
+      "loss": 1.2542,
+      "step": 639
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 16.466819763183594,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 1.336,
+      "step": 640
+    },
+    {
+      "epoch": 3.205,
+      "grad_norm": 15.50632381439209,
+      "learning_rate": 8.3975e-06,
+      "loss": 1.4517,
+      "step": 641
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 28.440776824951172,
+      "learning_rate": 8.395e-06,
+      "loss": 1.244,
+      "step": 642
+    },
+    {
+      "epoch": 3.215,
+      "grad_norm": 12.339152336120605,
+      "learning_rate": 8.392500000000001e-06,
+      "loss": 1.5973,
+      "step": 643
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 21.795631408691406,
+      "learning_rate": 8.390000000000001e-06,
+      "loss": 1.4586,
+      "step": 644
+    },
+    {
+      "epoch": 3.225,
+      "grad_norm": 14.898630142211914,
+      "learning_rate": 8.3875e-06,
+      "loss": 1.3226,
+      "step": 645
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 13.366578102111816,
+      "learning_rate": 8.385e-06,
+      "loss": 1.2608,
+      "step": 646
+    },
+    {
+      "epoch": 3.235,
+      "grad_norm": 13.639751434326172,
+      "learning_rate": 8.382500000000001e-06,
+      "loss": 1.3189,
+      "step": 647
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 19.03993034362793,
+      "learning_rate": 8.380000000000001e-06,
+      "loss": 1.2735,
+      "step": 648
+    },
+    {
+      "epoch": 3.245,
+      "grad_norm": 13.897819519042969,
+      "learning_rate": 8.3775e-06,
+      "loss": 1.2878,
+      "step": 649
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 25.421348571777344,
+      "learning_rate": 8.375e-06,
+      "loss": 2.3457,
+      "step": 650
+    },
+    {
+      "epoch": 3.255,
+      "grad_norm": 18.525129318237305,
+      "learning_rate": 8.372500000000002e-06,
+      "loss": 1.4613,
+      "step": 651
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 11.685342788696289,
+      "learning_rate": 8.370000000000001e-06,
+      "loss": 1.2192,
+      "step": 652
+    },
+    {
+      "epoch": 3.265,
+      "grad_norm": 16.042566299438477,
+      "learning_rate": 8.3675e-06,
+      "loss": 1.3736,
+      "step": 653
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 12.06123161315918,
+      "learning_rate": 8.365e-06,
+      "loss": 1.0659,
+      "step": 654
+    },
+    {
+      "epoch": 3.275,
+      "grad_norm": 13.24004077911377,
+      "learning_rate": 8.362500000000002e-06,
+      "loss": 1.0241,
+      "step": 655
+    },
+    {
+      "epoch": 3.2800000000000002,
+      "grad_norm": 13.70656681060791,
+      "learning_rate": 8.36e-06,
+      "loss": 1.1281,
+      "step": 656
+    },
+    {
+      "epoch": 3.285,
+      "grad_norm": 16.92243194580078,
+      "learning_rate": 8.357500000000001e-06,
+      "loss": 1.1999,
+      "step": 657
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 11.360891342163086,
+      "learning_rate": 8.355e-06,
+      "loss": 1.1346,
+      "step": 658
+    },
+    {
+      "epoch": 3.295,
+      "grad_norm": 9.794748306274414,
+      "learning_rate": 8.352500000000002e-06,
+      "loss": 1.2192,
+      "step": 659
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 12.57457160949707,
+      "learning_rate": 8.35e-06,
+      "loss": 1.1201,
+      "step": 660
+    },
+    {
+      "epoch": 3.305,
+      "grad_norm": 16.680103302001953,
+      "learning_rate": 8.347500000000001e-06,
+      "loss": 1.2061,
+      "step": 661
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 21.180166244506836,
+      "learning_rate": 8.345e-06,
+      "loss": 1.2447,
+      "step": 662
+    },
+    {
+      "epoch": 3.315,
+      "grad_norm": 12.157773971557617,
+      "learning_rate": 8.342500000000002e-06,
+      "loss": 1.1662,
+      "step": 663
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 12.90079402923584,
+      "learning_rate": 8.34e-06,
+      "loss": 1.2216,
+      "step": 664
+    },
+    {
+      "epoch": 3.325,
+      "grad_norm": 13.879995346069336,
+      "learning_rate": 8.337500000000001e-06,
+      "loss": 1.3771,
+      "step": 665
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 15.186230659484863,
+      "learning_rate": 8.335e-06,
+      "loss": 1.4604,
+      "step": 666
+    },
+    {
+      "epoch": 3.335,
+      "grad_norm": 12.18576431274414,
+      "learning_rate": 8.332500000000002e-06,
+      "loss": 1.4353,
+      "step": 667
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 24.858795166015625,
+      "learning_rate": 8.33e-06,
+      "loss": 1.579,
+      "step": 668
+    },
+    {
+      "epoch": 3.3449999999999998,
+      "grad_norm": 9.97230052947998,
+      "learning_rate": 8.327500000000001e-06,
+      "loss": 1.4105,
+      "step": 669
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 67.21975708007812,
+      "learning_rate": 8.325e-06,
+      "loss": 2.1563,
+      "step": 670
+    },
+    {
+      "epoch": 3.355,
+      "grad_norm": 10.43264389038086,
+      "learning_rate": 8.3225e-06,
+      "loss": 0.7232,
+      "step": 671
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 13.17708683013916,
+      "learning_rate": 8.32e-06,
+      "loss": 1.2024,
+      "step": 672
+    },
+    {
+      "epoch": 3.365,
+      "grad_norm": 11.955435752868652,
+      "learning_rate": 8.317500000000001e-06,
+      "loss": 1.3954,
+      "step": 673
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 16.89090347290039,
+      "learning_rate": 8.315000000000001e-06,
+      "loss": 2.0402,
+      "step": 674
+    },
+    {
+      "epoch": 3.375,
+      "grad_norm": 11.06239128112793,
+      "learning_rate": 8.3125e-06,
+      "loss": 0.8561,
+      "step": 675
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 14.577070236206055,
+      "learning_rate": 8.31e-06,
+      "loss": 1.2056,
+      "step": 676
+    },
+    {
+      "epoch": 3.385,
+      "grad_norm": 14.767572402954102,
+      "learning_rate": 8.307500000000001e-06,
+      "loss": 1.4661,
+      "step": 677
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 13.088774681091309,
+      "learning_rate": 8.305000000000001e-06,
+      "loss": 1.5603,
+      "step": 678
+    },
+    {
+      "epoch": 3.395,
+      "grad_norm": 9.9073486328125,
+      "learning_rate": 8.3025e-06,
+      "loss": 0.8386,
+      "step": 679
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 13.399995803833008,
+      "learning_rate": 8.3e-06,
+      "loss": 0.8437,
+      "step": 680
+    },
+    {
+      "epoch": 3.4050000000000002,
+      "grad_norm": 14.606317520141602,
+      "learning_rate": 8.2975e-06,
+      "loss": 1.0946,
+      "step": 681
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 15.816943168640137,
+      "learning_rate": 8.295000000000001e-06,
+      "loss": 1.2679,
+      "step": 682
+    },
+    {
+      "epoch": 3.415,
+      "grad_norm": 12.246903419494629,
+      "learning_rate": 8.2925e-06,
+      "loss": 1.4036,
+      "step": 683
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 12.374079704284668,
+      "learning_rate": 8.29e-06,
+      "loss": 1.2287,
+      "step": 684
+    },
+    {
+      "epoch": 3.425,
+      "grad_norm": 15.983515739440918,
+      "learning_rate": 8.2875e-06,
+      "loss": 1.6664,
+      "step": 685
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 10.002947807312012,
+      "learning_rate": 8.285000000000001e-06,
+      "loss": 0.8375,
+      "step": 686
+    },
+    {
+      "epoch": 3.435,
+      "grad_norm": 11.499504089355469,
+      "learning_rate": 8.2825e-06,
+      "loss": 1.7553,
+      "step": 687
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 10.58079719543457,
+      "learning_rate": 8.28e-06,
+      "loss": 0.7117,
+      "step": 688
+    },
+    {
+      "epoch": 3.445,
+      "grad_norm": 234.3501739501953,
+      "learning_rate": 8.2775e-06,
+      "loss": 2.6228,
+      "step": 689
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 15.132519721984863,
+      "learning_rate": 8.275000000000001e-06,
+      "loss": 1.1256,
+      "step": 690
+    },
+    {
+      "epoch": 3.455,
+      "grad_norm": 17.92886734008789,
+      "learning_rate": 8.2725e-06,
+      "loss": 1.2283,
+      "step": 691
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 9.650834083557129,
+      "learning_rate": 8.27e-06,
+      "loss": 1.3112,
+      "step": 692
+    },
+    {
+      "epoch": 3.465,
+      "grad_norm": 11.07042407989502,
+      "learning_rate": 8.2675e-06,
+      "loss": 1.2154,
+      "step": 693
+    },
+    {
+      "epoch": 3.4699999999999998,
+      "grad_norm": 14.875812530517578,
+      "learning_rate": 8.265000000000001e-06,
+      "loss": 1.375,
+      "step": 694
+    },
+    {
+      "epoch": 3.475,
+      "grad_norm": 16.761564254760742,
+      "learning_rate": 8.262500000000001e-06,
+      "loss": 1.5446,
+      "step": 695
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 11.524791717529297,
+      "learning_rate": 8.26e-06,
+      "loss": 1.2038,
+      "step": 696
+    },
+    {
+      "epoch": 3.485,
+      "grad_norm": 10.475284576416016,
+      "learning_rate": 8.2575e-06,
+      "loss": 0.8009,
+      "step": 697
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 11.565507888793945,
+      "learning_rate": 8.255000000000001e-06,
+      "loss": 1.7584,
+      "step": 698
+    },
+    {
+      "epoch": 3.495,
+      "grad_norm": 20.812192916870117,
+      "learning_rate": 8.252500000000001e-06,
+      "loss": 1.7469,
+      "step": 699
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 10.300994873046875,
+      "learning_rate": 8.25e-06,
+      "loss": 1.1747,
+      "step": 700
+    },
+    {
+      "epoch": 3.505,
+      "grad_norm": 13.302786827087402,
+      "learning_rate": 8.2475e-06,
+      "loss": 1.4933,
+      "step": 701
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 12.120156288146973,
+      "learning_rate": 8.245000000000002e-06,
+      "loss": 1.6572,
+      "step": 702
+    },
+    {
+      "epoch": 3.515,
+      "grad_norm": 10.178818702697754,
+      "learning_rate": 8.242500000000001e-06,
+      "loss": 1.0862,
+      "step": 703
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 136.6439971923828,
+      "learning_rate": 8.24e-06,
+      "loss": 2.6234,
+      "step": 704
+    },
+    {
+      "epoch": 3.525,
+      "grad_norm": 16.453821182250977,
+      "learning_rate": 8.2375e-06,
+      "loss": 1.6321,
+      "step": 705
+    },
+    {
+      "epoch": 3.5300000000000002,
+      "grad_norm": 19.14729881286621,
+      "learning_rate": 8.235e-06,
+      "loss": 1.2328,
+      "step": 706
+    },
+    {
+      "epoch": 3.535,
+      "grad_norm": 24.70721435546875,
+      "learning_rate": 8.232500000000001e-06,
+      "loss": 1.937,
+      "step": 707
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 38.77723693847656,
+      "learning_rate": 8.23e-06,
+      "loss": 1.3624,
+      "step": 708
+    },
+    {
+      "epoch": 3.545,
+      "grad_norm": 14.564453125,
+      "learning_rate": 8.2275e-06,
+      "loss": 1.1122,
+      "step": 709
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 10.613748550415039,
+      "learning_rate": 8.225e-06,
+      "loss": 1.3061,
+      "step": 710
+    },
+    {
+      "epoch": 3.555,
+      "grad_norm": 9.093697547912598,
+      "learning_rate": 8.222500000000001e-06,
+      "loss": 1.2645,
+      "step": 711
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 48.54293441772461,
+      "learning_rate": 8.220000000000001e-06,
+      "loss": 1.4538,
+      "step": 712
+    },
+    {
+      "epoch": 3.565,
+      "grad_norm": 12.92207145690918,
+      "learning_rate": 8.2175e-06,
+      "loss": 1.1249,
+      "step": 713
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 27.02292823791504,
+      "learning_rate": 8.215e-06,
+      "loss": 1.4033,
+      "step": 714
+    },
+    {
+      "epoch": 3.575,
+      "grad_norm": 13.046821594238281,
+      "learning_rate": 8.212500000000001e-06,
+      "loss": 1.029,
+      "step": 715
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 14.73383617401123,
+      "learning_rate": 8.210000000000001e-06,
+      "loss": 0.8824,
+      "step": 716
+    },
+    {
+      "epoch": 3.585,
+      "grad_norm": 9.930327415466309,
+      "learning_rate": 8.2075e-06,
+      "loss": 1.1423,
+      "step": 717
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 10.094903945922852,
+      "learning_rate": 8.205e-06,
+      "loss": 1.1584,
+      "step": 718
+    },
+    {
+      "epoch": 3.5949999999999998,
+      "grad_norm": 14.70242977142334,
+      "learning_rate": 8.202500000000002e-06,
+      "loss": 1.0576,
+      "step": 719
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 10.626029968261719,
+      "learning_rate": 8.2e-06,
+      "loss": 1.0774,
+      "step": 720
+    },
+    {
+      "epoch": 3.605,
+      "grad_norm": 9.646835327148438,
+      "learning_rate": 8.1975e-06,
+      "loss": 0.8959,
+      "step": 721
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 14.803915023803711,
+      "learning_rate": 8.195e-06,
+      "loss": 0.9732,
+      "step": 722
+    },
+    {
+      "epoch": 3.615,
+      "grad_norm": 14.44840145111084,
+      "learning_rate": 8.192500000000002e-06,
+      "loss": 1.0747,
+      "step": 723
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 17.86935043334961,
+      "learning_rate": 8.19e-06,
+      "loss": 1.4302,
+      "step": 724
+    },
+    {
+      "epoch": 3.625,
+      "grad_norm": 10.382486343383789,
+      "learning_rate": 8.1875e-06,
+      "loss": 1.1409,
+      "step": 725
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 9.77989387512207,
+      "learning_rate": 8.185e-06,
+      "loss": 0.9033,
+      "step": 726
+    },
+    {
+      "epoch": 3.635,
+      "grad_norm": 12.004661560058594,
+      "learning_rate": 8.182500000000002e-06,
+      "loss": 0.9718,
+      "step": 727
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 11.445658683776855,
+      "learning_rate": 8.18e-06,
+      "loss": 1.5589,
+      "step": 728
+    },
+    {
+      "epoch": 3.645,
+      "grad_norm": 8.887510299682617,
+      "learning_rate": 8.177500000000001e-06,
+      "loss": 0.9292,
+      "step": 729
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 9.02359676361084,
+      "learning_rate": 8.175e-06,
+      "loss": 1.1176,
+      "step": 730
+    },
+    {
+      "epoch": 3.6550000000000002,
+      "grad_norm": 10.504866600036621,
+      "learning_rate": 8.172500000000002e-06,
+      "loss": 1.4607,
+      "step": 731
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 10.499091148376465,
+      "learning_rate": 8.17e-06,
+      "loss": 0.734,
+      "step": 732
+    },
+    {
+      "epoch": 3.665,
+      "grad_norm": 9.88882827758789,
+      "learning_rate": 8.167500000000001e-06,
+      "loss": 1.0156,
+      "step": 733
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 17.604394912719727,
+      "learning_rate": 8.165e-06,
+      "loss": 1.2576,
+      "step": 734
+    },
+    {
+      "epoch": 3.675,
+      "grad_norm": 8.06059741973877,
+      "learning_rate": 8.1625e-06,
+      "loss": 0.8649,
+      "step": 735
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 12.10250186920166,
+      "learning_rate": 8.16e-06,
+      "loss": 1.1895,
+      "step": 736
+    },
+    {
+      "epoch": 3.685,
+      "grad_norm": 8.571789741516113,
+      "learning_rate": 8.157500000000001e-06,
+      "loss": 1.0085,
+      "step": 737
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 8.245233535766602,
+      "learning_rate": 8.155e-06,
+      "loss": 1.0284,
+      "step": 738
+    },
+    {
+      "epoch": 3.695,
+      "grad_norm": 18.235193252563477,
+      "learning_rate": 8.1525e-06,
+      "loss": 1.3031,
+      "step": 739
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 36.59899139404297,
+      "learning_rate": 8.15e-06,
+      "loss": 1.25,
+      "step": 740
+    },
+    {
+      "epoch": 3.705,
+      "grad_norm": 10.45309066772461,
+      "learning_rate": 8.147500000000001e-06,
+      "loss": 0.7625,
+      "step": 741
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 10.4016695022583,
+      "learning_rate": 8.145e-06,
+      "loss": 1.3311,
+      "step": 742
+    },
+    {
+      "epoch": 3.715,
+      "grad_norm": 8.789263725280762,
+      "learning_rate": 8.1425e-06,
+      "loss": 1.1153,
+      "step": 743
+    },
+    {
+      "epoch": 3.7199999999999998,
+      "grad_norm": 131.5764617919922,
+      "learning_rate": 8.14e-06,
+      "loss": 1.3931,
+      "step": 744
+    },
+    {
+      "epoch": 3.725,
+      "grad_norm": 16.24248695373535,
+      "learning_rate": 8.1375e-06,
+      "loss": 1.2868,
+      "step": 745
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 14.076769828796387,
+      "learning_rate": 8.135000000000001e-06,
+      "loss": 0.9616,
+      "step": 746
+    },
+    {
+      "epoch": 3.735,
+      "grad_norm": 10.77306079864502,
+      "learning_rate": 8.1325e-06,
+      "loss": 1.313,
+      "step": 747
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 9.290863990783691,
+      "learning_rate": 8.13e-06,
+      "loss": 0.907,
+      "step": 748
+    },
+    {
+      "epoch": 3.745,
+      "grad_norm": 14.74361515045166,
+      "learning_rate": 8.1275e-06,
+      "loss": 0.9915,
+      "step": 749
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 9.07702350616455,
+      "learning_rate": 8.125000000000001e-06,
+      "loss": 0.8145,
+      "step": 750
+    },
+    {
+      "epoch": 3.755,
+      "grad_norm": 7.710874557495117,
+      "learning_rate": 8.1225e-06,
+      "loss": 0.7717,
+      "step": 751
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 8.7726469039917,
+      "learning_rate": 8.120000000000002e-06,
+      "loss": 0.8307,
+      "step": 752
+    },
+    {
+      "epoch": 3.765,
+      "grad_norm": 11.597051620483398,
+      "learning_rate": 8.1175e-06,
+      "loss": 1.1448,
+      "step": 753
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 11.522122383117676,
+      "learning_rate": 8.115000000000001e-06,
+      "loss": 0.8955,
+      "step": 754
+    },
+    {
+      "epoch": 3.775,
+      "grad_norm": 8.009984970092773,
+      "learning_rate": 8.1125e-06,
+      "loss": 0.8092,
+      "step": 755
+    },
+    {
+      "epoch": 3.7800000000000002,
+      "grad_norm": 12.173160552978516,
+      "learning_rate": 8.110000000000002e-06,
+      "loss": 1.4352,
+      "step": 756
+    },
+    {
+      "epoch": 3.785,
+      "grad_norm": 12.024731636047363,
+      "learning_rate": 8.1075e-06,
+      "loss": 1.482,
+      "step": 757
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 7.088742256164551,
+      "learning_rate": 8.105000000000001e-06,
+      "loss": 0.8074,
+      "step": 758
+    },
+    {
+      "epoch": 3.795,
+      "grad_norm": 12.898659706115723,
+      "learning_rate": 8.1025e-06,
+      "loss": 1.1197,
+      "step": 759
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 20.777742385864258,
+      "learning_rate": 8.1e-06,
+      "loss": 0.8673,
+      "step": 760
+    },
+    {
+      "epoch": 3.805,
+      "grad_norm": 7.099936485290527,
+      "learning_rate": 8.0975e-06,
+      "loss": 0.6126,
+      "step": 761
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 11.543097496032715,
+      "learning_rate": 8.095000000000001e-06,
+      "loss": 0.8021,
+      "step": 762
+    },
+    {
+      "epoch": 3.815,
+      "grad_norm": 13.031311988830566,
+      "learning_rate": 8.092500000000001e-06,
+      "loss": 0.9424,
+      "step": 763
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 6.6431145668029785,
+      "learning_rate": 8.09e-06,
+      "loss": 1.1256,
+      "step": 764
+    },
+    {
+      "epoch": 3.825,
+      "grad_norm": 16.030258178710938,
+      "learning_rate": 8.0875e-06,
+      "loss": 1.4522,
+      "step": 765
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 11.678950309753418,
+      "learning_rate": 8.085000000000001e-06,
+      "loss": 1.0396,
+      "step": 766
+    },
+    {
+      "epoch": 3.835,
+      "grad_norm": 9.976725578308105,
+      "learning_rate": 8.082500000000001e-06,
+      "loss": 0.9611,
+      "step": 767
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 10.997763633728027,
+      "learning_rate": 8.08e-06,
+      "loss": 0.9145,
+      "step": 768
+    },
+    {
+      "epoch": 3.8449999999999998,
+      "grad_norm": 6.931461334228516,
+      "learning_rate": 8.0775e-06,
+      "loss": 0.9438,
+      "step": 769
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 11.558601379394531,
+      "learning_rate": 8.075000000000001e-06,
+      "loss": 1.0343,
+      "step": 770
+    },
+    {
+      "epoch": 3.855,
+      "grad_norm": 11.458104133605957,
+      "learning_rate": 8.072500000000001e-06,
+      "loss": 0.659,
+      "step": 771
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 11.247264862060547,
+      "learning_rate": 8.07e-06,
+      "loss": 1.2111,
+      "step": 772
+    },
+    {
+      "epoch": 3.865,
+      "grad_norm": 723.4169311523438,
+      "learning_rate": 8.0675e-06,
+      "loss": 7.1958,
+      "step": 773
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 15.80791187286377,
+      "learning_rate": 8.065e-06,
+      "loss": 0.8433,
+      "step": 774
+    },
+    {
+      "epoch": 3.875,
+      "grad_norm": 9.093989372253418,
+      "learning_rate": 8.062500000000001e-06,
+      "loss": 0.8974,
+      "step": 775
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 9.159106254577637,
+      "learning_rate": 8.06e-06,
+      "loss": 0.7953,
+      "step": 776
+    },
+    {
+      "epoch": 3.885,
+      "grad_norm": 15.91211223602295,
+      "learning_rate": 8.0575e-06,
+      "loss": 1.06,
+      "step": 777
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 8.19650936126709,
+      "learning_rate": 8.055e-06,
+      "loss": 0.7057,
+      "step": 778
+    },
+    {
+      "epoch": 3.895,
+      "grad_norm": 8.441150665283203,
+      "learning_rate": 8.052500000000001e-06,
+      "loss": 1.2897,
+      "step": 779
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 11.974963188171387,
+      "learning_rate": 8.050000000000001e-06,
+      "loss": 0.7241,
+      "step": 780
+    },
+    {
+      "epoch": 3.9050000000000002,
+      "grad_norm": 9.966224670410156,
+      "learning_rate": 8.0475e-06,
+      "loss": 1.3394,
+      "step": 781
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 7.388568878173828,
+      "learning_rate": 8.045e-06,
+      "loss": 0.5919,
+      "step": 782
+    },
+    {
+      "epoch": 3.915,
+      "grad_norm": 10.694477081298828,
+      "learning_rate": 8.042500000000001e-06,
+      "loss": 1.1477,
+      "step": 783
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 22.117881774902344,
+      "learning_rate": 8.040000000000001e-06,
+      "loss": 0.7965,
+      "step": 784
+    },
+    {
+      "epoch": 3.925,
+      "grad_norm": 10.859774589538574,
+      "learning_rate": 8.0375e-06,
+      "loss": 1.3497,
+      "step": 785
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 7.456951141357422,
+      "learning_rate": 8.035e-06,
+      "loss": 0.6957,
+      "step": 786
+    },
+    {
+      "epoch": 3.935,
+      "grad_norm": 7.884845733642578,
+      "learning_rate": 8.032500000000001e-06,
+      "loss": 0.6351,
+      "step": 787
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 8.390623092651367,
+      "learning_rate": 8.030000000000001e-06,
+      "loss": 0.7793,
+      "step": 788
+    },
+    {
+      "epoch": 3.945,
+      "grad_norm": 25.423852920532227,
+      "learning_rate": 8.0275e-06,
+      "loss": 0.9546,
+      "step": 789
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 12.56760311126709,
+      "learning_rate": 8.025e-06,
+      "loss": 0.9461,
+      "step": 790
+    },
+    {
+      "epoch": 3.955,
+      "grad_norm": 71.79685974121094,
+      "learning_rate": 8.022500000000002e-06,
+      "loss": 1.8767,
+      "step": 791
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 12.140376091003418,
+      "learning_rate": 8.020000000000001e-06,
+      "loss": 1.1709,
+      "step": 792
+    },
+    {
+      "epoch": 3.965,
+      "grad_norm": 11.511007308959961,
+      "learning_rate": 8.0175e-06,
+      "loss": 1.0747,
+      "step": 793
+    },
+    {
+      "epoch": 3.9699999999999998,
+      "grad_norm": 8.522571563720703,
+      "learning_rate": 8.015e-06,
+      "loss": 0.8514,
+      "step": 794
+    },
+    {
+      "epoch": 3.975,
+      "grad_norm": 8.838394165039062,
+      "learning_rate": 8.012500000000002e-06,
+      "loss": 0.9766,
+      "step": 795
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 6.454511642456055,
+      "learning_rate": 8.010000000000001e-06,
+      "loss": 0.5652,
+      "step": 796
+    },
+    {
+      "epoch": 3.985,
+      "grad_norm": 23.649662017822266,
+      "learning_rate": 8.0075e-06,
+      "loss": 0.9774,
+      "step": 797
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 6.850311279296875,
+      "learning_rate": 8.005e-06,
+      "loss": 0.7934,
+      "step": 798
+    },
+    {
+      "epoch": 3.995,
+      "grad_norm": 11.368961334228516,
+      "learning_rate": 8.0025e-06,
+      "loss": 0.8806,
+      "step": 799
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 10.509380340576172,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.0133,
+      "step": 800
+    },
+    {
+      "epoch": 4.0,
+      "eval_loss": 0.5585627555847168,
+      "eval_runtime": 15.1713,
+      "eval_samples_per_second": 52.599,
+      "eval_steps_per_second": 6.591,
+      "step": 800
+    },
+    {
+      "epoch": 4.005,
+      "grad_norm": 13.838993072509766,
+      "learning_rate": 7.997500000000001e-06,
+      "loss": 1.5727,
+      "step": 801
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 8.771504402160645,
+      "learning_rate": 7.995e-06,
+      "loss": 0.5541,
+      "step": 802
+    },
+    {
+      "epoch": 4.015,
+      "grad_norm": 11.329957008361816,
+      "learning_rate": 7.9925e-06,
+      "loss": 0.9736,
+      "step": 803
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 7.5490827560424805,
+      "learning_rate": 7.990000000000001e-06,
+      "loss": 0.9586,
+      "step": 804
+    },
+    {
+      "epoch": 4.025,
+      "grad_norm": 9.27094554901123,
+      "learning_rate": 7.987500000000001e-06,
+      "loss": 0.9663,
+      "step": 805
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 7.9206767082214355,
+      "learning_rate": 7.985e-06,
+      "loss": 0.9763,
+      "step": 806
+    },
+    {
+      "epoch": 4.035,
+      "grad_norm": 7.458390712738037,
+      "learning_rate": 7.9825e-06,
+      "loss": 0.5843,
+      "step": 807
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 6.256739139556885,
+      "learning_rate": 7.980000000000002e-06,
+      "loss": 0.4645,
+      "step": 808
+    },
+    {
+      "epoch": 4.045,
+      "grad_norm": 29.080068588256836,
+      "learning_rate": 7.9775e-06,
+      "loss": 0.4344,
+      "step": 809
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 8.608640670776367,
+      "learning_rate": 7.975e-06,
+      "loss": 0.5619,
+      "step": 810
+    },
+    {
+      "epoch": 4.055,
+      "grad_norm": 12.236395835876465,
+      "learning_rate": 7.9725e-06,
+      "loss": 0.8991,
+      "step": 811
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 9.644879341125488,
+      "learning_rate": 7.970000000000002e-06,
+      "loss": 1.2743,
+      "step": 812
+    },
+    {
+      "epoch": 4.065,
+      "grad_norm": 11.649674415588379,
+      "learning_rate": 7.9675e-06,
+      "loss": 0.912,
+      "step": 813
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 7.4158759117126465,
+      "learning_rate": 7.965e-06,
+      "loss": 0.8013,
+      "step": 814
+    },
+    {
+      "epoch": 4.075,
+      "grad_norm": 10.154393196105957,
+      "learning_rate": 7.9625e-06,
+      "loss": 0.8836,
+      "step": 815
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 6.346776485443115,
+      "learning_rate": 7.960000000000002e-06,
+      "loss": 0.6558,
+      "step": 816
+    },
+    {
+      "epoch": 4.085,
+      "grad_norm": 7.9860453605651855,
+      "learning_rate": 7.9575e-06,
+      "loss": 0.6322,
+      "step": 817
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 8.638710021972656,
+      "learning_rate": 7.955000000000001e-06,
+      "loss": 0.6606,
+      "step": 818
+    },
+    {
+      "epoch": 4.095,
+      "grad_norm": 15.537666320800781,
+      "learning_rate": 7.9525e-06,
+      "loss": 1.4751,
+      "step": 819
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 12.075648307800293,
+      "learning_rate": 7.950000000000002e-06,
+      "loss": 0.9565,
+      "step": 820
+    },
+    {
+      "epoch": 4.105,
+      "grad_norm": 9.073077201843262,
+      "learning_rate": 7.9475e-06,
+      "loss": 0.7414,
+      "step": 821
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 7.570252895355225,
+      "learning_rate": 7.945000000000001e-06,
+      "loss": 0.624,
+      "step": 822
+    },
+    {
+      "epoch": 4.115,
+      "grad_norm": 7.466359615325928,
+      "learning_rate": 7.9425e-06,
+      "loss": 0.6402,
+      "step": 823
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 7.846566200256348,
+      "learning_rate": 7.94e-06,
+      "loss": 0.8834,
+      "step": 824
+    },
+    {
+      "epoch": 4.125,
+      "grad_norm": 6.028357028961182,
+      "learning_rate": 7.9375e-06,
+      "loss": 0.541,
+      "step": 825
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 5.9524688720703125,
+      "learning_rate": 7.935000000000001e-06,
+      "loss": 0.5684,
+      "step": 826
+    },
+    {
+      "epoch": 4.135,
+      "grad_norm": 9.196531295776367,
+      "learning_rate": 7.9325e-06,
+      "loss": 1.0029,
+      "step": 827
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 9.987871170043945,
+      "learning_rate": 7.93e-06,
+      "loss": 1.3181,
+      "step": 828
+    },
+    {
+      "epoch": 4.145,
+      "grad_norm": 8.50416374206543,
+      "learning_rate": 7.9275e-06,
+      "loss": 1.4644,
+      "step": 829
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 14.648338317871094,
+      "learning_rate": 7.925000000000001e-06,
+      "loss": 0.9936,
+      "step": 830
+    },
+    {
+      "epoch": 4.155,
+      "grad_norm": 8.887319564819336,
+      "learning_rate": 7.9225e-06,
+      "loss": 0.8414,
+      "step": 831
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 8.118429183959961,
+      "learning_rate": 7.92e-06,
+      "loss": 0.6139,
+      "step": 832
+    },
+    {
+      "epoch": 4.165,
+      "grad_norm": 6.807774066925049,
+      "learning_rate": 7.9175e-06,
+      "loss": 0.7383,
+      "step": 833
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 6.264782428741455,
+      "learning_rate": 7.915000000000001e-06,
+      "loss": 0.7502,
+      "step": 834
+    },
+    {
+      "epoch": 4.175,
+      "grad_norm": 11.799866676330566,
+      "learning_rate": 7.912500000000001e-06,
+      "loss": 1.3401,
+      "step": 835
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 9.960253715515137,
+      "learning_rate": 7.91e-06,
+      "loss": 0.6108,
+      "step": 836
+    },
+    {
+      "epoch": 4.185,
+      "grad_norm": 5.822458267211914,
+      "learning_rate": 7.9075e-06,
+      "loss": 0.7158,
+      "step": 837
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 6.580624103546143,
+      "learning_rate": 7.905e-06,
+      "loss": 0.6817,
+      "step": 838
+    },
+    {
+      "epoch": 4.195,
+      "grad_norm": 7.986283302307129,
+      "learning_rate": 7.902500000000001e-06,
+      "loss": 0.8991,
+      "step": 839
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 5.431870937347412,
+      "learning_rate": 7.9e-06,
+      "loss": 0.4277,
+      "step": 840
+    },
+    {
+      "epoch": 4.205,
+      "grad_norm": 8.297979354858398,
+      "learning_rate": 7.8975e-06,
+      "loss": 1.0806,
+      "step": 841
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 7.770302772521973,
+      "learning_rate": 7.895e-06,
+      "loss": 0.9686,
+      "step": 842
+    },
+    {
+      "epoch": 4.215,
+      "grad_norm": 7.879178524017334,
+      "learning_rate": 7.892500000000001e-06,
+      "loss": 1.1393,
+      "step": 843
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 34.744911193847656,
+      "learning_rate": 7.89e-06,
+      "loss": 1.5025,
+      "step": 844
+    },
+    {
+      "epoch": 4.225,
+      "grad_norm": 11.846839904785156,
+      "learning_rate": 7.8875e-06,
+      "loss": 0.8516,
+      "step": 845
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 20.808643341064453,
+      "learning_rate": 7.885e-06,
+      "loss": 1.7091,
+      "step": 846
+    },
+    {
+      "epoch": 4.235,
+      "grad_norm": 18.16811180114746,
+      "learning_rate": 7.882500000000001e-06,
+      "loss": 0.7861,
+      "step": 847
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 16.109603881835938,
+      "learning_rate": 7.88e-06,
+      "loss": 0.5653,
+      "step": 848
+    },
+    {
+      "epoch": 4.245,
+      "grad_norm": 8.238914489746094,
+      "learning_rate": 7.8775e-06,
+      "loss": 0.5073,
+      "step": 849
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 7.80977201461792,
+      "learning_rate": 7.875e-06,
+      "loss": 1.1725,
+      "step": 850
+    },
+    {
+      "epoch": 4.255,
+      "grad_norm": 9.963988304138184,
+      "learning_rate": 7.872500000000001e-06,
+      "loss": 0.5111,
+      "step": 851
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 6.031205177307129,
+      "learning_rate": 7.870000000000001e-06,
+      "loss": 0.8074,
+      "step": 852
+    },
+    {
+      "epoch": 4.265,
+      "grad_norm": 7.428991794586182,
+      "learning_rate": 7.8675e-06,
+      "loss": 0.4875,
+      "step": 853
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 10.712441444396973,
+      "learning_rate": 7.865e-06,
+      "loss": 0.4775,
+      "step": 854
+    },
+    {
+      "epoch": 4.275,
+      "grad_norm": 6.611090183258057,
+      "learning_rate": 7.862500000000001e-06,
+      "loss": 0.6714,
+      "step": 855
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 9.8261137008667,
+      "learning_rate": 7.860000000000001e-06,
+      "loss": 0.56,
+      "step": 856
+    },
+    {
+      "epoch": 4.285,
+      "grad_norm": 48.70997619628906,
+      "learning_rate": 7.8575e-06,
+      "loss": 0.9612,
+      "step": 857
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 10.669054985046387,
+      "learning_rate": 7.855e-06,
+      "loss": 0.6933,
+      "step": 858
+    },
+    {
+      "epoch": 4.295,
+      "grad_norm": 9.595419883728027,
+      "learning_rate": 7.852500000000001e-06,
+      "loss": 0.4313,
+      "step": 859
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 10.332606315612793,
+      "learning_rate": 7.850000000000001e-06,
+      "loss": 1.0216,
+      "step": 860
+    },
+    {
+      "epoch": 4.305,
+      "grad_norm": 10.682745933532715,
+      "learning_rate": 7.8475e-06,
+      "loss": 0.7581,
+      "step": 861
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 5.491994857788086,
+      "learning_rate": 7.845e-06,
+      "loss": 0.6867,
+      "step": 862
+    },
+    {
+      "epoch": 4.315,
+      "grad_norm": 6.151000499725342,
+      "learning_rate": 7.8425e-06,
+      "loss": 0.799,
+      "step": 863
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 8.0258150100708,
+      "learning_rate": 7.840000000000001e-06,
+      "loss": 0.7104,
+      "step": 864
+    },
+    {
+      "epoch": 4.325,
+      "grad_norm": 17.792173385620117,
+      "learning_rate": 7.8375e-06,
+      "loss": 0.9917,
+      "step": 865
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 6.857790470123291,
+      "learning_rate": 7.835e-06,
+      "loss": 0.5869,
+      "step": 866
+    },
+    {
+      "epoch": 4.335,
+      "grad_norm": 7.594349384307861,
+      "learning_rate": 7.8325e-06,
+      "loss": 0.6549,
+      "step": 867
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 5.57966947555542,
+      "learning_rate": 7.830000000000001e-06,
+      "loss": 0.6993,
+      "step": 868
+    },
+    {
+      "epoch": 4.345,
+      "grad_norm": 10.504386901855469,
+      "learning_rate": 7.827500000000001e-06,
+      "loss": 0.8398,
+      "step": 869
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 6.704975605010986,
+      "learning_rate": 7.825e-06,
+      "loss": 0.9093,
+      "step": 870
+    },
+    {
+      "epoch": 4.355,
+      "grad_norm": 485.8310546875,
+      "learning_rate": 7.8225e-06,
+      "loss": 1.8782,
+      "step": 871
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 7.297395706176758,
+      "learning_rate": 7.820000000000001e-06,
+      "loss": 0.8585,
+      "step": 872
+    },
+    {
+      "epoch": 4.365,
+      "grad_norm": 7.293272018432617,
+      "learning_rate": 7.8175e-06,
+      "loss": 0.5548,
+      "step": 873
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 7.3904829025268555,
+      "learning_rate": 7.815e-06,
+      "loss": 1.1342,
+      "step": 874
+    },
+    {
+      "epoch": 4.375,
+      "grad_norm": 9.324162483215332,
+      "learning_rate": 7.8125e-06,
+      "loss": 0.6262,
+      "step": 875
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 8.746965408325195,
+      "learning_rate": 7.810000000000001e-06,
+      "loss": 0.7124,
+      "step": 876
+    },
+    {
+      "epoch": 4.385,
+      "grad_norm": 8.10885238647461,
+      "learning_rate": 7.807500000000001e-06,
+      "loss": 1.2354,
+      "step": 877
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 9.036296844482422,
+      "learning_rate": 7.805e-06,
+      "loss": 0.9454,
+      "step": 878
+    },
+    {
+      "epoch": 4.395,
+      "grad_norm": 7.633198261260986,
+      "learning_rate": 7.8025e-06,
+      "loss": 0.7919,
+      "step": 879
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 7.132240295410156,
+      "learning_rate": 7.800000000000002e-06,
+      "loss": 0.7076,
+      "step": 880
+    },
+    {
+      "epoch": 4.405,
+      "grad_norm": 6.450599193572998,
+      "learning_rate": 7.797500000000001e-06,
+      "loss": 0.6359,
+      "step": 881
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 6.664942264556885,
+      "learning_rate": 7.795e-06,
+      "loss": 0.9764,
+      "step": 882
+    },
+    {
+      "epoch": 4.415,
+      "grad_norm": 7.093107223510742,
+      "learning_rate": 7.7925e-06,
+      "loss": 1.2649,
+      "step": 883
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 7.043144226074219,
+      "learning_rate": 7.790000000000002e-06,
+      "loss": 0.6203,
+      "step": 884
+    },
+    {
+      "epoch": 4.425,
+      "grad_norm": 7.215663909912109,
+      "learning_rate": 7.787500000000001e-06,
+      "loss": 0.8837,
+      "step": 885
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 6.580089569091797,
+      "learning_rate": 7.785000000000001e-06,
+      "loss": 0.755,
+      "step": 886
+    },
+    {
+      "epoch": 4.435,
+      "grad_norm": 10.465328216552734,
+      "learning_rate": 7.7825e-06,
+      "loss": 0.5604,
+      "step": 887
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 6.5835137367248535,
+      "learning_rate": 7.78e-06,
+      "loss": 0.8127,
+      "step": 888
+    },
+    {
+      "epoch": 4.445,
+      "grad_norm": 7.070140361785889,
+      "learning_rate": 7.777500000000001e-06,
+      "loss": 0.6967,
+      "step": 889
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 9.254812240600586,
+      "learning_rate": 7.775000000000001e-06,
+      "loss": 0.8161,
+      "step": 890
+    },
+    {
+      "epoch": 4.455,
+      "grad_norm": 7.753115177154541,
+      "learning_rate": 7.7725e-06,
+      "loss": 0.7934,
+      "step": 891
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 11.864736557006836,
+      "learning_rate": 7.77e-06,
+      "loss": 1.1878,
+      "step": 892
+    },
+    {
+      "epoch": 4.465,
+      "grad_norm": 12.776810646057129,
+      "learning_rate": 7.767500000000001e-06,
+      "loss": 0.6883,
+      "step": 893
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 5.750730514526367,
+      "learning_rate": 7.765000000000001e-06,
+      "loss": 0.4594,
+      "step": 894
+    },
+    {
+      "epoch": 4.475,
+      "grad_norm": 23.227420806884766,
+      "learning_rate": 7.7625e-06,
+      "loss": 0.7727,
+      "step": 895
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 21.548477172851562,
+      "learning_rate": 7.76e-06,
+      "loss": 0.6231,
+      "step": 896
+    },
+    {
+      "epoch": 4.485,
+      "grad_norm": 9.685548782348633,
+      "learning_rate": 7.757500000000002e-06,
+      "loss": 0.9737,
+      "step": 897
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 8.175079345703125,
+      "learning_rate": 7.755000000000001e-06,
+      "loss": 0.7429,
+      "step": 898
+    },
+    {
+      "epoch": 4.495,
+      "grad_norm": 4.5502495765686035,
+      "learning_rate": 7.7525e-06,
+      "loss": 0.3891,
+      "step": 899
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 9.984167098999023,
+      "learning_rate": 7.75e-06,
+      "loss": 1.1427,
+      "step": 900
+    },
+    {
+      "epoch": 4.505,
+      "grad_norm": 5.199976444244385,
+      "learning_rate": 7.747500000000002e-06,
+      "loss": 0.5247,
+      "step": 901
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 11.98429012298584,
+      "learning_rate": 7.745e-06,
+      "loss": 0.7552,
+      "step": 902
+    },
+    {
+      "epoch": 4.515,
+      "grad_norm": 6.843251705169678,
+      "learning_rate": 7.7425e-06,
+      "loss": 0.7978,
+      "step": 903
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 6.33599328994751,
+      "learning_rate": 7.74e-06,
+      "loss": 0.3372,
+      "step": 904
+    },
+    {
+      "epoch": 4.525,
+      "grad_norm": 11.031352996826172,
+      "learning_rate": 7.737500000000002e-06,
+      "loss": 0.7817,
+      "step": 905
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 7.0565619468688965,
+      "learning_rate": 7.735e-06,
+      "loss": 0.5259,
+      "step": 906
+    },
+    {
+      "epoch": 4.535,
+      "grad_norm": 7.741739749908447,
+      "learning_rate": 7.732500000000001e-06,
+      "loss": 0.8119,
+      "step": 907
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 6.05742883682251,
+      "learning_rate": 7.73e-06,
+      "loss": 0.8334,
+      "step": 908
+    },
+    {
+      "epoch": 4.545,
+      "grad_norm": 7.659331321716309,
+      "learning_rate": 7.727500000000002e-06,
+      "loss": 1.1716,
+      "step": 909
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 6.693140983581543,
+      "learning_rate": 7.725e-06,
+      "loss": 0.5503,
+      "step": 910
+    },
+    {
+      "epoch": 4.555,
+      "grad_norm": 5.622475624084473,
+      "learning_rate": 7.722500000000001e-06,
+      "loss": 0.7317,
+      "step": 911
+    },
+    {
+      "epoch": 4.5600000000000005,
+      "grad_norm": 7.839320182800293,
+      "learning_rate": 7.72e-06,
+      "loss": 1.0013,
+      "step": 912
+    },
+    {
+      "epoch": 4.5649999999999995,
+      "grad_norm": 7.353050708770752,
+      "learning_rate": 7.717500000000002e-06,
+      "loss": 0.9737,
+      "step": 913
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 9.281424522399902,
+      "learning_rate": 7.715e-06,
+      "loss": 0.9555,
+      "step": 914
+    },
+    {
+      "epoch": 4.575,
+      "grad_norm": 8.326522827148438,
+      "learning_rate": 7.712500000000001e-06,
+      "loss": 0.9718,
+      "step": 915
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 10.096565246582031,
+      "learning_rate": 7.71e-06,
+      "loss": 0.6967,
+      "step": 916
+    },
+    {
+      "epoch": 4.585,
+      "grad_norm": 5.208967208862305,
+      "learning_rate": 7.7075e-06,
+      "loss": 1.1157,
+      "step": 917
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 6.28977108001709,
+      "learning_rate": 7.705e-06,
+      "loss": 0.6969,
+      "step": 918
+    },
+    {
+      "epoch": 4.595,
+      "grad_norm": 11.165891647338867,
+      "learning_rate": 7.702500000000001e-06,
+      "loss": 0.9135,
+      "step": 919
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 6.866476058959961,
+      "learning_rate": 7.7e-06,
+      "loss": 0.4429,
+      "step": 920
+    },
+    {
+      "epoch": 4.605,
+      "grad_norm": 9.071453094482422,
+      "learning_rate": 7.6975e-06,
+      "loss": 0.4437,
+      "step": 921
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 5.0336785316467285,
+      "learning_rate": 7.695e-06,
+      "loss": 0.6089,
+      "step": 922
+    },
+    {
+      "epoch": 4.615,
+      "grad_norm": 9.007340431213379,
+      "learning_rate": 7.692500000000001e-06,
+      "loss": 0.9301,
+      "step": 923
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 7.89274787902832,
+      "learning_rate": 7.690000000000001e-06,
+      "loss": 0.7974,
+      "step": 924
+    },
+    {
+      "epoch": 4.625,
+      "grad_norm": 9.030817031860352,
+      "learning_rate": 7.6875e-06,
+      "loss": 0.9197,
+      "step": 925
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 7.764341831207275,
+      "learning_rate": 7.685e-06,
+      "loss": 0.6966,
+      "step": 926
+    },
+    {
+      "epoch": 4.635,
+      "grad_norm": 13.567545890808105,
+      "learning_rate": 7.6825e-06,
+      "loss": 0.8618,
+      "step": 927
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 7.146632194519043,
+      "learning_rate": 7.680000000000001e-06,
+      "loss": 0.6834,
+      "step": 928
+    },
+    {
+      "epoch": 4.645,
+      "grad_norm": 5.738801956176758,
+      "learning_rate": 7.6775e-06,
+      "loss": 0.6694,
+      "step": 929
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 5.258743762969971,
+      "learning_rate": 7.675e-06,
+      "loss": 0.5751,
+      "step": 930
+    },
+    {
+      "epoch": 4.655,
+      "grad_norm": 6.266324043273926,
+      "learning_rate": 7.6725e-06,
+      "loss": 0.6656,
+      "step": 931
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 5.638592720031738,
+      "learning_rate": 7.670000000000001e-06,
+      "loss": 0.5344,
+      "step": 932
+    },
+    {
+      "epoch": 4.665,
+      "grad_norm": 8.314552307128906,
+      "learning_rate": 7.6675e-06,
+      "loss": 0.5518,
+      "step": 933
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 5.586596488952637,
+      "learning_rate": 7.665e-06,
+      "loss": 0.4989,
+      "step": 934
+    },
+    {
+      "epoch": 4.675,
+      "grad_norm": 6.009988307952881,
+      "learning_rate": 7.6625e-06,
+      "loss": 0.7549,
+      "step": 935
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 4.864471912384033,
+      "learning_rate": 7.660000000000001e-06,
+      "loss": 0.5008,
+      "step": 936
+    },
+    {
+      "epoch": 4.6850000000000005,
+      "grad_norm": 12.279769897460938,
+      "learning_rate": 7.6575e-06,
+      "loss": 0.591,
+      "step": 937
+    },
+    {
+      "epoch": 4.6899999999999995,
+      "grad_norm": 8.40523624420166,
+      "learning_rate": 7.655e-06,
+      "loss": 1.1309,
+      "step": 938
+    },
+    {
+      "epoch": 4.695,
+      "grad_norm": 5.364549160003662,
+      "learning_rate": 7.6525e-06,
+      "loss": 0.9289,
+      "step": 939
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 8.084023475646973,
+      "learning_rate": 7.650000000000001e-06,
+      "loss": 0.4936,
+      "step": 940
+    },
+    {
+      "epoch": 4.705,
+      "grad_norm": 6.338261127471924,
+      "learning_rate": 7.647500000000001e-06,
+      "loss": 0.6708,
+      "step": 941
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 20.893301010131836,
+      "learning_rate": 7.645e-06,
+      "loss": 1.3248,
+      "step": 942
+    },
+    {
+      "epoch": 4.715,
+      "grad_norm": 5.312731742858887,
+      "learning_rate": 7.6425e-06,
+      "loss": 0.5052,
+      "step": 943
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 8.227688789367676,
+      "learning_rate": 7.640000000000001e-06,
+      "loss": 0.7726,
+      "step": 944
+    },
+    {
+      "epoch": 4.725,
+      "grad_norm": 5.971982479095459,
+      "learning_rate": 7.637500000000001e-06,
+      "loss": 0.387,
+      "step": 945
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 4.514801502227783,
+      "learning_rate": 7.635e-06,
+      "loss": 0.6082,
+      "step": 946
+    },
+    {
+      "epoch": 4.735,
+      "grad_norm": 7.329568386077881,
+      "learning_rate": 7.6325e-06,
+      "loss": 0.5931,
+      "step": 947
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 8.544476509094238,
+      "learning_rate": 7.630000000000001e-06,
+      "loss": 0.8867,
+      "step": 948
+    },
+    {
+      "epoch": 4.745,
+      "grad_norm": 6.30322790145874,
+      "learning_rate": 7.627500000000001e-06,
+      "loss": 1.1539,
+      "step": 949
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 6.4409332275390625,
+      "learning_rate": 7.625e-06,
+      "loss": 0.5566,
+      "step": 950
+    },
+    {
+      "epoch": 4.755,
+      "grad_norm": 5.551914691925049,
+      "learning_rate": 7.6225e-06,
+      "loss": 0.6575,
+      "step": 951
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 9.450929641723633,
+      "learning_rate": 7.620000000000001e-06,
+      "loss": 0.7796,
+      "step": 952
+    },
+    {
+      "epoch": 4.765,
+      "grad_norm": 9.002180099487305,
+      "learning_rate": 7.617500000000001e-06,
+      "loss": 0.9793,
+      "step": 953
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 6.421311378479004,
+      "learning_rate": 7.615e-06,
+      "loss": 0.6812,
+      "step": 954
+    },
+    {
+      "epoch": 4.775,
+      "grad_norm": 6.8888139724731445,
+      "learning_rate": 7.6125e-06,
+      "loss": 0.8761,
+      "step": 955
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 9.849982261657715,
+      "learning_rate": 7.610000000000001e-06,
+      "loss": 0.8667,
+      "step": 956
+    },
+    {
+      "epoch": 4.785,
+      "grad_norm": 6.173455238342285,
+      "learning_rate": 7.607500000000001e-06,
+      "loss": 0.8259,
+      "step": 957
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 9.471467018127441,
+      "learning_rate": 7.605e-06,
+      "loss": 0.7598,
+      "step": 958
+    },
+    {
+      "epoch": 4.795,
+      "grad_norm": 5.789381980895996,
+      "learning_rate": 7.6025000000000005e-06,
+      "loss": 0.9815,
+      "step": 959
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 6.646718978881836,
+      "learning_rate": 7.600000000000001e-06,
+      "loss": 0.6353,
+      "step": 960
+    },
+    {
+      "epoch": 4.805,
+      "grad_norm": 5.5178751945495605,
+      "learning_rate": 7.597500000000001e-06,
+      "loss": 0.651,
+      "step": 961
+    },
+    {
+      "epoch": 4.8100000000000005,
+      "grad_norm": 14.98030948638916,
+      "learning_rate": 7.595e-06,
+      "loss": 1.0969,
+      "step": 962
+    },
+    {
+      "epoch": 4.8149999999999995,
+      "grad_norm": 6.984838008880615,
+      "learning_rate": 7.5925000000000006e-06,
+      "loss": 0.4359,
+      "step": 963
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 5.002830982208252,
+      "learning_rate": 7.590000000000001e-06,
+      "loss": 0.5469,
+      "step": 964
+    },
+    {
+      "epoch": 4.825,
+      "grad_norm": 8.776707649230957,
+      "learning_rate": 7.587500000000001e-06,
+      "loss": 0.6992,
+      "step": 965
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 9.907809257507324,
+      "learning_rate": 7.585e-06,
+      "loss": 0.6271,
+      "step": 966
+    },
+    {
+      "epoch": 4.835,
+      "grad_norm": 5.77224063873291,
+      "learning_rate": 7.582500000000001e-06,
+      "loss": 0.7432,
+      "step": 967
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 14.700920104980469,
+      "learning_rate": 7.58e-06,
+      "loss": 1.3781,
+      "step": 968
+    },
+    {
+      "epoch": 4.845,
+      "grad_norm": 6.946707248687744,
+      "learning_rate": 7.577500000000001e-06,
+      "loss": 1.319,
+      "step": 969
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 5.933041095733643,
+      "learning_rate": 7.575e-06,
+      "loss": 0.5965,
+      "step": 970
+    },
+    {
+      "epoch": 4.855,
+      "grad_norm": 7.6721320152282715,
+      "learning_rate": 7.572500000000001e-06,
+      "loss": 0.9051,
+      "step": 971
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 10.433431625366211,
+      "learning_rate": 7.57e-06,
+      "loss": 1.1911,
+      "step": 972
+    },
+    {
+      "epoch": 4.865,
+      "grad_norm": 7.691511154174805,
+      "learning_rate": 7.567500000000001e-06,
+      "loss": 0.4459,
+      "step": 973
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 20.308088302612305,
+      "learning_rate": 7.565e-06,
+      "loss": 0.3769,
+      "step": 974
+    },
+    {
+      "epoch": 4.875,
+      "grad_norm": 5.437723159790039,
+      "learning_rate": 7.5625e-06,
+      "loss": 0.785,
+      "step": 975
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 7.873809337615967,
+      "learning_rate": 7.5600000000000005e-06,
+      "loss": 0.6967,
+      "step": 976
+    },
+    {
+      "epoch": 4.885,
+      "grad_norm": 5.258434772491455,
+      "learning_rate": 7.557500000000001e-06,
+      "loss": 0.5594,
+      "step": 977
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 9.172933578491211,
+      "learning_rate": 7.5550000000000005e-06,
+      "loss": 0.6167,
+      "step": 978
+    },
+    {
+      "epoch": 4.895,
+      "grad_norm": 71.53560638427734,
+      "learning_rate": 7.5525e-06,
+      "loss": 1.8214,
+      "step": 979
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 5.724739074707031,
+      "learning_rate": 7.5500000000000006e-06,
+      "loss": 0.6524,
+      "step": 980
+    },
+    {
+      "epoch": 4.905,
+      "grad_norm": 4.90444803237915,
+      "learning_rate": 7.547500000000001e-06,
+      "loss": 0.7813,
+      "step": 981
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 6.017374515533447,
+      "learning_rate": 7.545e-06,
+      "loss": 1.0454,
+      "step": 982
+    },
+    {
+      "epoch": 4.915,
+      "grad_norm": 5.501596927642822,
+      "learning_rate": 7.5425e-06,
+      "loss": 0.6801,
+      "step": 983
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 4.529572486877441,
+      "learning_rate": 7.540000000000001e-06,
+      "loss": 0.6049,
+      "step": 984
+    },
+    {
+      "epoch": 4.925,
+      "grad_norm": 6.704092979431152,
+      "learning_rate": 7.537500000000001e-06,
+      "loss": 0.5803,
+      "step": 985
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 9.10400676727295,
+      "learning_rate": 7.535e-06,
+      "loss": 0.7328,
+      "step": 986
+    },
+    {
+      "epoch": 4.9350000000000005,
+      "grad_norm": 3.782054901123047,
+      "learning_rate": 7.5325e-06,
+      "loss": 0.5806,
+      "step": 987
+    },
+    {
+      "epoch": 4.9399999999999995,
+      "grad_norm": 13.161701202392578,
+      "learning_rate": 7.530000000000001e-06,
+      "loss": 0.8132,
+      "step": 988
+    },
+    {
+      "epoch": 4.945,
+      "grad_norm": 6.359323024749756,
+      "learning_rate": 7.527500000000001e-06,
+      "loss": 0.5623,
+      "step": 989
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 5.326925754547119,
+      "learning_rate": 7.525e-06,
+      "loss": 0.9189,
+      "step": 990
+    },
+    {
+      "epoch": 4.955,
+      "grad_norm": 3.474952220916748,
+      "learning_rate": 7.5225e-06,
+      "loss": 0.3923,
+      "step": 991
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 7.8486833572387695,
+      "learning_rate": 7.520000000000001e-06,
+      "loss": 0.9266,
+      "step": 992
+    },
+    {
+      "epoch": 4.965,
+      "grad_norm": 5.696709632873535,
+      "learning_rate": 7.517500000000001e-06,
+      "loss": 0.5579,
+      "step": 993
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 17.016956329345703,
+      "learning_rate": 7.515e-06,
+      "loss": 0.6911,
+      "step": 994
+    },
+    {
+      "epoch": 4.975,
+      "grad_norm": 16.98024559020996,
+      "learning_rate": 7.5125000000000005e-06,
+      "loss": 0.8292,
+      "step": 995
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 7.151941776275635,
+      "learning_rate": 7.510000000000001e-06,
+      "loss": 0.7377,
+      "step": 996
+    },
+    {
+      "epoch": 4.985,
+      "grad_norm": 4.528003215789795,
+      "learning_rate": 7.5075000000000005e-06,
+      "loss": 0.5026,
+      "step": 997
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 5.102850914001465,
+      "learning_rate": 7.505e-06,
+      "loss": 0.7528,
+      "step": 998
+    },
+    {
+      "epoch": 4.995,
+      "grad_norm": 5.543210506439209,
+      "learning_rate": 7.502500000000001e-06,
+      "loss": 0.4479,
+      "step": 999
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 5.714639663696289,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.5175,
+      "step": 1000
+    },
+    {
+      "epoch": 5.0,
+      "eval_loss": 0.4415563642978668,
+      "eval_runtime": 15.2515,
+      "eval_samples_per_second": 52.323,
+      "eval_steps_per_second": 6.557,
+      "step": 1000
+    },
+    {
+      "epoch": 5.005,
+      "grad_norm": 6.787521839141846,
+      "learning_rate": 7.497500000000001e-06,
+      "loss": 0.5499,
+      "step": 1001
+    },
+    {
+      "epoch": 5.01,
+      "grad_norm": 11.19926643371582,
+      "learning_rate": 7.495000000000001e-06,
+      "loss": 0.9259,
+      "step": 1002
+    },
+    {
+      "epoch": 5.015,
+      "grad_norm": 5.9822797775268555,
+      "learning_rate": 7.492500000000001e-06,
+      "loss": 0.91,
+      "step": 1003
+    },
+    {
+      "epoch": 5.02,
+      "grad_norm": 4.771771430969238,
+      "learning_rate": 7.49e-06,
+      "loss": 0.6687,
+      "step": 1004
+    },
+    {
+      "epoch": 5.025,
+      "grad_norm": 6.216484069824219,
+      "learning_rate": 7.487500000000001e-06,
+      "loss": 0.4382,
+      "step": 1005
+    },
+    {
+      "epoch": 5.03,
+      "grad_norm": 7.429642677307129,
+      "learning_rate": 7.485000000000001e-06,
+      "loss": 0.5299,
+      "step": 1006
+    },
+    {
+      "epoch": 5.035,
+      "grad_norm": 3.297884464263916,
+      "learning_rate": 7.4825e-06,
+      "loss": 0.3392,
+      "step": 1007
+    },
+    {
+      "epoch": 5.04,
+      "grad_norm": 7.149893760681152,
+      "learning_rate": 7.48e-06,
+      "loss": 0.7428,
+      "step": 1008
+    },
+    {
+      "epoch": 5.045,
+      "grad_norm": 4.738568305969238,
+      "learning_rate": 7.477500000000001e-06,
+      "loss": 0.6791,
+      "step": 1009
+    },
+    {
+      "epoch": 5.05,
+      "grad_norm": 5.036160945892334,
+      "learning_rate": 7.475000000000001e-06,
+      "loss": 0.6676,
+      "step": 1010
+    },
+    {
+      "epoch": 5.055,
+      "grad_norm": 4.8705525398254395,
+      "learning_rate": 7.4725e-06,
+      "loss": 0.4155,
+      "step": 1011
+    },
+    {
+      "epoch": 5.06,
+      "grad_norm": 4.042587757110596,
+      "learning_rate": 7.4700000000000005e-06,
+      "loss": 0.3965,
+      "step": 1012
+    },
+    {
+      "epoch": 5.065,
+      "grad_norm": 348.202880859375,
+      "learning_rate": 7.467500000000001e-06,
+      "loss": 1.419,
+      "step": 1013
+    },
+    {
+      "epoch": 5.07,
+      "grad_norm": 8.463419914245605,
+      "learning_rate": 7.465000000000001e-06,
+      "loss": 0.9332,
+      "step": 1014
+    },
+    {
+      "epoch": 5.075,
+      "grad_norm": 25.667293548583984,
+      "learning_rate": 7.4625e-06,
+      "loss": 0.5851,
+      "step": 1015
+    },
+    {
+      "epoch": 5.08,
+      "grad_norm": 4.099241256713867,
+      "learning_rate": 7.4600000000000006e-06,
+      "loss": 0.5093,
+      "step": 1016
+    },
+    {
+      "epoch": 5.085,
+      "grad_norm": 6.697269439697266,
+      "learning_rate": 7.457500000000001e-06,
+      "loss": 0.7548,
+      "step": 1017
+    },
+    {
+      "epoch": 5.09,
+      "grad_norm": 4.3906426429748535,
+      "learning_rate": 7.4550000000000015e-06,
+      "loss": 0.6866,
+      "step": 1018
+    },
+    {
+      "epoch": 5.095,
+      "grad_norm": 5.171369552612305,
+      "learning_rate": 7.4525e-06,
+      "loss": 0.3326,
+      "step": 1019
+    },
+    {
+      "epoch": 5.1,
+      "grad_norm": 8.67208480834961,
+      "learning_rate": 7.450000000000001e-06,
+      "loss": 1.2265,
+      "step": 1020
+    },
+    {
+      "epoch": 5.105,
+      "grad_norm": 4.345002174377441,
+      "learning_rate": 7.447500000000001e-06,
+      "loss": 0.5598,
+      "step": 1021
+    },
+    {
+      "epoch": 5.11,
+      "grad_norm": 5.751708984375,
+      "learning_rate": 7.445000000000001e-06,
+      "loss": 0.5184,
+      "step": 1022
+    },
+    {
+      "epoch": 5.115,
+      "grad_norm": 4.406136989593506,
+      "learning_rate": 7.4425e-06,
+      "loss": 0.4727,
+      "step": 1023
+    },
+    {
+      "epoch": 5.12,
+      "grad_norm": 6.1146955490112305,
+      "learning_rate": 7.440000000000001e-06,
+      "loss": 0.3679,
+      "step": 1024
+    },
+    {
+      "epoch": 5.125,
+      "grad_norm": 4.922518730163574,
+      "learning_rate": 7.437500000000001e-06,
+      "loss": 0.529,
+      "step": 1025
+    },
+    {
+      "epoch": 5.13,
+      "grad_norm": 4.6229143142700195,
+      "learning_rate": 7.435000000000001e-06,
+      "loss": 0.5755,
+      "step": 1026
+    },
+    {
+      "epoch": 5.135,
+      "grad_norm": 5.552960395812988,
+      "learning_rate": 7.4325e-06,
+      "loss": 0.7747,
+      "step": 1027
+    },
+    {
+      "epoch": 5.14,
+      "grad_norm": 6.791009426116943,
+      "learning_rate": 7.430000000000001e-06,
+      "loss": 0.7384,
+      "step": 1028
+    },
+    {
+      "epoch": 5.145,
+      "grad_norm": 6.661989688873291,
+      "learning_rate": 7.4275000000000005e-06,
+      "loss": 0.4395,
+      "step": 1029
+    },
+    {
+      "epoch": 5.15,
+      "grad_norm": 7.699005126953125,
+      "learning_rate": 7.425000000000001e-06,
+      "loss": 0.5362,
+      "step": 1030
+    },
+    {
+      "epoch": 5.155,
+      "grad_norm": 5.395854949951172,
+      "learning_rate": 7.4225000000000005e-06,
+      "loss": 0.5644,
+      "step": 1031
+    },
+    {
+      "epoch": 5.16,
+      "grad_norm": 6.368740081787109,
+      "learning_rate": 7.420000000000001e-06,
+      "loss": 0.8221,
+      "step": 1032
+    },
+    {
+      "epoch": 5.165,
+      "grad_norm": 4.649238109588623,
+      "learning_rate": 7.4175000000000006e-06,
+      "loss": 0.7489,
+      "step": 1033
+    },
+    {
+      "epoch": 5.17,
+      "grad_norm": 5.846208572387695,
+      "learning_rate": 7.415000000000001e-06,
+      "loss": 0.9053,
+      "step": 1034
+    },
+    {
+      "epoch": 5.175,
+      "grad_norm": 5.621824264526367,
+      "learning_rate": 7.412500000000001e-06,
+      "loss": 0.7327,
+      "step": 1035
+    },
+    {
+      "epoch": 5.18,
+      "grad_norm": 5.679276466369629,
+      "learning_rate": 7.41e-06,
+      "loss": 0.5236,
+      "step": 1036
+    },
+    {
+      "epoch": 5.185,
+      "grad_norm": 4.791536331176758,
+      "learning_rate": 7.407500000000001e-06,
+      "loss": 0.2833,
+      "step": 1037
+    },
+    {
+      "epoch": 5.19,
+      "grad_norm": 5.230154991149902,
+      "learning_rate": 7.405000000000001e-06,
+      "loss": 0.8433,
+      "step": 1038
+    },
+    {
+      "epoch": 5.195,
+      "grad_norm": 4.504428386688232,
+      "learning_rate": 7.4025e-06,
+      "loss": 0.2513,
+      "step": 1039
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 22.153135299682617,
+      "learning_rate": 7.4e-06,
+      "loss": 0.8918,
+      "step": 1040
+    },
+    {
+      "epoch": 5.205,
+      "grad_norm": 4.62979793548584,
+      "learning_rate": 7.397500000000001e-06,
+      "loss": 0.6821,
+      "step": 1041
+    },
+    {
+      "epoch": 5.21,
+      "grad_norm": 5.088130474090576,
+      "learning_rate": 7.395000000000001e-06,
+      "loss": 0.4094,
+      "step": 1042
+    },
+    {
+      "epoch": 5.215,
+      "grad_norm": 7.661623954772949,
+      "learning_rate": 7.3925e-06,
+      "loss": 0.6482,
+      "step": 1043
+    },
+    {
+      "epoch": 5.22,
+      "grad_norm": 9.418888092041016,
+      "learning_rate": 7.39e-06,
+      "loss": 0.7209,
+      "step": 1044
+    },
+    {
+      "epoch": 5.225,
+      "grad_norm": 4.2088141441345215,
+      "learning_rate": 7.387500000000001e-06,
+      "loss": 0.3114,
+      "step": 1045
+    },
+    {
+      "epoch": 5.23,
+      "grad_norm": 7.295292377471924,
+      "learning_rate": 7.385000000000001e-06,
+      "loss": 0.8772,
+      "step": 1046
+    },
+    {
+      "epoch": 5.235,
+      "grad_norm": 7.384597301483154,
+      "learning_rate": 7.3825e-06,
+      "loss": 0.6176,
+      "step": 1047
+    },
+    {
+      "epoch": 5.24,
+      "grad_norm": 4.947549819946289,
+      "learning_rate": 7.3800000000000005e-06,
+      "loss": 0.5258,
+      "step": 1048
+    },
+    {
+      "epoch": 5.245,
+      "grad_norm": 6.196619987487793,
+      "learning_rate": 7.377500000000001e-06,
+      "loss": 0.8717,
+      "step": 1049
+    },
+    {
+      "epoch": 5.25,
+      "grad_norm": 7.719471454620361,
+      "learning_rate": 7.375000000000001e-06,
+      "loss": 0.5993,
+      "step": 1050
+    },
+    {
+      "epoch": 5.255,
+      "grad_norm": 3.9146170616149902,
+      "learning_rate": 7.3725e-06,
+      "loss": 0.354,
+      "step": 1051
+    },
+    {
+      "epoch": 5.26,
+      "grad_norm": 7.464202404022217,
+      "learning_rate": 7.370000000000001e-06,
+      "loss": 0.3568,
+      "step": 1052
+    },
+    {
+      "epoch": 5.265,
+      "grad_norm": 24.17234230041504,
+      "learning_rate": 7.367500000000001e-06,
+      "loss": 0.4861,
+      "step": 1053
+    },
+    {
+      "epoch": 5.27,
+      "grad_norm": 5.440557479858398,
+      "learning_rate": 7.365000000000001e-06,
+      "loss": 0.8632,
+      "step": 1054
+    },
+    {
+      "epoch": 5.275,
+      "grad_norm": 6.242492198944092,
+      "learning_rate": 7.3625e-06,
+      "loss": 0.5586,
+      "step": 1055
+    },
+    {
+      "epoch": 5.28,
+      "grad_norm": 5.984451770782471,
+      "learning_rate": 7.360000000000001e-06,
+      "loss": 0.3614,
+      "step": 1056
+    },
+    {
+      "epoch": 5.285,
+      "grad_norm": 4.822052001953125,
+      "learning_rate": 7.357500000000001e-06,
+      "loss": 0.49,
+      "step": 1057
+    },
+    {
+      "epoch": 5.29,
+      "grad_norm": 7.827376365661621,
+      "learning_rate": 7.355000000000001e-06,
+      "loss": 0.8211,
+      "step": 1058
+    },
+    {
+      "epoch": 5.295,
+      "grad_norm": 4.992079734802246,
+      "learning_rate": 7.3525e-06,
+      "loss": 0.3602,
+      "step": 1059
+    },
+    {
+      "epoch": 5.3,
+      "grad_norm": 6.9616379737854,
+      "learning_rate": 7.350000000000001e-06,
+      "loss": 0.5532,
+      "step": 1060
+    },
+    {
+      "epoch": 5.305,
+      "grad_norm": 3.8527936935424805,
+      "learning_rate": 7.3475e-06,
+      "loss": 0.2756,
+      "step": 1061
+    },
+    {
+      "epoch": 5.31,
+      "grad_norm": 7.39367151260376,
+      "learning_rate": 7.345000000000001e-06,
+      "loss": 0.7052,
+      "step": 1062
+    },
+    {
+      "epoch": 5.315,
+      "grad_norm": 9.638630867004395,
+      "learning_rate": 7.3425000000000004e-06,
+      "loss": 0.2968,
+      "step": 1063
+    },
+    {
+      "epoch": 5.32,
+      "grad_norm": 4.041645526885986,
+      "learning_rate": 7.340000000000001e-06,
+      "loss": 0.2679,
+      "step": 1064
+    },
+    {
+      "epoch": 5.325,
+      "grad_norm": 6.004903793334961,
+      "learning_rate": 7.3375000000000005e-06,
+      "loss": 0.3562,
+      "step": 1065
+    },
+    {
+      "epoch": 5.33,
+      "grad_norm": 4.400599956512451,
+      "learning_rate": 7.335000000000001e-06,
+      "loss": 0.6988,
+      "step": 1066
+    },
+    {
