diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,10542 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 1500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.002,
+      "grad_norm": 226.99563598632812,
+      "learning_rate": 6.666666666666668e-08,
+      "loss": 21.9448,
+      "step": 1
+    },
+    {
+      "epoch": 0.004,
+      "grad_norm": 236.27398681640625,
+      "learning_rate": 1.3333333333333336e-07,
+      "loss": 22.6687,
+      "step": 2
+    },
+    {
+      "epoch": 0.006,
+      "grad_norm": 243.7906036376953,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 23.7182,
+      "step": 3
+    },
+    {
+      "epoch": 0.008,
+      "grad_norm": 233.99176025390625,
+      "learning_rate": 2.666666666666667e-07,
+      "loss": 22.4062,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 247.73858642578125,
+      "learning_rate": 3.3333333333333335e-07,
+      "loss": 23.4446,
+      "step": 5
+    },
+    {
+      "epoch": 0.012,
+      "grad_norm": 224.78480529785156,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 22.0816,
+      "step": 6
+    },
+    {
+      "epoch": 0.014,
+      "grad_norm": 239.08424377441406,
+      "learning_rate": 4.666666666666667e-07,
+      "loss": 22.2544,
+      "step": 7
+    },
+    {
+      "epoch": 0.016,
+      "grad_norm": 243.3548126220703,
+      "learning_rate": 5.333333333333335e-07,
+      "loss": 22.3676,
+      "step": 8
+    },
+    {
+      "epoch": 0.018,
+      "grad_norm": 228.88418579101562,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 21.8324,
+      "step": 9
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 255.19561767578125,
+      "learning_rate": 6.666666666666667e-07,
+      "loss": 21.7817,
+      "step": 10
+    },
+    {
+      "epoch": 0.022,
+      "grad_norm": 256.03631591796875,
+      "learning_rate": 7.333333333333334e-07,
+      "loss": 22.0901,
+      "step": 11
+    },
+    {
+      "epoch": 0.024,
+      "grad_norm": 256.8511047363281,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 22.1082,
+      "step": 12
+    },
+    {
+      "epoch": 0.026,
+      "grad_norm": 276.2884521484375,
+      "learning_rate": 8.666666666666668e-07,
+      "loss": 20.2445,
+      "step": 13
+    },
+    {
+      "epoch": 0.028,
+      "grad_norm": 274.6839599609375,
+      "learning_rate": 9.333333333333334e-07,
+      "loss": 18.7186,
+      "step": 14
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 282.9738464355469,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 19.2151,
+      "step": 15
+    },
+    {
+      "epoch": 0.032,
+      "grad_norm": 273.26934814453125,
+      "learning_rate": 1.066666666666667e-06,
+      "loss": 18.8715,
+      "step": 16
+    },
+    {
+      "epoch": 0.034,
+      "grad_norm": 222.71524047851562,
+      "learning_rate": 1.1333333333333334e-06,
+      "loss": 14.6642,
+      "step": 17
+    },
+    {
+      "epoch": 0.036,
+      "grad_norm": 218.20606994628906,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 13.549,
+      "step": 18
+    },
+    {
+      "epoch": 0.038,
+      "grad_norm": 229.51235961914062,
+      "learning_rate": 1.2666666666666669e-06,
+      "loss": 13.3927,
+      "step": 19
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 228.68344116210938,
+      "learning_rate": 1.3333333333333334e-06,
+      "loss": 12.2413,
+      "step": 20
+    },
+    {
+      "epoch": 0.042,
+      "grad_norm": 222.99270629882812,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 11.1357,
+      "step": 21
+    },
+    {
+      "epoch": 0.044,
+      "grad_norm": 236.81912231445312,
+      "learning_rate": 1.4666666666666669e-06,
+      "loss": 10.6309,
+      "step": 22
+    },
+    {
+      "epoch": 0.046,
+      "grad_norm": 253.3773193359375,
+      "learning_rate": 1.5333333333333334e-06,
+      "loss": 6.1736,
+      "step": 23
+    },
+    {
+      "epoch": 0.048,
+      "grad_norm": 132.65660095214844,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 3.7401,
+      "step": 24
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 171.21932983398438,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 3.2601,
+      "step": 25
+    },
+    {
+      "epoch": 0.052,
+      "grad_norm": 112.79869079589844,
+      "learning_rate": 1.7333333333333336e-06,
+      "loss": 2.6378,
+      "step": 26
+    },
+    {
+      "epoch": 0.054,
+      "grad_norm": 72.08856201171875,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 2.2122,
+      "step": 27
+    },
+    {
+      "epoch": 0.056,
+      "grad_norm": 47.903202056884766,
+      "learning_rate": 1.8666666666666669e-06,
+      "loss": 1.9545,
+      "step": 28
+    },
+    {
+      "epoch": 0.058,
+      "grad_norm": 42.68925094604492,
+      "learning_rate": 1.9333333333333336e-06,
+      "loss": 1.909,
+      "step": 29
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 37.55009460449219,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.7777,
+      "step": 30
+    },
+    {
+      "epoch": 0.062,
+      "grad_norm": 27.396568298339844,
+      "learning_rate": 2.0666666666666666e-06,
+      "loss": 1.8359,
+      "step": 31
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 30.148645401000977,
+      "learning_rate": 2.133333333333334e-06,
+      "loss": 1.7342,
+      "step": 32
+    },
+    {
+      "epoch": 0.066,
+      "grad_norm": 28.60604476928711,
+      "learning_rate": 2.2e-06,
+      "loss": 1.8821,
+      "step": 33
+    },
+    {
+      "epoch": 0.068,
+      "grad_norm": 22.30388641357422,
+      "learning_rate": 2.266666666666667e-06,
+      "loss": 1.6301,
+      "step": 34
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 28.496145248413086,
+      "learning_rate": 2.3333333333333336e-06,
+      "loss": 1.8361,
+      "step": 35
+    },
+    {
+      "epoch": 0.072,
+      "grad_norm": 34.518287658691406,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 1.6929,
+      "step": 36
+    },
+    {
+      "epoch": 0.074,
+      "grad_norm": 23.480363845825195,
+      "learning_rate": 2.466666666666667e-06,
+      "loss": 1.5912,
+      "step": 37
+    },
+    {
+      "epoch": 0.076,
+      "grad_norm": 26.371000289916992,
+      "learning_rate": 2.5333333333333338e-06,
+      "loss": 1.6452,
+      "step": 38
+    },
+    {
+      "epoch": 0.078,
+      "grad_norm": 26.6358585357666,
+      "learning_rate": 2.6e-06,
+      "loss": 1.7357,
+      "step": 39
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 21.163368225097656,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": 1.8135,
+      "step": 40
+    },
+    {
+      "epoch": 0.082,
+      "grad_norm": 38.06758499145508,
+      "learning_rate": 2.7333333333333336e-06,
+      "loss": 1.7483,
+      "step": 41
+    },
+    {
+      "epoch": 0.084,
+      "grad_norm": 20.108041763305664,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 1.7272,
+      "step": 42
+    },
+    {
+      "epoch": 0.086,
+      "grad_norm": 37.66496658325195,
+      "learning_rate": 2.866666666666667e-06,
+      "loss": 1.8263,
+      "step": 43
+    },
+    {
+      "epoch": 0.088,
+      "grad_norm": 17.51531219482422,
+      "learning_rate": 2.9333333333333338e-06,
+      "loss": 1.7223,
+      "step": 44
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 22.1541748046875,
+      "learning_rate": 3e-06,
+      "loss": 1.6006,
+      "step": 45
+    },
+    {
+      "epoch": 0.092,
+      "grad_norm": 19.2305908203125,
+      "learning_rate": 3.066666666666667e-06,
+      "loss": 1.6562,
+      "step": 46
+    },
+    {
+      "epoch": 0.094,
+      "grad_norm": 16.655160903930664,
+      "learning_rate": 3.133333333333334e-06,
+      "loss": 1.7491,
+      "step": 47
+    },
+    {
+      "epoch": 0.096,
+      "grad_norm": 20.247623443603516,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 1.6449,
+      "step": 48
+    },
+    {
+      "epoch": 0.098,
+      "grad_norm": 17.47651481628418,
+      "learning_rate": 3.266666666666667e-06,
+      "loss": 1.6761,
+      "step": 49
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 26.51736068725586,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.7986,
+      "step": 50
+    },
+    {
+      "epoch": 0.102,
+      "grad_norm": 21.225101470947266,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 1.6329,
+      "step": 51
+    },
+    {
+      "epoch": 0.104,
+      "grad_norm": 22.989187240600586,
+      "learning_rate": 3.4666666666666672e-06,
+      "loss": 1.7155,
+      "step": 52
+    },
+    {
+      "epoch": 0.106,
+      "grad_norm": 18.516536712646484,
+      "learning_rate": 3.5333333333333335e-06,
+      "loss": 1.6831,
+      "step": 53
+    },
+    {
+      "epoch": 0.108,
+      "grad_norm": 19.069442749023438,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 1.6225,
+      "step": 54
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 20.60556983947754,
+      "learning_rate": 3.6666666666666666e-06,
+      "loss": 1.6726,
+      "step": 55
+    },
+    {
+      "epoch": 0.112,
+      "grad_norm": 19.45859146118164,
+      "learning_rate": 3.7333333333333337e-06,
+      "loss": 1.6408,
+      "step": 56
+    },
+    {
+      "epoch": 0.114,
+      "grad_norm": 26.011234283447266,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 1.6057,
+      "step": 57
+    },
+    {
+      "epoch": 0.116,
+      "grad_norm": 17.240083694458008,
+      "learning_rate": 3.866666666666667e-06,
+      "loss": 1.677,
+      "step": 58
+    },
+    {
+      "epoch": 0.118,
+      "grad_norm": 22.089248657226562,
+      "learning_rate": 3.9333333333333335e-06,
+      "loss": 1.676,
+      "step": 59
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 33.735233306884766,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.6742,
+      "step": 60
+    },
+    {
+      "epoch": 0.122,
+      "grad_norm": 27.743314743041992,
+      "learning_rate": 4.066666666666667e-06,
+      "loss": 1.6956,
+      "step": 61
+    },
+    {
+      "epoch": 0.124,
+      "grad_norm": 25.890151977539062,
+      "learning_rate": 4.133333333333333e-06,
+      "loss": 1.56,
+      "step": 62
+    },
+    {
+      "epoch": 0.126,
+      "grad_norm": 16.895370483398438,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 1.5794,
+      "step": 63
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 19.82956314086914,
+      "learning_rate": 4.266666666666668e-06,
+      "loss": 1.4879,
+      "step": 64
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 15.571148872375488,
+      "learning_rate": 4.333333333333334e-06,
+      "loss": 1.7724,
+      "step": 65
+    },
+    {
+      "epoch": 0.132,
+      "grad_norm": 14.684285163879395,
+      "learning_rate": 4.4e-06,
+      "loss": 1.6943,
+      "step": 66
+    },
+    {
+      "epoch": 0.134,
+      "grad_norm": 13.911795616149902,
+      "learning_rate": 4.4666666666666665e-06,
+      "loss": 1.6627,
+      "step": 67
+    },
+    {
+      "epoch": 0.136,
+      "grad_norm": 15.752901077270508,
+      "learning_rate": 4.533333333333334e-06,
+      "loss": 1.7536,
+      "step": 68
+    },
+    {
+      "epoch": 0.138,
+      "grad_norm": 24.22060203552246,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 1.6225,
+      "step": 69
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.675880432128906,
+      "learning_rate": 4.666666666666667e-06,
+      "loss": 1.6696,
+      "step": 70
+    },
+    {
+      "epoch": 0.142,
+      "grad_norm": 20.69804573059082,
+      "learning_rate": 4.7333333333333335e-06,
+      "loss": 1.6295,
+      "step": 71
+    },
+    {
+      "epoch": 0.144,
+      "grad_norm": 28.758134841918945,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 1.6698,
+      "step": 72
+    },
+    {
+      "epoch": 0.146,
+      "grad_norm": 14.652379035949707,
+      "learning_rate": 4.866666666666667e-06,
+      "loss": 1.606,
+      "step": 73
+    },
+    {
+      "epoch": 0.148,
+      "grad_norm": 22.868322372436523,
+      "learning_rate": 4.933333333333334e-06,
+      "loss": 1.6168,
+      "step": 74
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 14.871725082397461,
+      "learning_rate": 5e-06,
+      "loss": 1.6506,
+      "step": 75
+    },
+    {
+      "epoch": 0.152,
+      "grad_norm": 21.552318572998047,
+      "learning_rate": 5.0666666666666676e-06,
+      "loss": 1.6366,
+      "step": 76
+    },
+    {
+      "epoch": 0.154,
+      "grad_norm": 15.857409477233887,
+      "learning_rate": 5.133333333333334e-06,
+      "loss": 1.6794,
+      "step": 77
+    },
+    {
+      "epoch": 0.156,
+      "grad_norm": 17.261104583740234,
+      "learning_rate": 5.2e-06,
+      "loss": 1.6165,
+      "step": 78
+    },
+    {
+      "epoch": 0.158,
+      "grad_norm": 14.66672420501709,
+      "learning_rate": 5.2666666666666665e-06,
+      "loss": 1.6562,
+      "step": 79
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 22.905376434326172,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 1.6468,
+      "step": 80
+    },
+    {
+      "epoch": 0.162,
+      "grad_norm": 17.427703857421875,
+      "learning_rate": 5.400000000000001e-06,
+      "loss": 1.6618,
+      "step": 81
+    },
+    {
+      "epoch": 0.164,
+      "grad_norm": 15.554149627685547,
+      "learning_rate": 5.466666666666667e-06,
+      "loss": 1.6368,
+      "step": 82
+    },
+    {
+      "epoch": 0.166,
+      "grad_norm": 16.746028900146484,
+      "learning_rate": 5.533333333333334e-06,
+      "loss": 1.6583,
+      "step": 83
+    },
+    {
+      "epoch": 0.168,
+      "grad_norm": 17.273025512695312,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 1.625,
+      "step": 84
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 20.172046661376953,
+      "learning_rate": 5.666666666666667e-06,
+      "loss": 1.6733,
+      "step": 85
+    },
+    {
+      "epoch": 0.172,
+      "grad_norm": 13.854218482971191,
+      "learning_rate": 5.733333333333334e-06,
+      "loss": 1.649,
+      "step": 86
+    },
+    {
+      "epoch": 0.174,
+      "grad_norm": 28.214906692504883,
+      "learning_rate": 5.8e-06,
+      "loss": 1.5396,
+      "step": 87
+    },
+    {
+      "epoch": 0.176,
+      "grad_norm": 17.772258758544922,
+      "learning_rate": 5.8666666666666675e-06,
+      "loss": 1.5781,
+      "step": 88
+    },
+    {
+      "epoch": 0.178,
+      "grad_norm": 14.36558723449707,
+      "learning_rate": 5.933333333333335e-06,
+      "loss": 1.6857,
+      "step": 89
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 15.606388092041016,
+      "learning_rate": 6e-06,
+      "loss": 1.6192,
+      "step": 90
+    },
+    {
+      "epoch": 0.182,
+      "grad_norm": 16.106149673461914,
+      "learning_rate": 6.066666666666667e-06,
+      "loss": 1.568,
+      "step": 91
+    },
+    {
+      "epoch": 0.184,
+      "grad_norm": 11.074296951293945,
+      "learning_rate": 6.133333333333334e-06,
+      "loss": 1.631,
+      "step": 92
+    },
+    {
+      "epoch": 0.186,
+      "grad_norm": 15.625429153442383,
+      "learning_rate": 6.200000000000001e-06,
+      "loss": 1.6064,
+      "step": 93
+    },
+    {
+      "epoch": 0.188,
+      "grad_norm": 15.9578218460083,
+      "learning_rate": 6.266666666666668e-06,
+      "loss": 1.6646,
+      "step": 94
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 19.56605339050293,
+      "learning_rate": 6.333333333333333e-06,
+      "loss": 1.5587,
+      "step": 95
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 14.848856925964355,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 1.6036,
+      "step": 96
+    },
+    {
+      "epoch": 0.194,
+      "grad_norm": 22.246797561645508,
+      "learning_rate": 6.466666666666667e-06,
+      "loss": 1.6709,
+      "step": 97
+    },
+    {