+      "epoch": 5.335,
+      "grad_norm": 5.240605354309082,
+      "learning_rate": 7.3325000000000005e-06,
+      "loss": 0.5638,
+      "step": 1067
+    },
+    {
+      "epoch": 5.34,
+      "grad_norm": 4.15360689163208,
+      "learning_rate": 7.33e-06,
+      "loss": 0.468,
+      "step": 1068
+    },
+    {
+      "epoch": 5.345,
+      "grad_norm": 5.678742408752441,
+      "learning_rate": 7.3275000000000006e-06,
+      "loss": 1.1706,
+      "step": 1069
+    },
+    {
+      "epoch": 5.35,
+      "grad_norm": 3.9543099403381348,
+      "learning_rate": 7.325000000000001e-06,
+      "loss": 0.564,
+      "step": 1070
+    },
+    {
+      "epoch": 5.355,
+      "grad_norm": 7.467982292175293,
+      "learning_rate": 7.3225e-06,
+      "loss": 0.7217,
+      "step": 1071
+    },
+    {
+      "epoch": 5.36,
+      "grad_norm": 7.043073654174805,
+      "learning_rate": 7.32e-06,
+      "loss": 0.4793,
+      "step": 1072
+    },
+    {
+      "epoch": 5.365,
+      "grad_norm": 5.870516300201416,
+      "learning_rate": 7.317500000000001e-06,
+      "loss": 0.4155,
+      "step": 1073
+    },
+    {
+      "epoch": 5.37,
+      "grad_norm": 5.947883129119873,
+      "learning_rate": 7.315000000000001e-06,
+      "loss": 0.69,
+      "step": 1074
+    },
+    {
+      "epoch": 5.375,
+      "grad_norm": 5.8306145668029785,
+      "learning_rate": 7.3125e-06,
+      "loss": 0.7111,
+      "step": 1075
+    },
+    {
+      "epoch": 5.38,
+      "grad_norm": 6.215963840484619,
+      "learning_rate": 7.31e-06,
+      "loss": 1.0054,
+      "step": 1076
+    },
+    {
+      "epoch": 5.385,
+      "grad_norm": 5.142808437347412,
+      "learning_rate": 7.307500000000001e-06,
+      "loss": 0.7519,
+      "step": 1077
+    },
+    {
+      "epoch": 5.39,
+      "grad_norm": 8.308109283447266,
+      "learning_rate": 7.305000000000001e-06,
+      "loss": 0.6698,
+      "step": 1078
+    },
+    {
+      "epoch": 5.395,
+      "grad_norm": 3.213364839553833,
+      "learning_rate": 7.3025e-06,
+      "loss": 0.1693,
+      "step": 1079
+    },
+    {
+      "epoch": 5.4,
+      "grad_norm": 8.754572868347168,
+      "learning_rate": 7.3e-06,
+      "loss": 1.0243,
+      "step": 1080
+    },
+    {
+      "epoch": 5.405,
+      "grad_norm": 7.617823600769043,
+      "learning_rate": 7.297500000000001e-06,
+      "loss": 1.0753,
+      "step": 1081
+    },
+    {
+      "epoch": 5.41,
+      "grad_norm": 5.803887367248535,
+      "learning_rate": 7.295000000000001e-06,
+      "loss": 0.9378,
+      "step": 1082
+    },
+    {
+      "epoch": 5.415,
+      "grad_norm": 5.234251022338867,
+      "learning_rate": 7.2925e-06,
+      "loss": 0.2593,
+      "step": 1083
+    },
+    {
+      "epoch": 5.42,
+      "grad_norm": 5.679897785186768,
+      "learning_rate": 7.2900000000000005e-06,
+      "loss": 0.4513,
+      "step": 1084
+    },
+    {
+      "epoch": 5.425,
+      "grad_norm": 5.814553260803223,
+      "learning_rate": 7.287500000000001e-06,
+      "loss": 0.5886,
+      "step": 1085
+    },
+    {
+      "epoch": 5.43,
+      "grad_norm": 3.6853599548339844,
+      "learning_rate": 7.2850000000000006e-06,
+      "loss": 0.5533,
+      "step": 1086
+    },
+    {
+      "epoch": 5.435,
+      "grad_norm": 5.131886005401611,
+      "learning_rate": 7.2825e-06,
+      "loss": 0.4354,
+      "step": 1087
+    },
+    {
+      "epoch": 5.44,
+      "grad_norm": 5.192765712738037,
+      "learning_rate": 7.280000000000001e-06,
+      "loss": 0.516,
+      "step": 1088
+    },
+    {
+      "epoch": 5.445,
+      "grad_norm": 8.660066604614258,
+      "learning_rate": 7.277500000000001e-06,
+      "loss": 0.9384,
+      "step": 1089
+    },
+    {
+      "epoch": 5.45,
+      "grad_norm": 4.082743167877197,
+      "learning_rate": 7.275000000000001e-06,
+      "loss": 0.4671,
+      "step": 1090
+    },
+    {
+      "epoch": 5.455,
+      "grad_norm": 6.213125705718994,
+      "learning_rate": 7.2725e-06,
+      "loss": 0.739,
+      "step": 1091
+    },
+    {
+      "epoch": 5.46,
+      "grad_norm": 5.036574840545654,
+      "learning_rate": 7.270000000000001e-06,
+      "loss": 0.5466,
+      "step": 1092
+    },
+    {
+      "epoch": 5.465,
+      "grad_norm": 9.438892364501953,
+      "learning_rate": 7.2675e-06,
+      "loss": 0.9895,
+      "step": 1093
+    },
+    {
+      "epoch": 5.47,
+      "grad_norm": 4.705496788024902,
+      "learning_rate": 7.265000000000001e-06,
+      "loss": 0.7294,
+      "step": 1094
+    },
+    {
+      "epoch": 5.475,
+      "grad_norm": 5.950634002685547,
+      "learning_rate": 7.2625e-06,
+      "loss": 0.789,
+      "step": 1095
+    },
+    {
+      "epoch": 5.48,
+      "grad_norm": 3.5333566665649414,
+      "learning_rate": 7.260000000000001e-06,
+      "loss": 0.3826,
+      "step": 1096
+    },
+    {
+      "epoch": 5.485,
+      "grad_norm": 5.232617378234863,
+      "learning_rate": 7.2575e-06,
+      "loss": 0.5171,
+      "step": 1097
+    },
+    {
+      "epoch": 5.49,
+      "grad_norm": 5.068689823150635,
+      "learning_rate": 7.255000000000001e-06,
+      "loss": 1.2072,
+      "step": 1098
+    },
+    {
+      "epoch": 5.495,
+      "grad_norm": 10.099414825439453,
+      "learning_rate": 7.2525000000000004e-06,
+      "loss": 0.8766,
+      "step": 1099
+    },
+    {
+      "epoch": 5.5,
+      "grad_norm": 4.5145158767700195,
+      "learning_rate": 7.25e-06,
+      "loss": 0.4547,
+      "step": 1100
+    },
+    {
+      "epoch": 5.505,
+      "grad_norm": 4.849261283874512,
+      "learning_rate": 7.2475000000000005e-06,
+      "loss": 0.324,
+      "step": 1101
+    },
+    {
+      "epoch": 5.51,
+      "grad_norm": 6.329558849334717,
+      "learning_rate": 7.245000000000001e-06,
+      "loss": 0.6101,
+      "step": 1102
+    },
+    {
+      "epoch": 5.515,
+      "grad_norm": 5.531838417053223,
+      "learning_rate": 7.2425e-06,
+      "loss": 0.4218,
+      "step": 1103
+    },
+    {
+      "epoch": 5.52,
+      "grad_norm": 6.789201736450195,
+      "learning_rate": 7.24e-06,
+      "loss": 0.4754,
+      "step": 1104
+    },
+    {
+      "epoch": 5.525,
+      "grad_norm": 18.579917907714844,
+      "learning_rate": 7.237500000000001e-06,
+      "loss": 0.6409,
+      "step": 1105
+    },
+    {
+      "epoch": 5.53,
+      "grad_norm": 23.4625186920166,
+      "learning_rate": 7.235000000000001e-06,
+      "loss": 0.643,
+      "step": 1106
+    },
+    {
+      "epoch": 5.535,
+      "grad_norm": 8.61853313446045,
+      "learning_rate": 7.2325e-06,
+      "loss": 0.8419,
+      "step": 1107
+    },
+    {
+      "epoch": 5.54,
+      "grad_norm": 5.996244430541992,
+      "learning_rate": 7.23e-06,
+      "loss": 0.6525,
+      "step": 1108
+    },
+    {
+      "epoch": 5.545,
+      "grad_norm": 7.86932373046875,
+      "learning_rate": 7.227500000000001e-06,
+      "loss": 0.7684,
+      "step": 1109
+    },
+    {
+      "epoch": 5.55,
+      "grad_norm": 8.651813507080078,
+      "learning_rate": 7.225000000000001e-06,
+      "loss": 0.5859,
+      "step": 1110
+    },
+    {
+      "epoch": 5.555,
+      "grad_norm": 12.120469093322754,
+      "learning_rate": 7.2225e-06,
+      "loss": 0.7529,
+      "step": 1111
+    },
+    {
+      "epoch": 5.5600000000000005,
+      "grad_norm": 3.6932249069213867,
+      "learning_rate": 7.22e-06,
+      "loss": 0.4418,
+      "step": 1112
+    },
+    {
+      "epoch": 5.5649999999999995,
+      "grad_norm": 9.313931465148926,
+      "learning_rate": 7.217500000000001e-06,
+      "loss": 1.0561,
+      "step": 1113
+    },
+    {
+      "epoch": 5.57,
+      "grad_norm": 6.078246116638184,
+      "learning_rate": 7.215000000000001e-06,
+      "loss": 0.5849,
+      "step": 1114
+    },
+    {
+      "epoch": 5.575,
+      "grad_norm": 4.7735490798950195,
+      "learning_rate": 7.2125e-06,
+      "loss": 0.4835,
+      "step": 1115
+    },
+    {
+      "epoch": 5.58,
+      "grad_norm": 4.161695957183838,
+      "learning_rate": 7.2100000000000004e-06,
+      "loss": 0.3342,
+      "step": 1116
+    },
+    {
+      "epoch": 5.585,
+      "grad_norm": 4.655792236328125,
+      "learning_rate": 7.207500000000001e-06,
+      "loss": 0.4437,
+      "step": 1117
+    },
+    {
+      "epoch": 5.59,
+      "grad_norm": 3.9782421588897705,
+      "learning_rate": 7.2050000000000005e-06,
+      "loss": 0.4474,
+      "step": 1118
+    },
+    {
+      "epoch": 5.595,
+      "grad_norm": 7.730457782745361,
+      "learning_rate": 7.2025e-06,
+      "loss": 0.671,
+      "step": 1119
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 8.157031059265137,
+      "learning_rate": 7.2000000000000005e-06,
+      "loss": 1.0783,
+      "step": 1120
+    },
+    {
+      "epoch": 5.605,
+      "grad_norm": 6.75458288192749,
+      "learning_rate": 7.197500000000001e-06,
+      "loss": 0.8674,
+      "step": 1121
+    },
+    {
+      "epoch": 5.61,
+      "grad_norm": 6.0980939865112305,
+      "learning_rate": 7.1950000000000006e-06,
+      "loss": 0.826,
+      "step": 1122
+    },
+    {
+      "epoch": 5.615,
+      "grad_norm": 6.093777656555176,
+      "learning_rate": 7.1925e-06,
+      "loss": 0.4686,
+      "step": 1123
+    },
+    {
+      "epoch": 5.62,
+      "grad_norm": 7.29619836807251,
+      "learning_rate": 7.190000000000001e-06,
+      "loss": 0.614,
+      "step": 1124
+    },
+    {
+      "epoch": 5.625,
+      "grad_norm": 6.660518646240234,
+      "learning_rate": 7.1875e-06,
+      "loss": 0.5114,
+      "step": 1125
+    },
+    {
+      "epoch": 5.63,
+      "grad_norm": 5.1991286277771,
+      "learning_rate": 7.185000000000001e-06,
+      "loss": 0.9036,
+      "step": 1126
+    },
+    {
+      "epoch": 5.635,
+      "grad_norm": 6.005958557128906,
+      "learning_rate": 7.182500000000001e-06,
+      "loss": 0.4837,
+      "step": 1127
+    },
+    {
+      "epoch": 5.64,
+      "grad_norm": 8.787395477294922,
+      "learning_rate": 7.180000000000001e-06,
+      "loss": 0.9858,
+      "step": 1128
+    },
+    {
+      "epoch": 5.645,
+      "grad_norm": 7.105536460876465,
+      "learning_rate": 7.1775e-06,
+      "loss": 0.4544,
+      "step": 1129
+    },
+    {
+      "epoch": 5.65,
+      "grad_norm": 6.821907043457031,
+      "learning_rate": 7.175000000000001e-06,
+      "loss": 0.2321,
+      "step": 1130
+    },
+    {
+      "epoch": 5.655,
+      "grad_norm": 6.5886759757995605,
+      "learning_rate": 7.172500000000001e-06,
+      "loss": 0.508,
+      "step": 1131
+    },
+    {
+      "epoch": 5.66,
+      "grad_norm": 7.134653568267822,
+      "learning_rate": 7.17e-06,
+      "loss": 0.6522,
+      "step": 1132
+    },
+    {
+      "epoch": 5.665,
+      "grad_norm": 6.263777732849121,
+      "learning_rate": 7.1675e-06,
+      "loss": 0.8691,
+      "step": 1133
+    },
+    {
+      "epoch": 5.67,
+      "grad_norm": 7.702743053436279,
+      "learning_rate": 7.165000000000001e-06,
+      "loss": 0.775,
+      "step": 1134
+    },
+    {
+      "epoch": 5.675,
+      "grad_norm": 5.8910441398620605,
+      "learning_rate": 7.162500000000001e-06,
+      "loss": 0.8515,
+      "step": 1135
+    },
+    {
+      "epoch": 5.68,
+      "grad_norm": 4.5243377685546875,
+      "learning_rate": 7.16e-06,
+      "loss": 0.5624,
+      "step": 1136
+    },
+    {
+      "epoch": 5.6850000000000005,
+      "grad_norm": 5.971381664276123,
+      "learning_rate": 7.1575000000000005e-06,
+      "loss": 0.5663,
+      "step": 1137
+    },
+    {
+      "epoch": 5.6899999999999995,
+      "grad_norm": 4.770837783813477,
+      "learning_rate": 7.155000000000001e-06,
+      "loss": 0.8121,
+      "step": 1138
+    },
+    {
+      "epoch": 5.695,
+      "grad_norm": 6.018206596374512,
+      "learning_rate": 7.152500000000001e-06,
+      "loss": 0.6482,
+      "step": 1139
+    },
+    {
+      "epoch": 5.7,
+      "grad_norm": 3.8003756999969482,
+      "learning_rate": 7.15e-06,
+      "loss": 0.3197,
+      "step": 1140
+    },
+    {
+      "epoch": 5.705,
+      "grad_norm": 6.136456489562988,
+      "learning_rate": 7.147500000000001e-06,
+      "loss": 0.5703,
+      "step": 1141
+    },
+    {
+      "epoch": 5.71,
+      "grad_norm": 7.011538982391357,
+      "learning_rate": 7.145000000000001e-06,
+      "loss": 0.5211,
+      "step": 1142
+    },
+    {
+      "epoch": 5.715,
+      "grad_norm": 6.128665447235107,
+      "learning_rate": 7.1425000000000015e-06,
+      "loss": 0.535,
+      "step": 1143
+    },
+    {
+      "epoch": 5.72,
+      "grad_norm": 4.94002103805542,
+      "learning_rate": 7.14e-06,
+      "loss": 0.5272,
+      "step": 1144
+    },
+    {
+      "epoch": 5.725,
+      "grad_norm": 107.81997680664062,
+      "learning_rate": 7.137500000000001e-06,
+      "loss": 0.9509,
+      "step": 1145
+    },
+    {
+      "epoch": 5.73,
+      "grad_norm": 36.92516326904297,
+      "learning_rate": 7.135000000000001e-06,
+      "loss": 0.9364,
+      "step": 1146
+    },
+    {
+      "epoch": 5.735,
+      "grad_norm": 4.975857734680176,
+      "learning_rate": 7.132500000000001e-06,
+      "loss": 0.6777,
+      "step": 1147
+    },
+    {
+      "epoch": 5.74,
+      "grad_norm": 3.8433215618133545,
+      "learning_rate": 7.13e-06,
+      "loss": 0.2975,
+      "step": 1148
+    },
+    {
+      "epoch": 5.745,
+      "grad_norm": 12.398770332336426,
+      "learning_rate": 7.127500000000001e-06,
+      "loss": 0.9269,
+      "step": 1149
+    },
+    {
+      "epoch": 5.75,
+      "grad_norm": 6.954431056976318,
+      "learning_rate": 7.125e-06,
+      "loss": 0.8911,
+      "step": 1150
+    },
+    {
+      "epoch": 5.755,
+      "grad_norm": 4.463475227355957,
+      "learning_rate": 7.122500000000001e-06,
+      "loss": 0.3016,
+      "step": 1151
+    },
+    {
+      "epoch": 5.76,
+      "grad_norm": 9.632946014404297,
+      "learning_rate": 7.1200000000000004e-06,
+      "loss": 0.5639,
+      "step": 1152
+    },
+    {
+      "epoch": 5.765,
+      "grad_norm": 53.941036224365234,
+      "learning_rate": 7.117500000000001e-06,
+      "loss": 0.5954,
+      "step": 1153
+    },
+    {
+      "epoch": 5.77,
+      "grad_norm": 3.773770809173584,
+      "learning_rate": 7.1150000000000005e-06,
+      "loss": 0.3921,
+      "step": 1154
+    },
+    {
+      "epoch": 5.775,
+      "grad_norm": 4.56242561340332,
+      "learning_rate": 7.112500000000001e-06,
+      "loss": 0.4858,
+      "step": 1155
+    },
+    {
+      "epoch": 5.78,
+      "grad_norm": 4.446235656738281,
+      "learning_rate": 7.1100000000000005e-06,
+      "loss": 0.4022,
+      "step": 1156
+    },
+    {
+      "epoch": 5.785,
+      "grad_norm": 3.7419331073760986,
+      "learning_rate": 7.1075e-06,
+      "loss": 0.3282,
+      "step": 1157
+    },
+    {
+      "epoch": 5.79,
+      "grad_norm": 8.221663475036621,
+      "learning_rate": 7.105000000000001e-06,
+      "loss": 0.2972,
+      "step": 1158
+    },
+    {
+      "epoch": 5.795,
+      "grad_norm": 5.5685529708862305,
+      "learning_rate": 7.102500000000001e-06,
+      "loss": 0.8065,
+      "step": 1159
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 3.722132444381714,
+      "learning_rate": 7.100000000000001e-06,
+      "loss": 0.45,
+      "step": 1160
+    },
+    {
+      "epoch": 5.805,
+      "grad_norm": 18.364704132080078,
+      "learning_rate": 7.0975e-06,
+      "loss": 0.7593,
+      "step": 1161
+    },
+    {
+      "epoch": 5.8100000000000005,
+      "grad_norm": 5.691896915435791,
+      "learning_rate": 7.095000000000001e-06,
+      "loss": 0.3082,
+      "step": 1162
+    },
+    {
+      "epoch": 5.8149999999999995,
+      "grad_norm": 6.393020153045654,
+      "learning_rate": 7.092500000000001e-06,
+      "loss": 1.0833,
+      "step": 1163
+    },
+    {
+      "epoch": 5.82,
+      "grad_norm": 7.112914085388184,
+      "learning_rate": 7.09e-06,
+      "loss": 0.8307,
+      "step": 1164
+    },
+    {
+      "epoch": 5.825,
+      "grad_norm": 5.782756328582764,
+      "learning_rate": 7.0875e-06,
+      "loss": 0.3478,
+      "step": 1165
+    },
+    {
+      "epoch": 5.83,
+      "grad_norm": 8.843560218811035,
+      "learning_rate": 7.085000000000001e-06,
+      "loss": 0.731,
+      "step": 1166
+    },
+    {
+      "epoch": 5.835,
+      "grad_norm": 4.861911296844482,
+      "learning_rate": 7.082500000000001e-06,
+      "loss": 0.4215,
+      "step": 1167
+    },
+    {
+      "epoch": 5.84,
+      "grad_norm": 4.119502067565918,
+      "learning_rate": 7.08e-06,
+      "loss": 0.3522,
+      "step": 1168
+    },
+    {
+      "epoch": 5.845,
+      "grad_norm": 6.957109451293945,
+      "learning_rate": 7.0775000000000004e-06,
+      "loss": 0.4197,
+      "step": 1169
+    },
+    {
+      "epoch": 5.85,
+      "grad_norm": 4.412446975708008,
+      "learning_rate": 7.075000000000001e-06,
+      "loss": 0.3531,
+      "step": 1170
+    },
+    {
+      "epoch": 5.855,
+      "grad_norm": 6.656820774078369,
+      "learning_rate": 7.072500000000001e-06,
+      "loss": 0.2977,
+      "step": 1171
+    },
+    {
+      "epoch": 5.86,
+      "grad_norm": 4.392847061157227,
+      "learning_rate": 7.07e-06,
+      "loss": 0.3854,
+      "step": 1172
+    },
+    {
+      "epoch": 5.865,
+      "grad_norm": 7.249214172363281,
+      "learning_rate": 7.0675000000000005e-06,
+      "loss": 0.6638,
+      "step": 1173
+    },
+    {
+      "epoch": 5.87,
+      "grad_norm": 13.473212242126465,
+      "learning_rate": 7.065000000000001e-06,
+      "loss": 0.9952,
+      "step": 1174
+    },
+    {
+      "epoch": 5.875,
+      "grad_norm": 6.404993534088135,
+      "learning_rate": 7.062500000000001e-06,
+      "loss": 0.8746,
+      "step": 1175
+    },
+    {
+      "epoch": 5.88,
+      "grad_norm": 6.613409996032715,
+      "learning_rate": 7.06e-06,
+      "loss": 0.5812,
+      "step": 1176
+    },
+    {
+      "epoch": 5.885,
+      "grad_norm": 5.925755977630615,
+      "learning_rate": 7.057500000000001e-06,
+      "loss": 0.5851,
+      "step": 1177
+    },
+    {
+      "epoch": 5.89,
+      "grad_norm": 5.744228839874268,
+      "learning_rate": 7.055000000000001e-06,
+      "loss": 0.4739,
+      "step": 1178
+    },
+    {
+      "epoch": 5.895,
+      "grad_norm": 6.181334495544434,
+      "learning_rate": 7.052500000000001e-06,
+      "loss": 0.532,
+      "step": 1179
+    },
+    {
+      "epoch": 5.9,
+      "grad_norm": 3.225562334060669,
+      "learning_rate": 7.05e-06,
+      "loss": 0.1749,
+      "step": 1180
+    },
+    {
+      "epoch": 5.905,
+      "grad_norm": 5.264067649841309,
+      "learning_rate": 7.047500000000001e-06,
+      "loss": 0.5075,
+      "step": 1181
+    },
+    {
+      "epoch": 5.91,
+      "grad_norm": 285.3383483886719,
+      "learning_rate": 7.045e-06,
+      "loss": 1.9927,
+      "step": 1182
+    },
+    {
+      "epoch": 5.915,
+      "grad_norm": 5.820057392120361,
+      "learning_rate": 7.042500000000001e-06,
+      "loss": 1.2021,
+      "step": 1183
+    },
+    {
+      "epoch": 5.92,
+      "grad_norm": 6.14114236831665,
+      "learning_rate": 7.04e-06,
+      "loss": 0.5488,
+      "step": 1184
+    },
+    {
+      "epoch": 5.925,
+      "grad_norm": 6.812324047088623,
+      "learning_rate": 7.037500000000001e-06,
+      "loss": 1.1204,
+      "step": 1185
+    },
+    {
+      "epoch": 5.93,
+      "grad_norm": 6.617299556732178,
+      "learning_rate": 7.035e-06,
+      "loss": 0.5714,
+      "step": 1186
+    },
+    {
+      "epoch": 5.9350000000000005,
+      "grad_norm": 5.627433776855469,
+      "learning_rate": 7.032500000000001e-06,
+      "loss": 0.6069,
+      "step": 1187
+    },
+    {
+      "epoch": 5.9399999999999995,
+      "grad_norm": 4.453884124755859,
+      "learning_rate": 7.0300000000000005e-06,
+      "loss": 0.4607,
+      "step": 1188
+    },
+    {
+      "epoch": 5.945,
+      "grad_norm": 25.149301528930664,
+      "learning_rate": 7.0275e-06,
+      "loss": 1.4699,
+      "step": 1189
+    },
+    {
+      "epoch": 5.95,
+      "grad_norm": 6.747213363647461,
+      "learning_rate": 7.0250000000000005e-06,
+      "loss": 0.5814,
+      "step": 1190
+    },
+    {
+      "epoch": 5.955,
+      "grad_norm": 5.607445240020752,
+      "learning_rate": 7.022500000000001e-06,
+      "loss": 0.5305,
+      "step": 1191
+    },
+    {
+      "epoch": 5.96,
+      "grad_norm": 5.7789130210876465,
+      "learning_rate": 7.0200000000000006e-06,
+      "loss": 0.4608,
+      "step": 1192
+    },
+    {
+      "epoch": 5.965,
+      "grad_norm": 6.444858074188232,
+      "learning_rate": 7.0175e-06,
+      "loss": 0.4157,
+      "step": 1193
+    },
+    {
+      "epoch": 5.97,
+      "grad_norm": 4.286928176879883,
+      "learning_rate": 7.015000000000001e-06,
+      "loss": 0.5013,
+      "step": 1194
+    },
+    {
+      "epoch": 5.975,
+      "grad_norm": 5.531680107116699,
+      "learning_rate": 7.012500000000001e-06,
+      "loss": 0.6023,
+      "step": 1195
+    },
+    {
+      "epoch": 5.98,
+      "grad_norm": 5.418078422546387,
+      "learning_rate": 7.01e-06,
+      "loss": 0.519,
+      "step": 1196
+    },
+    {
+      "epoch": 5.985,
+      "grad_norm": 4.688261032104492,
+      "learning_rate": 7.0075e-06,
+      "loss": 0.525,
+      "step": 1197
+    },
+    {
+      "epoch": 5.99,
+      "grad_norm": 8.170825958251953,
+      "learning_rate": 7.005000000000001e-06,
+      "loss": 1.2872,
+      "step": 1198
+    },
+    {
+      "epoch": 5.995,
+      "grad_norm": 5.593966960906982,
+      "learning_rate": 7.002500000000001e-06,
+      "loss": 0.4296,
+      "step": 1199
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 12.297003746032715,
+      "learning_rate": 7e-06,
+      "loss": 0.4436,
+      "step": 1200
+    },
+    {
+      "epoch": 6.0,
+      "eval_loss": 0.3700045049190521,
+      "eval_runtime": 15.2102,
+      "eval_samples_per_second": 52.465,
+      "eval_steps_per_second": 6.575,
+      "step": 1200
+    },
+    {
+      "epoch": 6.005,
+      "grad_norm": 4.4825310707092285,
+      "learning_rate": 6.9975e-06,
+      "loss": 0.5744,
+      "step": 1201
+    },
+    {
+      "epoch": 6.01,
+      "grad_norm": 4.36711311340332,
+      "learning_rate": 6.995000000000001e-06,
+      "loss": 0.417,
+      "step": 1202
+    },
+    {
+      "epoch": 6.015,
+      "grad_norm": 5.509248733520508,
+      "learning_rate": 6.992500000000001e-06,
+      "loss": 0.3878,
+      "step": 1203
+    },
+    {
+      "epoch": 6.02,
+      "grad_norm": 5.716054439544678,
+      "learning_rate": 6.99e-06,
+      "loss": 0.4578,
+      "step": 1204
+    },
+    {
+      "epoch": 6.025,
+      "grad_norm": 3.5962581634521484,
+      "learning_rate": 6.9875000000000004e-06,
+      "loss": 0.2228,
+      "step": 1205
+    },
+    {
+      "epoch": 6.03,
+      "grad_norm": 10.339753150939941,
+      "learning_rate": 6.985000000000001e-06,
+      "loss": 0.5625,
+      "step": 1206
+    },
+    {
+      "epoch": 6.035,
+      "grad_norm": 4.431244373321533,
+      "learning_rate": 6.982500000000001e-06,
+      "loss": 0.2803,
+      "step": 1207
+    },
+    {
+      "epoch": 6.04,
+      "grad_norm": 5.4503493309021,
+      "learning_rate": 6.98e-06,
+      "loss": 0.5286,
+      "step": 1208
+    },
+    {
+      "epoch": 6.045,
+      "grad_norm": 25.430767059326172,
+      "learning_rate": 6.9775000000000005e-06,
+      "loss": 0.6297,
+      "step": 1209
+    },
+    {
+      "epoch": 6.05,
+      "grad_norm": 4.579400539398193,
+      "learning_rate": 6.975000000000001e-06,
+      "loss": 0.4598,
+      "step": 1210
+    },
+    {
+      "epoch": 6.055,
+      "grad_norm": 6.865296363830566,
+      "learning_rate": 6.972500000000001e-06,
+      "loss": 0.5401,
+      "step": 1211
+    },
+    {
+      "epoch": 6.06,
+      "grad_norm": 5.276162624359131,
+      "learning_rate": 6.97e-06,
+      "loss": 0.4061,
+      "step": 1212
+    },
+    {
+      "epoch": 6.065,
+      "grad_norm": 6.396225929260254,
+      "learning_rate": 6.967500000000001e-06,
+      "loss": 0.4804,
+      "step": 1213
+    },
+    {
+      "epoch": 6.07,
+      "grad_norm": 3.600400686264038,
+      "learning_rate": 6.965e-06,
+      "loss": 0.5391,
+      "step": 1214
+    },
+    {
+      "epoch": 6.075,
+      "grad_norm": 9.624923706054688,
+      "learning_rate": 6.962500000000001e-06,
+      "loss": 0.2523,
+      "step": 1215
+    },
+    {
+      "epoch": 6.08,
+      "grad_norm": 6.754732131958008,
+      "learning_rate": 6.96e-06,
+      "loss": 0.5905,
+      "step": 1216
+    },
+    {
+      "epoch": 6.085,
+      "grad_norm": 7.601484775543213,
+      "learning_rate": 6.957500000000001e-06,
+      "loss": 0.4264,
+      "step": 1217
+    },
+    {
+      "epoch": 6.09,
+      "grad_norm": 7.815075397491455,
+      "learning_rate": 6.955e-06,
+      "loss": 0.7034,
+      "step": 1218
+    },
+    {
+      "epoch": 6.095,
+      "grad_norm": 6.407260894775391,
+      "learning_rate": 6.952500000000001e-06,
+      "loss": 0.4205,
+      "step": 1219
+    },
+    {
+      "epoch": 6.1,
+      "grad_norm": 6.0509419441223145,
+      "learning_rate": 6.95e-06,
+      "loss": 0.4166,
+      "step": 1220
+    },
+    {
+      "epoch": 6.105,
+      "grad_norm": 5.910727500915527,
+      "learning_rate": 6.9475e-06,
+      "loss": 0.7588,
+      "step": 1221
+    },
+    {
+      "epoch": 6.11,
+      "grad_norm": 5.999612331390381,
+      "learning_rate": 6.945e-06,
+      "loss": 0.3761,
+      "step": 1222
+    },
+    {
+      "epoch": 6.115,
+      "grad_norm": 7.258251667022705,
+      "learning_rate": 6.942500000000001e-06,
+      "loss": 0.4247,
+      "step": 1223
+    },
+    {
+      "epoch": 6.12,
+      "grad_norm": 41.502357482910156,
+      "learning_rate": 6.9400000000000005e-06,
+      "loss": 1.1388,
+      "step": 1224
+    },
+    {
+      "epoch": 6.125,
+      "grad_norm": 11.95021915435791,
+      "learning_rate": 6.9375e-06,
+      "loss": 0.9683,
+      "step": 1225
+    },
+    {
+      "epoch": 6.13,
+      "grad_norm": 4.492338180541992,
+      "learning_rate": 6.9350000000000005e-06,
+      "loss": 0.7903,
+      "step": 1226
+    },
+    {
+      "epoch": 6.135,
+      "grad_norm": 11.921549797058105,
+      "learning_rate": 6.932500000000001e-06,
+      "loss": 0.498,
+      "step": 1227
+    },
+    {
+      "epoch": 6.14,
+      "grad_norm": 3.1085989475250244,
+      "learning_rate": 6.93e-06,
+      "loss": 0.2042,
+      "step": 1228
+    },
+    {
+      "epoch": 6.145,
+      "grad_norm": 7.8389363288879395,
+      "learning_rate": 6.9275e-06,
+      "loss": 0.7476,
+      "step": 1229
+    },
+    {
+      "epoch": 6.15,
+      "grad_norm": 8.418336868286133,
+      "learning_rate": 6.925000000000001e-06,
+      "loss": 0.8947,
+      "step": 1230
+    },
+    {
+      "epoch": 6.155,
+      "grad_norm": 6.933560371398926,
+      "learning_rate": 6.922500000000001e-06,
+      "loss": 0.3881,
+      "step": 1231
+    },
+    {
+      "epoch": 6.16,
+      "grad_norm": 5.576759338378906,
+      "learning_rate": 6.92e-06,
+      "loss": 0.3986,
+      "step": 1232
+    },
+    {
+      "epoch": 6.165,
+      "grad_norm": 8.747431755065918,
+      "learning_rate": 6.9175e-06,
+      "loss": 0.6992,
+      "step": 1233
+    },
+    {
+      "epoch": 6.17,
+      "grad_norm": 5.323822498321533,
+      "learning_rate": 6.915000000000001e-06,
+      "loss": 0.566,
+      "step": 1234
+    },
+    {
+      "epoch": 6.175,
+      "grad_norm": 34.813690185546875,
+      "learning_rate": 6.912500000000001e-06,
+      "loss": 0.4627,
+      "step": 1235
+    },
+    {
+      "epoch": 6.18,
+      "grad_norm": 4.955639362335205,
+      "learning_rate": 6.91e-06,
+      "loss": 0.5486,
+      "step": 1236
+    },
+    {
+      "epoch": 6.185,
+      "grad_norm": 4.0058393478393555,
+      "learning_rate": 6.9075e-06,
+      "loss": 0.5378,
+      "step": 1237
+    },
+    {
+      "epoch": 6.19,
+      "grad_norm": 3.7979302406311035,
+      "learning_rate": 6.905000000000001e-06,
+      "loss": 0.6864,
+      "step": 1238
+    },
+    {
+      "epoch": 6.195,
+      "grad_norm": 4.497585296630859,
+      "learning_rate": 6.902500000000001e-06,
+      "loss": 0.4184,
+      "step": 1239
+    },
+    {
+      "epoch": 6.2,
+      "grad_norm": 4.18954610824585,
+      "learning_rate": 6.9e-06,
+      "loss": 0.6083,
+      "step": 1240
+    },
+    {
+      "epoch": 6.205,
+      "grad_norm": 6.625578880310059,
+      "learning_rate": 6.8975000000000005e-06,
+      "loss": 0.897,
+      "step": 1241
+    },
+    {
+      "epoch": 6.21,
+      "grad_norm": 4.3530497550964355,
+      "learning_rate": 6.895000000000001e-06,
+      "loss": 0.8167,
+      "step": 1242
+    },
+    {
+      "epoch": 6.215,
+      "grad_norm": 5.458725929260254,
+      "learning_rate": 6.8925000000000005e-06,
+      "loss": 0.356,
+      "step": 1243
+    },
+    {
+      "epoch": 6.22,
+      "grad_norm": 6.803143501281738,
+      "learning_rate": 6.89e-06,
+      "loss": 0.2914,
+      "step": 1244
+    },
+    {
+      "epoch": 6.225,
+      "grad_norm": 4.613215446472168,
+      "learning_rate": 6.8875000000000005e-06,
+      "loss": 0.4464,
+      "step": 1245
+    },
+    {
+      "epoch": 6.23,
+      "grad_norm": 8.706917762756348,
+      "learning_rate": 6.885e-06,
+      "loss": 0.8221,
+      "step": 1246
+    },
+    {
+      "epoch": 6.235,
+      "grad_norm": 4.9474287033081055,
+      "learning_rate": 6.882500000000001e-06,
+      "loss": 0.7083,
+      "step": 1247
+    },
+    {
+      "epoch": 6.24,
+      "grad_norm": 7.1159210205078125,
+      "learning_rate": 6.88e-06,
+      "loss": 0.4226,
+      "step": 1248
+    },
+    {
+      "epoch": 6.245,
+      "grad_norm": 5.2131028175354,
+      "learning_rate": 6.877500000000001e-06,
+      "loss": 0.4464,
+      "step": 1249
+    },
+    {
+      "epoch": 6.25,
+      "grad_norm": 8.915253639221191,
+      "learning_rate": 6.875e-06,
+      "loss": 0.6195,
+      "step": 1250
+    },
+    {
+      "epoch": 6.255,
+      "grad_norm": 6.6033034324646,
+      "learning_rate": 6.872500000000001e-06,
+      "loss": 0.6456,
+      "step": 1251
+    },
+    {
+      "epoch": 6.26,
+      "grad_norm": 5.440224647521973,
+      "learning_rate": 6.870000000000001e-06,
+      "loss": 0.6638,
+      "step": 1252
+    },
+    {
+      "epoch": 6.265,
+      "grad_norm": 3.549628257751465,
+      "learning_rate": 6.8675e-06,
+      "loss": 0.3291,
+      "step": 1253
+    },
+    {
+      "epoch": 6.27,
+      "grad_norm": 7.653029918670654,
+      "learning_rate": 6.865e-06,
+      "loss": 0.6568,
+      "step": 1254
+    },
+    {
+      "epoch": 6.275,
+      "grad_norm": 4.8423590660095215,
+      "learning_rate": 6.862500000000001e-06,
+      "loss": 0.4717,
+      "step": 1255
+    },
+    {
+      "epoch": 6.28,
+      "grad_norm": 4.571013450622559,
+      "learning_rate": 6.860000000000001e-06,
+      "loss": 0.8792,
+      "step": 1256
+    },
+    {
+      "epoch": 6.285,
+      "grad_norm": 8.791017532348633,
+      "learning_rate": 6.8575e-06,
+      "loss": 0.4477,
+      "step": 1257
+    },
+    {
+      "epoch": 6.29,
+      "grad_norm": 6.025150299072266,
+      "learning_rate": 6.8550000000000004e-06,
+      "loss": 0.5234,
+      "step": 1258
+    },
+    {
+      "epoch": 6.295,
+      "grad_norm": 4.552408218383789,
+      "learning_rate": 6.852500000000001e-06,
+      "loss": 0.7049,
+      "step": 1259
+    },
+    {
+      "epoch": 6.3,
+      "grad_norm": 5.322471618652344,
+      "learning_rate": 6.850000000000001e-06,
+      "loss": 0.3872,
+      "step": 1260
+    },
+    {
+      "epoch": 6.305,
+      "grad_norm": 4.06843376159668,
+      "learning_rate": 6.8475e-06,
+      "loss": 0.4771,
+      "step": 1261
+    },
+    {
+      "epoch": 6.31,
+      "grad_norm": 5.422154426574707,
+      "learning_rate": 6.8450000000000005e-06,
+      "loss": 0.2992,
+      "step": 1262
+    },
+    {
+      "epoch": 6.315,
+      "grad_norm": 5.077361106872559,
+      "learning_rate": 6.842500000000001e-06,
+      "loss": 0.3538,
+      "step": 1263
+    },
+    {
+      "epoch": 6.32,
+      "grad_norm": 3.5245907306671143,
+      "learning_rate": 6.8400000000000014e-06,
+      "loss": 0.2814,
+      "step": 1264
+    },
+    {
+      "epoch": 6.325,
+      "grad_norm": 4.9270100593566895,
+      "learning_rate": 6.8375e-06,
+      "loss": 0.5251,
+      "step": 1265
+    },
+    {
+      "epoch": 6.33,
+      "grad_norm": 3.9920599460601807,
+      "learning_rate": 6.835000000000001e-06,
+      "loss": 0.2506,
+      "step": 1266
+    },
+    {
+      "epoch": 6.335,
+      "grad_norm": 5.97065544128418,
+      "learning_rate": 6.832500000000001e-06,
+      "loss": 0.6823,
+      "step": 1267
+    },
+    {
+      "epoch": 6.34,
+      "grad_norm": 4.389867305755615,
+      "learning_rate": 6.830000000000001e-06,
+      "loss": 0.5607,
+      "step": 1268
+    },
+    {
+      "epoch": 6.345,
+      "grad_norm": 8.95691204071045,
+      "learning_rate": 6.8275e-06,
+      "loss": 0.8179,
+      "step": 1269
+    },
+    {
+      "epoch": 6.35,
+      "grad_norm": 4.985533237457275,
+      "learning_rate": 6.825000000000001e-06,
+      "loss": 0.3958,
+      "step": 1270
+    },
+    {
+      "epoch": 6.355,
+      "grad_norm": 7.16851806640625,
+      "learning_rate": 6.822500000000001e-06,
+      "loss": 0.9484,
+      "step": 1271
+    },
+    {
+      "epoch": 6.36,
+      "grad_norm": 3.7318403720855713,
+      "learning_rate": 6.820000000000001e-06,
+      "loss": 0.2293,
+      "step": 1272
+    },
+    {
+      "epoch": 6.365,
+      "grad_norm": 4.980233192443848,
+      "learning_rate": 6.8175e-06,
+      "loss": 0.8519,
+      "step": 1273
+    },
+    {
+      "epoch": 6.37,
+      "grad_norm": 5.376368999481201,
+      "learning_rate": 6.815000000000001e-06,
+      "loss": 0.3087,
+      "step": 1274
+    },
+    {
+      "epoch": 6.375,
+      "grad_norm": 4.691337585449219,
+      "learning_rate": 6.8125e-06,
+      "loss": 0.3766,
+      "step": 1275
+    },
+    {
+      "epoch": 6.38,
+      "grad_norm": 5.862613201141357,
+      "learning_rate": 6.810000000000001e-06,
+      "loss": 0.7366,
+      "step": 1276
+    },
+    {
+      "epoch": 6.385,
+      "grad_norm": 3.8882408142089844,
+      "learning_rate": 6.8075000000000005e-06,
+      "loss": 0.4114,
+      "step": 1277
+    },
+    {
+      "epoch": 6.39,
+      "grad_norm": 4.328252792358398,
+      "learning_rate": 6.805000000000001e-06,
+      "loss": 0.5976,
+      "step": 1278
+    },
+    {
+      "epoch": 6.395,
+      "grad_norm": 6.559730052947998,
+      "learning_rate": 6.8025000000000005e-06,
+      "loss": 0.6406,
+      "step": 1279
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 5.585719108581543,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 0.7763,
+      "step": 1280
+    },
+    {
+      "epoch": 6.405,
+      "grad_norm": 4.51055383682251,
+      "learning_rate": 6.7975000000000006e-06,
+      "loss": 0.5034,
+      "step": 1281
+    },
+    {
+      "epoch": 6.41,
+      "grad_norm": 5.906167507171631,
+      "learning_rate": 6.795e-06,
+      "loss": 0.8515,
+      "step": 1282
+    },
+    {
+      "epoch": 6.415,
+      "grad_norm": 4.232546806335449,
+      "learning_rate": 6.792500000000001e-06,
+      "loss": 0.3643,
+      "step": 1283
+    },
+    {
+      "epoch": 6.42,
+      "grad_norm": 9.062664985656738,
+      "learning_rate": 6.790000000000001e-06,
+      "loss": 1.071,
+      "step": 1284
+    },
+    {
+      "epoch": 6.425,
+      "grad_norm": 4.274650573730469,
+      "learning_rate": 6.7875e-06,
+      "loss": 0.6813,
+      "step": 1285
+    },
+    {
+      "epoch": 6.43,
+      "grad_norm": 5.163512229919434,
+      "learning_rate": 6.785e-06,
+      "loss": 0.3241,
+      "step": 1286
+    },
+    {
+      "epoch": 6.435,
+      "grad_norm": 7.453189373016357,
+      "learning_rate": 6.782500000000001e-06,
+      "loss": 0.6187,
+      "step": 1287
+    },
+    {
+      "epoch": 6.44,
+      "grad_norm": 8.773122787475586,
+      "learning_rate": 6.780000000000001e-06,
+      "loss": 0.586,
+      "step": 1288
+    },
+    {
+      "epoch": 6.445,
+      "grad_norm": 5.43891716003418,
+      "learning_rate": 6.7775e-06,
+      "loss": 1.0809,
+      "step": 1289
+    },
+    {
+      "epoch": 6.45,
+      "grad_norm": 6.651701927185059,
+      "learning_rate": 6.775e-06,
+      "loss": 0.4934,
+      "step": 1290
+    },
+    {
+      "epoch": 6.455,
+      "grad_norm": 7.890969753265381,
+      "learning_rate": 6.772500000000001e-06,
+      "loss": 0.8226,
+      "step": 1291
+    },
+    {
+      "epoch": 6.46,
+      "grad_norm": 63.54376220703125,
+      "learning_rate": 6.770000000000001e-06,
+      "loss": 0.6353,
+      "step": 1292
+    },
+    {
+      "epoch": 6.465,
+      "grad_norm": 5.521801948547363,
+      "learning_rate": 6.7675e-06,
+      "loss": 0.3087,
+      "step": 1293
+    },
+    {
+      "epoch": 6.47,
+      "grad_norm": 7.32753324508667,
+      "learning_rate": 6.7650000000000005e-06,
+      "loss": 0.5225,
+      "step": 1294
+    },
+    {
+      "epoch": 6.475,
+      "grad_norm": 7.318142890930176,
+      "learning_rate": 6.762500000000001e-06,
+      "loss": 0.5271,
+      "step": 1295
+    },
+    {
+      "epoch": 6.48,
+      "grad_norm": 5.703019142150879,
+      "learning_rate": 6.760000000000001e-06,
+      "loss": 0.4873,
+      "step": 1296
+    },
+    {
+      "epoch": 6.485,
+      "grad_norm": 7.614542484283447,
+      "learning_rate": 6.7575e-06,
+      "loss": 0.7169,
+      "step": 1297
+    },
+    {
+      "epoch": 6.49,
+      "grad_norm": 18.971420288085938,
+      "learning_rate": 6.7550000000000005e-06,
+      "loss": 0.51,
+      "step": 1298
+    },
+    {
+      "epoch": 6.495,
+      "grad_norm": 7.618166446685791,
+      "learning_rate": 6.752500000000001e-06,
+      "loss": 0.5645,
+      "step": 1299
+    },
+    {
+      "epoch": 6.5,
+      "grad_norm": 5.883105278015137,
+      "learning_rate": 6.750000000000001e-06,
+      "loss": 0.6861,
+      "step": 1300
+    },
+    {
+      "epoch": 6.505,
+      "grad_norm": 4.796202659606934,
+      "learning_rate": 6.7475e-06,
+      "loss": 0.464,
+      "step": 1301
+    },
+    {
+      "epoch": 6.51,
+      "grad_norm": 5.762648105621338,
+      "learning_rate": 6.745000000000001e-06,
+      "loss": 0.5118,
+      "step": 1302
+    },
+    {
+      "epoch": 6.515,
+      "grad_norm": 4.848067760467529,
+      "learning_rate": 6.742500000000001e-06,
+      "loss": 0.5312,
+      "step": 1303
+    },
+    {
+      "epoch": 6.52,
+      "grad_norm": 4.6505255699157715,
+      "learning_rate": 6.740000000000001e-06,
+      "loss": 0.1317,
+      "step": 1304
+    },
+    {
+      "epoch": 6.525,
+      "grad_norm": 9.25706672668457,
+      "learning_rate": 6.7375e-06,
+      "loss": 0.7033,
+      "step": 1305
+    },
+    {
+      "epoch": 6.53,
+      "grad_norm": 4.706851482391357,
+      "learning_rate": 6.735000000000001e-06,
+      "loss": 0.524,
+      "step": 1306
+    },
+    {
+      "epoch": 6.535,
+      "grad_norm": 17.441757202148438,
+      "learning_rate": 6.7325e-06,
+      "loss": 0.9961,
+      "step": 1307
+    },
+    {
+      "epoch": 6.54,
+      "grad_norm": 5.485891819000244,
+      "learning_rate": 6.730000000000001e-06,
+      "loss": 0.5555,
+      "step": 1308
+    },
+    {
+      "epoch": 6.545,
+      "grad_norm": 10.075560569763184,
+      "learning_rate": 6.7275e-06,
+      "loss": 0.709,
+      "step": 1309
+    },
+    {
+      "epoch": 6.55,
+      "grad_norm": 6.434652805328369,
+      "learning_rate": 6.725000000000001e-06,
+      "loss": 0.6089,
+      "step": 1310
+    },
+    {
+      "epoch": 6.555,
+      "grad_norm": 3.540315628051758,
+      "learning_rate": 6.7225000000000004e-06,
+      "loss": 0.4988,
+      "step": 1311
+    },
+    {
+      "epoch": 6.5600000000000005,
+      "grad_norm": 5.8414411544799805,
+      "learning_rate": 6.720000000000001e-06,
+      "loss": 0.7752,
+      "step": 1312
+    },
+    {
+      "epoch": 6.5649999999999995,
+      "grad_norm": 6.117225646972656,
+      "learning_rate": 6.7175000000000005e-06,
+      "loss": 0.5134,
+      "step": 1313
+    },
+    {
+      "epoch": 6.57,
+      "grad_norm": 4.164860725402832,
+      "learning_rate": 6.715e-06,
+      "loss": 0.3607,
+      "step": 1314
+    },
+    {
+      "epoch": 6.575,
+      "grad_norm": 3.228580951690674,
+      "learning_rate": 6.7125000000000005e-06,
+      "loss": 0.3495,
+      "step": 1315
+    },
+    {
+      "epoch": 6.58,
+      "grad_norm": 10.449295997619629,
+      "learning_rate": 6.710000000000001e-06,
+      "loss": 0.6086,
+      "step": 1316
+    },
+    {
+      "epoch": 6.585,
+      "grad_norm": 3.810534715652466,
+      "learning_rate": 6.7075e-06,
+      "loss": 0.3982,
+      "step": 1317
+    },
+    {
+      "epoch": 6.59,
+      "grad_norm": 5.032525539398193,
+      "learning_rate": 6.705e-06,
+      "loss": 0.4408,
+      "step": 1318
+    },
+    {
+      "epoch": 6.595,
+      "grad_norm": 4.009894371032715,
+      "learning_rate": 6.702500000000001e-06,
+      "loss": 0.2333,
+      "step": 1319
+    },
+    {
+      "epoch": 6.6,
+      "grad_norm": 5.812589645385742,
+      "learning_rate": 6.700000000000001e-06,
+      "loss": 0.8024,
+      "step": 1320
+    },
+    {
+      "epoch": 6.605,
+      "grad_norm": 5.306116580963135,
+      "learning_rate": 6.6975e-06,
+      "loss": 0.3349,
+      "step": 1321
+    },
+    {
+      "epoch": 6.61,
+      "grad_norm": 2.7185263633728027,
+      "learning_rate": 6.695e-06,
+      "loss": 0.2003,
+      "step": 1322
+    },
+    {
+      "epoch": 6.615,
+      "grad_norm": 4.070326805114746,
+      "learning_rate": 6.692500000000001e-06,
+      "loss": 0.4373,
+      "step": 1323
+    },
+    {
+      "epoch": 6.62,
+      "grad_norm": 6.27354097366333,
+      "learning_rate": 6.690000000000001e-06,
+      "loss": 0.4974,
+      "step": 1324
+    },
+    {
+      "epoch": 6.625,
+      "grad_norm": 5.582259178161621,
+      "learning_rate": 6.6875e-06,
+      "loss": 0.9021,
+      "step": 1325
+    },
+    {
+      "epoch": 6.63,
+      "grad_norm": 3.4716570377349854,
+      "learning_rate": 6.685e-06,
+      "loss": 0.3994,
+      "step": 1326
+    },
+    {
+      "epoch": 6.635,
+      "grad_norm": 53.414737701416016,
+      "learning_rate": 6.682500000000001e-06,
+      "loss": 0.6932,
+      "step": 1327
+    },
+    {
+      "epoch": 6.64,
+      "grad_norm": 6.558328151702881,
+      "learning_rate": 6.680000000000001e-06,
+      "loss": 0.3318,
+      "step": 1328
+    },
+    {
+      "epoch": 6.645,
+      "grad_norm": 5.8194966316223145,
+      "learning_rate": 6.6775e-06,
+      "loss": 0.1568,
+      "step": 1329
+    },
+    {
+      "epoch": 6.65,
+      "grad_norm": 4.380676746368408,
+      "learning_rate": 6.6750000000000005e-06,
+      "loss": 0.733,
+      "step": 1330
+    },
+    {
+      "epoch": 6.655,
+      "grad_norm": 5.084666728973389,
+      "learning_rate": 6.672500000000001e-06,
+      "loss": 0.4386,
+      "step": 1331
+    },
+    {
+      "epoch": 6.66,
+      "grad_norm": 5.832912921905518,
+      "learning_rate": 6.6700000000000005e-06,
+      "loss": 0.5896,
+      "step": 1332
+    },
+    {
+      "epoch": 6.665,
+      "grad_norm": 6.204337120056152,
+      "learning_rate": 6.6675e-06,
+      "loss": 0.6807,
+      "step": 1333
+    },
+    {
+      "epoch": 6.67,
+      "grad_norm": 4.444660663604736,
+      "learning_rate": 6.6650000000000006e-06,
+      "loss": 0.5509,
+      "step": 1334
+    },
+    {
+      "epoch": 6.675,
+      "grad_norm": 6.577089786529541,
+      "learning_rate": 6.662500000000001e-06,
+      "loss": 0.54,
+      "step": 1335
+    },
+    {
+      "epoch": 6.68,
+      "grad_norm": 4.057696342468262,
+      "learning_rate": 6.660000000000001e-06,
+      "loss": 0.3508,
+      "step": 1336
+    },
+    {
+      "epoch": 6.6850000000000005,
+      "grad_norm": 4.044608116149902,
+      "learning_rate": 6.6575e-06,
+      "loss": 0.5539,
+      "step": 1337
+    },
+    {
+      "epoch": 6.6899999999999995,
+      "grad_norm": 4.7123260498046875,
+      "learning_rate": 6.655000000000001e-06,
+      "loss": 0.5426,
+      "step": 1338
+    },
+    {
+      "epoch": 6.695,
+      "grad_norm": 5.284317493438721,
+      "learning_rate": 6.6525e-06,
+      "loss": 0.3373,
+      "step": 1339
+    },
+    {
+      "epoch": 6.7,
+      "grad_norm": 3.993507146835327,
+      "learning_rate": 6.650000000000001e-06,
+      "loss": 0.4123,
+      "step": 1340
+    },
+    {
+      "epoch": 6.705,
+      "grad_norm": 3.267998456954956,
+      "learning_rate": 6.6475e-06,
+      "loss": 0.1575,
+      "step": 1341
+    },
+    {
+      "epoch": 6.71,
+      "grad_norm": 4.154967308044434,
+      "learning_rate": 6.645000000000001e-06,
+      "loss": 0.5081,
+      "step": 1342
+    },
+    {
+      "epoch": 6.715,
+      "grad_norm": 6.5928874015808105,
+      "learning_rate": 6.6425e-06,
+      "loss": 1.0912,
+      "step": 1343
+    },
+    {
+      "epoch": 6.72,
+      "grad_norm": 3.58756422996521,
+      "learning_rate": 6.640000000000001e-06,
+      "loss": 0.1857,
+      "step": 1344
+    },
+    {
+      "epoch": 6.725,
+      "grad_norm": 5.610130310058594,
+      "learning_rate": 6.6375e-06,
+      "loss": 0.4652,
+      "step": 1345
+    },
+    {
+      "epoch": 6.73,
+      "grad_norm": 3.733987331390381,
+      "learning_rate": 6.635e-06,
+      "loss": 0.3886,
+      "step": 1346
+    },
+    {
+      "epoch": 6.735,
+      "grad_norm": 6.244208812713623,
+      "learning_rate": 6.6325000000000004e-06,
+      "loss": 0.5898,
+      "step": 1347
+    },
+    {
+      "epoch": 6.74,
+      "grad_norm": 2.1795804500579834,
+      "learning_rate": 6.630000000000001e-06,
+      "loss": 0.1677,
+      "step": 1348
+    },
+    {
+      "epoch": 6.745,
+      "grad_norm": 5.421024322509766,
+      "learning_rate": 6.6275e-06,
+      "loss": 0.8756,
+      "step": 1349
+    },
+    {
+      "epoch": 6.75,
+      "grad_norm": 43.66529083251953,
+      "learning_rate": 6.625e-06,
+      "loss": 0.8904,
+      "step": 1350
+    },
+    {
+      "epoch": 6.755,
+      "grad_norm": 4.4996562004089355,
+      "learning_rate": 6.6225000000000005e-06,
+      "loss": 0.3947,
+      "step": 1351
+    },
+    {
+      "epoch": 6.76,
+      "grad_norm": 3.4047160148620605,
+      "learning_rate": 6.620000000000001e-06,
+      "loss": 0.5019,
+      "step": 1352
+    },
+    {
+      "epoch": 6.765,
+      "grad_norm": 4.692281723022461,
+      "learning_rate": 6.6175e-06,
+      "loss": 0.627,
+      "step": 1353
+    },
+    {
+      "epoch": 6.77,
+      "grad_norm": 4.192058086395264,
+      "learning_rate": 6.615e-06,
+      "loss": 0.3971,
+      "step": 1354
+    },
+    {
+      "epoch": 6.775,
+      "grad_norm": 6.020359039306641,
+      "learning_rate": 6.612500000000001e-06,
+      "loss": 0.462,
+      "step": 1355
+    },
+    {
+      "epoch": 6.78,
+      "grad_norm": 6.246492862701416,
+      "learning_rate": 6.610000000000001e-06,
+      "loss": 0.7488,
+      "step": 1356
+    },
+    {
+      "epoch": 6.785,
+      "grad_norm": 7.669323921203613,
+      "learning_rate": 6.6075e-06,
+      "loss": 0.6218,
+      "step": 1357
+    },
+    {
+      "epoch": 6.79,
+      "grad_norm": 19.284364700317383,
+      "learning_rate": 6.605e-06,
+      "loss": 0.7193,
+      "step": 1358
+    },
+    {
+      "epoch": 6.795,
+      "grad_norm": 6.9484052658081055,
+      "learning_rate": 6.602500000000001e-06,
+      "loss": 0.7836,
+      "step": 1359
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 5.408597946166992,
+      "learning_rate": 6.600000000000001e-06,
+      "loss": 0.378,
+      "step": 1360
+    },
+    {
+      "epoch": 6.805,
+      "grad_norm": 7.371826171875,
+      "learning_rate": 6.5975e-06,
+      "loss": 0.9417,
+      "step": 1361
+    },
+    {
+      "epoch": 6.8100000000000005,
+      "grad_norm": 6.153540134429932,
+      "learning_rate": 6.595e-06,
+      "loss": 0.3824,
+      "step": 1362
+    },
+    {
+      "epoch": 6.8149999999999995,
+      "grad_norm": 3.14127779006958,
+      "learning_rate": 6.592500000000001e-06,
+      "loss": 0.4729,
+      "step": 1363
+    },
+    {
+      "epoch": 6.82,
+      "grad_norm": 6.176417350769043,
+      "learning_rate": 6.5900000000000004e-06,
+      "loss": 0.6242,
+      "step": 1364
+    },
+    {
+      "epoch": 6.825,
+      "grad_norm": 4.964938640594482,
+      "learning_rate": 6.5875e-06,
+      "loss": 0.2795,
+      "step": 1365
+    },
+    {
+      "epoch": 6.83,
+      "grad_norm": 5.460608005523682,
+      "learning_rate": 6.5850000000000005e-06,
+      "loss": 0.4675,
+      "step": 1366
+    },
+    {
+      "epoch": 6.835,
+      "grad_norm": 4.109920978546143,
+      "learning_rate": 6.582500000000001e-06,
+      "loss": 0.3887,
+      "step": 1367
+    },
+    {
+      "epoch": 6.84,
+      "grad_norm": 7.348577499389648,
+      "learning_rate": 6.5800000000000005e-06,
+      "loss": 0.563,
+      "step": 1368
+    },
+    {
+      "epoch": 6.845,
+      "grad_norm": 5.167625904083252,
+      "learning_rate": 6.5775e-06,
+      "loss": 0.5268,
+      "step": 1369
+    },
+    {
+      "epoch": 6.85,
+      "grad_norm": 7.408917427062988,
+      "learning_rate": 6.5750000000000006e-06,
+      "loss": 0.9471,
+      "step": 1370
+    },
+    {
+      "epoch": 6.855,
+      "grad_norm": 5.947335243225098,
+      "learning_rate": 6.5725e-06,
+      "loss": 0.3858,
+      "step": 1371
+    },
+    {
+      "epoch": 6.86,
+      "grad_norm": 5.7228827476501465,
+      "learning_rate": 6.570000000000001e-06,
+      "loss": 0.4304,
+      "step": 1372
+    },
+    {
+      "epoch": 6.865,
+      "grad_norm": 4.398152828216553,
+      "learning_rate": 6.5675e-06,
+      "loss": 0.1956,
+      "step": 1373
+    },
+    {
+      "epoch": 6.87,
+      "grad_norm": 5.479577541351318,
+      "learning_rate": 6.565000000000001e-06,
+      "loss": 0.3018,
+      "step": 1374
+    },
+    {
+      "epoch": 6.875,
+      "grad_norm": 6.460041046142578,
+      "learning_rate": 6.5625e-06,
+      "loss": 0.5693,
+      "step": 1375
+    },
+    {
+      "epoch": 6.88,
+      "grad_norm": 3.6492574214935303,
+      "learning_rate": 6.560000000000001e-06,
+      "loss": 0.1551,
+      "step": 1376
+    },
+    {
+      "epoch": 6.885,
+      "grad_norm": 5.34173059463501,
+      "learning_rate": 6.557500000000001e-06,
+      "loss": 0.5341,
+      "step": 1377
+    },
+    {
+      "epoch": 6.89,
+      "grad_norm": 5.74687385559082,
+      "learning_rate": 6.555e-06,
+      "loss": 0.354,
+      "step": 1378
+    },
+    {
+      "epoch": 6.895,
+      "grad_norm": 3.928020715713501,
+      "learning_rate": 6.5525e-06,
+      "loss": 0.2122,
+      "step": 1379
+    },
+    {
+      "epoch": 6.9,
+      "grad_norm": 5.352936267852783,
+      "learning_rate": 6.550000000000001e-06,
+      "loss": 0.4199,
+      "step": 1380
+    },
+    {
+      "epoch": 6.905,
+      "grad_norm": 12.345170974731445,
+      "learning_rate": 6.547500000000001e-06,
+      "loss": 1.1164,
+      "step": 1381
+    },
+    {
+      "epoch": 6.91,
+      "grad_norm": 7.133321285247803,
+      "learning_rate": 6.545e-06,
+      "loss": 0.8817,
+      "step": 1382
+    },
+    {
+      "epoch": 6.915,
+      "grad_norm": 4.556461334228516,
+      "learning_rate": 6.5425000000000005e-06,
+      "loss": 0.2885,
+      "step": 1383
+    },
+    {
+      "epoch": 6.92,
+      "grad_norm": 4.4646315574646,
+      "learning_rate": 6.540000000000001e-06,
+      "loss": 0.557,
+      "step": 1384
+    },
+    {
+      "epoch": 6.925,
+      "grad_norm": 6.292621612548828,
+      "learning_rate": 6.537500000000001e-06,
+      "loss": 0.511,
+      "step": 1385
+    },
+    {
+      "epoch": 6.93,
+      "grad_norm": 5.740077495574951,
+      "learning_rate": 6.535e-06,
+      "loss": 0.2949,
+      "step": 1386
+    },
+    {
+      "epoch": 6.9350000000000005,
+      "grad_norm": 4.484218597412109,
+      "learning_rate": 6.5325000000000006e-06,
+      "loss": 0.6377,
+      "step": 1387
+    },
+    {
+      "epoch": 6.9399999999999995,
+      "grad_norm": 5.000473499298096,
+      "learning_rate": 6.530000000000001e-06,
+      "loss": 0.2568,
+      "step": 1388
+    },
+    {
+      "epoch": 6.945,
+      "grad_norm": 6.672658920288086,
+      "learning_rate": 6.5275000000000015e-06,
+      "loss": 0.5863,
+      "step": 1389
+    },
+    {
+      "epoch": 6.95,
+      "grad_norm": 5.756199836730957,
+      "learning_rate": 6.525e-06,
+      "loss": 0.4405,
+      "step": 1390
+    },
+    {
+      "epoch": 6.955,
+      "grad_norm": 6.374634742736816,
+      "learning_rate": 6.522500000000001e-06,
+      "loss": 0.8482,
+      "step": 1391
+    },
+    {
+      "epoch": 6.96,
+      "grad_norm": 7.277703762054443,
+      "learning_rate": 6.520000000000001e-06,
+      "loss": 0.5084,
+      "step": 1392
+    },
+    {
+      "epoch": 6.965,
+      "grad_norm": 6.71511697769165,
+      "learning_rate": 6.517500000000001e-06,
+      "loss": 0.5782,
+      "step": 1393
+    },
+    {
+      "epoch": 6.97,
+      "grad_norm": 4.902310848236084,
+      "learning_rate": 6.515e-06,
+      "loss": 0.4962,
+      "step": 1394
+    },
+    {
+      "epoch": 6.975,
+      "grad_norm": 4.004740238189697,
+      "learning_rate": 6.512500000000001e-06,
+      "loss": 0.7565,
+      "step": 1395
+    },
+    {
+      "epoch": 6.98,
+      "grad_norm": 4.597345352172852,
+      "learning_rate": 6.51e-06,
+      "loss": 0.3662,
+      "step": 1396
+    },
+    {
+      "epoch": 6.985,
+      "grad_norm": 3.3693511486053467,
+      "learning_rate": 6.507500000000001e-06,
+      "loss": 0.5133,
+      "step": 1397
+    },
+    {
+      "epoch": 6.99,
+      "grad_norm": 2.8146605491638184,
+      "learning_rate": 6.505e-06,
+      "loss": 0.1184,
+      "step": 1398
+    },
+    {
+      "epoch": 6.995,
+      "grad_norm": 6.476680755615234,
+      "learning_rate": 6.502500000000001e-06,
+      "loss": 0.6637,
+      "step": 1399
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 8.291824340820312,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 0.2706,
+      "step": 1400
+    },
+    {
+      "epoch": 7.0,
+      "eval_loss": 0.32425156235694885,
+      "eval_runtime": 15.9523,
+      "eval_samples_per_second": 50.024,
+      "eval_steps_per_second": 6.269,
+      "step": 1400
+    },
+    {
+      "epoch": 7.005,
+      "grad_norm": 4.331314563751221,
+      "learning_rate": 6.497500000000001e-06,
+      "loss": 0.1823,
+      "step": 1401
+    },
+    {
+      "epoch": 7.01,
+      "grad_norm": 23.027589797973633,
+      "learning_rate": 6.4950000000000005e-06,
+      "loss": 0.766,
+      "step": 1402
+    },
+    {
+      "epoch": 7.015,
+      "grad_norm": 33.23447036743164,
+      "learning_rate": 6.4925e-06,
+      "loss": 0.4623,
+      "step": 1403
+    },
+    {
+      "epoch": 7.02,
+      "grad_norm": 5.248151779174805,
+      "learning_rate": 6.4900000000000005e-06,
+      "loss": 0.7515,
+      "step": 1404
+    },
+    {
+      "epoch": 7.025,
+      "grad_norm": 5.931012153625488,
+      "learning_rate": 6.487500000000001e-06,
+      "loss": 0.5184,
+      "step": 1405
+    },
+    {
+      "epoch": 7.03,
+      "grad_norm": 11.646405220031738,
+      "learning_rate": 6.485000000000001e-06,
+      "loss": 1.7132,
+      "step": 1406
+    },
+    {
+      "epoch": 7.035,
+      "grad_norm": 14.432633399963379,
+      "learning_rate": 6.4825e-06,
+      "loss": 0.8116,
+      "step": 1407
+    },
+    {
+      "epoch": 7.04,
+      "grad_norm": 8.943650245666504,
+      "learning_rate": 6.480000000000001e-06,
+      "loss": 0.6186,
+      "step": 1408
+    },
+    {
+      "epoch": 7.045,
+      "grad_norm": 4.816661834716797,
+      "learning_rate": 6.477500000000001e-06,
+      "loss": 0.3129,
+      "step": 1409
+    },
+    {
+      "epoch": 7.05,
+      "grad_norm": 7.699879169464111,
+      "learning_rate": 6.475e-06,
+      "loss": 0.4736,
+      "step": 1410
+    },
+    {
+      "epoch": 7.055,
+      "grad_norm": 10.32859992980957,
+      "learning_rate": 6.4725e-06,
+      "loss": 1.076,
+      "step": 1411
+    },
+    {
+      "epoch": 7.06,
+      "grad_norm": 5.455946445465088,
+      "learning_rate": 6.470000000000001e-06,
+      "loss": 0.4474,
+      "step": 1412
+    },
+    {
+      "epoch": 7.065,
+      "grad_norm": 60.967403411865234,
+      "learning_rate": 6.467500000000001e-06,
+      "loss": 1.0112,
+      "step": 1413
+    },
+    {
+      "epoch": 7.07,
+      "grad_norm": 5.848165512084961,
+      "learning_rate": 6.465e-06,
+      "loss": 0.6111,
+      "step": 1414
+    },
+    {
+      "epoch": 7.075,
+      "grad_norm": 7.0673933029174805,
+      "learning_rate": 6.4625e-06,
+      "loss": 0.6185,
+      "step": 1415
+    },
+    {
+      "epoch": 7.08,
+      "grad_norm": 5.367558479309082,
+      "learning_rate": 6.460000000000001e-06,
+      "loss": 0.168,
+      "step": 1416
+    },
+    {
+      "epoch": 7.085,
+      "grad_norm": 4.153028964996338,
+      "learning_rate": 6.457500000000001e-06,
+      "loss": 0.3579,
+      "step": 1417
+    },
+    {
+      "epoch": 7.09,
+      "grad_norm": 4.659906387329102,
+      "learning_rate": 6.455e-06,
+      "loss": 0.7093,
+      "step": 1418
+    },
+    {
+      "epoch": 7.095,
+      "grad_norm": 9.350465774536133,
+      "learning_rate": 6.4525000000000005e-06,
+      "loss": 1.1156,
+      "step": 1419
+    },
+    {
+      "epoch": 7.1,
+      "grad_norm": 5.823627948760986,
+      "learning_rate": 6.450000000000001e-06,
+      "loss": 0.66,
+      "step": 1420
+    },
+    {
+      "epoch": 7.105,
+      "grad_norm": 4.859367370605469,
+      "learning_rate": 6.447500000000001e-06,
+      "loss": 0.5556,
+      "step": 1421
+    },
+    {
+      "epoch": 7.11,
+      "grad_norm": 2.4092931747436523,
+      "learning_rate": 6.445e-06,
+      "loss": 0.3232,
+      "step": 1422
+    },
+    {
+      "epoch": 7.115,
+      "grad_norm": 5.1574387550354,
+      "learning_rate": 6.4425000000000006e-06,
+      "loss": 0.4379,
+      "step": 1423
+    },
+    {
+      "epoch": 7.12,
+      "grad_norm": 3.702406406402588,
+      "learning_rate": 6.440000000000001e-06,
+      "loss": 0.1804,
+      "step": 1424
+    },
+    {
+      "epoch": 7.125,
+      "grad_norm": 8.956673622131348,
+      "learning_rate": 6.437500000000001e-06,
+      "loss": 0.5338,
+      "step": 1425
+    },
+    {
+      "epoch": 7.13,
+      "grad_norm": 3.0138790607452393,
+      "learning_rate": 6.435e-06,
+      "loss": 0.2089,
+      "step": 1426
+    },
+    {
+      "epoch": 7.135,
+      "grad_norm": 43.30879211425781,
+      "learning_rate": 6.432500000000001e-06,
+      "loss": 0.7414,
+      "step": 1427
+    },
+    {
+      "epoch": 7.14,
+      "grad_norm": 3.2391490936279297,
+      "learning_rate": 6.43e-06,
+      "loss": 0.3899,
+      "step": 1428
+    },
+    {
+      "epoch": 7.145,
+      "grad_norm": 5.233663558959961,
+      "learning_rate": 6.427500000000001e-06,
+      "loss": 0.5741,
+      "step": 1429
+    },
+    {
+      "epoch": 7.15,
+      "grad_norm": 4.473627090454102,
+      "learning_rate": 6.425e-06,
+      "loss": 0.537,
+      "step": 1430
+    },
+    {
+      "epoch": 7.155,
+      "grad_norm": 4.597846984863281,
+      "learning_rate": 6.422500000000001e-06,
+      "loss": 0.5876,
+      "step": 1431
+    },
+    {
+      "epoch": 7.16,
+      "grad_norm": 16.961240768432617,
+      "learning_rate": 6.42e-06,
+      "loss": 0.641,
+      "step": 1432
+    },
+    {
+      "epoch": 7.165,
+      "grad_norm": 4.091169834136963,
+      "learning_rate": 6.417500000000001e-06,
+      "loss": 0.4536,
+      "step": 1433
+    },
+    {
+      "epoch": 7.17,
+      "grad_norm": 4.2891998291015625,
+      "learning_rate": 6.415e-06,
+      "loss": 0.4632,
+      "step": 1434
+    },
+    {
+      "epoch": 7.175,
+      "grad_norm": 3.7975497245788574,
+      "learning_rate": 6.4125e-06,
+      "loss": 0.3723,
+      "step": 1435
+    },
+    {
+      "epoch": 7.18,
+      "grad_norm": 4.054666042327881,
+      "learning_rate": 6.4100000000000005e-06,
+      "loss": 0.3205,
+      "step": 1436
+    },
+    {
+      "epoch": 7.185,
+      "grad_norm": 4.892772197723389,
+      "learning_rate": 6.407500000000001e-06,
+      "loss": 0.5379,
+      "step": 1437
+    },
+    {
+      "epoch": 7.19,
+      "grad_norm": 8.516433715820312,
+      "learning_rate": 6.4050000000000005e-06,
+      "loss": 0.7628,
+      "step": 1438
+    },
+    {
+      "epoch": 7.195,
+      "grad_norm": 4.729034423828125,
+      "learning_rate": 6.4025e-06,
+      "loss": 0.4755,
+      "step": 1439
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 7.213752269744873,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 0.2721,
+      "step": 1440
+    },
+    {
+      "epoch": 7.205,
+      "grad_norm": 5.3581037521362305,
+      "learning_rate": 6.397500000000001e-06,
+      "loss": 0.3441,
+      "step": 1441
+    },
+    {
+      "epoch": 7.21,
+      "grad_norm": 5.894812107086182,
+      "learning_rate": 6.395e-06,
+      "loss": 0.3513,
+      "step": 1442
+    },
+    {
+      "epoch": 7.215,
+      "grad_norm": 8.052431106567383,
+      "learning_rate": 6.3925e-06,
+      "loss": 0.5664,
+      "step": 1443
+    },
+    {
+      "epoch": 7.22,
+      "grad_norm": 6.2260565757751465,
+      "learning_rate": 6.390000000000001e-06,
+      "loss": 0.2744,
+      "step": 1444
+    },
+    {
+      "epoch": 7.225,
+      "grad_norm": 4.620178699493408,
+      "learning_rate": 6.387500000000001e-06,
+      "loss": 0.3842,
+      "step": 1445
+    },
+    {
+      "epoch": 7.23,
+      "grad_norm": 5.675449371337891,
+      "learning_rate": 6.385e-06,
+      "loss": 0.6521,
+      "step": 1446
+    },
+    {
+      "epoch": 7.235,
+      "grad_norm": 6.781020641326904,
+      "learning_rate": 6.3825e-06,
+      "loss": 0.4654,
+      "step": 1447
+    },
+    {
+      "epoch": 7.24,
+      "grad_norm": 4.464900016784668,
+      "learning_rate": 6.380000000000001e-06,
+      "loss": 0.3535,
+      "step": 1448
+    },
+    {
+      "epoch": 7.245,
+      "grad_norm": 4.859671115875244,
+      "learning_rate": 6.377500000000001e-06,
+      "loss": 0.4949,
+      "step": 1449
+    },
+    {
+      "epoch": 7.25,
+      "grad_norm": 3.718950033187866,
+      "learning_rate": 6.375e-06,
+      "loss": 0.511,
+      "step": 1450
+    },
+    {
+      "epoch": 7.255,
+      "grad_norm": 5.43745756149292,
+      "learning_rate": 6.3725e-06,
+      "loss": 0.4099,
+      "step": 1451
+    },
+    {
+      "epoch": 7.26,
+      "grad_norm": 4.613683700561523,
+      "learning_rate": 6.370000000000001e-06,
+      "loss": 0.1925,
+      "step": 1452
+    },
+    {
+      "epoch": 7.265,
+      "grad_norm": 4.260416507720947,
+      "learning_rate": 6.367500000000001e-06,
+      "loss": 0.2332,
+      "step": 1453
+    },
+    {
+      "epoch": 7.27,
+      "grad_norm": 5.336328983306885,
+      "learning_rate": 6.365e-06,
+      "loss": 0.4719,
+      "step": 1454
+    },
+    {
+      "epoch": 7.275,
+      "grad_norm": 4.176721572875977,
+      "learning_rate": 6.3625000000000005e-06,
+      "loss": 0.3268,
+      "step": 1455
+    },
+    {
+      "epoch": 7.28,
+      "grad_norm": 3.080766439437866,
+      "learning_rate": 6.360000000000001e-06,
+      "loss": 0.33,
+      "step": 1456
+    },
+    {
+      "epoch": 7.285,
+      "grad_norm": 4.627614498138428,
+      "learning_rate": 6.3575000000000005e-06,
+      "loss": 0.5107,
+      "step": 1457
+    },
+    {
+      "epoch": 7.29,
+      "grad_norm": 4.752645969390869,
+      "learning_rate": 6.355e-06,
+      "loss": 0.7327,
+      "step": 1458
+    },
+    {
+      "epoch": 7.295,
+      "grad_norm": 5.8618974685668945,
+      "learning_rate": 6.352500000000001e-06,
+      "loss": 0.4193,
+      "step": 1459
+    },
+    {
+      "epoch": 7.3,
+      "grad_norm": 3.5398154258728027,
+      "learning_rate": 6.35e-06,
+      "loss": 0.4467,
+      "step": 1460
+    },
+    {
+      "epoch": 7.305,
+      "grad_norm": 4.829983711242676,
+      "learning_rate": 6.347500000000001e-06,
+      "loss": 0.5332,
+      "step": 1461
+    },
+    {
+      "epoch": 7.31,
+      "grad_norm": 5.366067409515381,
+      "learning_rate": 6.345e-06,
+      "loss": 0.308,
+      "step": 1462
+    },
+    {
+      "epoch": 7.315,
+      "grad_norm": 7.315142631530762,
+      "learning_rate": 6.342500000000001e-06,
+      "loss": 0.7116,
+      "step": 1463
+    },
+    {
+      "epoch": 7.32,
+      "grad_norm": 5.300399303436279,
+      "learning_rate": 6.34e-06,
+      "loss": 0.3571,
+      "step": 1464
+    },
+    {
+      "epoch": 7.325,
+      "grad_norm": 4.849173069000244,
+      "learning_rate": 6.337500000000001e-06,
+      "loss": 0.4143,
+      "step": 1465
+    },
+    {
+      "epoch": 7.33,
+      "grad_norm": 7.185313701629639,
+      "learning_rate": 6.335e-06,
+      "loss": 0.6487,
+      "step": 1466
+    },
+    {
+      "epoch": 7.335,
+      "grad_norm": 4.1438069343566895,
+      "learning_rate": 6.3325e-06,
+      "loss": 0.4472,
+      "step": 1467
+    },
+    {
+      "epoch": 7.34,
+      "grad_norm": 5.969173908233643,
+      "learning_rate": 6.33e-06,
+      "loss": 0.2744,
+      "step": 1468
+    },
+    {
+      "epoch": 7.345,
+      "grad_norm": 3.7982192039489746,
+      "learning_rate": 6.327500000000001e-06,
+      "loss": 0.5689,
+      "step": 1469
+    },
+    {
+      "epoch": 7.35,
+      "grad_norm": 6.671392917633057,
+      "learning_rate": 6.3250000000000004e-06,
+      "loss": 0.6169,
+      "step": 1470
+    },
+    {
+      "epoch": 7.355,
+      "grad_norm": 3.6170287132263184,
+      "learning_rate": 6.3225e-06,
+      "loss": 0.5215,
+      "step": 1471
+    },
+    {
+      "epoch": 7.36,
+      "grad_norm": 6.451892852783203,
+      "learning_rate": 6.3200000000000005e-06,
+      "loss": 0.6023,
+      "step": 1472
+    },
+    {
+      "epoch": 7.365,
+      "grad_norm": 4.601886749267578,
+      "learning_rate": 6.317500000000001e-06,
+      "loss": 0.2924,
+      "step": 1473
+    },
+    {
+      "epoch": 7.37,
+      "grad_norm": 7.112369060516357,
+      "learning_rate": 6.315e-06,
+      "loss": 1.1339,
+      "step": 1474
+    },
+    {
+      "epoch": 7.375,
+      "grad_norm": 3.1813583374023438,
+      "learning_rate": 6.3125e-06,
+      "loss": 0.3239,
+      "step": 1475
+    },
+    {
+      "epoch": 7.38,
+      "grad_norm": 3.9900238513946533,
+      "learning_rate": 6.3100000000000006e-06,
+      "loss": 0.5045,
+      "step": 1476
+    },
+    {
+      "epoch": 7.385,
+      "grad_norm": 5.861837387084961,
+      "learning_rate": 6.307500000000001e-06,
+      "loss": 0.4855,
+      "step": 1477
+    },
+    {
+      "epoch": 7.39,
+      "grad_norm": 7.484753131866455,
+      "learning_rate": 6.305e-06,
+      "loss": 0.5439,
+      "step": 1478
+    },
+    {
+      "epoch": 7.395,
+      "grad_norm": 5.109606742858887,
+      "learning_rate": 6.3025e-06,
+      "loss": 0.4386,
+      "step": 1479
+    },
+    {
+      "epoch": 7.4,
+      "grad_norm": 5.093142986297607,
+      "learning_rate": 6.300000000000001e-06,
+      "loss": 0.3529,
+      "step": 1480
+    },
+    {
+      "epoch": 7.405,
+      "grad_norm": 3.9258766174316406,
+      "learning_rate": 6.297500000000001e-06,
+      "loss": 0.1939,
+      "step": 1481
+    },
+    {
+      "epoch": 7.41,
+      "grad_norm": 6.193164825439453,
+      "learning_rate": 6.295e-06,
+      "loss": 0.5582,
+      "step": 1482
+    },
+    {
+      "epoch": 7.415,
+      "grad_norm": 6.271528720855713,
+      "learning_rate": 6.2925e-06,
+      "loss": 0.378,
+      "step": 1483
+    },
+    {
+      "epoch": 7.42,
+      "grad_norm": 4.6442646980285645,
+      "learning_rate": 6.290000000000001e-06,
+      "loss": 0.5675,
+      "step": 1484
+    },
+    {
+      "epoch": 7.425,
+      "grad_norm": 4.060819149017334,
+      "learning_rate": 6.287500000000001e-06,
+      "loss": 0.464,
+      "step": 1485
+    },
+    {
+      "epoch": 7.43,
+      "grad_norm": 2.7939300537109375,
+      "learning_rate": 6.285e-06,
+      "loss": 0.261,
+      "step": 1486
+    },
+    {
+      "epoch": 7.435,
+      "grad_norm": 8.506726264953613,
+      "learning_rate": 6.2825e-06,
+      "loss": 0.5868,
+      "step": 1487
+    },
+    {
+      "epoch": 7.44,
+      "grad_norm": 5.71343994140625,
+      "learning_rate": 6.280000000000001e-06,
+      "loss": 0.3533,
+      "step": 1488
+    },
+    {
+      "epoch": 7.445,
+      "grad_norm": 6.951137542724609,
+      "learning_rate": 6.2775000000000005e-06,
+      "loss": 0.967,
+      "step": 1489
+    },
+    {
+      "epoch": 7.45,
+      "grad_norm": 9.001285552978516,
+      "learning_rate": 6.275e-06,
+      "loss": 0.4057,
+      "step": 1490
+    },
+    {
+      "epoch": 7.455,
+      "grad_norm": 3.9196596145629883,
+      "learning_rate": 6.2725000000000005e-06,
+      "loss": 0.1803,
+      "step": 1491
+    },
+    {
+      "epoch": 7.46,
+      "grad_norm": 3.361323833465576,
+      "learning_rate": 6.27e-06,
+      "loss": 0.4776,
+      "step": 1492
+    },
+    {
+      "epoch": 7.465,
+      "grad_norm": 8.999086380004883,
+      "learning_rate": 6.2675000000000006e-06,
+      "loss": 0.3201,
+      "step": 1493
+    },
+    {
+      "epoch": 7.47,
+      "grad_norm": 4.233700752258301,
+      "learning_rate": 6.265e-06,
+      "loss": 0.2376,
+      "step": 1494
+    },
+    {
+      "epoch": 7.475,
+      "grad_norm": 6.62499475479126,
+      "learning_rate": 6.262500000000001e-06,
+      "loss": 0.6047,
+      "step": 1495
+    },
+    {
+      "epoch": 7.48,
+      "grad_norm": 4.045609474182129,
+      "learning_rate": 6.26e-06,
+      "loss": 0.2327,
+      "step": 1496
+    },
+    {
+      "epoch": 7.485,
+      "grad_norm": 4.1038818359375,
+      "learning_rate": 6.257500000000001e-06,
+      "loss": 0.4287,
+      "step": 1497
+    },
+    {
+      "epoch": 7.49,
+      "grad_norm": 3.803050994873047,
+      "learning_rate": 6.255e-06,
+      "loss": 0.2529,
+      "step": 1498
+    },
+    {
+      "epoch": 7.495,
+      "grad_norm": 3.734079122543335,
+      "learning_rate": 6.2525e-06,
+      "loss": 0.3067,
+      "step": 1499
+    },
+    {
+      "epoch": 7.5,
+      "grad_norm": 4.518526554107666,
+      "learning_rate": 6.25e-06,
+      "loss": 0.459,
+      "step": 1500
+    },
+    {
+      "epoch": 7.505,
+      "grad_norm": 9.755220413208008,
+      "learning_rate": 6.247500000000001e-06,
+      "loss": 0.8462,
+      "step": 1501
+    },
+    {
+      "epoch": 7.51,
+      "grad_norm": 6.331623554229736,
+      "learning_rate": 6.245000000000001e-06,
+      "loss": 0.7631,
+      "step": 1502
+    },
+    {
+      "epoch": 7.515,
+      "grad_norm": 2.957162618637085,
+      "learning_rate": 6.2425e-06,
+      "loss": 0.1618,
+      "step": 1503
+    },
+    {
+      "epoch": 7.52,
+      "grad_norm": 6.043359279632568,
+      "learning_rate": 6.24e-06,
+      "loss": 0.2772,
+      "step": 1504
+    },
+    {
+      "epoch": 7.525,
+      "grad_norm": 5.4535675048828125,
+      "learning_rate": 6.237500000000001e-06,
+      "loss": 0.2317,
+      "step": 1505
+    },
+    {
+      "epoch": 7.53,
+      "grad_norm": 8.488757133483887,
+      "learning_rate": 6.235000000000001e-06,
+      "loss": 0.5556,
+      "step": 1506
+    },
+    {
+      "epoch": 7.535,
+      "grad_norm": 8.43838882446289,
+      "learning_rate": 6.2325e-06,
+      "loss": 0.3143,
+      "step": 1507
+    },
+    {
+      "epoch": 7.54,
+      "grad_norm": 3.8499398231506348,
+      "learning_rate": 6.2300000000000005e-06,
+      "loss": 0.3244,
+      "step": 1508
+    },
+    {
+      "epoch": 7.545,
+      "grad_norm": 3.861306667327881,
+      "learning_rate": 6.227500000000001e-06,
+      "loss": 0.4217,
+      "step": 1509
+    },
+    {
+      "epoch": 7.55,
+      "grad_norm": 5.0174384117126465,
+      "learning_rate": 6.225000000000001e-06,
+      "loss": 0.3718,
+      "step": 1510
+    },
+    {
+      "epoch": 7.555,
+      "grad_norm": 4.34398078918457,
+      "learning_rate": 6.2225e-06,
+      "loss": 0.2345,
+      "step": 1511
+    },
+    {
+      "epoch": 7.5600000000000005,
+      "grad_norm": 5.6488037109375,
+      "learning_rate": 6.220000000000001e-06,
+      "loss": 0.6219,
+      "step": 1512
+    },
+    {
+      "epoch": 7.5649999999999995,
+      "grad_norm": 8.369893074035645,
+      "learning_rate": 6.217500000000001e-06,
+      "loss": 0.7124,
+      "step": 1513
+    },
+    {
+      "epoch": 7.57,
+      "grad_norm": 4.1293840408325195,
+      "learning_rate": 6.215000000000001e-06,
+      "loss": 0.2691,
+      "step": 1514
+    },
+    {
+      "epoch": 7.575,
+      "grad_norm": 4.059218406677246,
+      "learning_rate": 6.2125e-06,
+      "loss": 0.3949,
+      "step": 1515
+    },
+    {
+      "epoch": 7.58,
+      "grad_norm": 5.032373428344727,
+      "learning_rate": 6.210000000000001e-06,
+      "loss": 0.7404,
+      "step": 1516
+    },
+    {
+      "epoch": 7.585,
+      "grad_norm": 11.159683227539062,
+      "learning_rate": 6.207500000000001e-06,
+      "loss": 1.2709,
+      "step": 1517
+    },
+    {
+      "epoch": 7.59,
+      "grad_norm": 4.629010200500488,
+      "learning_rate": 6.205000000000001e-06,
+      "loss": 0.3597,
+      "step": 1518
+    },
+    {
+      "epoch": 7.595,
+      "grad_norm": 5.1266021728515625,
+      "learning_rate": 6.2025e-06,
+      "loss": 0.587,
+      "step": 1519
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 19.878353118896484,
+      "learning_rate": 6.200000000000001e-06,
+      "loss": 0.6607,
+      "step": 1520
+    },
+    {
+      "epoch": 7.605,
+      "grad_norm": 6.475594520568848,
+      "learning_rate": 6.1975e-06,
+      "loss": 0.4445,
+      "step": 1521
+    },
+    {
+      "epoch": 7.61,
+      "grad_norm": 12.96800708770752,
+      "learning_rate": 6.195000000000001e-06,
+      "loss": 0.8162,
+      "step": 1522
+    },
+    {
+      "epoch": 7.615,
+      "grad_norm": 3.9075801372528076,
+      "learning_rate": 6.1925000000000004e-06,
+      "loss": 0.4651,
+      "step": 1523
+    },
+    {
+      "epoch": 7.62,
+      "grad_norm": 9.485398292541504,
+      "learning_rate": 6.190000000000001e-06,
+      "loss": 0.9293,
+      "step": 1524
+    },
+    {
+      "epoch": 7.625,
+      "grad_norm": 4.966853618621826,
+      "learning_rate": 6.1875000000000005e-06,
+      "loss": 0.4828,
+      "step": 1525
+    },
+    {
+      "epoch": 7.63,
+      "grad_norm": 4.854305267333984,
+      "learning_rate": 6.185000000000001e-06,
+      "loss": 0.4448,
+      "step": 1526
+    },
+    {
+      "epoch": 7.635,
+      "grad_norm": 31.39982032775879,
+      "learning_rate": 6.1825000000000005e-06,
+      "loss": 0.4254,
+      "step": 1527
+    },
+    {
+      "epoch": 7.64,
+      "grad_norm": 16.46684455871582,
+      "learning_rate": 6.18e-06,
+      "loss": 0.8842,
+      "step": 1528
+    },
+    {
+      "epoch": 7.645,
+      "grad_norm": 7.199646949768066,
+      "learning_rate": 6.1775000000000006e-06,
+      "loss": 0.3058,
+      "step": 1529
+    },
+    {
+      "epoch": 7.65,
+      "grad_norm": 4.856558799743652,
+      "learning_rate": 6.175000000000001e-06,
+      "loss": 0.3075,
+      "step": 1530
+    },
+    {
+      "epoch": 7.655,
+      "grad_norm": 5.092007160186768,
+      "learning_rate": 6.1725e-06,
+      "loss": 0.3751,
+      "step": 1531
+    },
+    {
+      "epoch": 7.66,
+      "grad_norm": 5.739753723144531,
+      "learning_rate": 6.17e-06,
+      "loss": 0.5891,
+      "step": 1532
+    },
+    {
+      "epoch": 7.665,
+      "grad_norm": 6.071251392364502,
+      "learning_rate": 6.167500000000001e-06,
+      "loss": 0.6937,
+      "step": 1533
+    },
+    {
+      "epoch": 7.67,
+      "grad_norm": 4.13346004486084,
+      "learning_rate": 6.165000000000001e-06,
+      "loss": 0.6876,
+      "step": 1534
+    },
+    {
+      "epoch": 7.675,
+      "grad_norm": 7.905163764953613,
+      "learning_rate": 6.1625e-06,
+      "loss": 0.7773,
+      "step": 1535
+    },
+    {
+      "epoch": 7.68,
+      "grad_norm": 5.325740337371826,
+      "learning_rate": 6.16e-06,
+      "loss": 0.3487,
+      "step": 1536
+    },
+    {
+      "epoch": 7.6850000000000005,
+      "grad_norm": 4.619688987731934,
+      "learning_rate": 6.157500000000001e-06,
+      "loss": 0.2789,
+      "step": 1537
+    },
+    {
+      "epoch": 7.6899999999999995,
+      "grad_norm": 3.681313991546631,
+      "learning_rate": 6.155000000000001e-06,
+      "loss": 0.3248,
+      "step": 1538
+    },
+    {
+      "epoch": 7.695,
+      "grad_norm": 3.969858169555664,
+      "learning_rate": 6.1525e-06,
+      "loss": 0.2068,
+      "step": 1539
+    },
+    {
+      "epoch": 7.7,
+      "grad_norm": 5.009308338165283,
+      "learning_rate": 6.15e-06,
+      "loss": 0.5491,
+      "step": 1540
+    },
+    {
+      "epoch": 7.705,
+      "grad_norm": 6.20375394821167,
+      "learning_rate": 6.147500000000001e-06,
+      "loss": 0.7372,
+      "step": 1541
+    },
+    {
+      "epoch": 7.71,
+      "grad_norm": 4.483755111694336,
+      "learning_rate": 6.145000000000001e-06,
+      "loss": 0.3237,
+      "step": 1542
+    },
+    {
+      "epoch": 7.715,
+      "grad_norm": 6.629380226135254,
+      "learning_rate": 6.1425e-06,
+      "loss": 0.6018,
+      "step": 1543
+    },
+    {
+      "epoch": 7.72,
+      "grad_norm": 4.270708084106445,
+      "learning_rate": 6.1400000000000005e-06,
+      "loss": 0.1423,
+      "step": 1544
+    },
+    {
+      "epoch": 7.725,
+      "grad_norm": 4.556453227996826,
+      "learning_rate": 6.137500000000001e-06,
+      "loss": 0.4509,
+      "step": 1545
+    },
+    {
+      "epoch": 7.73,
+      "grad_norm": 5.641851425170898,
+      "learning_rate": 6.1350000000000006e-06,
+      "loss": 0.6505,
+      "step": 1546
+    },
+    {
+      "epoch": 7.735,
+      "grad_norm": 6.288632392883301,
+      "learning_rate": 6.1325e-06,
+      "loss": 0.6075,
+      "step": 1547
+    },
+    {
+      "epoch": 7.74,
+      "grad_norm": 4.989062309265137,
+      "learning_rate": 6.130000000000001e-06,
+      "loss": 0.6031,
+      "step": 1548
+    },
+    {
+      "epoch": 7.745,
+      "grad_norm": 3.253988027572632,
+      "learning_rate": 6.127500000000001e-06,
+      "loss": 0.2168,
+      "step": 1549
+    },
+    {
+      "epoch": 7.75,
+      "grad_norm": 3.6411399841308594,
+      "learning_rate": 6.125000000000001e-06,
+      "loss": 0.4415,
+      "step": 1550
+    },
+    {
+      "epoch": 7.755,
+      "grad_norm": 13.75715446472168,
+      "learning_rate": 6.1225e-06,
+      "loss": 0.8918,
+      "step": 1551
+    },
+    {
+      "epoch": 7.76,
+      "grad_norm": 4.578426837921143,
+      "learning_rate": 6.120000000000001e-06,
+      "loss": 0.5603,
+      "step": 1552
+    },
+    {
+      "epoch": 7.765,
+      "grad_norm": 2.759368658065796,
+      "learning_rate": 6.1175e-06,
+      "loss": 0.2497,
+      "step": 1553
+    },
+    {
+      "epoch": 7.77,
+      "grad_norm": 6.976686954498291,
+      "learning_rate": 6.115000000000001e-06,
+      "loss": 0.8124,
+      "step": 1554
+    },
+    {
+      "epoch": 7.775,
+      "grad_norm": 3.9466745853424072,
+      "learning_rate": 6.1125e-06,
+      "loss": 0.5827,
+      "step": 1555
+    },
+    {
+      "epoch": 7.78,
+      "grad_norm": 3.46417498588562,
+      "learning_rate": 6.110000000000001e-06,
+      "loss": 0.3314,
+      "step": 1556
+    },
+    {
+      "epoch": 7.785,
+      "grad_norm": 6.931117534637451,
+      "learning_rate": 6.1075e-06,
+      "loss": 0.7602,
+      "step": 1557
+    },
+    {
+      "epoch": 7.79,
+      "grad_norm": 3.8627309799194336,
+      "learning_rate": 6.105000000000001e-06,
+      "loss": 0.2075,
+      "step": 1558
+    },
+    {
+      "epoch": 7.795,
+      "grad_norm": 5.289340972900391,
+      "learning_rate": 6.1025000000000004e-06,
+      "loss": 0.369,
+      "step": 1559
+    },
+    {
+      "epoch": 7.8,
+      "grad_norm": 13.306784629821777,
+      "learning_rate": 6.1e-06,
+      "loss": 1.0129,
+      "step": 1560
+    },
+    {
+      "epoch": 7.805,
+      "grad_norm": 8.380496978759766,
+      "learning_rate": 6.0975000000000005e-06,
+      "loss": 0.6984,
+      "step": 1561
+    },
+    {
+      "epoch": 7.8100000000000005,
+      "grad_norm": 5.159175395965576,
+      "learning_rate": 6.095000000000001e-06,
+      "loss": 0.3908,
+      "step": 1562
+    },
+    {
+      "epoch": 7.8149999999999995,
+      "grad_norm": 4.713963031768799,
+      "learning_rate": 6.0925e-06,
+      "loss": 0.3591,
+      "step": 1563
+    },
+    {
+      "epoch": 7.82,
+      "grad_norm": 6.00685977935791,
+      "learning_rate": 6.09e-06,
+      "loss": 0.5242,
+      "step": 1564
+    },
+    {
+      "epoch": 7.825,
+      "grad_norm": 2.9491894245147705,
+      "learning_rate": 6.087500000000001e-06,
+      "loss": 0.1737,
+      "step": 1565
+    },
+    {
+      "epoch": 7.83,
+      "grad_norm": 3.8897531032562256,
+      "learning_rate": 6.085000000000001e-06,
+      "loss": 0.309,
+      "step": 1566
+    },
+    {
+      "epoch": 7.835,
+      "grad_norm": 6.020238876342773,
+      "learning_rate": 6.0825e-06,
+      "loss": 0.5912,
+      "step": 1567
+    },
+    {
+      "epoch": 7.84,
+      "grad_norm": 3.9014713764190674,
+      "learning_rate": 6.08e-06,
+      "loss": 0.1425,
+      "step": 1568
+    },
+    {
+      "epoch": 7.845,
+      "grad_norm": 5.010612964630127,
+      "learning_rate": 6.077500000000001e-06,
+      "loss": 0.2064,
+      "step": 1569
+    },
+    {
+      "epoch": 7.85,
+      "grad_norm": 7.698296546936035,
+      "learning_rate": 6.075000000000001e-06,
+      "loss": 0.6469,
+      "step": 1570
+    },
+    {
+      "epoch": 7.855,
+      "grad_norm": 6.6042585372924805,
+      "learning_rate": 6.0725e-06,
+      "loss": 0.3043,
+      "step": 1571
+    },
+    {
+      "epoch": 7.86,
+      "grad_norm": 3.231539487838745,
+      "learning_rate": 6.07e-06,
+      "loss": 0.3017,
+      "step": 1572
+    },
+    {
+      "epoch": 7.865,
+      "grad_norm": 6.583184719085693,
+      "learning_rate": 6.067500000000001e-06,
+      "loss": 0.1306,
+      "step": 1573
+    },
+    {
+      "epoch": 7.87,
+      "grad_norm": 4.476734161376953,
+      "learning_rate": 6.065000000000001e-06,
+      "loss": 0.1549,
+      "step": 1574
+    },
+    {
+      "epoch": 7.875,
+      "grad_norm": 2.6369550228118896,
+      "learning_rate": 6.0625e-06,
+      "loss": 0.2601,
+      "step": 1575
+    },
+    {
+      "epoch": 7.88,
+      "grad_norm": 16.520837783813477,
+      "learning_rate": 6.0600000000000004e-06,
+      "loss": 0.269,
+      "step": 1576
+    },
+    {
+      "epoch": 7.885,
+      "grad_norm": 5.422561168670654,
+      "learning_rate": 6.057500000000001e-06,
+      "loss": 0.3467,
+      "step": 1577
+    },
+    {
+      "epoch": 7.89,
+      "grad_norm": 4.908959865570068,
+      "learning_rate": 6.0550000000000005e-06,
+      "loss": 0.2785,
+      "step": 1578
+    },
+    {
+      "epoch": 7.895,
+      "grad_norm": 3.5143966674804688,
+      "learning_rate": 6.0525e-06,
+      "loss": 0.1839,
+      "step": 1579
+    },
+    {
+      "epoch": 7.9,
+      "grad_norm": 3.1573448181152344,
+      "learning_rate": 6.0500000000000005e-06,
+      "loss": 0.1812,
+      "step": 1580
+    },
+    {
+      "epoch": 7.905,
+      "grad_norm": 5.420172691345215,
+      "learning_rate": 6.047500000000001e-06,
+      "loss": 1.04,
+      "step": 1581
+    },
+    {
+      "epoch": 7.91,
+      "grad_norm": 9.830706596374512,
+      "learning_rate": 6.0450000000000006e-06,
+      "loss": 0.6402,
+      "step": 1582
+    },
+    {
+      "epoch": 7.915,
+      "grad_norm": 3.7396488189697266,
+      "learning_rate": 6.0425e-06,
+      "loss": 0.3739,
+      "step": 1583
+    },
+    {
+      "epoch": 7.92,
+      "grad_norm": 3.252859115600586,
+      "learning_rate": 6.040000000000001e-06,
+      "loss": 0.2522,
+      "step": 1584
+    },
+    {
+      "epoch": 7.925,
+      "grad_norm": 4.1390814781188965,
+      "learning_rate": 6.0375e-06,
+      "loss": 0.4485,
+      "step": 1585
+    },
+    {
+      "epoch": 7.93,
+      "grad_norm": 3.7815802097320557,
+      "learning_rate": 6.035000000000001e-06,
+      "loss": 0.4189,
+      "step": 1586
+    },
+    {
+      "epoch": 7.9350000000000005,
+      "grad_norm": 3.5664005279541016,
+      "learning_rate": 6.0325e-06,
+      "loss": 0.6187,
+      "step": 1587
+    },
+    {
+      "epoch": 7.9399999999999995,
+      "grad_norm": 4.558847904205322,
+      "learning_rate": 6.030000000000001e-06,
+      "loss": 0.3862,
+      "step": 1588
+    },
+    {
+      "epoch": 7.945,
+      "grad_norm": 4.6947197914123535,
+      "learning_rate": 6.0275e-06,
+      "loss": 0.285,
+      "step": 1589
+    },
+    {
+      "epoch": 7.95,
+      "grad_norm": 5.713285446166992,
+      "learning_rate": 6.025000000000001e-06,
+      "loss": 0.4296,
+      "step": 1590
+    },
+    {
+      "epoch": 7.955,
+      "grad_norm": 6.391741752624512,
+      "learning_rate": 6.0225e-06,
+      "loss": 0.7541,
+      "step": 1591
+    },
+    {
+      "epoch": 7.96,
+      "grad_norm": 4.6010518074035645,
+      "learning_rate": 6.02e-06,
+      "loss": 0.4906,
+      "step": 1592
+    },
+    {
+      "epoch": 7.965,
+      "grad_norm": 5.531135082244873,
+      "learning_rate": 6.0175e-06,
+      "loss": 0.4183,
+      "step": 1593
+    },
+    {
+      "epoch": 7.97,
+      "grad_norm": 5.344430923461914,
+      "learning_rate": 6.015000000000001e-06,
+      "loss": 0.6549,
+      "step": 1594
+    },
+    {
+      "epoch": 7.975,
+      "grad_norm": 4.701644420623779,
+      "learning_rate": 6.0125e-06,
+      "loss": 0.4256,
+      "step": 1595
+    },
+    {
+      "epoch": 7.98,
+      "grad_norm": 3.9140567779541016,
+      "learning_rate": 6.01e-06,
+      "loss": 0.2285,
+      "step": 1596
+    },
+    {
+      "epoch": 7.985,
+      "grad_norm": 6.353307723999023,
+      "learning_rate": 6.0075000000000005e-06,
+      "loss": 0.3969,
+      "step": 1597
+    },
+    {
+      "epoch": 7.99,
+      "grad_norm": 4.474092483520508,
+      "learning_rate": 6.005000000000001e-06,
+      "loss": 0.3554,
+      "step": 1598
+    },
+    {
+      "epoch": 7.995,
+      "grad_norm": 5.053589344024658,
+      "learning_rate": 6.0025e-06,
+      "loss": 0.7431,
+      "step": 1599
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 5.900409698486328,
+      "learning_rate": 6e-06,
+      "loss": 0.4221,
+      "step": 1600
+    },
+    {
+      "epoch": 8.0,
+      "eval_loss": 0.29396384954452515,
+      "eval_runtime": 15.1921,
+      "eval_samples_per_second": 52.527,
+      "eval_steps_per_second": 6.582,
+      "step": 1600
+    },
+    {
+      "epoch": 8.005,
+      "grad_norm": 3.5353925228118896,
+      "learning_rate": 5.997500000000001e-06,
+      "loss": 0.4077,
+      "step": 1601
+    },
+    {
+      "epoch": 8.01,
+      "grad_norm": 2.9544553756713867,
+      "learning_rate": 5.995000000000001e-06,
+      "loss": 0.2128,
+      "step": 1602
+    },
+    {
+      "epoch": 8.015,
+      "grad_norm": 5.352968692779541,
+      "learning_rate": 5.9925e-06,
+      "loss": 0.305,
+      "step": 1603
+    },
+    {
+      "epoch": 8.02,
+      "grad_norm": 8.153446197509766,
+      "learning_rate": 5.99e-06,
+      "loss": 0.9553,
+      "step": 1604
+    },
+    {
+      "epoch": 8.025,
+      "grad_norm": 4.807647705078125,
+      "learning_rate": 5.987500000000001e-06,
+      "loss": 0.5507,
+      "step": 1605
+    },
+    {
+      "epoch": 8.03,
+      "grad_norm": 3.371746301651001,
+      "learning_rate": 5.985000000000001e-06,
+      "loss": 0.1173,
+      "step": 1606
+    },
+    {
+      "epoch": 8.035,
+      "grad_norm": 7.9792094230651855,
+      "learning_rate": 5.9825e-06,
+      "loss": 1.0455,
+      "step": 1607
+    },
+    {
+      "epoch": 8.04,
+      "grad_norm": 4.711618900299072,
+      "learning_rate": 5.98e-06,
+      "loss": 0.4044,
+      "step": 1608
+    },
+    {
+      "epoch": 8.045,
+      "grad_norm": 3.0883560180664062,
+      "learning_rate": 5.977500000000001e-06,
+      "loss": 0.2385,
+      "step": 1609
+    },
+    {
+      "epoch": 8.05,
+      "grad_norm": 3.4350533485412598,
+      "learning_rate": 5.975e-06,
+      "loss": 0.3678,
+      "step": 1610
+    },
+    {
+      "epoch": 8.055,
+      "grad_norm": 4.454967975616455,
+      "learning_rate": 5.9725e-06,
+      "loss": 0.3931,
+      "step": 1611
+    },
+    {
+      "epoch": 8.06,
+      "grad_norm": 5.610987186431885,
+      "learning_rate": 5.9700000000000004e-06,
+      "loss": 0.7828,
+      "step": 1612
+    },
+    {
+      "epoch": 8.065,
+      "grad_norm": 8.979826927185059,
+      "learning_rate": 5.967500000000001e-06,
+      "loss": 0.7575,
+      "step": 1613
+    },
+    {
+      "epoch": 8.07,
+      "grad_norm": 7.777012348175049,
+      "learning_rate": 5.9650000000000005e-06,
+      "loss": 0.6445,
+      "step": 1614
+    },
+    {
+      "epoch": 8.075,
+      "grad_norm": 6.797120094299316,
+      "learning_rate": 5.9625e-06,
+      "loss": 0.94,
+      "step": 1615
+    },
+    {
+      "epoch": 8.08,
+      "grad_norm": 7.317530632019043,
+      "learning_rate": 5.9600000000000005e-06,
+      "loss": 0.5106,
+      "step": 1616
+    },
+    {
+      "epoch": 8.085,
+      "grad_norm": 4.199662685394287,
+      "learning_rate": 5.9575e-06,
+      "loss": 0.3883,
+      "step": 1617
+    },
+    {
+      "epoch": 8.09,
+      "grad_norm": 17.127452850341797,
+      "learning_rate": 5.955000000000001e-06,
+      "loss": 0.4499,
+      "step": 1618
+    },
+    {
+      "epoch": 8.095,
+      "grad_norm": 5.7033209800720215,
+      "learning_rate": 5.9525e-06,
+      "loss": 0.4684,
+      "step": 1619
+    },
+    {
+      "epoch": 8.1,
+      "grad_norm": 6.2592926025390625,
+      "learning_rate": 5.950000000000001e-06,
+      "loss": 0.4454,
+      "step": 1620
+    },
+    {
+      "epoch": 8.105,
+      "grad_norm": 4.529707431793213,
+      "learning_rate": 5.9475e-06,
+      "loss": 0.6415,
+      "step": 1621
+    },
+    {
+      "epoch": 8.11,
+      "grad_norm": 5.618786811828613,
+      "learning_rate": 5.945000000000001e-06,
+      "loss": 0.5337,
+      "step": 1622
+    },
+    {
+      "epoch": 8.115,
+      "grad_norm": 16.066375732421875,
+      "learning_rate": 5.9425e-06,
+      "loss": 0.7873,
+      "step": 1623
+    },
+    {
+      "epoch": 8.12,
+      "grad_norm": 4.848687648773193,
+      "learning_rate": 5.94e-06,
+      "loss": 0.5523,
+      "step": 1624
+    },
+    {
+      "epoch": 8.125,
+      "grad_norm": 3.528921604156494,
+      "learning_rate": 5.9375e-06,
+      "loss": 0.1758,
+      "step": 1625
+    },
+    {
+      "epoch": 8.13,
+      "grad_norm": 4.488472938537598,
+      "learning_rate": 5.935000000000001e-06,
+      "loss": 0.373,
+      "step": 1626
+    },
+    {
+      "epoch": 8.135,
+      "grad_norm": 8.245428085327148,
+      "learning_rate": 5.932500000000001e-06,
+      "loss": 0.5638,
+      "step": 1627
+    },
+    {
+      "epoch": 8.14,
+      "grad_norm": 4.466881275177002,
+      "learning_rate": 5.93e-06,
+      "loss": 0.3925,
+      "step": 1628
+    },
+    {
+      "epoch": 8.145,
+      "grad_norm": 4.161615371704102,
+      "learning_rate": 5.9275e-06,
+      "loss": 0.357,
+      "step": 1629
+    },
+    {
+      "epoch": 8.15,
+      "grad_norm": 2.9932260513305664,
+      "learning_rate": 5.925000000000001e-06,
+      "loss": 0.1836,
+      "step": 1630
+    },
+    {
+      "epoch": 8.155,
+      "grad_norm": 4.168212890625,
+      "learning_rate": 5.922500000000001e-06,
+      "loss": 0.2598,
+      "step": 1631
+    },
+    {
+      "epoch": 8.16,
+      "grad_norm": 3.5720019340515137,
+      "learning_rate": 5.92e-06,
+      "loss": 0.496,
+      "step": 1632
+    },
+    {
+      "epoch": 8.165,
+      "grad_norm": 3.9904985427856445,
+      "learning_rate": 5.9175000000000005e-06,
+      "loss": 0.2652,
+      "step": 1633
+    },
+    {
+      "epoch": 8.17,
+      "grad_norm": 3.8955767154693604,
+      "learning_rate": 5.915000000000001e-06,
+      "loss": 0.5653,
+      "step": 1634
+    },
+    {
+      "epoch": 8.175,
+      "grad_norm": 4.211009979248047,
+      "learning_rate": 5.912500000000001e-06,
+      "loss": 0.5422,
+      "step": 1635
+    },
+    {
+      "epoch": 8.18,
+      "grad_norm": 4.91811990737915,
+      "learning_rate": 5.91e-06,
+      "loss": 0.4617,
+      "step": 1636
+    },
+    {
+      "epoch": 8.185,
+      "grad_norm": 9.31735610961914,
+      "learning_rate": 5.907500000000001e-06,
+      "loss": 0.5781,
+      "step": 1637
+    },
+    {
+      "epoch": 8.19,
+      "grad_norm": 5.391366004943848,
+      "learning_rate": 5.905000000000001e-06,
+      "loss": 0.6672,
+      "step": 1638
+    },
+    {
+      "epoch": 8.195,
+      "grad_norm": 6.391067028045654,
+      "learning_rate": 5.902500000000001e-06,
+      "loss": 0.1727,
+      "step": 1639
+    },
+    {
+      "epoch": 8.2,
+      "grad_norm": 5.489355564117432,
+      "learning_rate": 5.9e-06,
+      "loss": 0.224,
+      "step": 1640
+    },
+    {
+      "epoch": 8.205,
+      "grad_norm": 14.568402290344238,
+      "learning_rate": 5.897500000000001e-06,
+      "loss": 0.5199,
+      "step": 1641
+    },
+    {
+      "epoch": 8.21,
+      "grad_norm": 3.870760202407837,
+      "learning_rate": 5.895e-06,
+      "loss": 0.1946,
+      "step": 1642
+    },
+    {
+      "epoch": 8.215,
+      "grad_norm": 3.668199062347412,
+      "learning_rate": 5.892500000000001e-06,
+      "loss": 0.3025,
+      "step": 1643
+    },
+    {
+      "epoch": 8.22,
+      "grad_norm": 61.50282669067383,
+      "learning_rate": 5.89e-06,
+      "loss": 0.892,
+      "step": 1644
+    },
+    {
+      "epoch": 8.225,
+      "grad_norm": 5.699746131896973,
+      "learning_rate": 5.887500000000001e-06,
+      "loss": 0.5281,
+      "step": 1645
+    },
+    {
+      "epoch": 8.23,
+      "grad_norm": 3.55684757232666,
+      "learning_rate": 5.885e-06,
+      "loss": 0.2093,
+      "step": 1646
+    },
+    {
+      "epoch": 8.235,
+      "grad_norm": 13.582194328308105,
+      "learning_rate": 5.882500000000001e-06,
+      "loss": 0.79,
+      "step": 1647
+    },
+    {
+      "epoch": 8.24,
+      "grad_norm": 3.951651096343994,
+      "learning_rate": 5.8800000000000005e-06,
+      "loss": 0.2359,
+      "step": 1648
+    },
+    {
+      "epoch": 8.245,
+      "grad_norm": 4.39784574508667,
+      "learning_rate": 5.8775e-06,
+      "loss": 0.6916,
+      "step": 1649
+    },
+    {
+      "epoch": 8.25,
+      "grad_norm": 4.384176731109619,
+      "learning_rate": 5.8750000000000005e-06,
+      "loss": 0.262,
+      "step": 1650
+    },
+    {
+      "epoch": 8.255,
+      "grad_norm": 6.557911396026611,
+      "learning_rate": 5.872500000000001e-06,
+      "loss": 0.5353,
+      "step": 1651
+    },
+    {
+      "epoch": 8.26,
+      "grad_norm": 5.051352024078369,
+      "learning_rate": 5.8700000000000005e-06,
+      "loss": 0.5339,
+      "step": 1652
+    },
+    {
+      "epoch": 8.265,
+      "grad_norm": 4.22047233581543,
+      "learning_rate": 5.8675e-06,
+      "loss": 0.2851,
+      "step": 1653
+    },
+    {
+      "epoch": 8.27,
+      "grad_norm": 4.885536193847656,
+      "learning_rate": 5.865000000000001e-06,
+      "loss": 0.4766,
+      "step": 1654
+    },
+    {
+      "epoch": 8.275,
+      "grad_norm": 6.879123210906982,
+      "learning_rate": 5.862500000000001e-06,
+      "loss": 0.498,
+      "step": 1655
+    },
+    {
+      "epoch": 8.28,
+      "grad_norm": 7.969574928283691,
+      "learning_rate": 5.86e-06,
+      "loss": 0.7154,
+      "step": 1656
+    },
+    {
+      "epoch": 8.285,
+      "grad_norm": 4.172405242919922,
+      "learning_rate": 5.8575e-06,
+      "loss": 0.4068,
+      "step": 1657
+    },
+    {
+      "epoch": 8.29,
+      "grad_norm": 21.117387771606445,
+      "learning_rate": 5.855000000000001e-06,
+      "loss": 0.4692,
+      "step": 1658
+    },
+    {
+      "epoch": 8.295,
+      "grad_norm": 4.596335411071777,
+      "learning_rate": 5.852500000000001e-06,
+      "loss": 0.2914,
+      "step": 1659
+    },
+    {
+      "epoch": 8.3,
+      "grad_norm": 4.141327381134033,
+      "learning_rate": 5.85e-06,
+      "loss": 0.3146,
+      "step": 1660
+    },
+    {
+      "epoch": 8.305,
+      "grad_norm": 143.5450897216797,
+      "learning_rate": 5.8475e-06,
+      "loss": 1.6095,
+      "step": 1661
+    },
+    {
+      "epoch": 8.31,
+      "grad_norm": 4.569823265075684,
+      "learning_rate": 5.845000000000001e-06,
+      "loss": 0.4274,
+      "step": 1662
+    },
+    {
+      "epoch": 8.315,
+      "grad_norm": 6.136274337768555,
+      "learning_rate": 5.842500000000001e-06,
+      "loss": 0.3358,
+      "step": 1663
+    },
+    {
+      "epoch": 8.32,
+      "grad_norm": 4.218677043914795,
+      "learning_rate": 5.84e-06,
+      "loss": 0.363,
+      "step": 1664
+    },
+    {
+      "epoch": 8.325,
+      "grad_norm": 5.347954273223877,
+      "learning_rate": 5.8375000000000004e-06,
+      "loss": 0.6615,
+      "step": 1665
+    },
+    {
+      "epoch": 8.33,
+      "grad_norm": 6.443607330322266,
+      "learning_rate": 5.835000000000001e-06,
+      "loss": 0.6703,
+      "step": 1666
+    },
+    {
+      "epoch": 8.335,
+      "grad_norm": 3.8086256980895996,
+      "learning_rate": 5.832500000000001e-06,
+      "loss": 0.3247,
+      "step": 1667
+    },
+    {
+      "epoch": 8.34,
+      "grad_norm": 4.45775842666626,
+      "learning_rate": 5.83e-06,
+      "loss": 0.4195,
+      "step": 1668
+    },
+    {
+      "epoch": 8.345,
+      "grad_norm": 5.122772693634033,
+      "learning_rate": 5.8275000000000005e-06,
+      "loss": 0.4559,
+      "step": 1669
+    },
+    {
+      "epoch": 8.35,
+      "grad_norm": 6.156404495239258,
+      "learning_rate": 5.825000000000001e-06,
+      "loss": 0.5365,
+      "step": 1670
+    },
+    {
+      "epoch": 8.355,
+      "grad_norm": 5.240762710571289,
+      "learning_rate": 5.822500000000001e-06,
+      "loss": 0.6268,
+      "step": 1671
+    },
+    {
+      "epoch": 8.36,
+      "grad_norm": 5.975583076477051,
+      "learning_rate": 5.82e-06,
+      "loss": 0.8164,
+      "step": 1672
+    },
+    {
+      "epoch": 8.365,
+      "grad_norm": 4.931427001953125,
+      "learning_rate": 5.817500000000001e-06,
+      "loss": 0.2054,
+      "step": 1673
+    },
+    {
+      "epoch": 8.37,
+      "grad_norm": 8.416400909423828,
+      "learning_rate": 5.815e-06,
+      "loss": 0.2186,
+      "step": 1674
+    },
+    {
+      "epoch": 8.375,
+      "grad_norm": 57.26616668701172,
+      "learning_rate": 5.812500000000001e-06,
+      "loss": 0.8256,
+      "step": 1675
+    },
+    {
+      "epoch": 8.38,
+      "grad_norm": 8.521218299865723,
+      "learning_rate": 5.81e-06,
+      "loss": 0.5676,
+      "step": 1676
+    },
+    {
+      "epoch": 8.385,
+      "grad_norm": 6.498100757598877,
+      "learning_rate": 5.807500000000001e-06,
+      "loss": 0.5932,
+      "step": 1677
+    },
+    {
+      "epoch": 8.39,
+      "grad_norm": 4.140294075012207,
+      "learning_rate": 5.805e-06,
+      "loss": 0.4768,
+      "step": 1678
+    },
+    {
+      "epoch": 8.395,
+      "grad_norm": 5.1078619956970215,
+      "learning_rate": 5.802500000000001e-06,
+      "loss": 0.3306,
+      "step": 1679
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": 6.272132396697998,
+      "learning_rate": 5.8e-06,
+      "loss": 0.4951,
+      "step": 1680
+    },
+    {
+      "epoch": 8.405,
+      "grad_norm": 3.3870816230773926,
+      "learning_rate": 5.7975e-06,
+      "loss": 0.3605,
+      "step": 1681
+    },
+    {
+      "epoch": 8.41,
+      "grad_norm": 5.736361503601074,
+      "learning_rate": 5.795e-06,
+      "loss": 0.4505,
+      "step": 1682
+    },
+    {
+      "epoch": 8.415,
+      "grad_norm": 7.6952128410339355,
+      "learning_rate": 5.792500000000001e-06,
+      "loss": 0.5717,
+      "step": 1683
+    },
+    {
+      "epoch": 8.42,
+      "grad_norm": 4.581590175628662,
+      "learning_rate": 5.7900000000000005e-06,
+      "loss": 0.4422,
+      "step": 1684
+    },
+    {
+      "epoch": 8.425,
+      "grad_norm": 3.8910324573516846,
+      "learning_rate": 5.7875e-06,
+      "loss": 0.1216,
+      "step": 1685
+    },
+    {
+      "epoch": 8.43,
+      "grad_norm": 3.9917750358581543,
+      "learning_rate": 5.7850000000000005e-06,
+      "loss": 0.169,
+      "step": 1686
+    },
+    {
+      "epoch": 8.435,
+      "grad_norm": 4.86334228515625,
+      "learning_rate": 5.782500000000001e-06,
+      "loss": 0.6324,
+      "step": 1687
+    },
+    {
+      "epoch": 8.44,
+      "grad_norm": 5.24333381652832,
+      "learning_rate": 5.78e-06,
+      "loss": 0.5228,
+      "step": 1688
+    },
+    {
+      "epoch": 8.445,
+      "grad_norm": 5.162416458129883,
+      "learning_rate": 5.7775e-06,
+      "loss": 0.379,
+      "step": 1689
+    },
+    {
+      "epoch": 8.45,
+      "grad_norm": 5.488912105560303,
+      "learning_rate": 5.775000000000001e-06,
+      "loss": 0.5076,
+      "step": 1690
+    },
+    {
+      "epoch": 8.455,
+      "grad_norm": 6.416134834289551,
+      "learning_rate": 5.772500000000001e-06,
+      "loss": 0.5716,
+      "step": 1691
+    },
+    {
+      "epoch": 8.46,
+      "grad_norm": 6.805329322814941,
+      "learning_rate": 5.77e-06,
+      "loss": 0.288,
+      "step": 1692
+    },
+    {
+      "epoch": 8.465,
+      "grad_norm": 3.1456215381622314,
+      "learning_rate": 5.7675e-06,
+      "loss": 0.4705,
+      "step": 1693
+    },
+    {
+      "epoch": 8.47,
+      "grad_norm": 8.816915512084961,
+      "learning_rate": 5.765000000000001e-06,
+      "loss": 0.4709,
+      "step": 1694
+    },
+    {
+      "epoch": 8.475,
+      "grad_norm": 5.169362545013428,
+      "learning_rate": 5.762500000000001e-06,
+      "loss": 0.6501,
+      "step": 1695
+    },
+    {
+      "epoch": 8.48,
+      "grad_norm": 5.722634315490723,
+      "learning_rate": 5.76e-06,
+      "loss": 0.3398,
+      "step": 1696
+    },
+    {
+      "epoch": 8.485,
+      "grad_norm": 3.3464720249176025,
+      "learning_rate": 5.7575e-06,
+      "loss": 0.2508,
+      "step": 1697
+    },
+    {
+      "epoch": 8.49,
+      "grad_norm": 5.863711357116699,
+      "learning_rate": 5.755000000000001e-06,
+      "loss": 0.5636,
+      "step": 1698
+    },
+    {
+      "epoch": 8.495,
+      "grad_norm": 46.66291046142578,
+      "learning_rate": 5.752500000000001e-06,
+      "loss": 0.6952,
+      "step": 1699
+    },
+    {
+      "epoch": 8.5,
+      "grad_norm": 5.418575286865234,
+      "learning_rate": 5.75e-06,
+      "loss": 0.4627,
+      "step": 1700
+    },
+    {
+      "epoch": 8.505,
+      "grad_norm": 4.25799560546875,
+      "learning_rate": 5.7475000000000005e-06,
+      "loss": 0.5467,
+      "step": 1701
+    },
+    {
+      "epoch": 8.51,
+      "grad_norm": 3.5235164165496826,
+      "learning_rate": 5.745000000000001e-06,
+      "loss": 0.2043,
+      "step": 1702
+    },
+    {
+      "epoch": 8.515,
+      "grad_norm": 4.1685709953308105,
+      "learning_rate": 5.7425000000000005e-06,
+      "loss": 0.6839,
+      "step": 1703
+    },
+    {
+      "epoch": 8.52,
+      "grad_norm": 6.205177307128906,
+      "learning_rate": 5.74e-06,
+      "loss": 0.4052,
+      "step": 1704
+    },
+    {
+      "epoch": 8.525,
+      "grad_norm": 3.739806890487671,
+      "learning_rate": 5.7375000000000005e-06,
+      "loss": 0.3507,
+      "step": 1705
+    },
+    {
+      "epoch": 8.53,
+      "grad_norm": 7.384704113006592,
+      "learning_rate": 5.735e-06,
+      "loss": 0.4789,
+      "step": 1706
+    },
+    {
+      "epoch": 8.535,
+      "grad_norm": 3.9006104469299316,
+      "learning_rate": 5.732500000000001e-06,
+      "loss": 0.283,
+      "step": 1707
+    },
+    {
+      "epoch": 8.54,
+      "grad_norm": 3.139653205871582,
+      "learning_rate": 5.73e-06,
+      "loss": 0.3245,
+      "step": 1708
+    },
+    {
+      "epoch": 8.545,
+      "grad_norm": 27.305801391601562,
+      "learning_rate": 5.727500000000001e-06,
+      "loss": 0.455,
+      "step": 1709
+    },
+    {
+      "epoch": 8.55,
+      "grad_norm": 4.831796169281006,
+      "learning_rate": 5.725e-06,
+      "loss": 0.5076,
+      "step": 1710
+    },
+    {
+      "epoch": 8.555,
+      "grad_norm": 4.8325514793396,
+      "learning_rate": 5.722500000000001e-06,
+      "loss": 0.4518,
+      "step": 1711
+    },
+    {
+      "epoch": 8.56,
+      "grad_norm": 4.8336405754089355,
+      "learning_rate": 5.72e-06,
+      "loss": 0.1758,
+      "step": 1712
+    },
+    {
+      "epoch": 8.565,
+      "grad_norm": 7.2622599601745605,
+      "learning_rate": 5.7175e-06,
+      "loss": 0.0735,
+      "step": 1713
+    },
+    {
+      "epoch": 8.57,
+      "grad_norm": 2.844741106033325,
+      "learning_rate": 5.715e-06,
+      "loss": 0.1901,
+      "step": 1714
+    },
+    {
+      "epoch": 8.575,
+      "grad_norm": 4.059256553649902,
+      "learning_rate": 5.712500000000001e-06,
+      "loss": 0.3719,
+      "step": 1715
+    },
+    {
+      "epoch": 8.58,
+      "grad_norm": 3.1574742794036865,
+      "learning_rate": 5.71e-06,
+      "loss": 0.185,
+      "step": 1716
+    },
+    {
+      "epoch": 8.585,
+      "grad_norm": 2.4068524837493896,
+      "learning_rate": 5.7075e-06,
+      "loss": 0.2041,
+      "step": 1717
+    },
+    {
+      "epoch": 8.59,
+      "grad_norm": 3.7470247745513916,
+      "learning_rate": 5.7050000000000004e-06,
+      "loss": 0.4895,
+      "step": 1718
+    },
+    {
+      "epoch": 8.595,
+      "grad_norm": 4.513164043426514,
+      "learning_rate": 5.702500000000001e-06,
+      "loss": 0.332,
+      "step": 1719
+    },
+    {
+      "epoch": 8.6,
+      "grad_norm": 4.094332695007324,
+      "learning_rate": 5.7e-06,
+      "loss": 0.237,
+      "step": 1720
+    },
+    {
+      "epoch": 8.605,
+      "grad_norm": 6.826866149902344,
+      "learning_rate": 5.6975e-06,
+      "loss": 0.4972,
+      "step": 1721
+    },
+    {
+      "epoch": 8.61,
+      "grad_norm": 6.046241760253906,
+      "learning_rate": 5.6950000000000005e-06,
+      "loss": 0.6849,
+      "step": 1722
+    },
+    {
+      "epoch": 8.615,
+      "grad_norm": 7.580910682678223,
+      "learning_rate": 5.692500000000001e-06,
+      "loss": 0.434,
+      "step": 1723
+    },
+    {
+      "epoch": 8.62,
+      "grad_norm": 4.75798225402832,
+      "learning_rate": 5.69e-06,
+      "loss": 0.3307,
+      "step": 1724
+    },
+    {
+      "epoch": 8.625,
+      "grad_norm": 3.2493748664855957,
+      "learning_rate": 5.6875e-06,
+      "loss": 0.5661,
+      "step": 1725
+    },
+    {
+      "epoch": 8.63,
+      "grad_norm": 15.227035522460938,
+      "learning_rate": 5.685000000000001e-06,
+      "loss": 0.8864,
+      "step": 1726
+    },
+    {
+      "epoch": 8.635,
+      "grad_norm": 5.619315147399902,
+      "learning_rate": 5.682500000000001e-06,
+      "loss": 0.3174,
+      "step": 1727
+    },
+    {
+      "epoch": 8.64,
+      "grad_norm": 5.760064601898193,
+      "learning_rate": 5.68e-06,
+      "loss": 0.4395,
+      "step": 1728
+    },
+    {
+      "epoch": 8.645,
+      "grad_norm": 2.470125675201416,
+      "learning_rate": 5.6775e-06,
+      "loss": 0.2548,
+      "step": 1729
+    },
+    {
+      "epoch": 8.65,
+      "grad_norm": 2.959062337875366,
+      "learning_rate": 5.675000000000001e-06,
+      "loss": 0.3062,
+      "step": 1730
+    },
+    {
+      "epoch": 8.655,
+      "grad_norm": 9.449959754943848,
+      "learning_rate": 5.672500000000001e-06,
+      "loss": 0.741,
+      "step": 1731
+    },
+    {
+      "epoch": 8.66,
+      "grad_norm": 6.017218112945557,
+      "learning_rate": 5.67e-06,
+      "loss": 0.5384,
+      "step": 1732
+    },
+    {
+      "epoch": 8.665,
+      "grad_norm": 4.060155868530273,
+      "learning_rate": 5.6675e-06,
+      "loss": 0.3679,
+      "step": 1733
+    },
+    {
+      "epoch": 8.67,
+      "grad_norm": 3.8354756832122803,
+      "learning_rate": 5.665000000000001e-06,
+      "loss": 0.4076,
+      "step": 1734
+    },
+    {
+      "epoch": 8.675,
+      "grad_norm": 5.768906593322754,
+      "learning_rate": 5.6625e-06,
+      "loss": 0.4087,
+      "step": 1735
+    },
+    {
+      "epoch": 8.68,
+      "grad_norm": 12.54787540435791,
+      "learning_rate": 5.66e-06,
+      "loss": 0.2846,
+      "step": 1736
+    },
+    {
+      "epoch": 8.685,
+      "grad_norm": 3.9125475883483887,
+      "learning_rate": 5.6575000000000005e-06,
+      "loss": 0.2967,
+      "step": 1737
+    },
+    {
+      "epoch": 8.69,
+      "grad_norm": 5.826800346374512,
+      "learning_rate": 5.655e-06,
+      "loss": 0.5426,
+      "step": 1738
+    },
+    {
+      "epoch": 8.695,
+      "grad_norm": 2.427877426147461,
+      "learning_rate": 5.6525000000000005e-06,
+      "loss": 0.0876,
+      "step": 1739
+    },
+    {
+      "epoch": 8.7,
+      "grad_norm": 4.07373046875,
+      "learning_rate": 5.65e-06,
+      "loss": 0.5324,
+      "step": 1740
+    },
+    {
+      "epoch": 8.705,
+      "grad_norm": 3.847028970718384,
+      "learning_rate": 5.6475000000000006e-06,
+      "loss": 0.2022,
+      "step": 1741
+    },
+    {
+      "epoch": 8.71,
+      "grad_norm": 3.818915605545044,
+      "learning_rate": 5.645e-06,
+      "loss": 0.2584,
+      "step": 1742
+    },
+    {
+      "epoch": 8.715,
+      "grad_norm": 5.629948139190674,
+      "learning_rate": 5.642500000000001e-06,
+      "loss": 0.6561,
+      "step": 1743
+    },
+    {
+      "epoch": 8.72,
+      "grad_norm": 4.908345699310303,
+      "learning_rate": 5.64e-06,
+      "loss": 0.182,
+      "step": 1744
+    },
+    {
+      "epoch": 8.725,
+      "grad_norm": 6.73772668838501,
+      "learning_rate": 5.6375e-06,
+      "loss": 0.6104,
+      "step": 1745
+    },
+    {
+      "epoch": 8.73,
+      "grad_norm": 3.9958059787750244,
+      "learning_rate": 5.635e-06,
+      "loss": 0.3328,
+      "step": 1746
+    },
+    {
+      "epoch": 8.735,
+      "grad_norm": 3.988222360610962,
+      "learning_rate": 5.632500000000001e-06,
+      "loss": 0.3292,
+      "step": 1747
+    },
+    {
+      "epoch": 8.74,
+      "grad_norm": 5.679402828216553,
+      "learning_rate": 5.63e-06,
+      "loss": 0.4125,
+      "step": 1748
+    },
+    {
+      "epoch": 8.745,
+      "grad_norm": 4.933114528656006,
+      "learning_rate": 5.6275e-06,
+      "loss": 0.4772,
+      "step": 1749
+    },
+    {
+      "epoch": 8.75,
+      "grad_norm": 4.840892314910889,
+      "learning_rate": 5.625e-06,
+      "loss": 0.348,
+      "step": 1750
+    },
+    {
+      "epoch": 8.755,
+      "grad_norm": 2.8955938816070557,
+      "learning_rate": 5.622500000000001e-06,
+      "loss": 0.5761,
+      "step": 1751
+    },
+    {
+      "epoch": 8.76,
+      "grad_norm": 4.956716060638428,
+      "learning_rate": 5.620000000000001e-06,
+      "loss": 0.7474,
+      "step": 1752
+    },
+    {
+      "epoch": 8.765,
+      "grad_norm": 5.003101825714111,
+      "learning_rate": 5.6175e-06,
+      "loss": 0.1951,
+      "step": 1753
+    },
+    {
+      "epoch": 8.77,
+      "grad_norm": 8.809060096740723,
+      "learning_rate": 5.6150000000000005e-06,
+      "loss": 1.0925,
+      "step": 1754
+    },
+    {
+      "epoch": 8.775,
+      "grad_norm": 9.425941467285156,
+      "learning_rate": 5.612500000000001e-06,
+      "loss": 0.607,
+      "step": 1755
+    },
+    {
+      "epoch": 8.78,
+      "grad_norm": 5.388036727905273,
+      "learning_rate": 5.610000000000001e-06,
+      "loss": 0.3026,
+      "step": 1756
+    },
+    {
+      "epoch": 8.785,
+      "grad_norm": 5.409789562225342,
+      "learning_rate": 5.6075e-06,
+      "loss": 0.6179,
+      "step": 1757
+    },
+    {
+      "epoch": 8.79,
+      "grad_norm": 3.0721945762634277,
+      "learning_rate": 5.6050000000000005e-06,
+      "loss": 0.4697,
+      "step": 1758
+    },
+    {
+      "epoch": 8.795,
+      "grad_norm": 3.627573013305664,
+      "learning_rate": 5.602500000000001e-06,
+      "loss": 0.2577,
+      "step": 1759
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 5.136661529541016,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 0.4348,
+      "step": 1760
+    },
+    {
+      "epoch": 8.805,
+      "grad_norm": 8.43544864654541,
+      "learning_rate": 5.5975e-06,
+      "loss": 0.7802,
+      "step": 1761
+    },
+    {
+      "epoch": 8.81,
+      "grad_norm": 12.41855525970459,
+      "learning_rate": 5.595000000000001e-06,
+      "loss": 0.7716,
+      "step": 1762
+    },
+    {
+      "epoch": 8.815,
+      "grad_norm": 4.924271106719971,
+      "learning_rate": 5.592500000000001e-06,
+      "loss": 0.5584,
+      "step": 1763
+    },
+    {
+      "epoch": 8.82,
+      "grad_norm": 4.908925533294678,
+      "learning_rate": 5.590000000000001e-06,
+      "loss": 0.5225,
+      "step": 1764
+    },
+    {
+      "epoch": 8.825,
+      "grad_norm": 7.2566022872924805,
+      "learning_rate": 5.5875e-06,
+      "loss": 0.6361,
+      "step": 1765
+    },
+    {
+      "epoch": 8.83,
+      "grad_norm": 1.9353045225143433,
+      "learning_rate": 5.585000000000001e-06,
+      "loss": 0.0687,
+      "step": 1766
+    },
+    {
+      "epoch": 8.835,
+      "grad_norm": 3.7413973808288574,
+      "learning_rate": 5.5825e-06,
+      "loss": 0.281,
+      "step": 1767
+    },
+    {
+      "epoch": 8.84,
+      "grad_norm": 5.390810012817383,
+      "learning_rate": 5.580000000000001e-06,
+      "loss": 0.4364,
+      "step": 1768
+    },
+    {
+      "epoch": 8.845,
+      "grad_norm": 6.571136474609375,
+      "learning_rate": 5.5775e-06,
+      "loss": 0.3035,
+      "step": 1769
+    },
+    {
+      "epoch": 8.85,
+      "grad_norm": 4.84925651550293,
+      "learning_rate": 5.575000000000001e-06,
+      "loss": 0.53,
+      "step": 1770
+    },
+    {
+      "epoch": 8.855,
+      "grad_norm": 4.097307205200195,
+      "learning_rate": 5.5725000000000004e-06,
+      "loss": 0.4386,
+      "step": 1771
+    },
+    {
+      "epoch": 8.86,
+      "grad_norm": 3.302672863006592,
+      "learning_rate": 5.570000000000001e-06,
+      "loss": 0.3261,
+      "step": 1772
+    },
+    {
+      "epoch": 8.865,
+      "grad_norm": 2.2704436779022217,
+      "learning_rate": 5.5675000000000005e-06,
+      "loss": 0.2912,
+      "step": 1773
+    },
+    {
+      "epoch": 8.87,
+      "grad_norm": 3.314094066619873,
+      "learning_rate": 5.565e-06,
+      "loss": 0.4174,
+      "step": 1774
+    },
+    {
+      "epoch": 8.875,
+      "grad_norm": 3.21716570854187,
+      "learning_rate": 5.5625000000000005e-06,
+      "loss": 0.119,
+      "step": 1775
+    },
+    {
+      "epoch": 8.88,
+      "grad_norm": 5.689548015594482,
+      "learning_rate": 5.560000000000001e-06,
+      "loss": 0.9342,
+      "step": 1776
+    },
+    {
+      "epoch": 8.885,
+      "grad_norm": 6.471830368041992,
+      "learning_rate": 5.5575e-06,
+      "loss": 0.6537,
+      "step": 1777
+    },
+    {
+      "epoch": 8.89,
+      "grad_norm": 6.908535957336426,
+      "learning_rate": 5.555e-06,
+      "loss": 0.2966,
+      "step": 1778
+    },
+    {
+      "epoch": 8.895,
+      "grad_norm": 8.69045639038086,
+      "learning_rate": 5.552500000000001e-06,
+      "loss": 0.5905,
+      "step": 1779
+    },
+    {
+      "epoch": 8.9,
+      "grad_norm": 5.605594158172607,
+      "learning_rate": 5.550000000000001e-06,
+      "loss": 0.2222,
+      "step": 1780
+    },
+    {
+      "epoch": 8.905,
+      "grad_norm": 5.105831146240234,
+      "learning_rate": 5.5475e-06,
+      "loss": 0.4646,
+      "step": 1781
+    },
+    {
+      "epoch": 8.91,
+      "grad_norm": 5.8985514640808105,
+      "learning_rate": 5.545e-06,
+      "loss": 0.435,
+      "step": 1782
+    },
+    {
+      "epoch": 8.915,
+      "grad_norm": 4.812410831451416,
+      "learning_rate": 5.542500000000001e-06,
+      "loss": 0.429,
+      "step": 1783
+    },
+    {
+      "epoch": 8.92,
+      "grad_norm": 6.4889373779296875,
+      "learning_rate": 5.540000000000001e-06,
+      "loss": 0.436,
+      "step": 1784
+    },
+    {
+      "epoch": 8.925,
+      "grad_norm": 4.475999355316162,
+      "learning_rate": 5.5375e-06,
+      "loss": 0.4724,
+      "step": 1785
+    },
+    {
+      "epoch": 8.93,
+      "grad_norm": 4.099257469177246,
+      "learning_rate": 5.535e-06,
+      "loss": 0.5595,
+      "step": 1786
+    },
+    {
+      "epoch": 8.935,
+      "grad_norm": 4.0820488929748535,
+      "learning_rate": 5.532500000000001e-06,
+      "loss": 0.2818,
+      "step": 1787
+    },
+    {
+      "epoch": 8.94,
+      "grad_norm": 4.221375465393066,
+      "learning_rate": 5.530000000000001e-06,
+      "loss": 0.2467,
+      "step": 1788
+    },
+    {
+      "epoch": 8.945,
+      "grad_norm": 5.823844909667969,
+      "learning_rate": 5.5275e-06,
+      "loss": 0.3006,
+      "step": 1789
+    },
+    {
+      "epoch": 8.95,
+      "grad_norm": 3.5484721660614014,
+      "learning_rate": 5.5250000000000005e-06,
+      "loss": 0.2774,
+      "step": 1790
+    },
+    {
+      "epoch": 8.955,
+      "grad_norm": 4.8533244132995605,
+      "learning_rate": 5.522500000000001e-06,
+      "loss": 0.4321,
+      "step": 1791
+    },
+    {
+      "epoch": 8.96,
+      "grad_norm": 4.344070911407471,
+      "learning_rate": 5.5200000000000005e-06,
+      "loss": 0.3602,
+      "step": 1792
+    },
+    {
+      "epoch": 8.965,
+      "grad_norm": 4.83606481552124,
+      "learning_rate": 5.5175e-06,
+      "loss": 0.291,
+      "step": 1793
+    },
+    {
+      "epoch": 8.97,
+      "grad_norm": 4.262139797210693,
+      "learning_rate": 5.5150000000000006e-06,
+      "loss": 0.2583,
+      "step": 1794
+    },
+    {
+      "epoch": 8.975,
+      "grad_norm": 4.946857929229736,
+      "learning_rate": 5.512500000000001e-06,
+      "loss": 0.4056,
+      "step": 1795
+    },
+    {
+      "epoch": 8.98,
+      "grad_norm": 7.313800811767578,
+      "learning_rate": 5.510000000000001e-06,
+      "loss": 0.4318,
+      "step": 1796
+    },
+    {
+      "epoch": 8.985,
+      "grad_norm": 3.8988306522369385,
+      "learning_rate": 5.5075e-06,
+      "loss": 0.3587,
+      "step": 1797
+    },
+    {
+      "epoch": 8.99,
+      "grad_norm": 5.003207683563232,
+      "learning_rate": 5.505000000000001e-06,
+      "loss": 0.6536,
+      "step": 1798
+    },
+    {
+      "epoch": 8.995,
+      "grad_norm": 9.12913990020752,
+      "learning_rate": 5.5025e-06,
+      "loss": 0.3151,
+      "step": 1799
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 7.565154552459717,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 0.4273,
+      "step": 1800
+    },
+    {
+      "epoch": 9.0,
+      "eval_loss": 0.26947659254074097,
+      "eval_runtime": 15.2253,
+      "eval_samples_per_second": 52.413,
+      "eval_steps_per_second": 6.568,
+      "step": 1800
+    },
+    {
+      "epoch": 9.005,
+      "grad_norm": 9.554105758666992,
+      "learning_rate": 5.4975e-06,
+      "loss": 0.2311,
+      "step": 1801
+    },
+    {
+      "epoch": 9.01,
+      "grad_norm": 5.731585502624512,
+      "learning_rate": 5.495000000000001e-06,
+      "loss": 0.5291,
+      "step": 1802
+    },
+    {
+      "epoch": 9.015,
+      "grad_norm": 4.311106204986572,
+      "learning_rate": 5.4925e-06,
+      "loss": 0.2764,
+      "step": 1803
+    },
+    {
+      "epoch": 9.02,
+      "grad_norm": 4.2923502922058105,
+      "learning_rate": 5.490000000000001e-06,
+      "loss": 0.703,
+      "step": 1804
+    },
+    {
+      "epoch": 9.025,
+      "grad_norm": 138.34739685058594,
+      "learning_rate": 5.4875e-06,
+      "loss": 1.5205,
+      "step": 1805
+    },
+    {
+      "epoch": 9.03,
+      "grad_norm": 5.35724401473999,
+      "learning_rate": 5.485e-06,
+      "loss": 0.5406,
+      "step": 1806
+    },
+    {
+      "epoch": 9.035,
+      "grad_norm": 3.0489256381988525,
+      "learning_rate": 5.4825000000000004e-06,
+      "loss": 0.0908,
+      "step": 1807
+    },
+    {
+      "epoch": 9.04,
+      "grad_norm": 4.1995768547058105,
+      "learning_rate": 5.480000000000001e-06,
+      "loss": 0.2007,
+      "step": 1808
+    },
+    {
+      "epoch": 9.045,
+      "grad_norm": 5.404238700866699,
+      "learning_rate": 5.4775e-06,
+      "loss": 0.5671,
+      "step": 1809
+    },
+    {
+      "epoch": 9.05,
+      "grad_norm": 2.952890396118164,
+      "learning_rate": 5.475e-06,
+      "loss": 0.1841,
+      "step": 1810
+    },
+    {
+      "epoch": 9.055,
+      "grad_norm": 3.362837076187134,
+      "learning_rate": 5.4725000000000005e-06,
+      "loss": 0.2415,
+      "step": 1811
+    },
+    {
+      "epoch": 9.06,
+      "grad_norm": 3.7720820903778076,
+      "learning_rate": 5.470000000000001e-06,
+      "loss": 0.404,
+      "step": 1812
+    },
+    {
+      "epoch": 9.065,
+      "grad_norm": 8.109128952026367,
+      "learning_rate": 5.4675e-06,
+      "loss": 0.4795,
+      "step": 1813
+    },
+    {
+      "epoch": 9.07,
+      "grad_norm": 4.637774467468262,
+      "learning_rate": 5.465e-06,
+      "loss": 0.3255,
+      "step": 1814
+    },
+    {
+      "epoch": 9.075,
+      "grad_norm": 8.138349533081055,
+      "learning_rate": 5.462500000000001e-06,
+      "loss": 0.5987,
+      "step": 1815
+    },
+    {
+      "epoch": 9.08,
+      "grad_norm": 3.3280906677246094,
+      "learning_rate": 5.460000000000001e-06,
+      "loss": 0.2445,
+      "step": 1816
+    },
+    {
+      "epoch": 9.085,
+      "grad_norm": 4.046419620513916,
+      "learning_rate": 5.4575e-06,
+      "loss": 0.4409,
+      "step": 1817
+    },
+    {
+      "epoch": 9.09,
+      "grad_norm": 4.38001823425293,
+      "learning_rate": 5.455e-06,
+      "loss": 0.5672,
+      "step": 1818
+    },
+    {
+      "epoch": 9.095,
+      "grad_norm": 6.213986873626709,
+      "learning_rate": 5.452500000000001e-06,
+      "loss": 0.8049,
+      "step": 1819
+    },
+    {
+      "epoch": 9.1,
+      "grad_norm": 4.9193291664123535,
+      "learning_rate": 5.450000000000001e-06,
+      "loss": 0.4557,
+      "step": 1820
+    },
+    {
+      "epoch": 9.105,
+      "grad_norm": 5.113897800445557,
+      "learning_rate": 5.4475e-06,
+      "loss": 0.8327,
+      "step": 1821
+    },
+    {
+      "epoch": 9.11,
+      "grad_norm": 4.381991863250732,
+      "learning_rate": 5.445e-06,
+      "loss": 0.2575,
+      "step": 1822
+    },
+    {
+      "epoch": 9.115,
+      "grad_norm": 3.0280239582061768,
+      "learning_rate": 5.442500000000001e-06,
+      "loss": 0.0533,
+      "step": 1823
+    },
+    {
+      "epoch": 9.12,
+      "grad_norm": 6.0273613929748535,
+      "learning_rate": 5.4400000000000004e-06,
+      "loss": 0.512,
+      "step": 1824
+    },
+    {
+      "epoch": 9.125,
+      "grad_norm": 4.142186164855957,
+      "learning_rate": 5.4375e-06,
+      "loss": 0.2657,
+      "step": 1825
+    },
+    {
+      "epoch": 9.13,
+      "grad_norm": 3.7457308769226074,
+      "learning_rate": 5.4350000000000005e-06,
+      "loss": 0.1872,
+      "step": 1826
+    },
+    {
+      "epoch": 9.135,
+      "grad_norm": 2.8844082355499268,
+      "learning_rate": 5.432500000000001e-06,
+      "loss": 0.2283,
+      "step": 1827
+    },
+    {
+      "epoch": 9.14,
+      "grad_norm": 3.990661144256592,
+      "learning_rate": 5.4300000000000005e-06,
+      "loss": 0.499,
+      "step": 1828
+    },
+    {
+      "epoch": 9.145,
+      "grad_norm": 5.1265974044799805,
+      "learning_rate": 5.4275e-06,
+      "loss": 0.1909,
+      "step": 1829
+    },
+    {
+      "epoch": 9.15,
+      "grad_norm": 3.4633679389953613,
+      "learning_rate": 5.4250000000000006e-06,
+      "loss": 0.2695,
+      "step": 1830
+    },
+    {
+      "epoch": 9.155,
+      "grad_norm": 5.874523162841797,
+      "learning_rate": 5.4225e-06,
+      "loss": 0.6577,
+      "step": 1831
+    },
+    {
+      "epoch": 9.16,
+      "grad_norm": 3.0430357456207275,
+      "learning_rate": 5.420000000000001e-06,
+      "loss": 0.1821,
+      "step": 1832
+    },
+    {
+      "epoch": 9.165,
+      "grad_norm": 3.8242931365966797,
+      "learning_rate": 5.4175e-06,
+      "loss": 0.1561,
+      "step": 1833
+    },
+    {
+      "epoch": 9.17,
+      "grad_norm": 2.685910940170288,
+      "learning_rate": 5.415000000000001e-06,
+      "loss": 0.1693,
+      "step": 1834
+    },
+    {
+      "epoch": 9.175,
+      "grad_norm": 5.766054630279541,
+      "learning_rate": 5.4125e-06,
+      "loss": 0.2888,
+      "step": 1835
+    },
+    {
+      "epoch": 9.18,
+      "grad_norm": 4.765089988708496,
+      "learning_rate": 5.410000000000001e-06,
+      "loss": 0.2861,
+      "step": 1836
+    },
+    {
+      "epoch": 9.185,
+      "grad_norm": 4.221648693084717,
+      "learning_rate": 5.4075e-06,
+      "loss": 0.2556,
+      "step": 1837
+    },
+    {
+      "epoch": 9.19,
+      "grad_norm": 4.312774658203125,
+      "learning_rate": 5.405e-06,
+      "loss": 0.4271,
+      "step": 1838
+    },
+    {
+      "epoch": 9.195,
+      "grad_norm": 8.136211395263672,
+      "learning_rate": 5.4025e-06,
+      "loss": 0.4805,
+      "step": 1839
+    },
+    {
+      "epoch": 9.2,
+      "grad_norm": 3.7029025554656982,
+      "learning_rate": 5.400000000000001e-06,
+      "loss": 0.3079,
+      "step": 1840
+    },
+    {
+      "epoch": 9.205,
+      "grad_norm": 5.608863353729248,
+      "learning_rate": 5.3974999999999996e-06,
+      "loss": 0.469,
+      "step": 1841
+    },
+    {
+      "epoch": 9.21,
+      "grad_norm": 4.8934736251831055,
+      "learning_rate": 5.395e-06,
+      "loss": 0.2395,
+      "step": 1842
+    },
+    {
+      "epoch": 9.215,
+      "grad_norm": 2.930755376815796,
+      "learning_rate": 5.3925000000000005e-06,
+      "loss": 0.2178,
+      "step": 1843
+    },
+    {
+      "epoch": 9.22,
+      "grad_norm": 4.099243640899658,
+      "learning_rate": 5.390000000000001e-06,
+      "loss": 0.4302,
+      "step": 1844
+    },
+    {
+      "epoch": 9.225,
+      "grad_norm": 2.939011812210083,
+      "learning_rate": 5.3875e-06,
+      "loss": 0.1567,
+      "step": 1845
+    },
+    {
+      "epoch": 9.23,
+      "grad_norm": 3.384868860244751,
+      "learning_rate": 5.385e-06,
+      "loss": 0.1609,
+      "step": 1846
+    },
+    {
+      "epoch": 9.235,
+      "grad_norm": 2.247408866882324,
+      "learning_rate": 5.3825000000000006e-06,
+      "loss": 0.1256,
+      "step": 1847
+    },
+    {
+      "epoch": 9.24,
+      "grad_norm": 3.7575135231018066,
+      "learning_rate": 5.380000000000001e-06,
+      "loss": 0.235,
+      "step": 1848
+    },
+    {
+      "epoch": 9.245,
+      "grad_norm": 7.772483825683594,
+      "learning_rate": 5.3775e-06,
+      "loss": 0.4211,
+      "step": 1849
+    },
+    {
+      "epoch": 9.25,
+      "grad_norm": 6.026650428771973,
+      "learning_rate": 5.375e-06,
+      "loss": 0.352,
+      "step": 1850
+    },
+    {
+      "epoch": 9.255,
+      "grad_norm": 2.7513718605041504,
+      "learning_rate": 5.372500000000001e-06,
+      "loss": 0.1311,
+      "step": 1851
+    },
+    {
+      "epoch": 9.26,
+      "grad_norm": 3.658318042755127,
+      "learning_rate": 5.370000000000001e-06,
+      "loss": 0.3403,
+      "step": 1852
+    },
+    {
+      "epoch": 9.265,
+      "grad_norm": 5.6678547859191895,
+      "learning_rate": 5.3675e-06,
+      "loss": 0.5013,
+      "step": 1853
+    },
+    {
+      "epoch": 9.27,
+      "grad_norm": 3.059683322906494,
+      "learning_rate": 5.365e-06,
+      "loss": 0.1574,
+      "step": 1854
+    },
+    {
+      "epoch": 9.275,
+      "grad_norm": 4.372241020202637,
+      "learning_rate": 5.362500000000001e-06,
+      "loss": 0.5117,
+      "step": 1855
+    },
+    {
+      "epoch": 9.28,
+      "grad_norm": 6.942160129547119,
+      "learning_rate": 5.36e-06,
+      "loss": 0.4739,
+      "step": 1856
+    },
+    {
+      "epoch": 9.285,
+      "grad_norm": 5.159568786621094,
+      "learning_rate": 5.3575e-06,
+      "loss": 0.8298,
+      "step": 1857
+    },
+    {
+      "epoch": 9.29,
+      "grad_norm": 4.878772735595703,
+      "learning_rate": 5.355e-06,
+      "loss": 0.6413,
+      "step": 1858
+    },
+    {
+      "epoch": 9.295,
+      "grad_norm": 4.288748264312744,
+      "learning_rate": 5.352500000000001e-06,
+      "loss": 0.5644,
+      "step": 1859
+    },
+    {
+      "epoch": 9.3,
+      "grad_norm": 5.2212324142456055,
+      "learning_rate": 5.3500000000000004e-06,
+      "loss": 0.5927,
+      "step": 1860
+    },
+    {
+      "epoch": 9.305,
+      "grad_norm": 3.9715945720672607,
+      "learning_rate": 5.3475e-06,
+      "loss": 0.289,
+      "step": 1861
+    },
+    {
+      "epoch": 9.31,
+      "grad_norm": 8.6132173538208,
+      "learning_rate": 5.3450000000000005e-06,
+      "loss": 0.6325,
+      "step": 1862
+    },
+    {
+      "epoch": 9.315,
+      "grad_norm": 2.8956007957458496,
+      "learning_rate": 5.3425e-06,
+      "loss": 0.3391,
+      "step": 1863
+    },
+    {
+      "epoch": 9.32,
+      "grad_norm": 5.3188157081604,
+      "learning_rate": 5.3400000000000005e-06,
+      "loss": 0.3281,
+      "step": 1864
+    },
+    {
+      "epoch": 9.325,
+      "grad_norm": 4.818702697753906,
+      "learning_rate": 5.3375e-06,
+      "loss": 0.387,
+      "step": 1865
+    },
+    {
+      "epoch": 9.33,
+      "grad_norm": 8.189957618713379,
+      "learning_rate": 5.335000000000001e-06,
+      "loss": 0.5905,
+      "step": 1866
+    },
+    {
+      "epoch": 9.335,
+      "grad_norm": 4.412904262542725,
+      "learning_rate": 5.3325e-06,
+      "loss": 0.297,
+      "step": 1867
+    },
+    {
+      "epoch": 9.34,
+      "grad_norm": 6.219573020935059,
+      "learning_rate": 5.330000000000001e-06,
+      "loss": 0.4809,
+      "step": 1868
+    },
+    {
+      "epoch": 9.345,
+      "grad_norm": 3.9412217140197754,
+      "learning_rate": 5.3275e-06,
+      "loss": 0.4075,
+      "step": 1869
+    },
+    {
+      "epoch": 9.35,
+      "grad_norm": 3.397311210632324,
+      "learning_rate": 5.325e-06,
+      "loss": 0.1581,
+      "step": 1870
+    },
+    {
+      "epoch": 9.355,
+      "grad_norm": 4.278544902801514,
+      "learning_rate": 5.3225e-06,
+      "loss": 0.4075,
+      "step": 1871
+    },
+    {
+      "epoch": 9.36,
+      "grad_norm": 7.589515686035156,
+      "learning_rate": 5.320000000000001e-06,
+      "loss": 0.7918,
+      "step": 1872
+    },
+    {
+      "epoch": 9.365,
+      "grad_norm": 9.45683765411377,
+      "learning_rate": 5.3174999999999995e-06,
+      "loss": 0.5957,
+      "step": 1873
+    },
+    {
+      "epoch": 9.37,
+      "grad_norm": 4.645104885101318,
+      "learning_rate": 5.315e-06,
+      "loss": 0.4175,
+      "step": 1874
+    },
+    {
+      "epoch": 9.375,
+      "grad_norm": 4.477837085723877,
+      "learning_rate": 5.3125e-06,
+      "loss": 0.3139,
+      "step": 1875
+    },
+    {
+      "epoch": 9.38,
+      "grad_norm": 5.632585048675537,
+      "learning_rate": 5.310000000000001e-06,
+      "loss": 0.6124,
+      "step": 1876
+    },
+    {
+      "epoch": 9.385,
+      "grad_norm": 4.637319087982178,
+      "learning_rate": 5.307500000000001e-06,
+      "loss": 0.5742,
+      "step": 1877
+    },
+    {
+      "epoch": 9.39,
+      "grad_norm": 3.2166972160339355,
+      "learning_rate": 5.305e-06,
+      "loss": 0.2832,
+      "step": 1878
+    },
+    {
+      "epoch": 9.395,
+      "grad_norm": 3.885692834854126,
+      "learning_rate": 5.3025000000000005e-06,
+      "loss": 0.2128,
+      "step": 1879
+    },
+    {
+      "epoch": 9.4,
+      "grad_norm": 4.149622440338135,
+      "learning_rate": 5.300000000000001e-06,
+      "loss": 0.3662,
+      "step": 1880
+    },
+    {
+      "epoch": 9.405,
+      "grad_norm": 6.1054301261901855,
+      "learning_rate": 5.297500000000001e-06,
+      "loss": 0.7204,
+      "step": 1881
+    },
+    {
+      "epoch": 9.41,
+      "grad_norm": 4.150166034698486,
+      "learning_rate": 5.295e-06,
+      "loss": 0.1253,
+      "step": 1882
+    },
+    {
+      "epoch": 9.415,
+      "grad_norm": 10.560813903808594,
+      "learning_rate": 5.2925000000000006e-06,
+      "loss": 0.3529,
+      "step": 1883
+    },
+    {
+      "epoch": 9.42,
+      "grad_norm": 5.411450386047363,
+      "learning_rate": 5.290000000000001e-06,
+      "loss": 0.2877,
+      "step": 1884
+    },
+    {
+      "epoch": 9.425,
+      "grad_norm": 3.118422746658325,
+      "learning_rate": 5.287500000000001e-06,
+      "loss": 0.2446,
+      "step": 1885
+    },
+    {
+      "epoch": 9.43,
+      "grad_norm": 4.170166015625,
+      "learning_rate": 5.285e-06,
+      "loss": 0.3492,
+      "step": 1886
+    },
+    {
+      "epoch": 9.435,
+      "grad_norm": 8.93328857421875,
+      "learning_rate": 5.282500000000001e-06,
+      "loss": 0.7447,
+      "step": 1887
+    },
+    {
+      "epoch": 9.44,
+      "grad_norm": 6.413856029510498,
+      "learning_rate": 5.28e-06,
+      "loss": 0.3954,
+      "step": 1888
+    },
+    {
+      "epoch": 9.445,
+      "grad_norm": 6.343822002410889,
+      "learning_rate": 5.277500000000001e-06,
+      "loss": 0.4692,
+      "step": 1889
+    },
+    {
+      "epoch": 9.45,
+      "grad_norm": 5.056811332702637,
+      "learning_rate": 5.275e-06,
+      "loss": 0.2748,
+      "step": 1890
+    },
+    {
+      "epoch": 9.455,
+      "grad_norm": 5.198853015899658,
+      "learning_rate": 5.272500000000001e-06,
+      "loss": 0.4846,
+      "step": 1891
+    },
+    {
+      "epoch": 9.46,
+      "grad_norm": 14.727540969848633,
+      "learning_rate": 5.27e-06,
+      "loss": 0.7783,
+      "step": 1892
+    },
+    {
+      "epoch": 9.465,
+      "grad_norm": 3.6885440349578857,
+      "learning_rate": 5.267500000000001e-06,
+      "loss": 0.2175,
+      "step": 1893
+    },
+    {
+      "epoch": 9.47,
+      "grad_norm": 4.8815131187438965,
+      "learning_rate": 5.265e-06,
+      "loss": 0.1827,
+      "step": 1894
+    },
+    {
+      "epoch": 9.475,
+      "grad_norm": 4.054450988769531,
+      "learning_rate": 5.2625e-06,
+      "loss": 0.4616,
+      "step": 1895
+    },
+    {
+      "epoch": 9.48,
+      "grad_norm": 14.878830909729004,
+      "learning_rate": 5.2600000000000005e-06,
+      "loss": 0.6513,
+      "step": 1896
+    },
+    {
+      "epoch": 9.485,
+      "grad_norm": 3.4330379962921143,
+      "learning_rate": 5.257500000000001e-06,
+      "loss": 0.1231,
+      "step": 1897
+    },
+    {
+      "epoch": 9.49,
+      "grad_norm": 6.713959217071533,
+      "learning_rate": 5.2550000000000005e-06,
+      "loss": 0.5888,
+      "step": 1898
+    },
+    {
+      "epoch": 9.495,
+      "grad_norm": 6.676966667175293,
+      "learning_rate": 5.2525e-06,
+      "loss": 0.2973,
+      "step": 1899
+    },
+    {
+      "epoch": 9.5,
+      "grad_norm": 20.948755264282227,
+      "learning_rate": 5.2500000000000006e-06,
+      "loss": 0.1236,
+      "step": 1900
+    },
+    {
+      "epoch": 9.505,
+      "grad_norm": 3.6223652362823486,
+      "learning_rate": 5.247500000000001e-06,
+      "loss": 0.3759,
+      "step": 1901
+    },
+    {
+      "epoch": 9.51,
+      "grad_norm": 2.901864528656006,
+      "learning_rate": 5.245e-06,
+      "loss": 0.1077,
+      "step": 1902
+    },
+    {
+      "epoch": 9.515,
+      "grad_norm": 7.900261878967285,
+      "learning_rate": 5.2425e-06,
+      "loss": 0.914,
+      "step": 1903
+    },
+    {
+      "epoch": 9.52,
+      "grad_norm": 3.4141085147857666,
+      "learning_rate": 5.240000000000001e-06,
+      "loss": 0.3361,
+      "step": 1904
+    },
+    {
+      "epoch": 9.525,
+      "grad_norm": 3.4584081172943115,
+      "learning_rate": 5.237500000000001e-06,
+      "loss": 0.4417,