+      "epoch": 0.196,
+      "grad_norm": 18.220890045166016,
+      "learning_rate": 6.533333333333334e-06,
+      "loss": 1.6603,
+      "step": 98
+    },
+    {
+      "epoch": 0.198,
+      "grad_norm": 18.9663028717041,
+      "learning_rate": 6.600000000000001e-06,
+      "loss": 1.7576,
+      "step": 99
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 22.304872512817383,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.7278,
+      "step": 100
+    },
+    {
+      "epoch": 0.202,
+      "grad_norm": 22.000619888305664,
+      "learning_rate": 6.733333333333334e-06,
+      "loss": 1.7162,
+      "step": 101
+    },
+    {
+      "epoch": 0.204,
+      "grad_norm": 19.02189064025879,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 1.6294,
+      "step": 102
+    },
+    {
+      "epoch": 0.206,
+      "grad_norm": 17.57056999206543,
+      "learning_rate": 6.866666666666667e-06,
+      "loss": 1.651,
+      "step": 103
+    },
+    {
+      "epoch": 0.208,
+      "grad_norm": 20.44572639465332,
+      "learning_rate": 6.9333333333333344e-06,
+      "loss": 1.6798,
+      "step": 104
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 21.554231643676758,
+      "learning_rate": 7e-06,
+      "loss": 1.676,
+      "step": 105
+    },
+    {
+      "epoch": 0.212,
+      "grad_norm": 18.977502822875977,
+      "learning_rate": 7.066666666666667e-06,
+      "loss": 1.6736,
+      "step": 106
+    },
+    {
+      "epoch": 0.214,
+      "grad_norm": 18.957950592041016,
+      "learning_rate": 7.133333333333334e-06,
+      "loss": 1.7151,
+      "step": 107
+    },
+    {
+      "epoch": 0.216,
+      "grad_norm": 19.2664852142334,
+      "learning_rate": 7.2000000000000005e-06,
+      "loss": 1.6875,
+      "step": 108
+    },
+    {
+      "epoch": 0.218,
+      "grad_norm": 18.646236419677734,
+      "learning_rate": 7.266666666666668e-06,
+      "loss": 1.6509,
+      "step": 109
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 21.31477928161621,
+      "learning_rate": 7.333333333333333e-06,
+      "loss": 1.597,
+      "step": 110
+    },
+    {
+      "epoch": 0.222,
+      "grad_norm": 17.566238403320312,
+      "learning_rate": 7.4e-06,
+      "loss": 1.7574,
+      "step": 111
+    },
+    {
+      "epoch": 0.224,
+      "grad_norm": 13.890412330627441,
+      "learning_rate": 7.4666666666666675e-06,
+      "loss": 1.5845,
+      "step": 112
+    },
+    {
+      "epoch": 0.226,
+      "grad_norm": 16.55223846435547,
+      "learning_rate": 7.533333333333334e-06,
+      "loss": 1.5888,
+      "step": 113
+    },
+    {
+      "epoch": 0.228,
+      "grad_norm": 26.499820709228516,
+      "learning_rate": 7.600000000000001e-06,
+      "loss": 1.584,
+      "step": 114
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 31.17205810546875,
+      "learning_rate": 7.666666666666667e-06,
+      "loss": 1.6443,
+      "step": 115
+    },
+    {
+      "epoch": 0.232,
+      "grad_norm": 31.60078239440918,
+      "learning_rate": 7.733333333333334e-06,
+      "loss": 1.7377,
+      "step": 116
+    },
+    {
+      "epoch": 0.234,
+      "grad_norm": 27.893613815307617,
+      "learning_rate": 7.800000000000002e-06,
+      "loss": 1.779,
+      "step": 117
+    },
+    {
+      "epoch": 0.236,
+      "grad_norm": 16.482492446899414,
+      "learning_rate": 7.866666666666667e-06,
+      "loss": 1.7002,
+      "step": 118
+    },
+    {
+      "epoch": 0.238,
+      "grad_norm": 24.259340286254883,
+      "learning_rate": 7.933333333333334e-06,
+      "loss": 1.5989,
+      "step": 119
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 18.561723709106445,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.7336,
+      "step": 120
+    },
+    {
+      "epoch": 0.242,
+      "grad_norm": 15.461645126342773,
+      "learning_rate": 8.066666666666667e-06,
+      "loss": 1.7121,
+      "step": 121
+    },
+    {
+      "epoch": 0.244,
+      "grad_norm": 22.555328369140625,
+      "learning_rate": 8.133333333333334e-06,
+      "loss": 1.7554,
+      "step": 122
+    },
+    {
+      "epoch": 0.246,
+      "grad_norm": 14.69766902923584,
+      "learning_rate": 8.2e-06,
+      "loss": 1.6397,
+      "step": 123
+    },
+    {
+      "epoch": 0.248,
+      "grad_norm": 15.5711030960083,
+      "learning_rate": 8.266666666666667e-06,
+      "loss": 1.7016,
+      "step": 124
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.80448055267334,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 1.6544,
+      "step": 125
+    },
+    {
+      "epoch": 0.252,
+      "grad_norm": 18.54620361328125,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 1.623,
+      "step": 126
+    },
+    {
+      "epoch": 0.254,
+      "grad_norm": 34.518768310546875,
+      "learning_rate": 8.466666666666668e-06,
+      "loss": 1.6958,
+      "step": 127
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 21.02113151550293,
+      "learning_rate": 8.533333333333335e-06,
+      "loss": 1.7662,
+      "step": 128
+    },
+    {
+      "epoch": 0.258,
+      "grad_norm": 14.756684303283691,
+      "learning_rate": 8.6e-06,
+      "loss": 1.5814,
+      "step": 129
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 16.498056411743164,
+      "learning_rate": 8.666666666666668e-06,
+      "loss": 1.7248,
+      "step": 130
+    },
+    {
+      "epoch": 0.262,
+      "grad_norm": 19.80323028564453,
+      "learning_rate": 8.733333333333333e-06,
+      "loss": 1.7451,
+      "step": 131
+    },
+    {
+      "epoch": 0.264,
+      "grad_norm": 19.826833724975586,
+      "learning_rate": 8.8e-06,
+      "loss": 1.6475,
+      "step": 132
+    },
+    {
+      "epoch": 0.266,
+      "grad_norm": 15.338292121887207,
+      "learning_rate": 8.866666666666668e-06,
+      "loss": 1.623,
+      "step": 133
+    },
+    {
+      "epoch": 0.268,
+      "grad_norm": 12.683643341064453,
+      "learning_rate": 8.933333333333333e-06,
+      "loss": 1.6932,
+      "step": 134
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 15.072694778442383,
+      "learning_rate": 9e-06,
+      "loss": 1.6978,
+      "step": 135
+    },
+    {
+      "epoch": 0.272,
+      "grad_norm": 17.20955467224121,
+      "learning_rate": 9.066666666666667e-06,
+      "loss": 1.6464,
+      "step": 136
+    },
+    {
+      "epoch": 0.274,
+      "grad_norm": 16.128379821777344,
+      "learning_rate": 9.133333333333335e-06,
+      "loss": 1.6678,
+      "step": 137
+    },
+    {
+      "epoch": 0.276,
+      "grad_norm": 15.234736442565918,
+      "learning_rate": 9.200000000000002e-06,
+      "loss": 1.4748,
+      "step": 138
+    },
+    {
+      "epoch": 0.278,
+      "grad_norm": 12.608988761901855,
+      "learning_rate": 9.266666666666667e-06,
+      "loss": 1.6285,
+      "step": 139
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 20.843536376953125,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 1.699,
+      "step": 140
+    },
+    {
+      "epoch": 0.282,
+      "grad_norm": 18.524633407592773,
+      "learning_rate": 9.4e-06,
+      "loss": 1.6608,
+      "step": 141
+    },
+    {
+      "epoch": 0.284,
+      "grad_norm": 13.68975830078125,
+      "learning_rate": 9.466666666666667e-06,
+      "loss": 1.6153,
+      "step": 142
+    },
+    {
+      "epoch": 0.286,
+      "grad_norm": 14.197174072265625,
+      "learning_rate": 9.533333333333334e-06,
+      "loss": 1.7042,
+      "step": 143
+    },
+    {
+      "epoch": 0.288,
+      "grad_norm": 18.283809661865234,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 1.7929,
+      "step": 144
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 14.084982872009277,
+      "learning_rate": 9.666666666666667e-06,
+      "loss": 1.7448,
+      "step": 145
+    },
+    {
+      "epoch": 0.292,
+      "grad_norm": 15.236905097961426,
+      "learning_rate": 9.733333333333334e-06,
+      "loss": 1.7063,
+      "step": 146
+    },
+    {
+      "epoch": 0.294,
+      "grad_norm": 14.99785327911377,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 1.6991,
+      "step": 147
+    },
+    {
+      "epoch": 0.296,
+      "grad_norm": 18.394018173217773,
+      "learning_rate": 9.866666666666668e-06,
+      "loss": 1.5488,
+      "step": 148
+    },
+    {
+      "epoch": 0.298,
+      "grad_norm": 11.244776725769043,
+      "learning_rate": 9.933333333333334e-06,
+      "loss": 1.6293,
+      "step": 149
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 20.61504364013672,
+      "learning_rate": 1e-05,
+      "loss": 1.703,
+      "step": 150
+    },
+    {
+      "epoch": 0.302,
+      "grad_norm": 20.268144607543945,
+      "learning_rate": 9.99998646145412e-06,
+      "loss": 1.8211,
+      "step": 151
+    },
+    {
+      "epoch": 0.304,
+      "grad_norm": 17.08184242248535,
+      "learning_rate": 9.999945845889795e-06,
+      "loss": 1.5821,
+      "step": 152
+    },
+    {
+      "epoch": 0.306,
+      "grad_norm": 18.60841178894043,
+      "learning_rate": 9.999878153526974e-06,
+      "loss": 1.7683,
+      "step": 153
+    },
+    {
+      "epoch": 0.308,
+      "grad_norm": 11.097747802734375,
+      "learning_rate": 9.999783384732242e-06,
+      "loss": 1.6726,
+      "step": 154
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 24.11589241027832,
+      "learning_rate": 9.999661540018812e-06,
+      "loss": 1.7708,
+      "step": 155
+    },
+    {
+      "epoch": 0.312,
+      "grad_norm": 14.22172737121582,
+      "learning_rate": 9.999512620046523e-06,
+      "loss": 1.6218,
+      "step": 156
+    },
+    {
+      "epoch": 0.314,
+      "grad_norm": 15.114810943603516,
+      "learning_rate": 9.999336625621836e-06,
+      "loss": 1.7024,
+      "step": 157
+    },
+    {
+      "epoch": 0.316,
+      "grad_norm": 11.347399711608887,
+      "learning_rate": 9.99913355769784e-06,
+      "loss": 1.7575,
+      "step": 158
+    },
+    {
+      "epoch": 0.318,
+      "grad_norm": 13.826147079467773,
+      "learning_rate": 9.998903417374228e-06,
+      "loss": 1.727,
+      "step": 159
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 14.352263450622559,
+      "learning_rate": 9.99864620589731e-06,
+      "loss": 1.5935,
+      "step": 160
+    },
+    {
+      "epoch": 0.322,
+      "grad_norm": 12.488923072814941,
+      "learning_rate": 9.998361924659989e-06,
+      "loss": 1.5656,
+      "step": 161
+    },
+    {
+      "epoch": 0.324,
+      "grad_norm": 18.001474380493164,
+      "learning_rate": 9.998050575201772e-06,
+      "loss": 1.6436,
+      "step": 162
+    },
+    {
+      "epoch": 0.326,
+      "grad_norm": 16.588930130004883,
+      "learning_rate": 9.997712159208745e-06,
+      "loss": 1.6952,
+      "step": 163
+    },
+    {
+      "epoch": 0.328,
+      "grad_norm": 15.841776847839355,
+      "learning_rate": 9.99734667851357e-06,
+      "loss": 1.6058,
+      "step": 164
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 17.352310180664062,
+      "learning_rate": 9.99695413509548e-06,
+      "loss": 1.6962,
+      "step": 165
+    },
+    {
+      "epoch": 0.332,
+      "grad_norm": 12.201016426086426,
+      "learning_rate": 9.99653453108026e-06,
+      "loss": 1.5996,
+      "step": 166
+    },
+    {
+      "epoch": 0.334,
+      "grad_norm": 20.2280216217041,
+      "learning_rate": 9.996087868740244e-06,
+      "loss": 1.7222,
+      "step": 167
+    },
+    {
+      "epoch": 0.336,
+      "grad_norm": 15.186982154846191,
+      "learning_rate": 9.995614150494293e-06,
+      "loss": 1.7376,
+      "step": 168
+    },
+    {
+      "epoch": 0.338,
+      "grad_norm": 17.792768478393555,
+      "learning_rate": 9.995113378907791e-06,
+      "loss": 1.6865,
+      "step": 169
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.01579761505127,
+      "learning_rate": 9.994585556692624e-06,
+      "loss": 1.6595,
+      "step": 170
+    },
+    {
+      "epoch": 0.342,
+      "grad_norm": 16.191465377807617,
+      "learning_rate": 9.994030686707171e-06,
+      "loss": 1.5984,
+      "step": 171
+    },
+    {
+      "epoch": 0.344,
+      "grad_norm": 15.65661334991455,
+      "learning_rate": 9.993448771956285e-06,
+      "loss": 1.6659,
+      "step": 172
+    },
+    {
+      "epoch": 0.346,
+      "grad_norm": 673.6796264648438,
+      "learning_rate": 9.99283981559128e-06,
+      "loss": 2.1504,
+      "step": 173
+    },
+    {
+      "epoch": 0.348,
+      "grad_norm": 11.247122764587402,
+      "learning_rate": 9.992203820909906e-06,
+      "loss": 1.7253,
+      "step": 174
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 17.28847312927246,
+      "learning_rate": 9.991540791356342e-06,
+      "loss": 1.7098,
+      "step": 175
+    },
+    {
+      "epoch": 0.352,
+      "grad_norm": 13.881318092346191,
+      "learning_rate": 9.99085073052117e-06,
+      "loss": 1.5808,
+      "step": 176
+    },
+    {
+      "epoch": 0.354,
+      "grad_norm": 14.093522071838379,
+      "learning_rate": 9.990133642141359e-06,
+      "loss": 1.613,
+      "step": 177
+    },
+    {
+      "epoch": 0.356,
+      "grad_norm": 10.097461700439453,
+      "learning_rate": 9.989389530100242e-06,
+      "loss": 1.7024,
+      "step": 178
+    },
+    {
+      "epoch": 0.358,
+      "grad_norm": 12.82798957824707,
+      "learning_rate": 9.988618398427495e-06,
+      "loss": 1.6558,
+      "step": 179
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.32880687713623,
+      "learning_rate": 9.987820251299121e-06,
+      "loss": 1.4778,
+      "step": 180
+    },
+    {
+      "epoch": 0.362,
+      "grad_norm": 8.794655799865723,
+      "learning_rate": 9.986995093037422e-06,
+      "loss": 1.6181,
+      "step": 181
+    },
+    {
+      "epoch": 0.364,
+      "grad_norm": 11.556446075439453,
+      "learning_rate": 9.986142928110972e-06,
+      "loss": 1.7423,
+      "step": 182
+    },
+    {
+      "epoch": 0.366,
+      "grad_norm": 23.587114334106445,
+      "learning_rate": 9.985263761134602e-06,
+      "loss": 1.7441,
+      "step": 183
+    },
+    {
+      "epoch": 0.368,
+      "grad_norm": 14.703253746032715,
+      "learning_rate": 9.984357596869369e-06,
+      "loss": 1.6989,
+      "step": 184
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.022418975830078,
+      "learning_rate": 9.98342444022253e-06,
+      "loss": 1.6443,
+      "step": 185
+    },
+    {
+      "epoch": 0.372,
+      "grad_norm": 16.842967987060547,
+      "learning_rate": 9.982464296247523e-06,
+      "loss": 1.7194,
+      "step": 186
+    },
+    {
+      "epoch": 0.374,
+      "grad_norm": 14.422867774963379,
+      "learning_rate": 9.981477170143924e-06,
+      "loss": 1.6715,
+      "step": 187
+    },
+    {
+      "epoch": 0.376,
+      "grad_norm": 15.538463592529297,
+      "learning_rate": 9.980463067257437e-06,
+      "loss": 1.6339,
+      "step": 188
+    },
+    {
+      "epoch": 0.378,
+      "grad_norm": 10.891705513000488,
+      "learning_rate": 9.979421993079853e-06,
+      "loss": 1.6721,
+      "step": 189
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 12.947959899902344,
+      "learning_rate": 9.978353953249023e-06,
+      "loss": 1.7966,
+      "step": 190
+    },
+    {
+      "epoch": 0.382,
+      "grad_norm": 9.348876953125,
+      "learning_rate": 9.977258953548831e-06,
+      "loss": 1.6517,
+      "step": 191
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 8.993280410766602,
+      "learning_rate": 9.976136999909156e-06,
+      "loss": 1.6427,
+      "step": 192
+    },
+    {
+      "epoch": 0.386,
+      "grad_norm": 8.9512300491333,
+      "learning_rate": 9.97498809840585e-06,
+      "loss": 1.6192,
+      "step": 193
+    },
+    {
+      "epoch": 0.388,
+      "grad_norm": 9.874775886535645,
+      "learning_rate": 9.973812255260692e-06,
+      "loss": 1.6766,