+      "step": 1905
+    },
+    {
+      "epoch": 9.53,
+      "grad_norm": 6.58622407913208,
+      "learning_rate": 5.235e-06,
+      "loss": 0.521,
+      "step": 1906
+    },
+    {
+      "epoch": 9.535,
+      "grad_norm": 3.2912492752075195,
+      "learning_rate": 5.2325e-06,
+      "loss": 0.4097,
+      "step": 1907
+    },
+    {
+      "epoch": 9.54,
+      "grad_norm": 5.046441078186035,
+      "learning_rate": 5.230000000000001e-06,
+      "loss": 0.2137,
+      "step": 1908
+    },
+    {
+      "epoch": 9.545,
+      "grad_norm": 11.955282211303711,
+      "learning_rate": 5.227500000000001e-06,
+      "loss": 0.2097,
+      "step": 1909
+    },
+    {
+      "epoch": 9.55,
+      "grad_norm": 6.194656848907471,
+      "learning_rate": 5.225e-06,
+      "loss": 0.7813,
+      "step": 1910
+    },
+    {
+      "epoch": 9.555,
+      "grad_norm": 2.793523073196411,
+      "learning_rate": 5.2225e-06,
+      "loss": 0.3894,
+      "step": 1911
+    },
+    {
+      "epoch": 9.56,
+      "grad_norm": 3.7611589431762695,
+      "learning_rate": 5.220000000000001e-06,
+      "loss": 0.2645,
+      "step": 1912
+    },
+    {
+      "epoch": 9.565,
+      "grad_norm": 5.811321258544922,
+      "learning_rate": 5.217500000000001e-06,
+      "loss": 0.4395,
+      "step": 1913
+    },
+    {
+      "epoch": 9.57,
+      "grad_norm": 4.710318565368652,
+      "learning_rate": 5.215e-06,
+      "loss": 0.5585,
+      "step": 1914
+    },
+    {
+      "epoch": 9.575,
+      "grad_norm": 15.927265167236328,
+      "learning_rate": 5.2125000000000005e-06,
+      "loss": 0.5563,
+      "step": 1915
+    },
+    {
+      "epoch": 9.58,
+      "grad_norm": 7.260482311248779,
+      "learning_rate": 5.210000000000001e-06,
+      "loss": 0.5299,
+      "step": 1916
+    },
+    {
+      "epoch": 9.585,
+      "grad_norm": 9.602547645568848,
+      "learning_rate": 5.2075000000000005e-06,
+      "loss": 0.8566,
+      "step": 1917
+    },
+    {
+      "epoch": 9.59,
+      "grad_norm": 8.709080696105957,
+      "learning_rate": 5.205e-06,
+      "loss": 0.6886,
+      "step": 1918
+    },
+    {
+      "epoch": 9.595,
+      "grad_norm": 4.000442028045654,
+      "learning_rate": 5.202500000000001e-06,
+      "loss": 0.3829,
+      "step": 1919
+    },
+    {
+      "epoch": 9.6,
+      "grad_norm": 12.633285522460938,
+      "learning_rate": 5.2e-06,
+      "loss": 0.6304,
+      "step": 1920
+    },
+    {
+      "epoch": 9.605,
+      "grad_norm": 122.98428344726562,
+      "learning_rate": 5.197500000000001e-06,
+      "loss": 0.3144,
+      "step": 1921
+    },
+    {
+      "epoch": 9.61,
+      "grad_norm": 4.603846549987793,
+      "learning_rate": 5.195e-06,
+      "loss": 0.3459,
+      "step": 1922
+    },
+    {
+      "epoch": 9.615,
+      "grad_norm": 4.779256820678711,
+      "learning_rate": 5.192500000000001e-06,
+      "loss": 0.1789,
+      "step": 1923
+    },
+    {
+      "epoch": 9.62,
+      "grad_norm": 5.330276966094971,
+      "learning_rate": 5.19e-06,
+      "loss": 0.4345,
+      "step": 1924
+    },
+    {
+      "epoch": 9.625,
+      "grad_norm": 4.001300811767578,
+      "learning_rate": 5.187500000000001e-06,
+      "loss": 0.6937,
+      "step": 1925
+    },
+    {
+      "epoch": 9.63,
+      "grad_norm": 4.745081901550293,
+      "learning_rate": 5.185e-06,
+      "loss": 0.3757,
+      "step": 1926
+    },
+    {
+      "epoch": 9.635,
+      "grad_norm": 3.280548095703125,
+      "learning_rate": 5.1825e-06,
+      "loss": 0.2204,
+      "step": 1927
+    },
+    {
+      "epoch": 9.64,
+      "grad_norm": 3.275710344314575,
+      "learning_rate": 5.18e-06,
+      "loss": 0.3353,
+      "step": 1928
+    },
+    {
+      "epoch": 9.645,
+      "grad_norm": 4.557578086853027,
+      "learning_rate": 5.177500000000001e-06,
+      "loss": 0.4575,
+      "step": 1929
+    },
+    {
+      "epoch": 9.65,
+      "grad_norm": 7.695876121520996,
+      "learning_rate": 5.1750000000000004e-06,
+      "loss": 0.8023,
+      "step": 1930
+    },
+    {
+      "epoch": 9.655,
+      "grad_norm": 4.143624782562256,
+      "learning_rate": 5.1725e-06,
+      "loss": 0.1875,
+      "step": 1931
+    },
+    {
+      "epoch": 9.66,
+      "grad_norm": 3.7517449855804443,
+      "learning_rate": 5.1700000000000005e-06,
+      "loss": 0.445,
+      "step": 1932
+    },
+    {
+      "epoch": 9.665,
+      "grad_norm": 4.717888355255127,
+      "learning_rate": 5.167500000000001e-06,
+      "loss": 0.8337,
+      "step": 1933
+    },
+    {
+      "epoch": 9.67,
+      "grad_norm": 4.233070373535156,
+      "learning_rate": 5.165e-06,
+      "loss": 0.6676,
+      "step": 1934
+    },
+    {
+      "epoch": 9.675,
+      "grad_norm": 4.615227699279785,
+      "learning_rate": 5.1625e-06,
+      "loss": 0.5913,
+      "step": 1935
+    },
+    {
+      "epoch": 9.68,
+      "grad_norm": 3.741780996322632,
+      "learning_rate": 5.1600000000000006e-06,
+      "loss": 0.8163,
+      "step": 1936
+    },
+    {
+      "epoch": 9.685,
+      "grad_norm": 5.935297966003418,
+      "learning_rate": 5.157500000000001e-06,
+      "loss": 0.6823,
+      "step": 1937
+    },
+    {
+      "epoch": 9.69,
+      "grad_norm": 4.435423851013184,
+      "learning_rate": 5.155e-06,
+      "loss": 0.249,
+      "step": 1938
+    },
+    {
+      "epoch": 9.695,
+      "grad_norm": 5.649596214294434,
+      "learning_rate": 5.1525e-06,
+      "loss": 0.5953,
+      "step": 1939
+    },
+    {
+      "epoch": 9.7,
+      "grad_norm": 5.7872114181518555,
+      "learning_rate": 5.150000000000001e-06,
+      "loss": 0.5046,
+      "step": 1940
+    },
+    {
+      "epoch": 9.705,
+      "grad_norm": 1.8806332349777222,
+      "learning_rate": 5.147500000000001e-06,
+      "loss": 0.0592,
+      "step": 1941
+    },
+    {
+      "epoch": 9.71,
+      "grad_norm": 3.5933821201324463,
+      "learning_rate": 5.145e-06,
+      "loss": 0.4499,
+      "step": 1942
+    },
+    {
+      "epoch": 9.715,
+      "grad_norm": 2.683859348297119,
+      "learning_rate": 5.1425e-06,
+      "loss": 0.1447,
+      "step": 1943
+    },
+    {
+      "epoch": 9.72,
+      "grad_norm": 4.732725620269775,
+      "learning_rate": 5.140000000000001e-06,
+      "loss": 0.4358,
+      "step": 1944
+    },
+    {
+      "epoch": 9.725,
+      "grad_norm": 5.4364542961120605,
+      "learning_rate": 5.137500000000001e-06,
+      "loss": 0.6361,
+      "step": 1945
+    },
+    {
+      "epoch": 9.73,
+      "grad_norm": 5.239444732666016,
+      "learning_rate": 5.135e-06,
+      "loss": 0.4275,
+      "step": 1946
+    },
+    {
+      "epoch": 9.735,
+      "grad_norm": 10.067981719970703,
+      "learning_rate": 5.1325e-06,
+      "loss": 0.9874,
+      "step": 1947
+    },
+    {
+      "epoch": 9.74,
+      "grad_norm": 4.004093170166016,
+      "learning_rate": 5.130000000000001e-06,
+      "loss": 0.1473,
+      "step": 1948
+    },
+    {
+      "epoch": 9.745,
+      "grad_norm": 5.544090270996094,
+      "learning_rate": 5.1275000000000005e-06,
+      "loss": 0.3532,
+      "step": 1949
+    },
+    {
+      "epoch": 9.75,
+      "grad_norm": 4.347667694091797,
+      "learning_rate": 5.125e-06,
+      "loss": 0.5965,
+      "step": 1950
+    },
+    {
+      "epoch": 9.755,
+      "grad_norm": 6.379099369049072,
+      "learning_rate": 5.1225000000000005e-06,
+      "loss": 0.7988,
+      "step": 1951
+    },
+    {
+      "epoch": 9.76,
+      "grad_norm": 4.049731254577637,
+      "learning_rate": 5.12e-06,
+      "loss": 0.4572,
+      "step": 1952
+    },
+    {
+      "epoch": 9.765,
+      "grad_norm": 6.224698543548584,
+      "learning_rate": 5.1175000000000006e-06,
+      "loss": 0.3421,
+      "step": 1953
+    },
+    {
+      "epoch": 9.77,
+      "grad_norm": 8.514674186706543,
+      "learning_rate": 5.115e-06,
+      "loss": 0.6342,
+      "step": 1954
+    },
+    {
+      "epoch": 9.775,
+      "grad_norm": 3.5026698112487793,
+      "learning_rate": 5.112500000000001e-06,
+      "loss": 0.2571,
+      "step": 1955
+    },
+    {
+      "epoch": 9.78,
+      "grad_norm": 9.212541580200195,
+      "learning_rate": 5.11e-06,
+      "loss": 0.7818,
+      "step": 1956
+    },
+    {
+      "epoch": 9.785,
+      "grad_norm": 4.30979061126709,
+      "learning_rate": 5.107500000000001e-06,
+      "loss": 0.3729,
+      "step": 1957
+    },
+    {
+      "epoch": 9.79,
+      "grad_norm": 6.59328031539917,
+      "learning_rate": 5.105e-06,
+      "loss": 0.576,
+      "step": 1958
+    },
+    {
+      "epoch": 9.795,
+      "grad_norm": 7.057478427886963,
+      "learning_rate": 5.1025e-06,
+      "loss": 0.5111,
+      "step": 1959
+    },
+    {
+      "epoch": 9.8,
+      "grad_norm": 3.7829744815826416,
+      "learning_rate": 5.1e-06,
+      "loss": 0.4312,
+      "step": 1960
+    },
+    {
+      "epoch": 9.805,
+      "grad_norm": 6.042821407318115,
+      "learning_rate": 5.097500000000001e-06,
+      "loss": 0.279,
+      "step": 1961
+    },
+    {
+      "epoch": 9.81,
+      "grad_norm": 3.963782787322998,
+      "learning_rate": 5.095e-06,
+      "loss": 0.2144,
+      "step": 1962
+    },
+    {
+      "epoch": 9.815,
+      "grad_norm": 12.120972633361816,
+      "learning_rate": 5.0925e-06,
+      "loss": 0.6103,
+      "step": 1963
+    },
+    {
+      "epoch": 9.82,
+      "grad_norm": 6.759654521942139,
+      "learning_rate": 5.09e-06,
+      "loss": 0.6079,
+      "step": 1964
+    },
+    {
+      "epoch": 9.825,
+      "grad_norm": 13.75173568725586,
+      "learning_rate": 5.087500000000001e-06,
+      "loss": 0.6712,
+      "step": 1965
+    },
+    {
+      "epoch": 9.83,
+      "grad_norm": 3.555593252182007,
+      "learning_rate": 5.085e-06,
+      "loss": 0.2227,
+      "step": 1966
+    },
+    {
+      "epoch": 9.835,
+      "grad_norm": 4.190738677978516,
+      "learning_rate": 5.0825e-06,
+      "loss": 0.1825,
+      "step": 1967
+    },
+    {
+      "epoch": 9.84,
+      "grad_norm": 4.022721767425537,
+      "learning_rate": 5.0800000000000005e-06,
+      "loss": 0.3471,
+      "step": 1968
+    },
+    {
+      "epoch": 9.845,
+      "grad_norm": 3.522864818572998,
+      "learning_rate": 5.077500000000001e-06,
+      "loss": 0.3328,
+      "step": 1969
+    },
+    {
+      "epoch": 9.85,
+      "grad_norm": 4.256864070892334,
+      "learning_rate": 5.075e-06,
+      "loss": 0.2499,
+      "step": 1970
+    },
+    {
+      "epoch": 9.855,
+      "grad_norm": 4.1925506591796875,
+      "learning_rate": 5.0725e-06,
+      "loss": 0.2767,
+      "step": 1971
+    },
+    {
+      "epoch": 9.86,
+      "grad_norm": 6.286726951599121,
+      "learning_rate": 5.070000000000001e-06,
+      "loss": 0.5411,
+      "step": 1972
+    },
+    {
+      "epoch": 9.865,
+      "grad_norm": 3.210803985595703,
+      "learning_rate": 5.067500000000001e-06,
+      "loss": 0.5437,
+      "step": 1973
+    },
+    {
+      "epoch": 9.87,
+      "grad_norm": 6.139034271240234,
+      "learning_rate": 5.065e-06,
+      "loss": 0.3215,
+      "step": 1974
+    },
+    {
+      "epoch": 9.875,
+      "grad_norm": 4.8744797706604,
+      "learning_rate": 5.0625e-06,
+      "loss": 0.5565,
+      "step": 1975
+    },
+    {
+      "epoch": 9.88,
+      "grad_norm": 4.533969879150391,
+      "learning_rate": 5.060000000000001e-06,
+      "loss": 0.3812,
+      "step": 1976
+    },
+    {
+      "epoch": 9.885,
+      "grad_norm": 4.626248359680176,
+      "learning_rate": 5.057500000000001e-06,
+      "loss": 0.4421,
+      "step": 1977
+    },
+    {
+      "epoch": 9.89,
+      "grad_norm": 5.464611530303955,
+      "learning_rate": 5.055e-06,
+      "loss": 0.3407,
+      "step": 1978
+    },
+    {
+      "epoch": 9.895,
+      "grad_norm": 5.223963260650635,
+      "learning_rate": 5.0525e-06,
+      "loss": 0.5805,
+      "step": 1979
+    },
+    {
+      "epoch": 9.9,
+      "grad_norm": 3.7733609676361084,
+      "learning_rate": 5.050000000000001e-06,
+      "loss": 0.114,
+      "step": 1980
+    },
+    {
+      "epoch": 9.905,
+      "grad_norm": 10.090913772583008,
+      "learning_rate": 5.0475e-06,
+      "loss": 0.6395,
+      "step": 1981
+    },
+    {
+      "epoch": 9.91,
+      "grad_norm": 4.864150047302246,
+      "learning_rate": 5.045e-06,
+      "loss": 0.3474,
+      "step": 1982
+    },
+    {
+      "epoch": 9.915,
+      "grad_norm": 2.668281316757202,
+      "learning_rate": 5.0425000000000004e-06,
+      "loss": 0.2679,
+      "step": 1983
+    },
+    {
+      "epoch": 9.92,
+      "grad_norm": 4.866169452667236,
+      "learning_rate": 5.04e-06,
+      "loss": 0.3445,
+      "step": 1984
+    },
+    {
+      "epoch": 9.925,
+      "grad_norm": 3.4441475868225098,
+      "learning_rate": 5.0375000000000005e-06,
+      "loss": 0.2718,
+      "step": 1985
+    },
+    {
+      "epoch": 9.93,
+      "grad_norm": 3.1142284870147705,
+      "learning_rate": 5.035e-06,
+      "loss": 0.1144,
+      "step": 1986
+    },
+    {
+      "epoch": 9.935,
+      "grad_norm": 5.322275638580322,
+      "learning_rate": 5.0325000000000005e-06,
+      "loss": 0.2104,
+      "step": 1987
+    },
+    {
+      "epoch": 9.94,
+      "grad_norm": 3.6335620880126953,
+      "learning_rate": 5.03e-06,
+      "loss": 0.1043,
+      "step": 1988
+    },
+    {
+      "epoch": 9.945,
+      "grad_norm": 5.011532783508301,
+      "learning_rate": 5.0275000000000006e-06,
+      "loss": 0.2912,
+      "step": 1989
+    },
+    {
+      "epoch": 9.95,
+      "grad_norm": 6.007350444793701,
+      "learning_rate": 5.025e-06,
+      "loss": 0.5353,
+      "step": 1990
+    },
+    {
+      "epoch": 9.955,
+      "grad_norm": 3.383495807647705,
+      "learning_rate": 5.0225e-06,
+      "loss": 0.4764,
+      "step": 1991
+    },
+    {
+      "epoch": 9.96,
+      "grad_norm": 3.5531930923461914,
+      "learning_rate": 5.02e-06,
+      "loss": 0.2697,
+      "step": 1992
+    },
+    {
+      "epoch": 9.965,
+      "grad_norm": 4.557538986206055,
+      "learning_rate": 5.017500000000001e-06,
+      "loss": 0.3763,
+      "step": 1993
+    },
+    {
+      "epoch": 9.97,
+      "grad_norm": 5.343923091888428,
+      "learning_rate": 5.015e-06,
+      "loss": 0.5755,
+      "step": 1994
+    },
+    {
+      "epoch": 9.975,
+      "grad_norm": 3.443164825439453,
+      "learning_rate": 5.0125e-06,
+      "loss": 0.1294,
+      "step": 1995
+    },
+    {
+      "epoch": 9.98,
+      "grad_norm": 3.3768420219421387,
+      "learning_rate": 5.01e-06,
+      "loss": 0.2603,
+      "step": 1996
+    },
+    {
+      "epoch": 9.985,
+      "grad_norm": 13.33100414276123,
+      "learning_rate": 5.007500000000001e-06,
+      "loss": 1.163,
+      "step": 1997
+    },
+    {
+      "epoch": 9.99,
+      "grad_norm": 4.349503040313721,
+      "learning_rate": 5.0049999999999995e-06,
+      "loss": 0.3451,
+      "step": 1998
+    },
+    {
+      "epoch": 9.995,
+      "grad_norm": 8.976226806640625,
+      "learning_rate": 5.0025e-06,
+      "loss": 0.4259,
+      "step": 1999
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 21.96700096130371,
+      "learning_rate": 5e-06,
+      "loss": 1.1557,
+      "step": 2000
+    },
+    {
+      "epoch": 10.0,
+      "eval_loss": 0.25193285942077637,
+      "eval_runtime": 16.2078,
+      "eval_samples_per_second": 49.236,
+      "eval_steps_per_second": 6.17,
+      "step": 2000
+    },
+    {
+      "epoch": 10.005,
+      "grad_norm": 3.9602038860321045,
+      "learning_rate": 4.997500000000001e-06,
+      "loss": 0.2732,
+      "step": 2001
+    },
+    {
+      "epoch": 10.01,
+      "grad_norm": 5.106559753417969,
+      "learning_rate": 4.9950000000000005e-06,
+      "loss": 0.5215,
+      "step": 2002
+    },
+    {
+      "epoch": 10.015,
+      "grad_norm": 4.986967086791992,
+      "learning_rate": 4.992500000000001e-06,
+      "loss": 0.2817,
+      "step": 2003
+    },
+    {
+      "epoch": 10.02,
+      "grad_norm": 4.176296710968018,
+      "learning_rate": 4.9900000000000005e-06,
+      "loss": 0.1879,
+      "step": 2004
+    },
+    {
+      "epoch": 10.025,
+      "grad_norm": 5.558138847351074,
+      "learning_rate": 4.987500000000001e-06,
+      "loss": 0.4655,
+      "step": 2005
+    },
+    {
+      "epoch": 10.03,
+      "grad_norm": 5.982110977172852,
+      "learning_rate": 4.9850000000000006e-06,
+      "loss": 0.6474,
+      "step": 2006
+    },
+    {
+      "epoch": 10.035,
+      "grad_norm": 3.2876622676849365,
+      "learning_rate": 4.982500000000001e-06,
+      "loss": 0.3999,
+      "step": 2007
+    },
+    {
+      "epoch": 10.04,
+      "grad_norm": 10.164557456970215,
+      "learning_rate": 4.980000000000001e-06,
+      "loss": 0.7579,
+      "step": 2008
+    },
+    {
+      "epoch": 10.045,
+      "grad_norm": 5.994110584259033,
+      "learning_rate": 4.977500000000001e-06,
+      "loss": 0.3871,
+      "step": 2009
+    },
+    {
+      "epoch": 10.05,
+      "grad_norm": 3.662252187728882,
+      "learning_rate": 4.975000000000001e-06,
+      "loss": 0.1951,
+      "step": 2010
+    },
+    {
+      "epoch": 10.055,
+      "grad_norm": 4.704580307006836,
+      "learning_rate": 4.9725e-06,
+      "loss": 0.3312,
+      "step": 2011
+    },
+    {
+      "epoch": 10.06,
+      "grad_norm": 4.278799057006836,
+      "learning_rate": 4.970000000000001e-06,
+      "loss": 0.2507,
+      "step": 2012
+    },
+    {
+      "epoch": 10.065,
+      "grad_norm": 5.353710174560547,
+      "learning_rate": 4.9675e-06,
+      "loss": 0.3656,
+      "step": 2013
+    },
+    {
+      "epoch": 10.07,
+      "grad_norm": 3.5842180252075195,
+      "learning_rate": 4.965000000000001e-06,
+      "loss": 0.3869,
+      "step": 2014
+    },
+    {
+      "epoch": 10.075,
+      "grad_norm": 5.481183052062988,
+      "learning_rate": 4.9625e-06,
+      "loss": 0.2888,
+      "step": 2015
+    },
+    {
+      "epoch": 10.08,
+      "grad_norm": 5.122238636016846,
+      "learning_rate": 4.960000000000001e-06,
+      "loss": 0.5636,
+      "step": 2016
+    },
+    {
+      "epoch": 10.085,
+      "grad_norm": 5.606130123138428,
+      "learning_rate": 4.9575e-06,
+      "loss": 0.6327,
+      "step": 2017
+    },
+    {
+      "epoch": 10.09,
+      "grad_norm": 1.6847188472747803,
+      "learning_rate": 4.955e-06,
+      "loss": 0.0693,
+      "step": 2018
+    },
+    {
+      "epoch": 10.095,
+      "grad_norm": 5.812689781188965,
+      "learning_rate": 4.9525000000000004e-06,
+      "loss": 0.2567,
+      "step": 2019
+    },
+    {
+      "epoch": 10.1,
+      "grad_norm": 3.893354654312134,
+      "learning_rate": 4.95e-06,
+      "loss": 0.4272,
+      "step": 2020
+    },
+    {
+      "epoch": 10.105,
+      "grad_norm": 3.9031271934509277,
+      "learning_rate": 4.9475000000000005e-06,
+      "loss": 0.0957,
+      "step": 2021
+    },
+    {
+      "epoch": 10.11,
+      "grad_norm": 5.684651851654053,
+      "learning_rate": 4.945e-06,
+      "loss": 0.3731,
+      "step": 2022
+    },
+    {
+      "epoch": 10.115,
+      "grad_norm": 5.253362655639648,
+      "learning_rate": 4.9425000000000005e-06,
+      "loss": 0.4205,
+      "step": 2023
+    },
+    {
+      "epoch": 10.12,
+      "grad_norm": 2.9815926551818848,
+      "learning_rate": 4.94e-06,
+      "loss": 0.2386,
+      "step": 2024
+    },
+    {
+      "epoch": 10.125,
+      "grad_norm": 5.3686137199401855,
+      "learning_rate": 4.937500000000001e-06,
+      "loss": 0.5137,
+      "step": 2025
+    },
+    {
+      "epoch": 10.13,
+      "grad_norm": 3.5796005725860596,
+      "learning_rate": 4.935e-06,
+      "loss": 0.3133,
+      "step": 2026
+    },
+    {
+      "epoch": 10.135,
+      "grad_norm": 14.956930160522461,
+      "learning_rate": 4.932500000000001e-06,
+      "loss": 0.7664,
+      "step": 2027
+    },
+    {
+      "epoch": 10.14,
+      "grad_norm": 4.8340325355529785,
+      "learning_rate": 4.93e-06,
+      "loss": 0.195,
+      "step": 2028
+    },
+    {
+      "epoch": 10.145,
+      "grad_norm": 4.43747615814209,
+      "learning_rate": 4.927500000000001e-06,
+      "loss": 0.2777,
+      "step": 2029
+    },
+    {
+      "epoch": 10.15,
+      "grad_norm": 7.859208583831787,
+      "learning_rate": 4.925e-06,
+      "loss": 0.3923,
+      "step": 2030
+    },
+    {
+      "epoch": 10.155,
+      "grad_norm": 6.029805660247803,
+      "learning_rate": 4.922500000000001e-06,
+      "loss": 0.7275,
+      "step": 2031
+    },
+    {
+      "epoch": 10.16,
+      "grad_norm": 5.667142868041992,
+      "learning_rate": 4.92e-06,
+      "loss": 0.3163,
+      "step": 2032
+    },
+    {
+      "epoch": 10.165,
+      "grad_norm": 5.905954360961914,
+      "learning_rate": 4.917500000000001e-06,
+      "loss": 0.4253,
+      "step": 2033
+    },
+    {
+      "epoch": 10.17,
+      "grad_norm": 4.30299711227417,
+      "learning_rate": 4.915e-06,
+      "loss": 0.3875,
+      "step": 2034
+    },
+    {
+      "epoch": 10.175,
+      "grad_norm": 5.856101989746094,
+      "learning_rate": 4.912500000000001e-06,
+      "loss": 0.681,
+      "step": 2035
+    },
+    {
+      "epoch": 10.18,
+      "grad_norm": 4.844536304473877,
+      "learning_rate": 4.9100000000000004e-06,
+      "loss": 0.4126,
+      "step": 2036
+    },
+    {
+      "epoch": 10.185,
+      "grad_norm": 2.4630959033966064,
+      "learning_rate": 4.907500000000001e-06,
+      "loss": 0.0723,
+      "step": 2037
+    },
+    {
+      "epoch": 10.19,
+      "grad_norm": 4.6050238609313965,
+      "learning_rate": 4.9050000000000005e-06,
+      "loss": 0.391,
+      "step": 2038
+    },
+    {
+      "epoch": 10.195,
+      "grad_norm": 4.263057708740234,
+      "learning_rate": 4.902500000000001e-06,
+      "loss": 0.4582,
+      "step": 2039
+    },
+    {
+      "epoch": 10.2,
+      "grad_norm": 3.3101627826690674,
+      "learning_rate": 4.9000000000000005e-06,
+      "loss": 0.2381,
+      "step": 2040
+    },
+    {
+      "epoch": 10.205,
+      "grad_norm": 13.078474998474121,
+      "learning_rate": 4.897500000000001e-06,
+      "loss": 0.5931,
+      "step": 2041
+    },
+    {
+      "epoch": 10.21,
+      "grad_norm": 2.9836297035217285,
+      "learning_rate": 4.8950000000000006e-06,
+      "loss": 0.2433,
+      "step": 2042
+    },
+    {
+      "epoch": 10.215,
+      "grad_norm": 4.15477180480957,
+      "learning_rate": 4.8925e-06,
+      "loss": 0.369,
+      "step": 2043
+    },
+    {
+      "epoch": 10.22,
+      "grad_norm": 12.430608749389648,
+      "learning_rate": 4.890000000000001e-06,
+      "loss": 0.9001,
+      "step": 2044
+    },
+    {
+      "epoch": 10.225,
+      "grad_norm": 4.327898025512695,
+      "learning_rate": 4.8875e-06,
+      "loss": 0.74,
+      "step": 2045
+    },
+    {
+      "epoch": 10.23,
+      "grad_norm": 4.626647472381592,
+      "learning_rate": 4.885000000000001e-06,
+      "loss": 0.3964,
+      "step": 2046
+    },
+    {
+      "epoch": 10.235,
+      "grad_norm": 4.416995525360107,
+      "learning_rate": 4.8825e-06,
+      "loss": 0.2598,
+      "step": 2047
+    },
+    {
+      "epoch": 10.24,
+      "grad_norm": 3.5513739585876465,
+      "learning_rate": 4.880000000000001e-06,
+      "loss": 0.2932,
+      "step": 2048
+    },
+    {
+      "epoch": 10.245,
+      "grad_norm": 5.341632843017578,
+      "learning_rate": 4.8775e-06,
+      "loss": 0.4207,
+      "step": 2049
+    },
+    {
+      "epoch": 10.25,
+      "grad_norm": 5.528234004974365,
+      "learning_rate": 4.875e-06,
+      "loss": 0.5818,
+      "step": 2050
+    },
+    {
+      "epoch": 10.255,
+      "grad_norm": 3.776721715927124,
+      "learning_rate": 4.8725e-06,
+      "loss": 0.1835,
+      "step": 2051
+    },
+    {
+      "epoch": 10.26,
+      "grad_norm": 4.745205402374268,
+      "learning_rate": 4.87e-06,
+      "loss": 0.2904,
+      "step": 2052
+    },
+    {
+      "epoch": 10.265,
+      "grad_norm": 6.195372581481934,
+      "learning_rate": 4.8675e-06,
+      "loss": 0.4872,
+      "step": 2053
+    },
+    {
+      "epoch": 10.27,
+      "grad_norm": 3.399827718734741,
+      "learning_rate": 4.865e-06,
+      "loss": 0.2529,
+      "step": 2054
+    },
+    {
+      "epoch": 10.275,
+      "grad_norm": 2.5826122760772705,
+      "learning_rate": 4.8625000000000005e-06,
+      "loss": 0.1839,
+      "step": 2055
+    },
+    {
+      "epoch": 10.28,
+      "grad_norm": 5.116540431976318,
+      "learning_rate": 4.86e-06,
+      "loss": 0.3207,
+      "step": 2056
+    },
+    {
+      "epoch": 10.285,
+      "grad_norm": 3.655660629272461,
+      "learning_rate": 4.8575000000000005e-06,
+      "loss": 0.2072,
+      "step": 2057
+    },
+    {
+      "epoch": 10.29,
+      "grad_norm": 5.095080375671387,
+      "learning_rate": 4.855e-06,
+      "loss": 0.258,
+      "step": 2058
+    },
+    {
+      "epoch": 10.295,
+      "grad_norm": 6.436862945556641,
+      "learning_rate": 4.8525000000000006e-06,
+      "loss": 0.4421,
+      "step": 2059
+    },
+    {
+      "epoch": 10.3,
+      "grad_norm": 3.1079185009002686,
+      "learning_rate": 4.85e-06,
+      "loss": 0.2746,
+      "step": 2060
+    },
+    {
+      "epoch": 10.305,
+      "grad_norm": 3.6530158519744873,
+      "learning_rate": 4.847500000000001e-06,
+      "loss": 0.2502,
+      "step": 2061
+    },
+    {
+      "epoch": 10.31,
+      "grad_norm": 3.602027654647827,
+      "learning_rate": 4.845e-06,
+      "loss": 0.2623,
+      "step": 2062
+    },
+    {
+      "epoch": 10.315,
+      "grad_norm": 4.107232570648193,
+      "learning_rate": 4.842500000000001e-06,
+      "loss": 0.1243,
+      "step": 2063
+    },
+    {
+      "epoch": 10.32,
+      "grad_norm": 5.072595596313477,
+      "learning_rate": 4.84e-06,
+      "loss": 0.3138,
+      "step": 2064
+    },
+    {
+      "epoch": 10.325,
+      "grad_norm": 2.6169533729553223,
+      "learning_rate": 4.837500000000001e-06,
+      "loss": 0.1155,
+      "step": 2065
+    },
+    {
+      "epoch": 10.33,
+      "grad_norm": 719.6344604492188,
+      "learning_rate": 4.835e-06,
+      "loss": 1.7236,
+      "step": 2066
+    },
+    {
+      "epoch": 10.335,
+      "grad_norm": 5.256069183349609,
+      "learning_rate": 4.832500000000001e-06,
+      "loss": 0.6149,
+      "step": 2067
+    },
+    {
+      "epoch": 10.34,
+      "grad_norm": 3.910489082336426,
+      "learning_rate": 4.83e-06,
+      "loss": 0.219,
+      "step": 2068
+    },
+    {
+      "epoch": 10.345,
+      "grad_norm": 5.4245381355285645,
+      "learning_rate": 4.827500000000001e-06,
+      "loss": 0.4284,
+      "step": 2069
+    },
+    {
+      "epoch": 10.35,
+      "grad_norm": 9.79892635345459,
+      "learning_rate": 4.825e-06,
+      "loss": 0.8381,
+      "step": 2070
+    },
+    {
+      "epoch": 10.355,
+      "grad_norm": 4.817825794219971,
+      "learning_rate": 4.822500000000001e-06,
+      "loss": 0.3124,
+      "step": 2071
+    },
+    {
+      "epoch": 10.36,
+      "grad_norm": 2.83740234375,
+      "learning_rate": 4.8200000000000004e-06,
+      "loss": 0.2142,
+      "step": 2072
+    },
+    {
+      "epoch": 10.365,
+      "grad_norm": 5.456972122192383,
+      "learning_rate": 4.817500000000001e-06,
+      "loss": 0.9074,
+      "step": 2073
+    },
+    {
+      "epoch": 10.37,
+      "grad_norm": 35.035621643066406,
+      "learning_rate": 4.8150000000000005e-06,
+      "loss": 0.7177,
+      "step": 2074
+    },
+    {
+      "epoch": 10.375,
+      "grad_norm": 5.992659568786621,
+      "learning_rate": 4.8125e-06,
+      "loss": 0.892,
+      "step": 2075
+    },
+    {
+      "epoch": 10.38,
+      "grad_norm": 3.304826498031616,
+      "learning_rate": 4.8100000000000005e-06,
+      "loss": 0.2204,
+      "step": 2076
+    },
+    {
+      "epoch": 10.385,
+      "grad_norm": 4.513097286224365,
+      "learning_rate": 4.8075e-06,
+      "loss": 0.3633,
+      "step": 2077
+    },
+    {
+      "epoch": 10.39,
+      "grad_norm": 4.231469631195068,
+      "learning_rate": 4.805000000000001e-06,
+      "loss": 0.2871,
+      "step": 2078
+    },
+    {
+      "epoch": 10.395,
+      "grad_norm": 4.848938941955566,
+      "learning_rate": 4.8025e-06,
+      "loss": 0.3269,
+      "step": 2079
+    },
+    {
+      "epoch": 10.4,
+      "grad_norm": 14.516040802001953,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 0.6881,
+      "step": 2080
+    },
+    {
+      "epoch": 10.405,
+      "grad_norm": 2.6338951587677,
+      "learning_rate": 4.7975e-06,
+      "loss": 0.1417,
+      "step": 2081
+    },
+    {
+      "epoch": 10.41,
+      "grad_norm": 4.2142815589904785,
+      "learning_rate": 4.795e-06,
+      "loss": 0.3871,
+      "step": 2082
+    },
+    {
+      "epoch": 10.415,
+      "grad_norm": 9.282092094421387,
+      "learning_rate": 4.7925e-06,
+      "loss": 1.0097,
+      "step": 2083
+    },
+    {
+      "epoch": 10.42,
+      "grad_norm": 6.235434532165527,
+      "learning_rate": 4.79e-06,
+      "loss": 0.6605,
+      "step": 2084
+    },
+    {
+      "epoch": 10.425,
+      "grad_norm": 2.503915548324585,
+      "learning_rate": 4.7875e-06,
+      "loss": 0.3543,
+      "step": 2085
+    },
+    {
+      "epoch": 10.43,
+      "grad_norm": 19.252866744995117,
+      "learning_rate": 4.785e-06,
+      "loss": 0.5514,
+      "step": 2086
+    },
+    {
+      "epoch": 10.435,
+      "grad_norm": 5.912628650665283,
+      "learning_rate": 4.7825e-06,
+      "loss": 0.5284,
+      "step": 2087
+    },
+    {
+      "epoch": 10.44,
+      "grad_norm": 6.0194268226623535,
+      "learning_rate": 4.78e-06,
+      "loss": 0.2811,
+      "step": 2088
+    },
+    {
+      "epoch": 10.445,
+      "grad_norm": 6.182319164276123,
+      "learning_rate": 4.7775e-06,
+      "loss": 1.0876,
+      "step": 2089
+    },
+    {
+      "epoch": 10.45,
+      "grad_norm": 5.660794734954834,
+      "learning_rate": 4.775e-06,
+      "loss": 0.5828,
+      "step": 2090
+    },
+    {
+      "epoch": 10.455,
+      "grad_norm": 5.392878532409668,
+      "learning_rate": 4.7725000000000005e-06,
+      "loss": 0.2233,
+      "step": 2091
+    },
+    {
+      "epoch": 10.46,
+      "grad_norm": 5.065756320953369,
+      "learning_rate": 4.77e-06,
+      "loss": 0.3221,
+      "step": 2092
+    },
+    {
+      "epoch": 10.465,
+      "grad_norm": 3.910627841949463,
+      "learning_rate": 4.7675000000000005e-06,
+      "loss": 0.3387,
+      "step": 2093
+    },
+    {
+      "epoch": 10.47,
+      "grad_norm": 5.1016106605529785,
+      "learning_rate": 4.765e-06,
+      "loss": 0.2196,
+      "step": 2094
+    },
+    {
+      "epoch": 10.475,
+      "grad_norm": 4.595096111297607,
+      "learning_rate": 4.7625000000000006e-06,
+      "loss": 0.2858,
+      "step": 2095
+    },
+    {
+      "epoch": 10.48,
+      "grad_norm": 2.567638635635376,
+      "learning_rate": 4.76e-06,
+      "loss": 0.1896,
+      "step": 2096
+    },
+    {
+      "epoch": 10.485,
+      "grad_norm": 4.028942584991455,
+      "learning_rate": 4.757500000000001e-06,
+      "loss": 0.3377,
+      "step": 2097
+    },
+    {
+      "epoch": 10.49,
+      "grad_norm": 6.407106399536133,
+      "learning_rate": 4.755e-06,
+      "loss": 0.3352,
+      "step": 2098
+    },
+    {
+      "epoch": 10.495,
+      "grad_norm": 6.779458522796631,
+      "learning_rate": 4.752500000000001e-06,
+      "loss": 0.7186,
+      "step": 2099
+    },
+    {
+      "epoch": 10.5,
+      "grad_norm": 4.502418518066406,
+      "learning_rate": 4.75e-06,
+      "loss": 0.3252,
+      "step": 2100
+    },
+    {
+      "epoch": 10.505,
+      "grad_norm": 3.3509645462036133,
+      "learning_rate": 4.747500000000001e-06,
+      "loss": 0.2017,
+      "step": 2101
+    },
+    {
+      "epoch": 10.51,
+      "grad_norm": 3.6463065147399902,
+      "learning_rate": 4.745e-06,
+      "loss": 0.1099,
+      "step": 2102
+    },
+    {
+      "epoch": 10.515,
+      "grad_norm": 5.408427715301514,
+      "learning_rate": 4.742500000000001e-06,
+      "loss": 0.6924,
+      "step": 2103
+    },
+    {
+      "epoch": 10.52,
+      "grad_norm": 3.072255849838257,
+      "learning_rate": 4.74e-06,
+      "loss": 0.1758,
+      "step": 2104
+    },
+    {
+      "epoch": 10.525,
+      "grad_norm": 3.4298341274261475,
+      "learning_rate": 4.737500000000001e-06,
+      "loss": 0.6651,
+      "step": 2105
+    },
+    {
+      "epoch": 10.53,
+      "grad_norm": 5.576498031616211,
+      "learning_rate": 4.735e-06,
+      "loss": 0.4349,
+      "step": 2106
+    },
+    {
+      "epoch": 10.535,
+      "grad_norm": 4.733859539031982,
+      "learning_rate": 4.7325e-06,
+      "loss": 0.548,
+      "step": 2107
+    },
+    {
+      "epoch": 10.54,
+      "grad_norm": 3.098989963531494,
+      "learning_rate": 4.7300000000000005e-06,
+      "loss": 0.1844,
+      "step": 2108
+    },
+    {
+      "epoch": 10.545,
+      "grad_norm": 3.1821932792663574,
+      "learning_rate": 4.7275e-06,
+      "loss": 0.3244,
+      "step": 2109
+    },
+    {
+      "epoch": 10.55,
+      "grad_norm": 6.12773323059082,
+      "learning_rate": 4.7250000000000005e-06,
+      "loss": 0.438,
+      "step": 2110
+    },
+    {
+      "epoch": 10.555,
+      "grad_norm": 6.605343341827393,
+      "learning_rate": 4.7225e-06,
+      "loss": 0.4887,
+      "step": 2111
+    },
+    {
+      "epoch": 10.56,
+      "grad_norm": 5.57428503036499,
+      "learning_rate": 4.7200000000000005e-06,
+      "loss": 0.6872,
+      "step": 2112
+    },
+    {
+      "epoch": 10.565,
+      "grad_norm": 6.223629474639893,
+      "learning_rate": 4.7175e-06,
+      "loss": 0.5176,
+      "step": 2113
+    },
+    {
+      "epoch": 10.57,
+      "grad_norm": 4.843963623046875,
+      "learning_rate": 4.715e-06,
+      "loss": 0.4216,
+      "step": 2114
+    },
+    {
+      "epoch": 10.575,
+      "grad_norm": 4.061367988586426,
+      "learning_rate": 4.7125e-06,
+      "loss": 0.1598,
+      "step": 2115
+    },
+    {
+      "epoch": 10.58,
+      "grad_norm": 4.055842876434326,
+      "learning_rate": 4.71e-06,
+      "loss": 0.1711,
+      "step": 2116
+    },
+    {
+      "epoch": 10.585,
+      "grad_norm": 6.203118801116943,
+      "learning_rate": 4.7075e-06,
+      "loss": 0.6078,
+      "step": 2117
+    },
+    {
+      "epoch": 10.59,
+      "grad_norm": 4.871100425720215,
+      "learning_rate": 4.705e-06,
+      "loss": 0.4051,
+      "step": 2118
+    },
+    {
+      "epoch": 10.595,
+      "grad_norm": 4.861820220947266,
+      "learning_rate": 4.7025e-06,
+      "loss": 0.3574,
+      "step": 2119
+    },
+    {
+      "epoch": 10.6,
+      "grad_norm": 3.593967914581299,
+      "learning_rate": 4.7e-06,
+      "loss": 0.1874,
+      "step": 2120
+    },
+    {
+      "epoch": 10.605,
+      "grad_norm": 4.237033843994141,
+      "learning_rate": 4.6975e-06,
+      "loss": 0.3286,
+      "step": 2121
+    },
+    {
+      "epoch": 10.61,
+      "grad_norm": 3.3038394451141357,
+      "learning_rate": 4.695e-06,
+      "loss": 0.3527,
+      "step": 2122
+    },
+    {
+      "epoch": 10.615,
+      "grad_norm": 5.396274566650391,
+      "learning_rate": 4.6925e-06,
+      "loss": 0.5601,
+      "step": 2123
+    },
+    {
+      "epoch": 10.62,
+      "grad_norm": 4.299938201904297,
+      "learning_rate": 4.69e-06,
+      "loss": 0.1508,
+      "step": 2124
+    },
+    {
+      "epoch": 10.625,
+      "grad_norm": 36.992828369140625,
+      "learning_rate": 4.6875000000000004e-06,
+      "loss": 0.3127,
+      "step": 2125
+    },
+    {
+      "epoch": 10.63,
+      "grad_norm": 4.286873817443848,
+      "learning_rate": 4.685000000000001e-06,
+      "loss": 0.2648,
+      "step": 2126
+    },
+    {
+      "epoch": 10.635,
+      "grad_norm": 10.355278968811035,
+      "learning_rate": 4.6825000000000005e-06,
+      "loss": 0.2928,
+      "step": 2127
+    },
+    {
+      "epoch": 10.64,
+      "grad_norm": 2.0858328342437744,
+      "learning_rate": 4.680000000000001e-06,
+      "loss": 0.0527,
+      "step": 2128
+    },
+    {
+      "epoch": 10.645,
+      "grad_norm": 4.520831108093262,
+      "learning_rate": 4.6775000000000005e-06,
+      "loss": 0.206,
+      "step": 2129
+    },
+    {
+      "epoch": 10.65,
+      "grad_norm": 11.1453857421875,
+      "learning_rate": 4.675000000000001e-06,
+      "loss": 0.6276,
+      "step": 2130
+    },
+    {
+      "epoch": 10.655,
+      "grad_norm": 5.934189796447754,
+      "learning_rate": 4.672500000000001e-06,
+      "loss": 0.6181,
+      "step": 2131
+    },
+    {
+      "epoch": 10.66,
+      "grad_norm": 5.885476589202881,
+      "learning_rate": 4.670000000000001e-06,
+      "loss": 0.56,
+      "step": 2132
+    },
+    {
+      "epoch": 10.665,
+      "grad_norm": 10.607549667358398,
+      "learning_rate": 4.667500000000001e-06,
+      "loss": 0.4191,
+      "step": 2133
+    },
+    {
+      "epoch": 10.67,
+      "grad_norm": 3.6528143882751465,
+      "learning_rate": 4.665e-06,
+      "loss": 0.2134,
+      "step": 2134
+    },
+    {
+      "epoch": 10.675,
+      "grad_norm": 4.785233497619629,
+      "learning_rate": 4.662500000000001e-06,
+      "loss": 0.3191,
+      "step": 2135
+    },
+    {
+      "epoch": 10.68,
+      "grad_norm": 4.57125997543335,
+      "learning_rate": 4.66e-06,
+      "loss": 0.206,
+      "step": 2136
+    },
+    {
+      "epoch": 10.685,
+      "grad_norm": 4.283703327178955,
+      "learning_rate": 4.657500000000001e-06,
+      "loss": 0.2553,
+      "step": 2137
+    },
+    {
+      "epoch": 10.69,
+      "grad_norm": 7.024099349975586,
+      "learning_rate": 4.655e-06,
+      "loss": 0.5704,
+      "step": 2138
+    },
+    {
+      "epoch": 10.695,
+      "grad_norm": 6.5007405281066895,
+      "learning_rate": 4.652500000000001e-06,
+      "loss": 0.2141,
+      "step": 2139
+    },
+    {
+      "epoch": 10.7,
+      "grad_norm": 3.7273061275482178,
+      "learning_rate": 4.65e-06,
+      "loss": 0.4194,
+      "step": 2140
+    },
+    {
+      "epoch": 10.705,
+      "grad_norm": 5.659766674041748,
+      "learning_rate": 4.6475e-06,
+      "loss": 0.5643,
+      "step": 2141
+    },
+    {
+      "epoch": 10.71,
+      "grad_norm": 5.522407054901123,
+      "learning_rate": 4.645e-06,
+      "loss": 0.7079,
+      "step": 2142
+    },
+    {
+      "epoch": 10.715,
+      "grad_norm": 5.070888042449951,
+      "learning_rate": 4.6425e-06,
+      "loss": 0.3821,
+      "step": 2143
+    },
+    {
+      "epoch": 10.72,
+      "grad_norm": 5.120605945587158,
+      "learning_rate": 4.6400000000000005e-06,
+      "loss": 0.3457,
+      "step": 2144
+    },
+    {
+      "epoch": 10.725,
+      "grad_norm": 4.840078353881836,
+      "learning_rate": 4.6375e-06,
+      "loss": 0.4711,
+      "step": 2145
+    },
+    {
+      "epoch": 10.73,
+      "grad_norm": 4.545172214508057,
+      "learning_rate": 4.6350000000000005e-06,
+      "loss": 0.312,
+      "step": 2146
+    },
+    {
+      "epoch": 10.735,
+      "grad_norm": 2.711298942565918,
+      "learning_rate": 4.6325e-06,
+      "loss": 0.1926,
+      "step": 2147
+    },
+    {
+      "epoch": 10.74,
+      "grad_norm": 5.590420246124268,
+      "learning_rate": 4.6300000000000006e-06,
+      "loss": 0.6293,
+      "step": 2148
+    },
+    {
+      "epoch": 10.745,
+      "grad_norm": 12.013472557067871,
+      "learning_rate": 4.6275e-06,
+      "loss": 0.9329,
+      "step": 2149
+    },
+    {
+      "epoch": 10.75,
+      "grad_norm": 7.200877666473389,
+      "learning_rate": 4.625000000000001e-06,
+      "loss": 0.7628,
+      "step": 2150
+    },
+    {
+      "epoch": 10.755,
+      "grad_norm": 7.434020042419434,
+      "learning_rate": 4.6225e-06,
+      "loss": 0.618,
+      "step": 2151
+    },
+    {
+      "epoch": 10.76,
+      "grad_norm": 4.886632442474365,
+      "learning_rate": 4.620000000000001e-06,
+      "loss": 0.5574,
+      "step": 2152
+    },
+    {
+      "epoch": 10.765,
+      "grad_norm": 3.6396782398223877,
+      "learning_rate": 4.6175e-06,
+      "loss": 0.2978,
+      "step": 2153
+    },
+    {
+      "epoch": 10.77,
+      "grad_norm": 3.302323818206787,
+      "learning_rate": 4.615000000000001e-06,
+      "loss": 0.1282,
+      "step": 2154
+    },
+    {
+      "epoch": 10.775,
+      "grad_norm": 3.4387013912200928,
+      "learning_rate": 4.6125e-06,
+      "loss": 0.4099,
+      "step": 2155
+    },
+    {
+      "epoch": 10.78,
+      "grad_norm": 11.042346000671387,
+      "learning_rate": 4.610000000000001e-06,
+      "loss": 0.739,
+      "step": 2156
+    },
+    {
+      "epoch": 10.785,
+      "grad_norm": 3.8136332035064697,
+      "learning_rate": 4.6075e-06,
+      "loss": 0.2541,
+      "step": 2157
+    },
+    {
+      "epoch": 10.79,
+      "grad_norm": 5.057879447937012,
+      "learning_rate": 4.605000000000001e-06,
+      "loss": 0.3125,
+      "step": 2158
+    },
+    {
+      "epoch": 10.795,
+      "grad_norm": 7.281166076660156,
+      "learning_rate": 4.6025e-06,
+      "loss": 0.7776,
+      "step": 2159
+    },
+    {
+      "epoch": 10.8,
+      "grad_norm": 4.084758281707764,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 0.4002,
+      "step": 2160
+    },
+    {
+      "epoch": 10.805,
+      "grad_norm": 7.444750785827637,
+      "learning_rate": 4.5975000000000005e-06,
+      "loss": 0.3797,
+      "step": 2161
+    },
+    {
+      "epoch": 10.81,
+      "grad_norm": 3.8291075229644775,
+      "learning_rate": 4.595000000000001e-06,
+      "loss": 0.2503,
+      "step": 2162
+    },
+    {
+      "epoch": 10.815,
+      "grad_norm": 3.7204012870788574,
+      "learning_rate": 4.5925000000000005e-06,
+      "loss": 0.2976,
+      "step": 2163
+    },
+    {
+      "epoch": 10.82,
+      "grad_norm": 5.105237007141113,
+      "learning_rate": 4.590000000000001e-06,
+      "loss": 0.4838,
+      "step": 2164
+    },
+    {
+      "epoch": 10.825,
+      "grad_norm": 11.793466567993164,
+      "learning_rate": 4.5875000000000005e-06,
+      "loss": 0.9152,
+      "step": 2165
+    },
+    {
+      "epoch": 10.83,
+      "grad_norm": 4.9226250648498535,
+      "learning_rate": 4.585e-06,
+      "loss": 0.4385,
+      "step": 2166
+    },
+    {
+      "epoch": 10.835,
+      "grad_norm": 4.792433261871338,
+      "learning_rate": 4.582500000000001e-06,
+      "loss": 0.355,
+      "step": 2167
+    },
+    {
+      "epoch": 10.84,
+      "grad_norm": 4.086800575256348,
+      "learning_rate": 4.58e-06,
+      "loss": 0.2548,
+      "step": 2168
+    },
+    {
+      "epoch": 10.845,
+      "grad_norm": 4.503695964813232,
+      "learning_rate": 4.577500000000001e-06,
+      "loss": 0.4209,
+      "step": 2169
+    },
+    {
+      "epoch": 10.85,
+      "grad_norm": 4.808356761932373,
+      "learning_rate": 4.575e-06,
+      "loss": 0.2782,
+      "step": 2170
+    },
+    {
+      "epoch": 10.855,
+      "grad_norm": 4.278503894805908,
+      "learning_rate": 4.572500000000001e-06,
+      "loss": 0.2964,
+      "step": 2171
+    },
+    {
+      "epoch": 10.86,
+      "grad_norm": 4.5504069328308105,
+      "learning_rate": 4.57e-06,
+      "loss": 0.3173,
+      "step": 2172
+    },
+    {
+      "epoch": 10.865,
+      "grad_norm": 3.5673489570617676,
+      "learning_rate": 4.5675e-06,
+      "loss": 0.3441,
+      "step": 2173
+    },
+    {
+      "epoch": 10.87,
+      "grad_norm": 3.5556256771087646,
+      "learning_rate": 4.565e-06,
+      "loss": 0.1623,
+      "step": 2174
+    },
+    {
+      "epoch": 10.875,
+      "grad_norm": 4.961087703704834,
+      "learning_rate": 4.5625e-06,
+      "loss": 0.2198,
+      "step": 2175
+    },
+    {
+      "epoch": 10.88,
+      "grad_norm": 3.732306957244873,
+      "learning_rate": 4.56e-06,
+      "loss": 0.231,
+      "step": 2176
+    },
+    {
+      "epoch": 10.885,
+      "grad_norm": 3.991023302078247,
+      "learning_rate": 4.5575e-06,
+      "loss": 0.2224,
+      "step": 2177
+    },
+    {
+      "epoch": 10.89,
+      "grad_norm": 3.3262083530426025,
+      "learning_rate": 4.5550000000000004e-06,
+      "loss": 0.148,
+      "step": 2178
+    },
+    {
+      "epoch": 10.895,
+      "grad_norm": 3.3050827980041504,
+      "learning_rate": 4.5525e-06,
+      "loss": 0.3039,
+      "step": 2179
+    },
+    {
+      "epoch": 10.9,
+      "grad_norm": 3.484131336212158,
+      "learning_rate": 4.5500000000000005e-06,
+      "loss": 0.3695,
+      "step": 2180
+    },
+    {
+      "epoch": 10.905,
+      "grad_norm": 3.7504961490631104,
+      "learning_rate": 4.5475e-06,
+      "loss": 0.2787,
+      "step": 2181
+    },
+    {
+      "epoch": 10.91,
+      "grad_norm": 5.262529373168945,
+      "learning_rate": 4.5450000000000005e-06,
+      "loss": 0.3243,
+      "step": 2182
+    },
+    {
+      "epoch": 10.915,
+      "grad_norm": 4.565090179443359,
+      "learning_rate": 4.5425e-06,
+      "loss": 0.2351,
+      "step": 2183
+    },
+    {
+      "epoch": 10.92,
+      "grad_norm": 4.3347063064575195,
+      "learning_rate": 4.540000000000001e-06,
+      "loss": 0.2917,
+      "step": 2184
+    },
+    {
+      "epoch": 10.925,
+      "grad_norm": 5.82541561126709,
+      "learning_rate": 4.5375e-06,
+      "loss": 0.4009,
+      "step": 2185
+    },
+    {
+      "epoch": 10.93,
+      "grad_norm": 4.18768310546875,
+      "learning_rate": 4.535000000000001e-06,
+      "loss": 0.4509,
+      "step": 2186
+    },
+    {
+      "epoch": 10.935,
+      "grad_norm": 4.5996246337890625,
+      "learning_rate": 4.5325e-06,
+      "loss": 0.3619,
+      "step": 2187
+    },
+    {
+      "epoch": 10.94,
+      "grad_norm": 4.300445079803467,
+      "learning_rate": 4.530000000000001e-06,
+      "loss": 0.3522,
+      "step": 2188
+    },
+    {
+      "epoch": 10.945,
+      "grad_norm": 3.1735005378723145,
+      "learning_rate": 4.5275e-06,
+      "loss": 0.4124,
+      "step": 2189
+    },
+    {
+      "epoch": 10.95,
+      "grad_norm": 11.384907722473145,
+      "learning_rate": 4.525000000000001e-06,
+      "loss": 0.6592,
+      "step": 2190
+    },
+    {
+      "epoch": 10.955,
+      "grad_norm": 6.041499137878418,
+      "learning_rate": 4.5225e-06,
+      "loss": 0.7737,
+      "step": 2191
+    },
+    {
+      "epoch": 10.96,
+      "grad_norm": 5.701870918273926,
+      "learning_rate": 4.520000000000001e-06,
+      "loss": 0.6226,
+      "step": 2192
+    },
+    {
+      "epoch": 10.965,
+      "grad_norm": 2.69381046295166,
+      "learning_rate": 4.5175e-06,
+      "loss": 0.2342,
+      "step": 2193
+    },
+    {
+      "epoch": 10.97,
+      "grad_norm": 3.5829949378967285,
+      "learning_rate": 4.515000000000001e-06,
+      "loss": 0.2132,
+      "step": 2194
+    },
+    {
+      "epoch": 10.975,
+      "grad_norm": 3.163604974746704,
+      "learning_rate": 4.5125e-06,
+      "loss": 0.1034,
+      "step": 2195
+    },
+    {
+      "epoch": 10.98,
+      "grad_norm": 4.123225212097168,
+      "learning_rate": 4.510000000000001e-06,
+      "loss": 0.3287,
+      "step": 2196
+    },
+    {
+      "epoch": 10.985,
+      "grad_norm": 6.125432014465332,
+      "learning_rate": 4.5075000000000005e-06,
+      "loss": 0.4345,
+      "step": 2197
+    },
+    {
+      "epoch": 10.99,
+      "grad_norm": 6.796971797943115,
+      "learning_rate": 4.505e-06,
+      "loss": 0.2702,
+      "step": 2198
+    },
+    {
+      "epoch": 10.995,
+      "grad_norm": 7.182580947875977,
+      "learning_rate": 4.5025000000000005e-06,
+      "loss": 0.416,
+      "step": 2199
+    },
+    {
+      "epoch": 11.0,
+      "grad_norm": 3.9048402309417725,
+      "learning_rate": 4.5e-06,
+      "loss": 0.1617,
+      "step": 2200
+    },
+    {
+      "epoch": 11.0,
+      "eval_loss": 0.23965805768966675,
+      "eval_runtime": 15.9773,
+      "eval_samples_per_second": 49.946,
+      "eval_steps_per_second": 6.259,
+      "step": 2200
+    },
+    {
+      "epoch": 11.005,
+      "grad_norm": 4.818228721618652,
+      "learning_rate": 4.4975000000000006e-06,
+      "loss": 0.2545,
+      "step": 2201
+    },
+    {
+      "epoch": 11.01,
+      "grad_norm": 3.6002519130706787,
+      "learning_rate": 4.495e-06,
+      "loss": 0.1424,
+      "step": 2202
+    },
+    {
+      "epoch": 11.015,
+      "grad_norm": 2.879415273666382,
+      "learning_rate": 4.492500000000001e-06,
+      "loss": 0.19,
+      "step": 2203
+    },
+    {
+      "epoch": 11.02,
+      "grad_norm": 6.910661697387695,
+      "learning_rate": 4.49e-06,
+      "loss": 0.3838,
+      "step": 2204
+    },
+    {
+      "epoch": 11.025,
+      "grad_norm": 7.50560188293457,
+      "learning_rate": 4.4875e-06,
+      "loss": 0.4691,
+      "step": 2205
+    },
+    {
+      "epoch": 11.03,
+      "grad_norm": 3.6160504817962646,
+      "learning_rate": 4.485e-06,
+      "loss": 0.2874,
+      "step": 2206
+    },
+    {
+      "epoch": 11.035,
+      "grad_norm": 4.260619640350342,
+      "learning_rate": 4.4825e-06,
+      "loss": 0.3736,
+      "step": 2207
+    },
+    {
+      "epoch": 11.04,
+      "grad_norm": 8.293294906616211,
+      "learning_rate": 4.48e-06,
+      "loss": 0.5132,
+      "step": 2208
+    },
+    {
+      "epoch": 11.045,
+      "grad_norm": 4.423193454742432,
+      "learning_rate": 4.4775e-06,
+      "loss": 0.2743,
+      "step": 2209
+    },
+    {
+      "epoch": 11.05,
+      "grad_norm": 3.9291443824768066,
+      "learning_rate": 4.475e-06,
+      "loss": 0.1498,
+      "step": 2210
+    },
+    {
+      "epoch": 11.055,
+      "grad_norm": 5.721567153930664,
+      "learning_rate": 4.4725e-06,
+      "loss": 0.2553,
+      "step": 2211
+    },
+    {
+      "epoch": 11.06,
+      "grad_norm": 3.8018274307250977,
+      "learning_rate": 4.47e-06,
+      "loss": 0.458,
+      "step": 2212
+    },
+    {
+      "epoch": 11.065,
+      "grad_norm": 4.295903205871582,
+      "learning_rate": 4.4675e-06,
+      "loss": 0.4284,
+      "step": 2213
+    },
+    {
+      "epoch": 11.07,
+      "grad_norm": 5.214331150054932,
+      "learning_rate": 4.4650000000000004e-06,
+      "loss": 0.4574,
+      "step": 2214
+    },
+    {
+      "epoch": 11.075,
+      "grad_norm": 1.9817272424697876,
+      "learning_rate": 4.4625e-06,
+      "loss": 0.048,
+      "step": 2215
+    },
+    {
+      "epoch": 11.08,
+      "grad_norm": 9.845718383789062,
+      "learning_rate": 4.4600000000000005e-06,
+      "loss": 0.1773,
+      "step": 2216
+    },
+    {
+      "epoch": 11.085,
+      "grad_norm": 3.769869089126587,
+      "learning_rate": 4.4575e-06,
+      "loss": 0.2387,
+      "step": 2217
+    },
+    {
+      "epoch": 11.09,
+      "grad_norm": 4.819910049438477,
+      "learning_rate": 4.4550000000000005e-06,
+      "loss": 0.3983,
+      "step": 2218
+    },
+    {
+      "epoch": 11.095,
+      "grad_norm": 7.21164083480835,
+      "learning_rate": 4.4525e-06,
+      "loss": 0.4152,
+      "step": 2219
+    },
+    {
+      "epoch": 11.1,
+      "grad_norm": 4.018520832061768,
+      "learning_rate": 4.450000000000001e-06,
+      "loss": 0.1641,
+      "step": 2220
+    },
+    {
+      "epoch": 11.105,
+      "grad_norm": 5.574578762054443,
+      "learning_rate": 4.4475e-06,
+      "loss": 0.6912,
+      "step": 2221
+    },
+    {
+      "epoch": 11.11,
+      "grad_norm": 4.258738040924072,
+      "learning_rate": 4.445000000000001e-06,
+      "loss": 0.3036,
+      "step": 2222
+    },
+    {
+      "epoch": 11.115,
+      "grad_norm": 4.83868932723999,
+      "learning_rate": 4.4425e-06,
+      "loss": 0.4539,
+      "step": 2223
+    },
+    {
+      "epoch": 11.12,
+      "grad_norm": 3.5753378868103027,
+      "learning_rate": 4.440000000000001e-06,
+      "loss": 0.1722,
+      "step": 2224
+    },
+    {
+      "epoch": 11.125,
+      "grad_norm": 4.223999977111816,
+      "learning_rate": 4.4375e-06,
+      "loss": 0.2151,
+      "step": 2225
+    },
+    {
+      "epoch": 11.13,
+      "grad_norm": 5.0814208984375,
+      "learning_rate": 4.435000000000001e-06,
+      "loss": 0.5294,
+      "step": 2226
+    },
+    {
+      "epoch": 11.135,
+      "grad_norm": 6.96180534362793,
+      "learning_rate": 4.4325e-06,
+      "loss": 0.5936,
+      "step": 2227
+    },
+    {
+      "epoch": 11.14,
+      "grad_norm": 4.053010940551758,
+      "learning_rate": 4.430000000000001e-06,
+      "loss": 0.6401,
+      "step": 2228
+    },
+    {
+      "epoch": 11.145,
+      "grad_norm": 3.7808470726013184,
+      "learning_rate": 4.4275e-06,
+      "loss": 0.3754,
+      "step": 2229
+    },
+    {
+      "epoch": 11.15,
+      "grad_norm": 4.531346797943115,
+      "learning_rate": 4.425e-06,
+      "loss": 0.2416,
+      "step": 2230
+    },
+    {
+      "epoch": 11.155,
+      "grad_norm": 4.5211873054504395,
+      "learning_rate": 4.4225000000000004e-06,
+      "loss": 0.4252,
+      "step": 2231
+    },
+    {
+      "epoch": 11.16,
+      "grad_norm": 4.181341171264648,
+      "learning_rate": 4.42e-06,
+      "loss": 0.4183,
+      "step": 2232
+    },
+    {
+      "epoch": 11.165,
+      "grad_norm": 4.11368465423584,
+      "learning_rate": 4.4175000000000005e-06,
+      "loss": 0.3271,
+      "step": 2233
+    },
+    {
+      "epoch": 11.17,
+      "grad_norm": 4.21378755569458,
+      "learning_rate": 4.415e-06,
+      "loss": 0.3673,
+      "step": 2234
+    },
+    {
+      "epoch": 11.175,
+      "grad_norm": 4.606934547424316,
+      "learning_rate": 4.4125000000000005e-06,
+      "loss": 0.1622,
+      "step": 2235
+    },
+    {
+      "epoch": 11.18,
+      "grad_norm": 4.074502944946289,
+      "learning_rate": 4.41e-06,
+      "loss": 0.2639,
+      "step": 2236
+    },
+    {
+      "epoch": 11.185,
+      "grad_norm": 4.500039100646973,
+      "learning_rate": 4.4075e-06,
+      "loss": 0.4723,
+      "step": 2237
+    },
+    {
+      "epoch": 11.19,
+      "grad_norm": 2.5529093742370605,
+      "learning_rate": 4.405e-06,
+      "loss": 0.2011,
+      "step": 2238
+    },
+    {
+      "epoch": 11.195,
+      "grad_norm": 2.8829782009124756,
+      "learning_rate": 4.4025e-06,
+      "loss": 0.2931,
+      "step": 2239
+    },
+    {
+      "epoch": 11.2,
+      "grad_norm": 4.489750862121582,
+      "learning_rate": 4.4e-06,
+      "loss": 0.1626,
+      "step": 2240
+    },
+    {
+      "epoch": 11.205,
+      "grad_norm": 13.069165229797363,
+      "learning_rate": 4.3975e-06,
+      "loss": 0.9695,
+      "step": 2241
+    },
+    {
+      "epoch": 11.21,
+      "grad_norm": 4.709218502044678,
+      "learning_rate": 4.395e-06,
+      "loss": 0.1716,
+      "step": 2242
+    },
+    {
+      "epoch": 11.215,
+      "grad_norm": 5.72643518447876,
+      "learning_rate": 4.3925e-06,
+      "loss": 0.3493,
+      "step": 2243
+    },
+    {
+      "epoch": 11.22,
+      "grad_norm": 6.240438938140869,
+      "learning_rate": 4.39e-06,
+      "loss": 0.2711,
+      "step": 2244
+    },
+    {
+      "epoch": 11.225,
+      "grad_norm": 2.788637161254883,
+      "learning_rate": 4.3875e-06,
+      "loss": 0.1462,
+      "step": 2245
+    },
+    {
+      "epoch": 11.23,
+      "grad_norm": 6.433646202087402,
+      "learning_rate": 4.385e-06,
+      "loss": 0.2566,
+      "step": 2246
+    },
+    {
+      "epoch": 11.235,
+      "grad_norm": 15.927465438842773,
+      "learning_rate": 4.3825e-06,
+      "loss": 0.2111,
+      "step": 2247
+    },
+    {
+      "epoch": 11.24,
+      "grad_norm": 3.339094638824463,
+      "learning_rate": 4.38e-06,
+      "loss": 0.1302,
+      "step": 2248
+    },
+    {
+      "epoch": 11.245,
+      "grad_norm": 5.308197021484375,
+      "learning_rate": 4.3775e-06,
+      "loss": 0.2214,
+      "step": 2249
+    },
+    {
+      "epoch": 11.25,
+      "grad_norm": 4.219059467315674,
+      "learning_rate": 4.3750000000000005e-06,
+      "loss": 0.4063,
+      "step": 2250
+    },
+    {
+      "epoch": 11.255,
+      "grad_norm": 6.414569854736328,
+      "learning_rate": 4.372500000000001e-06,
+      "loss": 0.6542,
+      "step": 2251
+    },
+    {
+      "epoch": 11.26,
+      "grad_norm": 12.879244804382324,
+      "learning_rate": 4.3700000000000005e-06,
+      "loss": 1.0715,
+      "step": 2252
+    },
+    {
+      "epoch": 11.265,
+      "grad_norm": 4.9547271728515625,
+      "learning_rate": 4.367500000000001e-06,
+      "loss": 0.3675,
+      "step": 2253
+    },
+    {
+      "epoch": 11.27,
+      "grad_norm": 4.765869617462158,
+      "learning_rate": 4.3650000000000006e-06,
+      "loss": 0.525,
+      "step": 2254
+    },
+    {
+      "epoch": 11.275,
+      "grad_norm": 5.108703136444092,
+      "learning_rate": 4.362500000000001e-06,
+      "loss": 0.3993,
+      "step": 2255
+    },
+    {
+      "epoch": 11.28,
+      "grad_norm": 10.423528671264648,
+      "learning_rate": 4.360000000000001e-06,
+      "loss": 1.3329,
+      "step": 2256
+    },
+    {
+      "epoch": 11.285,
+      "grad_norm": 5.517879962921143,
+      "learning_rate": 4.3575e-06,
+      "loss": 0.4825,
+      "step": 2257
+    },
+    {
+      "epoch": 11.29,
+      "grad_norm": 3.1662089824676514,
+      "learning_rate": 4.355000000000001e-06,
+      "loss": 0.1992,
+      "step": 2258
+    },
+    {
+      "epoch": 11.295,
+      "grad_norm": 3.091914653778076,
+      "learning_rate": 4.3525e-06,
+      "loss": 0.1718,
+      "step": 2259
+    },
+    {
+      "epoch": 11.3,
+      "grad_norm": 3.3534984588623047,
+      "learning_rate": 4.350000000000001e-06,
+      "loss": 0.2291,
+      "step": 2260
+    },
+    {
+      "epoch": 11.305,
+      "grad_norm": 3.0282843112945557,
+      "learning_rate": 4.3475e-06,
+      "loss": 0.1827,
+      "step": 2261
+    },
+    {
+      "epoch": 11.31,
+      "grad_norm": 6.626020431518555,
+      "learning_rate": 4.345000000000001e-06,
+      "loss": 0.6197,
+      "step": 2262
+    },
+    {
+      "epoch": 11.315,
+      "grad_norm": 4.06280517578125,
+      "learning_rate": 4.3425e-06,
+      "loss": 0.3989,
+      "step": 2263
+    },
+    {
+      "epoch": 11.32,
+      "grad_norm": 4.157721042633057,
+      "learning_rate": 4.34e-06,
+      "loss": 0.2734,
+      "step": 2264
+    },
+    {
+      "epoch": 11.325,
+      "grad_norm": 4.3070387840271,
+      "learning_rate": 4.3375e-06,
+      "loss": 0.2456,
+      "step": 2265
+    },
+    {
+      "epoch": 11.33,
+      "grad_norm": 5.310956001281738,
+      "learning_rate": 4.335e-06,
+      "loss": 0.6845,
+      "step": 2266
+    },
+    {
+      "epoch": 11.335,
+      "grad_norm": 2.673309803009033,
+      "learning_rate": 4.3325000000000004e-06,
+      "loss": 0.1364,
+      "step": 2267
+    },
+    {
+      "epoch": 11.34,
+      "grad_norm": 17.121192932128906,
+      "learning_rate": 4.33e-06,
+      "loss": 0.3932,
+      "step": 2268
+    },
+    {
+      "epoch": 11.345,
+      "grad_norm": 8.0404634475708,
+      "learning_rate": 4.3275000000000005e-06,
+      "loss": 0.7887,
+      "step": 2269
+    },
+    {
+      "epoch": 11.35,
+      "grad_norm": 10.471095085144043,
+      "learning_rate": 4.325e-06,
+      "loss": 0.8049,
+      "step": 2270
+    },
+    {
+      "epoch": 11.355,
+      "grad_norm": 4.282561302185059,
+      "learning_rate": 4.3225000000000005e-06,
+      "loss": 0.3108,
+      "step": 2271
+    },
+    {
+      "epoch": 11.36,
+      "grad_norm": 3.7814695835113525,
+      "learning_rate": 4.32e-06,
+      "loss": 0.1939,
+      "step": 2272
+    },
+    {
+      "epoch": 11.365,
+      "grad_norm": 4.249180793762207,
+      "learning_rate": 4.317500000000001e-06,
+      "loss": 0.4686,
+      "step": 2273
+    },
+    {
+      "epoch": 11.37,
+      "grad_norm": 3.262373685836792,
+      "learning_rate": 4.315e-06,
+      "loss": 0.1909,
+      "step": 2274
+    },
+    {
+      "epoch": 11.375,
+      "grad_norm": 6.9596266746521,
+      "learning_rate": 4.312500000000001e-06,
+      "loss": 0.855,
+      "step": 2275
+    },
+    {
+      "epoch": 11.38,
+      "grad_norm": 3.397547721862793,
+      "learning_rate": 4.31e-06,
+      "loss": 0.2387,
+      "step": 2276
+    },
+    {
+      "epoch": 11.385,
+      "grad_norm": 4.92538595199585,
+      "learning_rate": 4.307500000000001e-06,
+      "loss": 0.2707,
+      "step": 2277
+    },
+    {
+      "epoch": 11.39,
+      "grad_norm": 2.114370822906494,
+      "learning_rate": 4.305e-06,
+      "loss": 0.1443,
+      "step": 2278
+    },
+    {
+      "epoch": 11.395,
+      "grad_norm": 7.048013210296631,
+      "learning_rate": 4.302500000000001e-06,
+      "loss": 0.4319,
+      "step": 2279
+    },
+    {
+      "epoch": 11.4,
+      "grad_norm": 5.883264541625977,
+      "learning_rate": 4.3e-06,
+      "loss": 0.6438,
+      "step": 2280
+    },
+    {
+      "epoch": 11.405,
+      "grad_norm": 5.028811454772949,
+      "learning_rate": 4.297500000000001e-06,
+      "loss": 0.4277,
+      "step": 2281
+    },
+    {
+      "epoch": 11.41,
+      "grad_norm": 4.68692684173584,
+      "learning_rate": 4.295e-06,
+      "loss": 0.4481,
+      "step": 2282
+    },
+    {
+      "epoch": 11.415,
+      "grad_norm": 4.5374040603637695,
+      "learning_rate": 4.292500000000001e-06,
+      "loss": 0.1169,
+      "step": 2283
+    },
+    {
+      "epoch": 11.42,
+      "grad_norm": 4.615859508514404,
+      "learning_rate": 4.2900000000000004e-06,
+      "loss": 0.1943,
+      "step": 2284
+    },
+    {
+      "epoch": 11.425,
+      "grad_norm": 5.435596466064453,
+      "learning_rate": 4.287500000000001e-06,
+      "loss": 0.3012,
+      "step": 2285
+    },
+    {
+      "epoch": 11.43,
+      "grad_norm": 21.19466209411621,
+      "learning_rate": 4.2850000000000005e-06,
+      "loss": 0.9894,
+      "step": 2286
+    },
+    {
+      "epoch": 11.435,
+      "grad_norm": 5.5940752029418945,
+      "learning_rate": 4.282500000000001e-06,
+      "loss": 0.2751,
+      "step": 2287
+    },
+    {
+      "epoch": 11.44,
+      "grad_norm": 9.273931503295898,
+      "learning_rate": 4.2800000000000005e-06,
+      "loss": 0.696,
+      "step": 2288
+    },
+    {
+      "epoch": 11.445,
+      "grad_norm": 3.7645857334136963,
+      "learning_rate": 4.2775e-06,
+      "loss": 0.2415,
+      "step": 2289
+    },
+    {
+      "epoch": 11.45,
+      "grad_norm": 5.528829097747803,
+      "learning_rate": 4.2750000000000006e-06,
+      "loss": 0.4048,
+      "step": 2290
+    },
+    {
+      "epoch": 11.455,
+      "grad_norm": 5.803488254547119,
+      "learning_rate": 4.2725e-06,
+      "loss": 0.2045,
+      "step": 2291
+    },
+    {
+      "epoch": 11.46,
+      "grad_norm": 3.3351376056671143,
+      "learning_rate": 4.270000000000001e-06,
+      "loss": 0.4494,
+      "step": 2292
+    },
+    {
+      "epoch": 11.465,
+      "grad_norm": 2.708376407623291,
+      "learning_rate": 4.2675e-06,
+      "loss": 0.0904,
+      "step": 2293
+    },
+    {
+      "epoch": 11.47,
+      "grad_norm": 6.012026309967041,
+      "learning_rate": 4.265000000000001e-06,
+      "loss": 0.4924,
+      "step": 2294
+    },
+    {
+      "epoch": 11.475,
+      "grad_norm": 3.3292744159698486,
+      "learning_rate": 4.2625e-06,
+      "loss": 0.2446,
+      "step": 2295
+    },
+    {
+      "epoch": 11.48,
+      "grad_norm": 3.8096694946289062,
+      "learning_rate": 4.26e-06,
+      "loss": 0.2417,
+      "step": 2296
+    },
+    {
+      "epoch": 11.485,
+      "grad_norm": 5.135099411010742,
+      "learning_rate": 4.2575e-06,
+      "loss": 0.4597,
+      "step": 2297
+    },
+    {
+      "epoch": 11.49,
+      "grad_norm": 3.8893628120422363,
+      "learning_rate": 4.255e-06,
+      "loss": 0.2008,
+      "step": 2298
+    },
+    {
+      "epoch": 11.495,
+      "grad_norm": 3.5078275203704834,
+      "learning_rate": 4.2525e-06,
+      "loss": 0.3571,
+      "step": 2299
+    },
+    {
+      "epoch": 11.5,
+      "grad_norm": 1.9357882738113403,
+      "learning_rate": 4.25e-06,
+      "loss": 0.1456,
+      "step": 2300
+    },
+    {
+      "epoch": 11.505,
+      "grad_norm": 3.5266592502593994,
+      "learning_rate": 4.2475e-06,
+      "loss": 0.2467,
+      "step": 2301
+    },
+    {
+      "epoch": 11.51,
+      "grad_norm": 3.0292744636535645,
+      "learning_rate": 4.245e-06,
+      "loss": 0.339,
+      "step": 2302
+    },
+    {
+      "epoch": 11.515,
+      "grad_norm": 9.980597496032715,
+      "learning_rate": 4.2425000000000005e-06,
+      "loss": 0.3037,
+      "step": 2303
+    },
+    {
+      "epoch": 11.52,
+      "grad_norm": 4.3424177169799805,
+      "learning_rate": 4.24e-06,
+      "loss": 0.2654,
+      "step": 2304
+    },
+    {
+      "epoch": 11.525,
+      "grad_norm": 6.575417518615723,
+      "learning_rate": 4.2375000000000005e-06,
+      "loss": 0.5836,
+      "step": 2305
+    },
+    {
+      "epoch": 11.53,
+      "grad_norm": 9.216874122619629,
+      "learning_rate": 4.235e-06,
+      "loss": 0.256,
+      "step": 2306
+    },
+    {
+      "epoch": 11.535,
+      "grad_norm": 4.97996711730957,
+      "learning_rate": 4.2325000000000006e-06,
+      "loss": 0.3881,
+      "step": 2307
+    },
+    {
+      "epoch": 11.54,
+      "grad_norm": 3.889450788497925,
+      "learning_rate": 4.23e-06,
+      "loss": 0.4359,
+      "step": 2308
+    },
+    {
+      "epoch": 11.545,
+      "grad_norm": 2.1790573596954346,
+      "learning_rate": 4.227500000000001e-06,
+      "loss": 0.2035,
+      "step": 2309
+    },
+    {
+      "epoch": 11.55,
+      "grad_norm": 3.7957513332366943,
+      "learning_rate": 4.225e-06,
+      "loss": 0.3621,
+      "step": 2310
+    },
+    {
+      "epoch": 11.555,
+      "grad_norm": 3.9984192848205566,
+      "learning_rate": 4.222500000000001e-06,
+      "loss": 0.3028,
+      "step": 2311
+    },
+    {
+      "epoch": 11.56,
+      "grad_norm": 16.344308853149414,
+      "learning_rate": 4.22e-06,
+      "loss": 0.9855,
+      "step": 2312
+    },
+    {
+      "epoch": 11.565,
+      "grad_norm": 5.1493024826049805,
+      "learning_rate": 4.217500000000001e-06,
+      "loss": 0.3415,
+      "step": 2313
+    },
+    {
+      "epoch": 11.57,
+      "grad_norm": 3.5031473636627197,
+      "learning_rate": 4.215e-06,
+      "loss": 0.2584,
+      "step": 2314
+    },
+    {
+      "epoch": 11.575,
+      "grad_norm": 5.803607940673828,
+      "learning_rate": 4.212500000000001e-06,
+      "loss": 0.7998,
+      "step": 2315
+    },
+    {
+      "epoch": 11.58,
+      "grad_norm": 4.164289474487305,
+      "learning_rate": 4.21e-06,
+      "loss": 0.2746,
+      "step": 2316
+    },
+    {
+      "epoch": 11.585,
+      "grad_norm": 4.619289875030518,
+      "learning_rate": 4.207500000000001e-06,
+      "loss": 0.3893,
+      "step": 2317
+    },
+    {
+      "epoch": 11.59,
+      "grad_norm": 2.6296751499176025,
+      "learning_rate": 4.205e-06,
+      "loss": 0.3638,
+      "step": 2318
+    },
+    {
+      "epoch": 11.595,
+      "grad_norm": 13.854787826538086,
+      "learning_rate": 4.202500000000001e-06,
+      "loss": 0.191,
+      "step": 2319
+    },
+    {
+      "epoch": 11.6,
+      "grad_norm": 6.276731967926025,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 0.3921,
+      "step": 2320
+    },
+    {
+      "epoch": 11.605,
+      "grad_norm": 4.362667560577393,
+      "learning_rate": 4.1975e-06,
+      "loss": 0.2815,
+      "step": 2321
+    },
+    {
+      "epoch": 11.61,
+      "grad_norm": 6.1637420654296875,
+      "learning_rate": 4.1950000000000005e-06,
+      "loss": 0.4413,
+      "step": 2322
+    },
+    {
+      "epoch": 11.615,
+      "grad_norm": 3.6981089115142822,
+      "learning_rate": 4.1925e-06,
+      "loss": 0.2253,
+      "step": 2323
+    },
+    {
+      "epoch": 11.62,
+      "grad_norm": 2.8664724826812744,
+      "learning_rate": 4.1900000000000005e-06,
+      "loss": 0.121,
+      "step": 2324
+    },
+    {
+      "epoch": 11.625,
+      "grad_norm": 5.081115245819092,
+      "learning_rate": 4.1875e-06,
+      "loss": 0.5237,
+      "step": 2325
+    },
+    {
+      "epoch": 11.63,
+      "grad_norm": 2.85467791557312,
+      "learning_rate": 4.185000000000001e-06,
+      "loss": 0.387,
+      "step": 2326
+    },
+    {
+      "epoch": 11.635,
+      "grad_norm": 4.326765537261963,
+      "learning_rate": 4.1825e-06,
+      "loss": 0.5064,
+      "step": 2327
+    },
+    {
+      "epoch": 11.64,
+      "grad_norm": 4.01115083694458,
+      "learning_rate": 4.18e-06,
+      "loss": 0.5102,
+      "step": 2328
+    },
+    {
+      "epoch": 11.645,
+      "grad_norm": 4.983065128326416,
+      "learning_rate": 4.1775e-06,
+      "loss": 0.6415,
+      "step": 2329
+    },
+    {
+      "epoch": 11.65,
+      "grad_norm": 9.806619644165039,
+      "learning_rate": 4.175e-06,
+      "loss": 0.4682,
+      "step": 2330
+    },
+    {
+      "epoch": 11.655,
+      "grad_norm": 7.778748035430908,
+      "learning_rate": 4.1725e-06,
+      "loss": 0.5898,
+      "step": 2331
+    },
+    {
+      "epoch": 11.66,
+      "grad_norm": 5.065155982971191,
+      "learning_rate": 4.17e-06,
+      "loss": 0.3331,
+      "step": 2332
+    },
+    {
+      "epoch": 11.665,
+      "grad_norm": 3.8532779216766357,
+      "learning_rate": 4.1675e-06,
+      "loss": 0.3018,
+      "step": 2333
+    },
+    {
+      "epoch": 11.67,
+      "grad_norm": 5.820866107940674,
+      "learning_rate": 4.165e-06,
+      "loss": 0.3254,
+      "step": 2334
+    },
+    {
+      "epoch": 11.675,
+      "grad_norm": 5.615697860717773,
+      "learning_rate": 4.1625e-06,
+      "loss": 0.3127,
+      "step": 2335
+    },
+    {
+      "epoch": 11.68,
+      "grad_norm": 3.552272319793701,
+      "learning_rate": 4.16e-06,
+      "loss": 0.1538,
+      "step": 2336
+    },
+    {
+      "epoch": 11.685,
+      "grad_norm": 3.405289649963379,
+      "learning_rate": 4.1575000000000004e-06,
+      "loss": 0.4219,
+      "step": 2337
+    },
+    {
+      "epoch": 11.69,
+      "grad_norm": 3.168076753616333,
+      "learning_rate": 4.155e-06,
+      "loss": 0.1854,
+      "step": 2338
+    },
+    {
+      "epoch": 11.695,
+      "grad_norm": 7.662829875946045,
+      "learning_rate": 4.1525000000000005e-06,
+      "loss": 0.6255,
+      "step": 2339
+    },
+    {
+      "epoch": 11.7,
+      "grad_norm": 7.669622421264648,
+      "learning_rate": 4.15e-06,
+      "loss": 0.6135,
+      "step": 2340
+    },
+    {
+      "epoch": 11.705,
+      "grad_norm": 4.416104316711426,
+      "learning_rate": 4.1475000000000005e-06,
+      "loss": 0.2886,
+      "step": 2341
+    },
+    {
+      "epoch": 11.71,
+      "grad_norm": 4.219148635864258,
+      "learning_rate": 4.145e-06,
+      "loss": 0.5741,
+      "step": 2342
+    },
+    {
+      "epoch": 11.715,
+      "grad_norm": 5.837990760803223,
+      "learning_rate": 4.1425000000000006e-06,
+      "loss": 0.3748,
+      "step": 2343
+    },
+    {
+      "epoch": 11.72,
+      "grad_norm": 3.1593775749206543,
+      "learning_rate": 4.14e-06,
+      "loss": 0.1696,
+      "step": 2344
+    },
+    {
+      "epoch": 11.725,
+      "grad_norm": 4.43656063079834,
+      "learning_rate": 4.137500000000001e-06,
+      "loss": 0.3201,
+      "step": 2345
+    },
+    {
+      "epoch": 11.73,
+      "grad_norm": 4.518340587615967,
+      "learning_rate": 4.135e-06,
+      "loss": 0.199,
+      "step": 2346
+    },
+    {
+      "epoch": 11.735,
+      "grad_norm": 1.5564370155334473,
+      "learning_rate": 4.132500000000001e-06,
+      "loss": 0.0997,
+      "step": 2347
+    },
+    {
+      "epoch": 11.74,
+      "grad_norm": 8.209750175476074,
+      "learning_rate": 4.13e-06,
+      "loss": 0.5912,
+      "step": 2348
+    },
+    {
+      "epoch": 11.745,
+      "grad_norm": 2.8856821060180664,
+      "learning_rate": 4.127500000000001e-06,
+      "loss": 0.0947,
+      "step": 2349
+    },
+    {
+      "epoch": 11.75,
+      "grad_norm": 12.122101783752441,
+      "learning_rate": 4.125e-06,
+      "loss": 0.67,
+      "step": 2350
+    },
+    {
+      "epoch": 11.755,
+      "grad_norm": 4.647777557373047,
+      "learning_rate": 4.122500000000001e-06,
+      "loss": 0.139,
+      "step": 2351
+    },
+    {
+      "epoch": 11.76,
+      "grad_norm": 3.5065090656280518,
+      "learning_rate": 4.12e-06,
+      "loss": 0.2812,
+      "step": 2352
+    },
+    {
+      "epoch": 11.765,
+      "grad_norm": 2.9833123683929443,
+      "learning_rate": 4.1175e-06,
+      "loss": 0.1064,
+      "step": 2353
+    },
+    {
+      "epoch": 11.77,
+      "grad_norm": 3.7749814987182617,
+      "learning_rate": 4.115e-06,
+      "loss": 0.3192,
+      "step": 2354
+    },
+    {
+      "epoch": 11.775,
+      "grad_norm": 7.490254878997803,
+      "learning_rate": 4.1125e-06,
+      "loss": 0.5117,
+      "step": 2355
+    },
+    {
+      "epoch": 11.78,
+      "grad_norm": 6.808143615722656,
+      "learning_rate": 4.1100000000000005e-06,
+      "loss": 0.6317,
+      "step": 2356
+    },
+    {
+      "epoch": 11.785,
+      "grad_norm": 7.931894302368164,
+      "learning_rate": 4.1075e-06,
+      "loss": 0.7912,
+      "step": 2357
+    },
+    {
+      "epoch": 11.79,
+      "grad_norm": 7.922944068908691,
+      "learning_rate": 4.1050000000000005e-06,
+      "loss": 0.4472,
+      "step": 2358
+    },
+    {
+      "epoch": 11.795,
+      "grad_norm": 5.047360897064209,
+      "learning_rate": 4.1025e-06,
+      "loss": 0.6414,
+      "step": 2359
+    },
+    {
+      "epoch": 11.8,
+      "grad_norm": 6.630589485168457,
+      "learning_rate": 4.1e-06,
+      "loss": 0.3234,
+      "step": 2360
+    },
+    {
+      "epoch": 11.805,
+      "grad_norm": 6.972137451171875,
+      "learning_rate": 4.0975e-06,
+      "loss": 0.2843,
+      "step": 2361
+    },
+    {
+      "epoch": 11.81,
+      "grad_norm": 10.212103843688965,
+      "learning_rate": 4.095e-06,
+      "loss": 1.0432,
+      "step": 2362
+    },
+    {
+      "epoch": 11.815,
+      "grad_norm": 3.390244483947754,
+      "learning_rate": 4.0925e-06,
+      "loss": 0.2189,
+      "step": 2363
+    },
+    {
+      "epoch": 11.82,
+      "grad_norm": 3.4476006031036377,
+      "learning_rate": 4.09e-06,
+      "loss": 0.428,
+      "step": 2364
+    },
+    {
+      "epoch": 11.825,
+      "grad_norm": 3.9801177978515625,
+      "learning_rate": 4.0875e-06,
+      "loss": 0.1476,
+      "step": 2365
+    },
+    {
+      "epoch": 11.83,
+      "grad_norm": 6.634037017822266,
+      "learning_rate": 4.085e-06,
+      "loss": 0.832,
+      "step": 2366
+    },
+    {
+      "epoch": 11.835,
+      "grad_norm": 124.75836181640625,
+      "learning_rate": 4.0825e-06,
+      "loss": 1.4329,
+      "step": 2367
+    },
+    {
+      "epoch": 11.84,
+      "grad_norm": 4.701694488525391,
+      "learning_rate": 4.08e-06,
+      "loss": 0.2547,
+      "step": 2368
+    },
+    {
+      "epoch": 11.845,
+      "grad_norm": 4.675734519958496,
+      "learning_rate": 4.0775e-06,
+      "loss": 0.4933,
+      "step": 2369
+    },
+    {
+      "epoch": 11.85,
+      "grad_norm": 988.65576171875,
+      "learning_rate": 4.075e-06,
+      "loss": 2.0567,
+      "step": 2370
+    },
+    {
+      "epoch": 11.855,
+      "grad_norm": 4.727938652038574,
+      "learning_rate": 4.0725e-06,
+      "loss": 0.4632,
+      "step": 2371
+    },
+    {
+      "epoch": 11.86,
+      "grad_norm": 2.282249927520752,
+      "learning_rate": 4.07e-06,
+      "loss": 0.0789,
+      "step": 2372
+    },
+    {
+      "epoch": 11.865,
+      "grad_norm": 2.9680278301239014,
+      "learning_rate": 4.0675000000000004e-06,
+      "loss": 0.3314,
+      "step": 2373
+    },
+    {
+      "epoch": 11.87,
+      "grad_norm": 3.3804068565368652,
+      "learning_rate": 4.065e-06,
+      "loss": 0.3634,
+      "step": 2374
+    },
+    {
+      "epoch": 11.875,
+      "grad_norm": 4.0782270431518555,
+      "learning_rate": 4.0625000000000005e-06,
+      "loss": 0.2237,
+      "step": 2375
+    },
+    {
+      "epoch": 11.88,
+      "grad_norm": 5.009937286376953,
+      "learning_rate": 4.060000000000001e-06,
+      "loss": 0.2431,
+      "step": 2376
+    },
+    {
+      "epoch": 11.885,
+      "grad_norm": 8.83266544342041,
+      "learning_rate": 4.0575000000000005e-06,
+      "loss": 0.5372,
+      "step": 2377
+    },
+    {
+      "epoch": 11.89,
+      "grad_norm": 4.7235918045043945,
+      "learning_rate": 4.055000000000001e-06,
+      "loss": 0.539,
+      "step": 2378
+    },
+    {
+      "epoch": 11.895,
+      "grad_norm": 2.0739548206329346,
+      "learning_rate": 4.052500000000001e-06,
+      "loss": 0.1065,
+      "step": 2379
+    },
+    {
+      "epoch": 11.9,
+      "grad_norm": 6.982193946838379,
+      "learning_rate": 4.05e-06,
+      "loss": 0.716,
+      "step": 2380
+    },
+    {
+      "epoch": 11.905,
+      "grad_norm": 0.9975789785385132,
+      "learning_rate": 4.047500000000001e-06,
+      "loss": 0.0201,
+      "step": 2381
+    },
+    {
+      "epoch": 11.91,
+      "grad_norm": 7.101102352142334,
+      "learning_rate": 4.045e-06,
+      "loss": 0.4531,
+      "step": 2382
+    },
+    {
+      "epoch": 11.915,
+      "grad_norm": 3.739859104156494,
+      "learning_rate": 4.042500000000001e-06,
+      "loss": 0.3249,
+      "step": 2383
+    },
+    {
+      "epoch": 11.92,
+      "grad_norm": 5.76193904876709,
+      "learning_rate": 4.04e-06,
+      "loss": 0.3118,
+      "step": 2384
+    },
+    {
+      "epoch": 11.925,
+      "grad_norm": 4.511212348937988,
+      "learning_rate": 4.037500000000001e-06,
+      "loss": 0.2506,
+      "step": 2385
+    },
+    {
+      "epoch": 11.93,
+      "grad_norm": 3.3247878551483154,
+      "learning_rate": 4.035e-06,
+      "loss": 0.1743,
+      "step": 2386
+    },
+    {
+      "epoch": 11.935,
+      "grad_norm": 2.817157506942749,
+      "learning_rate": 4.0325e-06,
+      "loss": 0.364,
+      "step": 2387
+    },
+    {
+      "epoch": 11.94,
+      "grad_norm": 4.632575035095215,
+      "learning_rate": 4.03e-06,
+      "loss": 0.1946,
+      "step": 2388
+    },
+    {
+      "epoch": 11.945,
+      "grad_norm": 4.842473030090332,
+      "learning_rate": 4.0275e-06,
+      "loss": 0.2771,
+      "step": 2389
+    },
+    {
+      "epoch": 11.95,
+      "grad_norm": 3.5947372913360596,
+      "learning_rate": 4.0250000000000004e-06,
+      "loss": 0.4142,
+      "step": 2390
+    },
+    {
+      "epoch": 11.955,
+      "grad_norm": 3.3107411861419678,
+      "learning_rate": 4.0225e-06,
+      "loss": 0.115,
+      "step": 2391
+    },
+    {
+      "epoch": 11.96,
+      "grad_norm": 5.029871940612793,
+      "learning_rate": 4.0200000000000005e-06,
+      "loss": 0.6418,
+      "step": 2392
+    },
+    {
+      "epoch": 11.965,
+      "grad_norm": 5.792545318603516,
+      "learning_rate": 4.0175e-06,
+      "loss": 0.4181,
+      "step": 2393
+    },
+    {
+      "epoch": 11.97,
+      "grad_norm": 3.9975132942199707,
+      "learning_rate": 4.0150000000000005e-06,
+      "loss": 0.2738,
+      "step": 2394
+    },
+    {
+      "epoch": 11.975,
+      "grad_norm": 10.398273468017578,
+      "learning_rate": 4.0125e-06,
+      "loss": 0.3484,
+      "step": 2395
+    },
+    {
+      "epoch": 11.98,
+      "grad_norm": 3.2295308113098145,
+      "learning_rate": 4.0100000000000006e-06,
+      "loss": 0.1396,
+      "step": 2396
+    },
+    {
+      "epoch": 11.985,
+      "grad_norm": 5.58898401260376,
+      "learning_rate": 4.0075e-06,
+      "loss": 0.5282,
+      "step": 2397
+    },
+    {
+      "epoch": 11.99,
+      "grad_norm": 7.6075849533081055,
+      "learning_rate": 4.005000000000001e-06,
+      "loss": 0.2245,
+      "step": 2398
+    },
+    {
+      "epoch": 11.995,
+      "grad_norm": 3.4022302627563477,
+      "learning_rate": 4.0025e-06,
+      "loss": 0.3693,
+      "step": 2399
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 3.8441624641418457,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.066,
+      "step": 2400
+    },
+    {
+      "epoch": 12.0,
+      "eval_loss": 0.2261846661567688,
+      "eval_runtime": 16.0445,
+      "eval_samples_per_second": 49.737,
+      "eval_steps_per_second": 6.233,
+      "step": 2400
+    },
+    {
+      "epoch": 12.005,
+      "grad_norm": 3.476515054702759,
+      "learning_rate": 3.9975e-06,
+      "loss": 0.2987,
+      "step": 2401
+    },
+    {
+      "epoch": 12.01,
+      "grad_norm": 7.534771919250488,
+      "learning_rate": 3.995000000000001e-06,
+      "loss": 0.7934,
+      "step": 2402
+    },
+    {
+      "epoch": 12.015,
+      "grad_norm": 3.664865493774414,
+      "learning_rate": 3.9925e-06,
+      "loss": 0.2385,
+      "step": 2403
+    },
+    {
+      "epoch": 12.02,
+      "grad_norm": 4.090984344482422,
+      "learning_rate": 3.990000000000001e-06,
+      "loss": 0.259,
+      "step": 2404
+    },
+    {
+      "epoch": 12.025,
+      "grad_norm": 4.56601619720459,
+      "learning_rate": 3.9875e-06,
+      "loss": 0.5211,
+      "step": 2405
+    },
+    {
+      "epoch": 12.03,
+      "grad_norm": 3.3471474647521973,
+      "learning_rate": 3.985000000000001e-06,
+      "loss": 0.1791,
+      "step": 2406
+    },
+    {
+      "epoch": 12.035,
+      "grad_norm": 3.619697332382202,
+      "learning_rate": 3.9825e-06,
+      "loss": 0.1079,
+      "step": 2407
+    },
+    {
+      "epoch": 12.04,
+      "grad_norm": 4.274447917938232,
+      "learning_rate": 3.980000000000001e-06,
+      "loss": 0.6252,
+      "step": 2408
+    },
+    {
+      "epoch": 12.045,
+      "grad_norm": 5.825752258300781,
+      "learning_rate": 3.9775000000000005e-06,
+      "loss": 0.5227,
+      "step": 2409
+    },
+    {
+      "epoch": 12.05,
+      "grad_norm": 3.639138698577881,
+      "learning_rate": 3.975000000000001e-06,
+      "loss": 0.2573,
+      "step": 2410
+    },
+    {
+      "epoch": 12.055,
+      "grad_norm": 3.6322240829467773,
+      "learning_rate": 3.9725000000000005e-06,
+      "loss": 0.3168,
+      "step": 2411
+    },
+    {
+      "epoch": 12.06,
+      "grad_norm": 4.3578691482543945,
+      "learning_rate": 3.97e-06,
+      "loss": 0.2125,
+      "step": 2412
+    },
+    {
+      "epoch": 12.065,
+      "grad_norm": 7.293939590454102,
+      "learning_rate": 3.9675000000000006e-06,
+      "loss": 0.54,
+      "step": 2413
+    },
+    {
+      "epoch": 12.07,
+      "grad_norm": 5.767150402069092,
+      "learning_rate": 3.965e-06,
+      "loss": 0.2183,
+      "step": 2414
+    },
+    {
+      "epoch": 12.075,
+      "grad_norm": 3.1661341190338135,
+      "learning_rate": 3.962500000000001e-06,
+      "loss": 0.1528,
+      "step": 2415
+    },
+    {
+      "epoch": 12.08,
+      "grad_norm": 3.2233824729919434,
+      "learning_rate": 3.96e-06,
+      "loss": 0.3552,
+      "step": 2416
+    },
+    {
+      "epoch": 12.085,
+      "grad_norm": 2.9981627464294434,
+      "learning_rate": 3.957500000000001e-06,
+      "loss": 0.2014,
+      "step": 2417
+    },
+    {
+      "epoch": 12.09,
+      "grad_norm": 3.3674001693725586,
+      "learning_rate": 3.955e-06,
+      "loss": 0.1503,
+      "step": 2418
+    },
+    {
+      "epoch": 12.095,
+      "grad_norm": 1.7284117937088013,
+      "learning_rate": 3.9525e-06,
+      "loss": 0.0702,
+      "step": 2419
+    },
+    {
+      "epoch": 12.1,
+      "grad_norm": 3.4924397468566895,
+      "learning_rate": 3.95e-06,
+      "loss": 0.4496,
+      "step": 2420
+    },
+    {
+      "epoch": 12.105,
+      "grad_norm": 4.650517463684082,
+      "learning_rate": 3.9475e-06,
+      "loss": 0.4587,
+      "step": 2421
+    },
+    {
+      "epoch": 12.11,
+      "grad_norm": 5.932701110839844,
+      "learning_rate": 3.945e-06,
+      "loss": 0.4906,
+      "step": 2422
+    },
+    {
+      "epoch": 12.115,
+      "grad_norm": 6.421353340148926,
+      "learning_rate": 3.9425e-06,
+      "loss": 0.5921,
+      "step": 2423
+    },
+    {
+      "epoch": 12.12,
+      "grad_norm": 11.128769874572754,
+      "learning_rate": 3.94e-06,
+      "loss": 0.424,
+      "step": 2424
+    },
+    {
+      "epoch": 12.125,
+      "grad_norm": 2.6439945697784424,
+      "learning_rate": 3.9375e-06,
+      "loss": 0.061,
+      "step": 2425
+    },
+    {
+      "epoch": 12.13,
+      "grad_norm": 6.735964298248291,
+      "learning_rate": 3.9350000000000004e-06,
+      "loss": 0.3435,
+      "step": 2426
+    },
+    {
+      "epoch": 12.135,
+      "grad_norm": 3.7680532932281494,
+      "learning_rate": 3.9325e-06,
+      "loss": 0.607,
+      "step": 2427
+    },
+    {
+      "epoch": 12.14,
+      "grad_norm": 4.679652690887451,
+      "learning_rate": 3.9300000000000005e-06,
+      "loss": 0.321,
+      "step": 2428
+    },
+    {
+      "epoch": 12.145,
+      "grad_norm": 4.177921295166016,
+      "learning_rate": 3.9275e-06,
+      "loss": 0.1731,
+      "step": 2429
+    },
+    {
+      "epoch": 12.15,
+      "grad_norm": 5.049252510070801,
+      "learning_rate": 3.9250000000000005e-06,
+      "loss": 0.5351,
+      "step": 2430
+    },
+    {
+      "epoch": 12.155,
+      "grad_norm": 6.223732948303223,
+      "learning_rate": 3.9225e-06,
+      "loss": 0.4077,
+      "step": 2431
+    },
+    {
+      "epoch": 12.16,
+      "grad_norm": 5.338696002960205,
+      "learning_rate": 3.920000000000001e-06,
+      "loss": 0.1613,
+      "step": 2432
+    },
+    {
+      "epoch": 12.165,
+      "grad_norm": 3.6500754356384277,
+      "learning_rate": 3.9175e-06,
+      "loss": 0.446,
+      "step": 2433
+    },
+    {
+      "epoch": 12.17,
+      "grad_norm": 10.379395484924316,
+      "learning_rate": 3.915000000000001e-06,
+      "loss": 0.9746,
+      "step": 2434
+    },
+    {
+      "epoch": 12.175,
+      "grad_norm": 3.70862078666687,
+      "learning_rate": 3.9125e-06,
+      "loss": 0.2113,
+      "step": 2435
+    },
+    {
+      "epoch": 12.18,
+      "grad_norm": 4.9443769454956055,
+      "learning_rate": 3.910000000000001e-06,
+      "loss": 0.3474,
+      "step": 2436
+    },
+    {
+      "epoch": 12.185,
+      "grad_norm": 6.082023620605469,
+      "learning_rate": 3.9075e-06,
+      "loss": 0.9065,
+      "step": 2437
+    },
+    {
+      "epoch": 12.19,
+      "grad_norm": 4.623384952545166,
+      "learning_rate": 3.905000000000001e-06,
+      "loss": 0.3219,
+      "step": 2438
+    },
+    {
+      "epoch": 12.195,
+      "grad_norm": 3.3660264015197754,
+      "learning_rate": 3.9025e-06,
+      "loss": 0.1574,
+      "step": 2439
+    },
+    {
+      "epoch": 12.2,
+      "grad_norm": 4.600410461425781,
+      "learning_rate": 3.900000000000001e-06,
+      "loss": 0.3026,
+      "step": 2440
+    },
+    {
+      "epoch": 12.205,
+      "grad_norm": 2.531733989715576,
+      "learning_rate": 3.8975e-06,
+      "loss": 0.0592,
+      "step": 2441
+    },
+    {
+      "epoch": 12.21,
+      "grad_norm": 3.099790334701538,
+      "learning_rate": 3.895000000000001e-06,
+      "loss": 0.1417,
+      "step": 2442
+    },
+    {
+      "epoch": 12.215,
+      "grad_norm": 5.944413185119629,
+      "learning_rate": 3.8925000000000004e-06,
+      "loss": 0.5555,
+      "step": 2443
+    },
+    {
+      "epoch": 12.22,
+      "grad_norm": 3.4926860332489014,
+      "learning_rate": 3.89e-06,
+      "loss": 0.1776,
+      "step": 2444
+    },
+    {
+      "epoch": 12.225,
+      "grad_norm": 4.8463616371154785,
+      "learning_rate": 3.8875000000000005e-06,
+      "loss": 0.5837,
+      "step": 2445
+    },
+    {
+      "epoch": 12.23,
+      "grad_norm": 4.29119348526001,
+      "learning_rate": 3.885e-06,
+      "loss": 0.4596,
+      "step": 2446
+    },
+    {
+      "epoch": 12.235,
+      "grad_norm": 5.39113187789917,
+      "learning_rate": 3.8825000000000005e-06,
+      "loss": 0.4406,
+      "step": 2447
+    },
+    {
+      "epoch": 12.24,
+      "grad_norm": 10.20838737487793,
+      "learning_rate": 3.88e-06,
+      "loss": 0.4395,
+      "step": 2448
+    },
+    {
+      "epoch": 12.245,
+      "grad_norm": 5.101496696472168,
+      "learning_rate": 3.8775000000000006e-06,
+      "loss": 0.3936,
+      "step": 2449
+    },
+    {
+      "epoch": 12.25,
+      "grad_norm": 3.2225875854492188,
+      "learning_rate": 3.875e-06,
+      "loss": 0.1083,
+      "step": 2450
+    },
+    {
+      "epoch": 12.255,
+      "grad_norm": 8.815666198730469,
+      "learning_rate": 3.8725e-06,
+      "loss": 0.6817,
+      "step": 2451
+    },
+    {
+      "epoch": 12.26,
+      "grad_norm": 9.043973922729492,
+      "learning_rate": 3.87e-06,
+      "loss": 0.1447,
+      "step": 2452
+    },
+    {
+      "epoch": 12.265,
+      "grad_norm": 7.027723789215088,
+      "learning_rate": 3.8675e-06,
+      "loss": 0.5909,
+      "step": 2453
+    },
+    {
+      "epoch": 12.27,
+      "grad_norm": 5.624725341796875,
+      "learning_rate": 3.865e-06,
+      "loss": 0.3614,
+      "step": 2454
+    },
+    {
+      "epoch": 12.275,
+      "grad_norm": 4.797810077667236,
+      "learning_rate": 3.8625e-06,
+      "loss": 0.5267,
+      "step": 2455
+    },
+    {
+      "epoch": 12.28,
+      "grad_norm": 3.8676393032073975,
+      "learning_rate": 3.86e-06,
+      "loss": 0.2007,
+      "step": 2456
+    },
+    {
+      "epoch": 12.285,
+      "grad_norm": 5.423714637756348,
+      "learning_rate": 3.8575e-06,
+      "loss": 0.2102,
+      "step": 2457
+    },
+    {
+      "epoch": 12.29,
+      "grad_norm": 3.3590707778930664,
+      "learning_rate": 3.855e-06,
+      "loss": 0.3087,
+      "step": 2458
+    },
+    {
+      "epoch": 12.295,
+      "grad_norm": 5.719491481781006,
+      "learning_rate": 3.8525e-06,
+      "loss": 0.8735,
+      "step": 2459
+    },
+    {
+      "epoch": 12.3,
+      "grad_norm": 3.96392822265625,
+      "learning_rate": 3.85e-06,
+      "loss": 0.434,
+      "step": 2460
+    },
+    {
+      "epoch": 12.305,
+      "grad_norm": 3.24631404876709,
+      "learning_rate": 3.8475e-06,
+      "loss": 0.201,
+      "step": 2461
+    },
+    {
+      "epoch": 12.31,
+      "grad_norm": 4.7879252433776855,
+      "learning_rate": 3.8450000000000005e-06,
+      "loss": 0.3489,
+      "step": 2462
+    },
+    {
+      "epoch": 12.315,
+      "grad_norm": 6.1969146728515625,
+      "learning_rate": 3.8425e-06,
+      "loss": 0.4964,
+      "step": 2463
+    },
+    {
+      "epoch": 12.32,
+      "grad_norm": 2.95414400100708,
+      "learning_rate": 3.8400000000000005e-06,
+      "loss": 0.0989,
+      "step": 2464
+    },
+    {
+      "epoch": 12.325,
+      "grad_norm": 10.95142936706543,
+      "learning_rate": 3.8375e-06,
+      "loss": 0.3144,
+      "step": 2465
+    },
+    {
+      "epoch": 12.33,
+      "grad_norm": 9.397826194763184,
+      "learning_rate": 3.8350000000000006e-06,
+      "loss": 0.411,
+      "step": 2466
+    },
+    {
+      "epoch": 12.335,
+      "grad_norm": 3.312406063079834,
+      "learning_rate": 3.8325e-06,
+      "loss": 0.1632,
+      "step": 2467
+    },
+    {
+      "epoch": 12.34,
+      "grad_norm": 3.240449905395508,
+      "learning_rate": 3.830000000000001e-06,
+      "loss": 0.2192,
+      "step": 2468
+    },
+    {
+      "epoch": 12.345,
+      "grad_norm": 6.7621870040893555,
+      "learning_rate": 3.8275e-06,
+      "loss": 0.4029,
+      "step": 2469
+    },
+    {
+      "epoch": 12.35,
+      "grad_norm": 3.9045205116271973,
+      "learning_rate": 3.825000000000001e-06,
+      "loss": 0.6516,
+      "step": 2470
+    },
+    {
+      "epoch": 12.355,
+      "grad_norm": 4.117351055145264,
+      "learning_rate": 3.8225e-06,
+      "loss": 0.1773,
+      "step": 2471
+    },
+    {
+      "epoch": 12.36,
+      "grad_norm": 1.4342386722564697,
+      "learning_rate": 3.820000000000001e-06,
+      "loss": 0.0503,
+      "step": 2472
+    },
+    {
+      "epoch": 12.365,
+      "grad_norm": 3.9351234436035156,
+      "learning_rate": 3.8175e-06,
+      "loss": 0.2277,
+      "step": 2473
+    },
+    {
+      "epoch": 12.37,
+      "grad_norm": 9.061785697937012,
+      "learning_rate": 3.815000000000001e-06,
+      "loss": 0.7707,
+      "step": 2474
+    },
+    {
+      "epoch": 12.375,
+      "grad_norm": 5.2592926025390625,
+      "learning_rate": 3.8125e-06,
+      "loss": 0.6174,
+      "step": 2475
+    },
+    {
+      "epoch": 12.38,
+      "grad_norm": 3.1703150272369385,
+      "learning_rate": 3.8100000000000004e-06,
+      "loss": 0.2098,
+      "step": 2476
+    },
+    {
+      "epoch": 12.385,
+      "grad_norm": 3.7314414978027344,
+      "learning_rate": 3.8075e-06,
+      "loss": 0.2471,
+      "step": 2477
+    },
+    {
+      "epoch": 12.39,
+      "grad_norm": 2.6589150428771973,
+      "learning_rate": 3.8050000000000004e-06,
+      "loss": 0.0498,
+      "step": 2478
+    },
+    {
+      "epoch": 12.395,
+      "grad_norm": 1.779199242591858,
+      "learning_rate": 3.8025e-06,
+      "loss": 0.0717,
+      "step": 2479
+    },
+    {
+      "epoch": 12.4,
+      "grad_norm": 2.85843563079834,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 0.1994,
+      "step": 2480
+    },
+    {
+      "epoch": 12.405,
+      "grad_norm": 5.831678867340088,
+      "learning_rate": 3.7975e-06,
+      "loss": 0.5922,
+      "step": 2481
+    },
+    {
+      "epoch": 12.41,
+      "grad_norm": 5.710839748382568,
+      "learning_rate": 3.7950000000000005e-06,
+      "loss": 0.1702,
+      "step": 2482
+    },
+    {
+      "epoch": 12.415,
+      "grad_norm": 7.364334583282471,
+      "learning_rate": 3.7925e-06,
+      "loss": 0.4517,
+      "step": 2483
+    },
+    {
+      "epoch": 12.42,
+      "grad_norm": 5.067105293273926,
+      "learning_rate": 3.79e-06,
+      "loss": 0.5045,
+      "step": 2484
+    },
+    {
+      "epoch": 12.425,
+      "grad_norm": 3.1936073303222656,
+      "learning_rate": 3.7875e-06,
+      "loss": 0.4,
+      "step": 2485
+    },
+    {
+      "epoch": 12.43,
+      "grad_norm": 5.957182884216309,
+      "learning_rate": 3.785e-06,
+      "loss": 0.2806,
+      "step": 2486
+    },
+    {
+      "epoch": 12.435,
+      "grad_norm": 4.410153865814209,
+      "learning_rate": 3.7825e-06,
+      "loss": 0.3068,
+      "step": 2487
+    },
+    {
+      "epoch": 12.44,
+      "grad_norm": 3.0283303260803223,
+      "learning_rate": 3.7800000000000002e-06,
+      "loss": 0.2134,
+      "step": 2488
+    },
+    {
+      "epoch": 12.445,
+      "grad_norm": 4.918923854827881,
+      "learning_rate": 3.7775000000000003e-06,
+      "loss": 0.5969,
+      "step": 2489
+    },
+    {
+      "epoch": 12.45,
+      "grad_norm": 4.513157367706299,
+      "learning_rate": 3.7750000000000003e-06,
+      "loss": 0.22,
+      "step": 2490
+    },
+    {
+      "epoch": 12.455,
+      "grad_norm": 8.156756401062012,
+      "learning_rate": 3.7725e-06,
+      "loss": 0.6502,
+      "step": 2491
+    },
+    {
+      "epoch": 12.46,
+      "grad_norm": 10.012717247009277,
+      "learning_rate": 3.7700000000000003e-06,
+      "loss": 1.0648,
+      "step": 2492
+    },
+    {
+      "epoch": 12.465,
+      "grad_norm": 19.343761444091797,
+      "learning_rate": 3.7675e-06,
+      "loss": 0.2983,
+      "step": 2493
+    },
+    {
+      "epoch": 12.47,
+      "grad_norm": 4.603684902191162,
+      "learning_rate": 3.7650000000000004e-06,
+      "loss": 0.6964,
+      "step": 2494
+    },
+    {
+      "epoch": 12.475,
+      "grad_norm": 5.471191883087158,
+      "learning_rate": 3.7625e-06,
+      "loss": 0.4097,
+      "step": 2495
+    },
+    {
+      "epoch": 12.48,
+      "grad_norm": 3.5714986324310303,
+      "learning_rate": 3.7600000000000004e-06,
+      "loss": 0.5263,
+      "step": 2496
+    },
+    {
+      "epoch": 12.485,
+      "grad_norm": 2.9785377979278564,
+      "learning_rate": 3.7575e-06,
+      "loss": 0.1158,
+      "step": 2497
+    },
+    {
+      "epoch": 12.49,
+      "grad_norm": 22.77909278869629,
+      "learning_rate": 3.7550000000000005e-06,
+      "loss": 0.2494,
+      "step": 2498
+    },
+    {
+      "epoch": 12.495,
+      "grad_norm": 5.2543416023254395,
+      "learning_rate": 3.7525e-06,
+      "loss": 0.1212,
+      "step": 2499
+    },
+    {
+      "epoch": 12.5,
+      "grad_norm": 3.82944393157959,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.1635,
+      "step": 2500
+    },
+    {
+      "epoch": 12.505,
+      "grad_norm": 7.59887170791626,
+      "learning_rate": 3.7475000000000005e-06,
+      "loss": 0.5132,
+      "step": 2501
+    },
+    {
+      "epoch": 12.51,
+      "grad_norm": 7.780153274536133,
+      "learning_rate": 3.745e-06,
+      "loss": 0.5893,
+      "step": 2502
+    },
+    {
+      "epoch": 12.515,
+      "grad_norm": 3.37327241897583,
+      "learning_rate": 3.7425000000000006e-06,
+      "loss": 0.2688,
+      "step": 2503
+    },
+    {
+      "epoch": 12.52,
+      "grad_norm": 4.723874568939209,
+      "learning_rate": 3.74e-06,
+      "loss": 0.3386,
+      "step": 2504
+    },
+    {
+      "epoch": 12.525,
+      "grad_norm": 3.9527359008789062,
+      "learning_rate": 3.7375000000000006e-06,
+      "loss": 0.4114,
+      "step": 2505
+    },
+    {
+      "epoch": 12.53,
+      "grad_norm": 2.327986240386963,
+      "learning_rate": 3.7350000000000002e-06,
+      "loss": 0.1646,
+      "step": 2506
+    },
+    {
+      "epoch": 12.535,
+      "grad_norm": 4.764712333679199,
+      "learning_rate": 3.7325000000000007e-06,
+      "loss": 0.5161,
+      "step": 2507
+    },
+    {
+      "epoch": 12.54,
+      "grad_norm": 8.827037811279297,
+      "learning_rate": 3.7300000000000003e-06,
+      "loss": 0.2569,
+      "step": 2508
+    },
+    {
+      "epoch": 12.545,
+      "grad_norm": 4.572065830230713,
+      "learning_rate": 3.7275000000000007e-06,
+      "loss": 0.7418,
+      "step": 2509
+    },
+    {
+      "epoch": 12.55,
+      "grad_norm": 5.875421524047852,
+      "learning_rate": 3.7250000000000003e-06,
+      "loss": 0.5464,
+      "step": 2510
+    },
+    {
+      "epoch": 12.555,
+      "grad_norm": 5.372376918792725,
+      "learning_rate": 3.7225000000000004e-06,
+      "loss": 0.2595,
+      "step": 2511
+    },
+    {
+      "epoch": 12.56,
+      "grad_norm": 10.581452369689941,
+      "learning_rate": 3.7200000000000004e-06,
+      "loss": 0.6486,
+      "step": 2512
+    },
+    {
+      "epoch": 12.565,
+      "grad_norm": 8.040359497070312,
+      "learning_rate": 3.7175000000000004e-06,
+      "loss": 0.5642,
+      "step": 2513
+    },
+    {
+      "epoch": 12.57,
+      "grad_norm": 3.138930320739746,
+      "learning_rate": 3.7150000000000004e-06,
+      "loss": 0.1173,
+      "step": 2514
+    },
+    {
+      "epoch": 12.575,
+      "grad_norm": 3.50618052482605,
+      "learning_rate": 3.7125000000000005e-06,
+      "loss": 0.2739,
+      "step": 2515
+    },
+    {
+      "epoch": 12.58,
+      "grad_norm": 6.6320576667785645,
+      "learning_rate": 3.7100000000000005e-06,
+      "loss": 0.3711,
+      "step": 2516
+    },
+    {
+      "epoch": 12.585,
+      "grad_norm": 2.5286452770233154,
+      "learning_rate": 3.7075000000000005e-06,
+      "loss": 0.2342,
+      "step": 2517
+    },
+    {
+      "epoch": 12.59,
+      "grad_norm": 3.745680332183838,
+      "learning_rate": 3.705e-06,
+      "loss": 0.319,
+      "step": 2518
+    },
+    {
+      "epoch": 12.595,
+      "grad_norm": 3.4215164184570312,
+      "learning_rate": 3.7025000000000005e-06,
+      "loss": 0.3613,
+      "step": 2519
+    },
+    {
+      "epoch": 12.6,
+      "grad_norm": 3.7860350608825684,
+      "learning_rate": 3.7e-06,
+      "loss": 0.3275,
+      "step": 2520
+    },
+    {
+      "epoch": 12.605,
+      "grad_norm": 6.23767614364624,
+      "learning_rate": 3.6975000000000006e-06,
+      "loss": 0.3215,
+      "step": 2521
+    },
+    {
+      "epoch": 12.61,
+      "grad_norm": 4.677371501922607,
+      "learning_rate": 3.695e-06,
+      "loss": 0.3266,
+      "step": 2522
+    },
+    {
+      "epoch": 12.615,
+      "grad_norm": 2.850188970565796,
+      "learning_rate": 3.6925000000000006e-06,
+      "loss": 0.2084,
+      "step": 2523
+    },
+    {
+      "epoch": 12.62,
+      "grad_norm": 6.311230182647705,
+      "learning_rate": 3.6900000000000002e-06,
+      "loss": 0.6191,
+      "step": 2524
+    },
+    {
+      "epoch": 12.625,
+      "grad_norm": 2.8153622150421143,
+      "learning_rate": 3.6875000000000007e-06,
+      "loss": 0.1679,
+      "step": 2525
+    },
+    {
+      "epoch": 12.63,
+      "grad_norm": 4.119739055633545,
+      "learning_rate": 3.6850000000000003e-06,
+      "loss": 0.1696,
+      "step": 2526
+    },
+    {
+      "epoch": 12.635,
+      "grad_norm": 5.033112049102783,
+      "learning_rate": 3.6825000000000003e-06,
+      "loss": 0.3354,
+      "step": 2527
+    },
+    {
+      "epoch": 12.64,
+      "grad_norm": 4.327576160430908,
+      "learning_rate": 3.6800000000000003e-06,
+      "loss": 0.1829,
+      "step": 2528
+    },
+    {
+      "epoch": 12.645,
+      "grad_norm": 4.768941402435303,
+      "learning_rate": 3.6775000000000004e-06,
+      "loss": 0.384,
+      "step": 2529
+    },
+    {
+      "epoch": 12.65,
+      "grad_norm": 4.385491371154785,
+      "learning_rate": 3.6750000000000004e-06,
+      "loss": 0.3351,
+      "step": 2530
+    },
+    {
+      "epoch": 12.655,
+      "grad_norm": 2.7953200340270996,
+      "learning_rate": 3.6725000000000004e-06,
+      "loss": 0.4392,
+      "step": 2531
+    },
+    {
+      "epoch": 12.66,
+      "grad_norm": 3.8814423084259033,
+      "learning_rate": 3.6700000000000004e-06,
+      "loss": 0.1696,
+      "step": 2532
+    },
+    {
+      "epoch": 12.665,
+      "grad_norm": 7.4518723487854,
+      "learning_rate": 3.6675000000000005e-06,
+      "loss": 0.5572,
+      "step": 2533
+    },
+    {
+      "epoch": 12.67,
+      "grad_norm": 5.325657367706299,
+      "learning_rate": 3.665e-06,
+      "loss": 0.4932,
+      "step": 2534
+    },
+    {
+      "epoch": 12.675,
+      "grad_norm": 2.6530144214630127,
+      "learning_rate": 3.6625000000000005e-06,
+      "loss": 0.2295,
+      "step": 2535
+    },
+    {
+      "epoch": 12.68,
+      "grad_norm": 3.9954841136932373,
+      "learning_rate": 3.66e-06,
+      "loss": 0.2671,
+      "step": 2536
+    },
+    {
+      "epoch": 12.685,
+      "grad_norm": 2.9491353034973145,
+      "learning_rate": 3.6575000000000006e-06,
+      "loss": 0.0859,
+      "step": 2537
+    },
+    {
+      "epoch": 12.69,
+      "grad_norm": 15.393409729003906,
+      "learning_rate": 3.655e-06,
+      "loss": 0.6401,
+      "step": 2538
+    },
+    {
+      "epoch": 12.695,
+      "grad_norm": 5.529659271240234,
+      "learning_rate": 3.6525000000000006e-06,
+      "loss": 0.4166,
+      "step": 2539
+    },
+    {
+      "epoch": 12.7,
+      "grad_norm": 10.201495170593262,
+      "learning_rate": 3.65e-06,
+      "loss": 0.2146,
+      "step": 2540
+    },
+    {
+      "epoch": 12.705,
+      "grad_norm": 5.9331135749816895,
+      "learning_rate": 3.6475000000000007e-06,
+      "loss": 0.3684,
+      "step": 2541
+    },
+    {
+      "epoch": 12.71,
+      "grad_norm": 6.820456027984619,
+      "learning_rate": 3.6450000000000003e-06,
+      "loss": 0.4726,
+      "step": 2542
+    },
+    {
+      "epoch": 12.715,
+      "grad_norm": 7.3526530265808105,
+      "learning_rate": 3.6425000000000003e-06,
+      "loss": 0.7496,
+      "step": 2543
+    },
+    {
+      "epoch": 12.72,
+      "grad_norm": 3.7125093936920166,
+      "learning_rate": 3.6400000000000003e-06,
+      "loss": 0.2651,
+      "step": 2544
+    },
+    {
+      "epoch": 12.725,
+      "grad_norm": 6.252248287200928,
+      "learning_rate": 3.6375000000000003e-06,
+      "loss": 0.426,
+      "step": 2545
+    },
+    {
+      "epoch": 12.73,
+      "grad_norm": 3.7096357345581055,
+      "learning_rate": 3.6350000000000003e-06,
+      "loss": 0.3226,
+      "step": 2546
+    },
+    {
+      "epoch": 12.735,
+      "grad_norm": 8.98271369934082,
+      "learning_rate": 3.6325000000000004e-06,
+      "loss": 0.7302,
+      "step": 2547
+    },
+    {
+      "epoch": 12.74,
+      "grad_norm": 5.488955020904541,
+      "learning_rate": 3.6300000000000004e-06,
+      "loss": 0.3132,
+      "step": 2548
+    },
+    {
+      "epoch": 12.745,
+      "grad_norm": 4.299445152282715,
+      "learning_rate": 3.6275000000000004e-06,
+      "loss": 0.2426,
+      "step": 2549
+    },
+    {
+      "epoch": 12.75,
+      "grad_norm": 4.000606060028076,
+      "learning_rate": 3.625e-06,
+      "loss": 0.4208,
+      "step": 2550
+    },
+    {
+      "epoch": 12.755,
+      "grad_norm": 4.710710525512695,
+      "learning_rate": 3.6225000000000005e-06,
+      "loss": 0.3986,
+      "step": 2551
+    },
+    {
+      "epoch": 12.76,
+      "grad_norm": 4.096498489379883,
+      "learning_rate": 3.62e-06,
+      "loss": 0.3169,
+      "step": 2552
+    },
+    {
+      "epoch": 12.765,
+      "grad_norm": 6.277150630950928,
+      "learning_rate": 3.6175000000000005e-06,
+      "loss": 0.4489,
+      "step": 2553
+    },
+    {
+      "epoch": 12.77,
+      "grad_norm": 5.711903095245361,
+      "learning_rate": 3.615e-06,
+      "loss": 0.0838,
+      "step": 2554
+    },
+    {
+      "epoch": 12.775,
+      "grad_norm": 3.5505242347717285,
+      "learning_rate": 3.6125000000000006e-06,
+      "loss": 0.1921,
+      "step": 2555
+    },
+    {
+      "epoch": 12.78,
+      "grad_norm": 5.033021926879883,
+      "learning_rate": 3.61e-06,
+      "loss": 0.3731,
+      "step": 2556
+    },
+    {
+      "epoch": 12.785,
+      "grad_norm": 4.8283257484436035,
+      "learning_rate": 3.6075000000000006e-06,
+      "loss": 0.2323,
+      "step": 2557
+    },
+    {
+      "epoch": 12.79,
+      "grad_norm": 2.5534188747406006,
+      "learning_rate": 3.6050000000000002e-06,
+      "loss": 0.1428,
+      "step": 2558
+    },
+    {
+      "epoch": 12.795,
+      "grad_norm": 3.5825068950653076,
+      "learning_rate": 3.6025000000000002e-06,
+      "loss": 0.2501,
+      "step": 2559
+    },
+    {
+      "epoch": 12.8,
+      "grad_norm": 4.186033248901367,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 0.1759,
+      "step": 2560
+    },
+    {
+      "epoch": 12.805,
+      "grad_norm": 4.2932963371276855,
+      "learning_rate": 3.5975000000000003e-06,
+      "loss": 0.2715,
+      "step": 2561
+    },
+    {
+      "epoch": 12.81,
+      "grad_norm": 4.045590877532959,
+      "learning_rate": 3.5950000000000003e-06,
+      "loss": 0.2091,
+      "step": 2562
+    },
+    {
+      "epoch": 12.815,
+      "grad_norm": 5.730029582977295,
+      "learning_rate": 3.5925000000000003e-06,
+      "loss": 0.387,
+      "step": 2563
+    },
+    {
+      "epoch": 12.82,
+      "grad_norm": 3.460638999938965,
+      "learning_rate": 3.5900000000000004e-06,
+      "loss": 0.2362,
+      "step": 2564
+    },
+    {
+      "epoch": 12.825,
+      "grad_norm": 5.7042036056518555,
+      "learning_rate": 3.5875000000000004e-06,
+      "loss": 0.4479,
+      "step": 2565
+    },
+    {
+      "epoch": 12.83,
+      "grad_norm": 3.2396886348724365,
+      "learning_rate": 3.585e-06,
+      "loss": 0.0587,
+      "step": 2566
+    },
+    {
+      "epoch": 12.835,
+      "grad_norm": 5.003210067749023,
+      "learning_rate": 3.5825000000000004e-06,
+      "loss": 0.3961,
+      "step": 2567
+    },
+    {
+      "epoch": 12.84,
+      "grad_norm": 5.7366943359375,
+      "learning_rate": 3.58e-06,
+      "loss": 0.6725,
+      "step": 2568
+    },
+    {
+      "epoch": 12.845,
+      "grad_norm": 162.47035217285156,
+      "learning_rate": 3.5775000000000005e-06,
+      "loss": 0.9766,
+      "step": 2569
+    },
+    {
+      "epoch": 12.85,
+      "grad_norm": 3.1198208332061768,
+      "learning_rate": 3.575e-06,
+      "loss": 0.3196,
+      "step": 2570
+    },
+    {
+      "epoch": 12.855,
+      "grad_norm": 2.306438446044922,
+      "learning_rate": 3.5725000000000005e-06,
+      "loss": 0.1058,
+      "step": 2571
+    },
+    {
+      "epoch": 12.86,
+      "grad_norm": 4.77396297454834,
+      "learning_rate": 3.57e-06,
+      "loss": 0.2888,
+      "step": 2572
+    },
+    {
+      "epoch": 12.865,
+      "grad_norm": 4.047329902648926,
+      "learning_rate": 3.5675000000000006e-06,
+      "loss": 0.2642,
+      "step": 2573
+    },
+    {
+      "epoch": 12.87,
+      "grad_norm": 5.200380802154541,
+      "learning_rate": 3.565e-06,
+      "loss": 0.521,
+      "step": 2574
+    },
+    {
+      "epoch": 12.875,
+      "grad_norm": 2.8099706172943115,
+      "learning_rate": 3.5625e-06,
+      "loss": 0.264,
+      "step": 2575
+    },
+    {
+      "epoch": 12.88,
+      "grad_norm": 6.882323265075684,
+      "learning_rate": 3.5600000000000002e-06,
+      "loss": 0.5242,
+      "step": 2576
+    },
+    {
+      "epoch": 12.885,
+      "grad_norm": 5.92142915725708,
+      "learning_rate": 3.5575000000000002e-06,
+      "loss": 0.3966,
+      "step": 2577
+    },
+    {
+      "epoch": 12.89,
+      "grad_norm": 5.33410120010376,
+      "learning_rate": 3.5550000000000003e-06,
+      "loss": 0.2859,
+      "step": 2578
+    },
+    {
+      "epoch": 12.895,
+      "grad_norm": 4.7160420417785645,
+      "learning_rate": 3.5525000000000003e-06,
+      "loss": 0.2797,
+      "step": 2579
+    },
+    {
+      "epoch": 12.9,
+      "grad_norm": 3.8806099891662598,
+      "learning_rate": 3.5500000000000003e-06,
+      "loss": 0.393,
+      "step": 2580
+    },
+    {
+      "epoch": 12.905,
+      "grad_norm": 6.762237548828125,
+      "learning_rate": 3.5475000000000003e-06,
+      "loss": 0.4598,
+      "step": 2581
+    },
+    {
+      "epoch": 12.91,
+      "grad_norm": 3.8168752193450928,
+      "learning_rate": 3.545e-06,
+      "loss": 0.1068,
+      "step": 2582
+    },
+    {
+      "epoch": 12.915,
+      "grad_norm": 38.69218063354492,
+      "learning_rate": 3.5425000000000004e-06,
+      "loss": 0.252,
+      "step": 2583
+    },
+    {
+      "epoch": 12.92,
+      "grad_norm": 5.468113422393799,
+      "learning_rate": 3.54e-06,
+      "loss": 0.2204,
+      "step": 2584
+    },
+    {
+      "epoch": 12.925,
+      "grad_norm": 8.902981758117676,
+      "learning_rate": 3.5375000000000004e-06,
+      "loss": 0.3627,
+      "step": 2585
+    },
+    {
+      "epoch": 12.93,
+      "grad_norm": 4.087343215942383,
+      "learning_rate": 3.535e-06,
+      "loss": 0.4641,
+      "step": 2586
+    },
+    {
+      "epoch": 12.935,
+      "grad_norm": 5.373355865478516,
+      "learning_rate": 3.5325000000000005e-06,
+      "loss": 0.5639,
+      "step": 2587
+    },
+    {
+      "epoch": 12.94,
+      "grad_norm": 5.7207255363464355,
+      "learning_rate": 3.53e-06,
+      "loss": 0.5561,
+      "step": 2588
+    },
+    {
+      "epoch": 12.945,
+      "grad_norm": 4.294384956359863,
+      "learning_rate": 3.5275000000000005e-06,
+      "loss": 0.3339,
+      "step": 2589
+    },
+    {
+      "epoch": 12.95,
+      "grad_norm": 3.391033172607422,
+      "learning_rate": 3.525e-06,
+      "loss": 0.1331,
+      "step": 2590
+    },
+    {
+      "epoch": 12.955,
+      "grad_norm": 4.324766635894775,
+      "learning_rate": 3.5225e-06,
+      "loss": 0.2522,
+      "step": 2591
+    },
+    {
+      "epoch": 12.96,
+      "grad_norm": 5.999277591705322,
+      "learning_rate": 3.52e-06,
+      "loss": 0.6045,
+      "step": 2592
+    },
+    {
+      "epoch": 12.965,
+      "grad_norm": 3.9514262676239014,
+      "learning_rate": 3.5175e-06,
+      "loss": 0.2044,
+      "step": 2593
+    },
+    {
+      "epoch": 12.97,
+      "grad_norm": 3.5547804832458496,
+      "learning_rate": 3.5150000000000002e-06,
+      "loss": 0.2915,
+      "step": 2594
+    },
+    {
+      "epoch": 12.975,
+      "grad_norm": 4.776151657104492,
+      "learning_rate": 3.5125000000000003e-06,
+      "loss": 0.2718,
+      "step": 2595
+    },
+    {
+      "epoch": 12.98,
+      "grad_norm": 8.932281494140625,
+      "learning_rate": 3.5100000000000003e-06,
+      "loss": 0.6357,
+      "step": 2596
+    },
+    {
+      "epoch": 12.985,
+      "grad_norm": 4.184885501861572,
+      "learning_rate": 3.5075000000000003e-06,
+      "loss": 0.2163,
+      "step": 2597
+    },
+    {
+      "epoch": 12.99,
+      "grad_norm": 3.6633529663085938,
+      "learning_rate": 3.505e-06,
+      "loss": 0.3669,
+      "step": 2598
+    },
+    {
+      "epoch": 12.995,
+      "grad_norm": 6.877824306488037,
+      "learning_rate": 3.5025000000000003e-06,
+      "loss": 0.5,
+      "step": 2599
+    },
+    {
+      "epoch": 13.0,
+      "grad_norm": 1.1057254076004028,
+      "learning_rate": 3.5e-06,
+      "loss": 0.0222,
+      "step": 2600
+    },
+    {
+      "epoch": 13.0,
+      "eval_loss": 0.21740847826004028,
+      "eval_runtime": 15.6583,
+      "eval_samples_per_second": 50.963,
+      "eval_steps_per_second": 6.386,
+      "step": 2600
+    },
+    {
+      "epoch": 13.005,
+      "grad_norm": 2.3351714611053467,
+      "learning_rate": 3.4975000000000004e-06,
+      "loss": 0.1493,
+      "step": 2601
+    },
+    {
+      "epoch": 13.01,
+      "grad_norm": 3.5202529430389404,
+      "learning_rate": 3.495e-06,
+      "loss": 0.1877,
+      "step": 2602
+    },
+    {
+      "epoch": 13.015,
+      "grad_norm": 5.538907051086426,
+      "learning_rate": 3.4925000000000004e-06,
+      "loss": 0.4134,
+      "step": 2603
+    },
+    {
+      "epoch": 13.02,
+      "grad_norm": 3.633805274963379,
+      "learning_rate": 3.49e-06,
+      "loss": 0.1489,
+      "step": 2604
+    },
+    {
+      "epoch": 13.025,
+      "grad_norm": 3.295497179031372,
+      "learning_rate": 3.4875000000000005e-06,
+      "loss": 0.1584,
+      "step": 2605
+    },
+    {
+      "epoch": 13.03,
+      "grad_norm": 10.646966934204102,
+      "learning_rate": 3.485e-06,
+      "loss": 0.7161,
+      "step": 2606
+    },
+    {
+      "epoch": 13.035,
+      "grad_norm": 6.279256343841553,
+      "learning_rate": 3.4825e-06,
+      "loss": 0.593,
+      "step": 2607
+    },
+    {
+      "epoch": 13.04,
+      "grad_norm": 4.029870986938477,
+      "learning_rate": 3.48e-06,
+      "loss": 0.225,
+      "step": 2608
+    },
+    {
+      "epoch": 13.045,
+      "grad_norm": 4.961399555206299,
+      "learning_rate": 3.4775e-06,
+      "loss": 0.434,
+      "step": 2609
+    },
+    {
+      "epoch": 13.05,
+      "grad_norm": 3.076810359954834,
+      "learning_rate": 3.475e-06,
+      "loss": 0.277,
+      "step": 2610
+    },
+    {
+      "epoch": 13.055,
+      "grad_norm": 4.897925853729248,
+      "learning_rate": 3.4725e-06,
+      "loss": 0.2791,
+      "step": 2611
+    },
+    {
+      "epoch": 13.06,
+      "grad_norm": 5.143693447113037,
+      "learning_rate": 3.4700000000000002e-06,
+      "loss": 0.4468,
+      "step": 2612
+    },
+    {
+      "epoch": 13.065,
+      "grad_norm": 4.237023830413818,
+      "learning_rate": 3.4675000000000003e-06,
+      "loss": 0.1314,
+      "step": 2613
+    },
+    {
+      "epoch": 13.07,
+      "grad_norm": 2.99906325340271,
+      "learning_rate": 3.465e-06,
+      "loss": 0.0638,
+      "step": 2614
+    },
+    {
+      "epoch": 13.075,
+      "grad_norm": 3.9287221431732178,
+      "learning_rate": 3.4625000000000003e-06,
+      "loss": 0.3068,
+      "step": 2615
+    },
+    {
+      "epoch": 13.08,
+      "grad_norm": 3.5814008712768555,
+      "learning_rate": 3.46e-06,
+      "loss": 0.2295,
+      "step": 2616
+    },
+    {
+      "epoch": 13.085,
+      "grad_norm": 7.304368495941162,
+      "learning_rate": 3.4575000000000004e-06,
+      "loss": 0.7139,
+      "step": 2617
+    },
+    {
+      "epoch": 13.09,
+      "grad_norm": 6.295146465301514,
+      "learning_rate": 3.455e-06,
+      "loss": 0.6818,
+      "step": 2618
+    },
+    {
+      "epoch": 13.095,
+      "grad_norm": 7.6342010498046875,
+      "learning_rate": 3.4525000000000004e-06,
+      "loss": 0.6644,
+      "step": 2619
+    },
+    {
+      "epoch": 13.1,
+      "grad_norm": 3.2951509952545166,
+      "learning_rate": 3.45e-06,
+      "loss": 0.1323,
+      "step": 2620
+    },
+    {
+      "epoch": 13.105,
+      "grad_norm": 6.0835957527160645,
+      "learning_rate": 3.4475000000000005e-06,
+      "loss": 0.3493,
+      "step": 2621
+    },
+    {
+      "epoch": 13.11,
+      "grad_norm": 7.716336727142334,
+      "learning_rate": 3.445e-06,
+      "loss": 0.4253,
+      "step": 2622
+    },
+    {
+      "epoch": 13.115,
+      "grad_norm": 4.401556015014648,
+      "learning_rate": 3.4425e-06,
+      "loss": 0.3979,
+      "step": 2623
+    },
+    {
+      "epoch": 13.12,
+      "grad_norm": 4.6217122077941895,
+      "learning_rate": 3.44e-06,
+      "loss": 0.3785,
+      "step": 2624
+    },
+    {
+      "epoch": 13.125,
+      "grad_norm": 4.029857158660889,
+      "learning_rate": 3.4375e-06,
+      "loss": 0.4355,
+      "step": 2625
+    },
+    {
+      "epoch": 13.13,
+      "grad_norm": 5.796063423156738,
+      "learning_rate": 3.4350000000000006e-06,
+      "loss": 0.3565,
+      "step": 2626
+    },
+    {
+      "epoch": 13.135,
+      "grad_norm": 6.488113880157471,
+      "learning_rate": 3.4325e-06,
+      "loss": 0.5645,
+      "step": 2627
+    },
+    {
+      "epoch": 13.14,
+      "grad_norm": 4.610037326812744,
+      "learning_rate": 3.4300000000000006e-06,
+      "loss": 0.3409,
+      "step": 2628
+    },
+    {
+      "epoch": 13.145,
+      "grad_norm": 2.7086803913116455,
+      "learning_rate": 3.4275000000000002e-06,
+      "loss": 0.1833,
+      "step": 2629
+    },
+    {
+      "epoch": 13.15,
+      "grad_norm": 4.708641529083252,
+      "learning_rate": 3.4250000000000007e-06,
+      "loss": 0.3197,
+      "step": 2630
+    },
+    {
+      "epoch": 13.155,
+      "grad_norm": 3.0031087398529053,
+      "learning_rate": 3.4225000000000003e-06,
+      "loss": 0.095,
+      "step": 2631
+    },
+    {
+      "epoch": 13.16,
+      "grad_norm": 5.291317939758301,
+      "learning_rate": 3.4200000000000007e-06,
+      "loss": 0.7254,
+      "step": 2632
+    },
+    {
+      "epoch": 13.165,
+      "grad_norm": 6.509015083312988,
+      "learning_rate": 3.4175000000000003e-06,
+      "loss": 0.3953,
+      "step": 2633
+    },
+    {
+      "epoch": 13.17,
+      "grad_norm": 2.136404275894165,
+      "learning_rate": 3.4150000000000003e-06,
+      "loss": 0.0843,
+      "step": 2634
+    },
+    {
+      "epoch": 13.175,
+      "grad_norm": 3.625977039337158,
+      "learning_rate": 3.4125000000000004e-06,
+      "loss": 0.2383,
+      "step": 2635
+    },
+    {
+      "epoch": 13.18,
+      "grad_norm": 5.4440598487854,
+      "learning_rate": 3.4100000000000004e-06,
+      "loss": 0.5989,
+      "step": 2636
+    },
+    {
+      "epoch": 13.185,
+      "grad_norm": 2.613140344619751,
+      "learning_rate": 3.4075000000000004e-06,
+      "loss": 0.0852,
+      "step": 2637
+    },
+    {
+      "epoch": 13.19,
+      "grad_norm": 3.5298361778259277,
+      "learning_rate": 3.4050000000000004e-06,
+      "loss": 0.3615,
+      "step": 2638
+    },
+    {
+      "epoch": 13.195,
+      "grad_norm": 4.0747857093811035,
+      "learning_rate": 3.4025000000000005e-06,
+      "loss": 0.1837,
+      "step": 2639
+    },
+    {
+      "epoch": 13.2,
+      "grad_norm": 3.9930715560913086,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 0.4293,
+      "step": 2640
+    },
+    {
+      "epoch": 13.205,
+      "grad_norm": 3.3911590576171875,
+      "learning_rate": 3.3975e-06,
+      "loss": 0.2859,
+      "step": 2641
+    },
+    {
+      "epoch": 13.21,
+      "grad_norm": 2.9776017665863037,
+      "learning_rate": 3.3950000000000005e-06,
+      "loss": 0.1561,
+      "step": 2642
+    },
+    {
+      "epoch": 13.215,
+      "grad_norm": 3.8941338062286377,
+      "learning_rate": 3.3925e-06,
+      "loss": 0.2506,
+      "step": 2643
+    },
+    {
+      "epoch": 13.22,
+      "grad_norm": 3.079866409301758,
+      "learning_rate": 3.3900000000000006e-06,
+      "loss": 0.135,
+      "step": 2644
+    },
+    {
+      "epoch": 13.225,
+      "grad_norm": 4.343544006347656,
+      "learning_rate": 3.3875e-06,
+      "loss": 0.1294,
+      "step": 2645
+    },
+    {
+      "epoch": 13.23,
+      "grad_norm": 7.8805389404296875,
+      "learning_rate": 3.3850000000000006e-06,
+      "loss": 0.8191,
+      "step": 2646
+    },
+    {
+      "epoch": 13.235,
+      "grad_norm": 6.149213790893555,
+      "learning_rate": 3.3825000000000002e-06,
+      "loss": 0.3755,
+      "step": 2647
+    },
+    {
+      "epoch": 13.24,
+      "grad_norm": 3.914381504058838,
+      "learning_rate": 3.3800000000000007e-06,
+      "loss": 0.4796,
+      "step": 2648
+    },
+    {
+      "epoch": 13.245,
+      "grad_norm": 2.36649751663208,
+      "learning_rate": 3.3775000000000003e-06,
+      "loss": 0.1726,
+      "step": 2649
+    },
+    {
+      "epoch": 13.25,
+      "grad_norm": 25.46281623840332,
+      "learning_rate": 3.3750000000000003e-06,
+      "loss": 0.6316,
+      "step": 2650
+    },
+    {
+      "epoch": 13.255,
+      "grad_norm": 2.4677343368530273,
+      "learning_rate": 3.3725000000000003e-06,
+      "loss": 0.0931,
+      "step": 2651
+    },
+    {
+      "epoch": 13.26,
+      "grad_norm": 3.1744086742401123,
+      "learning_rate": 3.3700000000000003e-06,
+      "loss": 0.2285,
+      "step": 2652
+    },
+    {
+      "epoch": 13.265,
+      "grad_norm": 2.9443745613098145,
+      "learning_rate": 3.3675000000000004e-06,
+      "loss": 0.2106,
+      "step": 2653
+    },
+    {
+      "epoch": 13.27,
+      "grad_norm": 5.407256603240967,
+      "learning_rate": 3.3650000000000004e-06,
+      "loss": 0.2259,
+      "step": 2654
+    },
+    {
+      "epoch": 13.275,
+      "grad_norm": 4.463200569152832,
+      "learning_rate": 3.3625000000000004e-06,
+      "loss": 0.3075,
+      "step": 2655
+    },
+    {
+      "epoch": 13.28,
+      "grad_norm": 4.433103084564209,
+      "learning_rate": 3.3600000000000004e-06,
+      "loss": 0.557,
+      "step": 2656
+    },
+    {
+      "epoch": 13.285,
+      "grad_norm": 4.375494480133057,
+      "learning_rate": 3.3575e-06,
+      "loss": 0.2941,
+      "step": 2657
+    },
+    {
+      "epoch": 13.29,
+      "grad_norm": 6.884410858154297,
+      "learning_rate": 3.3550000000000005e-06,
+      "loss": 0.4521,
+      "step": 2658
+    },
+    {
+      "epoch": 13.295,
+      "grad_norm": 4.55253267288208,
+      "learning_rate": 3.3525e-06,
+      "loss": 0.3371,
+      "step": 2659
+    },
+    {
+      "epoch": 13.3,
+      "grad_norm": 2.1417627334594727,
+      "learning_rate": 3.3500000000000005e-06,
+      "loss": 0.2245,
+      "step": 2660
+    },
+    {
+      "epoch": 13.305,
+      "grad_norm": 3.1329843997955322,
+      "learning_rate": 3.3475e-06,
+      "loss": 0.177,
+      "step": 2661
+    },
+    {
+      "epoch": 13.31,
+      "grad_norm": 2.486212730407715,
+      "learning_rate": 3.3450000000000006e-06,
+      "loss": 0.0672,
+      "step": 2662
+    },
+    {
+      "epoch": 13.315,
+      "grad_norm": 3.357534646987915,
+      "learning_rate": 3.3425e-06,
+      "loss": 0.1391,
+      "step": 2663
+    },
+    {
+      "epoch": 13.32,
+      "grad_norm": 5.549729347229004,
+      "learning_rate": 3.3400000000000006e-06,
+      "loss": 0.2412,
+      "step": 2664
+    },
+    {
+      "epoch": 13.325,
+      "grad_norm": 3.9570395946502686,
+      "learning_rate": 3.3375000000000002e-06,
+      "loss": 0.4546,
+      "step": 2665
+    },
+    {
+      "epoch": 13.33,
+      "grad_norm": 12.097053527832031,
+      "learning_rate": 3.3350000000000003e-06,
+      "loss": 0.6734,
+      "step": 2666
+    },
+    {
+      "epoch": 13.335,
+      "grad_norm": 4.281240940093994,
+      "learning_rate": 3.3325000000000003e-06,
+      "loss": 0.2267,
+      "step": 2667
+    },
+    {
+      "epoch": 13.34,
+      "grad_norm": 4.326976776123047,
+      "learning_rate": 3.3300000000000003e-06,
+      "loss": 0.512,
+      "step": 2668
+    },
+    {
+      "epoch": 13.345,
+      "grad_norm": 3.510136842727661,
+      "learning_rate": 3.3275000000000003e-06,
+      "loss": 0.3161,
+      "step": 2669
+    },
+    {
+      "epoch": 13.35,
+      "grad_norm": 3.8528075218200684,
+      "learning_rate": 3.3250000000000004e-06,
+      "loss": 0.489,
+      "step": 2670
+    },
+    {
+      "epoch": 13.355,
+      "grad_norm": 5.670360088348389,
+      "learning_rate": 3.3225000000000004e-06,
+      "loss": 0.3437,
+      "step": 2671
+    },
+    {
+      "epoch": 13.36,
+      "grad_norm": 5.210115909576416,
+      "learning_rate": 3.3200000000000004e-06,
+      "loss": 0.9602,
+      "step": 2672
+    },
+    {
+      "epoch": 13.365,
+      "grad_norm": 3.446441650390625,
+      "learning_rate": 3.3175e-06,
+      "loss": 0.3176,
+      "step": 2673
+    },
+    {
+      "epoch": 13.37,
+      "grad_norm": 6.4163103103637695,
+      "learning_rate": 3.3150000000000004e-06,
+      "loss": 0.5011,
+      "step": 2674
+    },
+    {
+      "epoch": 13.375,
+      "grad_norm": 4.014662265777588,
+      "learning_rate": 3.3125e-06,
+      "loss": 0.2722,
+      "step": 2675
+    },
+    {
+      "epoch": 13.38,
+      "grad_norm": 4.364996433258057,
+      "learning_rate": 3.3100000000000005e-06,
+      "loss": 0.5726,
+      "step": 2676
+    },
+    {
+      "epoch": 13.385,
+      "grad_norm": 5.254627227783203,
+      "learning_rate": 3.3075e-06,
+      "loss": 0.3202,
+      "step": 2677
+    },
+    {
+      "epoch": 13.39,
+      "grad_norm": 8.804628372192383,
+      "learning_rate": 3.3050000000000005e-06,
+      "loss": 0.4869,
+      "step": 2678
+    },
+    {
+      "epoch": 13.395,
+      "grad_norm": 3.4403374195098877,
+      "learning_rate": 3.3025e-06,
+      "loss": 0.3378,
+      "step": 2679
+    },
+    {
+      "epoch": 13.4,
+      "grad_norm": 4.856618881225586,
+      "learning_rate": 3.3000000000000006e-06,
+      "loss": 0.335,
+      "step": 2680
+    },
+    {
+      "epoch": 13.405,
+      "grad_norm": 2.4180331230163574,
+      "learning_rate": 3.2975e-06,
+      "loss": 0.0957,
+      "step": 2681
+    },
+    {
+      "epoch": 13.41,
+      "grad_norm": 8.02914810180664,
+      "learning_rate": 3.2950000000000002e-06,
+      "loss": 0.468,
+      "step": 2682
+    },
+    {
+      "epoch": 13.415,
+      "grad_norm": 4.505415916442871,
+      "learning_rate": 3.2925000000000002e-06,
+      "loss": 0.3048,
+      "step": 2683
+    },
+    {
+      "epoch": 13.42,
+      "grad_norm": 17.221214294433594,
+      "learning_rate": 3.2900000000000003e-06,
+      "loss": 0.3663,
+      "step": 2684
+    },
+    {
+      "epoch": 13.425,
+      "grad_norm": 3.744819402694702,
+      "learning_rate": 3.2875000000000003e-06,
+      "loss": 0.1829,
+      "step": 2685
+    },
+    {
+      "epoch": 13.43,
+      "grad_norm": 4.308400630950928,
+      "learning_rate": 3.2850000000000003e-06,
+      "loss": 0.2517,
+      "step": 2686
+    },
+    {
+      "epoch": 13.435,
+      "grad_norm": 4.122700214385986,
+      "learning_rate": 3.2825000000000003e-06,
+      "loss": 0.2471,
+      "step": 2687
+    },
+    {
+      "epoch": 13.44,
+      "grad_norm": 6.057083606719971,
+      "learning_rate": 3.2800000000000004e-06,
+      "loss": 0.9794,
+      "step": 2688
+    },
+    {
+      "epoch": 13.445,
+      "grad_norm": 2.1911020278930664,
+      "learning_rate": 3.2775e-06,
+      "loss": 0.0527,
+      "step": 2689
+    },
+    {
+      "epoch": 13.45,
+      "grad_norm": 3.445913791656494,
+      "learning_rate": 3.2750000000000004e-06,
+      "loss": 0.3848,
+      "step": 2690
+    },
+    {
+      "epoch": 13.455,
+      "grad_norm": 3.394446611404419,
+      "learning_rate": 3.2725e-06,
+      "loss": 0.2351,
+      "step": 2691
+    },
+    {
+      "epoch": 13.46,
+      "grad_norm": 3.297323226928711,
+      "learning_rate": 3.2700000000000005e-06,
+      "loss": 0.253,
+      "step": 2692
+    },
+    {
+      "epoch": 13.465,
+      "grad_norm": 3.544461488723755,
+      "learning_rate": 3.2675e-06,
+      "loss": 0.2229,
+      "step": 2693
+    },
+    {
+      "epoch": 13.47,
+      "grad_norm": 2.157289743423462,
+      "learning_rate": 3.2650000000000005e-06,
+      "loss": 0.1106,
+      "step": 2694
+    },
+    {
+      "epoch": 13.475,
+      "grad_norm": 5.264073371887207,
+      "learning_rate": 3.2625e-06,
+      "loss": 0.4629,
+      "step": 2695
+    },
+    {
+      "epoch": 13.48,
+      "grad_norm": 7.2001566886901855,
+      "learning_rate": 3.2600000000000006e-06,
+      "loss": 0.4739,
+      "step": 2696
+    },
+    {
+      "epoch": 13.485,
+      "grad_norm": 3.1903605461120605,
+      "learning_rate": 3.2575e-06,
+      "loss": 0.1738,
+      "step": 2697
+    },
+    {
+      "epoch": 13.49,
+      "grad_norm": 3.355541467666626,
+      "learning_rate": 3.255e-06,
+      "loss": 0.2102,
+      "step": 2698
+    },
+    {
+      "epoch": 13.495,
+      "grad_norm": 4.697750091552734,
+      "learning_rate": 3.2525e-06,
+      "loss": 0.2821,
+      "step": 2699
+    },
+    {
+      "epoch": 13.5,
+      "grad_norm": 3.5834336280822754,
+      "learning_rate": 3.2500000000000002e-06,
+      "loss": 0.3121,
+      "step": 2700
+    },
+    {
+      "epoch": 13.505,
+      "grad_norm": 4.7487473487854,
+      "learning_rate": 3.2475000000000002e-06,
+      "loss": 0.3343,
+      "step": 2701
+    },
+    {
+      "epoch": 13.51,
+      "grad_norm": 5.141551971435547,
+      "learning_rate": 3.2450000000000003e-06,
+      "loss": 0.6302,
+      "step": 2702
+    },
+    {
+      "epoch": 13.515,
+      "grad_norm": 5.297160625457764,
+      "learning_rate": 3.2425000000000003e-06,
+      "loss": 1.006,
+      "step": 2703
+    },
+    {
+      "epoch": 13.52,
+      "grad_norm": 6.223616600036621,
+      "learning_rate": 3.2400000000000003e-06,
+      "loss": 0.4215,
+      "step": 2704
+    },
+    {
+      "epoch": 13.525,
+      "grad_norm": 10.025032997131348,
+      "learning_rate": 3.2375e-06,
+      "loss": 0.607,
+      "step": 2705
+    },
+    {
+      "epoch": 13.53,
+      "grad_norm": 5.424351692199707,
+      "learning_rate": 3.2350000000000004e-06,
+      "loss": 0.7088,
+      "step": 2706
+    },
+    {
+      "epoch": 13.535,
+      "grad_norm": 2.998936891555786,
+      "learning_rate": 3.2325e-06,
+      "loss": 0.0906,
+      "step": 2707
+    },
+    {
+      "epoch": 13.54,
+      "grad_norm": 2.6324141025543213,
+      "learning_rate": 3.2300000000000004e-06,
+      "loss": 0.188,
+      "step": 2708
+    },
+    {
+      "epoch": 13.545,
+      "grad_norm": 3.835195541381836,
+      "learning_rate": 3.2275e-06,
+      "loss": 0.1268,
+      "step": 2709
+    },
+    {
+      "epoch": 13.55,
+      "grad_norm": 4.012763500213623,
+      "learning_rate": 3.2250000000000005e-06,
+      "loss": 0.3734,
+      "step": 2710
+    },
+    {
+      "epoch": 13.555,
+      "grad_norm": 4.8824663162231445,
+      "learning_rate": 3.2225e-06,
+      "loss": 0.6108,
+      "step": 2711
+    },
+    {
+      "epoch": 13.56,
+      "grad_norm": 3.9833850860595703,
+      "learning_rate": 3.2200000000000005e-06,
+      "loss": 0.3133,
+      "step": 2712
+    },
+    {
+      "epoch": 13.565,
+      "grad_norm": 14.29844856262207,
+      "learning_rate": 3.2175e-06,
+      "loss": 0.6903,
+      "step": 2713
+    },
+    {
+      "epoch": 13.57,
+      "grad_norm": 4.837995529174805,
+      "learning_rate": 3.215e-06,
+      "loss": 0.1886,
+      "step": 2714
+    },
+    {
+      "epoch": 13.575,
+      "grad_norm": 5.560044288635254,
+      "learning_rate": 3.2125e-06,
+      "loss": 0.4169,
+      "step": 2715
+    },
+    {
+      "epoch": 13.58,
+      "grad_norm": 7.9728546142578125,
+      "learning_rate": 3.21e-06,
+      "loss": 0.4585,
+      "step": 2716
+    },
+    {
+      "epoch": 13.585,
+      "grad_norm": 4.939574718475342,
+      "learning_rate": 3.2075e-06,
+      "loss": 0.4967,
+      "step": 2717
+    },
+    {
+      "epoch": 13.59,
+      "grad_norm": 4.826523303985596,
+      "learning_rate": 3.2050000000000002e-06,
+      "loss": 0.2065,
+      "step": 2718
+    },
+    {
+      "epoch": 13.595,
+      "grad_norm": 4.9271650314331055,
+      "learning_rate": 3.2025000000000003e-06,
+      "loss": 0.8053,
+      "step": 2719
+    },
+    {
+      "epoch": 13.6,
+      "grad_norm": 4.4050469398498535,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 0.2896,
+      "step": 2720
+    },
+    {
+      "epoch": 13.605,
+      "grad_norm": 6.031970500946045,
+      "learning_rate": 3.1975e-06,
+      "loss": 0.225,
+      "step": 2721
+    },
+    {
+      "epoch": 13.61,
+      "grad_norm": 5.65727424621582,
+      "learning_rate": 3.1950000000000003e-06,
+      "loss": 0.367,
+      "step": 2722
+    },
+    {
+      "epoch": 13.615,
+      "grad_norm": 4.936824798583984,
+      "learning_rate": 3.1925e-06,
+      "loss": 0.3763,
+      "step": 2723
+    },
+    {
+      "epoch": 13.62,
+      "grad_norm": 9.859992027282715,
+      "learning_rate": 3.1900000000000004e-06,
+      "loss": 0.7882,
+      "step": 2724
+    },
+    {
+      "epoch": 13.625,
+      "grad_norm": 2.5502331256866455,
+      "learning_rate": 3.1875e-06,
+      "loss": 0.2031,
+      "step": 2725
+    },
+    {
+      "epoch": 13.63,
+      "grad_norm": 3.961629629135132,
+      "learning_rate": 3.1850000000000004e-06,
+      "loss": 0.1961,
+      "step": 2726
+    },
+    {
+      "epoch": 13.635,
+      "grad_norm": 3.3742082118988037,
+      "learning_rate": 3.1825e-06,
+      "loss": 0.2249,
+      "step": 2727
+    },
+    {
+      "epoch": 13.64,
+      "grad_norm": 5.1020379066467285,
+      "learning_rate": 3.1800000000000005e-06,
+      "loss": 0.4829,
+      "step": 2728
+    },
+    {
+      "epoch": 13.645,
+      "grad_norm": 2.7706644535064697,
+      "learning_rate": 3.1775e-06,
+      "loss": 0.1671,
+      "step": 2729
+    },
+    {
+      "epoch": 13.65,
+      "grad_norm": 5.066681385040283,
+      "learning_rate": 3.175e-06,
+      "loss": 0.3439,
+      "step": 2730
+    },
+    {
+      "epoch": 13.655,
+      "grad_norm": 3.722487211227417,
+      "learning_rate": 3.1725e-06,
+      "loss": 0.3525,
+      "step": 2731
+    },
+    {
+      "epoch": 13.66,
+      "grad_norm": 5.672784805297852,
+      "learning_rate": 3.17e-06,
+      "loss": 0.3151,
+      "step": 2732
+    },
+    {
+      "epoch": 13.665,
+      "grad_norm": 2.16601300239563,
+      "learning_rate": 3.1675e-06,
+      "loss": 0.0606,
+      "step": 2733
+    },
+    {
+      "epoch": 13.67,
+      "grad_norm": 4.874353408813477,
+      "learning_rate": 3.165e-06,
+      "loss": 0.2521,
+      "step": 2734
+    },
+    {
+      "epoch": 13.675,
+      "grad_norm": 4.165945053100586,
+      "learning_rate": 3.1625000000000002e-06,
+      "loss": 0.1999,
+      "step": 2735
+    },
+    {
+      "epoch": 13.68,
+      "grad_norm": 5.493929386138916,
+      "learning_rate": 3.1600000000000002e-06,
+      "loss": 0.4019,
+      "step": 2736
+    },
+    {
+      "epoch": 13.685,
+      "grad_norm": 4.465583324432373,
+      "learning_rate": 3.1575e-06,
+      "loss": 0.0962,
+      "step": 2737
+    },
+    {
+      "epoch": 13.69,
+      "grad_norm": 3.037649154663086,
+      "learning_rate": 3.1550000000000003e-06,
+      "loss": 0.1129,
+      "step": 2738
+    },
+    {
+      "epoch": 13.695,
+      "grad_norm": 5.382497787475586,
+      "learning_rate": 3.1525e-06,
+      "loss": 0.3361,
+      "step": 2739
+    },
+    {
+      "epoch": 13.7,
+      "grad_norm": 2.6108553409576416,
+      "learning_rate": 3.1500000000000003e-06,
+      "loss": 0.0601,
+      "step": 2740
+    },
+    {
+      "epoch": 13.705,
+      "grad_norm": 114.65782165527344,
+      "learning_rate": 3.1475e-06,
+      "loss": 0.3491,
+      "step": 2741
+    },
+    {
+      "epoch": 13.71,
+      "grad_norm": 3.5241737365722656,
+      "learning_rate": 3.1450000000000004e-06,
+      "loss": 0.4148,
+      "step": 2742
+    },
+    {
+      "epoch": 13.715,
+      "grad_norm": 3.2684850692749023,
+      "learning_rate": 3.1425e-06,
+      "loss": 0.2641,
+      "step": 2743
+    },
+    {
+      "epoch": 13.72,
+      "grad_norm": 3.0700430870056152,
+      "learning_rate": 3.1400000000000004e-06,
+      "loss": 0.2269,
+      "step": 2744
+    },
+    {
+      "epoch": 13.725,
+      "grad_norm": 7.483908176422119,
+      "learning_rate": 3.1375e-06,
+      "loss": 0.3685,