+      "step": 194
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.946427345275879,
+      "learning_rate": 9.972609476841368e-06,
+      "loss": 1.6853,
+      "step": 195
+    },
+    {
+      "epoch": 0.392,
+      "grad_norm": 11.508259773254395,
+      "learning_rate": 9.971379769661422e-06,
+      "loss": 1.6911,
+      "step": 196
+    },
+    {
+      "epoch": 0.394,
+      "grad_norm": 10.715835571289062,
+      "learning_rate": 9.970123140380237e-06,
+      "loss": 1.6015,
+      "step": 197
+    },
+    {
+      "epoch": 0.396,
+      "grad_norm": 11.727666854858398,
+      "learning_rate": 9.968839595802982e-06,
+      "loss": 1.6955,
+      "step": 198
+    },
+    {
+      "epoch": 0.398,
+      "grad_norm": 10.18913745880127,
+      "learning_rate": 9.967529142880592e-06,
+      "loss": 1.5734,
+      "step": 199
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.491742134094238,
+      "learning_rate": 9.966191788709716e-06,
+      "loss": 1.659,
+      "step": 200
+    },
+    {
+      "epoch": 0.402,
+      "grad_norm": 8.228020668029785,
+      "learning_rate": 9.964827540532685e-06,
+      "loss": 1.5536,
+      "step": 201
+    },
+    {
+      "epoch": 0.404,
+      "grad_norm": 15.777739524841309,
+      "learning_rate": 9.963436405737476e-06,
+      "loss": 1.536,
+      "step": 202
+    },
+    {
+      "epoch": 0.406,
+      "grad_norm": 11.057550430297852,
+      "learning_rate": 9.962018391857665e-06,
+      "loss": 1.583,
+      "step": 203
+    },
+    {
+      "epoch": 0.408,
+      "grad_norm": 11.679120063781738,
+      "learning_rate": 9.960573506572391e-06,
+      "loss": 1.5931,
+      "step": 204
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 18.90032386779785,
+      "learning_rate": 9.959101757706308e-06,
+      "loss": 1.7411,
+      "step": 205
+    },
+    {
+      "epoch": 0.412,
+      "grad_norm": 18.84972381591797,
+      "learning_rate": 9.957603153229559e-06,
+      "loss": 1.6154,
+      "step": 206
+    },
+    {
+      "epoch": 0.414,
+      "grad_norm": 13.331605911254883,
+      "learning_rate": 9.95607770125771e-06,
+      "loss": 1.677,
+      "step": 207
+    },
+    {
+      "epoch": 0.416,
+      "grad_norm": 13.559438705444336,
+      "learning_rate": 9.95452541005172e-06,
+      "loss": 1.6305,
+      "step": 208
+    },
+    {
+      "epoch": 0.418,
+      "grad_norm": 14.361797332763672,
+      "learning_rate": 9.952946288017899e-06,
+      "loss": 1.6408,
+      "step": 209
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 16.02930450439453,
+      "learning_rate": 9.951340343707852e-06,
+      "loss": 1.7221,
+      "step": 210
+    },
+    {
+      "epoch": 0.422,
+      "grad_norm": 15.836488723754883,
+      "learning_rate": 9.94970758581844e-06,
+      "loss": 1.5559,
+      "step": 211
+    },
+    {
+      "epoch": 0.424,
+      "grad_norm": 13.252622604370117,
+      "learning_rate": 9.948048023191728e-06,
+      "loss": 1.6659,
+      "step": 212
+    },
+    {
+      "epoch": 0.426,
+      "grad_norm": 10.738286972045898,
+      "learning_rate": 9.946361664814942e-06,
+      "loss": 1.5846,
+      "step": 213
+    },
+    {
+      "epoch": 0.428,
+      "grad_norm": 12.482677459716797,
+      "learning_rate": 9.94464851982042e-06,
+      "loss": 1.7871,
+      "step": 214
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.500411987304688,
+      "learning_rate": 9.942908597485558e-06,
+      "loss": 1.7273,
+      "step": 215
+    },
+    {
+      "epoch": 0.432,
+      "grad_norm": 10.43774700164795,
+      "learning_rate": 9.941141907232766e-06,
+      "loss": 1.7749,
+      "step": 216
+    },
+    {
+      "epoch": 0.434,
+      "grad_norm": 10.52614688873291,
+      "learning_rate": 9.939348458629406e-06,
+      "loss": 1.6396,
+      "step": 217
+    },
+    {
+      "epoch": 0.436,
+      "grad_norm": 9.133844375610352,
+      "learning_rate": 9.937528261387753e-06,
+      "loss": 1.6775,
+      "step": 218
+    },
+    {
+      "epoch": 0.438,
+      "grad_norm": 10.179513931274414,
+      "learning_rate": 9.93568132536494e-06,
+      "loss": 1.6362,
+      "step": 219
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.565604209899902,
+      "learning_rate": 9.933807660562898e-06,
+      "loss": 1.6125,
+      "step": 220
+    },
+    {
+      "epoch": 0.442,
+      "grad_norm": 13.952414512634277,
+      "learning_rate": 9.9319072771283e-06,
+      "loss": 1.6762,
+      "step": 221
+    },
+    {
+      "epoch": 0.444,
+      "grad_norm": 13.159649848937988,
+      "learning_rate": 9.929980185352525e-06,
+      "loss": 1.7111,
+      "step": 222
+    },
+    {
+      "epoch": 0.446,
+      "grad_norm": 15.363605499267578,
+      "learning_rate": 9.928026395671577e-06,
+      "loss": 1.687,
+      "step": 223
+    },
+    {
+      "epoch": 0.448,
+      "grad_norm": 12.805254936218262,
+      "learning_rate": 9.926045918666045e-06,
+      "loss": 1.6426,
+      "step": 224
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.896926879882812,
+      "learning_rate": 9.924038765061042e-06,
+      "loss": 1.657,
+      "step": 225
+    },
+    {
+      "epoch": 0.452,
+      "grad_norm": 8.098036766052246,
+      "learning_rate": 9.92200494572614e-06,
+      "loss": 1.5689,
+      "step": 226
+    },
+    {
+      "epoch": 0.454,
+      "grad_norm": 9.297266960144043,
+      "learning_rate": 9.919944471675328e-06,
+      "loss": 1.5641,
+      "step": 227
+    },
+    {
+      "epoch": 0.456,
+      "grad_norm": 11.86732292175293,
+      "learning_rate": 9.91785735406693e-06,
+      "loss": 1.5646,
+      "step": 228
+    },
+    {
+      "epoch": 0.458,
+      "grad_norm": 15.542577743530273,
+      "learning_rate": 9.915743604203563e-06,
+      "loss": 1.7098,
+      "step": 229
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 15.652758598327637,
+      "learning_rate": 9.913603233532067e-06,
+      "loss": 1.6008,
+      "step": 230
+    },
+    {
+      "epoch": 0.462,
+      "grad_norm": 14.435830116271973,
+      "learning_rate": 9.911436253643445e-06,
+      "loss": 1.6086,
+      "step": 231
+    },
+    {
+      "epoch": 0.464,
+      "grad_norm": 12.681479454040527,
+      "learning_rate": 9.909242676272797e-06,
+      "loss": 1.7047,
+      "step": 232
+    },
+    {
+      "epoch": 0.466,
+      "grad_norm": 13.065205574035645,
+      "learning_rate": 9.907022513299264e-06,
+      "loss": 1.6552,
+      "step": 233
+    },
+    {
+      "epoch": 0.468,
+      "grad_norm": 10.368819236755371,
+      "learning_rate": 9.904775776745959e-06,
+      "loss": 1.6101,
+      "step": 234
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 17.001413345336914,
+      "learning_rate": 9.902502478779897e-06,
+      "loss": 1.7077,
+      "step": 235
+    },
+    {
+      "epoch": 0.472,
+      "grad_norm": 9.429941177368164,
+      "learning_rate": 9.90020263171194e-06,
+      "loss": 1.5966,
+      "step": 236
+    },
+    {
+      "epoch": 0.474,
+      "grad_norm": 9.692448616027832,
+      "learning_rate": 9.89787624799672e-06,
+      "loss": 1.6065,
+      "step": 237
+    },
+    {
+      "epoch": 0.476,
+      "grad_norm": 12.412395477294922,
+      "learning_rate": 9.89552334023258e-06,
+      "loss": 1.8045,
+      "step": 238
+    },
+    {
+      "epoch": 0.478,
+      "grad_norm": 10.290336608886719,
+      "learning_rate": 9.893143921161501e-06,
+      "loss": 1.5883,
+      "step": 239
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 16.823965072631836,
+      "learning_rate": 9.890738003669029e-06,
+      "loss": 1.6098,
+      "step": 240
+    },
+    {
+      "epoch": 0.482,
+      "grad_norm": 10.46012020111084,
+      "learning_rate": 9.888305600784217e-06,
+      "loss": 1.6483,
+      "step": 241
+    },
+    {
+      "epoch": 0.484,
+      "grad_norm": 9.59673023223877,
+      "learning_rate": 9.88584672567954e-06,
+      "loss": 1.5618,
+      "step": 242
+    },
+    {
+      "epoch": 0.486,
+      "grad_norm": 11.56871223449707,
+      "learning_rate": 9.883361391670841e-06,
+      "loss": 1.698,
+      "step": 243
+    },
+    {
+      "epoch": 0.488,
+      "grad_norm": 6.657703399658203,
+      "learning_rate": 9.880849612217238e-06,
+      "loss": 1.689,
+      "step": 244
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 11.670310020446777,
+      "learning_rate": 9.878311400921072e-06,
+      "loss": 1.694,
+      "step": 245
+    },
+    {
+      "epoch": 0.492,
+      "grad_norm": 9.138250350952148,
+      "learning_rate": 9.875746771527817e-06,
+      "loss": 1.6979,
+      "step": 246
+    },
+    {
+      "epoch": 0.494,
+      "grad_norm": 10.32927131652832,
+      "learning_rate": 9.873155737926014e-06,
+      "loss": 1.6198,
+      "step": 247
+    },
+    {
+      "epoch": 0.496,
+      "grad_norm": 8.178282737731934,
+      "learning_rate": 9.870538314147194e-06,
+      "loss": 1.6388,
+      "step": 248
+    },
+    {
+      "epoch": 0.498,
+      "grad_norm": 10.270970344543457,
+      "learning_rate": 9.867894514365802e-06,
+      "loss": 1.5972,
+      "step": 249
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.778549194335938,
+      "learning_rate": 9.86522435289912e-06,
+      "loss": 1.7337,
+      "step": 250
+    },
+    {
+      "epoch": 0.502,
+      "grad_norm": 7.87909460067749,
+      "learning_rate": 9.862527844207189e-06,
+      "loss": 1.6102,
+      "step": 251
+    },
+    {
+      "epoch": 0.504,
+      "grad_norm": 9.182847023010254,
+      "learning_rate": 9.859805002892733e-06,
+      "loss": 1.7548,
+      "step": 252
+    },
+    {
+      "epoch": 0.506,
+      "grad_norm": 7.970770835876465,
+      "learning_rate": 9.857055843701073e-06,
+      "loss": 1.6482,
+      "step": 253
+    },
+    {
+      "epoch": 0.508,
+      "grad_norm": 13.17197322845459,
+      "learning_rate": 9.85428038152006e-06,
+      "loss": 1.663,
+      "step": 254
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.727519035339355,
+      "learning_rate": 9.851478631379982e-06,
+      "loss": 1.613,
+      "step": 255
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 11.023371696472168,
+      "learning_rate": 9.84865060845349e-06,
+      "loss": 1.6383,
+      "step": 256
+    },
+    {
+      "epoch": 0.514,
+      "grad_norm": 11.075227737426758,
+      "learning_rate": 9.845796328055505e-06,
+      "loss": 1.6506,
+      "step": 257
+    },
+    {
+      "epoch": 0.516,
+      "grad_norm": 10.624147415161133,
+      "learning_rate": 9.842915805643156e-06,
+      "loss": 1.6063,
+      "step": 258
+    },
+    {
+      "epoch": 0.518,
+      "grad_norm": 8.243653297424316,
+      "learning_rate": 9.840009056815674e-06,
+      "loss": 1.6393,
+      "step": 259
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.168231964111328,
+      "learning_rate": 9.83707609731432e-06,
+      "loss": 1.5481,
+      "step": 260
+    },
+    {
+      "epoch": 0.522,
+      "grad_norm": 10.238672256469727,
+      "learning_rate": 9.834116943022299e-06,
+      "loss": 1.703,
+      "step": 261
+    },
+    {
+      "epoch": 0.524,
+      "grad_norm": 6.614677906036377,
+      "learning_rate": 9.831131609964664e-06,
+      "loss": 1.5414,
+      "step": 262
+    },
+    {
+      "epoch": 0.526,
+      "grad_norm": 7.963540554046631,
+      "learning_rate": 9.828120114308248e-06,
+      "loss": 1.603,
+      "step": 263
+    },
+    {
+      "epoch": 0.528,
+      "grad_norm": 11.000412940979004,
+      "learning_rate": 9.825082472361558e-06,
+      "loss": 1.7098,
+      "step": 264
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.786978721618652,
+      "learning_rate": 9.822018700574696e-06,
+      "loss": 1.6766,
+      "step": 265
+    },
+    {
+      "epoch": 0.532,
+      "grad_norm": 9.823018074035645,
+      "learning_rate": 9.818928815539266e-06,
+      "loss": 1.6362,
+      "step": 266
+    },
+    {
+      "epoch": 0.534,
+      "grad_norm": 10.396644592285156,
+      "learning_rate": 9.815812833988292e-06,
+      "loss": 1.8225,
+      "step": 267
+    },
+    {
+      "epoch": 0.536,
+      "grad_norm": 11.278525352478027,
+      "learning_rate": 9.812670772796113e-06,
+      "loss": 1.7868,
+      "step": 268
+    },
+    {
+      "epoch": 0.538,
+      "grad_norm": 8.690960884094238,
+      "learning_rate": 9.809502648978311e-06,
+      "loss": 1.567,
+      "step": 269
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.479389190673828,
+      "learning_rate": 9.806308479691595e-06,
+      "loss": 1.6175,
+      "step": 270
+    },
+    {
+      "epoch": 0.542,
+      "grad_norm": 8.428929328918457,
+      "learning_rate": 9.803088282233733e-06,
+      "loss": 1.6575,
+      "step": 271
+    },
+    {
+      "epoch": 0.544,
+      "grad_norm": 8.638998985290527,
+      "learning_rate": 9.799842074043438e-06,
+      "loss": 1.6002,
+      "step": 272
+    },
+    {
+      "epoch": 0.546,
+      "grad_norm": 9.341121673583984,
+      "learning_rate": 9.796569872700287e-06,
+      "loss": 1.672,
+      "step": 273
+    },
+    {
+      "epoch": 0.548,
+      "grad_norm": 9.511893272399902,
+      "learning_rate": 9.793271695924621e-06,
+      "loss": 1.7315,
+      "step": 274
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 12.620855331420898,
+      "learning_rate": 9.789947561577445e-06,
+      "loss": 1.63,
+      "step": 275
+    },
+    {
+      "epoch": 0.552,
+      "grad_norm": 14.644619941711426,
+      "learning_rate": 9.786597487660336e-06,
+      "loss": 1.649,
+      "step": 276
+    },
+    {
+      "epoch": 0.554,
+      "grad_norm": 8.164085388183594,
+      "learning_rate": 9.78322149231535e-06,
+      "loss": 1.6249,
+      "step": 277
+    },
+    {
+      "epoch": 0.556,
+      "grad_norm": 12.925362586975098,
+      "learning_rate": 9.779819593824909e-06,
+      "loss": 1.572,
+      "step": 278
+    },
+    {
+      "epoch": 0.558,
+      "grad_norm": 10.382940292358398,
+      "learning_rate": 9.776391810611719e-06,
+      "loss": 1.6225,
+      "step": 279
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 14.114727973937988,
+      "learning_rate": 9.77293816123866e-06,
+      "loss": 1.4772,
+      "step": 280
+    },
+    {
+      "epoch": 0.562,
+      "grad_norm": 9.955523490905762,
+      "learning_rate": 9.769458664408689e-06,
+      "loss": 1.6424,
+      "step": 281
+    },
+    {
+      "epoch": 0.564,
+      "grad_norm": 10.805880546569824,
+      "learning_rate": 9.765953338964736e-06,
+      "loss": 1.5621,
+      "step": 282
+    },
+    {
+      "epoch": 0.566,
+      "grad_norm": 9.07789134979248,
+      "learning_rate": 9.762422203889604e-06,
+      "loss": 1.6892,
+      "step": 283
+    },
+    {
+      "epoch": 0.568,
+      "grad_norm": 13.484134674072266,
+      "learning_rate": 9.75886527830587e-06,
+      "loss": 1.7468,
+      "step": 284
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.758011817932129,
+      "learning_rate": 9.755282581475769e-06,
+      "loss": 1.699,
+      "step": 285
+    },
+    {
+      "epoch": 0.572,
+      "grad_norm": 8.382722854614258,
+      "learning_rate": 9.751674132801106e-06,
+      "loss": 1.5586,
+      "step": 286
+    },
+    {
+      "epoch": 0.574,
+      "grad_norm": 7.959819316864014,
+      "learning_rate": 9.748039951823141e-06,
+      "loss": 1.5625,
+      "step": 287
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 8.60888671875,
+      "learning_rate": 9.744380058222483e-06,
+      "loss": 1.5166,
+      "step": 288
+    },
+    {
+      "epoch": 0.578,
+      "grad_norm": 10.173426628112793,
+      "learning_rate": 9.740694471818988e-06,
+      "loss": 1.7414,
+      "step": 289
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.694028854370117,
+      "learning_rate": 9.736983212571646e-06,
+      "loss": 1.5719,
+      "step": 290
+    },
+    {
+      "epoch": 0.582,
+      "grad_norm": 12.50915241241455,
+      "learning_rate": 9.733246300578482e-06,
+      "loss": 1.5379,
+      "step": 291
+    },
+    {
+      "epoch": 0.584,
+      "grad_norm": 7.476858615875244,
+      "learning_rate": 9.729483756076436e-06,
+      "loss": 1.5798,
+      "step": 292
+    },
+    {
+      "epoch": 0.586,
+      "grad_norm": 8.575016975402832,
+      "learning_rate": 9.72569559944126e-06,
+      "loss": 1.6348,
+      "step": 293
+    },
+    {
+      "epoch": 0.588,
+      "grad_norm": 10.078662872314453,
+      "learning_rate": 9.721881851187406e-06,
+      "loss": 1.6831,
+      "step": 294
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 12.232658386230469,
+      "learning_rate": 9.718042531967918e-06,
+      "loss": 1.687,
+      "step": 295
+    },
+    {
+      "epoch": 0.592,
+      "grad_norm": 10.713579177856445,
+      "learning_rate": 9.714177662574316e-06,
+      "loss": 1.6168,
+      "step": 296
+    },
+    {
+      "epoch": 0.594,
+      "grad_norm": 10.275739669799805,
+      "learning_rate": 9.710287263936485e-06,
+      "loss": 1.7538,
+      "step": 297
+    },
+    {
+      "epoch": 0.596,
+      "grad_norm": 7.381926536560059,
+      "learning_rate": 9.70637135712256e-06,
+      "loss": 1.5313,
+      "step": 298
+    },
+    {
+      "epoch": 0.598,
+      "grad_norm": 10.051726341247559,
+      "learning_rate": 9.702429963338812e-06,
+      "loss": 1.7786,
+      "step": 299
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 10.447134017944336,
+      "learning_rate": 9.698463103929542e-06,
+      "loss": 1.6418,
+      "step": 300
+    },
+    {
+      "epoch": 0.602,
+      "grad_norm": 13.121293067932129,
+      "learning_rate": 9.694470800376951e-06,
+      "loss": 1.7257,
+      "step": 301
+    },
+    {
+      "epoch": 0.604,
+      "grad_norm": 8.965837478637695,
+      "learning_rate": 9.690453074301035e-06,
+      "loss": 1.6921,
+      "step": 302
+    },
+    {
+      "epoch": 0.606,
+      "grad_norm": 13.510847091674805,
+      "learning_rate": 9.68640994745946e-06,
+      "loss": 1.7176,
+      "step": 303
+    },
+    {
+      "epoch": 0.608,
+      "grad_norm": 10.887725830078125,
+      "learning_rate": 9.682341441747446e-06,
+      "loss": 1.7061,
+      "step": 304
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.40693473815918,
+      "learning_rate": 9.678247579197658e-06,
+      "loss": 1.6644,
+      "step": 305
+    },
+    {
+      "epoch": 0.612,
+      "grad_norm": 11.772154808044434,
+      "learning_rate": 9.674128381980073e-06,
+      "loss": 1.7006,
+      "step": 306
+    },
+    {
+      "epoch": 0.614,
+      "grad_norm": 10.446290969848633,
+      "learning_rate": 9.669983872401868e-06,
+      "loss": 1.7206,
+      "step": 307
+    },
+    {
+      "epoch": 0.616,
+      "grad_norm": 10.989177703857422,
+      "learning_rate": 9.665814072907293e-06,
+      "loss": 1.6525,
+      "step": 308
+    },
+    {
+      "epoch": 0.618,
+      "grad_norm": 11.617728233337402,
+      "learning_rate": 9.661619006077562e-06,
+      "loss": 1.6366,
+      "step": 309
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 12.805769920349121,
+      "learning_rate": 9.657398694630713e-06,
+      "loss": 1.7126,
+      "step": 310
+    },
+    {
+      "epoch": 0.622,
+      "grad_norm": 10.547633171081543,
+      "learning_rate": 9.653153161421497e-06,
+      "loss": 1.6447,
+      "step": 311
+    },
+    {
+      "epoch": 0.624,
+      "grad_norm": 11.312098503112793,
+      "learning_rate": 9.648882429441258e-06,
+      "loss": 1.732,
+      "step": 312
+    },
+    {
+      "epoch": 0.626,
+      "grad_norm": 11.163432121276855,
+      "learning_rate": 9.644586521817792e-06,
+      "loss": 1.5886,
+      "step": 313
+    },
+    {
+      "epoch": 0.628,
+      "grad_norm": 14.074267387390137,
+      "learning_rate": 9.640265461815235e-06,
+      "loss": 1.6658,
+      "step": 314
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.643535614013672,
+      "learning_rate": 9.635919272833938e-06,
+      "loss": 1.5441,
+      "step": 315
+    },
+    {
+      "epoch": 0.632,
+      "grad_norm": 17.20092010498047,
+      "learning_rate": 9.63154797841033e-06,
+      "loss": 1.6208,
+      "step": 316
+    },
+    {
+      "epoch": 0.634,
+      "grad_norm": 9.415007591247559,
+      "learning_rate": 9.627151602216801e-06,
+      "loss": 1.5145,
+      "step": 317
+    },
+    {
+      "epoch": 0.636,
+      "grad_norm": 14.709664344787598,
+      "learning_rate": 9.622730168061568e-06,
+      "loss": 1.7034,
+      "step": 318
+    },
+    {
+      "epoch": 0.638,
+      "grad_norm": 13.328878402709961,
+      "learning_rate": 9.618283699888543e-06,
+      "loss": 1.7108,
+      "step": 319
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.980976104736328,
+      "learning_rate": 9.613812221777212e-06,
+      "loss": 1.6241,
+      "step": 320
+    },
+    {
+      "epoch": 0.642,
+      "grad_norm": 11.059935569763184,
+      "learning_rate": 9.609315757942504e-06,
+      "loss": 1.7884,
+      "step": 321
+    },
+    {
+      "epoch": 0.644,
+      "grad_norm": 8.424722671508789,
+      "learning_rate": 9.604794332734647e-06,
+      "loss": 1.5701,
+      "step": 322
+    },
+    {
+      "epoch": 0.646,
+      "grad_norm": 13.05691146850586,
+      "learning_rate": 9.600247970639053e-06,
+      "loss": 1.6983,
+      "step": 323
+    },
+    {
+      "epoch": 0.648,
+      "grad_norm": 10.937281608581543,
+      "learning_rate": 9.595676696276173e-06,
+      "loss": 1.7072,
+      "step": 324
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.963797569274902,
+      "learning_rate": 9.591080534401371e-06,
+      "loss": 1.7535,
+      "step": 325
+    },
+    {
+      "epoch": 0.652,
+      "grad_norm": 11.561660766601562,
+      "learning_rate": 9.586459509904786e-06,
+      "loss": 1.564,
+      "step": 326
+    },
+    {
+      "epoch": 0.654,
+      "grad_norm": 9.038650512695312,
+      "learning_rate": 9.581813647811199e-06,
+      "loss": 1.7599,
+      "step": 327
+    },
+    {
+      "epoch": 0.656,
+      "grad_norm": 9.181880950927734,
+      "learning_rate": 9.577142973279896e-06,
+      "loss": 1.5822,
+      "step": 328
+    },
+    {
+      "epoch": 0.658,
+      "grad_norm": 9.09460735321045,
+      "learning_rate": 9.572447511604536e-06,
+      "loss": 1.6547,
+      "step": 329
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.515859603881836,
+      "learning_rate": 9.567727288213005e-06,
+      "loss": 1.59,
+      "step": 330
+    },
+    {
+      "epoch": 0.662,
+      "grad_norm": 11.761797904968262,
+      "learning_rate": 9.56298232866729e-06,
+      "loss": 1.7411,
+      "step": 331
+    },
+    {
+      "epoch": 0.664,
+      "grad_norm": 10.45743179321289,
+      "learning_rate": 9.55821265866333e-06,
+      "loss": 1.6905,
+      "step": 332
+    },
+    {
+      "epoch": 0.666,
+      "grad_norm": 8.767599105834961,
+      "learning_rate": 9.553418304030886e-06,
+      "loss": 1.624,
+      "step": 333
+    },
+    {
+      "epoch": 0.668,
+      "grad_norm": 8.311175346374512,
+      "learning_rate": 9.548599290733393e-06,
+      "loss": 1.7281,
+      "step": 334
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.42666244506836,
+      "learning_rate": 9.543755644867823e-06,
+      "loss": 1.6878,
+      "step": 335
+    },
+    {
+      "epoch": 0.672,
+      "grad_norm": 7.798083782196045,
+      "learning_rate": 9.538887392664544e-06,
+      "loss": 1.623,
+      "step": 336
+    },
+    {
+      "epoch": 0.674,
+      "grad_norm": 10.574175834655762,
+      "learning_rate": 9.53399456048718e-06,
+      "loss": 1.6386,
+      "step": 337
+    },
+    {
+      "epoch": 0.676,
+      "grad_norm": 8.909551620483398,
+      "learning_rate": 9.529077174832466e-06,
+      "loss": 1.7324,
+      "step": 338
+    },
+    {
+      "epoch": 0.678,
+      "grad_norm": 7.710360527038574,
+      "learning_rate": 9.524135262330098e-06,
+      "loss": 1.6373,
+      "step": 339
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 13.496674537658691,
+      "learning_rate": 9.519168849742603e-06,
+      "loss": 1.6202,
+      "step": 340
+    },
+    {
+      "epoch": 0.682,
+      "grad_norm": 7.529291152954102,
+      "learning_rate": 9.514177963965181e-06,
+      "loss": 1.7156,
+      "step": 341
+    },
+    {
+      "epoch": 0.684,
+      "grad_norm": 7.229175567626953,
+      "learning_rate": 9.50916263202557e-06,
+      "loss": 1.663,
+      "step": 342
+    },
+    {
+      "epoch": 0.686,
+      "grad_norm": 12.523305892944336,
+      "learning_rate": 9.504122881083886e-06,
+      "loss": 1.6902,
+      "step": 343
+    },
+    {
+      "epoch": 0.688,
+      "grad_norm": 10.700648307800293,
+      "learning_rate": 9.499058738432492e-06,
+      "loss": 1.6298,
+      "step": 344
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.542520523071289,
+      "learning_rate": 9.493970231495836e-06,
+      "loss": 1.5829,
+      "step": 345
+    },
+    {
+      "epoch": 0.692,
+      "grad_norm": 9.304771423339844,
+      "learning_rate": 9.488857387830315e-06,
+      "loss": 1.6467,
+      "step": 346
+    },
+    {
+      "epoch": 0.694,
+      "grad_norm": 11.418816566467285,
+      "learning_rate": 9.483720235124113e-06,
+      "loss": 1.7727,
+      "step": 347
+    },
+    {
+      "epoch": 0.696,
+      "grad_norm": 7.398611068725586,
+      "learning_rate": 9.478558801197065e-06,
+      "loss": 1.5846,
+      "step": 348
+    },
+    {
+      "epoch": 0.698,
+      "grad_norm": 10.512649536132812,
+      "learning_rate": 9.473373114000493e-06,
+      "loss": 1.7182,
+      "step": 349
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.7409029006958,
+      "learning_rate": 9.468163201617063e-06,
+      "loss": 1.6869,
+      "step": 350
+    },
+    {
+      "epoch": 0.702,
+      "grad_norm": 11.419679641723633,
+      "learning_rate": 9.46292909226063e-06,
+      "loss": 1.6018,
+      "step": 351
+    },
+    {
+      "epoch": 0.704,
+      "grad_norm": 12.171987533569336,
+      "learning_rate": 9.457670814276083e-06,
+      "loss": 1.5906,
+      "step": 352
+    },
+    {
+      "epoch": 0.706,
+      "grad_norm": 10.171733856201172,
+      "learning_rate": 9.452388396139202e-06,
+      "loss": 1.6307,
+      "step": 353
+    },
+    {
+      "epoch": 0.708,
+      "grad_norm": 11.288997650146484,
+      "learning_rate": 9.44708186645649e-06,
+      "loss": 1.627,
+      "step": 354
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.578997611999512,
+      "learning_rate": 9.441751253965022e-06,
+      "loss": 1.7024,
+      "step": 355
+    },
+    {
+      "epoch": 0.712,
+      "grad_norm": 8.29975414276123,
+      "learning_rate": 9.436396587532297e-06,
+      "loss": 1.7009,
+      "step": 356
+    },
+    {
+      "epoch": 0.714,
+      "grad_norm": 9.676756858825684,
+      "learning_rate": 9.431017896156074e-06,
+      "loss": 1.6421,
+      "step": 357
+    },
+    {
+      "epoch": 0.716,
+      "grad_norm": 8.530579566955566,
+      "learning_rate": 9.425615208964217e-06,
+      "loss": 1.619,
+      "step": 358
+    },
+    {
+      "epoch": 0.718,
+      "grad_norm": 6.746725082397461,
+      "learning_rate": 9.420188555214537e-06,
+      "loss": 1.5287,
+      "step": 359
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.753213405609131,
+      "learning_rate": 9.414737964294636e-06,
+      "loss": 1.53,
+      "step": 360
+    },
+    {
+      "epoch": 0.722,
+      "grad_norm": 7.940869331359863,
+      "learning_rate": 9.40926346572174e-06,
+      "loss": 1.5859,
+      "step": 361
+    },
+    {
+      "epoch": 0.724,
+      "grad_norm": 8.257153511047363,
+      "learning_rate": 9.403765089142554e-06,
+      "loss": 1.559,
+      "step": 362
+    },
+    {
+      "epoch": 0.726,
+      "grad_norm": 8.980607032775879,
+      "learning_rate": 9.398242864333084e-06,
+      "loss": 1.6408,
+      "step": 363
+    },
+    {
+      "epoch": 0.728,
+      "grad_norm": 12.678767204284668,
+      "learning_rate": 9.392696821198488e-06,
+      "loss": 1.6613,
+      "step": 364
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.08625316619873,
+      "learning_rate": 9.38712698977291e-06,
+      "loss": 1.6557,
+      "step": 365
+    },
+    {
+      "epoch": 0.732,
+      "grad_norm": 9.90092658996582,
+      "learning_rate": 9.381533400219319e-06,
+      "loss": 1.7143,
+      "step": 366
+    },
+    {
+      "epoch": 0.734,
+      "grad_norm": 15.684215545654297,
+      "learning_rate": 9.375916082829341e-06,
+      "loss": 1.6149,
+      "step": 367
+    },
+    {
+      "epoch": 0.736,
+      "grad_norm": 8.9537935256958,
+      "learning_rate": 9.370275068023097e-06,
+      "loss": 1.5268,
+      "step": 368
+    },
+    {
+      "epoch": 0.738,
+      "grad_norm": 10.356677055358887,
+      "learning_rate": 9.364610386349048e-06,
+      "loss": 1.7228,
+      "step": 369
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.680842399597168,
+      "learning_rate": 9.358922068483813e-06,
+      "loss": 1.762,
+      "step": 370
+    },
+    {
+      "epoch": 0.742,
+      "grad_norm": 9.59675121307373,
+      "learning_rate": 9.35321014523201e-06,
+      "loss": 1.6749,
+      "step": 371
+    },
+    {
+      "epoch": 0.744,
+      "grad_norm": 11.908790588378906,
+      "learning_rate": 9.347474647526095e-06,
+      "loss": 1.67,
+      "step": 372
+    },
+    {
+      "epoch": 0.746,
+      "grad_norm": 10.074137687683105,
+      "learning_rate": 9.34171560642619e-06,
+      "loss": 1.6049,
+      "step": 373
+    },
+    {
+      "epoch": 0.748,
+      "grad_norm": 9.519111633300781,
+      "learning_rate": 9.335933053119906e-06,
+      "loss": 1.7383,
+      "step": 374
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.124157905578613,
+      "learning_rate": 9.330127018922195e-06,
+      "loss": 1.5772,
+      "step": 375
+    },
+    {
+      "epoch": 0.752,
+      "grad_norm": 9.715572357177734,
+      "learning_rate": 9.324297535275156e-06,
+      "loss": 1.6024,
+      "step": 376
+    },
+    {
+      "epoch": 0.754,
+      "grad_norm": 14.774503707885742,
+      "learning_rate": 9.318444633747884e-06,
+      "loss": 1.614,
+      "step": 377
+    },
+    {
+      "epoch": 0.756,
+      "grad_norm": 15.559951782226562,
+      "learning_rate": 9.312568346036288e-06,
+      "loss": 1.6546,
+      "step": 378
+    },
+    {
+      "epoch": 0.758,
+      "grad_norm": 9.326455116271973,
+      "learning_rate": 9.306668703962927e-06,
+      "loss": 1.6567,
+      "step": 379
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.113893508911133,
+      "learning_rate": 9.30074573947683e-06,
+      "loss": 1.6268,
+      "step": 380
+    },
+    {
+      "epoch": 0.762,
+      "grad_norm": 10.376815795898438,
+      "learning_rate": 9.294799484653323e-06,
+      "loss": 1.649,
+      "step": 381
+    },
+    {
+      "epoch": 0.764,
+      "grad_norm": 12.478555679321289,
+      "learning_rate": 9.288829971693869e-06,
+      "loss": 1.5325,
+      "step": 382
+    },
+    {
+      "epoch": 0.766,
+      "grad_norm": 8.862922668457031,
+      "learning_rate": 9.282837232925876e-06,
+      "loss": 1.5611,
+      "step": 383
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 9.65459156036377,
+      "learning_rate": 9.276821300802535e-06,
+      "loss": 1.6238,
+      "step": 384
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.751141548156738,
+      "learning_rate": 9.27078220790263e-06,
+      "loss": 1.629,
+      "step": 385
+    },
+    {
+      "epoch": 0.772,
+      "grad_norm": 11.15700912475586,