+      "step": 2745
+    },
+    {
+      "epoch": 13.73,
+      "grad_norm": 5.249300479888916,
+      "learning_rate": 3.135e-06,
+      "loss": 0.4574,
+      "step": 2746
+    },
+    {
+      "epoch": 13.735,
+      "grad_norm": 3.5041451454162598,
+      "learning_rate": 3.1325e-06,
+      "loss": 0.3638,
+      "step": 2747
+    },
+    {
+      "epoch": 13.74,
+      "grad_norm": 8.526005744934082,
+      "learning_rate": 3.13e-06,
+      "loss": 0.5912,
+      "step": 2748
+    },
+    {
+      "epoch": 13.745,
+      "grad_norm": 9.01889419555664,
+      "learning_rate": 3.1275e-06,
+      "loss": 0.3528,
+      "step": 2749
+    },
+    {
+      "epoch": 13.75,
+      "grad_norm": 4.308925151824951,
+      "learning_rate": 3.125e-06,
+      "loss": 0.2215,
+      "step": 2750
+    },
+    {
+      "epoch": 13.755,
+      "grad_norm": 3.677945613861084,
+      "learning_rate": 3.1225000000000006e-06,
+      "loss": 0.1789,
+      "step": 2751
+    },
+    {
+      "epoch": 13.76,
+      "grad_norm": 3.600541591644287,
+      "learning_rate": 3.12e-06,
+      "loss": 0.3828,
+      "step": 2752
+    },
+    {
+      "epoch": 13.765,
+      "grad_norm": 2.628596067428589,
+      "learning_rate": 3.1175000000000006e-06,
+      "loss": 0.1123,
+      "step": 2753
+    },
+    {
+      "epoch": 13.77,
+      "grad_norm": 5.657008171081543,
+      "learning_rate": 3.1150000000000002e-06,
+      "loss": 0.4438,
+      "step": 2754
+    },
+    {
+      "epoch": 13.775,
+      "grad_norm": 3.393906831741333,
+      "learning_rate": 3.1125000000000007e-06,
+      "loss": 0.1791,
+      "step": 2755
+    },
+    {
+      "epoch": 13.78,
+      "grad_norm": 3.5643527507781982,
+      "learning_rate": 3.1100000000000003e-06,
+      "loss": 0.2569,
+      "step": 2756
+    },
+    {
+      "epoch": 13.785,
+      "grad_norm": 4.410267353057861,
+      "learning_rate": 3.1075000000000003e-06,
+      "loss": 0.3151,
+      "step": 2757
+    },
+    {
+      "epoch": 13.79,
+      "grad_norm": 5.563354969024658,
+      "learning_rate": 3.1050000000000003e-06,
+      "loss": 0.337,
+      "step": 2758
+    },
+    {
+      "epoch": 13.795,
+      "grad_norm": 3.192826747894287,
+      "learning_rate": 3.1025000000000004e-06,
+      "loss": 0.1894,
+      "step": 2759
+    },
+    {
+      "epoch": 13.8,
+      "grad_norm": 5.754162311553955,
+      "learning_rate": 3.1000000000000004e-06,
+      "loss": 0.2925,
+      "step": 2760
+    },
+    {
+      "epoch": 13.805,
+      "grad_norm": 5.789515018463135,
+      "learning_rate": 3.0975000000000004e-06,
+      "loss": 0.2207,
+      "step": 2761
+    },
+    {
+      "epoch": 13.81,
+      "grad_norm": 5.546463966369629,
+      "learning_rate": 3.0950000000000004e-06,
+      "loss": 0.3367,
+      "step": 2762
+    },
+    {
+      "epoch": 13.815,
+      "grad_norm": 3.8791401386260986,
+      "learning_rate": 3.0925000000000005e-06,
+      "loss": 0.4548,
+      "step": 2763
+    },
+    {
+      "epoch": 13.82,
+      "grad_norm": 4.269449710845947,
+      "learning_rate": 3.09e-06,
+      "loss": 0.7155,
+      "step": 2764
+    },
+    {
+      "epoch": 13.825,
+      "grad_norm": 4.424363613128662,
+      "learning_rate": 3.0875000000000005e-06,
+      "loss": 0.6501,
+      "step": 2765
+    },
+    {
+      "epoch": 13.83,
+      "grad_norm": 6.146673679351807,
+      "learning_rate": 3.085e-06,
+      "loss": 0.4118,
+      "step": 2766
+    },
+    {
+      "epoch": 13.835,
+      "grad_norm": 8.60024642944336,
+      "learning_rate": 3.0825000000000006e-06,
+      "loss": 0.4786,
+      "step": 2767
+    },
+    {
+      "epoch": 13.84,
+      "grad_norm": 5.119856834411621,
+      "learning_rate": 3.08e-06,
+      "loss": 0.4546,
+      "step": 2768
+    },
+    {
+      "epoch": 13.845,
+      "grad_norm": 4.460896015167236,
+      "learning_rate": 3.0775000000000006e-06,
+      "loss": 0.1992,
+      "step": 2769
+    },
+    {
+      "epoch": 13.85,
+      "grad_norm": 5.615173816680908,
+      "learning_rate": 3.075e-06,
+      "loss": 0.1586,
+      "step": 2770
+    },
+    {
+      "epoch": 13.855,
+      "grad_norm": 3.6638989448547363,
+      "learning_rate": 3.0725000000000007e-06,
+      "loss": 0.5622,
+      "step": 2771
+    },
+    {
+      "epoch": 13.86,
+      "grad_norm": 5.286365985870361,
+      "learning_rate": 3.0700000000000003e-06,
+      "loss": 0.2922,
+      "step": 2772
+    },
+    {
+      "epoch": 13.865,
+      "grad_norm": 4.394322872161865,
+      "learning_rate": 3.0675000000000003e-06,
+      "loss": 0.3078,
+      "step": 2773
+    },
+    {
+      "epoch": 13.87,
+      "grad_norm": 3.3733880519866943,
+      "learning_rate": 3.0650000000000003e-06,
+      "loss": 0.2666,
+      "step": 2774
+    },
+    {
+      "epoch": 13.875,
+      "grad_norm": 3.725795269012451,
+      "learning_rate": 3.0625000000000003e-06,
+      "loss": 0.1878,
+      "step": 2775
+    },
+    {
+      "epoch": 13.88,
+      "grad_norm": 11.321919441223145,
+      "learning_rate": 3.0600000000000003e-06,
+      "loss": 0.909,
+      "step": 2776
+    },
+    {
+      "epoch": 13.885,
+      "grad_norm": 4.446516513824463,
+      "learning_rate": 3.0575000000000004e-06,
+      "loss": 0.2069,
+      "step": 2777
+    },
+    {
+      "epoch": 13.89,
+      "grad_norm": 3.8478379249572754,
+      "learning_rate": 3.0550000000000004e-06,
+      "loss": 0.1998,
+      "step": 2778
+    },
+    {
+      "epoch": 13.895,
+      "grad_norm": 3.1270041465759277,
+      "learning_rate": 3.0525000000000004e-06,
+      "loss": 0.1864,
+      "step": 2779
+    },
+    {
+      "epoch": 13.9,
+      "grad_norm": 6.413837432861328,
+      "learning_rate": 3.05e-06,
+      "loss": 0.4383,
+      "step": 2780
+    },
+    {
+      "epoch": 13.905,
+      "grad_norm": 2.921950340270996,
+      "learning_rate": 3.0475000000000005e-06,
+      "loss": 0.161,
+      "step": 2781
+    },
+    {
+      "epoch": 13.91,
+      "grad_norm": 3.4269487857818604,
+      "learning_rate": 3.045e-06,
+      "loss": 0.1634,
+      "step": 2782
+    },
+    {
+      "epoch": 13.915,
+      "grad_norm": 4.3338470458984375,
+      "learning_rate": 3.0425000000000005e-06,
+      "loss": 0.3954,
+      "step": 2783
+    },
+    {
+      "epoch": 13.92,
+      "grad_norm": 5.959560394287109,
+      "learning_rate": 3.04e-06,
+      "loss": 0.3665,
+      "step": 2784
+    },
+    {
+      "epoch": 13.925,
+      "grad_norm": 4.5074076652526855,
+      "learning_rate": 3.0375000000000006e-06,
+      "loss": 0.1496,
+      "step": 2785
+    },
+    {
+      "epoch": 13.93,
+      "grad_norm": 2.289268970489502,
+      "learning_rate": 3.035e-06,
+      "loss": 0.1968,
+      "step": 2786
+    },
+    {
+      "epoch": 13.935,
+      "grad_norm": 5.289707660675049,
+      "learning_rate": 3.0325000000000006e-06,
+      "loss": 0.3258,
+      "step": 2787
+    },
+    {
+      "epoch": 13.94,
+      "grad_norm": 4.607162952423096,
+      "learning_rate": 3.0300000000000002e-06,
+      "loss": 0.463,
+      "step": 2788
+    },
+    {
+      "epoch": 13.945,
+      "grad_norm": 3.3369274139404297,
+      "learning_rate": 3.0275000000000002e-06,
+      "loss": 0.3358,
+      "step": 2789
+    },
+    {
+      "epoch": 13.95,
+      "grad_norm": 6.468296051025391,
+      "learning_rate": 3.0250000000000003e-06,
+      "loss": 0.4016,
+      "step": 2790
+    },
+    {
+      "epoch": 13.955,
+      "grad_norm": 4.887318134307861,
+      "learning_rate": 3.0225000000000003e-06,
+      "loss": 0.6652,
+      "step": 2791
+    },
+    {
+      "epoch": 13.96,
+      "grad_norm": 3.470227003097534,
+      "learning_rate": 3.0200000000000003e-06,
+      "loss": 0.2174,
+      "step": 2792
+    },
+    {
+      "epoch": 13.965,
+      "grad_norm": 8.786201477050781,
+      "learning_rate": 3.0175000000000003e-06,
+      "loss": 0.4575,
+      "step": 2793
+    },
+    {
+      "epoch": 13.97,
+      "grad_norm": 3.7398910522460938,
+      "learning_rate": 3.0150000000000004e-06,
+      "loss": 0.5431,
+      "step": 2794
+    },
+    {
+      "epoch": 13.975,
+      "grad_norm": 3.6543006896972656,
+      "learning_rate": 3.0125000000000004e-06,
+      "loss": 0.2876,
+      "step": 2795
+    },
+    {
+      "epoch": 13.98,
+      "grad_norm": 4.363594055175781,
+      "learning_rate": 3.01e-06,
+      "loss": 0.4234,
+      "step": 2796
+    },
+    {
+      "epoch": 13.985,
+      "grad_norm": 4.8803324699401855,
+      "learning_rate": 3.0075000000000004e-06,
+      "loss": 0.3149,
+      "step": 2797
+    },
+    {
+      "epoch": 13.99,
+      "grad_norm": 3.938737154006958,
+      "learning_rate": 3.005e-06,
+      "loss": 0.2535,
+      "step": 2798
+    },
+    {
+      "epoch": 13.995,
+      "grad_norm": 4.415417671203613,
+      "learning_rate": 3.0025000000000005e-06,
+      "loss": 0.4422,
+      "step": 2799
+    },
+    {
+      "epoch": 14.0,
+      "grad_norm": 5.837874412536621,
+      "learning_rate": 3e-06,
+      "loss": 0.2436,
+      "step": 2800
+    },
+    {
+      "epoch": 14.0,
+      "eval_loss": 0.20994079113006592,
+      "eval_runtime": 15.6513,
+      "eval_samples_per_second": 50.986,
+      "eval_steps_per_second": 6.389,
+      "step": 2800
+    },
+    {
+      "epoch": 14.005,
+      "grad_norm": 5.01359224319458,
+      "learning_rate": 2.9975000000000005e-06,
+      "loss": 0.1743,
+      "step": 2801
+    },
+    {
+      "epoch": 14.01,
+      "grad_norm": 2.639007329940796,
+      "learning_rate": 2.995e-06,
+      "loss": 0.3207,
+      "step": 2802
+    },
+    {
+      "epoch": 14.015,
+      "grad_norm": 4.276294231414795,
+      "learning_rate": 2.9925000000000006e-06,
+      "loss": 0.2943,
+      "step": 2803
+    },
+    {
+      "epoch": 14.02,
+      "grad_norm": 2.784391164779663,
+      "learning_rate": 2.99e-06,
+      "loss": 0.2505,
+      "step": 2804
+    },
+    {
+      "epoch": 14.025,
+      "grad_norm": 6.044054985046387,
+      "learning_rate": 2.9875e-06,
+      "loss": 0.1026,
+      "step": 2805
+    },
+    {
+      "epoch": 14.03,
+      "grad_norm": 2.551337718963623,
+      "learning_rate": 2.9850000000000002e-06,
+      "loss": 0.2405,
+      "step": 2806
+    },
+    {
+      "epoch": 14.035,
+      "grad_norm": 4.802229404449463,
+      "learning_rate": 2.9825000000000002e-06,
+      "loss": 0.6306,
+      "step": 2807
+    },
+    {
+      "epoch": 14.04,
+      "grad_norm": 3.224566698074341,
+      "learning_rate": 2.9800000000000003e-06,
+      "loss": 0.3837,
+      "step": 2808
+    },
+    {
+      "epoch": 14.045,
+      "grad_norm": 4.355310916900635,
+      "learning_rate": 2.9775000000000003e-06,
+      "loss": 0.2965,
+      "step": 2809
+    },
+    {
+      "epoch": 14.05,
+      "grad_norm": 4.9176177978515625,
+      "learning_rate": 2.9750000000000003e-06,
+      "loss": 0.5024,
+      "step": 2810
+    },
+    {
+      "epoch": 14.055,
+      "grad_norm": 3.84378719329834,
+      "learning_rate": 2.9725000000000003e-06,
+      "loss": 0.2897,
+      "step": 2811
+    },
+    {
+      "epoch": 14.06,
+      "grad_norm": 1.8339954614639282,
+      "learning_rate": 2.97e-06,
+      "loss": 0.0707,
+      "step": 2812
+    },
+    {
+      "epoch": 14.065,
+      "grad_norm": 3.194967746734619,
+      "learning_rate": 2.9675000000000004e-06,
+      "loss": 0.125,
+      "step": 2813
+    },
+    {
+      "epoch": 14.07,
+      "grad_norm": 3.3036513328552246,
+      "learning_rate": 2.965e-06,
+      "loss": 0.2429,
+      "step": 2814
+    },
+    {
+      "epoch": 14.075,
+      "grad_norm": 5.395280838012695,
+      "learning_rate": 2.9625000000000004e-06,
+      "loss": 0.699,
+      "step": 2815
+    },
+    {
+      "epoch": 14.08,
+      "grad_norm": 5.59435510635376,
+      "learning_rate": 2.96e-06,
+      "loss": 0.5877,
+      "step": 2816
+    },
+    {
+      "epoch": 14.085,
+      "grad_norm": 3.082697868347168,
+      "learning_rate": 2.9575000000000005e-06,
+      "loss": 0.3256,
+      "step": 2817
+    },
+    {
+      "epoch": 14.09,
+      "grad_norm": 4.451581001281738,
+      "learning_rate": 2.955e-06,
+      "loss": 0.265,
+      "step": 2818
+    },
+    {
+      "epoch": 14.095,
+      "grad_norm": 4.798580169677734,
+      "learning_rate": 2.9525000000000005e-06,
+      "loss": 0.3005,
+      "step": 2819
+    },
+    {
+      "epoch": 14.1,
+      "grad_norm": 2.6561508178710938,
+      "learning_rate": 2.95e-06,
+      "loss": 0.0866,
+      "step": 2820
+    },
+    {
+      "epoch": 14.105,
+      "grad_norm": 3.660815954208374,
+      "learning_rate": 2.9475e-06,
+      "loss": 0.3475,
+      "step": 2821
+    },
+    {
+      "epoch": 14.11,
+      "grad_norm": 2.507158041000366,
+      "learning_rate": 2.945e-06,
+      "loss": 0.167,
+      "step": 2822
+    },
+    {
+      "epoch": 14.115,
+      "grad_norm": 2.442861557006836,
+      "learning_rate": 2.9425e-06,
+      "loss": 0.1932,
+      "step": 2823
+    },
+    {
+      "epoch": 14.12,
+      "grad_norm": 4.054699420928955,
+      "learning_rate": 2.9400000000000002e-06,
+      "loss": 0.2087,
+      "step": 2824
+    },
+    {
+      "epoch": 14.125,
+      "grad_norm": 4.421236515045166,
+      "learning_rate": 2.9375000000000003e-06,
+      "loss": 0.2725,
+      "step": 2825
+    },
+    {
+      "epoch": 14.13,
+      "grad_norm": 3.780001401901245,
+      "learning_rate": 2.9350000000000003e-06,
+      "loss": 0.2068,
+      "step": 2826
+    },
+    {
+      "epoch": 14.135,
+      "grad_norm": 5.034428119659424,
+      "learning_rate": 2.9325000000000003e-06,
+      "loss": 0.4766,
+      "step": 2827
+    },
+    {
+      "epoch": 14.14,
+      "grad_norm": 2.785090208053589,
+      "learning_rate": 2.93e-06,
+      "loss": 0.0484,
+      "step": 2828
+    },
+    {
+      "epoch": 14.145,
+      "grad_norm": 4.416884422302246,
+      "learning_rate": 2.9275000000000003e-06,
+      "loss": 0.1985,
+      "step": 2829
+    },
+    {
+      "epoch": 14.15,
+      "grad_norm": 7.123900413513184,
+      "learning_rate": 2.925e-06,
+      "loss": 0.4832,
+      "step": 2830
+    },
+    {
+      "epoch": 14.155,
+      "grad_norm": 4.456885814666748,
+      "learning_rate": 2.9225000000000004e-06,
+      "loss": 0.2135,
+      "step": 2831
+    },
+    {
+      "epoch": 14.16,
+      "grad_norm": 9.616790771484375,
+      "learning_rate": 2.92e-06,
+      "loss": 0.8607,
+      "step": 2832
+    },
+    {
+      "epoch": 14.165,
+      "grad_norm": 2.050347089767456,
+      "learning_rate": 2.9175000000000004e-06,
+      "loss": 0.2224,
+      "step": 2833
+    },
+    {
+      "epoch": 14.17,
+      "grad_norm": 5.433361530303955,
+      "learning_rate": 2.915e-06,
+      "loss": 0.315,
+      "step": 2834
+    },
+    {
+      "epoch": 14.175,
+      "grad_norm": 5.712324619293213,
+      "learning_rate": 2.9125000000000005e-06,
+      "loss": 0.4743,
+      "step": 2835
+    },
+    {
+      "epoch": 14.18,
+      "grad_norm": 3.9418728351593018,
+      "learning_rate": 2.91e-06,
+      "loss": 0.3998,
+      "step": 2836
+    },
+    {
+      "epoch": 14.185,
+      "grad_norm": 16.07750129699707,
+      "learning_rate": 2.9075e-06,
+      "loss": 0.3302,
+      "step": 2837
+    },
+    {
+      "epoch": 14.19,
+      "grad_norm": 16.15114974975586,
+      "learning_rate": 2.905e-06,
+      "loss": 1.2726,
+      "step": 2838
+    },
+    {
+      "epoch": 14.195,
+      "grad_norm": 3.043646812438965,
+      "learning_rate": 2.9025e-06,
+      "loss": 0.0842,
+      "step": 2839
+    },
+    {
+      "epoch": 14.2,
+      "grad_norm": 6.526980876922607,
+      "learning_rate": 2.9e-06,
+      "loss": 0.4141,
+      "step": 2840
+    },
+    {
+      "epoch": 14.205,
+      "grad_norm": 5.75487756729126,
+      "learning_rate": 2.8975e-06,
+      "loss": 0.251,
+      "step": 2841
+    },
+    {
+      "epoch": 14.21,
+      "grad_norm": 7.653066158294678,
+      "learning_rate": 2.8950000000000002e-06,
+      "loss": 0.4919,
+      "step": 2842
+    },
+    {
+      "epoch": 14.215,
+      "grad_norm": 2.9612858295440674,
+      "learning_rate": 2.8925000000000003e-06,
+      "loss": 0.1655,
+      "step": 2843
+    },
+    {
+      "epoch": 14.22,
+      "grad_norm": 3.858091115951538,
+      "learning_rate": 2.89e-06,
+      "loss": 0.3499,
+      "step": 2844
+    },
+    {
+      "epoch": 14.225,
+      "grad_norm": 4.152839660644531,
+      "learning_rate": 2.8875000000000003e-06,
+      "loss": 0.1338,
+      "step": 2845
+    },
+    {
+      "epoch": 14.23,
+      "grad_norm": 109.11841583251953,
+      "learning_rate": 2.885e-06,
+      "loss": 1.0403,
+      "step": 2846
+    },
+    {
+      "epoch": 14.235,
+      "grad_norm": 5.4302802085876465,
+      "learning_rate": 2.8825000000000004e-06,
+      "loss": 0.4928,
+      "step": 2847
+    },
+    {
+      "epoch": 14.24,
+      "grad_norm": 4.435436725616455,
+      "learning_rate": 2.88e-06,
+      "loss": 0.3745,
+      "step": 2848
+    },
+    {
+      "epoch": 14.245,
+      "grad_norm": 4.73663854598999,
+      "learning_rate": 2.8775000000000004e-06,
+      "loss": 0.2284,
+      "step": 2849
+    },
+    {
+      "epoch": 14.25,
+      "grad_norm": 3.9541730880737305,
+      "learning_rate": 2.875e-06,
+      "loss": 0.166,
+      "step": 2850
+    },
+    {
+      "epoch": 14.255,
+      "grad_norm": 4.068276405334473,
+      "learning_rate": 2.8725000000000004e-06,
+      "loss": 0.1351,
+      "step": 2851
+    },
+    {
+      "epoch": 14.26,
+      "grad_norm": 5.920307636260986,
+      "learning_rate": 2.87e-06,
+      "loss": 0.5167,
+      "step": 2852
+    },
+    {
+      "epoch": 14.265,
+      "grad_norm": 3.046992063522339,
+      "learning_rate": 2.8675e-06,
+      "loss": 0.1681,
+      "step": 2853
+    },
+    {
+      "epoch": 14.27,
+      "grad_norm": 3.253187656402588,
+      "learning_rate": 2.865e-06,
+      "loss": 0.1737,
+      "step": 2854
+    },
+    {
+      "epoch": 14.275,
+      "grad_norm": 3.7747397422790527,
+      "learning_rate": 2.8625e-06,
+      "loss": 0.1729,
+      "step": 2855
+    },
+    {
+      "epoch": 14.28,
+      "grad_norm": 6.188694000244141,
+      "learning_rate": 2.86e-06,
+      "loss": 0.3365,
+      "step": 2856
+    },
+    {
+      "epoch": 14.285,
+      "grad_norm": 5.016397476196289,
+      "learning_rate": 2.8575e-06,
+      "loss": 0.3376,
+      "step": 2857
+    },
+    {
+      "epoch": 14.29,
+      "grad_norm": 3.83217716217041,
+      "learning_rate": 2.855e-06,
+      "loss": 0.3918,
+      "step": 2858
+    },
+    {
+      "epoch": 14.295,
+      "grad_norm": 4.064434051513672,
+      "learning_rate": 2.8525000000000002e-06,
+      "loss": 0.0731,
+      "step": 2859
+    },
+    {
+      "epoch": 14.3,
+      "grad_norm": 6.552209377288818,
+      "learning_rate": 2.85e-06,
+      "loss": 0.3666,
+      "step": 2860
+    },
+    {
+      "epoch": 14.305,
+      "grad_norm": 3.186866283416748,
+      "learning_rate": 2.8475000000000003e-06,
+      "loss": 0.2232,
+      "step": 2861
+    },
+    {
+      "epoch": 14.31,
+      "grad_norm": 4.299386501312256,
+      "learning_rate": 2.845e-06,
+      "loss": 0.4734,
+      "step": 2862
+    },
+    {
+      "epoch": 14.315,
+      "grad_norm": 6.168147563934326,
+      "learning_rate": 2.8425000000000003e-06,
+      "loss": 0.73,
+      "step": 2863
+    },
+    {
+      "epoch": 14.32,
+      "grad_norm": 5.09070348739624,
+      "learning_rate": 2.84e-06,
+      "loss": 0.2325,
+      "step": 2864
+    },
+    {
+      "epoch": 14.325,
+      "grad_norm": 3.2935614585876465,
+      "learning_rate": 2.8375000000000004e-06,
+      "loss": 0.2505,
+      "step": 2865
+    },
+    {
+      "epoch": 14.33,
+      "grad_norm": 5.599834442138672,
+      "learning_rate": 2.835e-06,
+      "loss": 0.5593,
+      "step": 2866
+    },
+    {
+      "epoch": 14.335,
+      "grad_norm": 5.485595226287842,
+      "learning_rate": 2.8325000000000004e-06,
+      "loss": 0.4119,
+      "step": 2867
+    },
+    {
+      "epoch": 14.34,
+      "grad_norm": 3.4383649826049805,
+      "learning_rate": 2.83e-06,
+      "loss": 0.4087,
+      "step": 2868
+    },
+    {
+      "epoch": 14.345,
+      "grad_norm": 4.463839530944824,
+      "learning_rate": 2.8275e-06,
+      "loss": 0.2469,
+      "step": 2869
+    },
+    {
+      "epoch": 14.35,
+      "grad_norm": 11.599230766296387,
+      "learning_rate": 2.825e-06,
+      "loss": 0.5604,
+      "step": 2870
+    },
+    {
+      "epoch": 14.355,
+      "grad_norm": 7.849985122680664,
+      "learning_rate": 2.8225e-06,
+      "loss": 1.0186,
+      "step": 2871
+    },
+    {
+      "epoch": 14.36,
+      "grad_norm": 6.265213966369629,
+      "learning_rate": 2.82e-06,
+      "loss": 0.3471,
+      "step": 2872
+    },
+    {
+      "epoch": 14.365,
+      "grad_norm": 3.6639747619628906,
+      "learning_rate": 2.8175e-06,
+      "loss": 0.1603,
+      "step": 2873
+    },
+    {
+      "epoch": 14.37,
+      "grad_norm": 5.183992385864258,
+      "learning_rate": 2.815e-06,
+      "loss": 0.261,
+      "step": 2874
+    },
+    {
+      "epoch": 14.375,
+      "grad_norm": 2.124757766723633,
+      "learning_rate": 2.8125e-06,
+      "loss": 0.1083,
+      "step": 2875
+    },
+    {
+      "epoch": 14.38,
+      "grad_norm": 3.1801185607910156,
+      "learning_rate": 2.8100000000000006e-06,
+      "loss": 0.2053,
+      "step": 2876
+    },
+    {
+      "epoch": 14.385,
+      "grad_norm": 4.897256374359131,
+      "learning_rate": 2.8075000000000002e-06,
+      "loss": 0.4351,
+      "step": 2877
+    },
+    {
+      "epoch": 14.39,
+      "grad_norm": 4.534748554229736,
+      "learning_rate": 2.8050000000000007e-06,
+      "loss": 0.4921,
+      "step": 2878
+    },
+    {
+      "epoch": 14.395,
+      "grad_norm": 3.5778791904449463,
+      "learning_rate": 2.8025000000000003e-06,
+      "loss": 0.2973,
+      "step": 2879
+    },
+    {
+      "epoch": 14.4,
+      "grad_norm": 3.631765127182007,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 0.2245,
+      "step": 2880
+    },
+    {
+      "epoch": 14.405,
+      "grad_norm": 3.3186404705047607,
+      "learning_rate": 2.7975000000000003e-06,
+      "loss": 0.2353,
+      "step": 2881
+    },
+    {
+      "epoch": 14.41,
+      "grad_norm": 3.3155155181884766,
+      "learning_rate": 2.7950000000000003e-06,
+      "loss": 0.1822,
+      "step": 2882
+    },
+    {
+      "epoch": 14.415,
+      "grad_norm": 5.866283893585205,
+      "learning_rate": 2.7925000000000004e-06,
+      "loss": 0.2277,
+      "step": 2883
+    },
+    {
+      "epoch": 14.42,
+      "grad_norm": 3.149761438369751,
+      "learning_rate": 2.7900000000000004e-06,
+      "loss": 0.2215,
+      "step": 2884
+    },
+    {
+      "epoch": 14.425,
+      "grad_norm": 5.893641471862793,
+      "learning_rate": 2.7875000000000004e-06,
+      "loss": 0.4232,
+      "step": 2885
+    },
+    {
+      "epoch": 14.43,
+      "grad_norm": 3.9935860633850098,
+      "learning_rate": 2.7850000000000004e-06,
+      "loss": 0.2515,
+      "step": 2886
+    },
+    {
+      "epoch": 14.435,
+      "grad_norm": 8.18681812286377,
+      "learning_rate": 2.7825e-06,
+      "loss": 0.4632,
+      "step": 2887
+    },
+    {
+      "epoch": 14.44,
+      "grad_norm": 4.235905170440674,
+      "learning_rate": 2.7800000000000005e-06,
+      "loss": 0.1823,
+      "step": 2888
+    },
+    {
+      "epoch": 14.445,
+      "grad_norm": 2.0676558017730713,
+      "learning_rate": 2.7775e-06,
+      "loss": 0.1405,
+      "step": 2889
+    },
+    {
+      "epoch": 14.45,
+      "grad_norm": 6.326505184173584,
+      "learning_rate": 2.7750000000000005e-06,
+      "loss": 0.5712,
+      "step": 2890
+    },
+    {
+      "epoch": 14.455,
+      "grad_norm": 3.400968551635742,
+      "learning_rate": 2.7725e-06,
+      "loss": 0.3677,
+      "step": 2891
+    },
+    {
+      "epoch": 14.46,
+      "grad_norm": 6.130422115325928,
+      "learning_rate": 2.7700000000000006e-06,
+      "loss": 0.5928,
+      "step": 2892
+    },
+    {
+      "epoch": 14.465,
+      "grad_norm": 2.9719674587249756,
+      "learning_rate": 2.7675e-06,
+      "loss": 0.1056,
+      "step": 2893
+    },
+    {
+      "epoch": 14.47,
+      "grad_norm": 3.428532600402832,
+      "learning_rate": 2.7650000000000006e-06,
+      "loss": 0.2387,
+      "step": 2894
+    },
+    {
+      "epoch": 14.475,
+      "grad_norm": 7.73337984085083,
+      "learning_rate": 2.7625000000000002e-06,
+      "loss": 0.3998,
+      "step": 2895
+    },
+    {
+      "epoch": 14.48,
+      "grad_norm": 4.3975443840026855,
+      "learning_rate": 2.7600000000000003e-06,
+      "loss": 0.2328,
+      "step": 2896
+    },
+    {
+      "epoch": 14.485,
+      "grad_norm": 3.372309446334839,
+      "learning_rate": 2.7575000000000003e-06,
+      "loss": 0.2518,
+      "step": 2897
+    },
+    {
+      "epoch": 14.49,
+      "grad_norm": 3.587087392807007,
+      "learning_rate": 2.7550000000000003e-06,
+      "loss": 0.2452,
+      "step": 2898
+    },
+    {
+      "epoch": 14.495,
+      "grad_norm": 3.6475183963775635,
+      "learning_rate": 2.7525000000000003e-06,
+      "loss": 0.1078,
+      "step": 2899
+    },
+    {
+      "epoch": 14.5,
+      "grad_norm": 3.5961222648620605,
+      "learning_rate": 2.7500000000000004e-06,
+      "loss": 0.1234,
+      "step": 2900
+    },
+    {
+      "epoch": 14.505,
+      "grad_norm": 5.708836555480957,
+      "learning_rate": 2.7475000000000004e-06,
+      "loss": 0.4441,
+      "step": 2901
+    },
+    {
+      "epoch": 14.51,
+      "grad_norm": 4.050062656402588,
+      "learning_rate": 2.7450000000000004e-06,
+      "loss": 0.3456,
+      "step": 2902
+    },
+    {
+      "epoch": 14.515,
+      "grad_norm": 5.57611083984375,
+      "learning_rate": 2.7425e-06,
+      "loss": 0.2613,
+      "step": 2903
+    },
+    {
+      "epoch": 14.52,
+      "grad_norm": 4.473504066467285,
+      "learning_rate": 2.7400000000000004e-06,
+      "loss": 0.3004,
+      "step": 2904
+    },
+    {
+      "epoch": 14.525,
+      "grad_norm": 11.805594444274902,
+      "learning_rate": 2.7375e-06,
+      "loss": 0.6336,
+      "step": 2905
+    },
+    {
+      "epoch": 14.53,
+      "grad_norm": 12.832369804382324,
+      "learning_rate": 2.7350000000000005e-06,
+      "loss": 0.5723,
+      "step": 2906
+    },
+    {
+      "epoch": 14.535,
+      "grad_norm": 4.725528717041016,
+      "learning_rate": 2.7325e-06,
+      "loss": 0.3692,
+      "step": 2907
+    },
+    {
+      "epoch": 14.54,
+      "grad_norm": 2.986116647720337,
+      "learning_rate": 2.7300000000000005e-06,
+      "loss": 0.2664,
+      "step": 2908
+    },
+    {
+      "epoch": 14.545,
+      "grad_norm": 4.817147731781006,
+      "learning_rate": 2.7275e-06,
+      "loss": 0.3258,
+      "step": 2909
+    },
+    {
+      "epoch": 14.55,
+      "grad_norm": 2.3819682598114014,
+      "learning_rate": 2.7250000000000006e-06,
+      "loss": 0.1231,
+      "step": 2910
+    },
+    {
+      "epoch": 14.555,
+      "grad_norm": 3.731292963027954,
+      "learning_rate": 2.7225e-06,
+      "loss": 0.3075,
+      "step": 2911
+    },
+    {
+      "epoch": 14.56,
+      "grad_norm": 3.8343050479888916,
+      "learning_rate": 2.7200000000000002e-06,
+      "loss": 0.496,
+      "step": 2912
+    },
+    {
+      "epoch": 14.565,
+      "grad_norm": 2.6631815433502197,
+      "learning_rate": 2.7175000000000002e-06,
+      "loss": 0.1891,
+      "step": 2913
+    },
+    {
+      "epoch": 14.57,
+      "grad_norm": 5.217584609985352,
+      "learning_rate": 2.7150000000000003e-06,
+      "loss": 0.25,
+      "step": 2914
+    },
+    {
+      "epoch": 14.575,
+      "grad_norm": 3.385634422302246,
+      "learning_rate": 2.7125000000000003e-06,
+      "loss": 0.0272,
+      "step": 2915
+    },
+    {
+      "epoch": 14.58,
+      "grad_norm": 3.713735818862915,
+      "learning_rate": 2.7100000000000003e-06,
+      "loss": 0.1125,
+      "step": 2916
+    },
+    {
+      "epoch": 14.585,
+      "grad_norm": 4.0107245445251465,
+      "learning_rate": 2.7075000000000003e-06,
+      "loss": 0.1972,
+      "step": 2917
+    },
+    {
+      "epoch": 14.59,
+      "grad_norm": 6.135714530944824,
+      "learning_rate": 2.7050000000000004e-06,
+      "loss": 0.5578,
+      "step": 2918
+    },
+    {
+      "epoch": 14.595,
+      "grad_norm": 6.022823333740234,
+      "learning_rate": 2.7025e-06,
+      "loss": 0.6465,
+      "step": 2919
+    },
+    {
+      "epoch": 14.6,
+      "grad_norm": 3.901965618133545,
+      "learning_rate": 2.7000000000000004e-06,
+      "loss": 0.1992,
+      "step": 2920
+    },
+    {
+      "epoch": 14.605,
+      "grad_norm": 2.233861207962036,
+      "learning_rate": 2.6975e-06,
+      "loss": 0.1134,
+      "step": 2921
+    },
+    {
+      "epoch": 14.61,
+      "grad_norm": 4.574281692504883,
+      "learning_rate": 2.6950000000000005e-06,
+      "loss": 0.1906,
+      "step": 2922
+    },
+    {
+      "epoch": 14.615,
+      "grad_norm": 4.76069450378418,
+      "learning_rate": 2.6925e-06,
+      "loss": 0.235,
+      "step": 2923
+    },
+    {
+      "epoch": 14.62,
+      "grad_norm": 6.2615532875061035,
+      "learning_rate": 2.6900000000000005e-06,
+      "loss": 0.5436,
+      "step": 2924
+    },
+    {
+      "epoch": 14.625,
+      "grad_norm": 3.6986448764801025,
+      "learning_rate": 2.6875e-06,
+      "loss": 0.33,
+      "step": 2925
+    },
+    {
+      "epoch": 14.63,
+      "grad_norm": 6.145036697387695,
+      "learning_rate": 2.6850000000000006e-06,
+      "loss": 0.3446,
+      "step": 2926
+    },
+    {
+      "epoch": 14.635,
+      "grad_norm": 5.884594917297363,
+      "learning_rate": 2.6825e-06,
+      "loss": 1.0038,
+      "step": 2927
+    },
+    {
+      "epoch": 14.64,
+      "grad_norm": 3.4012131690979004,
+      "learning_rate": 2.68e-06,
+      "loss": 0.2096,
+      "step": 2928
+    },
+    {
+      "epoch": 14.645,
+      "grad_norm": 4.130620956420898,
+      "learning_rate": 2.6775e-06,
+      "loss": 0.1667,
+      "step": 2929
+    },
+    {
+      "epoch": 14.65,
+      "grad_norm": 4.691809177398682,
+      "learning_rate": 2.6750000000000002e-06,
+      "loss": 0.2902,
+      "step": 2930
+    },
+    {
+      "epoch": 14.655,
+      "grad_norm": 4.130764961242676,
+      "learning_rate": 2.6725000000000002e-06,
+      "loss": 0.3708,
+      "step": 2931
+    },
+    {
+      "epoch": 14.66,
+      "grad_norm": 4.483701705932617,
+      "learning_rate": 2.6700000000000003e-06,
+      "loss": 0.2027,
+      "step": 2932
+    },
+    {
+      "epoch": 14.665,
+      "grad_norm": 4.669301986694336,
+      "learning_rate": 2.6675000000000003e-06,
+      "loss": 0.5704,
+      "step": 2933
+    },
+    {
+      "epoch": 14.67,
+      "grad_norm": 2.9664230346679688,
+      "learning_rate": 2.6650000000000003e-06,
+      "loss": 0.1409,
+      "step": 2934
+    },
+    {
+      "epoch": 14.675,
+      "grad_norm": 9.559643745422363,
+      "learning_rate": 2.6625e-06,
+      "loss": 0.854,
+      "step": 2935
+    },
+    {
+      "epoch": 14.68,
+      "grad_norm": 5.868269443511963,
+      "learning_rate": 2.6600000000000004e-06,
+      "loss": 0.4681,
+      "step": 2936
+    },
+    {
+      "epoch": 14.685,
+      "grad_norm": 6.734318733215332,
+      "learning_rate": 2.6575e-06,
+      "loss": 0.4056,
+      "step": 2937
+    },
+    {
+      "epoch": 14.69,
+      "grad_norm": 4.015400409698486,
+      "learning_rate": 2.6550000000000004e-06,
+      "loss": 0.3031,
+      "step": 2938
+    },
+    {
+      "epoch": 14.695,
+      "grad_norm": 4.237788200378418,
+      "learning_rate": 2.6525e-06,
+      "loss": 0.3527,
+      "step": 2939
+    },
+    {
+      "epoch": 14.7,
+      "grad_norm": 3.542008638381958,
+      "learning_rate": 2.6500000000000005e-06,
+      "loss": 0.1847,
+      "step": 2940
+    },
+    {
+      "epoch": 14.705,
+      "grad_norm": 3.1657683849334717,
+      "learning_rate": 2.6475e-06,
+      "loss": 0.1629,
+      "step": 2941
+    },
+    {
+      "epoch": 14.71,
+      "grad_norm": 9.93850326538086,
+      "learning_rate": 2.6450000000000005e-06,
+      "loss": 0.5753,
+      "step": 2942
+    },
+    {
+      "epoch": 14.715,
+      "grad_norm": 4.277157783508301,
+      "learning_rate": 2.6425e-06,
+      "loss": 0.3428,
+      "step": 2943
+    },
+    {
+      "epoch": 14.72,
+      "grad_norm": 3.781299591064453,
+      "learning_rate": 2.64e-06,
+      "loss": 0.2695,
+      "step": 2944
+    },
+    {
+      "epoch": 14.725,
+      "grad_norm": 1.532659888267517,
+      "learning_rate": 2.6375e-06,
+      "loss": 0.0389,
+      "step": 2945
+    },
+    {
+      "epoch": 14.73,
+      "grad_norm": 5.310700416564941,
+      "learning_rate": 2.635e-06,
+      "loss": 0.3104,
+      "step": 2946
+    },
+    {
+      "epoch": 14.735,
+      "grad_norm": 2.7025699615478516,
+      "learning_rate": 2.6325e-06,
+      "loss": 0.3849,
+      "step": 2947
+    },
+    {
+      "epoch": 14.74,
+      "grad_norm": 2.551384687423706,
+      "learning_rate": 2.6300000000000002e-06,
+      "loss": 0.2239,
+      "step": 2948
+    },
+    {
+      "epoch": 14.745,
+      "grad_norm": 9.906868934631348,
+      "learning_rate": 2.6275000000000003e-06,
+      "loss": 0.5513,
+      "step": 2949
+    },
+    {
+      "epoch": 14.75,
+      "grad_norm": 3.7779667377471924,
+      "learning_rate": 2.6250000000000003e-06,
+      "loss": 0.321,
+      "step": 2950
+    },
+    {
+      "epoch": 14.755,
+      "grad_norm": 97.15838623046875,
+      "learning_rate": 2.6225e-06,
+      "loss": 0.6905,
+      "step": 2951
+    },
+    {
+      "epoch": 14.76,
+      "grad_norm": 4.150561809539795,
+      "learning_rate": 2.6200000000000003e-06,
+      "loss": 0.306,
+      "step": 2952
+    },
+    {
+      "epoch": 14.765,
+      "grad_norm": 3.7971608638763428,
+      "learning_rate": 2.6175e-06,
+      "loss": 0.1688,
+      "step": 2953
+    },
+    {
+      "epoch": 14.77,
+      "grad_norm": 4.498149394989014,
+      "learning_rate": 2.6150000000000004e-06,
+      "loss": 0.1714,
+      "step": 2954
+    },
+    {
+      "epoch": 14.775,
+      "grad_norm": 2.5333163738250732,
+      "learning_rate": 2.6125e-06,
+      "loss": 0.0924,
+      "step": 2955
+    },
+    {
+      "epoch": 14.78,
+      "grad_norm": 2.8399128913879395,
+      "learning_rate": 2.6100000000000004e-06,
+      "loss": 0.1967,
+      "step": 2956
+    },
+    {
+      "epoch": 14.785,
+      "grad_norm": 4.420997142791748,
+      "learning_rate": 2.6075e-06,
+      "loss": 0.2923,
+      "step": 2957
+    },
+    {
+      "epoch": 14.79,
+      "grad_norm": 6.664902210235596,
+      "learning_rate": 2.6050000000000005e-06,
+      "loss": 0.4315,
+      "step": 2958
+    },
+    {
+      "epoch": 14.795,
+      "grad_norm": 6.604458808898926,
+      "learning_rate": 2.6025e-06,
+      "loss": 0.5429,
+      "step": 2959
+    },
+    {
+      "epoch": 14.8,
+      "grad_norm": 3.2393696308135986,
+      "learning_rate": 2.6e-06,
+      "loss": 0.2099,
+      "step": 2960
+    },
+    {
+      "epoch": 14.805,
+      "grad_norm": 9.887956619262695,
+      "learning_rate": 2.5975e-06,
+      "loss": 0.4893,
+      "step": 2961
+    },
+    {
+      "epoch": 14.81,
+      "grad_norm": 4.350278854370117,
+      "learning_rate": 2.595e-06,
+      "loss": 0.3818,
+      "step": 2962
+    },
+    {
+      "epoch": 14.815,
+      "grad_norm": 7.744044303894043,
+      "learning_rate": 2.5925e-06,
+      "loss": 0.3578,
+      "step": 2963
+    },
+    {
+      "epoch": 14.82,
+      "grad_norm": 9.987229347229004,
+      "learning_rate": 2.59e-06,
+      "loss": 0.565,
+      "step": 2964
+    },
+    {
+      "epoch": 14.825,
+      "grad_norm": 4.109292507171631,
+      "learning_rate": 2.5875000000000002e-06,
+      "loss": 0.1474,
+      "step": 2965
+    },
+    {
+      "epoch": 14.83,
+      "grad_norm": 7.630754470825195,
+      "learning_rate": 2.5850000000000002e-06,
+      "loss": 0.7976,
+      "step": 2966
+    },
+    {
+      "epoch": 14.835,
+      "grad_norm": 1.9726728200912476,
+      "learning_rate": 2.5825e-06,
+      "loss": 0.1759,
+      "step": 2967
+    },
+    {
+      "epoch": 14.84,
+      "grad_norm": 5.120915412902832,
+      "learning_rate": 2.5800000000000003e-06,
+      "loss": 0.2684,
+      "step": 2968
+    },
+    {
+      "epoch": 14.845,
+      "grad_norm": 2.580174207687378,
+      "learning_rate": 2.5775e-06,
+      "loss": 0.0691,
+      "step": 2969
+    },
+    {
+      "epoch": 14.85,
+      "grad_norm": 5.163429260253906,
+      "learning_rate": 2.5750000000000003e-06,
+      "loss": 0.3275,
+      "step": 2970
+    },
+    {
+      "epoch": 14.855,
+      "grad_norm": 3.1634650230407715,
+      "learning_rate": 2.5725e-06,
+      "loss": 0.358,
+      "step": 2971
+    },
+    {
+      "epoch": 14.86,
+      "grad_norm": 2.9550273418426514,
+      "learning_rate": 2.5700000000000004e-06,
+      "loss": 0.1296,
+      "step": 2972
+    },
+    {
+      "epoch": 14.865,
+      "grad_norm": 5.008945465087891,
+      "learning_rate": 2.5675e-06,
+      "loss": 0.4985,
+      "step": 2973
+    },
+    {
+      "epoch": 14.87,
+      "grad_norm": 7.707421779632568,
+      "learning_rate": 2.5650000000000004e-06,
+      "loss": 0.6473,
+      "step": 2974
+    },
+    {
+      "epoch": 14.875,
+      "grad_norm": 3.834484338760376,
+      "learning_rate": 2.5625e-06,
+      "loss": 0.2044,
+      "step": 2975
+    },
+    {
+      "epoch": 14.88,
+      "grad_norm": 6.532800674438477,
+      "learning_rate": 2.56e-06,
+      "loss": 0.2966,
+      "step": 2976
+    },
+    {
+      "epoch": 14.885,
+      "grad_norm": 10.070513725280762,
+      "learning_rate": 2.5575e-06,
+      "loss": 0.5975,
+      "step": 2977
+    },
+    {
+      "epoch": 14.89,
+      "grad_norm": 7.73294734954834,
+      "learning_rate": 2.555e-06,
+      "loss": 0.5435,
+      "step": 2978
+    },
+    {
+      "epoch": 14.895,
+      "grad_norm": 3.982605457305908,
+      "learning_rate": 2.5525e-06,
+      "loss": 0.2237,
+      "step": 2979
+    },
+    {
+      "epoch": 14.9,
+      "grad_norm": 3.9253714084625244,
+      "learning_rate": 2.55e-06,
+      "loss": 0.14,
+      "step": 2980
+    },
+    {
+      "epoch": 14.905,
+      "grad_norm": 5.402847766876221,
+      "learning_rate": 2.5475e-06,
+      "loss": 0.1921,
+      "step": 2981
+    },
+    {
+      "epoch": 14.91,
+      "grad_norm": 4.493337631225586,
+      "learning_rate": 2.545e-06,
+      "loss": 0.3276,
+      "step": 2982
+    },
+    {
+      "epoch": 14.915,
+      "grad_norm": 7.478642463684082,
+      "learning_rate": 2.5425e-06,
+      "loss": 0.3678,
+      "step": 2983
+    },
+    {
+      "epoch": 14.92,
+      "grad_norm": 3.557004451751709,
+      "learning_rate": 2.5400000000000002e-06,
+      "loss": 0.3344,
+      "step": 2984
+    },
+    {
+      "epoch": 14.925,
+      "grad_norm": 3.0203752517700195,
+      "learning_rate": 2.5375e-06,
+      "loss": 0.147,
+      "step": 2985
+    },
+    {
+      "epoch": 14.93,
+      "grad_norm": 4.17352819442749,
+      "learning_rate": 2.5350000000000003e-06,
+      "loss": 0.4143,
+      "step": 2986
+    },
+    {
+      "epoch": 14.935,
+      "grad_norm": 3.129591703414917,
+      "learning_rate": 2.5325e-06,
+      "loss": 0.1856,
+      "step": 2987
+    },
+    {
+      "epoch": 14.94,
+      "grad_norm": 14.81112003326416,
+      "learning_rate": 2.5300000000000003e-06,
+      "loss": 0.6848,
+      "step": 2988
+    },
+    {
+      "epoch": 14.945,
+      "grad_norm": 8.879931449890137,
+      "learning_rate": 2.5275e-06,
+      "loss": 0.3589,
+      "step": 2989
+    },
+    {
+      "epoch": 14.95,
+      "grad_norm": 7.179346084594727,
+      "learning_rate": 2.5250000000000004e-06,
+      "loss": 0.7032,
+      "step": 2990
+    },
+    {
+      "epoch": 14.955,
+      "grad_norm": 5.148601531982422,
+      "learning_rate": 2.5225e-06,
+      "loss": 0.328,
+      "step": 2991
+    },
+    {
+      "epoch": 14.96,
+      "grad_norm": 4.435116767883301,
+      "learning_rate": 2.52e-06,
+      "loss": 0.2195,
+      "step": 2992
+    },
+    {
+      "epoch": 14.965,
+      "grad_norm": 12.129100799560547,
+      "learning_rate": 2.5175e-06,
+      "loss": 0.7488,
+      "step": 2993
+    },
+    {
+      "epoch": 14.97,
+      "grad_norm": 14.786237716674805,
+      "learning_rate": 2.515e-06,
+      "loss": 0.8216,
+      "step": 2994
+    },
+    {
+      "epoch": 14.975,
+      "grad_norm": 5.465200901031494,
+      "learning_rate": 2.5125e-06,
+      "loss": 0.3183,
+      "step": 2995
+    },
+    {
+      "epoch": 14.98,
+      "grad_norm": 3.7007482051849365,
+      "learning_rate": 2.51e-06,
+      "loss": 0.2598,
+      "step": 2996
+    },
+    {
+      "epoch": 14.985,
+      "grad_norm": 5.134115219116211,
+      "learning_rate": 2.5075e-06,
+      "loss": 0.4758,
+      "step": 2997
+    },
+    {
+      "epoch": 14.99,
+      "grad_norm": 5.178332805633545,
+      "learning_rate": 2.505e-06,
+      "loss": 0.2468,
+      "step": 2998
+    },
+    {
+      "epoch": 14.995,
+      "grad_norm": 58.34430694580078,
+      "learning_rate": 2.5024999999999998e-06,
+      "loss": 0.6462,
+      "step": 2999
+    },
+    {
+      "epoch": 15.0,
+      "grad_norm": 10.602553367614746,
+      "learning_rate": 2.5e-06,
+      "loss": 1.1728,
+      "step": 3000
+    },
+    {
+      "epoch": 15.0,
+      "eval_loss": 0.20335747301578522,
+      "eval_runtime": 15.3679,
+      "eval_samples_per_second": 51.927,
+      "eval_steps_per_second": 6.507,
+      "step": 3000
+    },
+    {
+      "epoch": 15.005,
+      "grad_norm": 3.655702829360962,
+      "learning_rate": 2.4975000000000002e-06,
+      "loss": 0.2366,
+      "step": 3001
+    },
+    {
+      "epoch": 15.01,
+      "grad_norm": 2.451974868774414,
+      "learning_rate": 2.4950000000000003e-06,
+      "loss": 0.2847,
+      "step": 3002
+    },
+    {
+      "epoch": 15.015,
+      "grad_norm": 3.2273175716400146,
+      "learning_rate": 2.4925000000000003e-06,
+      "loss": 0.3262,
+      "step": 3003
+    },
+    {
+      "epoch": 15.02,
+      "grad_norm": 5.280306816101074,
+      "learning_rate": 2.4900000000000003e-06,
+      "loss": 0.5878,
+      "step": 3004
+    },
+    {
+      "epoch": 15.025,
+      "grad_norm": 19.149995803833008,
+      "learning_rate": 2.4875000000000003e-06,
+      "loss": 0.6354,
+      "step": 3005
+    },
+    {
+      "epoch": 15.03,
+      "grad_norm": 3.063966751098633,
+      "learning_rate": 2.4850000000000003e-06,
+      "loss": 0.1793,
+      "step": 3006
+    },
+    {
+      "epoch": 15.035,
+      "grad_norm": 3.9570693969726562,
+      "learning_rate": 2.4825000000000004e-06,
+      "loss": 0.1921,
+      "step": 3007
+    },
+    {
+      "epoch": 15.04,
+      "grad_norm": 5.406646251678467,
+      "learning_rate": 2.4800000000000004e-06,
+      "loss": 0.4544,
+      "step": 3008
+    },
+    {
+      "epoch": 15.045,
+      "grad_norm": 8.194955825805664,
+      "learning_rate": 2.4775e-06,
+      "loss": 0.5085,
+      "step": 3009
+    },
+    {
+      "epoch": 15.05,
+      "grad_norm": 3.8690025806427,
+      "learning_rate": 2.475e-06,
+      "loss": 0.1301,
+      "step": 3010
+    },
+    {
+      "epoch": 15.055,
+      "grad_norm": 3.0510995388031006,
+      "learning_rate": 2.4725e-06,
+      "loss": 0.1896,
+      "step": 3011
+    },
+    {
+      "epoch": 15.06,
+      "grad_norm": 4.348089218139648,
+      "learning_rate": 2.47e-06,
+      "loss": 0.359,
+      "step": 3012
+    },
+    {
+      "epoch": 15.065,
+      "grad_norm": 2.70632266998291,
+      "learning_rate": 2.4675e-06,
+      "loss": 0.1191,
+      "step": 3013
+    },
+    {
+      "epoch": 15.07,
+      "grad_norm": 4.988892555236816,
+      "learning_rate": 2.465e-06,
+      "loss": 0.3607,
+      "step": 3014
+    },
+    {
+      "epoch": 15.075,
+      "grad_norm": 3.6979446411132812,
+      "learning_rate": 2.4625e-06,
+      "loss": 0.1332,
+      "step": 3015
+    },
+    {
+      "epoch": 15.08,
+      "grad_norm": 4.453253269195557,
+      "learning_rate": 2.46e-06,
+      "loss": 0.4334,
+      "step": 3016
+    },
+    {
+      "epoch": 15.085,
+      "grad_norm": 4.78302001953125,
+      "learning_rate": 2.4575e-06,
+      "loss": 0.4609,
+      "step": 3017
+    },
+    {
+      "epoch": 15.09,
+      "grad_norm": 3.090935230255127,
+      "learning_rate": 2.4550000000000002e-06,
+      "loss": 0.1259,
+      "step": 3018
+    },
+    {
+      "epoch": 15.095,
+      "grad_norm": 5.700923919677734,
+      "learning_rate": 2.4525000000000002e-06,
+      "loss": 0.6785,
+      "step": 3019
+    },
+    {
+      "epoch": 15.1,
+      "grad_norm": 5.009494304656982,
+      "learning_rate": 2.4500000000000003e-06,
+      "loss": 0.3438,
+      "step": 3020
+    },
+    {
+      "epoch": 15.105,
+      "grad_norm": 3.938011884689331,
+      "learning_rate": 2.4475000000000003e-06,
+      "loss": 0.2572,
+      "step": 3021
+    },
+    {
+      "epoch": 15.11,
+      "grad_norm": 5.388587951660156,
+      "learning_rate": 2.4450000000000003e-06,
+      "loss": 0.553,
+      "step": 3022
+    },
+    {
+      "epoch": 15.115,
+      "grad_norm": 4.634590148925781,
+      "learning_rate": 2.4425000000000003e-06,
+      "loss": 0.2109,
+      "step": 3023
+    },
+    {
+      "epoch": 15.12,
+      "grad_norm": 3.248481035232544,
+      "learning_rate": 2.4400000000000004e-06,
+      "loss": 0.3273,
+      "step": 3024
+    },
+    {
+      "epoch": 15.125,
+      "grad_norm": 4.283886909484863,
+      "learning_rate": 2.4375e-06,
+      "loss": 0.3722,
+      "step": 3025
+    },
+    {
+      "epoch": 15.13,
+      "grad_norm": 4.2436041831970215,
+      "learning_rate": 2.435e-06,
+      "loss": 0.3909,
+      "step": 3026
+    },
+    {
+      "epoch": 15.135,
+      "grad_norm": 6.356209754943848,
+      "learning_rate": 2.4325e-06,
+      "loss": 0.4502,
+      "step": 3027
+    },
+    {
+      "epoch": 15.14,
+      "grad_norm": 4.773648262023926,
+      "learning_rate": 2.43e-06,
+      "loss": 0.3935,
+      "step": 3028
+    },
+    {
+      "epoch": 15.145,
+      "grad_norm": 4.603784084320068,
+      "learning_rate": 2.4275e-06,
+      "loss": 0.2792,
+      "step": 3029
+    },
+    {
+      "epoch": 15.15,
+      "grad_norm": 4.480953693389893,
+      "learning_rate": 2.425e-06,
+      "loss": 0.3062,
+      "step": 3030
+    },
+    {
+      "epoch": 15.155,
+      "grad_norm": 8.494505882263184,
+      "learning_rate": 2.4225e-06,
+      "loss": 0.3787,
+      "step": 3031
+    },
+    {
+      "epoch": 15.16,
+      "grad_norm": 5.387947082519531,
+      "learning_rate": 2.42e-06,
+      "loss": 0.3358,
+      "step": 3032
+    },
+    {
+      "epoch": 15.165,
+      "grad_norm": 5.089359283447266,
+      "learning_rate": 2.4175e-06,
+      "loss": 0.1267,
+      "step": 3033
+    },
+    {
+      "epoch": 15.17,
+      "grad_norm": 6.907346248626709,
+      "learning_rate": 2.415e-06,
+      "loss": 0.4949,
+      "step": 3034
+    },
+    {
+      "epoch": 15.175,
+      "grad_norm": 3.8406240940093994,
+      "learning_rate": 2.4125e-06,
+      "loss": 0.2058,
+      "step": 3035
+    },
+    {
+      "epoch": 15.18,
+      "grad_norm": 9.053852081298828,
+      "learning_rate": 2.4100000000000002e-06,
+      "loss": 0.2012,
+      "step": 3036
+    },
+    {
+      "epoch": 15.185,
+      "grad_norm": 1.731350064277649,
+      "learning_rate": 2.4075000000000002e-06,
+      "loss": 0.0963,
+      "step": 3037
+    },
+    {
+      "epoch": 15.19,
+      "grad_norm": 6.082643985748291,
+      "learning_rate": 2.4050000000000003e-06,
+      "loss": 0.2417,
+      "step": 3038
+    },
+    {
+      "epoch": 15.195,
+      "grad_norm": 4.203982353210449,
+      "learning_rate": 2.4025000000000003e-06,
+      "loss": 0.4479,
+      "step": 3039
+    },
+    {
+      "epoch": 15.2,
+      "grad_norm": 4.470425605773926,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 0.3224,
+      "step": 3040
+    },
+    {
+      "epoch": 15.205,
+      "grad_norm": 2.5686800479888916,
+      "learning_rate": 2.3975e-06,
+      "loss": 0.2595,
+      "step": 3041
+    },
+    {
+      "epoch": 15.21,
+      "grad_norm": 2.6989030838012695,
+      "learning_rate": 2.395e-06,
+      "loss": 0.2489,
+      "step": 3042
+    },
+    {
+      "epoch": 15.215,
+      "grad_norm": 4.441350936889648,
+      "learning_rate": 2.3925e-06,
+      "loss": 0.2215,
+      "step": 3043
+    },
+    {
+      "epoch": 15.22,
+      "grad_norm": 5.279112815856934,
+      "learning_rate": 2.39e-06,
+      "loss": 0.4106,
+      "step": 3044
+    },
+    {
+      "epoch": 15.225,
+      "grad_norm": 3.037468671798706,
+      "learning_rate": 2.3875e-06,
+      "loss": 0.2792,
+      "step": 3045
+    },
+    {
+      "epoch": 15.23,
+      "grad_norm": 2.339402914047241,
+      "learning_rate": 2.385e-06,
+      "loss": 0.0948,
+      "step": 3046
+    },
+    {
+      "epoch": 15.235,
+      "grad_norm": 3.6543102264404297,
+      "learning_rate": 2.3825e-06,
+      "loss": 0.5664,
+      "step": 3047
+    },
+    {
+      "epoch": 15.24,
+      "grad_norm": 1.1036807298660278,
+      "learning_rate": 2.38e-06,
+      "loss": 0.0173,
+      "step": 3048
+    },
+    {
+      "epoch": 15.245,
+      "grad_norm": 6.524346828460693,
+      "learning_rate": 2.3775e-06,
+      "loss": 0.4957,
+      "step": 3049
+    },
+    {
+      "epoch": 15.25,
+      "grad_norm": 6.945590019226074,
+      "learning_rate": 2.375e-06,
+      "loss": 0.5742,
+      "step": 3050
+    },
+    {
+      "epoch": 15.255,
+      "grad_norm": 5.461936950683594,
+      "learning_rate": 2.3725e-06,
+      "loss": 0.382,
+      "step": 3051
+    },
+    {
+      "epoch": 15.26,
+      "grad_norm": 2.413191556930542,
+      "learning_rate": 2.37e-06,
+      "loss": 0.132,
+      "step": 3052
+    },
+    {
+      "epoch": 15.265,
+      "grad_norm": 3.205247163772583,
+      "learning_rate": 2.3675e-06,
+      "loss": 0.1486,
+      "step": 3053
+    },
+    {
+      "epoch": 15.27,
+      "grad_norm": 8.999713897705078,
+      "learning_rate": 2.3650000000000002e-06,
+      "loss": 0.4356,
+      "step": 3054
+    },
+    {
+      "epoch": 15.275,
+      "grad_norm": 3.087646245956421,
+      "learning_rate": 2.3625000000000003e-06,
+      "loss": 0.3506,
+      "step": 3055
+    },
+    {
+      "epoch": 15.28,
+      "grad_norm": 5.838226795196533,
+      "learning_rate": 2.3600000000000003e-06,
+      "loss": 0.2441,
+      "step": 3056
+    },
+    {
+      "epoch": 15.285,
+      "grad_norm": 5.782450199127197,
+      "learning_rate": 2.3575e-06,
+      "loss": 0.4981,
+      "step": 3057
+    },
+    {
+      "epoch": 15.29,
+      "grad_norm": 3.42917537689209,
+      "learning_rate": 2.355e-06,
+      "loss": 0.2665,
+      "step": 3058
+    },
+    {
+      "epoch": 15.295,
+      "grad_norm": 3.563577890396118,
+      "learning_rate": 2.3525e-06,
+      "loss": 0.1161,
+      "step": 3059
+    },
+    {
+      "epoch": 15.3,
+      "grad_norm": 3.220884084701538,
+      "learning_rate": 2.35e-06,
+      "loss": 0.1476,
+      "step": 3060
+    },
+    {
+      "epoch": 15.305,
+      "grad_norm": 9.673454284667969,
+      "learning_rate": 2.3475e-06,
+      "loss": 0.5737,
+      "step": 3061
+    },
+    {
+      "epoch": 15.31,
+      "grad_norm": 2.966461658477783,
+      "learning_rate": 2.345e-06,
+      "loss": 0.0642,
+      "step": 3062
+    },
+    {
+      "epoch": 15.315,
+      "grad_norm": 5.834527969360352,
+      "learning_rate": 2.3425000000000004e-06,
+      "loss": 0.699,
+      "step": 3063
+    },
+    {
+      "epoch": 15.32,
+      "grad_norm": 2.7250776290893555,
+      "learning_rate": 2.3400000000000005e-06,
+      "loss": 0.3725,
+      "step": 3064
+    },
+    {
+      "epoch": 15.325,
+      "grad_norm": 4.776270866394043,
+      "learning_rate": 2.3375000000000005e-06,
+      "loss": 0.588,
+      "step": 3065
+    },
+    {
+      "epoch": 15.33,
+      "grad_norm": 5.584927082061768,
+      "learning_rate": 2.3350000000000005e-06,
+      "loss": 0.5738,
+      "step": 3066
+    },
+    {
+      "epoch": 15.335,
+      "grad_norm": 2.558159351348877,
+      "learning_rate": 2.3325e-06,
+      "loss": 0.1995,
+      "step": 3067
+    },
+    {
+      "epoch": 15.34,
+      "grad_norm": 5.797266006469727,
+      "learning_rate": 2.33e-06,
+      "loss": 0.3877,
+      "step": 3068
+    },
+    {
+      "epoch": 15.345,
+      "grad_norm": 4.067357063293457,
+      "learning_rate": 2.3275e-06,
+      "loss": 0.2502,
+      "step": 3069
+    },
+    {
+      "epoch": 15.35,
+      "grad_norm": 4.991832733154297,
+      "learning_rate": 2.325e-06,
+      "loss": 0.5369,
+      "step": 3070
+    },
+    {
+      "epoch": 15.355,
+      "grad_norm": 4.256237030029297,
+      "learning_rate": 2.3225e-06,
+      "loss": 0.5845,
+      "step": 3071
+    },
+    {
+      "epoch": 15.36,
+      "grad_norm": 4.7166428565979,
+      "learning_rate": 2.3200000000000002e-06,
+      "loss": 0.1112,
+      "step": 3072
+    },
+    {
+      "epoch": 15.365,
+      "grad_norm": 5.0812788009643555,
+      "learning_rate": 2.3175000000000003e-06,
+      "loss": 0.1804,
+      "step": 3073
+    },
+    {
+      "epoch": 15.37,
+      "grad_norm": 4.766618728637695,
+      "learning_rate": 2.3150000000000003e-06,
+      "loss": 0.325,
+      "step": 3074
+    },
+    {
+      "epoch": 15.375,
+      "grad_norm": 3.6853365898132324,
+      "learning_rate": 2.3125000000000003e-06,
+      "loss": 0.0914,
+      "step": 3075
+    },
+    {
+      "epoch": 15.38,
+      "grad_norm": 4.531446933746338,
+      "learning_rate": 2.3100000000000003e-06,
+      "loss": 0.254,
+      "step": 3076
+    },
+    {
+      "epoch": 15.385,
+      "grad_norm": 2.790302276611328,
+      "learning_rate": 2.3075000000000004e-06,
+      "loss": 0.1968,
+      "step": 3077
+    },
+    {
+      "epoch": 15.39,
+      "grad_norm": 2.0093204975128174,
+      "learning_rate": 2.3050000000000004e-06,
+      "loss": 0.0776,
+      "step": 3078
+    },
+    {
+      "epoch": 15.395,
+      "grad_norm": 4.001646041870117,
+      "learning_rate": 2.3025000000000004e-06,
+      "loss": 0.2506,
+      "step": 3079
+    },
+    {
+      "epoch": 15.4,
+      "grad_norm": 4.944500923156738,
+      "learning_rate": 2.3000000000000004e-06,
+      "loss": 0.437,
+      "step": 3080
+    },
+    {
+      "epoch": 15.405,
+      "grad_norm": 4.405821323394775,
+      "learning_rate": 2.2975000000000004e-06,
+      "loss": 0.4922,
+      "step": 3081
+    },
+    {
+      "epoch": 15.41,
+      "grad_norm": 4.068020820617676,
+      "learning_rate": 2.2950000000000005e-06,
+      "loss": 0.3449,
+      "step": 3082
+    },
+    {
+      "epoch": 15.415,
+      "grad_norm": 4.433620452880859,
+      "learning_rate": 2.2925e-06,
+      "loss": 0.2138,
+      "step": 3083
+    },
+    {
+      "epoch": 15.42,
+      "grad_norm": 3.966602325439453,
+      "learning_rate": 2.29e-06,
+      "loss": 0.3982,
+      "step": 3084
+    },
+    {
+      "epoch": 15.425,
+      "grad_norm": 3.598874568939209,
+      "learning_rate": 2.2875e-06,
+      "loss": 0.5231,
+      "step": 3085
+    },
+    {
+      "epoch": 15.43,
+      "grad_norm": 4.647474765777588,
+      "learning_rate": 2.285e-06,
+      "loss": 0.2759,
+      "step": 3086
+    },
+    {
+      "epoch": 15.435,
+      "grad_norm": 3.9509329795837402,
+      "learning_rate": 2.2825e-06,
+      "loss": 0.1782,
+      "step": 3087
+    },
+    {
+      "epoch": 15.44,
+      "grad_norm": 6.365646839141846,
+      "learning_rate": 2.28e-06,
+      "loss": 0.4212,
+      "step": 3088
+    },
+    {
+      "epoch": 15.445,
+      "grad_norm": 2.5152533054351807,
+      "learning_rate": 2.2775000000000002e-06,
+      "loss": 0.0661,
+      "step": 3089
+    },
+    {
+      "epoch": 15.45,
+      "grad_norm": 3.834230661392212,
+      "learning_rate": 2.2750000000000002e-06,
+      "loss": 0.2625,
+      "step": 3090
+    },
+    {
+      "epoch": 15.455,
+      "grad_norm": 2.1329407691955566,
+      "learning_rate": 2.2725000000000003e-06,
+      "loss": 0.1351,
+      "step": 3091
+    },
+    {
+      "epoch": 15.46,
+      "grad_norm": 5.297748565673828,
+      "learning_rate": 2.2700000000000003e-06,
+      "loss": 0.6742,
+      "step": 3092
+    },
+    {
+      "epoch": 15.465,
+      "grad_norm": 3.4902470111846924,
+      "learning_rate": 2.2675000000000003e-06,
+      "loss": 0.3273,
+      "step": 3093
+    },
+    {
+      "epoch": 15.47,
+      "grad_norm": 10.504515647888184,
+      "learning_rate": 2.2650000000000003e-06,
+      "loss": 0.5688,
+      "step": 3094
+    },
+    {
+      "epoch": 15.475,
+      "grad_norm": 3.7388157844543457,
+      "learning_rate": 2.2625000000000004e-06,
+      "loss": 0.3037,
+      "step": 3095
+    },
+    {
+      "epoch": 15.48,
+      "grad_norm": 4.067750453948975,
+      "learning_rate": 2.2600000000000004e-06,
+      "loss": 0.5877,
+      "step": 3096
+    },
+    {
+      "epoch": 15.485,
+      "grad_norm": 260.2000427246094,
+      "learning_rate": 2.2575000000000004e-06,
+      "loss": 0.3791,
+      "step": 3097
+    },
+    {
+      "epoch": 15.49,
+      "grad_norm": 7.443624019622803,
+      "learning_rate": 2.2550000000000004e-06,
+      "loss": 0.8124,
+      "step": 3098
+    },
+    {
+      "epoch": 15.495,
+      "grad_norm": 4.725823402404785,
+      "learning_rate": 2.2525e-06,
+      "loss": 0.5624,
+      "step": 3099
+    },
+    {
+      "epoch": 15.5,
+      "grad_norm": 4.212906837463379,
+      "learning_rate": 2.25e-06,
+      "loss": 0.2856,
+      "step": 3100
+    },
+    {
+      "epoch": 15.505,
+      "grad_norm": 12.581936836242676,
+      "learning_rate": 2.2475e-06,
+      "loss": 0.9387,
+      "step": 3101
+    },
+    {
+      "epoch": 15.51,
+      "grad_norm": 6.142890930175781,
+      "learning_rate": 2.245e-06,
+      "loss": 0.328,
+      "step": 3102
+    },
+    {
+      "epoch": 15.515,
+      "grad_norm": 2.2868151664733887,
+      "learning_rate": 2.2425e-06,
+      "loss": 0.0277,
+      "step": 3103
+    },
+    {
+      "epoch": 15.52,
+      "grad_norm": 3.3269705772399902,
+      "learning_rate": 2.24e-06,
+      "loss": 0.1698,
+      "step": 3104
+    },
+    {
+      "epoch": 15.525,
+      "grad_norm": 5.219757080078125,
+      "learning_rate": 2.2375e-06,
+      "loss": 0.4788,
+      "step": 3105
+    },
+    {
+      "epoch": 15.53,
+      "grad_norm": 5.078825950622559,
+      "learning_rate": 2.235e-06,
+      "loss": 0.2925,
+      "step": 3106
+    },
+    {
+      "epoch": 15.535,
+      "grad_norm": 4.462192535400391,
+      "learning_rate": 2.2325000000000002e-06,
+      "loss": 0.4073,
+      "step": 3107
+    },
+    {
+      "epoch": 15.54,
+      "grad_norm": 2.7567050457000732,
+      "learning_rate": 2.2300000000000002e-06,
+      "loss": 0.1528,
+      "step": 3108
+    },
+    {
+      "epoch": 15.545,
+      "grad_norm": 3.695281744003296,
+      "learning_rate": 2.2275000000000003e-06,
+      "loss": 0.1647,
+      "step": 3109
+    },
+    {
+      "epoch": 15.55,
+      "grad_norm": 7.101161956787109,
+      "learning_rate": 2.2250000000000003e-06,
+      "loss": 0.6802,
+      "step": 3110
+    },
+    {
+      "epoch": 15.555,
+      "grad_norm": 3.2206151485443115,
+      "learning_rate": 2.2225000000000003e-06,
+      "loss": 0.1953,
+      "step": 3111
+    },
+    {
+      "epoch": 15.56,
+      "grad_norm": 7.38597297668457,
+      "learning_rate": 2.2200000000000003e-06,
+      "loss": 0.531,
+      "step": 3112
+    },
+    {
+      "epoch": 15.565,
+      "grad_norm": 5.759932994842529,
+      "learning_rate": 2.2175000000000004e-06,
+      "loss": 0.3748,
+      "step": 3113
+    },
+    {
+      "epoch": 15.57,
+      "grad_norm": 6.2886457443237305,
+      "learning_rate": 2.2150000000000004e-06,
+      "loss": 0.5574,
+      "step": 3114
+    },
+    {
+      "epoch": 15.575,
+      "grad_norm": 3.5693233013153076,
+      "learning_rate": 2.2125e-06,
+      "loss": 0.2949,
+      "step": 3115
+    },
+    {
+      "epoch": 15.58,
+      "grad_norm": 3.4980831146240234,
+      "learning_rate": 2.21e-06,
+      "loss": 0.2623,
+      "step": 3116
+    },
+    {
+      "epoch": 15.585,
+      "grad_norm": 2.792644739151001,
+      "learning_rate": 2.2075e-06,
+      "loss": 0.2916,
+      "step": 3117
+    },
+    {
+      "epoch": 15.59,
+      "grad_norm": 3.8246428966522217,
+      "learning_rate": 2.205e-06,
+      "loss": 0.3906,
+      "step": 3118
+    },
+    {
+      "epoch": 15.595,
+      "grad_norm": 3.790663957595825,
+      "learning_rate": 2.2025e-06,
+      "loss": 0.2126,
+      "step": 3119
+    },
+    {
+      "epoch": 15.6,
+      "grad_norm": 3.60733699798584,
+      "learning_rate": 2.2e-06,
+      "loss": 0.2367,
+      "step": 3120
+    },
+    {
+      "epoch": 15.605,
+      "grad_norm": 4.902060508728027,
+      "learning_rate": 2.1975e-06,
+      "loss": 0.2083,
+      "step": 3121
+    },
+    {
+      "epoch": 15.61,
+      "grad_norm": 6.764699935913086,
+      "learning_rate": 2.195e-06,
+      "loss": 0.5573,
+      "step": 3122
+    },
+    {
+      "epoch": 15.615,
+      "grad_norm": 3.108431816101074,
+      "learning_rate": 2.1925e-06,
+      "loss": 0.1747,
+      "step": 3123
+    },
+    {
+      "epoch": 15.62,
+      "grad_norm": 4.220947265625,
+      "learning_rate": 2.19e-06,
+      "loss": 0.2044,
+      "step": 3124
+    },
+    {
+      "epoch": 15.625,
+      "grad_norm": 5.1478590965271,
+      "learning_rate": 2.1875000000000002e-06,
+      "loss": 0.0998,
+      "step": 3125
+    },
+    {
+      "epoch": 15.63,
+      "grad_norm": 7.7752604484558105,
+      "learning_rate": 2.1850000000000003e-06,
+      "loss": 0.5516,
+      "step": 3126
+    },
+    {
+      "epoch": 15.635,
+      "grad_norm": 3.662782907485962,
+      "learning_rate": 2.1825000000000003e-06,
+      "loss": 0.1346,
+      "step": 3127
+    },
+    {
+      "epoch": 15.64,
+      "grad_norm": 2.7775628566741943,
+      "learning_rate": 2.1800000000000003e-06,
+      "loss": 0.2747,
+      "step": 3128
+    },
+    {
+      "epoch": 15.645,
+      "grad_norm": 10.91114330291748,
+      "learning_rate": 2.1775000000000003e-06,
+      "loss": 0.3122,
+      "step": 3129
+    },
+    {
+      "epoch": 15.65,
+      "grad_norm": 3.6708970069885254,
+      "learning_rate": 2.1750000000000004e-06,
+      "loss": 0.1018,
+      "step": 3130
+    },
+    {
+      "epoch": 15.655,
+      "grad_norm": 8.447978019714355,
+      "learning_rate": 2.1725000000000004e-06,
+      "loss": 0.573,
+      "step": 3131
+    },
+    {
+      "epoch": 15.66,
+      "grad_norm": 3.3864943981170654,
+      "learning_rate": 2.17e-06,
+      "loss": 0.2422,
+      "step": 3132
+    },
+    {
+      "epoch": 15.665,
+      "grad_norm": 3.7010810375213623,
+      "learning_rate": 2.1675e-06,
+      "loss": 0.3073,
+      "step": 3133
+    },
+    {
+      "epoch": 15.67,
+      "grad_norm": 5.281312465667725,
+      "learning_rate": 2.165e-06,
+      "loss": 0.2157,
+      "step": 3134
+    },
+    {
+      "epoch": 15.675,
+      "grad_norm": 3.2229299545288086,
+      "learning_rate": 2.1625e-06,
+      "loss": 0.2371,
+      "step": 3135
+    },
+    {
+      "epoch": 15.68,
+      "grad_norm": 4.059622764587402,
+      "learning_rate": 2.16e-06,
+      "loss": 0.3715,
+      "step": 3136
+    },
+    {
+      "epoch": 15.685,
+      "grad_norm": 3.1468262672424316,
+      "learning_rate": 2.1575e-06,
+      "loss": 0.464,
+      "step": 3137
+    },
+    {
+      "epoch": 15.69,
+      "grad_norm": 3.9346096515655518,
+      "learning_rate": 2.155e-06,
+      "loss": 0.2623,
+      "step": 3138
+    },
+    {
+      "epoch": 15.695,
+      "grad_norm": 3.7853074073791504,
+      "learning_rate": 2.1525e-06,
+      "loss": 0.2492,
+      "step": 3139
+    },
+    {
+      "epoch": 15.7,
+      "grad_norm": 3.7260003089904785,
+      "learning_rate": 2.15e-06,
+      "loss": 0.278,
+      "step": 3140
+    },
+    {
+      "epoch": 15.705,
+      "grad_norm": 7.160543441772461,
+      "learning_rate": 2.1475e-06,
+      "loss": 0.4476,
+      "step": 3141
+    },
+    {
+      "epoch": 15.71,
+      "grad_norm": 6.385192394256592,
+      "learning_rate": 2.1450000000000002e-06,
+      "loss": 0.2352,
+      "step": 3142
+    },
+    {
+      "epoch": 15.715,
+      "grad_norm": 3.641991138458252,
+      "learning_rate": 2.1425000000000002e-06,
+      "loss": 0.2207,
+      "step": 3143
+    },
+    {
+      "epoch": 15.72,
+      "grad_norm": 6.031627178192139,
+      "learning_rate": 2.1400000000000003e-06,
+      "loss": 0.3815,
+      "step": 3144
+    },
+    {
+      "epoch": 15.725,
+      "grad_norm": 2.7314064502716064,
+      "learning_rate": 2.1375000000000003e-06,
+      "loss": 0.0779,
+      "step": 3145
+    },
+    {
+      "epoch": 15.73,
+      "grad_norm": 4.6987199783325195,
+      "learning_rate": 2.1350000000000003e-06,
+      "loss": 0.3657,
+      "step": 3146
+    },
+    {
+      "epoch": 15.735,
+      "grad_norm": 3.43854022026062,
+      "learning_rate": 2.1325000000000003e-06,
+      "loss": 0.135,
+      "step": 3147
+    },
+    {
+      "epoch": 15.74,
+      "grad_norm": 2.2921388149261475,
+      "learning_rate": 2.13e-06,
+      "loss": 0.0985,
+      "step": 3148
+    },
+    {
+      "epoch": 15.745,
+      "grad_norm": 3.0667850971221924,
+      "learning_rate": 2.1275e-06,
+      "loss": 0.0542,
+      "step": 3149
+    },
+    {
+      "epoch": 15.75,
+      "grad_norm": 4.806392192840576,
+      "learning_rate": 2.125e-06,
+      "loss": 0.2343,
+      "step": 3150
+    },
+    {
+      "epoch": 15.755,
+      "grad_norm": 4.268637657165527,
+      "learning_rate": 2.1225e-06,
+      "loss": 0.3344,
+      "step": 3151
+    },
+    {
+      "epoch": 15.76,
+      "grad_norm": 6.2913408279418945,
+      "learning_rate": 2.12e-06,
+      "loss": 0.2998,
+      "step": 3152
+    },
+    {
+      "epoch": 15.765,
+      "grad_norm": 5.333639144897461,
+      "learning_rate": 2.1175e-06,
+      "loss": 0.5013,
+      "step": 3153
+    },
+    {
+      "epoch": 15.77,
+      "grad_norm": 4.938657760620117,
+      "learning_rate": 2.115e-06,
+      "loss": 0.2919,
+      "step": 3154
+    },
+    {
+      "epoch": 15.775,
+      "grad_norm": 4.009632110595703,
+      "learning_rate": 2.1125e-06,
+      "loss": 0.285,
+      "step": 3155
+    },
+    {
+      "epoch": 15.78,
+      "grad_norm": 649.7720336914062,
+      "learning_rate": 2.11e-06,
+      "loss": 1.6306,
+      "step": 3156
+    },
+    {
+      "epoch": 15.785,
+      "grad_norm": 2.7265512943267822,
+      "learning_rate": 2.1075e-06,
+      "loss": 0.2378,
+      "step": 3157
+    },
+    {
+      "epoch": 15.79,
+      "grad_norm": 4.267139434814453,
+      "learning_rate": 2.105e-06,
+      "loss": 0.324,
+      "step": 3158
+    },
+    {
+      "epoch": 15.795,
+      "grad_norm": 4.12963342666626,
+      "learning_rate": 2.1025e-06,
+      "loss": 0.231,
+      "step": 3159
+    },
+    {
+      "epoch": 15.8,
+      "grad_norm": 4.0534515380859375,
+      "learning_rate": 2.1000000000000002e-06,
+      "loss": 0.3931,
+      "step": 3160
+    },
+    {
+      "epoch": 15.805,
+      "grad_norm": 4.073664665222168,
+      "learning_rate": 2.0975000000000002e-06,
+      "loss": 0.4782,
+      "step": 3161
+    },
+    {
+      "epoch": 15.81,
+      "grad_norm": 4.429361820220947,
+      "learning_rate": 2.0950000000000003e-06,
+      "loss": 0.2134,
+      "step": 3162
+    },
+    {
+      "epoch": 15.815,
+      "grad_norm": 4.311758995056152,
+      "learning_rate": 2.0925000000000003e-06,
+      "loss": 0.2692,
+      "step": 3163
+    },
+    {
+      "epoch": 15.82,
+      "grad_norm": 4.750184059143066,
+      "learning_rate": 2.09e-06,
+      "loss": 0.2267,
+      "step": 3164
+    },
+    {
+      "epoch": 15.825,