+      "learning_rate": 9.264719986930376e-06,
+      "loss": 1.757,
+      "step": 386
+    },
+    {
+      "epoch": 0.774,
+      "grad_norm": 11.11645221710205,
+      "learning_rate": 9.25863467071524e-06,
+      "loss": 1.7262,
+      "step": 387
+    },
+    {
+      "epoch": 0.776,
+      "grad_norm": 9.191835403442383,
+      "learning_rate": 9.25252629221175e-06,
+      "loss": 1.5161,
+      "step": 388
+    },
+    {
+      "epoch": 0.778,
+      "grad_norm": 8.911935806274414,
+      "learning_rate": 9.246394884499334e-06,
+      "loss": 1.6185,
+      "step": 389
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.155871391296387,
+      "learning_rate": 9.24024048078213e-06,
+      "loss": 1.6151,
+      "step": 390
+    },
+    {
+      "epoch": 0.782,
+      "grad_norm": 7.377295970916748,
+      "learning_rate": 9.234063114388809e-06,
+      "loss": 1.5232,
+      "step": 391
+    },
+    {
+      "epoch": 0.784,
+      "grad_norm": 10.363668441772461,
+      "learning_rate": 9.227862818772392e-06,
+      "loss": 1.6332,
+      "step": 392
+    },
+    {
+      "epoch": 0.786,
+      "grad_norm": 8.449118614196777,
+      "learning_rate": 9.221639627510076e-06,
+      "loss": 1.5769,
+      "step": 393
+    },
+    {
+      "epoch": 0.788,
+      "grad_norm": 8.908105850219727,
+      "learning_rate": 9.215393574303043e-06,
+      "loss": 1.7008,
+      "step": 394
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.45577335357666,
+      "learning_rate": 9.209124692976287e-06,
+      "loss": 1.6707,
+      "step": 395
+    },
+    {
+      "epoch": 0.792,
+      "grad_norm": 9.608206748962402,
+      "learning_rate": 9.202833017478421e-06,
+      "loss": 1.6652,
+      "step": 396
+    },
+    {
+      "epoch": 0.794,
+      "grad_norm": 9.177596092224121,
+      "learning_rate": 9.196518581881502e-06,
+      "loss": 1.7067,
+      "step": 397
+    },
+    {
+      "epoch": 0.796,
+      "grad_norm": 5.568012714385986,
+      "learning_rate": 9.190181420380838e-06,
+      "loss": 1.6233,
+      "step": 398
+    },
+    {
+      "epoch": 0.798,
+      "grad_norm": 10.35659122467041,
+      "learning_rate": 9.18382156729481e-06,
+      "loss": 1.6694,
+      "step": 399
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.523914813995361,
+      "learning_rate": 9.177439057064684e-06,
+      "loss": 1.6281,
+      "step": 400
+    },
+    {
+      "epoch": 0.802,
+      "grad_norm": 7.155726432800293,
+      "learning_rate": 9.171033924254421e-06,
+      "loss": 1.6411,
+      "step": 401
+    },
+    {
+      "epoch": 0.804,
+      "grad_norm": 11.25847339630127,
+      "learning_rate": 9.164606203550498e-06,
+      "loss": 1.5891,
+      "step": 402
+    },
+    {
+      "epoch": 0.806,
+      "grad_norm": 12.162601470947266,
+      "learning_rate": 9.15815592976171e-06,
+      "loss": 1.5304,
+      "step": 403
+    },
+    {
+      "epoch": 0.808,
+      "grad_norm": 9.000397682189941,
+      "learning_rate": 9.151683137818989e-06,
+      "loss": 1.6407,
+      "step": 404
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.760917663574219,
+      "learning_rate": 9.145187862775208e-06,
+      "loss": 1.6462,
+      "step": 405
+    },
+    {
+      "epoch": 0.812,
+      "grad_norm": 9.060429573059082,
+      "learning_rate": 9.138670139805004e-06,
+      "loss": 1.6845,
+      "step": 406
+    },
+    {
+      "epoch": 0.814,
+      "grad_norm": 8.54736614227295,
+      "learning_rate": 9.132130004204569e-06,
+      "loss": 1.7051,
+      "step": 407
+    },
+    {
+      "epoch": 0.816,
+      "grad_norm": 10.724129676818848,
+      "learning_rate": 9.125567491391476e-06,
+      "loss": 1.6058,
+      "step": 408
+    },
+    {
+      "epoch": 0.818,
+      "grad_norm": 9.786903381347656,
+      "learning_rate": 9.118982636904476e-06,
+      "loss": 1.6181,
+      "step": 409
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.084239959716797,
+      "learning_rate": 9.112375476403313e-06,
+      "loss": 1.7077,
+      "step": 410
+    },
+    {
+      "epoch": 0.822,
+      "grad_norm": 8.961618423461914,
+      "learning_rate": 9.10574604566852e-06,
+      "loss": 1.6431,
+      "step": 411
+    },
+    {
+      "epoch": 0.824,
+      "grad_norm": 8.100645065307617,
+      "learning_rate": 9.099094380601244e-06,
+      "loss": 1.6346,
+      "step": 412
+    },
+    {
+      "epoch": 0.826,
+      "grad_norm": 10.302638053894043,
+      "learning_rate": 9.09242051722303e-06,
+      "loss": 1.728,
+      "step": 413
+    },
+    {
+      "epoch": 0.828,
+      "grad_norm": 8.645998001098633,
+      "learning_rate": 9.085724491675642e-06,
+      "loss": 1.6298,
+      "step": 414
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.636702060699463,
+      "learning_rate": 9.079006340220862e-06,
+      "loss": 1.7458,
+      "step": 415
+    },
+    {
+      "epoch": 0.832,
+      "grad_norm": 9.560812950134277,
+      "learning_rate": 9.072266099240286e-06,
+      "loss": 1.6252,
+      "step": 416
+    },
+    {
+      "epoch": 0.834,
+      "grad_norm": 8.170851707458496,
+      "learning_rate": 9.065503805235139e-06,
+      "loss": 1.5759,
+      "step": 417
+    },
+    {
+      "epoch": 0.836,
+      "grad_norm": 10.425228118896484,
+      "learning_rate": 9.058719494826076e-06,
+      "loss": 1.7297,
+      "step": 418
+    },
+    {
+      "epoch": 0.838,
+      "grad_norm": 8.824040412902832,
+      "learning_rate": 9.051913204752972e-06,
+      "loss": 1.7309,
+      "step": 419
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.032853126525879,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 1.5986,
+      "step": 420
+    },
+    {
+      "epoch": 0.842,
+      "grad_norm": 8.494746208190918,
+      "learning_rate": 9.03823483316911e-06,
+      "loss": 1.6812,
+      "step": 421
+    },
+    {
+      "epoch": 0.844,
+      "grad_norm": 9.491890907287598,
+      "learning_rate": 9.031362825732456e-06,
+      "loss": 1.6809,
+      "step": 422
+    },
+    {
+      "epoch": 0.846,
+      "grad_norm": 10.895613670349121,
+      "learning_rate": 9.02446898677957e-06,
+      "loss": 1.6665,
+      "step": 423
+    },
+    {
+      "epoch": 0.848,
+      "grad_norm": 6.491031169891357,
+      "learning_rate": 9.017553353643479e-06,
+      "loss": 1.6723,
+      "step": 424
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.402596473693848,
+      "learning_rate": 9.01061596377522e-06,
+      "loss": 1.5644,
+      "step": 425
+    },
+    {
+      "epoch": 0.852,
+      "grad_norm": 8.312848091125488,
+      "learning_rate": 9.003656854743667e-06,
+      "loss": 1.5704,
+      "step": 426
+    },
+    {
+      "epoch": 0.854,
+      "grad_norm": 10.550932884216309,
+      "learning_rate": 8.996676064235308e-06,
+      "loss": 1.6347,
+      "step": 427
+    },
+    {
+      "epoch": 0.856,
+      "grad_norm": 7.069148540496826,
+      "learning_rate": 8.989673630054044e-06,
+      "loss": 1.5883,
+      "step": 428
+    },
+    {
+      "epoch": 0.858,
+      "grad_norm": 8.47697639465332,
+      "learning_rate": 8.982649590120982e-06,
+      "loss": 1.6156,
+      "step": 429
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.076146125793457,
+      "learning_rate": 8.97560398247424e-06,
+      "loss": 1.6816,
+      "step": 430
+    },
+    {
+      "epoch": 0.862,
+      "grad_norm": 9.539214134216309,
+      "learning_rate": 8.96853684526873e-06,
+      "loss": 1.6875,
+      "step": 431
+    },
+    {
+      "epoch": 0.864,
+      "grad_norm": 10.413458824157715,
+      "learning_rate": 8.961448216775955e-06,
+      "loss": 1.6266,
+      "step": 432
+    },
+    {
+      "epoch": 0.866,
+      "grad_norm": 9.216937065124512,
+      "learning_rate": 8.954338135383804e-06,
+      "loss": 1.6485,
+      "step": 433
+    },
+    {
+      "epoch": 0.868,
+      "grad_norm": 8.924684524536133,
+      "learning_rate": 8.947206639596346e-06,
+      "loss": 1.5875,
+      "step": 434
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.03341007232666,
+      "learning_rate": 8.94005376803361e-06,
+      "loss": 1.562,
+      "step": 435
+    },
+    {
+      "epoch": 0.872,
+      "grad_norm": 7.471339225769043,
+      "learning_rate": 8.932879559431392e-06,
+      "loss": 1.5646,
+      "step": 436
+    },
+    {
+      "epoch": 0.874,
+      "grad_norm": 12.005831718444824,
+      "learning_rate": 8.925684052641027e-06,
+      "loss": 1.7528,
+      "step": 437
+    },
+    {
+      "epoch": 0.876,
+      "grad_norm": 7.764755725860596,
+      "learning_rate": 8.9184672866292e-06,
+      "loss": 1.6857,
+      "step": 438
+    },
+    {
+      "epoch": 0.878,
+      "grad_norm": 7.975648403167725,
+      "learning_rate": 8.911229300477716e-06,
+      "loss": 1.7224,
+      "step": 439
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.063853740692139,
+      "learning_rate": 8.903970133383297e-06,
+      "loss": 1.67,
+      "step": 440
+    },
+    {
+      "epoch": 0.882,
+      "grad_norm": 9.282867431640625,
+      "learning_rate": 8.896689824657371e-06,
+      "loss": 1.5403,
+      "step": 441
+    },
+    {
+      "epoch": 0.884,
+      "grad_norm": 7.045912265777588,
+      "learning_rate": 8.889388413725857e-06,
+      "loss": 1.5697,
+      "step": 442
+    },
+    {
+      "epoch": 0.886,
+      "grad_norm": 7.684753894805908,
+      "learning_rate": 8.882065940128946e-06,
+      "loss": 1.6201,
+      "step": 443
+    },
+    {
+      "epoch": 0.888,
+      "grad_norm": 11.638023376464844,
+      "learning_rate": 8.874722443520898e-06,
+      "loss": 1.6565,
+      "step": 444
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.81474781036377,
+      "learning_rate": 8.867357963669821e-06,
+      "loss": 1.6399,
+      "step": 445
+    },
+    {
+      "epoch": 0.892,
+      "grad_norm": 9.439010620117188,
+      "learning_rate": 8.859972540457451e-06,
+      "loss": 1.7231,
+      "step": 446
+    },
+    {
+      "epoch": 0.894,
+      "grad_norm": 8.762112617492676,
+      "learning_rate": 8.852566213878947e-06,
+      "loss": 1.5841,
+      "step": 447
+    },
+    {
+      "epoch": 0.896,
+      "grad_norm": 8.807167053222656,
+      "learning_rate": 8.845139024042664e-06,
+      "loss": 1.6177,
+      "step": 448
+    },
+    {
+      "epoch": 0.898,
+      "grad_norm": 7.984807014465332,
+      "learning_rate": 8.837691011169944e-06,
+      "loss": 1.6911,
+      "step": 449
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.620468139648438,
+      "learning_rate": 8.83022221559489e-06,
+      "loss": 1.6385,
+      "step": 450
+    },
+    {
+      "epoch": 0.902,
+      "grad_norm": 8.629189491271973,
+      "learning_rate": 8.822732677764158e-06,
+      "loss": 1.6133,
+      "step": 451
+    },
+    {
+      "epoch": 0.904,
+      "grad_norm": 10.831542015075684,
+      "learning_rate": 8.815222438236726e-06,
+      "loss": 1.655,
+      "step": 452
+    },
+    {
+      "epoch": 0.906,
+      "grad_norm": 7.558828353881836,
+      "learning_rate": 8.807691537683685e-06,
+      "loss": 1.5384,
+      "step": 453
+    },
+    {
+      "epoch": 0.908,
+      "grad_norm": 10.0075044631958,
+      "learning_rate": 8.800140016888009e-06,
+      "loss": 1.6663,
+      "step": 454
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.153804779052734,
+      "learning_rate": 8.792567916744346e-06,
+      "loss": 1.73,
+      "step": 455
+    },
+    {
+      "epoch": 0.912,
+      "grad_norm": 7.489542007446289,
+      "learning_rate": 8.784975278258783e-06,
+      "loss": 1.6016,
+      "step": 456
+    },
+    {
+      "epoch": 0.914,
+      "grad_norm": 6.917328357696533,
+      "learning_rate": 8.777362142548636e-06,
+      "loss": 1.6948,
+      "step": 457
+    },
+    {
+      "epoch": 0.916,
+      "grad_norm": 6.830845832824707,
+      "learning_rate": 8.769728550842217e-06,
+      "loss": 1.6222,
+      "step": 458
+    },
+    {
+      "epoch": 0.918,
+      "grad_norm": 7.163386821746826,
+      "learning_rate": 8.762074544478622e-06,
+      "loss": 1.6015,
+      "step": 459
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.106656074523926,
+      "learning_rate": 8.754400164907496e-06,
+      "loss": 1.6755,
+      "step": 460
+    },
+    {
+      "epoch": 0.922,
+      "grad_norm": 8.358505249023438,
+      "learning_rate": 8.746705453688815e-06,
+      "loss": 1.5817,
+      "step": 461
+    },
+    {
+      "epoch": 0.924,
+      "grad_norm": 10.56636905670166,
+      "learning_rate": 8.73899045249266e-06,
+      "loss": 1.6605,
+      "step": 462
+    },
+    {
+      "epoch": 0.926,
+      "grad_norm": 8.793217658996582,
+      "learning_rate": 8.73125520309899e-06,
+      "loss": 1.6739,
+      "step": 463
+    },
+    {
+      "epoch": 0.928,
+      "grad_norm": 7.576605319976807,
+      "learning_rate": 8.723499747397415e-06,
+      "loss": 1.5703,
+      "step": 464
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.332236289978027,
+      "learning_rate": 8.715724127386971e-06,
+      "loss": 1.5906,
+      "step": 465
+    },
+    {
+      "epoch": 0.932,
+      "grad_norm": 8.388971328735352,
+      "learning_rate": 8.707928385175898e-06,
+      "loss": 1.6549,
+      "step": 466
+    },
+    {
+      "epoch": 0.934,
+      "grad_norm": 7.761672019958496,
+      "learning_rate": 8.700112562981398e-06,
+      "loss": 1.5616,
+      "step": 467
+    },
+    {
+      "epoch": 0.936,
+      "grad_norm": 9.178117752075195,
+      "learning_rate": 8.692276703129421e-06,
+      "loss": 1.7963,
+      "step": 468
+    },
+    {
+      "epoch": 0.938,
+      "grad_norm": 6.7894287109375,
+      "learning_rate": 8.68442084805442e-06,
+      "loss": 1.579,
+      "step": 469
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.14854907989502,
+      "learning_rate": 8.676545040299145e-06,
+      "loss": 1.7024,
+      "step": 470
+    },
+    {
+      "epoch": 0.942,
+      "grad_norm": 8.014143943786621,
+      "learning_rate": 8.668649322514382e-06,
+      "loss": 1.6969,
+      "step": 471
+    },
+    {
+      "epoch": 0.944,
+      "grad_norm": 7.022289752960205,
+      "learning_rate": 8.660733737458751e-06,
+      "loss": 1.5763,
+      "step": 472
+    },
+    {
+      "epoch": 0.946,
+      "grad_norm": 7.455463886260986,
+      "learning_rate": 8.652798327998458e-06,
+      "loss": 1.6751,
+      "step": 473
+    },
+    {
+      "epoch": 0.948,
+      "grad_norm": 10.453346252441406,
+      "learning_rate": 8.644843137107058e-06,
+      "loss": 1.6393,
+      "step": 474
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.011568069458008,
+      "learning_rate": 8.636868207865244e-06,
+      "loss": 1.6452,
+      "step": 475
+    },
+    {
+      "epoch": 0.952,
+      "grad_norm": 10.156243324279785,
+      "learning_rate": 8.628873583460593e-06,
+      "loss": 1.6624,
+      "step": 476
+    },
+    {
+      "epoch": 0.954,
+      "grad_norm": 8.829718589782715,
+      "learning_rate": 8.620859307187339e-06,
+      "loss": 1.6908,
+      "step": 477
+    },
+    {
+      "epoch": 0.956,
+      "grad_norm": 10.530731201171875,
+      "learning_rate": 8.61282542244614e-06,
+      "loss": 1.6371,
+      "step": 478
+    },
+    {
+      "epoch": 0.958,
+      "grad_norm": 9.92332935333252,
+      "learning_rate": 8.604771972743848e-06,
+      "loss": 1.6542,
+      "step": 479
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.439817905426025,
+      "learning_rate": 8.596699001693257e-06,
+      "loss": 1.7083,
+      "step": 480
+    },
+    {
+      "epoch": 0.962,
+      "grad_norm": 7.941191673278809,
+      "learning_rate": 8.588606553012884e-06,
+      "loss": 1.5819,
+      "step": 481
+    },
+    {
+      "epoch": 0.964,
+      "grad_norm": 7.217868328094482,
+      "learning_rate": 8.580494670526725e-06,
+      "loss": 1.7362,
+      "step": 482
+    },
+    {
+      "epoch": 0.966,
+      "grad_norm": 8.529379844665527,
+      "learning_rate": 8.572363398164017e-06,
+      "loss": 1.7196,
+      "step": 483
+    },
+    {
+      "epoch": 0.968,
+      "grad_norm": 8.774826049804688,
+      "learning_rate": 8.564212779959003e-06,
+      "loss": 1.6536,
+      "step": 484
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.855049133300781,
+      "learning_rate": 8.556042860050686e-06,
+      "loss": 1.6759,
+      "step": 485
+    },
+    {
+      "epoch": 0.972,
+      "grad_norm": 7.4477620124816895,
+      "learning_rate": 8.547853682682605e-06,
+      "loss": 1.6095,
+      "step": 486
+    },
+    {
+      "epoch": 0.974,
+      "grad_norm": 8.155046463012695,
+      "learning_rate": 8.539645292202579e-06,
+      "loss": 1.6397,
+      "step": 487
+    },
+    {
+      "epoch": 0.976,
+      "grad_norm": 9.956945419311523,
+      "learning_rate": 8.531417733062476e-06,
+      "loss": 1.5748,
+      "step": 488
+    },
+    {
+      "epoch": 0.978,
+      "grad_norm": 9.320968627929688,
+      "learning_rate": 8.523171049817974e-06,
+      "loss": 1.6588,
+      "step": 489
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.276108264923096,
+      "learning_rate": 8.51490528712831e-06,
+      "loss": 1.6456,
+      "step": 490
+    },
+    {
+      "epoch": 0.982,
+      "grad_norm": 11.987957954406738,
+      "learning_rate": 8.506620489756045e-06,
+      "loss": 1.6061,
+      "step": 491
+    },
+    {
+      "epoch": 0.984,
+      "grad_norm": 9.098200798034668,
+      "learning_rate": 8.498316702566828e-06,
+      "loss": 1.6903,
+      "step": 492
+    },
+    {
+      "epoch": 0.986,
+      "grad_norm": 7.814845561981201,
+      "learning_rate": 8.489993970529137e-06,
+      "loss": 1.4855,
+      "step": 493
+    },
+    {
+      "epoch": 0.988,
+      "grad_norm": 7.096703052520752,
+      "learning_rate": 8.481652338714048e-06,
+      "loss": 1.6221,
+      "step": 494
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.307745456695557,
+      "learning_rate": 8.473291852294986e-06,
+      "loss": 1.5903,
+      "step": 495
+    },
+    {
+      "epoch": 0.992,
+      "grad_norm": 7.393770694732666,
+      "learning_rate": 8.464912556547486e-06,
+      "loss": 1.5778,
+      "step": 496
+    },
+    {
+      "epoch": 0.994,
+      "grad_norm": 21.961185455322266,
+      "learning_rate": 8.456514496848938e-06,
+      "loss": 1.7559,
+      "step": 497
+    },
+    {
+      "epoch": 0.996,
+      "grad_norm": 12.971484184265137,
+      "learning_rate": 8.44809771867835e-06,
+      "loss": 1.6592,
+      "step": 498
+    },
+    {
+      "epoch": 0.998,
+      "grad_norm": 12.674059867858887,
+      "learning_rate": 8.439662267616093e-06,
+      "loss": 1.6422,
+      "step": 499
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.149175643920898,
+      "learning_rate": 8.43120818934367e-06,
+      "loss": 1.6264,
+      "step": 500
+    },
+    {
+      "epoch": 1.002,
+      "grad_norm": 14.11556625366211,
+      "learning_rate": 8.422735529643445e-06,
+      "loss": 1.5558,
+      "step": 501
+    },
+    {
+      "epoch": 1.004,
+      "grad_norm": 10.902586936950684,
+      "learning_rate": 8.414244334398418e-06,
+      "loss": 1.6304,
+      "step": 502
+    },
+    {
+      "epoch": 1.006,
+      "grad_norm": 9.201044082641602,
+      "learning_rate": 8.405734649591964e-06,
+      "loss": 1.7151,
+      "step": 503
+    },
+    {
+      "epoch": 1.008,
+      "grad_norm": 8.613140106201172,
+      "learning_rate": 8.397206521307584e-06,
+      "loss": 1.6446,
+      "step": 504
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 11.33934497833252,
+      "learning_rate": 8.388659995728662e-06,
+      "loss": 1.5727,
+      "step": 505
+    },
+    {
+      "epoch": 1.012,
+      "grad_norm": 9.204126358032227,
+      "learning_rate": 8.380095119138209e-06,
+      "loss": 1.6427,
+      "step": 506
+    },
+    {
+      "epoch": 1.014,
+      "grad_norm": 13.722872734069824,
+      "learning_rate": 8.371511937918616e-06,
+      "loss": 1.5922,
+      "step": 507
+    },
+    {
+      "epoch": 1.016,
+      "grad_norm": 9.983489036560059,
+      "learning_rate": 8.362910498551402e-06,
+      "loss": 1.5911,
+      "step": 508
+    },
+    {
+      "epoch": 1.018,
+      "grad_norm": 6.286109447479248,
+      "learning_rate": 8.354290847616954e-06,
+      "loss": 1.581,
+      "step": 509
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 8.981342315673828,
+      "learning_rate": 8.345653031794292e-06,
+      "loss": 1.6099,
+      "step": 510
+    },
+    {
+      "epoch": 1.022,
+      "grad_norm": 10.48516845703125,
+      "learning_rate": 8.3369970978608e-06,
+      "loss": 1.4918,
+      "step": 511
+    },
+    {
+      "epoch": 1.024,
+      "grad_norm": 10.683873176574707,
+      "learning_rate": 8.328323092691985e-06,
+      "loss": 1.7634,
+      "step": 512
+    },
+    {
+      "epoch": 1.026,
+      "grad_norm": 9.128839492797852,
+      "learning_rate": 8.319631063261209e-06,
+      "loss": 1.5248,
+      "step": 513
+    },
+    {
+      "epoch": 1.028,
+      "grad_norm": 7.288445949554443,
+      "learning_rate": 8.310921056639451e-06,
+      "loss": 1.597,
+      "step": 514
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 7.43460750579834,
+      "learning_rate": 8.302193119995038e-06,
+      "loss": 1.469,
+      "step": 515
+    },
+    {
+      "epoch": 1.032,
+      "grad_norm": 11.071965217590332,
+      "learning_rate": 8.293447300593402e-06,
+      "loss": 1.6112,
+      "step": 516
+    },
+    {
+      "epoch": 1.034,
+      "grad_norm": 9.424286842346191,
+      "learning_rate": 8.284683645796814e-06,
+      "loss": 1.6145,
+      "step": 517
+    },
+    {
+      "epoch": 1.036,
+      "grad_norm": 10.420470237731934,
+      "learning_rate": 8.275902203064125e-06,
+      "loss": 1.7171,
+      "step": 518
+    },
+    {
+      "epoch": 1.038,
+      "grad_norm": 9.002021789550781,
+      "learning_rate": 8.267103019950529e-06,
+      "loss": 1.603,
+      "step": 519
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 12.653961181640625,
+      "learning_rate": 8.258286144107277e-06,
+      "loss": 1.6203,
+      "step": 520
+    },
+    {
+      "epoch": 1.042,
+      "grad_norm": 7.975899696350098,
+      "learning_rate": 8.249451623281444e-06,
+      "loss": 1.5552,
+      "step": 521
+    },
+    {
+      "epoch": 1.044,
+      "grad_norm": 7.326491832733154,
+      "learning_rate": 8.240599505315656e-06,
+      "loss": 1.6063,
+      "step": 522
+    },
+    {
+      "epoch": 1.046,
+      "grad_norm": 7.846243381500244,
+      "learning_rate": 8.231729838147833e-06,
+      "loss": 1.5635,
+      "step": 523
+    },
+    {
+      "epoch": 1.048,
+      "grad_norm": 9.442638397216797,
+      "learning_rate": 8.222842669810936e-06,
+      "loss": 1.5853,
+      "step": 524
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 8.490560531616211,
+      "learning_rate": 8.213938048432697e-06,
+      "loss": 1.5719,
+      "step": 525
+    },
+    {
+      "epoch": 1.052,
+      "grad_norm": 9.718538284301758,
+      "learning_rate": 8.205016022235368e-06,
+      "loss": 1.6211,
+      "step": 526
+    },
+    {
+      "epoch": 1.054,
+      "grad_norm": 8.703619956970215,
+      "learning_rate": 8.196076639535453e-06,
+      "loss": 1.6081,
+      "step": 527
+    },
+    {
+      "epoch": 1.056,
+      "grad_norm": 11.14815902709961,
+      "learning_rate": 8.18711994874345e-06,
+      "loss": 1.6158,
+      "step": 528
+    },
+    {
+      "epoch": 1.058,
+      "grad_norm": 9.741596221923828,
+      "learning_rate": 8.178145998363585e-06,
+      "loss": 1.5314,
+      "step": 529
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 9.530781745910645,
+      "learning_rate": 8.16915483699355e-06,
+      "loss": 1.5866,
+      "step": 530
+    },
+    {
+      "epoch": 1.062,
+      "grad_norm": 8.731345176696777,
+      "learning_rate": 8.160146513324256e-06,
+      "loss": 1.5666,
+      "step": 531
+    },
+    {
+      "epoch": 1.064,
+      "grad_norm": 11.338554382324219,
+      "learning_rate": 8.151121076139534e-06,
+      "loss": 1.5834,
+      "step": 532
+    },
+    {
+      "epoch": 1.066,
+      "grad_norm": 20.54311752319336,
+      "learning_rate": 8.142078574315907e-06,
+      "loss": 1.5652,
+      "step": 533
+    },
+    {
+      "epoch": 1.068,
+      "grad_norm": 11.977243423461914,
+      "learning_rate": 8.133019056822303e-06,
+      "loss": 1.4832,
+      "step": 534
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 10.514939308166504,
+      "learning_rate": 8.123942572719801e-06,
+      "loss": 1.6011,
+      "step": 535
+    },
+    {
+      "epoch": 1.072,
+      "grad_norm": 9.276664733886719,
+      "learning_rate": 8.11484917116136e-06,
+      "loss": 1.5909,
+      "step": 536
+    },
+    {
+      "epoch": 1.074,
+      "grad_norm": 11.073919296264648,
+      "learning_rate": 8.105738901391553e-06,
+      "loss": 1.5826,
+      "step": 537
+    },
+    {
+      "epoch": 1.076,
+      "grad_norm": 9.078001022338867,
+      "learning_rate": 8.096611812746302e-06,
+      "loss": 1.641,
+      "step": 538
+    },
+    {
+      "epoch": 1.078,
+      "grad_norm": 10.105594635009766,
+      "learning_rate": 8.087467954652608e-06,
+      "loss": 1.6196,
+      "step": 539
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 13.397587776184082,
+      "learning_rate": 8.078307376628292e-06,
+      "loss": 1.562,
+      "step": 540
+    },
+    {
+      "epoch": 1.082,
+      "grad_norm": 13.389184951782227,
+      "learning_rate": 8.069130128281714e-06,
+      "loss": 1.5884,
+      "step": 541
+    },
+    {
+      "epoch": 1.084,
+      "grad_norm": 10.954330444335938,
+      "learning_rate": 8.059936259311514e-06,
+      "loss": 1.6104,
+      "step": 542
+    },
+    {
+      "epoch": 1.086,
+      "grad_norm": 8.864795684814453,
+      "learning_rate": 8.05072581950634e-06,
+      "loss": 1.6024,
+      "step": 543
+    },
+    {
+      "epoch": 1.088,
+      "grad_norm": 9.769396781921387,
+      "learning_rate": 8.041498858744572e-06,
+      "loss": 1.553,
+      "step": 544
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.69478988647461,
+      "learning_rate": 8.032255426994069e-06,
+      "loss": 1.6915,
+      "step": 545
+    },
+    {
+      "epoch": 1.092,
+      "grad_norm": 8.354500770568848,
+      "learning_rate": 8.022995574311876e-06,
+      "loss": 1.5773,
+      "step": 546
+    },
+    {
+      "epoch": 1.094,
+      "grad_norm": 9.368426322937012,
+      "learning_rate": 8.013719350843969e-06,
+      "loss": 1.6233,
+      "step": 547
+    },
+    {
+      "epoch": 1.096,
+      "grad_norm": 8.086129188537598,
+      "learning_rate": 8.004426806824985e-06,
+      "loss": 1.4905,
+      "step": 548
+    },
+    {
+      "epoch": 1.098,
+      "grad_norm": 8.991674423217773,
+      "learning_rate": 7.99511799257793e-06,
+      "loss": 1.4947,
+      "step": 549
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 9.440938949584961,
+      "learning_rate": 7.985792958513932e-06,
+      "loss": 1.6062,
+      "step": 550
+    },
+    {
+      "epoch": 1.102,
+      "grad_norm": 7.645297527313232,
+      "learning_rate": 7.97645175513195e-06,
+      "loss": 1.5178,
+      "step": 551
+    },
+    {
+      "epoch": 1.104,
+      "grad_norm": 8.402973175048828,
+      "learning_rate": 7.967094433018508e-06,
+      "loss": 1.571,
+      "step": 552
+    },
+    {
+      "epoch": 1.106,
+      "grad_norm": 9.366331100463867,
+      "learning_rate": 7.95772104284742e-06,
+      "loss": 1.4459,
+      "step": 553
+    },
+    {
+      "epoch": 1.108,
+      "grad_norm": 9.664446830749512,
+      "learning_rate": 7.948331635379517e-06,
+      "loss": 1.5755,
+      "step": 554
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 10.556148529052734,
+      "learning_rate": 7.938926261462366e-06,
+      "loss": 1.5738,
+      "step": 555
+    },
+    {
+      "epoch": 1.112,
+      "grad_norm": 10.46329116821289,
+      "learning_rate": 7.929504972030003e-06,
+      "loss": 1.7149,
+      "step": 556
+    },
+    {
+      "epoch": 1.114,
+      "grad_norm": 9.829599380493164,
+      "learning_rate": 7.920067818102652e-06,
+      "loss": 1.5629,
+      "step": 557
+    },
+    {
+      "epoch": 1.116,
+      "grad_norm": 8.370344161987305,
+      "learning_rate": 7.910614850786448e-06,
+      "loss": 1.6013,
+      "step": 558
+    },
+    {
+      "epoch": 1.1179999999999999,
+      "grad_norm": 9.646671295166016,
+      "learning_rate": 7.901146121273165e-06,
+      "loss": 1.6073,
+      "step": 559
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 8.669224739074707,
+      "learning_rate": 7.891661680839932e-06,
+      "loss": 1.5163,
+      "step": 560
+    },
+    {
+      "epoch": 1.1219999999999999,
+      "grad_norm": 11.22138500213623,
+      "learning_rate": 7.882161580848966e-06,
+      "loss": 1.6723,
+      "step": 561
+    },
+    {
+      "epoch": 1.124,
+      "grad_norm": 9.086870193481445,
+      "learning_rate": 7.872645872747281e-06,
+      "loss": 1.4885,
+      "step": 562
+    },
+    {
+      "epoch": 1.126,
+      "grad_norm": 11.861745834350586,
+      "learning_rate": 7.863114608066417e-06,
+      "loss": 1.6129,
+      "step": 563
+    },
+    {
+      "epoch": 1.1280000000000001,
+      "grad_norm": 7.978770732879639,
+      "learning_rate": 7.85356783842216e-06,
+      "loss": 1.6875,
+      "step": 564
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.279024124145508,
+      "learning_rate": 7.84400561551426e-06,
+      "loss": 1.5874,
+      "step": 565
+    },
+    {
+      "epoch": 1.1320000000000001,
+      "grad_norm": 10.196725845336914,
+      "learning_rate": 7.834427991126155e-06,
+      "loss": 1.5669,
+      "step": 566
+    },
+    {
+      "epoch": 1.134,
+      "grad_norm": 12.994475364685059,
+      "learning_rate": 7.82483501712469e-06,
+      "loss": 1.6626,
+      "step": 567
+    },
+    {
+      "epoch": 1.1360000000000001,
+      "grad_norm": 11.740025520324707,
+      "learning_rate": 7.815226745459831e-06,
+      "loss": 1.5378,
+      "step": 568
+    },
+    {
+      "epoch": 1.138,
+      "grad_norm": 8.01264762878418,
+      "learning_rate": 7.80560322816439e-06,
+      "loss": 1.592,
+      "step": 569
+    },
+    {
+      "epoch": 1.1400000000000001,
+      "grad_norm": 9.302233695983887,
+      "learning_rate": 7.795964517353734e-06,
+      "loss": 1.5832,
+      "step": 570
+    },
+    {
+      "epoch": 1.142,
+      "grad_norm": 9.458878517150879,
+      "learning_rate": 7.786310665225522e-06,
+      "loss": 1.6201,
+      "step": 571
+    },
+    {
+      "epoch": 1.144,
+      "grad_norm": 19.037525177001953,
+      "learning_rate": 7.776641724059398e-06,
+      "loss": 1.7447,
+      "step": 572
+    },
+    {
+      "epoch": 1.146,
+      "grad_norm": 11.674372673034668,
+      "learning_rate": 7.76695774621672e-06,
+      "loss": 1.6556,
+      "step": 573
+    },
+    {
+      "epoch": 1.148,
+      "grad_norm": 11.437299728393555,
+      "learning_rate": 7.757258784140286e-06,
+      "loss": 1.6109,
+      "step": 574
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.853487014770508,
+      "learning_rate": 7.747544890354031e-06,
+      "loss": 1.5528,
+      "step": 575
+    },
+    {
+      "epoch": 1.152,
+      "grad_norm": 6.265218734741211,
+      "learning_rate": 7.737816117462752e-06,
+      "loss": 1.5691,
+      "step": 576
+    },
+    {
+      "epoch": 1.154,
+      "grad_norm": 8.988601684570312,
+      "learning_rate": 7.728072518151826e-06,