+      "grad_norm": 2.8843016624450684,
+      "learning_rate": 2.0875e-06,
+      "loss": 0.0717,
+      "step": 3165
+    },
+    {
+      "epoch": 15.83,
+      "grad_norm": 2.542543888092041,
+      "learning_rate": 2.085e-06,
+      "loss": 0.1294,
+      "step": 3166
+    },
+    {
+      "epoch": 15.835,
+      "grad_norm": 1.7800548076629639,
+      "learning_rate": 2.0825e-06,
+      "loss": 0.0826,
+      "step": 3167
+    },
+    {
+      "epoch": 15.84,
+      "grad_norm": 4.715466499328613,
+      "learning_rate": 2.08e-06,
+      "loss": 0.4155,
+      "step": 3168
+    },
+    {
+      "epoch": 15.845,
+      "grad_norm": 7.1504011154174805,
+      "learning_rate": 2.0775e-06,
+      "loss": 0.3617,
+      "step": 3169
+    },
+    {
+      "epoch": 15.85,
+      "grad_norm": 3.673304557800293,
+      "learning_rate": 2.075e-06,
+      "loss": 0.2535,
+      "step": 3170
+    },
+    {
+      "epoch": 15.855,
+      "grad_norm": 6.388810634613037,
+      "learning_rate": 2.0725e-06,
+      "loss": 0.3236,
+      "step": 3171
+    },
+    {
+      "epoch": 15.86,
+      "grad_norm": 6.968552589416504,
+      "learning_rate": 2.07e-06,
+      "loss": 0.5439,
+      "step": 3172
+    },
+    {
+      "epoch": 15.865,
+      "grad_norm": 4.365705490112305,
+      "learning_rate": 2.0675e-06,
+      "loss": 0.152,
+      "step": 3173
+    },
+    {
+      "epoch": 15.87,
+      "grad_norm": 3.60847806930542,
+      "learning_rate": 2.065e-06,
+      "loss": 0.4294,
+      "step": 3174
+    },
+    {
+      "epoch": 15.875,
+      "grad_norm": 4.026469707489014,
+      "learning_rate": 2.0625e-06,
+      "loss": 0.1949,
+      "step": 3175
+    },
+    {
+      "epoch": 15.88,
+      "grad_norm": 3.5806896686553955,
+      "learning_rate": 2.06e-06,
+      "loss": 0.3364,
+      "step": 3176
+    },
+    {
+      "epoch": 15.885,
+      "grad_norm": 3.6777052879333496,
+      "learning_rate": 2.0575e-06,
+      "loss": 0.1691,
+      "step": 3177
+    },
+    {
+      "epoch": 15.89,
+      "grad_norm": 3.401668071746826,
+      "learning_rate": 2.0550000000000002e-06,
+      "loss": 0.5726,
+      "step": 3178
+    },
+    {
+      "epoch": 15.895,
+      "grad_norm": 5.075344085693359,
+      "learning_rate": 2.0525000000000003e-06,
+      "loss": 0.2522,
+      "step": 3179
+    },
+    {
+      "epoch": 15.9,
+      "grad_norm": 9.302205085754395,
+      "learning_rate": 2.05e-06,
+      "loss": 0.4768,
+      "step": 3180
+    },
+    {
+      "epoch": 15.905,
+      "grad_norm": 4.4560956954956055,
+      "learning_rate": 2.0475e-06,
+      "loss": 0.2954,
+      "step": 3181
+    },
+    {
+      "epoch": 15.91,
+      "grad_norm": 5.434938430786133,
+      "learning_rate": 2.045e-06,
+      "loss": 0.1552,
+      "step": 3182
+    },
+    {
+      "epoch": 15.915,
+      "grad_norm": 5.970384120941162,
+      "learning_rate": 2.0425e-06,
+      "loss": 0.5503,
+      "step": 3183
+    },
+    {
+      "epoch": 15.92,
+      "grad_norm": 6.739590644836426,
+      "learning_rate": 2.04e-06,
+      "loss": 0.4471,
+      "step": 3184
+    },
+    {
+      "epoch": 15.925,
+      "grad_norm": 18.867116928100586,
+      "learning_rate": 2.0375e-06,
+      "loss": 0.4856,
+      "step": 3185
+    },
+    {
+      "epoch": 15.93,
+      "grad_norm": 6.885401725769043,
+      "learning_rate": 2.035e-06,
+      "loss": 0.8085,
+      "step": 3186
+    },
+    {
+      "epoch": 15.935,
+      "grad_norm": 3.671773672103882,
+      "learning_rate": 2.0325e-06,
+      "loss": 0.2403,
+      "step": 3187
+    },
+    {
+      "epoch": 15.94,
+      "grad_norm": 31.94533920288086,
+      "learning_rate": 2.0300000000000005e-06,
+      "loss": 0.3865,
+      "step": 3188
+    },
+    {
+      "epoch": 15.945,
+      "grad_norm": 6.052237033843994,
+      "learning_rate": 2.0275000000000005e-06,
+      "loss": 0.6851,
+      "step": 3189
+    },
+    {
+      "epoch": 15.95,
+      "grad_norm": 3.690936326980591,
+      "learning_rate": 2.025e-06,
+      "loss": 0.1483,
+      "step": 3190
+    },
+    {
+      "epoch": 15.955,
+      "grad_norm": 4.0976128578186035,
+      "learning_rate": 2.0225e-06,
+      "loss": 0.2715,
+      "step": 3191
+    },
+    {
+      "epoch": 15.96,
+      "grad_norm": 3.018514394760132,
+      "learning_rate": 2.02e-06,
+      "loss": 0.3517,
+      "step": 3192
+    },
+    {
+      "epoch": 15.965,
+      "grad_norm": 3.306885242462158,
+      "learning_rate": 2.0175e-06,
+      "loss": 0.1719,
+      "step": 3193
+    },
+    {
+      "epoch": 15.97,
+      "grad_norm": 5.869591236114502,
+      "learning_rate": 2.015e-06,
+      "loss": 0.1731,
+      "step": 3194
+    },
+    {
+      "epoch": 15.975,
+      "grad_norm": 6.387131214141846,
+      "learning_rate": 2.0125000000000002e-06,
+      "loss": 0.5182,
+      "step": 3195
+    },
+    {
+      "epoch": 15.98,
+      "grad_norm": 4.7032790184021,
+      "learning_rate": 2.0100000000000002e-06,
+      "loss": 0.3378,
+      "step": 3196
+    },
+    {
+      "epoch": 15.985,
+      "grad_norm": 3.7517316341400146,
+      "learning_rate": 2.0075000000000003e-06,
+      "loss": 0.284,
+      "step": 3197
+    },
+    {
+      "epoch": 15.99,
+      "grad_norm": 13.434398651123047,
+      "learning_rate": 2.0050000000000003e-06,
+      "loss": 0.2432,
+      "step": 3198
+    },
+    {
+      "epoch": 15.995,
+      "grad_norm": 4.905867576599121,
+      "learning_rate": 2.0025000000000003e-06,
+      "loss": 0.2801,
+      "step": 3199
+    },
+    {
+      "epoch": 16.0,
+      "grad_norm": 3.7303099632263184,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.232,
+      "step": 3200
+    },
+    {
+      "epoch": 16.0,
+      "eval_loss": 0.19819729030132294,
+      "eval_runtime": 15.2145,
+      "eval_samples_per_second": 52.45,
+      "eval_steps_per_second": 6.573,
+      "step": 3200
+    },
+    {
+      "epoch": 16.005,
+      "grad_norm": 3.4478981494903564,
+      "learning_rate": 1.9975000000000004e-06,
+      "loss": 0.3068,
+      "step": 3201
+    },
+    {
+      "epoch": 16.01,
+      "grad_norm": 3.8169233798980713,
+      "learning_rate": 1.9950000000000004e-06,
+      "loss": 0.4189,
+      "step": 3202
+    },
+    {
+      "epoch": 16.015,
+      "grad_norm": 10.544240951538086,
+      "learning_rate": 1.9925000000000004e-06,
+      "loss": 0.1757,
+      "step": 3203
+    },
+    {
+      "epoch": 16.02,
+      "grad_norm": 6.238424777984619,
+      "learning_rate": 1.9900000000000004e-06,
+      "loss": 0.2056,
+      "step": 3204
+    },
+    {
+      "epoch": 16.025,
+      "grad_norm": 7.078701972961426,
+      "learning_rate": 1.9875000000000005e-06,
+      "loss": 0.274,
+      "step": 3205
+    },
+    {
+      "epoch": 16.03,
+      "grad_norm": 4.549793243408203,
+      "learning_rate": 1.985e-06,
+      "loss": 0.289,
+      "step": 3206
+    },
+    {
+      "epoch": 16.035,
+      "grad_norm": 4.049411296844482,
+      "learning_rate": 1.9825e-06,
+      "loss": 0.3597,
+      "step": 3207
+    },
+    {
+      "epoch": 16.04,
+      "grad_norm": 6.2245283126831055,
+      "learning_rate": 1.98e-06,
+      "loss": 0.5751,
+      "step": 3208
+    },
+    {
+      "epoch": 16.045,
+      "grad_norm": 7.372754096984863,
+      "learning_rate": 1.9775e-06,
+      "loss": 0.5231,
+      "step": 3209
+    },
+    {
+      "epoch": 16.05,
+      "grad_norm": 2.690129041671753,
+      "learning_rate": 1.975e-06,
+      "loss": 0.4385,
+      "step": 3210
+    },
+    {
+      "epoch": 16.055,
+      "grad_norm": 4.057859897613525,
+      "learning_rate": 1.9725e-06,
+      "loss": 0.2854,
+      "step": 3211
+    },
+    {
+      "epoch": 16.06,
+      "grad_norm": 3.7713522911071777,
+      "learning_rate": 1.97e-06,
+      "loss": 0.1896,
+      "step": 3212
+    },
+    {
+      "epoch": 16.065,
+      "grad_norm": 3.207489252090454,
+      "learning_rate": 1.9675000000000002e-06,
+      "loss": 0.2833,
+      "step": 3213
+    },
+    {
+      "epoch": 16.07,
+      "grad_norm": 3.238175868988037,
+      "learning_rate": 1.9650000000000002e-06,
+      "loss": 0.2207,
+      "step": 3214
+    },
+    {
+      "epoch": 16.075,
+      "grad_norm": 2.2949297428131104,
+      "learning_rate": 1.9625000000000003e-06,
+      "loss": 0.2725,
+      "step": 3215
+    },
+    {
+      "epoch": 16.08,
+      "grad_norm": 4.465222358703613,
+      "learning_rate": 1.9600000000000003e-06,
+      "loss": 0.3959,
+      "step": 3216
+    },
+    {
+      "epoch": 16.085,
+      "grad_norm": 3.3759946823120117,
+      "learning_rate": 1.9575000000000003e-06,
+      "loss": 0.1093,
+      "step": 3217
+    },
+    {
+      "epoch": 16.09,
+      "grad_norm": 3.6745927333831787,
+      "learning_rate": 1.9550000000000003e-06,
+      "loss": 0.219,
+      "step": 3218
+    },
+    {
+      "epoch": 16.095,
+      "grad_norm": 5.156643390655518,
+      "learning_rate": 1.9525000000000004e-06,
+      "loss": 0.3489,
+      "step": 3219
+    },
+    {
+      "epoch": 16.1,
+      "grad_norm": 7.829514980316162,
+      "learning_rate": 1.9500000000000004e-06,
+      "loss": 0.4058,
+      "step": 3220
+    },
+    {
+      "epoch": 16.105,
+      "grad_norm": 1.9332938194274902,
+      "learning_rate": 1.9475000000000004e-06,
+      "loss": 0.0445,
+      "step": 3221
+    },
+    {
+      "epoch": 16.11,
+      "grad_norm": 3.5957601070404053,
+      "learning_rate": 1.945e-06,
+      "loss": 0.1636,
+      "step": 3222
+    },
+    {
+      "epoch": 16.115,
+      "grad_norm": 5.341385364532471,
+      "learning_rate": 1.9425e-06,
+      "loss": 0.4766,
+      "step": 3223
+    },
+    {
+      "epoch": 16.12,
+      "grad_norm": 5.976744174957275,
+      "learning_rate": 1.94e-06,
+      "loss": 0.2623,
+      "step": 3224
+    },
+    {
+      "epoch": 16.125,
+      "grad_norm": 8.002298355102539,
+      "learning_rate": 1.9375e-06,
+      "loss": 0.6703,
+      "step": 3225
+    },
+    {
+      "epoch": 16.13,
+      "grad_norm": 6.137474536895752,
+      "learning_rate": 1.935e-06,
+      "loss": 0.3186,
+      "step": 3226
+    },
+    {
+      "epoch": 16.135,
+      "grad_norm": 5.355105400085449,
+      "learning_rate": 1.9325e-06,
+      "loss": 0.3742,
+      "step": 3227
+    },
+    {
+      "epoch": 16.14,
+      "grad_norm": 2.5555434226989746,
+      "learning_rate": 1.93e-06,
+      "loss": 0.3932,
+      "step": 3228
+    },
+    {
+      "epoch": 16.145,
+      "grad_norm": 3.583042621612549,
+      "learning_rate": 1.9275e-06,
+      "loss": 0.3242,
+      "step": 3229
+    },
+    {
+      "epoch": 16.15,
+      "grad_norm": 3.4275858402252197,
+      "learning_rate": 1.925e-06,
+      "loss": 0.1375,
+      "step": 3230
+    },
+    {
+      "epoch": 16.155,
+      "grad_norm": 3.510256767272949,
+      "learning_rate": 1.9225000000000002e-06,
+      "loss": 0.215,
+      "step": 3231
+    },
+    {
+      "epoch": 16.16,
+      "grad_norm": 4.154227256774902,
+      "learning_rate": 1.9200000000000003e-06,
+      "loss": 0.2326,
+      "step": 3232
+    },
+    {
+      "epoch": 16.165,
+      "grad_norm": 6.276437759399414,
+      "learning_rate": 1.9175000000000003e-06,
+      "loss": 0.3589,
+      "step": 3233
+    },
+    {
+      "epoch": 16.17,
+      "grad_norm": 5.121100425720215,
+      "learning_rate": 1.9150000000000003e-06,
+      "loss": 0.4361,
+      "step": 3234
+    },
+    {
+      "epoch": 16.175,
+      "grad_norm": 6.244390964508057,
+      "learning_rate": 1.9125000000000003e-06,
+      "loss": 0.4546,
+      "step": 3235
+    },
+    {
+      "epoch": 16.18,
+      "grad_norm": 9.673368453979492,
+      "learning_rate": 1.9100000000000003e-06,
+      "loss": 0.1205,
+      "step": 3236
+    },
+    {
+      "epoch": 16.185,
+      "grad_norm": 3.4378128051757812,
+      "learning_rate": 1.9075000000000004e-06,
+      "loss": 0.1699,
+      "step": 3237
+    },
+    {
+      "epoch": 16.19,
+      "grad_norm": 1.8403854370117188,
+      "learning_rate": 1.9050000000000002e-06,
+      "loss": 0.0418,
+      "step": 3238
+    },
+    {
+      "epoch": 16.195,
+      "grad_norm": 4.064254283905029,
+      "learning_rate": 1.9025000000000002e-06,
+      "loss": 0.3514,
+      "step": 3239
+    },
+    {
+      "epoch": 16.2,
+      "grad_norm": 4.841102123260498,
+      "learning_rate": 1.9000000000000002e-06,
+      "loss": 0.5217,
+      "step": 3240
+    },
+    {
+      "epoch": 16.205,
+      "grad_norm": 7.344009876251221,
+      "learning_rate": 1.8975000000000003e-06,
+      "loss": 0.4896,
+      "step": 3241
+    },
+    {
+      "epoch": 16.21,
+      "grad_norm": 2.723459482192993,
+      "learning_rate": 1.895e-06,
+      "loss": 0.1675,
+      "step": 3242
+    },
+    {
+      "epoch": 16.215,
+      "grad_norm": 4.058963775634766,
+      "learning_rate": 1.8925e-06,
+      "loss": 0.1549,
+      "step": 3243
+    },
+    {
+      "epoch": 16.22,
+      "grad_norm": 9.402939796447754,
+      "learning_rate": 1.8900000000000001e-06,
+      "loss": 0.4886,
+      "step": 3244
+    },
+    {
+      "epoch": 16.225,
+      "grad_norm": 7.74470853805542,
+      "learning_rate": 1.8875000000000001e-06,
+      "loss": 0.3842,
+      "step": 3245
+    },
+    {
+      "epoch": 16.23,
+      "grad_norm": 3.7326676845550537,
+      "learning_rate": 1.8850000000000002e-06,
+      "loss": 0.1583,
+      "step": 3246
+    },
+    {
+      "epoch": 16.235,
+      "grad_norm": 2.4392449855804443,
+      "learning_rate": 1.8825000000000002e-06,
+      "loss": 0.0867,
+      "step": 3247
+    },
+    {
+      "epoch": 16.24,
+      "grad_norm": 3.8829855918884277,
+      "learning_rate": 1.8800000000000002e-06,
+      "loss": 0.5797,
+      "step": 3248
+    },
+    {
+      "epoch": 16.245,
+      "grad_norm": 7.10105562210083,
+      "learning_rate": 1.8775000000000002e-06,
+      "loss": 0.4636,
+      "step": 3249
+    },
+    {
+      "epoch": 16.25,
+      "grad_norm": 3.2992286682128906,
+      "learning_rate": 1.8750000000000003e-06,
+      "loss": 0.2976,
+      "step": 3250
+    },
+    {
+      "epoch": 16.255,
+      "grad_norm": 4.053872108459473,
+      "learning_rate": 1.8725e-06,
+      "loss": 0.318,
+      "step": 3251
+    },
+    {
+      "epoch": 16.26,
+      "grad_norm": 10.315462112426758,
+      "learning_rate": 1.87e-06,
+      "loss": 0.6008,
+      "step": 3252
+    },
+    {
+      "epoch": 16.265,
+      "grad_norm": 2.4355626106262207,
+      "learning_rate": 1.8675000000000001e-06,
+      "loss": 0.0663,
+      "step": 3253
+    },
+    {
+      "epoch": 16.27,
+      "grad_norm": 4.35316276550293,
+      "learning_rate": 1.8650000000000001e-06,
+      "loss": 0.1977,
+      "step": 3254
+    },
+    {
+      "epoch": 16.275,
+      "grad_norm": 4.498222827911377,
+      "learning_rate": 1.8625000000000002e-06,
+      "loss": 0.2784,
+      "step": 3255
+    },
+    {
+      "epoch": 16.28,
+      "grad_norm": 3.8926422595977783,
+      "learning_rate": 1.8600000000000002e-06,
+      "loss": 0.2489,
+      "step": 3256
+    },
+    {
+      "epoch": 16.285,
+      "grad_norm": 3.9124815464019775,
+      "learning_rate": 1.8575000000000002e-06,
+      "loss": 0.2819,
+      "step": 3257
+    },
+    {
+      "epoch": 16.29,
+      "grad_norm": 3.3898565769195557,
+      "learning_rate": 1.8550000000000002e-06,
+      "loss": 0.2189,
+      "step": 3258
+    },
+    {
+      "epoch": 16.295,
+      "grad_norm": 2.8204894065856934,
+      "learning_rate": 1.8525e-06,
+      "loss": 0.2873,
+      "step": 3259
+    },
+    {
+      "epoch": 16.3,
+      "grad_norm": 2.3646838665008545,
+      "learning_rate": 1.85e-06,
+      "loss": 0.2251,
+      "step": 3260
+    },
+    {
+      "epoch": 16.305,
+      "grad_norm": 4.555030822753906,
+      "learning_rate": 1.8475e-06,
+      "loss": 0.1594,
+      "step": 3261
+    },
+    {
+      "epoch": 16.31,
+      "grad_norm": 5.801641941070557,
+      "learning_rate": 1.8450000000000001e-06,
+      "loss": 0.2548,
+      "step": 3262
+    },
+    {
+      "epoch": 16.315,
+      "grad_norm": 2.6225337982177734,
+      "learning_rate": 1.8425000000000001e-06,
+      "loss": 0.1755,
+      "step": 3263
+    },
+    {
+      "epoch": 16.32,
+      "grad_norm": 7.140406608581543,
+      "learning_rate": 1.8400000000000002e-06,
+      "loss": 0.6067,
+      "step": 3264
+    },
+    {
+      "epoch": 16.325,
+      "grad_norm": 3.0261030197143555,
+      "learning_rate": 1.8375000000000002e-06,
+      "loss": 0.1083,
+      "step": 3265
+    },
+    {
+      "epoch": 16.33,
+      "grad_norm": 3.718174457550049,
+      "learning_rate": 1.8350000000000002e-06,
+      "loss": 0.2378,
+      "step": 3266
+    },
+    {
+      "epoch": 16.335,
+      "grad_norm": 2.781790018081665,
+      "learning_rate": 1.8325e-06,
+      "loss": 0.4304,
+      "step": 3267
+    },
+    {
+      "epoch": 16.34,
+      "grad_norm": 3.2054054737091064,
+      "learning_rate": 1.83e-06,
+      "loss": 0.3259,
+      "step": 3268
+    },
+    {
+      "epoch": 16.345,
+      "grad_norm": 4.927772045135498,
+      "learning_rate": 1.8275e-06,
+      "loss": 0.3293,
+      "step": 3269
+    },
+    {
+      "epoch": 16.35,
+      "grad_norm": 2.9013285636901855,
+      "learning_rate": 1.825e-06,
+      "loss": 0.3608,
+      "step": 3270
+    },
+    {
+      "epoch": 16.355,
+      "grad_norm": 2.9312877655029297,
+      "learning_rate": 1.8225000000000001e-06,
+      "loss": 0.2192,
+      "step": 3271
+    },
+    {
+      "epoch": 16.36,
+      "grad_norm": 4.376914978027344,
+      "learning_rate": 1.8200000000000002e-06,
+      "loss": 0.2068,
+      "step": 3272
+    },
+    {
+      "epoch": 16.365,
+      "grad_norm": 8.309657096862793,
+      "learning_rate": 1.8175000000000002e-06,
+      "loss": 0.6735,
+      "step": 3273
+    },
+    {
+      "epoch": 16.37,
+      "grad_norm": 4.021854877471924,
+      "learning_rate": 1.8150000000000002e-06,
+      "loss": 0.3579,
+      "step": 3274
+    },
+    {
+      "epoch": 16.375,
+      "grad_norm": 20.36651039123535,
+      "learning_rate": 1.8125e-06,
+      "loss": 0.7959,
+      "step": 3275
+    },
+    {
+      "epoch": 16.38,
+      "grad_norm": 5.842319488525391,
+      "learning_rate": 1.81e-06,
+      "loss": 0.2651,
+      "step": 3276
+    },
+    {
+      "epoch": 16.385,
+      "grad_norm": 5.426783561706543,
+      "learning_rate": 1.8075e-06,
+      "loss": 0.291,
+      "step": 3277
+    },
+    {
+      "epoch": 16.39,
+      "grad_norm": 4.858299732208252,
+      "learning_rate": 1.805e-06,
+      "loss": 0.2191,
+      "step": 3278
+    },
+    {
+      "epoch": 16.395,
+      "grad_norm": 2.2117466926574707,
+      "learning_rate": 1.8025000000000001e-06,
+      "loss": 0.0803,
+      "step": 3279
+    },
+    {
+      "epoch": 16.4,
+      "grad_norm": 4.840840816497803,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 0.1693,
+      "step": 3280
+    },
+    {
+      "epoch": 16.405,
+      "grad_norm": 5.898733139038086,
+      "learning_rate": 1.7975000000000002e-06,
+      "loss": 0.2937,
+      "step": 3281
+    },
+    {
+      "epoch": 16.41,
+      "grad_norm": 4.352290153503418,
+      "learning_rate": 1.7950000000000002e-06,
+      "loss": 0.2044,
+      "step": 3282
+    },
+    {
+      "epoch": 16.415,
+      "grad_norm": 3.455841064453125,
+      "learning_rate": 1.7925e-06,
+      "loss": 0.1294,
+      "step": 3283
+    },
+    {
+      "epoch": 16.42,
+      "grad_norm": 4.711461544036865,
+      "learning_rate": 1.79e-06,
+      "loss": 0.2859,
+      "step": 3284
+    },
+    {
+      "epoch": 16.425,
+      "grad_norm": 5.745739459991455,
+      "learning_rate": 1.7875e-06,
+      "loss": 0.5767,
+      "step": 3285
+    },
+    {
+      "epoch": 16.43,
+      "grad_norm": 5.111008644104004,
+      "learning_rate": 1.785e-06,
+      "loss": 0.4217,
+      "step": 3286
+    },
+    {
+      "epoch": 16.435,
+      "grad_norm": 5.080211162567139,
+      "learning_rate": 1.7825e-06,
+      "loss": 0.5622,
+      "step": 3287
+    },
+    {
+      "epoch": 16.44,
+      "grad_norm": 4.839994430541992,
+      "learning_rate": 1.7800000000000001e-06,
+      "loss": 0.2202,
+      "step": 3288
+    },
+    {
+      "epoch": 16.445,
+      "grad_norm": 3.5078511238098145,
+      "learning_rate": 1.7775000000000001e-06,
+      "loss": 0.2921,
+      "step": 3289
+    },
+    {
+      "epoch": 16.45,
+      "grad_norm": 6.41483736038208,
+      "learning_rate": 1.7750000000000002e-06,
+      "loss": 0.5494,
+      "step": 3290
+    },
+    {
+      "epoch": 16.455,
+      "grad_norm": 2.2354838848114014,
+      "learning_rate": 1.7725e-06,
+      "loss": 0.0564,
+      "step": 3291
+    },
+    {
+      "epoch": 16.46,
+      "grad_norm": 5.539938926696777,
+      "learning_rate": 1.77e-06,
+      "loss": 0.7224,
+      "step": 3292
+    },
+    {
+      "epoch": 16.465,
+      "grad_norm": 2.223759651184082,
+      "learning_rate": 1.7675e-06,
+      "loss": 0.3181,
+      "step": 3293
+    },
+    {
+      "epoch": 16.47,
+      "grad_norm": 2.3818132877349854,
+      "learning_rate": 1.765e-06,
+      "loss": 0.1386,
+      "step": 3294
+    },
+    {
+      "epoch": 16.475,
+      "grad_norm": 3.7334539890289307,
+      "learning_rate": 1.7625e-06,
+      "loss": 0.2891,
+      "step": 3295
+    },
+    {
+      "epoch": 16.48,
+      "grad_norm": 4.152212142944336,
+      "learning_rate": 1.76e-06,
+      "loss": 0.2158,
+      "step": 3296
+    },
+    {
+      "epoch": 16.485,
+      "grad_norm": 5.936916828155518,
+      "learning_rate": 1.7575000000000001e-06,
+      "loss": 0.5492,
+      "step": 3297
+    },
+    {
+      "epoch": 16.49,
+      "grad_norm": 1.5779609680175781,
+      "learning_rate": 1.7550000000000001e-06,
+      "loss": 0.0789,
+      "step": 3298
+    },
+    {
+      "epoch": 16.495,
+      "grad_norm": 5.098733901977539,
+      "learning_rate": 1.7525e-06,
+      "loss": 0.4965,
+      "step": 3299
+    },
+    {
+      "epoch": 16.5,
+      "grad_norm": 6.155126094818115,
+      "learning_rate": 1.75e-06,
+      "loss": 0.4724,
+      "step": 3300
+    },
+    {
+      "epoch": 16.505,
+      "grad_norm": 2.998126745223999,
+      "learning_rate": 1.7475e-06,
+      "loss": 0.131,
+      "step": 3301
+    },
+    {
+      "epoch": 16.51,
+      "grad_norm": 4.56244421005249,
+      "learning_rate": 1.745e-06,
+      "loss": 0.308,
+      "step": 3302
+    },
+    {
+      "epoch": 16.515,
+      "grad_norm": 4.542159080505371,
+      "learning_rate": 1.7425e-06,
+      "loss": 0.2056,
+      "step": 3303
+    },
+    {
+      "epoch": 16.52,
+      "grad_norm": 4.396841049194336,
+      "learning_rate": 1.74e-06,
+      "loss": 0.2399,
+      "step": 3304
+    },
+    {
+      "epoch": 16.525,
+      "grad_norm": 4.553340911865234,
+      "learning_rate": 1.7375e-06,
+      "loss": 0.3893,
+      "step": 3305
+    },
+    {
+      "epoch": 16.53,
+      "grad_norm": 18.358760833740234,
+      "learning_rate": 1.7350000000000001e-06,
+      "loss": 0.8999,
+      "step": 3306
+    },
+    {
+      "epoch": 16.535,
+      "grad_norm": 2.1946053504943848,
+      "learning_rate": 1.7325e-06,
+      "loss": 0.0388,
+      "step": 3307
+    },
+    {
+      "epoch": 16.54,
+      "grad_norm": 5.189298152923584,
+      "learning_rate": 1.73e-06,
+      "loss": 0.2969,
+      "step": 3308
+    },
+    {
+      "epoch": 16.545,
+      "grad_norm": 5.030030250549316,
+      "learning_rate": 1.7275e-06,
+      "loss": 0.3203,
+      "step": 3309
+    },
+    {
+      "epoch": 16.55,
+      "grad_norm": 4.398233890533447,
+      "learning_rate": 1.725e-06,
+      "loss": 0.242,
+      "step": 3310
+    },
+    {
+      "epoch": 16.555,
+      "grad_norm": 3.4126675128936768,
+      "learning_rate": 1.7225e-06,
+      "loss": 0.157,
+      "step": 3311
+    },
+    {
+      "epoch": 16.56,
+      "grad_norm": 4.920454502105713,
+      "learning_rate": 1.72e-06,
+      "loss": 0.151,
+      "step": 3312
+    },
+    {
+      "epoch": 16.565,
+      "grad_norm": 8.223849296569824,
+      "learning_rate": 1.7175000000000003e-06,
+      "loss": 0.4392,
+      "step": 3313
+    },
+    {
+      "epoch": 16.57,
+      "grad_norm": 6.308965682983398,
+      "learning_rate": 1.7150000000000003e-06,
+      "loss": 0.4101,
+      "step": 3314
+    },
+    {
+      "epoch": 16.575,
+      "grad_norm": 10.77561092376709,
+      "learning_rate": 1.7125000000000003e-06,
+      "loss": 0.2675,
+      "step": 3315
+    },
+    {
+      "epoch": 16.58,
+      "grad_norm": 3.767138957977295,
+      "learning_rate": 1.7100000000000004e-06,
+      "loss": 0.1418,
+      "step": 3316
+    },
+    {
+      "epoch": 16.585,
+      "grad_norm": 3.736222267150879,
+      "learning_rate": 1.7075000000000002e-06,
+      "loss": 0.1759,
+      "step": 3317
+    },
+    {
+      "epoch": 16.59,
+      "grad_norm": 4.276198387145996,
+      "learning_rate": 1.7050000000000002e-06,
+      "loss": 0.2882,
+      "step": 3318
+    },
+    {
+      "epoch": 16.595,
+      "grad_norm": 4.637908935546875,
+      "learning_rate": 1.7025000000000002e-06,
+      "loss": 0.5841,
+      "step": 3319
+    },
+    {
+      "epoch": 16.6,
+      "grad_norm": 5.584635257720947,
+      "learning_rate": 1.7000000000000002e-06,
+      "loss": 0.3764,
+      "step": 3320
+    },
+    {
+      "epoch": 16.605,
+      "grad_norm": 6.699474334716797,
+      "learning_rate": 1.6975000000000003e-06,
+      "loss": 0.3461,
+      "step": 3321
+    },
+    {
+      "epoch": 16.61,
+      "grad_norm": 4.561931133270264,
+      "learning_rate": 1.6950000000000003e-06,
+      "loss": 0.511,
+      "step": 3322
+    },
+    {
+      "epoch": 16.615,
+      "grad_norm": 3.841766357421875,
+      "learning_rate": 1.6925000000000003e-06,
+      "loss": 0.6992,
+      "step": 3323
+    },
+    {
+      "epoch": 16.62,
+      "grad_norm": 4.760002136230469,
+      "learning_rate": 1.6900000000000003e-06,
+      "loss": 0.3915,
+      "step": 3324
+    },
+    {
+      "epoch": 16.625,
+      "grad_norm": 3.47341251373291,
+      "learning_rate": 1.6875000000000001e-06,
+      "loss": 0.1571,
+      "step": 3325
+    },
+    {
+      "epoch": 16.63,
+      "grad_norm": 3.881911516189575,
+      "learning_rate": 1.6850000000000002e-06,
+      "loss": 0.2028,
+      "step": 3326
+    },
+    {
+      "epoch": 16.635,
+      "grad_norm": 3.664778470993042,
+      "learning_rate": 1.6825000000000002e-06,
+      "loss": 0.4163,
+      "step": 3327
+    },
+    {
+      "epoch": 16.64,
+      "grad_norm": 4.0560994148254395,
+      "learning_rate": 1.6800000000000002e-06,
+      "loss": 0.2066,
+      "step": 3328
+    },
+    {
+      "epoch": 16.645,
+      "grad_norm": 2.580767869949341,
+      "learning_rate": 1.6775000000000002e-06,
+      "loss": 0.2531,
+      "step": 3329
+    },
+    {
+      "epoch": 16.65,
+      "grad_norm": 3.780874490737915,
+      "learning_rate": 1.6750000000000003e-06,
+      "loss": 0.2753,
+      "step": 3330
+    },
+    {
+      "epoch": 16.655,
+      "grad_norm": 9.407083511352539,
+      "learning_rate": 1.6725000000000003e-06,
+      "loss": 0.3794,
+      "step": 3331
+    },
+    {
+      "epoch": 16.66,
+      "grad_norm": 2.6181418895721436,
+      "learning_rate": 1.6700000000000003e-06,
+      "loss": 0.0466,
+      "step": 3332
+    },
+    {
+      "epoch": 16.665,
+      "grad_norm": 3.4600296020507812,
+      "learning_rate": 1.6675000000000001e-06,
+      "loss": 0.2892,
+      "step": 3333
+    },
+    {
+      "epoch": 16.67,
+      "grad_norm": 7.713902950286865,
+      "learning_rate": 1.6650000000000002e-06,
+      "loss": 0.4739,
+      "step": 3334
+    },
+    {
+      "epoch": 16.675,
+      "grad_norm": 4.81476354598999,
+      "learning_rate": 1.6625000000000002e-06,
+      "loss": 0.2816,
+      "step": 3335
+    },
+    {
+      "epoch": 16.68,
+      "grad_norm": 6.36476469039917,
+      "learning_rate": 1.6600000000000002e-06,
+      "loss": 0.4451,
+      "step": 3336
+    },
+    {
+      "epoch": 16.685,
+      "grad_norm": 3.0283100605010986,
+      "learning_rate": 1.6575000000000002e-06,
+      "loss": 0.2746,
+      "step": 3337
+    },
+    {
+      "epoch": 16.69,
+      "grad_norm": 3.018043041229248,
+      "learning_rate": 1.6550000000000002e-06,
+      "loss": 0.2169,
+      "step": 3338
+    },
+    {
+      "epoch": 16.695,
+      "grad_norm": 4.029423713684082,
+      "learning_rate": 1.6525000000000003e-06,
+      "loss": 0.2814,
+      "step": 3339
+    },
+    {
+      "epoch": 16.7,
+      "grad_norm": 2.061889886856079,
+      "learning_rate": 1.6500000000000003e-06,
+      "loss": 0.0487,
+      "step": 3340
+    },
+    {
+      "epoch": 16.705,
+      "grad_norm": 3.510653018951416,
+      "learning_rate": 1.6475000000000001e-06,
+      "loss": 0.1834,
+      "step": 3341
+    },
+    {
+      "epoch": 16.71,
+      "grad_norm": 6.757889747619629,
+      "learning_rate": 1.6450000000000001e-06,
+      "loss": 0.5764,
+      "step": 3342
+    },
+    {
+      "epoch": 16.715,
+      "grad_norm": 8.3269624710083,
+      "learning_rate": 1.6425000000000002e-06,
+      "loss": 0.6033,
+      "step": 3343
+    },
+    {
+      "epoch": 16.72,
+      "grad_norm": 4.307374000549316,
+      "learning_rate": 1.6400000000000002e-06,
+      "loss": 0.1684,
+      "step": 3344
+    },
+    {
+      "epoch": 16.725,
+      "grad_norm": 4.898263931274414,
+      "learning_rate": 1.6375000000000002e-06,
+      "loss": 0.208,
+      "step": 3345
+    },
+    {
+      "epoch": 16.73,
+      "grad_norm": 2.75921368598938,
+      "learning_rate": 1.6350000000000002e-06,
+      "loss": 0.0856,
+      "step": 3346
+    },
+    {
+      "epoch": 16.735,
+      "grad_norm": 3.8106815814971924,
+      "learning_rate": 1.6325000000000003e-06,
+      "loss": 0.5824,
+      "step": 3347
+    },
+    {
+      "epoch": 16.74,
+      "grad_norm": 3.623408555984497,
+      "learning_rate": 1.6300000000000003e-06,
+      "loss": 0.3561,
+      "step": 3348
+    },
+    {
+      "epoch": 16.745,
+      "grad_norm": 3.177215814590454,
+      "learning_rate": 1.6275e-06,
+      "loss": 0.2628,
+      "step": 3349
+    },
+    {
+      "epoch": 16.75,
+      "grad_norm": 3.569906711578369,
+      "learning_rate": 1.6250000000000001e-06,
+      "loss": 0.2189,
+      "step": 3350
+    },
+    {
+      "epoch": 16.755,
+      "grad_norm": 5.109379291534424,
+      "learning_rate": 1.6225000000000001e-06,
+      "loss": 0.1733,
+      "step": 3351
+    },
+    {
+      "epoch": 16.76,
+      "grad_norm": 7.938911437988281,
+      "learning_rate": 1.6200000000000002e-06,
+      "loss": 0.7251,
+      "step": 3352
+    },
+    {
+      "epoch": 16.765,
+      "grad_norm": 2.427959442138672,
+      "learning_rate": 1.6175000000000002e-06,
+      "loss": 0.081,
+      "step": 3353
+    },
+    {
+      "epoch": 16.77,
+      "grad_norm": 3.8375320434570312,
+      "learning_rate": 1.6150000000000002e-06,
+      "loss": 0.2511,
+      "step": 3354
+    },
+    {
+      "epoch": 16.775,
+      "grad_norm": 5.430028438568115,
+      "learning_rate": 1.6125000000000002e-06,
+      "loss": 0.5144,
+      "step": 3355
+    },
+    {
+      "epoch": 16.78,
+      "grad_norm": 4.40625524520874,
+      "learning_rate": 1.6100000000000003e-06,
+      "loss": 0.3184,
+      "step": 3356
+    },
+    {
+      "epoch": 16.785,
+      "grad_norm": 9.8374605178833,
+      "learning_rate": 1.6075e-06,
+      "loss": 0.3697,
+      "step": 3357
+    },
+    {
+      "epoch": 16.79,
+      "grad_norm": 8.751962661743164,
+      "learning_rate": 1.605e-06,
+      "loss": 0.7771,
+      "step": 3358
+    },
+    {
+      "epoch": 16.795,
+      "grad_norm": 5.366896629333496,
+      "learning_rate": 1.6025000000000001e-06,
+      "loss": 0.5207,
+      "step": 3359
+    },
+    {
+      "epoch": 16.8,
+      "grad_norm": 2.922800064086914,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 0.1059,
+      "step": 3360
+    },
+    {
+      "epoch": 16.805,
+      "grad_norm": 2.9987196922302246,
+      "learning_rate": 1.5975000000000002e-06,
+      "loss": 0.3419,
+      "step": 3361
+    },
+    {
+      "epoch": 16.81,
+      "grad_norm": 5.300695419311523,
+      "learning_rate": 1.5950000000000002e-06,
+      "loss": 0.2414,
+      "step": 3362
+    },
+    {
+      "epoch": 16.815,
+      "grad_norm": 5.255833625793457,
+      "learning_rate": 1.5925000000000002e-06,
+      "loss": 0.2885,
+      "step": 3363
+    },
+    {
+      "epoch": 16.82,
+      "grad_norm": 3.1727163791656494,
+      "learning_rate": 1.5900000000000002e-06,
+      "loss": 0.1406,
+      "step": 3364
+    },
+    {
+      "epoch": 16.825,
+      "grad_norm": 4.986965179443359,
+      "learning_rate": 1.5875e-06,
+      "loss": 0.2192,
+      "step": 3365
+    },
+    {
+      "epoch": 16.83,
+      "grad_norm": 5.399299144744873,
+      "learning_rate": 1.585e-06,
+      "loss": 0.3828,
+      "step": 3366
+    },
+    {
+      "epoch": 16.835,
+      "grad_norm": 4.643341064453125,
+      "learning_rate": 1.5825e-06,
+      "loss": 0.5404,
+      "step": 3367
+    },
+    {
+      "epoch": 16.84,
+      "grad_norm": 4.139863014221191,
+      "learning_rate": 1.5800000000000001e-06,
+      "loss": 0.4206,
+      "step": 3368
+    },
+    {
+      "epoch": 16.845,
+      "grad_norm": 6.267672061920166,
+      "learning_rate": 1.5775000000000001e-06,
+      "loss": 0.5294,
+      "step": 3369
+    },
+    {
+      "epoch": 16.85,
+      "grad_norm": 2.6592345237731934,
+      "learning_rate": 1.5750000000000002e-06,
+      "loss": 0.3484,
+      "step": 3370
+    },
+    {
+      "epoch": 16.855,
+      "grad_norm": 3.8271725177764893,
+      "learning_rate": 1.5725000000000002e-06,
+      "loss": 0.1262,
+      "step": 3371
+    },
+    {
+      "epoch": 16.86,
+      "grad_norm": 3.1400415897369385,
+      "learning_rate": 1.5700000000000002e-06,
+      "loss": 0.1647,
+      "step": 3372
+    },
+    {
+      "epoch": 16.865,
+      "grad_norm": 3.6939187049865723,
+      "learning_rate": 1.5675e-06,
+      "loss": 0.2745,
+      "step": 3373
+    },
+    {
+      "epoch": 16.87,
+      "grad_norm": 3.6110997200012207,
+      "learning_rate": 1.565e-06,
+      "loss": 0.4966,
+      "step": 3374
+    },
+    {
+      "epoch": 16.875,
+      "grad_norm": 6.412832260131836,
+      "learning_rate": 1.5625e-06,
+      "loss": 0.4686,
+      "step": 3375
+    },
+    {
+      "epoch": 16.88,
+      "grad_norm": 12.13860034942627,
+      "learning_rate": 1.56e-06,
+      "loss": 0.616,
+      "step": 3376
+    },
+    {
+      "epoch": 16.885,
+      "grad_norm": 2.8702049255371094,
+      "learning_rate": 1.5575000000000001e-06,
+      "loss": 0.3631,
+      "step": 3377
+    },
+    {
+      "epoch": 16.89,
+      "grad_norm": 4.390155792236328,
+      "learning_rate": 1.5550000000000001e-06,
+      "loss": 0.1698,
+      "step": 3378
+    },
+    {
+      "epoch": 16.895,
+      "grad_norm": 5.024360179901123,
+      "learning_rate": 1.5525000000000002e-06,
+      "loss": 0.3749,
+      "step": 3379
+    },
+    {
+      "epoch": 16.9,
+      "grad_norm": 6.758533000946045,
+      "learning_rate": 1.5500000000000002e-06,
+      "loss": 0.2574,
+      "step": 3380
+    },
+    {
+      "epoch": 16.905,
+      "grad_norm": 2.6670968532562256,
+      "learning_rate": 1.5475000000000002e-06,
+      "loss": 0.2004,
+      "step": 3381
+    },
+    {
+      "epoch": 16.91,
+      "grad_norm": 4.190546035766602,
+      "learning_rate": 1.545e-06,
+      "loss": 0.1779,
+      "step": 3382
+    },
+    {
+      "epoch": 16.915,
+      "grad_norm": 6.017910480499268,
+      "learning_rate": 1.5425e-06,
+      "loss": 0.2128,
+      "step": 3383
+    },
+    {
+      "epoch": 16.92,
+      "grad_norm": 4.900136947631836,
+      "learning_rate": 1.54e-06,
+      "loss": 0.2266,
+      "step": 3384
+    },
+    {
+      "epoch": 16.925,
+      "grad_norm": 6.530525207519531,
+      "learning_rate": 1.5375e-06,
+      "loss": 0.7791,
+      "step": 3385
+    },
+    {
+      "epoch": 16.93,
+      "grad_norm": 4.728362560272217,
+      "learning_rate": 1.5350000000000001e-06,
+      "loss": 0.5358,
+      "step": 3386
+    },
+    {
+      "epoch": 16.935,
+      "grad_norm": 4.729607105255127,
+      "learning_rate": 1.5325000000000002e-06,
+      "loss": 0.5013,
+      "step": 3387
+    },
+    {
+      "epoch": 16.94,
+      "grad_norm": 7.058508396148682,
+      "learning_rate": 1.5300000000000002e-06,
+      "loss": 0.5955,
+      "step": 3388
+    },
+    {
+      "epoch": 16.945,
+      "grad_norm": 2.7840518951416016,
+      "learning_rate": 1.5275000000000002e-06,
+      "loss": 0.1263,
+      "step": 3389
+    },
+    {
+      "epoch": 16.95,
+      "grad_norm": 7.263626575469971,
+      "learning_rate": 1.525e-06,
+      "loss": 0.5498,
+      "step": 3390
+    },
+    {
+      "epoch": 16.955,
+      "grad_norm": 5.092872619628906,
+      "learning_rate": 1.5225e-06,
+      "loss": 0.4895,
+      "step": 3391
+    },
+    {
+      "epoch": 16.96,
+      "grad_norm": 3.7699272632598877,
+      "learning_rate": 1.52e-06,
+      "loss": 0.617,
+      "step": 3392
+    },
+    {
+      "epoch": 16.965,
+      "grad_norm": 4.53580379486084,
+      "learning_rate": 1.5175e-06,
+      "loss": 0.3373,
+      "step": 3393
+    },
+    {
+      "epoch": 16.97,
+      "grad_norm": 4.466193199157715,
+      "learning_rate": 1.5150000000000001e-06,
+      "loss": 0.3776,
+      "step": 3394
+    },
+    {
+      "epoch": 16.975,
+      "grad_norm": 3.712405204772949,
+      "learning_rate": 1.5125000000000001e-06,
+      "loss": 0.0939,
+      "step": 3395
+    },
+    {
+      "epoch": 16.98,
+      "grad_norm": 3.1014857292175293,
+      "learning_rate": 1.5100000000000002e-06,
+      "loss": 0.1173,
+      "step": 3396
+    },
+    {
+      "epoch": 16.985,
+      "grad_norm": 4.102566242218018,
+      "learning_rate": 1.5075000000000002e-06,
+      "loss": 0.3972,
+      "step": 3397
+    },
+    {
+      "epoch": 16.99,
+      "grad_norm": 4.366480827331543,
+      "learning_rate": 1.505e-06,
+      "loss": 0.2209,
+      "step": 3398
+    },
+    {
+      "epoch": 16.995,
+      "grad_norm": 3.611368417739868,
+      "learning_rate": 1.5025e-06,
+      "loss": 0.3165,
+      "step": 3399
+    },
+    {
+      "epoch": 17.0,
+      "grad_norm": 5.137276649475098,
+      "learning_rate": 1.5e-06,
+      "loss": 0.2077,
+      "step": 3400
+    },
+    {
+      "epoch": 17.0,
+      "eval_loss": 0.19446712732315063,
+      "eval_runtime": 15.5213,
+      "eval_samples_per_second": 51.413,
+      "eval_steps_per_second": 6.443,
+      "step": 3400
+    },
+    {
+      "epoch": 17.005,
+      "grad_norm": 3.882573366165161,
+      "learning_rate": 1.4975e-06,
+      "loss": 0.2766,
+      "step": 3401
+    },
+    {
+      "epoch": 17.01,
+      "grad_norm": 2.9682796001434326,
+      "learning_rate": 1.495e-06,
+      "loss": 0.2504,
+      "step": 3402
+    },
+    {
+      "epoch": 17.015,
+      "grad_norm": 3.3439695835113525,
+      "learning_rate": 1.4925000000000001e-06,
+      "loss": 0.2193,
+      "step": 3403
+    },
+    {
+      "epoch": 17.02,
+      "grad_norm": 3.566612482070923,
+      "learning_rate": 1.4900000000000001e-06,
+      "loss": 0.205,
+      "step": 3404
+    },
+    {
+      "epoch": 17.025,
+      "grad_norm": 4.3815531730651855,
+      "learning_rate": 1.4875000000000002e-06,
+      "loss": 0.1659,
+      "step": 3405
+    },
+    {
+      "epoch": 17.03,
+      "grad_norm": 12.198838233947754,
+      "learning_rate": 1.485e-06,
+      "loss": 0.6116,
+      "step": 3406
+    },
+    {
+      "epoch": 17.035,
+      "grad_norm": 3.024465560913086,
+      "learning_rate": 1.4825e-06,
+      "loss": 0.2353,
+      "step": 3407
+    },
+    {
+      "epoch": 17.04,
+      "grad_norm": 5.118997573852539,
+      "learning_rate": 1.48e-06,
+      "loss": 0.3107,
+      "step": 3408
+    },
+    {
+      "epoch": 17.045,
+      "grad_norm": 5.240654945373535,
+      "learning_rate": 1.4775e-06,
+      "loss": 0.2309,
+      "step": 3409
+    },
+    {
+      "epoch": 17.05,
+      "grad_norm": 4.817141532897949,
+      "learning_rate": 1.475e-06,
+      "loss": 0.2134,
+      "step": 3410
+    },
+    {
+      "epoch": 17.055,
+      "grad_norm": 6.374105930328369,
+      "learning_rate": 1.4725e-06,
+      "loss": 0.5803,
+      "step": 3411
+    },
+    {
+      "epoch": 17.06,
+      "grad_norm": 3.816504955291748,
+      "learning_rate": 1.4700000000000001e-06,
+      "loss": 0.1063,
+      "step": 3412
+    },
+    {
+      "epoch": 17.065,
+      "grad_norm": 3.752039909362793,
+      "learning_rate": 1.4675000000000001e-06,
+      "loss": 0.2506,
+      "step": 3413
+    },
+    {
+      "epoch": 17.07,
+      "grad_norm": 3.4115569591522217,
+      "learning_rate": 1.465e-06,
+      "loss": 0.2055,
+      "step": 3414
+    },
+    {
+      "epoch": 17.075,
+      "grad_norm": 4.999746322631836,
+      "learning_rate": 1.4625e-06,
+      "loss": 0.4273,
+      "step": 3415
+    },
+    {
+      "epoch": 17.08,
+      "grad_norm": 3.0715651512145996,
+      "learning_rate": 1.46e-06,
+      "loss": 0.1787,
+      "step": 3416
+    },
+    {
+      "epoch": 17.085,
+      "grad_norm": 5.321597576141357,
+      "learning_rate": 1.4575e-06,
+      "loss": 0.4451,
+      "step": 3417
+    },
+    {
+      "epoch": 17.09,
+      "grad_norm": 3.4731502532958984,
+      "learning_rate": 1.455e-06,
+      "loss": 0.1488,
+      "step": 3418
+    },
+    {
+      "epoch": 17.095,
+      "grad_norm": 3.7567055225372314,
+      "learning_rate": 1.4525e-06,
+      "loss": 0.2604,
+      "step": 3419
+    },
+    {
+      "epoch": 17.1,
+      "grad_norm": 4.553384304046631,
+      "learning_rate": 1.45e-06,
+      "loss": 0.3776,
+      "step": 3420
+    },
+    {
+      "epoch": 17.105,
+      "grad_norm": 4.569341659545898,
+      "learning_rate": 1.4475000000000001e-06,
+      "loss": 0.2531,
+      "step": 3421
+    },
+    {
+      "epoch": 17.11,
+      "grad_norm": 4.970749378204346,
+      "learning_rate": 1.445e-06,
+      "loss": 0.2811,
+      "step": 3422
+    },
+    {
+      "epoch": 17.115,
+      "grad_norm": 4.749880790710449,
+      "learning_rate": 1.4425e-06,
+      "loss": 0.3516,
+      "step": 3423
+    },
+    {
+      "epoch": 17.12,
+      "grad_norm": 7.778641700744629,
+      "learning_rate": 1.44e-06,
+      "loss": 0.6184,
+      "step": 3424
+    },
+    {
+      "epoch": 17.125,
+      "grad_norm": 5.796271324157715,
+      "learning_rate": 1.4375e-06,
+      "loss": 0.3924,
+      "step": 3425
+    },
+    {
+      "epoch": 17.13,
+      "grad_norm": 4.103795051574707,
+      "learning_rate": 1.435e-06,
+      "loss": 0.2111,
+      "step": 3426
+    },
+    {
+      "epoch": 17.135,
+      "grad_norm": 5.665992259979248,
+      "learning_rate": 1.4325e-06,
+      "loss": 0.7401,
+      "step": 3427
+    },
+    {
+      "epoch": 17.14,
+      "grad_norm": 6.363750457763672,
+      "learning_rate": 1.43e-06,
+      "loss": 0.6675,
+      "step": 3428
+    },
+    {
+      "epoch": 17.145,
+      "grad_norm": 4.592513084411621,
+      "learning_rate": 1.4275e-06,
+      "loss": 0.3217,
+      "step": 3429
+    },
+    {
+      "epoch": 17.15,
+      "grad_norm": 2.6420068740844727,
+      "learning_rate": 1.425e-06,
+      "loss": 0.096,
+      "step": 3430
+    },
+    {
+      "epoch": 17.155,
+      "grad_norm": 3.6999831199645996,
+      "learning_rate": 1.4225e-06,
+      "loss": 0.3735,
+      "step": 3431
+    },
+    {
+      "epoch": 17.16,
+      "grad_norm": 3.2805466651916504,
+      "learning_rate": 1.42e-06,
+      "loss": 0.209,
+      "step": 3432
+    },
+    {
+      "epoch": 17.165,
+      "grad_norm": 4.490758419036865,
+      "learning_rate": 1.4175e-06,
+      "loss": 0.4058,
+      "step": 3433
+    },
+    {
+      "epoch": 17.17,
+      "grad_norm": 3.500405788421631,
+      "learning_rate": 1.415e-06,
+      "loss": 0.562,
+      "step": 3434
+    },
+    {
+      "epoch": 17.175,
+      "grad_norm": 1.7692235708236694,
+      "learning_rate": 1.4125e-06,
+      "loss": 0.1672,
+      "step": 3435
+    },
+    {
+      "epoch": 17.18,
+      "grad_norm": 7.840676784515381,
+      "learning_rate": 1.41e-06,
+      "loss": 1.177,
+      "step": 3436
+    },
+    {
+      "epoch": 17.185,
+      "grad_norm": 4.515567779541016,
+      "learning_rate": 1.4075e-06,
+      "loss": 0.1781,
+      "step": 3437
+    },
+    {
+      "epoch": 17.19,
+      "grad_norm": 2.746764659881592,
+      "learning_rate": 1.4050000000000003e-06,
+      "loss": 0.1639,
+      "step": 3438
+    },
+    {
+      "epoch": 17.195,
+      "grad_norm": 2.5430116653442383,
+      "learning_rate": 1.4025000000000003e-06,
+      "loss": 0.1129,
+      "step": 3439
+    },
+    {
+      "epoch": 17.2,
+      "grad_norm": 2.946214199066162,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 0.1466,
+      "step": 3440
+    },
+    {
+      "epoch": 17.205,
+      "grad_norm": 6.146472930908203,
+      "learning_rate": 1.3975000000000002e-06,
+      "loss": 0.3854,
+      "step": 3441
+    },
+    {
+      "epoch": 17.21,
+      "grad_norm": 3.5039591789245605,
+      "learning_rate": 1.3950000000000002e-06,
+      "loss": 0.1947,
+      "step": 3442
+    },
+    {
+      "epoch": 17.215,
+      "grad_norm": 14.526341438293457,
+      "learning_rate": 1.3925000000000002e-06,
+      "loss": 0.5981,
+      "step": 3443
+    },
+    {
+      "epoch": 17.22,
+      "grad_norm": 4.883888244628906,
+      "learning_rate": 1.3900000000000002e-06,
+      "loss": 0.1867,
+      "step": 3444
+    },
+    {
+      "epoch": 17.225,
+      "grad_norm": 6.55780553817749,
+      "learning_rate": 1.3875000000000003e-06,
+      "loss": 0.5846,
+      "step": 3445
+    },
+    {
+      "epoch": 17.23,
+      "grad_norm": 6.657224655151367,
+      "learning_rate": 1.3850000000000003e-06,
+      "loss": 0.7324,
+      "step": 3446
+    },
+    {
+      "epoch": 17.235,
+      "grad_norm": 4.503376007080078,
+      "learning_rate": 1.3825000000000003e-06,
+      "loss": 0.3316,
+      "step": 3447
+    },
+    {
+      "epoch": 17.24,
+      "grad_norm": 1.704339623451233,
+      "learning_rate": 1.3800000000000001e-06,
+      "loss": 0.0406,
+      "step": 3448
+    },
+    {
+      "epoch": 17.245,
+      "grad_norm": 5.064183712005615,
+      "learning_rate": 1.3775000000000002e-06,
+      "loss": 0.2111,
+      "step": 3449
+    },
+    {
+      "epoch": 17.25,
+      "grad_norm": 3.247853994369507,
+      "learning_rate": 1.3750000000000002e-06,
+      "loss": 0.122,
+      "step": 3450
+    },
+    {
+      "epoch": 17.255,
+      "grad_norm": 4.86765718460083,
+      "learning_rate": 1.3725000000000002e-06,
+      "loss": 0.1931,
+      "step": 3451
+    },
+    {
+      "epoch": 17.26,
+      "grad_norm": 3.945019245147705,
+      "learning_rate": 1.3700000000000002e-06,
+      "loss": 0.4431,
+      "step": 3452
+    },
+    {
+      "epoch": 17.265,
+      "grad_norm": 4.379837989807129,
+      "learning_rate": 1.3675000000000002e-06,
+      "loss": 0.1764,
+      "step": 3453
+    },
+    {
+      "epoch": 17.27,
+      "grad_norm": 3.640855312347412,
+      "learning_rate": 1.3650000000000003e-06,
+      "loss": 0.2471,
+      "step": 3454
+    },
+    {
+      "epoch": 17.275,
+      "grad_norm": 4.39068603515625,
+      "learning_rate": 1.3625000000000003e-06,
+      "loss": 0.1898,
+      "step": 3455
+    },
+    {
+      "epoch": 17.28,
+      "grad_norm": 5.143912315368652,
+      "learning_rate": 1.3600000000000001e-06,
+      "loss": 0.3224,
+      "step": 3456
+    },
+    {
+      "epoch": 17.285,
+      "grad_norm": 3.2194507122039795,
+      "learning_rate": 1.3575000000000001e-06,
+      "loss": 0.2256,
+      "step": 3457
+    },
+    {
+      "epoch": 17.29,
+      "grad_norm": 5.75006103515625,
+      "learning_rate": 1.3550000000000002e-06,
+      "loss": 0.3213,
+      "step": 3458
+    },
+    {
+      "epoch": 17.295,
+      "grad_norm": 2.988380193710327,
+      "learning_rate": 1.3525000000000002e-06,
+      "loss": 0.1341,
+      "step": 3459
+    },
+    {
+      "epoch": 17.3,
+      "grad_norm": 8.820841789245605,
+      "learning_rate": 1.3500000000000002e-06,
+      "loss": 1.0638,
+      "step": 3460
+    },
+    {
+      "epoch": 17.305,
+      "grad_norm": 8.199127197265625,
+      "learning_rate": 1.3475000000000002e-06,
+      "loss": 0.8503,
+      "step": 3461
+    },
+    {
+      "epoch": 17.31,
+      "grad_norm": 8.224039077758789,
+      "learning_rate": 1.3450000000000003e-06,
+      "loss": 0.5272,
+      "step": 3462
+    },
+    {
+      "epoch": 17.315,
+      "grad_norm": 5.946857452392578,
+      "learning_rate": 1.3425000000000003e-06,
+      "loss": 0.534,
+      "step": 3463
+    },
+    {
+      "epoch": 17.32,
+      "grad_norm": 5.567616939544678,
+      "learning_rate": 1.34e-06,
+      "loss": 0.4004,
+      "step": 3464
+    },
+    {
+      "epoch": 17.325,
+      "grad_norm": 2.725923538208008,
+      "learning_rate": 1.3375000000000001e-06,
+      "loss": 0.0895,
+      "step": 3465
+    },
+    {
+      "epoch": 17.33,
+      "grad_norm": 3.792078971862793,
+      "learning_rate": 1.3350000000000001e-06,
+      "loss": 0.1563,
+      "step": 3466
+    },
+    {
+      "epoch": 17.335,
+      "grad_norm": 2.1296656131744385,
+      "learning_rate": 1.3325000000000002e-06,
+      "loss": 0.1068,
+      "step": 3467
+    },
+    {
+      "epoch": 17.34,
+      "grad_norm": 5.3928022384643555,
+      "learning_rate": 1.3300000000000002e-06,
+      "loss": 0.2773,
+      "step": 3468
+    },
+    {
+      "epoch": 17.345,
+      "grad_norm": 4.050804615020752,
+      "learning_rate": 1.3275000000000002e-06,
+      "loss": 0.1696,
+      "step": 3469
+    },
+    {
+      "epoch": 17.35,
+      "grad_norm": 7.533737659454346,
+      "learning_rate": 1.3250000000000002e-06,
+      "loss": 0.7054,
+      "step": 3470
+    },
+    {
+      "epoch": 17.355,
+      "grad_norm": 4.947542667388916,
+      "learning_rate": 1.3225000000000003e-06,
+      "loss": 0.2627,
+      "step": 3471
+    },
+    {
+      "epoch": 17.36,
+      "grad_norm": 5.3484039306640625,
+      "learning_rate": 1.32e-06,
+      "loss": 0.3181,
+      "step": 3472
+    },
+    {
+      "epoch": 17.365,
+      "grad_norm": 5.698040008544922,
+      "learning_rate": 1.3175e-06,
+      "loss": 0.8175,
+      "step": 3473
+    },
+    {
+      "epoch": 17.37,
+      "grad_norm": 2.4396255016326904,
+      "learning_rate": 1.3150000000000001e-06,
+      "loss": 0.2016,
+      "step": 3474
+    },
+    {
+      "epoch": 17.375,
+      "grad_norm": 5.307563781738281,
+      "learning_rate": 1.3125000000000001e-06,
+      "loss": 0.4339,
+      "step": 3475
+    },
+    {
+      "epoch": 17.38,
+      "grad_norm": 3.5963034629821777,
+      "learning_rate": 1.3100000000000002e-06,
+      "loss": 0.2453,
+      "step": 3476
+    },
+    {
+      "epoch": 17.385,
+      "grad_norm": 3.974271297454834,
+      "learning_rate": 1.3075000000000002e-06,
+      "loss": 0.2179,
+      "step": 3477
+    },
+    {
+      "epoch": 17.39,
+      "grad_norm": 3.954867124557495,
+      "learning_rate": 1.3050000000000002e-06,
+      "loss": 0.2138,
+      "step": 3478
+    },
+    {
+      "epoch": 17.395,
+      "grad_norm": 8.501402854919434,
+      "learning_rate": 1.3025000000000002e-06,
+      "loss": 0.5201,
+      "step": 3479
+    },
+    {
+      "epoch": 17.4,
+      "grad_norm": 5.32069206237793,
+      "learning_rate": 1.3e-06,
+      "loss": 0.298,
+      "step": 3480
+    },
+    {
+      "epoch": 17.405,
+      "grad_norm": 1.8387954235076904,
+      "learning_rate": 1.2975e-06,
+      "loss": 0.0874,
+      "step": 3481
+    },
+    {
+      "epoch": 17.41,
+      "grad_norm": 4.971597194671631,
+      "learning_rate": 1.295e-06,
+      "loss": 0.1314,
+      "step": 3482
+    },
+    {
+      "epoch": 17.415,
+      "grad_norm": 5.760146617889404,
+      "learning_rate": 1.2925000000000001e-06,
+      "loss": 0.4946,
+      "step": 3483
+    },
+    {
+      "epoch": 17.42,
+      "grad_norm": 3.4041361808776855,
+      "learning_rate": 1.2900000000000001e-06,
+      "loss": 0.1885,
+      "step": 3484
+    },
+    {
+      "epoch": 17.425,
+      "grad_norm": 3.7039568424224854,
+      "learning_rate": 1.2875000000000002e-06,
+      "loss": 0.2376,
+      "step": 3485
+    },
+    {
+      "epoch": 17.43,
+      "grad_norm": 8.94361400604248,
+      "learning_rate": 1.2850000000000002e-06,
+      "loss": 0.4923,
+      "step": 3486
+    },
+    {
+      "epoch": 17.435,
+      "grad_norm": 3.0983850955963135,
+      "learning_rate": 1.2825000000000002e-06,
+      "loss": 0.0374,
+      "step": 3487
+    },
+    {
+      "epoch": 17.44,
+      "grad_norm": 4.638023853302002,
+      "learning_rate": 1.28e-06,
+      "loss": 0.396,
+      "step": 3488
+    },
+    {
+      "epoch": 17.445,
+      "grad_norm": 3.5828001499176025,
+      "learning_rate": 1.2775e-06,
+      "loss": 0.2949,
+      "step": 3489
+    },
+    {
+      "epoch": 17.45,
+      "grad_norm": 8.87285041809082,
+      "learning_rate": 1.275e-06,
+      "loss": 0.6431,
+      "step": 3490
+    },
+    {
+      "epoch": 17.455,
+      "grad_norm": 4.595953464508057,
+      "learning_rate": 1.2725e-06,
+      "loss": 0.408,
+      "step": 3491
+    },
+    {
+      "epoch": 17.46,
+      "grad_norm": 5.200317859649658,
+      "learning_rate": 1.2700000000000001e-06,
+      "loss": 0.5282,
+      "step": 3492
+    },
+    {
+      "epoch": 17.465,
+      "grad_norm": 4.058686256408691,
+      "learning_rate": 1.2675000000000001e-06,
+      "loss": 0.328,
+      "step": 3493
+    },
+    {
+      "epoch": 17.47,
+      "grad_norm": 3.902787685394287,
+      "learning_rate": 1.2650000000000002e-06,
+      "loss": 0.2,
+      "step": 3494
+    },
+    {
+      "epoch": 17.475,
+      "grad_norm": 4.310050010681152,
+      "learning_rate": 1.2625000000000002e-06,
+      "loss": 0.4924,
+      "step": 3495
+    },
+    {
+      "epoch": 17.48,
+      "grad_norm": 2.564716100692749,
+      "learning_rate": 1.26e-06,
+      "loss": 0.197,
+      "step": 3496
+    },
+    {
+      "epoch": 17.485,
+      "grad_norm": 5.677400588989258,
+      "learning_rate": 1.2575e-06,
+      "loss": 0.3364,
+      "step": 3497
+    },
+    {
+      "epoch": 17.49,
+      "grad_norm": 3.2230818271636963,
+      "learning_rate": 1.255e-06,
+      "loss": 0.1633,
+      "step": 3498
+    },
+    {
+      "epoch": 17.495,
+      "grad_norm": 3.7361018657684326,
+      "learning_rate": 1.2525e-06,
+      "loss": 0.2129,
+      "step": 3499
+    },
+    {
+      "epoch": 17.5,
+      "grad_norm": 3.3416073322296143,
+      "learning_rate": 1.25e-06,
+      "loss": 0.2392,
+      "step": 3500
+    },
+    {
+      "epoch": 17.505,
+      "grad_norm": 22.86296844482422,
+      "learning_rate": 1.2475000000000001e-06,
+      "loss": 0.385,
+      "step": 3501
+    },
+    {
+      "epoch": 17.51,
+      "grad_norm": 3.2887656688690186,
+      "learning_rate": 1.2450000000000002e-06,
+      "loss": 0.2875,
+      "step": 3502
+    },
+    {
+      "epoch": 17.515,
+      "grad_norm": 3.815303087234497,
+      "learning_rate": 1.2425000000000002e-06,
+      "loss": 0.1899,
+      "step": 3503
+    },
+    {
+      "epoch": 17.52,
+      "grad_norm": 3.542574167251587,
+      "learning_rate": 1.2400000000000002e-06,
+      "loss": 0.2047,
+      "step": 3504
+    },
+    {
+      "epoch": 17.525,
+      "grad_norm": 4.92686128616333,
+      "learning_rate": 1.2375e-06,
+      "loss": 0.187,
+      "step": 3505
+    },
+    {
+      "epoch": 17.53,
+      "grad_norm": 1.8266156911849976,
+      "learning_rate": 1.235e-06,
+      "loss": 0.0781,
+      "step": 3506
+    },
+    {
+      "epoch": 17.535,
+      "grad_norm": 2.860957145690918,
+      "learning_rate": 1.2325e-06,
+      "loss": 0.3319,
+      "step": 3507
+    },
+    {
+      "epoch": 17.54,
+      "grad_norm": 5.883620262145996,
+      "learning_rate": 1.23e-06,
+      "loss": 0.391,
+      "step": 3508
+    },
+    {
+      "epoch": 17.545,
+      "grad_norm": 3.6476218700408936,
+      "learning_rate": 1.2275000000000001e-06,
+      "loss": 0.5852,
+      "step": 3509
+    },
+    {
+      "epoch": 17.55,
+      "grad_norm": 7.6689229011535645,
+      "learning_rate": 1.2250000000000001e-06,
+      "loss": 0.261,
+      "step": 3510
+    },
+    {
+      "epoch": 17.555,
+      "grad_norm": 4.628355979919434,
+      "learning_rate": 1.2225000000000002e-06,
+      "loss": 0.4962,
+      "step": 3511
+    },
+    {
+      "epoch": 17.56,
+      "grad_norm": 4.30551815032959,
+      "learning_rate": 1.2200000000000002e-06,
+      "loss": 0.4737,
+      "step": 3512
+    },
+    {
+      "epoch": 17.565,
+      "grad_norm": 4.2342987060546875,
+      "learning_rate": 1.2175e-06,
+      "loss": 0.3465,
+      "step": 3513
+    },
+    {
+      "epoch": 17.57,
+      "grad_norm": 3.951742172241211,
+      "learning_rate": 1.215e-06,
+      "loss": 0.3673,
+      "step": 3514
+    },
+    {
+      "epoch": 17.575,
+      "grad_norm": 7.576996326446533,
+      "learning_rate": 1.2125e-06,
+      "loss": 0.4491,
+      "step": 3515
+    },
+    {
+      "epoch": 17.58,
+      "grad_norm": 6.888933181762695,
+      "learning_rate": 1.21e-06,
+      "loss": 0.4931,
+      "step": 3516
+    },
+    {
+      "epoch": 17.585,
+      "grad_norm": 5.126704216003418,
+      "learning_rate": 1.2075e-06,
+      "loss": 0.3847,
+      "step": 3517
+    },
+    {
+      "epoch": 17.59,
+      "grad_norm": 5.120598316192627,
+      "learning_rate": 1.2050000000000001e-06,
+      "loss": 0.2803,
+      "step": 3518
+    },
+    {
+      "epoch": 17.595,
+      "grad_norm": 3.4318132400512695,
+      "learning_rate": 1.2025000000000001e-06,
+      "loss": 0.2022,
+      "step": 3519
+    },
+    {
+      "epoch": 17.6,
+      "grad_norm": 4.16720724105835,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 0.1661,
+      "step": 3520
+    },
+    {
+      "epoch": 17.605,
+      "grad_norm": 4.478663921356201,
+      "learning_rate": 1.1975e-06,
+      "loss": 0.1375,
+      "step": 3521
+    },
+    {
+      "epoch": 17.61,
+      "grad_norm": 8.510655403137207,
+      "learning_rate": 1.195e-06,
+      "loss": 0.3771,
+      "step": 3522
+    },
+    {
+      "epoch": 17.615,
+      "grad_norm": 5.538172721862793,
+      "learning_rate": 1.1925e-06,
+      "loss": 0.3241,
+      "step": 3523
+    },
+    {
+      "epoch": 17.62,
+      "grad_norm": 4.872572898864746,
+      "learning_rate": 1.19e-06,
+      "loss": 0.2302,
+      "step": 3524
+    },
+    {
+      "epoch": 17.625,
+      "grad_norm": 5.964207172393799,
+      "learning_rate": 1.1875e-06,
+      "loss": 0.2821,
+      "step": 3525
+    },
+    {
+      "epoch": 17.63,
+      "grad_norm": 4.612212181091309,
+      "learning_rate": 1.185e-06,
+      "loss": 0.4514,
+      "step": 3526
+    },
+    {
+      "epoch": 17.635,
+      "grad_norm": 4.089216232299805,
+      "learning_rate": 1.1825000000000001e-06,
+      "loss": 0.3098,
+      "step": 3527
+    },
+    {
+      "epoch": 17.64,
+      "grad_norm": 9.317346572875977,
+      "learning_rate": 1.1800000000000001e-06,
+      "loss": 0.6735,
+      "step": 3528
+    },
+    {
+      "epoch": 17.645,
+      "grad_norm": 5.726574420928955,
+      "learning_rate": 1.1775e-06,
+      "loss": 0.573,
+      "step": 3529
+    },
+    {
+      "epoch": 17.65,
+      "grad_norm": 3.408263921737671,
+      "learning_rate": 1.175e-06,
+      "loss": 0.4266,
+      "step": 3530
+    },
+    {
+      "epoch": 17.655,
+      "grad_norm": 5.021317958831787,
+      "learning_rate": 1.1725e-06,
+      "loss": 0.3031,
+      "step": 3531
+    },
+    {
+      "epoch": 17.66,
+      "grad_norm": 5.618551731109619,
+      "learning_rate": 1.1700000000000002e-06,
+      "loss": 0.6891,
+      "step": 3532
+    },
+    {
+      "epoch": 17.665,
+      "grad_norm": 8.518710136413574,
+      "learning_rate": 1.1675000000000003e-06,
+      "loss": 0.2692,
+      "step": 3533
+    },
+    {
+      "epoch": 17.67,
+      "grad_norm": 4.41873025894165,
+      "learning_rate": 1.165e-06,
+      "loss": 0.3608,
+      "step": 3534
+    },
+    {
+      "epoch": 17.675,
+      "grad_norm": 5.112030029296875,
+      "learning_rate": 1.1625e-06,
+      "loss": 0.5147,
+      "step": 3535
+    },
+    {
+      "epoch": 17.68,
+      "grad_norm": 5.903056621551514,
+      "learning_rate": 1.1600000000000001e-06,
+      "loss": 0.5775,
+      "step": 3536
+    },
+    {
+      "epoch": 17.685,
+      "grad_norm": 3.868605136871338,
+      "learning_rate": 1.1575000000000001e-06,
+      "loss": 0.4398,
+      "step": 3537
+    },
+    {
+      "epoch": 17.69,
+      "grad_norm": 5.562639236450195,
+      "learning_rate": 1.1550000000000002e-06,
+      "loss": 0.1769,
+      "step": 3538
+    },
+    {
+      "epoch": 17.695,
+      "grad_norm": 2.774164915084839,
+      "learning_rate": 1.1525000000000002e-06,
+      "loss": 0.1446,
+      "step": 3539
+    },
+    {
+      "epoch": 17.7,
+      "grad_norm": 6.027912139892578,
+      "learning_rate": 1.1500000000000002e-06,
+      "loss": 0.361,
+      "step": 3540
+    },
+    {
+      "epoch": 17.705,
+      "grad_norm": 3.6880881786346436,
+      "learning_rate": 1.1475000000000002e-06,
+      "loss": 0.3667,
+      "step": 3541
+    },
+    {
+      "epoch": 17.71,
+      "grad_norm": 3.9484100341796875,
+      "learning_rate": 1.145e-06,
+      "loss": 0.0954,
+      "step": 3542
+    },
+    {
+      "epoch": 17.715,
+      "grad_norm": 4.018896579742432,
+      "learning_rate": 1.1425e-06,
+      "loss": 0.2168,
+      "step": 3543
+    },
+    {
+      "epoch": 17.72,
+      "grad_norm": 3.90238618850708,
+      "learning_rate": 1.14e-06,
+      "loss": 0.2983,
+      "step": 3544
+    },
+    {
+      "epoch": 17.725,
+      "grad_norm": 3.012153387069702,
+      "learning_rate": 1.1375000000000001e-06,
+      "loss": 0.2576,
+      "step": 3545
+    },
+    {
+      "epoch": 17.73,
+      "grad_norm": 4.09089469909668,
+      "learning_rate": 1.1350000000000001e-06,
+      "loss": 0.3987,
+      "step": 3546
+    },
+    {
+      "epoch": 17.735,
+      "grad_norm": 1.4479118585586548,
+      "learning_rate": 1.1325000000000002e-06,
+      "loss": 0.032,
+      "step": 3547
+    },
+    {
+      "epoch": 17.74,
+      "grad_norm": 4.09706974029541,
+      "learning_rate": 1.1300000000000002e-06,
+      "loss": 0.2883,
+      "step": 3548
+    },
+    {
+      "epoch": 17.745,
+      "grad_norm": 4.880789756774902,
+      "learning_rate": 1.1275000000000002e-06,
+      "loss": 0.1992,
+      "step": 3549
+    },
+    {
+      "epoch": 17.75,
+      "grad_norm": 4.993986129760742,
+      "learning_rate": 1.125e-06,
+      "loss": 0.5068,
+      "step": 3550
+    },
+    {
+      "epoch": 17.755,
+      "grad_norm": 3.733489751815796,
+      "learning_rate": 1.1225e-06,
+      "loss": 0.3517,
+      "step": 3551
+    },
+    {
+      "epoch": 17.76,
+      "grad_norm": 4.264455318450928,
+      "learning_rate": 1.12e-06,
+      "loss": 0.1228,
+      "step": 3552
+    },
+    {
+      "epoch": 17.765,
+      "grad_norm": 3.14739990234375,
+      "learning_rate": 1.1175e-06,
+      "loss": 0.1146,
+      "step": 3553
+    },
+    {
+      "epoch": 17.77,
+      "grad_norm": 3.076986074447632,
+      "learning_rate": 1.1150000000000001e-06,
+      "loss": 0.3105,
+      "step": 3554
+    },
+    {
+      "epoch": 17.775,
+      "grad_norm": 5.034000873565674,
+      "learning_rate": 1.1125000000000001e-06,
+      "loss": 0.39,
+      "step": 3555
+    },
+    {
+      "epoch": 17.78,
+      "grad_norm": 5.384341716766357,
+      "learning_rate": 1.1100000000000002e-06,
+      "loss": 0.1285,
+      "step": 3556
+    },
+    {
+      "epoch": 17.785,
+      "grad_norm": 4.269554615020752,
+      "learning_rate": 1.1075000000000002e-06,
+      "loss": 0.2053,
+      "step": 3557
+    },
+    {
+      "epoch": 17.79,
+      "grad_norm": 3.1643424034118652,
+      "learning_rate": 1.105e-06,
+      "loss": 0.1183,
+      "step": 3558
+    },
+    {
+      "epoch": 17.795,
+      "grad_norm": 4.0334625244140625,
+      "learning_rate": 1.1025e-06,
+      "loss": 0.2838,
+      "step": 3559
+    },
+    {
+      "epoch": 17.8,
+      "grad_norm": 2.957672119140625,
+      "learning_rate": 1.1e-06,
+      "loss": 0.0681,
+      "step": 3560
+    },
+    {
+      "epoch": 17.805,
+      "grad_norm": 6.300900459289551,
+      "learning_rate": 1.0975e-06,
+      "loss": 0.5641,
+      "step": 3561
+    },
+    {
+      "epoch": 17.81,
+      "grad_norm": 10.119474411010742,
+      "learning_rate": 1.095e-06,
+      "loss": 0.5901,
+      "step": 3562
+    },
+    {
+      "epoch": 17.815,
+      "grad_norm": 6.088560581207275,
+      "learning_rate": 1.0925000000000001e-06,
+      "loss": 0.2428,
+      "step": 3563
+    },
+    {
+      "epoch": 17.82,
+      "grad_norm": 5.725935935974121,
+      "learning_rate": 1.0900000000000002e-06,
+      "loss": 0.3926,
+      "step": 3564
+    },
+    {
+      "epoch": 17.825,
+      "grad_norm": 2.666710138320923,
+      "learning_rate": 1.0875000000000002e-06,
+      "loss": 0.0667,
+      "step": 3565
+    },
+    {
+      "epoch": 17.83,
+      "grad_norm": 7.405109882354736,
+      "learning_rate": 1.085e-06,
+      "loss": 0.3801,
+      "step": 3566
+    },
+    {
+      "epoch": 17.835,
+      "grad_norm": 8.38760757446289,
+      "learning_rate": 1.0825e-06,
+      "loss": 0.6513,
+      "step": 3567
+    },
+    {
+      "epoch": 17.84,
+      "grad_norm": 2.7117276191711426,
+      "learning_rate": 1.08e-06,
+      "loss": 0.1715,
+      "step": 3568
+    },
+    {
+      "epoch": 17.845,
+      "grad_norm": 3.261204242706299,
+      "learning_rate": 1.0775e-06,
+      "loss": 0.2783,
+      "step": 3569
+    },
+    {
+      "epoch": 17.85,
+      "grad_norm": 4.056581974029541,
+      "learning_rate": 1.075e-06,
+      "loss": 0.2186,
+      "step": 3570
+    },
+    {
+      "epoch": 17.855,
+      "grad_norm": 5.020595550537109,
+      "learning_rate": 1.0725000000000001e-06,
+      "loss": 0.191,
+      "step": 3571
+    },
+    {
+      "epoch": 17.86,
+      "grad_norm": 3.854360818862915,
+      "learning_rate": 1.0700000000000001e-06,
+      "loss": 0.3805,
+      "step": 3572
+    },
+    {
+      "epoch": 17.865,
+      "grad_norm": 7.411772727966309,
+      "learning_rate": 1.0675000000000002e-06,
+      "loss": 0.3597,
+      "step": 3573
+    },
+    {
+      "epoch": 17.87,
+      "grad_norm": 4.204338073730469,
+      "learning_rate": 1.065e-06,
+      "loss": 0.3177,
+      "step": 3574
+    },
+    {
+      "epoch": 17.875,
+      "grad_norm": 2.0991995334625244,
+      "learning_rate": 1.0625e-06,
+      "loss": 0.0373,
+      "step": 3575
+    },
+    {
+      "epoch": 17.88,
+      "grad_norm": 3.5615155696868896,
+      "learning_rate": 1.06e-06,
+      "loss": 0.1528,
+      "step": 3576
+    },
+    {
+      "epoch": 17.885,
+      "grad_norm": 3.096902847290039,
+      "learning_rate": 1.0575e-06,
+      "loss": 0.1821,
+      "step": 3577
+    },
+    {
+      "epoch": 17.89,
+      "grad_norm": 4.306049823760986,
+      "learning_rate": 1.055e-06,
+      "loss": 0.0977,
+      "step": 3578
+    },
+    {
+      "epoch": 17.895,
+      "grad_norm": 5.882034778594971,
+      "learning_rate": 1.0525e-06,
+      "loss": 0.306,
+      "step": 3579
+    },
+    {
+      "epoch": 17.9,
+      "grad_norm": 4.9777302742004395,
+      "learning_rate": 1.0500000000000001e-06,
+      "loss": 0.3713,
+      "step": 3580
+    },
+    {
+      "epoch": 17.905,
+      "grad_norm": 2.8412909507751465,
+      "learning_rate": 1.0475000000000001e-06,
+      "loss": 0.1661,
+      "step": 3581
+    },
+    {
+      "epoch": 17.91,
+      "grad_norm": 2.5934154987335205,
+      "learning_rate": 1.045e-06,
+      "loss": 0.0713,
+      "step": 3582
+    },
+    {
+      "epoch": 17.915,