+      "loss": 1.646,
+      "step": 577
+    },
+    {
+      "epoch": 1.156,
+      "grad_norm": 10.684486389160156,
+      "learning_rate": 7.718314145186918e-06,
+      "loss": 1.6239,
+      "step": 578
+    },
+    {
+      "epoch": 1.158,
+      "grad_norm": 8.397745132446289,
+      "learning_rate": 7.7085410514137e-06,
+      "loss": 1.5766,
+      "step": 579
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 8.635762214660645,
+      "learning_rate": 7.698753289757565e-06,
+      "loss": 1.5302,
+      "step": 580
+    },
+    {
+      "epoch": 1.162,
+      "grad_norm": 11.5667085647583,
+      "learning_rate": 7.688950913223336e-06,
+      "loss": 1.5859,
+      "step": 581
+    },
+    {
+      "epoch": 1.164,
+      "grad_norm": 6.988083839416504,
+      "learning_rate": 7.679133974894984e-06,
+      "loss": 1.6023,
+      "step": 582
+    },
+    {
+      "epoch": 1.166,
+      "grad_norm": 8.85438346862793,
+      "learning_rate": 7.669302527935334e-06,
+      "loss": 1.7383,
+      "step": 583
+    },
+    {
+      "epoch": 1.168,
+      "grad_norm": 8.871450424194336,
+      "learning_rate": 7.65945662558579e-06,
+      "loss": 1.5443,
+      "step": 584
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 6.740708827972412,
+      "learning_rate": 7.649596321166024e-06,
+      "loss": 1.581,
+      "step": 585
+    },
+    {
+      "epoch": 1.172,
+      "grad_norm": 7.056929111480713,
+      "learning_rate": 7.639721668073718e-06,
+      "loss": 1.4831,
+      "step": 586
+    },
+    {
+      "epoch": 1.174,
+      "grad_norm": 9.177947998046875,
+      "learning_rate": 7.629832719784245e-06,
+      "loss": 1.5719,
+      "step": 587
+    },
+    {
+      "epoch": 1.176,
+      "grad_norm": 9.50424861907959,
+      "learning_rate": 7.619929529850397e-06,
+      "loss": 1.5951,
+      "step": 588
+    },
+    {
+      "epoch": 1.178,
+      "grad_norm": 7.7955498695373535,
+      "learning_rate": 7.610012151902091e-06,
+      "loss": 1.6338,
+      "step": 589
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.096726417541504,
+      "learning_rate": 7.600080639646077e-06,
+      "loss": 1.6591,
+      "step": 590
+    },
+    {
+      "epoch": 1.182,
+      "grad_norm": 8.232474327087402,
+      "learning_rate": 7.590135046865652e-06,
+      "loss": 1.4726,
+      "step": 591
+    },
+    {
+      "epoch": 1.184,
+      "grad_norm": 8.038969039916992,
+      "learning_rate": 7.580175427420358e-06,
+      "loss": 1.7216,
+      "step": 592
+    },
+    {
+      "epoch": 1.186,
+      "grad_norm": 10.49866771697998,
+      "learning_rate": 7.570201835245703e-06,
+      "loss": 1.55,
+      "step": 593
+    },
+    {
+      "epoch": 1.188,
+      "grad_norm": 8.377806663513184,
+      "learning_rate": 7.560214324352858e-06,
+      "loss": 1.5407,
+      "step": 594
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 11.32930850982666,
+      "learning_rate": 7.550212948828377e-06,
+      "loss": 1.5192,
+      "step": 595
+    },
+    {
+      "epoch": 1.192,
+      "grad_norm": 9.1795015335083,
+      "learning_rate": 7.54019776283389e-06,
+      "loss": 1.5734,
+      "step": 596
+    },
+    {
+      "epoch": 1.194,
+      "grad_norm": 8.019397735595703,
+      "learning_rate": 7.530168820605819e-06,
+      "loss": 1.5216,
+      "step": 597
+    },
+    {
+      "epoch": 1.196,
+      "grad_norm": 25.26226234436035,
+      "learning_rate": 7.520126176455084e-06,
+      "loss": 1.6976,
+      "step": 598
+    },
+    {
+      "epoch": 1.198,
+      "grad_norm": 9.354122161865234,
+      "learning_rate": 7.510069884766802e-06,
+      "loss": 1.5146,
+      "step": 599
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 6.69480037689209,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 1.5678,
+      "step": 600
+    },
+    {
+      "epoch": 1.202,
+      "grad_norm": 8.49975299835205,
+      "learning_rate": 7.489916576687318e-06,
+      "loss": 1.5507,
+      "step": 601
+    },
+    {
+      "epoch": 1.204,
+      "grad_norm": 6.0479416847229,
+      "learning_rate": 7.479819669434712e-06,
+      "loss": 1.5298,
+      "step": 602
+    },
+    {
+      "epoch": 1.206,
+      "grad_norm": 12.379949569702148,
+      "learning_rate": 7.469709332921155e-06,
+      "loss": 1.5703,
+      "step": 603
+    },
+    {
+      "epoch": 1.208,
+      "grad_norm": 9.304234504699707,
+      "learning_rate": 7.459585621898353e-06,
+      "loss": 1.4965,
+      "step": 604
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 8.749626159667969,
+      "learning_rate": 7.449448591190436e-06,
+      "loss": 1.5667,
+      "step": 605
+    },
+    {
+      "epoch": 1.212,
+      "grad_norm": 12.647852897644043,
+      "learning_rate": 7.4392982956936644e-06,
+      "loss": 1.518,
+      "step": 606
+    },
+    {
+      "epoch": 1.214,
+      "grad_norm": 11.38586139678955,
+      "learning_rate": 7.429134790376136e-06,
+      "loss": 1.5827,
+      "step": 607
+    },
+    {
+      "epoch": 1.216,
+      "grad_norm": 9.004302978515625,
+      "learning_rate": 7.418958130277483e-06,
+      "loss": 1.5788,
+      "step": 608
+    },
+    {
+      "epoch": 1.218,
+      "grad_norm": 12.129622459411621,
+      "learning_rate": 7.408768370508577e-06,
+      "loss": 1.6698,
+      "step": 609
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 8.174181938171387,
+      "learning_rate": 7.398565566251232e-06,
+      "loss": 1.6104,
+      "step": 610
+    },
+    {
+      "epoch": 1.222,
+      "grad_norm": 9.542710304260254,
+      "learning_rate": 7.3883497727579e-06,
+      "loss": 1.527,
+      "step": 611
+    },
+    {
+      "epoch": 1.224,
+      "grad_norm": 8.519256591796875,
+      "learning_rate": 7.378121045351378e-06,
+      "loss": 1.6646,
+      "step": 612
+    },
+    {
+      "epoch": 1.226,
+      "grad_norm": 10.969525337219238,
+      "learning_rate": 7.3678794394245e-06,
+      "loss": 1.5838,
+      "step": 613
+    },
+    {
+      "epoch": 1.228,
+      "grad_norm": 12.213068008422852,
+      "learning_rate": 7.357625010439853e-06,
+      "loss": 1.5068,
+      "step": 614
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 14.603283882141113,
+      "learning_rate": 7.347357813929455e-06,
+      "loss": 1.7072,
+      "step": 615
+    },
+    {
+      "epoch": 1.232,
+      "grad_norm": 9.047625541687012,
+      "learning_rate": 7.337077905494472e-06,
+      "loss": 1.5516,
+      "step": 616
+    },
+    {
+      "epoch": 1.234,
+      "grad_norm": 8.961601257324219,
+      "learning_rate": 7.326785340804908e-06,
+      "loss": 1.6053,
+      "step": 617
+    },
+    {
+      "epoch": 1.236,
+      "grad_norm": 10.050729751586914,
+      "learning_rate": 7.31648017559931e-06,
+      "loss": 1.5849,
+      "step": 618
+    },
+    {
+      "epoch": 1.238,
+      "grad_norm": 6.469438076019287,
+      "learning_rate": 7.3061624656844544e-06,
+      "loss": 1.6075,
+      "step": 619
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 11.897672653198242,
+      "learning_rate": 7.295832266935059e-06,
+      "loss": 1.5038,
+      "step": 620
+    },
+    {
+      "epoch": 1.242,
+      "grad_norm": 7.534898281097412,
+      "learning_rate": 7.285489635293472e-06,
+      "loss": 1.5344,
+      "step": 621
+    },
+    {
+      "epoch": 1.244,
+      "grad_norm": 8.457494735717773,
+      "learning_rate": 7.275134626769369e-06,
+      "loss": 1.6063,
+      "step": 622
+    },
+    {
+      "epoch": 1.246,
+      "grad_norm": 7.499537467956543,
+      "learning_rate": 7.264767297439455e-06,
+      "loss": 1.5974,
+      "step": 623
+    },
+    {
+      "epoch": 1.248,
+      "grad_norm": 10.001435279846191,
+      "learning_rate": 7.254387703447154e-06,
+      "loss": 1.6185,
+      "step": 624
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.439517974853516,
+      "learning_rate": 7.243995901002312e-06,
+      "loss": 1.6371,
+      "step": 625
+    },
+    {
+      "epoch": 1.252,
+      "grad_norm": 8.107089042663574,
+      "learning_rate": 7.233591946380884e-06,
+      "loss": 1.5729,
+      "step": 626
+    },
+    {
+      "epoch": 1.254,
+      "grad_norm": 9.131059646606445,
+      "learning_rate": 7.223175895924638e-06,
+      "loss": 1.5677,
+      "step": 627
+    },
+    {
+      "epoch": 1.256,
+      "grad_norm": 6.757850646972656,
+      "learning_rate": 7.212747806040845e-06,
+      "loss": 1.5922,
+      "step": 628
+    },
+    {
+      "epoch": 1.258,
+      "grad_norm": 9.472972869873047,
+      "learning_rate": 7.2023077332019755e-06,
+      "loss": 1.6055,
+      "step": 629
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 7.784339427947998,
+      "learning_rate": 7.191855733945388e-06,
+      "loss": 1.461,
+      "step": 630
+    },
+    {
+      "epoch": 1.262,
+      "grad_norm": 9.098430633544922,
+      "learning_rate": 7.181391864873034e-06,
+      "loss": 1.5214,
+      "step": 631
+    },
+    {
+      "epoch": 1.264,
+      "grad_norm": 8.37144947052002,
+      "learning_rate": 7.170916182651141e-06,
+      "loss": 1.5766,
+      "step": 632
+    },
+    {
+      "epoch": 1.266,
+      "grad_norm": 9.366813659667969,
+      "learning_rate": 7.160428744009913e-06,
+      "loss": 1.5955,
+      "step": 633
+    },
+    {
+      "epoch": 1.268,
+      "grad_norm": 11.189432144165039,
+      "learning_rate": 7.149929605743214e-06,
+      "loss": 1.6196,
+      "step": 634
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.027175903320312,
+      "learning_rate": 7.1394188247082715e-06,
+      "loss": 1.5104,
+      "step": 635
+    },
+    {
+      "epoch": 1.272,
+      "grad_norm": 7.792656898498535,
+      "learning_rate": 7.128896457825364e-06,
+      "loss": 1.6022,
+      "step": 636
+    },
+    {
+      "epoch": 1.274,
+      "grad_norm": 7.127891540527344,
+      "learning_rate": 7.118362562077508e-06,
+      "loss": 1.624,
+      "step": 637
+    },
+    {
+      "epoch": 1.276,
+      "grad_norm": 7.897225379943848,
+      "learning_rate": 7.107817194510157e-06,
+      "loss": 1.6961,
+      "step": 638
+    },
+    {
+      "epoch": 1.278,
+      "grad_norm": 8.84518051147461,
+      "learning_rate": 7.0972604122308865e-06,
+      "loss": 1.4578,
+      "step": 639
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 7.849598407745361,
+      "learning_rate": 7.08669227240909e-06,
+      "loss": 1.6101,
+      "step": 640
+    },
+    {
+      "epoch": 1.282,
+      "grad_norm": 9.6469144821167,
+      "learning_rate": 7.076112832275667e-06,
+      "loss": 1.6093,
+      "step": 641
+    },
+    {
+      "epoch": 1.284,
+      "grad_norm": 5.705495834350586,
+      "learning_rate": 7.06552214912271e-06,
+      "loss": 1.5104,
+      "step": 642
+    },
+    {
+      "epoch": 1.286,
+      "grad_norm": 7.658092975616455,
+      "learning_rate": 7.054920280303199e-06,
+      "loss": 1.666,
+      "step": 643
+    },
+    {
+      "epoch": 1.288,
+      "grad_norm": 7.542886734008789,
+      "learning_rate": 7.04430728323069e-06,
+      "loss": 1.6318,
+      "step": 644
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.100532531738281,
+      "learning_rate": 7.033683215379002e-06,
+      "loss": 1.6983,
+      "step": 645
+    },
+    {
+      "epoch": 1.292,
+      "grad_norm": 7.658532619476318,
+      "learning_rate": 7.023048134281907e-06,
+      "loss": 1.5868,
+      "step": 646
+    },
+    {
+      "epoch": 1.294,
+      "grad_norm": 10.516380310058594,
+      "learning_rate": 7.012402097532815e-06,
+      "loss": 1.489,
+      "step": 647
+    },
+    {
+      "epoch": 1.296,
+      "grad_norm": 7.941004276275635,
+      "learning_rate": 7.0017451627844765e-06,
+      "loss": 1.6114,
+      "step": 648
+    },
+    {
+      "epoch": 1.298,
+      "grad_norm": 8.073655128479004,
+      "learning_rate": 6.991077387748643e-06,
+      "loss": 1.6023,
+      "step": 649
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 8.82116985321045,
+      "learning_rate": 6.980398830195785e-06,
+      "loss": 1.5371,
+      "step": 650
+    },
+    {
+      "epoch": 1.302,
+      "grad_norm": 10.482192039489746,
+      "learning_rate": 6.9697095479547564e-06,
+      "loss": 1.6303,
+      "step": 651
+    },
+    {
+      "epoch": 1.304,
+      "grad_norm": 12.892633438110352,
+      "learning_rate": 6.959009598912493e-06,
+      "loss": 1.4729,
+      "step": 652
+    },
+    {
+      "epoch": 1.306,
+      "grad_norm": 7.209946632385254,
+      "learning_rate": 6.948299041013695e-06,
+      "loss": 1.493,
+      "step": 653
+    },
+    {
+      "epoch": 1.308,
+      "grad_norm": 7.52318811416626,
+      "learning_rate": 6.9375779322605154e-06,
+      "loss": 1.5806,
+      "step": 654
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 10.265365600585938,
+      "learning_rate": 6.9268463307122425e-06,
+      "loss": 1.5159,
+      "step": 655
+    },
+    {
+      "epoch": 1.312,
+      "grad_norm": 10.420729637145996,
+      "learning_rate": 6.916104294484988e-06,
+      "loss": 1.5097,
+      "step": 656
+    },
+    {
+      "epoch": 1.314,
+      "grad_norm": 7.465059280395508,
+      "learning_rate": 6.905351881751372e-06,
+      "loss": 1.6499,
+      "step": 657
+    },
+    {
+      "epoch": 1.316,
+      "grad_norm": 9.007606506347656,
+      "learning_rate": 6.8945891507402075e-06,
+      "loss": 1.5783,
+      "step": 658
+    },
+    {
+      "epoch": 1.318,
+      "grad_norm": 11.8942289352417,
+      "learning_rate": 6.883816159736187e-06,
+      "loss": 1.6481,
+      "step": 659
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 10.41662311553955,
+      "learning_rate": 6.873032967079562e-06,
+      "loss": 1.6615,
+      "step": 660
+    },
+    {
+      "epoch": 1.322,
+      "grad_norm": 11.450822830200195,
+      "learning_rate": 6.862239631165831e-06,
+      "loss": 1.6254,
+      "step": 661
+    },
+    {
+      "epoch": 1.324,
+      "grad_norm": 8.604270935058594,
+      "learning_rate": 6.851436210445427e-06,
+      "loss": 1.6191,
+      "step": 662
+    },
+    {
+      "epoch": 1.326,
+      "grad_norm": 7.488931655883789,
+      "learning_rate": 6.840622763423391e-06,
+      "loss": 1.6588,
+      "step": 663
+    },
+    {
+      "epoch": 1.328,
+      "grad_norm": 9.091219902038574,
+      "learning_rate": 6.829799348659061e-06,
+      "loss": 1.6639,
+      "step": 664
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 8.698543548583984,
+      "learning_rate": 6.818966024765758e-06,
+      "loss": 1.5772,
+      "step": 665
+    },
+    {
+      "epoch": 1.332,
+      "grad_norm": 13.658153533935547,
+      "learning_rate": 6.808122850410461e-06,
+      "loss": 1.4672,
+      "step": 666
+    },
+    {
+      "epoch": 1.334,
+      "grad_norm": 26.972639083862305,
+      "learning_rate": 6.7972698843135e-06,
+      "loss": 1.6459,
+      "step": 667
+    },
+    {
+      "epoch": 1.336,
+      "grad_norm": 9.072081565856934,
+      "learning_rate": 6.7864071852482205e-06,
+      "loss": 1.6711,
+      "step": 668
+    },
+    {
+      "epoch": 1.338,
+      "grad_norm": 8.552204132080078,
+      "learning_rate": 6.775534812040686e-06,
+      "loss": 1.5927,
+      "step": 669
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.554261207580566,
+      "learning_rate": 6.7646528235693445e-06,
+      "loss": 1.6078,
+      "step": 670
+    },
+    {
+      "epoch": 1.342,
+      "grad_norm": 9.100646018981934,
+      "learning_rate": 6.753761278764719e-06,
+      "loss": 1.6746,
+      "step": 671
+    },
+    {
+      "epoch": 1.3439999999999999,
+      "grad_norm": 10.204363822937012,
+      "learning_rate": 6.7428602366090764e-06,
+      "loss": 1.626,
+      "step": 672
+    },
+    {
+      "epoch": 1.346,
+      "grad_norm": 10.063047409057617,