+      "grad_norm": 16.809934616088867,
+      "learning_rate": 1.0425e-06,
+      "loss": 0.6057,
+      "step": 3583
+    },
+    {
+      "epoch": 17.92,
+      "grad_norm": 4.879302978515625,
+      "learning_rate": 1.04e-06,
+      "loss": 0.4875,
+      "step": 3584
+    },
+    {
+      "epoch": 17.925,
+      "grad_norm": 3.7760324478149414,
+      "learning_rate": 1.0375e-06,
+      "loss": 0.1721,
+      "step": 3585
+    },
+    {
+      "epoch": 17.93,
+      "grad_norm": 5.390283107757568,
+      "learning_rate": 1.035e-06,
+      "loss": 0.5088,
+      "step": 3586
+    },
+    {
+      "epoch": 17.935,
+      "grad_norm": 3.9657561779022217,
+      "learning_rate": 1.0325e-06,
+      "loss": 0.2029,
+      "step": 3587
+    },
+    {
+      "epoch": 17.94,
+      "grad_norm": 3.0941410064697266,
+      "learning_rate": 1.03e-06,
+      "loss": 0.1432,
+      "step": 3588
+    },
+    {
+      "epoch": 17.945,
+      "grad_norm": 4.832656383514404,
+      "learning_rate": 1.0275000000000001e-06,
+      "loss": 0.2464,
+      "step": 3589
+    },
+    {
+      "epoch": 17.95,
+      "grad_norm": 3.399916648864746,
+      "learning_rate": 1.025e-06,
+      "loss": 0.1462,
+      "step": 3590
+    },
+    {
+      "epoch": 17.955,
+      "grad_norm": 3.8467509746551514,
+      "learning_rate": 1.0225e-06,
+      "loss": 0.2292,
+      "step": 3591
+    },
+    {
+      "epoch": 17.96,
+      "grad_norm": 3.8830339908599854,
+      "learning_rate": 1.02e-06,
+      "loss": 0.2991,
+      "step": 3592
+    },
+    {
+      "epoch": 17.965,
+      "grad_norm": 3.3385915756225586,
+      "learning_rate": 1.0175e-06,
+      "loss": 0.1621,
+      "step": 3593
+    },
+    {
+      "epoch": 17.97,
+      "grad_norm": 5.3569536209106445,
+      "learning_rate": 1.0150000000000002e-06,
+      "loss": 0.3425,
+      "step": 3594
+    },
+    {
+      "epoch": 17.975,
+      "grad_norm": 5.815585136413574,
+      "learning_rate": 1.0125e-06,
+      "loss": 0.3836,
+      "step": 3595
+    },
+    {
+      "epoch": 17.98,
+      "grad_norm": 5.238308906555176,
+      "learning_rate": 1.01e-06,
+      "loss": 0.4469,
+      "step": 3596
+    },
+    {
+      "epoch": 17.985,
+      "grad_norm": 1.0525665283203125,
+      "learning_rate": 1.0075e-06,
+      "loss": 0.0337,
+      "step": 3597
+    },
+    {
+      "epoch": 17.99,
+      "grad_norm": 6.449350833892822,
+      "learning_rate": 1.0050000000000001e-06,
+      "loss": 0.2802,
+      "step": 3598
+    },
+    {
+      "epoch": 17.995,
+      "grad_norm": 10.151460647583008,
+      "learning_rate": 1.0025000000000001e-06,
+      "loss": 0.4975,
+      "step": 3599
+    },
+    {
+      "epoch": 18.0,
+      "grad_norm": 6.603348731994629,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.0445,
+      "step": 3600
+    },
+    {
+      "epoch": 18.0,
+      "eval_loss": 0.19213047623634338,
+      "eval_runtime": 15.4626,
+      "eval_samples_per_second": 51.608,
+      "eval_steps_per_second": 6.467,
+      "step": 3600
+    },
+    {
+      "epoch": 18.005,
+      "grad_norm": 6.414360523223877,
+      "learning_rate": 9.975000000000002e-07,
+      "loss": 0.4058,
+      "step": 3601
+    },
+    {
+      "epoch": 18.01,
+      "grad_norm": 2.657835006713867,
+      "learning_rate": 9.950000000000002e-07,
+      "loss": 0.4087,
+      "step": 3602
+    },
+    {
+      "epoch": 18.015,
+      "grad_norm": 8.133830070495605,
+      "learning_rate": 9.925e-07,
+      "loss": 0.2561,
+      "step": 3603
+    },
+    {
+      "epoch": 18.02,
+      "grad_norm": 5.337889194488525,
+      "learning_rate": 9.9e-07,
+      "loss": 0.1936,
+      "step": 3604
+    },
+    {
+      "epoch": 18.025,
+      "grad_norm": 4.144707202911377,
+      "learning_rate": 9.875e-07,
+      "loss": 0.3255,
+      "step": 3605
+    },
+    {
+      "epoch": 18.03,
+      "grad_norm": 5.087593078613281,
+      "learning_rate": 9.85e-07,
+      "loss": 0.337,
+      "step": 3606
+    },
+    {
+      "epoch": 18.035,
+      "grad_norm": 3.8073902130126953,
+      "learning_rate": 9.825000000000001e-07,
+      "loss": 0.6005,
+      "step": 3607
+    },
+    {
+      "epoch": 18.04,
+      "grad_norm": 3.551114797592163,
+      "learning_rate": 9.800000000000001e-07,
+      "loss": 0.2166,
+      "step": 3608
+    },
+    {
+      "epoch": 18.045,
+      "grad_norm": 3.5221590995788574,
+      "learning_rate": 9.775000000000002e-07,
+      "loss": 0.2246,
+      "step": 3609
+    },
+    {
+      "epoch": 18.05,
+      "grad_norm": 3.9775664806365967,
+      "learning_rate": 9.750000000000002e-07,
+      "loss": 0.4648,
+      "step": 3610
+    },
+    {
+      "epoch": 18.055,
+      "grad_norm": 4.745052337646484,
+      "learning_rate": 9.725e-07,
+      "loss": 0.4447,
+      "step": 3611
+    },
+    {
+      "epoch": 18.06,
+      "grad_norm": 2.0482115745544434,
+      "learning_rate": 9.7e-07,
+      "loss": 0.1355,
+      "step": 3612
+    },
+    {
+      "epoch": 18.065,
+      "grad_norm": 5.979750633239746,
+      "learning_rate": 9.675e-07,
+      "loss": 0.3135,
+      "step": 3613
+    },
+    {
+      "epoch": 18.07,
+      "grad_norm": 3.1865017414093018,
+      "learning_rate": 9.65e-07,
+      "loss": 0.1516,
+      "step": 3614
+    },
+    {
+      "epoch": 18.075,
+      "grad_norm": 4.504761219024658,
+      "learning_rate": 9.625e-07,
+      "loss": 0.1838,
+      "step": 3615
+    },
+    {
+      "epoch": 18.08,
+      "grad_norm": 4.08765172958374,
+      "learning_rate": 9.600000000000001e-07,
+      "loss": 0.2802,
+      "step": 3616
+    },
+    {
+      "epoch": 18.085,
+      "grad_norm": 4.596279144287109,
+      "learning_rate": 9.575000000000001e-07,
+      "loss": 0.1536,
+      "step": 3617
+    },
+    {
+      "epoch": 18.09,
+      "grad_norm": 5.3212056159973145,
+      "learning_rate": 9.550000000000002e-07,
+      "loss": 0.2893,
+      "step": 3618
+    },
+    {
+      "epoch": 18.095,
+      "grad_norm": 4.766326427459717,
+      "learning_rate": 9.525000000000001e-07,
+      "loss": 0.199,
+      "step": 3619
+    },
+    {
+      "epoch": 18.1,
+      "grad_norm": 3.9234867095947266,
+      "learning_rate": 9.500000000000001e-07,
+      "loss": 0.2126,
+      "step": 3620
+    },
+    {
+      "epoch": 18.105,
+      "grad_norm": 4.269250869750977,
+      "learning_rate": 9.475e-07,
+      "loss": 0.4242,
+      "step": 3621
+    },
+    {
+      "epoch": 18.11,
+      "grad_norm": 4.1503143310546875,
+      "learning_rate": 9.450000000000001e-07,
+      "loss": 0.7232,
+      "step": 3622
+    },
+    {
+      "epoch": 18.115,
+      "grad_norm": 4.380398273468018,
+      "learning_rate": 9.425000000000001e-07,
+      "loss": 0.1794,
+      "step": 3623
+    },
+    {
+      "epoch": 18.12,
+      "grad_norm": 2.6925508975982666,
+      "learning_rate": 9.400000000000001e-07,
+      "loss": 0.264,
+      "step": 3624
+    },
+    {
+      "epoch": 18.125,
+      "grad_norm": 17.73287582397461,
+      "learning_rate": 9.375000000000001e-07,
+      "loss": 0.6877,
+      "step": 3625
+    },
+    {
+      "epoch": 18.13,
+      "grad_norm": 3.6111087799072266,
+      "learning_rate": 9.35e-07,
+      "loss": 0.3508,
+      "step": 3626
+    },
+    {
+      "epoch": 18.135,
+      "grad_norm": 4.601571083068848,
+      "learning_rate": 9.325000000000001e-07,
+      "loss": 0.4023,
+      "step": 3627
+    },
+    {
+      "epoch": 18.14,
+      "grad_norm": 2.7724802494049072,
+      "learning_rate": 9.300000000000001e-07,
+      "loss": 0.2849,
+      "step": 3628
+    },
+    {
+      "epoch": 18.145,
+      "grad_norm": 2.6019840240478516,
+      "learning_rate": 9.275000000000001e-07,
+      "loss": 0.1289,
+      "step": 3629
+    },
+    {
+      "epoch": 18.15,
+      "grad_norm": 2.948263168334961,
+      "learning_rate": 9.25e-07,
+      "loss": 0.2446,
+      "step": 3630
+    },
+    {
+      "epoch": 18.155,
+      "grad_norm": 2.4856152534484863,
+      "learning_rate": 9.225000000000001e-07,
+      "loss": 0.0898,
+      "step": 3631
+    },
+    {
+      "epoch": 18.16,
+      "grad_norm": 3.568239688873291,
+      "learning_rate": 9.200000000000001e-07,
+      "loss": 0.1377,
+      "step": 3632
+    },
+    {
+      "epoch": 18.165,
+      "grad_norm": 6.75287389755249,
+      "learning_rate": 9.175000000000001e-07,
+      "loss": 0.3093,
+      "step": 3633
+    },
+    {
+      "epoch": 18.17,
+      "grad_norm": 4.305637836456299,
+      "learning_rate": 9.15e-07,
+      "loss": 0.3809,
+      "step": 3634
+    },
+    {
+      "epoch": 18.175,
+      "grad_norm": 3.7961175441741943,
+      "learning_rate": 9.125e-07,
+      "loss": 0.1321,
+      "step": 3635
+    },
+    {
+      "epoch": 18.18,
+      "grad_norm": 4.931452751159668,
+      "learning_rate": 9.100000000000001e-07,
+      "loss": 0.2254,
+      "step": 3636
+    },
+    {
+      "epoch": 18.185,
+      "grad_norm": 3.471118450164795,
+      "learning_rate": 9.075000000000001e-07,
+      "loss": 0.1498,
+      "step": 3637
+    },
+    {
+      "epoch": 18.19,
+      "grad_norm": 5.757644176483154,
+      "learning_rate": 9.05e-07,
+      "loss": 0.5887,
+      "step": 3638
+    },
+    {
+      "epoch": 18.195,
+      "grad_norm": 33.592708587646484,
+      "learning_rate": 9.025e-07,
+      "loss": 0.6114,
+      "step": 3639
+    },
+    {
+      "epoch": 18.2,
+      "grad_norm": 7.0438432693481445,
+      "learning_rate": 9.000000000000001e-07,
+      "loss": 0.5709,
+      "step": 3640
+    },
+    {
+      "epoch": 18.205,
+      "grad_norm": 2.9734466075897217,
+      "learning_rate": 8.975000000000001e-07,
+      "loss": 0.1287,
+      "step": 3641
+    },
+    {
+      "epoch": 18.21,
+      "grad_norm": 2.9040613174438477,
+      "learning_rate": 8.95e-07,
+      "loss": 0.1956,
+      "step": 3642
+    },
+    {
+      "epoch": 18.215,
+      "grad_norm": 5.7859272956848145,
+      "learning_rate": 8.925e-07,
+      "loss": 0.2582,
+      "step": 3643
+    },
+    {
+      "epoch": 18.22,
+      "grad_norm": 3.8436577320098877,
+      "learning_rate": 8.900000000000001e-07,
+      "loss": 0.2419,
+      "step": 3644
+    },
+    {
+      "epoch": 18.225,
+      "grad_norm": 9.9724760055542,
+      "learning_rate": 8.875000000000001e-07,
+      "loss": 0.404,
+      "step": 3645
+    },
+    {
+      "epoch": 18.23,
+      "grad_norm": 7.809504985809326,
+      "learning_rate": 8.85e-07,
+      "loss": 0.3246,
+      "step": 3646
+    },
+    {
+      "epoch": 18.235,
+      "grad_norm": 6.094870090484619,
+      "learning_rate": 8.825e-07,
+      "loss": 0.4623,
+      "step": 3647
+    },
+    {
+      "epoch": 18.24,
+      "grad_norm": 4.512402534484863,
+      "learning_rate": 8.8e-07,
+      "loss": 0.3352,
+      "step": 3648
+    },
+    {
+      "epoch": 18.245,
+      "grad_norm": 4.568089962005615,
+      "learning_rate": 8.775000000000001e-07,
+      "loss": 0.2809,
+      "step": 3649
+    },
+    {
+      "epoch": 18.25,
+      "grad_norm": 4.1774067878723145,
+      "learning_rate": 8.75e-07,
+      "loss": 0.1629,
+      "step": 3650
+    },
+    {
+      "epoch": 18.255,
+      "grad_norm": 3.028980255126953,
+      "learning_rate": 8.725e-07,
+      "loss": 0.2988,
+      "step": 3651
+    },
+    {
+      "epoch": 18.26,
+      "grad_norm": 5.228294372558594,
+      "learning_rate": 8.7e-07,
+      "loss": 0.4544,
+      "step": 3652
+    },
+    {
+      "epoch": 18.265,
+      "grad_norm": 8.223569869995117,
+      "learning_rate": 8.675000000000001e-07,
+      "loss": 0.5494,
+      "step": 3653
+    },
+    {
+      "epoch": 18.27,
+      "grad_norm": 6.205587387084961,
+      "learning_rate": 8.65e-07,
+      "loss": 1.0821,
+      "step": 3654
+    },
+    {
+      "epoch": 18.275,
+      "grad_norm": 3.081597089767456,
+      "learning_rate": 8.625e-07,
+      "loss": 0.1975,
+      "step": 3655
+    },
+    {
+      "epoch": 18.28,
+      "grad_norm": 3.5202319622039795,
+      "learning_rate": 8.6e-07,
+      "loss": 0.1939,
+      "step": 3656
+    },
+    {
+      "epoch": 18.285,
+      "grad_norm": 4.107989311218262,
+      "learning_rate": 8.575000000000002e-07,
+      "loss": 0.2162,
+      "step": 3657
+    },
+    {
+      "epoch": 18.29,
+      "grad_norm": 4.328415870666504,
+      "learning_rate": 8.550000000000002e-07,
+      "loss": 0.286,
+      "step": 3658
+    },
+    {
+      "epoch": 18.295,
+      "grad_norm": 4.390509128570557,
+      "learning_rate": 8.525000000000001e-07,
+      "loss": 0.3257,
+      "step": 3659
+    },
+    {
+      "epoch": 18.3,
+      "grad_norm": 2.9701716899871826,
+      "learning_rate": 8.500000000000001e-07,
+      "loss": 0.1303,
+      "step": 3660
+    },
+    {
+      "epoch": 18.305,
+      "grad_norm": 2.9118125438690186,
+      "learning_rate": 8.475000000000001e-07,
+      "loss": 0.37,
+      "step": 3661
+    },
+    {
+      "epoch": 18.31,
+      "grad_norm": 3.1001288890838623,
+      "learning_rate": 8.450000000000002e-07,
+      "loss": 0.1734,
+      "step": 3662
+    },
+    {
+      "epoch": 18.315,
+      "grad_norm": 2.7583963871002197,
+      "learning_rate": 8.425000000000001e-07,
+      "loss": 0.1189,
+      "step": 3663
+    },
+    {
+      "epoch": 18.32,
+      "grad_norm": 2.0983667373657227,
+      "learning_rate": 8.400000000000001e-07,
+      "loss": 0.1472,
+      "step": 3664
+    },
+    {
+      "epoch": 18.325,
+      "grad_norm": 30.662343978881836,
+      "learning_rate": 8.375000000000001e-07,
+      "loss": 1.0658,
+      "step": 3665
+    },
+    {
+      "epoch": 18.33,
+      "grad_norm": 3.248748779296875,
+      "learning_rate": 8.350000000000002e-07,
+      "loss": 0.4267,
+      "step": 3666
+    },
+    {
+      "epoch": 18.335,
+      "grad_norm": 1.9023979902267456,
+      "learning_rate": 8.325000000000001e-07,
+      "loss": 0.1178,
+      "step": 3667
+    },
+    {
+      "epoch": 18.34,
+      "grad_norm": 5.478327751159668,
+      "learning_rate": 8.300000000000001e-07,
+      "loss": 0.6152,
+      "step": 3668
+    },
+    {
+      "epoch": 18.345,
+      "grad_norm": 3.321002960205078,
+      "learning_rate": 8.275000000000001e-07,
+      "loss": 0.1579,
+      "step": 3669
+    },
+    {
+      "epoch": 18.35,
+      "grad_norm": 3.9427974224090576,
+      "learning_rate": 8.250000000000001e-07,
+      "loss": 0.3455,
+      "step": 3670
+    },
+    {
+      "epoch": 18.355,
+      "grad_norm": 6.803521156311035,
+      "learning_rate": 8.225000000000001e-07,
+      "loss": 0.3754,
+      "step": 3671
+    },
+    {
+      "epoch": 18.36,
+      "grad_norm": 4.849568843841553,
+      "learning_rate": 8.200000000000001e-07,
+      "loss": 0.3793,
+      "step": 3672
+    },
+    {
+      "epoch": 18.365,
+      "grad_norm": 4.915162563323975,
+      "learning_rate": 8.175000000000001e-07,
+      "loss": 0.4638,
+      "step": 3673
+    },
+    {
+      "epoch": 18.37,
+      "grad_norm": 5.784596920013428,
+      "learning_rate": 8.150000000000001e-07,
+      "loss": 0.5757,
+      "step": 3674
+    },
+    {
+      "epoch": 18.375,
+      "grad_norm": 3.469790458679199,
+      "learning_rate": 8.125000000000001e-07,
+      "loss": 0.2977,
+      "step": 3675
+    },
+    {
+      "epoch": 18.38,
+      "grad_norm": 4.9008893966674805,
+      "learning_rate": 8.100000000000001e-07,
+      "loss": 0.2094,
+      "step": 3676
+    },
+    {
+      "epoch": 18.385,
+      "grad_norm": 4.558713436126709,
+      "learning_rate": 8.075000000000001e-07,
+      "loss": 0.2792,
+      "step": 3677
+    },
+    {
+      "epoch": 18.39,
+      "grad_norm": 4.522171974182129,
+      "learning_rate": 8.050000000000001e-07,
+      "loss": 0.1964,
+      "step": 3678
+    },
+    {
+      "epoch": 18.395,
+      "grad_norm": 2.8531434535980225,
+      "learning_rate": 8.025e-07,
+      "loss": 0.071,
+      "step": 3679
+    },
+    {
+      "epoch": 18.4,
+      "grad_norm": 2.3239169120788574,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 0.2436,
+      "step": 3680
+    },
+    {
+      "epoch": 18.405,
+      "grad_norm": 2.8895602226257324,
+      "learning_rate": 7.975000000000001e-07,
+      "loss": 0.1199,
+      "step": 3681
+    },
+    {
+      "epoch": 18.41,
+      "grad_norm": 5.752717971801758,
+      "learning_rate": 7.950000000000001e-07,
+      "loss": 0.1923,
+      "step": 3682
+    },
+    {
+      "epoch": 18.415,
+      "grad_norm": 7.250768184661865,
+      "learning_rate": 7.925e-07,
+      "loss": 0.3125,
+      "step": 3683
+    },
+    {
+      "epoch": 18.42,
+      "grad_norm": 3.901494026184082,
+      "learning_rate": 7.900000000000001e-07,
+      "loss": 0.3028,
+      "step": 3684
+    },
+    {
+      "epoch": 18.425,
+      "grad_norm": 5.809044361114502,
+      "learning_rate": 7.875000000000001e-07,
+      "loss": 0.2886,
+      "step": 3685
+    },
+    {
+      "epoch": 18.43,
+      "grad_norm": 3.8569176197052,
+      "learning_rate": 7.850000000000001e-07,
+      "loss": 0.2875,
+      "step": 3686
+    },
+    {
+      "epoch": 18.435,
+      "grad_norm": 4.525756359100342,
+      "learning_rate": 7.825e-07,
+      "loss": 0.3939,
+      "step": 3687
+    },
+    {
+      "epoch": 18.44,
+      "grad_norm": 4.387591361999512,
+      "learning_rate": 7.8e-07,
+      "loss": 0.2592,
+      "step": 3688
+    },
+    {
+      "epoch": 18.445,
+      "grad_norm": 4.122879505157471,
+      "learning_rate": 7.775000000000001e-07,
+      "loss": 0.1539,
+      "step": 3689
+    },
+    {
+      "epoch": 18.45,
+      "grad_norm": 7.389387130737305,
+      "learning_rate": 7.750000000000001e-07,
+      "loss": 0.6516,
+      "step": 3690
+    },
+    {
+      "epoch": 18.455,
+      "grad_norm": 5.360770225524902,
+      "learning_rate": 7.725e-07,
+      "loss": 0.2471,
+      "step": 3691
+    },
+    {
+      "epoch": 18.46,
+      "grad_norm": 4.596419811248779,
+      "learning_rate": 7.7e-07,
+      "loss": 0.313,
+      "step": 3692
+    },
+    {
+      "epoch": 18.465,
+      "grad_norm": 5.193201065063477,
+      "learning_rate": 7.675000000000001e-07,
+      "loss": 0.4794,
+      "step": 3693
+    },
+    {
+      "epoch": 18.47,
+      "grad_norm": 5.482877731323242,
+      "learning_rate": 7.650000000000001e-07,
+      "loss": 0.2806,
+      "step": 3694
+    },
+    {
+      "epoch": 18.475,
+      "grad_norm": 2.9124603271484375,
+      "learning_rate": 7.625e-07,
+      "loss": 0.2728,
+      "step": 3695
+    },
+    {
+      "epoch": 18.48,
+      "grad_norm": 9.124374389648438,
+      "learning_rate": 7.6e-07,
+      "loss": 0.314,
+      "step": 3696
+    },
+    {
+      "epoch": 18.485,
+      "grad_norm": 3.836876153945923,
+      "learning_rate": 7.575000000000001e-07,
+      "loss": 0.3154,
+      "step": 3697
+    },
+    {
+      "epoch": 18.49,
+      "grad_norm": 12.354578018188477,
+      "learning_rate": 7.550000000000001e-07,
+      "loss": 0.4847,
+      "step": 3698
+    },
+    {
+      "epoch": 18.495,
+      "grad_norm": 4.552191734313965,
+      "learning_rate": 7.525e-07,
+      "loss": 0.1582,
+      "step": 3699
+    },
+    {
+      "epoch": 18.5,
+      "grad_norm": 9.311527252197266,
+      "learning_rate": 7.5e-07,
+      "loss": 0.4714,
+      "step": 3700
+    },
+    {
+      "epoch": 18.505,
+      "grad_norm": 5.861787796020508,
+      "learning_rate": 7.475e-07,
+      "loss": 0.38,
+      "step": 3701
+    },
+    {
+      "epoch": 18.51,
+      "grad_norm": 7.12423849105835,
+      "learning_rate": 7.450000000000001e-07,
+      "loss": 0.5733,
+      "step": 3702
+    },
+    {
+      "epoch": 18.515,
+      "grad_norm": 1.9372780323028564,
+      "learning_rate": 7.425e-07,
+      "loss": 0.1381,
+      "step": 3703
+    },
+    {
+      "epoch": 18.52,
+      "grad_norm": 6.54016637802124,
+      "learning_rate": 7.4e-07,
+      "loss": 0.4519,
+      "step": 3704
+    },
+    {
+      "epoch": 18.525,
+      "grad_norm": 2.47084379196167,
+      "learning_rate": 7.375e-07,
+      "loss": 0.1014,
+      "step": 3705
+    },
+    {
+      "epoch": 18.53,
+      "grad_norm": 2.261693239212036,
+      "learning_rate": 7.350000000000001e-07,
+      "loss": 0.1055,
+      "step": 3706
+    },
+    {
+      "epoch": 18.535,
+      "grad_norm": 2.6924045085906982,
+      "learning_rate": 7.325e-07,
+      "loss": 0.2469,
+      "step": 3707
+    },
+    {
+      "epoch": 18.54,
+      "grad_norm": 2.495431423187256,
+      "learning_rate": 7.3e-07,
+      "loss": 0.1937,
+      "step": 3708
+    },
+    {
+      "epoch": 18.545,
+      "grad_norm": 4.074650287628174,
+      "learning_rate": 7.275e-07,
+      "loss": 0.268,
+      "step": 3709
+    },
+    {
+      "epoch": 18.55,
+      "grad_norm": 3.7677969932556152,
+      "learning_rate": 7.25e-07,
+      "loss": 0.4081,
+      "step": 3710
+    },
+    {
+      "epoch": 18.555,
+      "grad_norm": 4.246646881103516,
+      "learning_rate": 7.225e-07,
+      "loss": 0.3235,
+      "step": 3711
+    },
+    {
+      "epoch": 18.56,
+      "grad_norm": 2.45059871673584,
+      "learning_rate": 7.2e-07,
+      "loss": 0.0503,
+      "step": 3712
+    },
+    {
+      "epoch": 18.565,
+      "grad_norm": 4.093562126159668,
+      "learning_rate": 7.175e-07,
+      "loss": 0.3361,
+      "step": 3713
+    },
+    {
+      "epoch": 18.57,
+      "grad_norm": 5.89413595199585,
+      "learning_rate": 7.15e-07,
+      "loss": 0.4016,
+      "step": 3714
+    },
+    {
+      "epoch": 18.575,
+      "grad_norm": 6.102001190185547,
+      "learning_rate": 7.125e-07,
+      "loss": 0.4403,
+      "step": 3715
+    },
+    {
+      "epoch": 18.58,
+      "grad_norm": 7.075068950653076,
+      "learning_rate": 7.1e-07,
+      "loss": 0.5499,
+      "step": 3716
+    },
+    {
+      "epoch": 18.585,
+      "grad_norm": 5.699973106384277,
+      "learning_rate": 7.075e-07,
+      "loss": 0.5226,
+      "step": 3717
+    },
+    {
+      "epoch": 18.59,
+      "grad_norm": 2.322159767150879,
+      "learning_rate": 7.05e-07,
+      "loss": 0.2291,
+      "step": 3718
+    },
+    {
+      "epoch": 18.595,
+      "grad_norm": 251.3951873779297,
+      "learning_rate": 7.025000000000002e-07,
+      "loss": 1.0919,
+      "step": 3719
+    },
+    {
+      "epoch": 18.6,
+      "grad_norm": 3.704434394836426,
+      "learning_rate": 7.000000000000001e-07,
+      "loss": 0.6082,
+      "step": 3720
+    },
+    {
+      "epoch": 18.605,
+      "grad_norm": 9.250022888183594,
+      "learning_rate": 6.975000000000001e-07,
+      "loss": 1.0421,
+      "step": 3721
+    },
+    {
+      "epoch": 18.61,
+      "grad_norm": 3.485954523086548,
+      "learning_rate": 6.950000000000001e-07,
+      "loss": 0.2706,
+      "step": 3722
+    },
+    {
+      "epoch": 18.615,
+      "grad_norm": 3.4246373176574707,
+      "learning_rate": 6.925000000000001e-07,
+      "loss": 0.1347,
+      "step": 3723
+    },
+    {
+      "epoch": 18.62,
+      "grad_norm": 5.477091312408447,
+      "learning_rate": 6.900000000000001e-07,
+      "loss": 0.4078,
+      "step": 3724
+    },
+    {
+      "epoch": 18.625,
+      "grad_norm": 2.583685874938965,
+      "learning_rate": 6.875000000000001e-07,
+      "loss": 0.0874,
+      "step": 3725
+    },
+    {
+      "epoch": 18.63,
+      "grad_norm": 6.358933925628662,
+      "learning_rate": 6.850000000000001e-07,
+      "loss": 0.5121,
+      "step": 3726
+    },
+    {
+      "epoch": 18.635,
+      "grad_norm": 3.789268732070923,
+      "learning_rate": 6.825000000000001e-07,
+      "loss": 0.3411,
+      "step": 3727
+    },
+    {
+      "epoch": 18.64,
+      "grad_norm": 2.9707624912261963,
+      "learning_rate": 6.800000000000001e-07,
+      "loss": 0.1516,
+      "step": 3728
+    },
+    {
+      "epoch": 18.645,
+      "grad_norm": 8.62167739868164,
+      "learning_rate": 6.775000000000001e-07,
+      "loss": 0.4809,
+      "step": 3729
+    },
+    {
+      "epoch": 18.65,
+      "grad_norm": 4.493370532989502,
+      "learning_rate": 6.750000000000001e-07,
+      "loss": 0.2413,
+      "step": 3730
+    },
+    {
+      "epoch": 18.655,
+      "grad_norm": 1.924551010131836,
+      "learning_rate": 6.725000000000001e-07,
+      "loss": 0.12,
+      "step": 3731
+    },
+    {
+      "epoch": 18.66,
+      "grad_norm": 4.705450057983398,
+      "learning_rate": 6.7e-07,
+      "loss": 0.3439,
+      "step": 3732
+    },
+    {
+      "epoch": 18.665,
+      "grad_norm": 1.6782103776931763,
+      "learning_rate": 6.675000000000001e-07,
+      "loss": 0.0633,
+      "step": 3733
+    },
+    {
+      "epoch": 18.67,
+      "grad_norm": 2.4559223651885986,
+      "learning_rate": 6.650000000000001e-07,
+      "loss": 0.1487,
+      "step": 3734
+    },
+    {
+      "epoch": 18.675,
+      "grad_norm": 5.954574108123779,
+      "learning_rate": 6.625000000000001e-07,
+      "loss": 0.3607,
+      "step": 3735
+    },
+    {
+      "epoch": 18.68,
+      "grad_norm": 16.010211944580078,
+      "learning_rate": 6.6e-07,
+      "loss": 0.5969,
+      "step": 3736
+    },
+    {
+      "epoch": 18.685,
+      "grad_norm": 4.446964740753174,
+      "learning_rate": 6.575000000000001e-07,
+      "loss": 0.2693,
+      "step": 3737
+    },
+    {
+      "epoch": 18.69,
+      "grad_norm": 3.6463518142700195,
+      "learning_rate": 6.550000000000001e-07,
+      "loss": 0.1789,
+      "step": 3738
+    },
+    {
+      "epoch": 18.695,
+      "grad_norm": 2.9095544815063477,
+      "learning_rate": 6.525000000000001e-07,
+      "loss": 0.3209,
+      "step": 3739
+    },
+    {
+      "epoch": 18.7,
+      "grad_norm": 4.031222820281982,
+      "learning_rate": 6.5e-07,
+      "loss": 0.4446,
+      "step": 3740
+    },
+    {
+      "epoch": 18.705,
+      "grad_norm": 3.880913496017456,
+      "learning_rate": 6.475e-07,
+      "loss": 0.2547,
+      "step": 3741
+    },
+    {
+      "epoch": 18.71,
+      "grad_norm": 3.7417075634002686,
+      "learning_rate": 6.450000000000001e-07,
+      "loss": 0.193,
+      "step": 3742
+    },
+    {
+      "epoch": 18.715,
+      "grad_norm": 4.219126224517822,
+      "learning_rate": 6.425000000000001e-07,
+      "loss": 0.1159,
+      "step": 3743
+    },
+    {
+      "epoch": 18.72,
+      "grad_norm": 6.142454147338867,
+      "learning_rate": 6.4e-07,
+      "loss": 0.6623,
+      "step": 3744
+    },
+    {
+      "epoch": 18.725,
+      "grad_norm": 2.516157865524292,
+      "learning_rate": 6.375e-07,
+      "loss": 0.0879,
+      "step": 3745
+    },
+    {
+      "epoch": 18.73,
+      "grad_norm": 7.940515041351318,
+      "learning_rate": 6.350000000000001e-07,
+      "loss": 0.7118,
+      "step": 3746
+    },
+    {
+      "epoch": 18.735,
+      "grad_norm": 3.3893935680389404,
+      "learning_rate": 6.325000000000001e-07,
+      "loss": 0.1317,
+      "step": 3747
+    },
+    {
+      "epoch": 18.74,
+      "grad_norm": 3.930238723754883,
+      "learning_rate": 6.3e-07,
+      "loss": 0.4831,
+      "step": 3748
+    },
+    {
+      "epoch": 18.745,
+      "grad_norm": 4.979517459869385,
+      "learning_rate": 6.275e-07,
+      "loss": 0.1907,
+      "step": 3749
+    },
+    {
+      "epoch": 18.75,
+      "grad_norm": 3.1787681579589844,
+      "learning_rate": 6.25e-07,
+      "loss": 0.2815,
+      "step": 3750
+    },
+    {
+      "epoch": 18.755,
+      "grad_norm": 4.042114734649658,
+      "learning_rate": 6.225000000000001e-07,
+      "loss": 0.2986,
+      "step": 3751
+    },
+    {
+      "epoch": 18.76,
+      "grad_norm": 3.8174567222595215,
+      "learning_rate": 6.200000000000001e-07,
+      "loss": 0.1404,
+      "step": 3752
+    },
+    {
+      "epoch": 18.765,
+      "grad_norm": 4.696832656860352,
+      "learning_rate": 6.175e-07,
+      "loss": 0.3683,
+      "step": 3753
+    },
+    {
+      "epoch": 18.77,
+      "grad_norm": 2.76373553276062,
+      "learning_rate": 6.15e-07,
+      "loss": 0.1372,
+      "step": 3754
+    },
+    {
+      "epoch": 18.775,
+      "grad_norm": 3.8718221187591553,
+      "learning_rate": 6.125000000000001e-07,
+      "loss": 0.2745,
+      "step": 3755
+    },
+    {
+      "epoch": 18.78,
+      "grad_norm": 5.7584147453308105,
+      "learning_rate": 6.100000000000001e-07,
+      "loss": 0.1522,
+      "step": 3756
+    },
+    {
+      "epoch": 18.785,
+      "grad_norm": 2.5119876861572266,
+      "learning_rate": 6.075e-07,
+      "loss": 0.0709,
+      "step": 3757
+    },
+    {
+      "epoch": 18.79,
+      "grad_norm": 5.278165340423584,
+      "learning_rate": 6.05e-07,
+      "loss": 0.1556,
+      "step": 3758
+    },
+    {
+      "epoch": 18.795,
+      "grad_norm": 5.974693775177002,
+      "learning_rate": 6.025000000000001e-07,
+      "loss": 0.293,
+      "step": 3759
+    },
+    {
+      "epoch": 18.8,
+      "grad_norm": 5.086860656738281,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 0.4169,
+      "step": 3760
+    },
+    {
+      "epoch": 18.805,
+      "grad_norm": 5.714186668395996,
+      "learning_rate": 5.975e-07,
+      "loss": 0.3703,
+      "step": 3761
+    },
+    {
+      "epoch": 18.81,
+      "grad_norm": 3.9842753410339355,
+      "learning_rate": 5.95e-07,
+      "loss": 0.1873,
+      "step": 3762
+    },
+    {
+      "epoch": 18.815,
+      "grad_norm": 6.323047161102295,
+      "learning_rate": 5.925e-07,
+      "loss": 0.4329,
+      "step": 3763
+    },
+    {
+      "epoch": 18.82,
+      "grad_norm": 3.383909225463867,
+      "learning_rate": 5.900000000000001e-07,
+      "loss": 0.1775,
+      "step": 3764
+    },
+    {
+      "epoch": 18.825,
+      "grad_norm": 4.274167060852051,
+      "learning_rate": 5.875e-07,
+      "loss": 0.4046,
+      "step": 3765
+    },
+    {
+      "epoch": 18.83,
+      "grad_norm": 7.98857307434082,
+      "learning_rate": 5.850000000000001e-07,
+      "loss": 0.1182,
+      "step": 3766
+    },
+    {
+      "epoch": 18.835,
+      "grad_norm": 3.573542594909668,
+      "learning_rate": 5.825e-07,
+      "loss": 0.2159,
+      "step": 3767
+    },
+    {
+      "epoch": 18.84,
+      "grad_norm": 4.1028337478637695,
+      "learning_rate": 5.800000000000001e-07,
+      "loss": 0.3397,
+      "step": 3768
+    },
+    {
+      "epoch": 18.845,
+      "grad_norm": 15.797172546386719,
+      "learning_rate": 5.775000000000001e-07,
+      "loss": 0.3719,
+      "step": 3769
+    },
+    {
+      "epoch": 18.85,
+      "grad_norm": 270.8778381347656,
+      "learning_rate": 5.750000000000001e-07,
+      "loss": 0.7884,
+      "step": 3770
+    },
+    {
+      "epoch": 18.855,
+      "grad_norm": 3.9505276679992676,
+      "learning_rate": 5.725e-07,
+      "loss": 0.3362,
+      "step": 3771
+    },
+    {
+      "epoch": 18.86,
+      "grad_norm": 1.8637144565582275,
+      "learning_rate": 5.7e-07,
+      "loss": 0.0514,
+      "step": 3772
+    },
+    {
+      "epoch": 18.865,
+      "grad_norm": 52.386043548583984,
+      "learning_rate": 5.675000000000001e-07,
+      "loss": 1.0546,
+      "step": 3773
+    },
+    {
+      "epoch": 18.87,
+      "grad_norm": 5.708125591278076,
+      "learning_rate": 5.650000000000001e-07,
+      "loss": 0.395,
+      "step": 3774
+    },
+    {
+      "epoch": 18.875,
+      "grad_norm": 3.134474515914917,
+      "learning_rate": 5.625e-07,
+      "loss": 0.2898,
+      "step": 3775
+    },
+    {
+      "epoch": 18.88,
+      "grad_norm": 8.156974792480469,
+      "learning_rate": 5.6e-07,
+      "loss": 0.6514,
+      "step": 3776
+    },
+    {
+      "epoch": 18.885,
+      "grad_norm": 6.068454742431641,
+      "learning_rate": 5.575000000000001e-07,
+      "loss": 0.9069,
+      "step": 3777
+    },
+    {
+      "epoch": 18.89,
+      "grad_norm": 5.987872123718262,
+      "learning_rate": 5.550000000000001e-07,
+      "loss": 0.4743,
+      "step": 3778
+    },
+    {
+      "epoch": 18.895,
+      "grad_norm": 2.3744711875915527,
+      "learning_rate": 5.525e-07,
+      "loss": 0.1265,
+      "step": 3779
+    },
+    {
+      "epoch": 18.9,
+      "grad_norm": 102.964599609375,
+      "learning_rate": 5.5e-07,
+      "loss": 0.6113,
+      "step": 3780
+    },
+    {
+      "epoch": 18.905,
+      "grad_norm": 5.274792671203613,
+      "learning_rate": 5.475e-07,
+      "loss": 0.1671,
+      "step": 3781
+    },
+    {
+      "epoch": 18.91,
+      "grad_norm": 3.024134874343872,
+      "learning_rate": 5.450000000000001e-07,
+      "loss": 0.332,
+      "step": 3782
+    },
+    {
+      "epoch": 18.915,
+      "grad_norm": 2.2969558238983154,
+      "learning_rate": 5.425e-07,
+      "loss": 0.2392,
+      "step": 3783
+    },
+    {
+      "epoch": 18.92,
+      "grad_norm": 2.763758420944214,
+      "learning_rate": 5.4e-07,
+      "loss": 0.1402,
+      "step": 3784
+    },
+    {
+      "epoch": 18.925,
+      "grad_norm": 1.6174405813217163,
+      "learning_rate": 5.375e-07,
+      "loss": 0.0372,
+      "step": 3785
+    },
+    {
+      "epoch": 18.93,
+      "grad_norm": 4.813601016998291,
+      "learning_rate": 5.350000000000001e-07,
+      "loss": 0.1469,
+      "step": 3786
+    },
+    {
+      "epoch": 18.935,
+      "grad_norm": 5.153240203857422,
+      "learning_rate": 5.325e-07,
+      "loss": 0.4905,
+      "step": 3787
+    },
+    {
+      "epoch": 18.94,
+      "grad_norm": 3.5327141284942627,
+      "learning_rate": 5.3e-07,
+      "loss": 0.2579,
+      "step": 3788
+    },
+    {
+      "epoch": 18.945,
+      "grad_norm": 3.932525157928467,
+      "learning_rate": 5.275e-07,
+      "loss": 0.13,
+      "step": 3789
+    },
+    {
+      "epoch": 18.95,
+      "grad_norm": 10.065864562988281,
+      "learning_rate": 5.250000000000001e-07,
+      "loss": 0.3853,
+      "step": 3790
+    },
+    {
+      "epoch": 18.955,
+      "grad_norm": 2.4893903732299805,
+      "learning_rate": 5.225e-07,
+      "loss": 0.1096,
+      "step": 3791
+    },
+    {
+      "epoch": 18.96,
+      "grad_norm": 4.847314834594727,
+      "learning_rate": 5.2e-07,
+      "loss": 0.3665,
+      "step": 3792
+    },
+    {
+      "epoch": 18.965,
+      "grad_norm": 9.319063186645508,
+      "learning_rate": 5.175e-07,
+      "loss": 0.8814,
+      "step": 3793
+    },
+    {
+      "epoch": 18.97,
+      "grad_norm": 3.289945363998413,
+      "learning_rate": 5.15e-07,
+      "loss": 0.2003,
+      "step": 3794
+    },
+    {
+      "epoch": 18.975,
+      "grad_norm": 5.4716081619262695,
+      "learning_rate": 5.125e-07,
+      "loss": 0.5917,
+      "step": 3795
+    },
+    {
+      "epoch": 18.98,
+      "grad_norm": 3.385580539703369,
+      "learning_rate": 5.1e-07,
+      "loss": 0.5971,
+      "step": 3796
+    },
+    {
+      "epoch": 18.985,
+      "grad_norm": 2.7456722259521484,
+      "learning_rate": 5.075000000000001e-07,
+      "loss": 0.1009,
+      "step": 3797
+    },
+    {
+      "epoch": 18.99,
+      "grad_norm": 4.28514289855957,
+      "learning_rate": 5.05e-07,
+      "loss": 0.4078,
+      "step": 3798
+    },
+    {
+      "epoch": 18.995,
+      "grad_norm": 3.7169647216796875,
+      "learning_rate": 5.025000000000001e-07,
+      "loss": 0.1883,
+      "step": 3799
+    },
+    {
+      "epoch": 19.0,
+      "grad_norm": 6.687422275543213,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 0.3158,
+      "step": 3800
+    },
+    {
+      "epoch": 19.0,
+      "eval_loss": 0.19036567211151123,
+      "eval_runtime": 15.3892,
+      "eval_samples_per_second": 51.854,
+      "eval_steps_per_second": 6.498,
+      "step": 3800
+    },
+    {
+      "epoch": 19.005,
+      "grad_norm": 4.590823173522949,
+      "learning_rate": 4.975000000000001e-07,
+      "loss": 0.2755,
+      "step": 3801
+    },
+    {
+      "epoch": 19.01,
+      "grad_norm": 2.485339879989624,
+      "learning_rate": 4.95e-07,
+      "loss": 0.0958,
+      "step": 3802
+    },
+    {
+      "epoch": 19.015,
+      "grad_norm": 3.471675395965576,
+      "learning_rate": 4.925e-07,
+      "loss": 0.3521,
+      "step": 3803
+    },
+    {
+      "epoch": 19.02,
+      "grad_norm": 16.085906982421875,
+      "learning_rate": 4.900000000000001e-07,
+      "loss": 0.6986,
+      "step": 3804
+    },
+    {
+      "epoch": 19.025,
+      "grad_norm": 8.475992202758789,
+      "learning_rate": 4.875000000000001e-07,
+      "loss": 0.5149,
+      "step": 3805
+    },
+    {
+      "epoch": 19.03,
+      "grad_norm": 6.070219039916992,
+      "learning_rate": 4.85e-07,
+      "loss": 0.1882,
+      "step": 3806
+    },
+    {
+      "epoch": 19.035,
+      "grad_norm": 7.363964557647705,
+      "learning_rate": 4.825e-07,
+      "loss": 0.8411,
+      "step": 3807
+    },
+    {
+      "epoch": 19.04,
+      "grad_norm": 5.156578063964844,
+      "learning_rate": 4.800000000000001e-07,
+      "loss": 0.1579,
+      "step": 3808
+    },
+    {
+      "epoch": 19.045,
+      "grad_norm": 6.714890956878662,
+      "learning_rate": 4.775000000000001e-07,
+      "loss": 0.3215,
+      "step": 3809
+    },
+    {
+      "epoch": 19.05,
+      "grad_norm": 5.222243785858154,
+      "learning_rate": 4.7500000000000006e-07,
+      "loss": 0.2872,
+      "step": 3810
+    },
+    {
+      "epoch": 19.055,
+      "grad_norm": 5.404577255249023,
+      "learning_rate": 4.7250000000000003e-07,
+      "loss": 0.1399,
+      "step": 3811
+    },
+    {
+      "epoch": 19.06,
+      "grad_norm": 9.171982765197754,
+      "learning_rate": 4.7000000000000005e-07,
+      "loss": 0.5869,
+      "step": 3812
+    },
+    {
+      "epoch": 19.065,
+      "grad_norm": 3.7528862953186035,
+      "learning_rate": 4.675e-07,
+      "loss": 0.1467,
+      "step": 3813
+    },
+    {
+      "epoch": 19.07,
+      "grad_norm": 8.56230354309082,
+      "learning_rate": 4.6500000000000005e-07,
+      "loss": 0.5809,
+      "step": 3814
+    },
+    {
+      "epoch": 19.075,
+      "grad_norm": 6.122354507446289,
+      "learning_rate": 4.625e-07,
+      "loss": 0.3493,
+      "step": 3815
+    },
+    {
+      "epoch": 19.08,
+      "grad_norm": 2.4050674438476562,
+      "learning_rate": 4.6000000000000004e-07,
+      "loss": 0.1314,
+      "step": 3816
+    },
+    {
+      "epoch": 19.085,
+      "grad_norm": 6.048617362976074,
+      "learning_rate": 4.575e-07,
+      "loss": 0.4251,
+      "step": 3817
+    },
+    {
+      "epoch": 19.09,
+      "grad_norm": 6.893596172332764,
+      "learning_rate": 4.5500000000000004e-07,
+      "loss": 0.419,
+      "step": 3818
+    },
+    {
+      "epoch": 19.095,
+      "grad_norm": 4.327513217926025,
+      "learning_rate": 4.525e-07,
+      "loss": 0.238,
+      "step": 3819
+    },
+    {
+      "epoch": 19.1,
+      "grad_norm": 5.779255390167236,
+      "learning_rate": 4.5000000000000003e-07,
+      "loss": 0.353,
+      "step": 3820
+    },
+    {
+      "epoch": 19.105,
+      "grad_norm": 3.7132699489593506,
+      "learning_rate": 4.475e-07,
+      "loss": 0.1518,
+      "step": 3821
+    },
+    {
+      "epoch": 19.11,
+      "grad_norm": 5.017323970794678,
+      "learning_rate": 4.4500000000000003e-07,
+      "loss": 0.5271,
+      "step": 3822
+    },
+    {
+      "epoch": 19.115,
+      "grad_norm": 4.98216438293457,
+      "learning_rate": 4.425e-07,
+      "loss": 0.5621,
+      "step": 3823
+    },
+    {
+      "epoch": 19.12,
+      "grad_norm": 8.0313720703125,
+      "learning_rate": 4.4e-07,
+      "loss": 0.5697,
+      "step": 3824
+    },
+    {
+      "epoch": 19.125,
+      "grad_norm": 3.6557774543762207,
+      "learning_rate": 4.375e-07,
+      "loss": 0.2399,
+      "step": 3825
+    },
+    {
+      "epoch": 19.13,
+      "grad_norm": 3.7783775329589844,
+      "learning_rate": 4.35e-07,
+      "loss": 0.2393,
+      "step": 3826
+    },
+    {
+      "epoch": 19.135,
+      "grad_norm": 3.596280336380005,
+      "learning_rate": 4.325e-07,
+      "loss": 0.4012,
+      "step": 3827
+    },
+    {
+      "epoch": 19.14,
+      "grad_norm": 5.155918598175049,
+      "learning_rate": 4.3e-07,
+      "loss": 0.1606,
+      "step": 3828
+    },
+    {
+      "epoch": 19.145,
+      "grad_norm": 1.8245155811309814,
+      "learning_rate": 4.275000000000001e-07,
+      "loss": 0.0476,
+      "step": 3829
+    },
+    {
+      "epoch": 19.15,
+      "grad_norm": 3.16603422164917,
+      "learning_rate": 4.2500000000000006e-07,
+      "loss": 0.1753,
+      "step": 3830
+    },
+    {
+      "epoch": 19.155,
+      "grad_norm": 3.3752894401550293,
+      "learning_rate": 4.225000000000001e-07,
+      "loss": 0.3358,
+      "step": 3831
+    },
+    {
+      "epoch": 19.16,
+      "grad_norm": 4.750109672546387,
+      "learning_rate": 4.2000000000000006e-07,
+      "loss": 0.2344,
+      "step": 3832
+    },
+    {
+      "epoch": 19.165,
+      "grad_norm": 4.481292724609375,
+      "learning_rate": 4.175000000000001e-07,
+      "loss": 0.3107,
+      "step": 3833
+    },
+    {
+      "epoch": 19.17,
+      "grad_norm": 5.8020734786987305,
+      "learning_rate": 4.1500000000000005e-07,
+      "loss": 0.2872,
+      "step": 3834
+    },
+    {
+      "epoch": 19.175,
+      "grad_norm": 4.9224629402160645,
+      "learning_rate": 4.125000000000001e-07,
+      "loss": 0.4377,
+      "step": 3835
+    },
+    {
+      "epoch": 19.18,
+      "grad_norm": 5.376506328582764,
+      "learning_rate": 4.1000000000000004e-07,
+      "loss": 0.7042,
+      "step": 3836
+    },
+    {
+      "epoch": 19.185,
+      "grad_norm": 6.038681507110596,
+      "learning_rate": 4.0750000000000007e-07,
+      "loss": 0.6443,
+      "step": 3837
+    },
+    {
+      "epoch": 19.19,
+      "grad_norm": 8.549304008483887,
+      "learning_rate": 4.0500000000000004e-07,
+      "loss": 0.3322,
+      "step": 3838
+    },
+    {
+      "epoch": 19.195,
+      "grad_norm": 6.1060309410095215,
+      "learning_rate": 4.0250000000000006e-07,
+      "loss": 0.7803,
+      "step": 3839
+    },
+    {
+      "epoch": 19.2,
+      "grad_norm": 4.264254570007324,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 0.1862,
+      "step": 3840
+    },
+    {
+      "epoch": 19.205,
+      "grad_norm": 4.843519687652588,
+      "learning_rate": 3.9750000000000006e-07,
+      "loss": 0.2196,
+      "step": 3841
+    },
+    {
+      "epoch": 19.21,
+      "grad_norm": 4.124020576477051,
+      "learning_rate": 3.9500000000000003e-07,
+      "loss": 0.2084,
+      "step": 3842
+    },
+    {
+      "epoch": 19.215,
+      "grad_norm": 7.157625675201416,
+      "learning_rate": 3.9250000000000005e-07,
+      "loss": 0.5308,
+      "step": 3843
+    },
+    {
+      "epoch": 19.22,
+      "grad_norm": 5.037046432495117,
+      "learning_rate": 3.9e-07,
+      "loss": 0.3071,
+      "step": 3844
+    },
+    {
+      "epoch": 19.225,
+      "grad_norm": 3.0660996437072754,
+      "learning_rate": 3.8750000000000005e-07,
+      "loss": 0.1525,
+      "step": 3845
+    },
+    {
+      "epoch": 19.23,
+      "grad_norm": 3.795856237411499,
+      "learning_rate": 3.85e-07,
+      "loss": 0.2231,
+      "step": 3846
+    },
+    {
+      "epoch": 19.235,
+      "grad_norm": 6.268131732940674,
+      "learning_rate": 3.8250000000000004e-07,
+      "loss": 0.2545,
+      "step": 3847
+    },
+    {
+      "epoch": 19.24,
+      "grad_norm": 3.662970781326294,
+      "learning_rate": 3.8e-07,
+      "loss": 0.2797,
+      "step": 3848
+    },
+    {
+      "epoch": 19.245,
+      "grad_norm": 4.584615230560303,
+      "learning_rate": 3.7750000000000004e-07,
+      "loss": 0.2693,
+      "step": 3849
+    },
+    {
+      "epoch": 19.25,
+      "grad_norm": 6.840015411376953,
+      "learning_rate": 3.75e-07,
+      "loss": 0.5427,
+      "step": 3850
+    },
+    {
+      "epoch": 19.255,
+      "grad_norm": 4.420175075531006,
+      "learning_rate": 3.7250000000000003e-07,
+      "loss": 0.3051,
+      "step": 3851
+    },
+    {
+      "epoch": 19.26,
+      "grad_norm": 3.1015872955322266,
+      "learning_rate": 3.7e-07,
+      "loss": 0.2063,
+      "step": 3852
+    },
+    {
+      "epoch": 19.265,
+      "grad_norm": 4.38470458984375,
+      "learning_rate": 3.6750000000000003e-07,
+      "loss": 0.3866,
+      "step": 3853
+    },
+    {
+      "epoch": 19.27,
+      "grad_norm": 2.342325210571289,
+      "learning_rate": 3.65e-07,
+      "loss": 0.0376,
+      "step": 3854
+    },
+    {
+      "epoch": 19.275,
+      "grad_norm": 4.9687395095825195,
+      "learning_rate": 3.625e-07,
+      "loss": 0.5279,
+      "step": 3855
+    },
+    {
+      "epoch": 19.28,
+      "grad_norm": 2.7299537658691406,
+      "learning_rate": 3.6e-07,
+      "loss": 0.1004,
+      "step": 3856
+    },
+    {
+      "epoch": 19.285,
+      "grad_norm": 4.4709954261779785,
+      "learning_rate": 3.575e-07,
+      "loss": 0.2141,
+      "step": 3857
+    },
+    {
+      "epoch": 19.29,
+      "grad_norm": 4.248438835144043,
+      "learning_rate": 3.55e-07,
+      "loss": 0.2362,
+      "step": 3858
+    },
+    {
+      "epoch": 19.295,
+      "grad_norm": 3.045637369155884,
+      "learning_rate": 3.525e-07,
+      "loss": 0.1465,
+      "step": 3859
+    },
+    {
+      "epoch": 19.3,
+      "grad_norm": 3.602736473083496,
+      "learning_rate": 3.5000000000000004e-07,
+      "loss": 0.2476,
+      "step": 3860
+    },
+    {
+      "epoch": 19.305,
+      "grad_norm": 3.8818912506103516,
+      "learning_rate": 3.4750000000000006e-07,
+      "loss": 0.069,
+      "step": 3861
+    },
+    {
+      "epoch": 19.31,
+      "grad_norm": 5.097687244415283,
+      "learning_rate": 3.4500000000000003e-07,
+      "loss": 0.5342,
+      "step": 3862
+    },
+    {
+      "epoch": 19.315,
+      "grad_norm": 5.558239936828613,
+      "learning_rate": 3.4250000000000006e-07,
+      "loss": 0.524,
+      "step": 3863
+    },
+    {
+      "epoch": 19.32,
+      "grad_norm": 4.600433349609375,
+      "learning_rate": 3.4000000000000003e-07,
+      "loss": 0.258,
+      "step": 3864
+    },
+    {
+      "epoch": 19.325,
+      "grad_norm": 5.282414436340332,
+      "learning_rate": 3.3750000000000005e-07,
+      "loss": 0.2212,
+      "step": 3865
+    },
+    {
+      "epoch": 19.33,
+      "grad_norm": 6.102233409881592,
+      "learning_rate": 3.35e-07,
+      "loss": 0.3704,
+      "step": 3866
+    },
+    {
+      "epoch": 19.335,
+      "grad_norm": 2.792508602142334,
+      "learning_rate": 3.3250000000000005e-07,
+      "loss": 0.1363,
+      "step": 3867
+    },
+    {
+      "epoch": 19.34,
+      "grad_norm": 7.1083455085754395,
+      "learning_rate": 3.3e-07,
+      "loss": 0.4282,
+      "step": 3868
+    },
+    {
+      "epoch": 19.345,
+      "grad_norm": 4.861268043518066,
+      "learning_rate": 3.2750000000000004e-07,
+      "loss": 0.2408,
+      "step": 3869
+    },
+    {
+      "epoch": 19.35,
+      "grad_norm": 7.079010963439941,
+      "learning_rate": 3.25e-07,
+      "loss": 0.4404,
+      "step": 3870
+    },
+    {
+      "epoch": 19.355,
+      "grad_norm": 3.179600477218628,
+      "learning_rate": 3.2250000000000004e-07,
+      "loss": 0.2627,
+      "step": 3871
+    },
+    {
+      "epoch": 19.36,
+      "grad_norm": 5.526487350463867,
+      "learning_rate": 3.2e-07,
+      "loss": 0.3692,
+      "step": 3872
+    },
+    {
+      "epoch": 19.365,
+      "grad_norm": 4.398512363433838,
+      "learning_rate": 3.1750000000000003e-07,
+      "loss": 0.454,
+      "step": 3873
+    },
+    {
+      "epoch": 19.37,
+      "grad_norm": 5.618788719177246,
+      "learning_rate": 3.15e-07,
+      "loss": 0.7557,
+      "step": 3874
+    },
+    {
+      "epoch": 19.375,
+      "grad_norm": 4.0343403816223145,
+      "learning_rate": 3.125e-07,
+      "loss": 0.4325,
+      "step": 3875
+    },
+    {
+      "epoch": 19.38,
+      "grad_norm": 2.383394956588745,
+      "learning_rate": 3.1000000000000005e-07,
+      "loss": 0.1459,
+      "step": 3876
+    },
+    {
+      "epoch": 19.385,
+      "grad_norm": 4.266153812408447,
+      "learning_rate": 3.075e-07,
+      "loss": 0.1047,
+      "step": 3877
+    },
+    {
+      "epoch": 19.39,
+      "grad_norm": 6.467111110687256,
+      "learning_rate": 3.0500000000000004e-07,
+      "loss": 0.4929,
+      "step": 3878
+    },
+    {
+      "epoch": 19.395,
+      "grad_norm": 6.976054668426514,
+      "learning_rate": 3.025e-07,
+      "loss": 0.475,
+      "step": 3879
+    },
+    {
+      "epoch": 19.4,
+      "grad_norm": 6.112469673156738,
+      "learning_rate": 3.0000000000000004e-07,
+      "loss": 0.3065,
+      "step": 3880
+    },
+    {
+      "epoch": 19.405,
+      "grad_norm": 4.561709880828857,
+      "learning_rate": 2.975e-07,
+      "loss": 0.1483,
+      "step": 3881
+    },
+    {
+      "epoch": 19.41,
+      "grad_norm": 4.0725321769714355,
+      "learning_rate": 2.9500000000000003e-07,
+      "loss": 0.414,
+      "step": 3882
+    },
+    {
+      "epoch": 19.415,
+      "grad_norm": 3.869568109512329,
+      "learning_rate": 2.9250000000000006e-07,
+      "loss": 0.1958,
+      "step": 3883
+    },
+    {
+      "epoch": 19.42,
+      "grad_norm": 3.5765485763549805,
+      "learning_rate": 2.9000000000000003e-07,
+      "loss": 0.365,
+      "step": 3884
+    },
+    {
+      "epoch": 19.425,
+      "grad_norm": 4.512336254119873,
+      "learning_rate": 2.8750000000000005e-07,
+      "loss": 0.283,
+      "step": 3885
+    },
+    {
+      "epoch": 19.43,
+      "grad_norm": 6.432215213775635,
+      "learning_rate": 2.85e-07,
+      "loss": 0.3383,
+      "step": 3886
+    },
+    {
+      "epoch": 19.435,
+      "grad_norm": 3.865532159805298,
+      "learning_rate": 2.8250000000000005e-07,
+      "loss": 0.12,
+      "step": 3887
+    },
+    {
+      "epoch": 19.44,
+      "grad_norm": 4.304607391357422,
+      "learning_rate": 2.8e-07,
+      "loss": 0.2481,
+      "step": 3888
+    },
+    {
+      "epoch": 19.445,
+      "grad_norm": 3.79744553565979,
+      "learning_rate": 2.7750000000000004e-07,
+      "loss": 0.3466,
+      "step": 3889
+    },
+    {
+      "epoch": 19.45,
+      "grad_norm": 3.7063965797424316,
+      "learning_rate": 2.75e-07,
+      "loss": 0.1633,
+      "step": 3890
+    },
+    {
+      "epoch": 19.455,
+      "grad_norm": 5.186378479003906,
+      "learning_rate": 2.7250000000000004e-07,
+      "loss": 0.1914,
+      "step": 3891
+    },
+    {
+      "epoch": 19.46,
+      "grad_norm": 2.6742851734161377,
+      "learning_rate": 2.7e-07,
+      "loss": 0.1457,
+      "step": 3892
+    },
+    {
+      "epoch": 19.465,
+      "grad_norm": 2.9154419898986816,
+      "learning_rate": 2.6750000000000003e-07,
+      "loss": 0.1249,
+      "step": 3893
+    },
+    {
+      "epoch": 19.47,
+      "grad_norm": 3.163067579269409,
+      "learning_rate": 2.65e-07,
+      "loss": 0.2153,
+      "step": 3894
+    },
+    {
+      "epoch": 19.475,
+      "grad_norm": 2.653581142425537,
+      "learning_rate": 2.6250000000000003e-07,
+      "loss": 0.0884,
+      "step": 3895
+    },
+    {
+      "epoch": 19.48,
+      "grad_norm": 3.3983139991760254,
+      "learning_rate": 2.6e-07,
+      "loss": 0.2836,
+      "step": 3896
+    },
+    {
+      "epoch": 19.485,
+      "grad_norm": 3.3833844661712646,
+      "learning_rate": 2.575e-07,
+      "loss": 0.218,
+      "step": 3897
+    },
+    {
+      "epoch": 19.49,
+      "grad_norm": 4.85869026184082,
+      "learning_rate": 2.55e-07,
+      "loss": 0.3032,
+      "step": 3898
+    },
+    {
+      "epoch": 19.495,
+      "grad_norm": 5.693041801452637,
+      "learning_rate": 2.525e-07,
+      "loss": 0.3203,
+      "step": 3899
+    },
+    {
+      "epoch": 19.5,
+      "grad_norm": 2.181387186050415,
+      "learning_rate": 2.5000000000000004e-07,
+      "loss": 0.0688,
+      "step": 3900
+    },
+    {
+      "epoch": 19.505,
+      "grad_norm": 3.104853391647339,
+      "learning_rate": 2.475e-07,
+      "loss": 0.33,
+      "step": 3901
+    },
+    {
+      "epoch": 19.51,
+      "grad_norm": 2.6698997020721436,
+      "learning_rate": 2.4500000000000004e-07,
+      "loss": 0.214,
+      "step": 3902
+    },
+    {
+      "epoch": 19.515,
+      "grad_norm": 2.098999261856079,
+      "learning_rate": 2.425e-07,
+      "loss": 0.1977,
+      "step": 3903
+    },
+    {
+      "epoch": 19.52,
+      "grad_norm": 13.75471305847168,
+      "learning_rate": 2.4000000000000003e-07,
+      "loss": 0.7372,
+      "step": 3904
+    },
+    {
+      "epoch": 19.525,
+      "grad_norm": 6.624691486358643,
+      "learning_rate": 2.3750000000000003e-07,
+      "loss": 0.4542,
+      "step": 3905
+    },
+    {
+      "epoch": 19.53,
+      "grad_norm": 6.444518089294434,
+      "learning_rate": 2.3500000000000003e-07,
+      "loss": 0.354,
+      "step": 3906
+    },
+    {
+      "epoch": 19.535,
+      "grad_norm": 4.935043811798096,
+      "learning_rate": 2.3250000000000002e-07,
+      "loss": 0.2338,
+      "step": 3907
+    },
+    {
+      "epoch": 19.54,
+      "grad_norm": 4.317294120788574,
+      "learning_rate": 2.3000000000000002e-07,
+      "loss": 0.2291,
+      "step": 3908
+    },
+    {
+      "epoch": 19.545,
+      "grad_norm": 21.709857940673828,
+      "learning_rate": 2.2750000000000002e-07,
+      "loss": 0.7788,
+      "step": 3909
+    },
+    {
+      "epoch": 19.55,
+      "grad_norm": 5.721822261810303,
+      "learning_rate": 2.2500000000000002e-07,
+      "loss": 0.2782,
+      "step": 3910
+    },
+    {
+      "epoch": 19.555,
+      "grad_norm": 3.9527273178100586,
+      "learning_rate": 2.2250000000000001e-07,
+      "loss": 0.3124,
+      "step": 3911
+    },
+    {
+      "epoch": 19.56,
+      "grad_norm": 7.8763580322265625,
+      "learning_rate": 2.2e-07,
+      "loss": 0.4035,
+      "step": 3912
+    },
+    {
+      "epoch": 19.565,
+      "grad_norm": 3.761298179626465,
+      "learning_rate": 2.175e-07,
+      "loss": 0.1965,
+      "step": 3913
+    },
+    {
+      "epoch": 19.57,
+      "grad_norm": 3.252838373184204,
+      "learning_rate": 2.15e-07,
+      "loss": 0.1834,
+      "step": 3914
+    },
+    {
+      "epoch": 19.575,
+      "grad_norm": 3.480520009994507,
+      "learning_rate": 2.1250000000000003e-07,
+      "loss": 0.131,
+      "step": 3915
+    },
+    {
+      "epoch": 19.58,
+      "grad_norm": 2.271131753921509,
+      "learning_rate": 2.1000000000000003e-07,
+      "loss": 0.1192,
+      "step": 3916
+    },
+    {
+      "epoch": 19.585,
+      "grad_norm": 3.336682081222534,
+      "learning_rate": 2.0750000000000003e-07,
+      "loss": 0.2294,
+      "step": 3917
+    },
+    {
+      "epoch": 19.59,
+      "grad_norm": 3.378654718399048,
+      "learning_rate": 2.0500000000000002e-07,
+      "loss": 0.134,
+      "step": 3918
+    },
+    {
+      "epoch": 19.595,
+      "grad_norm": 5.437769889831543,
+      "learning_rate": 2.0250000000000002e-07,
+      "loss": 0.2332,
+      "step": 3919
+    },
+    {
+      "epoch": 19.6,
+      "grad_norm": 5.206194877624512,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 0.3686,
+      "step": 3920
+    },
+    {
+      "epoch": 19.605,
+      "grad_norm": 4.42988920211792,
+      "learning_rate": 1.9750000000000001e-07,
+      "loss": 0.621,
+      "step": 3921
+    },
+    {
+      "epoch": 19.61,
+      "grad_norm": 3.259613275527954,
+      "learning_rate": 1.95e-07,
+      "loss": 0.336,
+      "step": 3922
+    },
+    {
+      "epoch": 19.615,
+      "grad_norm": 3.7597978115081787,
+      "learning_rate": 1.925e-07,
+      "loss": 0.0854,
+      "step": 3923
+    },
+    {
+      "epoch": 19.62,
+      "grad_norm": 7.700922012329102,
+      "learning_rate": 1.9e-07,
+      "loss": 0.4735,
+      "step": 3924
+    },
+    {
+      "epoch": 19.625,
+      "grad_norm": 4.547913551330566,
+      "learning_rate": 1.875e-07,
+      "loss": 0.2563,
+      "step": 3925
+    },
+    {
+      "epoch": 19.63,
+      "grad_norm": 4.393975257873535,
+      "learning_rate": 1.85e-07,
+      "loss": 0.1534,
+      "step": 3926
+    },
+    {
+      "epoch": 19.635,
+      "grad_norm": 3.716777801513672,
+      "learning_rate": 1.825e-07,
+      "loss": 0.2132,
+      "step": 3927
+    },
+    {
+      "epoch": 19.64,
+      "grad_norm": 4.547081470489502,
+      "learning_rate": 1.8e-07,
+      "loss": 0.3739,
+      "step": 3928
+    },
+    {
+      "epoch": 19.645,
+      "grad_norm": 3.9950966835021973,
+      "learning_rate": 1.775e-07,
+      "loss": 0.1936,
+      "step": 3929
+    },
+    {
+      "epoch": 19.65,
+      "grad_norm": 3.5846245288848877,
+      "learning_rate": 1.7500000000000002e-07,
+      "loss": 0.2484,
+      "step": 3930
+    },
+    {
+      "epoch": 19.655,
+      "grad_norm": 4.755615711212158,
+      "learning_rate": 1.7250000000000002e-07,
+      "loss": 0.3243,
+      "step": 3931
+    },
+    {
+      "epoch": 19.66,
+      "grad_norm": 3.9170844554901123,
+      "learning_rate": 1.7000000000000001e-07,
+      "loss": 0.1503,
+      "step": 3932
+    },
+    {
+      "epoch": 19.665,
+      "grad_norm": 2.588578701019287,
+      "learning_rate": 1.675e-07,
+      "loss": 0.2382,
+      "step": 3933
+    },
+    {
+      "epoch": 19.67,
+      "grad_norm": 5.185120582580566,
+      "learning_rate": 1.65e-07,
+      "loss": 0.644,
+      "step": 3934
+    },
+    {
+      "epoch": 19.675,
+      "grad_norm": 3.0452635288238525,
+      "learning_rate": 1.625e-07,
+      "loss": 0.1195,
+      "step": 3935
+    },
+    {
+      "epoch": 19.68,
+      "grad_norm": 5.1596360206604,
+      "learning_rate": 1.6e-07,
+      "loss": 0.2444,
+      "step": 3936
+    },
+    {
+      "epoch": 19.685,
+      "grad_norm": 7.291883945465088,
+      "learning_rate": 1.575e-07,
+      "loss": 0.4891,
+      "step": 3937
+    },
+    {
+      "epoch": 19.69,
+      "grad_norm": 3.4564836025238037,
+      "learning_rate": 1.5500000000000002e-07,
+      "loss": 0.3225,
+      "step": 3938
+    },
+    {
+      "epoch": 19.695,
+      "grad_norm": 2.5129737854003906,
+      "learning_rate": 1.5250000000000002e-07,
+      "loss": 0.2489,
+      "step": 3939
+    },
+    {
+      "epoch": 19.7,
+      "grad_norm": 3.22450590133667,
+      "learning_rate": 1.5000000000000002e-07,
+      "loss": 0.2791,
+      "step": 3940
+    },
+    {
+      "epoch": 19.705,
+      "grad_norm": 3.476330041885376,
+      "learning_rate": 1.4750000000000002e-07,
+      "loss": 0.2034,
+      "step": 3941
+    },
+    {
+      "epoch": 19.71,
+      "grad_norm": 1.868434190750122,
+      "learning_rate": 1.4500000000000001e-07,
+      "loss": 0.0979,
+      "step": 3942
+    },
+    {
+      "epoch": 19.715,
+      "grad_norm": 4.662663459777832,
+      "learning_rate": 1.425e-07,
+      "loss": 0.1609,
+      "step": 3943
+    },
+    {
+      "epoch": 19.72,
+      "grad_norm": 3.34397292137146,
+      "learning_rate": 1.4e-07,
+      "loss": 0.17,
+      "step": 3944
+    },
+    {
+      "epoch": 19.725,
+      "grad_norm": 3.977994203567505,
+      "learning_rate": 1.375e-07,
+      "loss": 0.1536,
+      "step": 3945
+    },
+    {
+      "epoch": 19.73,
+      "grad_norm": 3.9934370517730713,
+      "learning_rate": 1.35e-07,
+      "loss": 0.3025,
+      "step": 3946
+    },
+    {
+      "epoch": 19.735,
+      "grad_norm": 8.027313232421875,
+      "learning_rate": 1.325e-07,
+      "loss": 0.3615,
+      "step": 3947
+    },
+    {
+      "epoch": 19.74,
+      "grad_norm": 6.847579002380371,
+      "learning_rate": 1.3e-07,
+      "loss": 0.4189,
+      "step": 3948
+    },
+    {
+      "epoch": 19.745,
+      "grad_norm": 5.594716548919678,
+      "learning_rate": 1.275e-07,
+      "loss": 0.5431,
+      "step": 3949
+    },
+    {
+      "epoch": 19.75,
+      "grad_norm": 5.246018409729004,
+      "learning_rate": 1.2500000000000002e-07,
+      "loss": 0.6131,
+      "step": 3950
+    },
+    {
+      "epoch": 19.755,
+      "grad_norm": 7.684979438781738,
+      "learning_rate": 1.2250000000000002e-07,
+      "loss": 0.4145,
+      "step": 3951
+    },
+    {
+      "epoch": 19.76,
+      "grad_norm": 2.6393275260925293,
+      "learning_rate": 1.2000000000000002e-07,
+      "loss": 0.1334,
+      "step": 3952
+    },
+    {
+      "epoch": 19.765,
+      "grad_norm": 3.8045854568481445,
+      "learning_rate": 1.1750000000000001e-07,
+      "loss": 0.1746,
+      "step": 3953
+    },
+    {
+      "epoch": 19.77,
+      "grad_norm": 4.224969387054443,
+      "learning_rate": 1.1500000000000001e-07,
+      "loss": 0.2748,
+      "step": 3954
+    },
+    {
+      "epoch": 19.775,
+      "grad_norm": 10.931328773498535,
+      "learning_rate": 1.1250000000000001e-07,
+      "loss": 0.5587,
+      "step": 3955
+    },
+    {
+      "epoch": 19.78,
+      "grad_norm": 3.6320457458496094,
+      "learning_rate": 1.1e-07,
+      "loss": 0.5001,
+      "step": 3956
+    },
+    {
+      "epoch": 19.785,
+      "grad_norm": 12.500187873840332,
+      "learning_rate": 1.075e-07,
+      "loss": 0.4597,
+      "step": 3957
+    },
+    {
+      "epoch": 19.79,
+      "grad_norm": 4.303757190704346,
+      "learning_rate": 1.0500000000000001e-07,
+      "loss": 0.5066,
+      "step": 3958
+    },
+    {
+      "epoch": 19.795,
+      "grad_norm": 4.641019344329834,
+      "learning_rate": 1.0250000000000001e-07,
+      "loss": 0.4497,
+      "step": 3959
+    },
+    {
+      "epoch": 19.8,
+      "grad_norm": 3.0587615966796875,
+      "learning_rate": 1.0000000000000001e-07,
+      "loss": 0.3665,
+      "step": 3960
+    },
+    {
+      "epoch": 19.805,
+      "grad_norm": 4.5718231201171875,
+      "learning_rate": 9.75e-08,
+      "loss": 0.2832,
+      "step": 3961
+    },
+    {
+      "epoch": 19.81,
+      "grad_norm": 5.733880519866943,
+      "learning_rate": 9.5e-08,
+      "loss": 0.2934,
+      "step": 3962
+    },
+    {
+      "epoch": 19.815,
+      "grad_norm": 4.962919235229492,
+      "learning_rate": 9.25e-08,
+      "loss": 0.4046,
+      "step": 3963
+    },
+    {
+      "epoch": 19.82,
+      "grad_norm": 4.370629787445068,
+      "learning_rate": 9e-08,
+      "loss": 0.266,
+      "step": 3964
+    },
+    {
+      "epoch": 19.825,
+      "grad_norm": 3.665912389755249,
+      "learning_rate": 8.750000000000001e-08,
+      "loss": 0.3129,
+      "step": 3965
+    },
+    {
+      "epoch": 19.83,
+      "grad_norm": 3.1528871059417725,
+      "learning_rate": 8.500000000000001e-08,
+      "loss": 0.1119,
+      "step": 3966
+    },
+    {
+      "epoch": 19.835,
+      "grad_norm": 3.492469072341919,
+      "learning_rate": 8.25e-08,
+      "loss": 0.3586,
+      "step": 3967
+    },
+    {
+      "epoch": 19.84,
+      "grad_norm": 2.070781707763672,
+      "learning_rate": 8e-08,
+      "loss": 0.0334,
+      "step": 3968
+    },
+    {
+      "epoch": 19.845,
+      "grad_norm": 3.790229320526123,
+      "learning_rate": 7.750000000000001e-08,
+      "loss": 0.3649,
+      "step": 3969
+    },
+    {
+      "epoch": 19.85,
+      "grad_norm": 5.246890068054199,
+      "learning_rate": 7.500000000000001e-08,
+      "loss": 0.2737,
+      "step": 3970
+    },
+    {
+      "epoch": 19.855,
+      "grad_norm": 3.5979814529418945,
+      "learning_rate": 7.250000000000001e-08,
+      "loss": 0.2784,
+      "step": 3971
+    },
+    {
+      "epoch": 19.86,
+      "grad_norm": 74.49684143066406,
+      "learning_rate": 7e-08,
+      "loss": 0.916,
+      "step": 3972
+    },
+    {
+      "epoch": 19.865,
+      "grad_norm": 5.321796894073486,
+      "learning_rate": 6.75e-08,
+      "loss": 0.435,
+      "step": 3973
+    },
+    {
+      "epoch": 19.87,
+      "grad_norm": 6.188290119171143,
+      "learning_rate": 6.5e-08,
+      "loss": 0.77,
+      "step": 3974
+    },
+    {
+      "epoch": 19.875,
+      "grad_norm": 5.913536548614502,
+      "learning_rate": 6.250000000000001e-08,
+      "loss": 0.1588,
+      "step": 3975
+    },
+    {
+      "epoch": 19.88,
+      "grad_norm": 3.6910040378570557,
+      "learning_rate": 6.000000000000001e-08,
+      "loss": 0.3125,
+      "step": 3976
+    },
+    {
+      "epoch": 19.885,
+      "grad_norm": 6.543662071228027,
+      "learning_rate": 5.7500000000000005e-08,
+      "loss": 0.3236,
+      "step": 3977
+    },
+    {
+      "epoch": 19.89,
+      "grad_norm": 7.070521831512451,
+      "learning_rate": 5.5e-08,
+      "loss": 0.3209,
+      "step": 3978
+    },
+    {
+      "epoch": 19.895,
+      "grad_norm": 4.187525272369385,
+      "learning_rate": 5.250000000000001e-08,
+      "loss": 0.1418,
+      "step": 3979
+    },
+    {
+      "epoch": 19.9,
+      "grad_norm": 3.9989395141601562,
+      "learning_rate": 5.0000000000000004e-08,
+      "loss": 0.3331,
+      "step": 3980
+    },
+    {
+      "epoch": 19.905,
+      "grad_norm": 5.990621089935303,
+      "learning_rate": 4.75e-08,
+      "loss": 0.4282,
+      "step": 3981
+    },
+    {
+      "epoch": 19.91,
+      "grad_norm": 1.9905527830123901,
+      "learning_rate": 4.5e-08,
+      "loss": 0.0263,
+      "step": 3982
+    },
+    {
+      "epoch": 19.915,
+      "grad_norm": 3.465538263320923,
+      "learning_rate": 4.2500000000000003e-08,
+      "loss": 0.2468,
+      "step": 3983
+    },
+    {
+      "epoch": 19.92,
+      "grad_norm": 2.7771060466766357,
+      "learning_rate": 4e-08,
+      "loss": 0.1053,
+      "step": 3984
+    },
+    {
+      "epoch": 19.925,
+      "grad_norm": 4.31032657623291,
+      "learning_rate": 3.7500000000000005e-08,
+      "loss": 0.1662,
+      "step": 3985
+    },
+    {
+      "epoch": 19.93,
+      "grad_norm": 5.227555751800537,
+      "learning_rate": 3.5e-08,
+      "loss": 0.2419,
+      "step": 3986
+    },
+    {
+      "epoch": 19.935,
+      "grad_norm": 10.32761287689209,
+      "learning_rate": 3.25e-08,
+      "loss": 1.1244,
+      "step": 3987
+    },
+    {
+      "epoch": 19.94,
+      "grad_norm": 3.3771324157714844,
+      "learning_rate": 3.0000000000000004e-08,
+      "loss": 0.2004,
+      "step": 3988
+    },
+    {
+      "epoch": 19.945,
+      "grad_norm": 5.193995952606201,
+      "learning_rate": 2.75e-08,
+      "loss": 0.4238,
+      "step": 3989
+    },
+    {
+      "epoch": 19.95,
+      "grad_norm": 7.835831642150879,
+      "learning_rate": 2.5000000000000002e-08,
+      "loss": 0.1367,
+      "step": 3990
+    },
+    {
+      "epoch": 19.955,
+      "grad_norm": 2.9515206813812256,
+      "learning_rate": 2.25e-08,
+      "loss": 0.1653,
+      "step": 3991
+    },
+    {
+      "epoch": 19.96,
+      "grad_norm": 3.9092938899993896,
+      "learning_rate": 2e-08,
+      "loss": 0.1443,
+      "step": 3992
+    },
+    {
+      "epoch": 19.965,
+      "grad_norm": 5.684747219085693,
+      "learning_rate": 1.75e-08,
+      "loss": 0.4249,
+      "step": 3993
+    },
+    {
+      "epoch": 19.97,
+      "grad_norm": 3.9162096977233887,
+      "learning_rate": 1.5000000000000002e-08,
+      "loss": 0.2724,
+      "step": 3994
+    },
+    {
+      "epoch": 19.975,
+      "grad_norm": 5.637531757354736,
+      "learning_rate": 1.2500000000000001e-08,
+      "loss": 0.3384,
+      "step": 3995
+    },
+    {
+      "epoch": 19.98,
+      "grad_norm": 12.666444778442383,
+      "learning_rate": 1e-08,
+      "loss": 0.472,
+      "step": 3996
+    },
+    {
+      "epoch": 19.985,
+      "grad_norm": 4.503374099731445,
+      "learning_rate": 7.500000000000001e-09,
+      "loss": 0.4283,
+      "step": 3997
+    },
+    {
+      "epoch": 19.99,
+      "grad_norm": 5.100307464599609,
+      "learning_rate": 5e-09,
+      "loss": 0.5686,
+      "step": 3998
+    },
+    {
+      "epoch": 19.995,
+      "grad_norm": 4.2250776290893555,
+      "learning_rate": 2.5e-09,
+      "loss": 0.4675,
+      "step": 3999
+    },
+    {
+      "epoch": 20.0,
+      "grad_norm": 9.18569278717041,
+      "learning_rate": 0.0,
+      "loss": 0.3562,
+      "step": 4000
+    },
+    {
+      "epoch": 20.0,
+      "eval_loss": 0.18977239727973938,
+      "eval_runtime": 15.3656,
+      "eval_samples_per_second": 51.934,
+      "eval_steps_per_second": 6.508,
+      "step": 4000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 4000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 20,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 768426141450240.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}