+      "learning_rate": 6.7319497561361245e-06,
+      "loss": 1.6601,
+      "step": 673
+    },
+    {
+      "epoch": 1.3479999999999999,
+      "grad_norm": 13.726625442504883,
+      "learning_rate": 6.721029896430678e-06,
+      "loss": 1.583,
+      "step": 674
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 11.784052848815918,
+      "learning_rate": 6.710100716628345e-06,
+      "loss": 1.6688,
+      "step": 675
+    },
+    {
+      "epoch": 1.3519999999999999,
+      "grad_norm": 8.559849739074707,
+      "learning_rate": 6.699162275915208e-06,
+      "loss": 1.6578,
+      "step": 676
+    },
+    {
+      "epoch": 1.354,
+      "grad_norm": 9.396211624145508,
+      "learning_rate": 6.6882146335274955e-06,
+      "loss": 1.6294,
+      "step": 677
+    },
+    {
+      "epoch": 1.3559999999999999,
+      "grad_norm": 7.74012565612793,
+      "learning_rate": 6.677257848751276e-06,
+      "loss": 1.5466,
+      "step": 678
+    },
+    {
+      "epoch": 1.358,
+      "grad_norm": 10.365443229675293,
+      "learning_rate": 6.666291980922122e-06,
+      "loss": 1.5296,
+      "step": 679
+    },
+    {
+      "epoch": 1.3599999999999999,
+      "grad_norm": 7.977329254150391,
+      "learning_rate": 6.655317089424791e-06,
+      "loss": 1.5837,
+      "step": 680
+    },
+    {
+      "epoch": 1.362,
+      "grad_norm": 8.795341491699219,
+      "learning_rate": 6.644333233692917e-06,
+      "loss": 1.5884,
+      "step": 681
+    },
+    {
+      "epoch": 1.3639999999999999,
+      "grad_norm": 13.247668266296387,
+      "learning_rate": 6.633340473208673e-06,
+      "loss": 1.5977,
+      "step": 682
+    },
+    {
+      "epoch": 1.366,
+      "grad_norm": 8.479852676391602,
+      "learning_rate": 6.622338867502452e-06,
+      "loss": 1.5983,
+      "step": 683
+    },
+    {
+      "epoch": 1.3679999999999999,
+      "grad_norm": 8.384832382202148,
+      "learning_rate": 6.611328476152557e-06,
+      "loss": 1.5377,
+      "step": 684
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 7.661881923675537,
+      "learning_rate": 6.600309358784858e-06,
+      "loss": 1.4785,
+      "step": 685
+    },
+    {
+      "epoch": 1.3719999999999999,
+      "grad_norm": 9.14194107055664,
+      "learning_rate": 6.58928157507249e-06,
+      "loss": 1.5439,
+      "step": 686
+    },
+    {
+      "epoch": 1.374,
+      "grad_norm": 8.849069595336914,
+      "learning_rate": 6.578245184735513e-06,
+      "loss": 1.5466,
+      "step": 687
+    },
+    {
+      "epoch": 1.376,
+      "grad_norm": 8.54023551940918,
+      "learning_rate": 6.567200247540599e-06,
+      "loss": 1.5738,
+      "step": 688
+    },
+    {
+      "epoch": 1.3780000000000001,
+      "grad_norm": 13.907352447509766,
+      "learning_rate": 6.556146823300701e-06,
+      "loss": 1.6541,
+      "step": 689
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 10.435159683227539,
+      "learning_rate": 6.545084971874738e-06,
+      "loss": 1.6579,
+      "step": 690
+    },
+    {
+      "epoch": 1.3820000000000001,
+      "grad_norm": 9.772438049316406,
+      "learning_rate": 6.534014753167263e-06,
+      "loss": 1.6995,
+      "step": 691
+    },
+    {
+      "epoch": 1.384,
+      "grad_norm": 9.34780502319336,
+      "learning_rate": 6.522936227128139e-06,
+      "loss": 1.5157,
+      "step": 692
+    },
+    {
+      "epoch": 1.3860000000000001,
+      "grad_norm": 8.92451000213623,
+      "learning_rate": 6.5118494537522235e-06,
+      "loss": 1.6148,
+      "step": 693
+    },
+    {
+      "epoch": 1.388,
+      "grad_norm": 6.892347812652588,
+      "learning_rate": 6.500754493079029e-06,
+      "loss": 1.5511,
+      "step": 694
+    },
+    {
+      "epoch": 1.3900000000000001,
+      "grad_norm": 10.89332389831543,
+      "learning_rate": 6.48965140519241e-06,
+      "loss": 1.5804,
+      "step": 695
+    },
+    {
+      "epoch": 1.392,
+      "grad_norm": 7.6322221755981445,
+      "learning_rate": 6.4785402502202345e-06,
+      "loss": 1.6086,
+      "step": 696
+    },
+    {
+      "epoch": 1.3940000000000001,
+      "grad_norm": 8.542652130126953,
+      "learning_rate": 6.467421088334052e-06,
+      "loss": 1.6496,
+      "step": 697
+    },
+    {
+      "epoch": 1.396,
+      "grad_norm": 8.015932083129883,
+      "learning_rate": 6.456293979748778e-06,
+      "loss": 1.617,
+      "step": 698
+    },
+    {
+      "epoch": 1.3980000000000001,
+      "grad_norm": 7.57454776763916,
+      "learning_rate": 6.445158984722358e-06,
+      "loss": 1.567,
+      "step": 699
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 8.950959205627441,
+      "learning_rate": 6.434016163555452e-06,
+      "loss": 1.5682,
+      "step": 700
+    },
+    {
+      "epoch": 1.4020000000000001,
+      "grad_norm": 8.991364479064941,
+      "learning_rate": 6.422865576591096e-06,
+      "loss": 1.5012,
+      "step": 701
+    },
+    {
+      "epoch": 1.404,
+      "grad_norm": 7.7304182052612305,
+      "learning_rate": 6.411707284214384e-06,
+      "loss": 1.5961,
+      "step": 702
+    },
+    {
+      "epoch": 1.4060000000000001,
+      "grad_norm": 7.197605609893799,
+      "learning_rate": 6.400541346852136e-06,
+      "loss": 1.6654,
+      "step": 703
+    },
+    {
+      "epoch": 1.408,
+      "grad_norm": 7.552187442779541,
+      "learning_rate": 6.389367824972575e-06,
+      "loss": 1.5137,
+      "step": 704
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.499285697937012,
+      "learning_rate": 6.378186779084996e-06,
+      "loss": 1.5841,
+      "step": 705
+    },
+    {
+      "epoch": 1.412,
+      "grad_norm": 9.511728286743164,
+      "learning_rate": 6.366998269739442e-06,
+      "loss": 1.5504,
+      "step": 706
+    },
+    {
+      "epoch": 1.414,
+      "grad_norm": 8.45433521270752,
+      "learning_rate": 6.35580235752637e-06,
+      "loss": 1.5983,
+      "step": 707
+    },
+    {
+      "epoch": 1.416,
+      "grad_norm": 9.685602188110352,
+      "learning_rate": 6.344599103076329e-06,
+      "loss": 1.5785,
+      "step": 708
+    },
+    {
+      "epoch": 1.418,
+      "grad_norm": 9.533397674560547,
+      "learning_rate": 6.3333885670596285e-06,
+      "loss": 1.5921,
+      "step": 709
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 10.1636381149292,
+      "learning_rate": 6.322170810186013e-06,
+      "loss": 1.6896,
+      "step": 710
+    },
+    {
+      "epoch": 1.422,
+      "grad_norm": 9.969033241271973,
+      "learning_rate": 6.310945893204324e-06,
+      "loss": 1.6277,
+      "step": 711
+    },
+    {
+      "epoch": 1.424,
+      "grad_norm": 13.378399848937988,
+      "learning_rate": 6.299713876902188e-06,
+      "loss": 1.6618,
+      "step": 712
+    },
+    {
+      "epoch": 1.426,
+      "grad_norm": 7.862060546875,
+      "learning_rate": 6.28847482210567e-06,
+      "loss": 1.5287,
+      "step": 713
+    },
+    {
+      "epoch": 1.428,
+      "grad_norm": 8.11763858795166,
+      "learning_rate": 6.277228789678953e-06,
+      "loss": 1.6417,
+      "step": 714
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 8.063044548034668,
+      "learning_rate": 6.26597584052401e-06,
+      "loss": 1.6309,
+      "step": 715
+    },
+    {
+      "epoch": 1.432,
+      "grad_norm": 6.405232906341553,
+      "learning_rate": 6.254716035580264e-06,
+      "loss": 1.5913,
+      "step": 716
+    },
+    {
+      "epoch": 1.434,
+      "grad_norm": 6.900762557983398,
+      "learning_rate": 6.243449435824276e-06,
+      "loss": 1.5563,
+      "step": 717
+    },
+    {
+      "epoch": 1.436,
+      "grad_norm": 10.574906349182129,
+      "learning_rate": 6.23217610226939e-06,
+      "loss": 1.601,
+      "step": 718
+    },
+    {
+      "epoch": 1.438,
+      "grad_norm": 7.271671295166016,
+      "learning_rate": 6.220896095965428e-06,
+      "loss": 1.5809,
+      "step": 719
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 8.701942443847656,
+      "learning_rate": 6.209609477998339e-06,
+      "loss": 1.588,
+      "step": 720
+    },
+    {
+      "epoch": 1.442,
+      "grad_norm": 8.735617637634277,
+      "learning_rate": 6.198316309489886e-06,
+      "loss": 1.5435,
+      "step": 721
+    },
+    {
+      "epoch": 1.444,
+      "grad_norm": 7.150242328643799,
+      "learning_rate": 6.187016651597299e-06,
+      "loss": 1.6309,
+      "step": 722
+    },
+    {
+      "epoch": 1.446,
+      "grad_norm": 9.113722801208496,
+      "learning_rate": 6.17571056551295e-06,
+      "loss": 1.6128,
+      "step": 723
+    },
+    {
+      "epoch": 1.448,
+      "grad_norm": 9.70335865020752,
+      "learning_rate": 6.16439811246403e-06,
+      "loss": 1.5812,
+      "step": 724
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 7.478845596313477,
+      "learning_rate": 6.153079353712201e-06,
+      "loss": 1.653,
+      "step": 725
+    },
+    {
+      "epoch": 1.452,
+      "grad_norm": 7.009204864501953,
+      "learning_rate": 6.141754350553279e-06,
+      "loss": 1.432,
+      "step": 726
+    },
+    {
+      "epoch": 1.454,
+      "grad_norm": 8.384693145751953,
+      "learning_rate": 6.130423164316893e-06,
+      "loss": 1.6201,
+      "step": 727
+    },
+    {
+      "epoch": 1.456,
+      "grad_norm": 12.53249454498291,
+      "learning_rate": 6.119085856366158e-06,
+      "loss": 1.6526,
+      "step": 728
+    },
+    {
+      "epoch": 1.458,
+      "grad_norm": 7.682470321655273,
+      "learning_rate": 6.107742488097338e-06,
+      "loss": 1.5562,
+      "step": 729
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 10.752216339111328,
+      "learning_rate": 6.0963931209395165e-06,
+      "loss": 1.6478,
+      "step": 730
+    },
+    {
+      "epoch": 1.462,
+      "grad_norm": 9.087852478027344,
+      "learning_rate": 6.085037816354269e-06,
+      "loss": 1.5929,
+      "step": 731
+    },
+    {
+      "epoch": 1.464,
+      "grad_norm": 7.526221752166748,
+      "learning_rate": 6.073676635835317e-06,
+      "loss": 1.588,
+      "step": 732
+    },
+    {
+      "epoch": 1.466,
+      "grad_norm": 7.148344993591309,
+      "learning_rate": 6.062309640908206e-06,
+      "loss": 1.575,
+      "step": 733
+    },
+    {
+      "epoch": 1.468,
+      "grad_norm": 10.243959426879883,
+      "learning_rate": 6.05093689312997e-06,
+      "loss": 1.5805,
+      "step": 734
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 8.489969253540039,
+      "learning_rate": 6.039558454088796e-06,
+      "loss": 1.4469,
+      "step": 735
+    },
+    {
+      "epoch": 1.472,
+      "grad_norm": 7.80888557434082,
+      "learning_rate": 6.028174385403693e-06,
+      "loss": 1.6703,
+      "step": 736
+    },
+    {
+      "epoch": 1.474,
+      "grad_norm": 8.337100982666016,
+      "learning_rate": 6.016784748724153e-06,
+      "loss": 1.5589,
+      "step": 737
+    },
+    {
+      "epoch": 1.476,
+      "grad_norm": 7.308524131774902,
+      "learning_rate": 6.005389605729824e-06,
+      "loss": 1.5458,
+      "step": 738
+    },
+    {
+      "epoch": 1.478,
+      "grad_norm": 10.208418846130371,
+      "learning_rate": 5.993989018130173e-06,
+      "loss": 1.5389,
+      "step": 739
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 6.792656421661377,
+      "learning_rate": 5.982583047664151e-06,
+      "loss": 1.574,
+      "step": 740
+    },
+    {
+      "epoch": 1.482,
+      "grad_norm": 7.216954708099365,
+      "learning_rate": 5.97117175609986e-06,
+      "loss": 1.6285,
+      "step": 741
+    },
+    {
+      "epoch": 1.484,
+      "grad_norm": 7.5151238441467285,
+      "learning_rate": 5.9597552052342174e-06,
+      "loss": 1.536,
+      "step": 742
+    },
+    {
+      "epoch": 1.486,
+      "grad_norm": 8.329843521118164,
+      "learning_rate": 5.948333456892624e-06,
+      "loss": 1.564,
+      "step": 743
+    },
+    {
+      "epoch": 1.488,
+      "grad_norm": 7.076045036315918,
+      "learning_rate": 5.936906572928625e-06,
+      "loss": 1.5722,
+      "step": 744
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 8.269755363464355,
+      "learning_rate": 5.925474615223573e-06,
+      "loss": 1.6673,
+      "step": 745
+    },
+    {
+      "epoch": 1.492,
+      "grad_norm": 9.766873359680176,
+      "learning_rate": 5.914037645686308e-06,
+      "loss": 1.4451,
+      "step": 746
+    },
+    {
+      "epoch": 1.494,
+      "grad_norm": 7.705737590789795,
+      "learning_rate": 5.902595726252801e-06,
+      "loss": 1.7146,
+      "step": 747
+    },
+    {
+      "epoch": 1.496,
+      "grad_norm": 6.665014266967773,
+      "learning_rate": 5.891148918885834e-06,
+      "loss": 1.498,
+      "step": 748
+    },
+    {
+      "epoch": 1.498,
+      "grad_norm": 8.403691291809082,
+      "learning_rate": 5.879697285574655e-06,
+      "loss": 1.6839,
+      "step": 749
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 6.893798351287842,
+      "learning_rate": 5.8682408883346535e-06,
+      "loss": 1.535,
+      "step": 750
+    },
+    {
+      "epoch": 1.502,
+      "grad_norm": 7.8562421798706055,
+      "learning_rate": 5.85677978920701e-06,
+      "loss": 1.5439,
+      "step": 751
+    },
+    {
+      "epoch": 1.504,
+      "grad_norm": 9.73377513885498,
+      "learning_rate": 5.84531405025837e-06,
+      "loss": 1.576,
+      "step": 752
+    },
+    {
+      "epoch": 1.506,
+      "grad_norm": 9.49201774597168,
+      "learning_rate": 5.8338437335805124e-06,
+      "loss": 1.7001,
+      "step": 753
+    },
+    {
+      "epoch": 1.508,
+      "grad_norm": 7.50467586517334,
+      "learning_rate": 5.8223689012899945e-06,
+      "loss": 1.6096,
+      "step": 754
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 7.861015319824219,
+      "learning_rate": 5.810889615527839e-06,
+      "loss": 1.6206,
+      "step": 755
+    },
+    {
+      "epoch": 1.512,
+      "grad_norm": 9.48017692565918,
+      "learning_rate": 5.799405938459175e-06,
+      "loss": 1.5395,
+      "step": 756
+    },
+    {
+      "epoch": 1.514,
+      "grad_norm": 8.29472827911377,
+      "learning_rate": 5.787917932272922e-06,
+      "loss": 1.567,
+      "step": 757
+    },
+    {
+      "epoch": 1.516,
+      "grad_norm": 9.242623329162598,
+      "learning_rate": 5.776425659181438e-06,
+      "loss": 1.4826,
+      "step": 758
+    },
+    {
+      "epoch": 1.518,
+      "grad_norm": 8.55409049987793,
+      "learning_rate": 5.764929181420191e-06,
+      "loss": 1.462,
+      "step": 759
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 8.088449478149414,
+      "learning_rate": 5.753428561247416e-06,
+      "loss": 1.6413,
+      "step": 760
+    },
+    {
+      "epoch": 1.522,
+      "grad_norm": 7.811099529266357,
+      "learning_rate": 5.741923860943783e-06,
+      "loss": 1.5257,
+      "step": 761
+    },
+    {
+      "epoch": 1.524,
+      "grad_norm": 7.643144130706787,
+      "learning_rate": 5.730415142812059e-06,
+      "loss": 1.638,
+      "step": 762
+    },
+    {
+      "epoch": 1.526,
+      "grad_norm": 8.378085136413574,
+      "learning_rate": 5.718902469176765e-06,
+      "loss": 1.6547,
+      "step": 763
+    },
+    {
+      "epoch": 1.528,
+      "grad_norm": 8.745218276977539,
+      "learning_rate": 5.707385902383845e-06,
+      "loss": 1.5213,
+      "step": 764
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 11.519359588623047,
+      "learning_rate": 5.695865504800328e-06,
+      "loss": 1.6252,
+      "step": 765
+    },
+    {
+      "epoch": 1.532,
+      "grad_norm": 9.585226058959961,
+      "learning_rate": 5.684341338813986e-06,
+      "loss": 1.5071,
+      "step": 766
+    },
+    {
+      "epoch": 1.534,
+      "grad_norm": 7.059016704559326,
+      "learning_rate": 5.672813466832998e-06,
+      "loss": 1.6355,
+      "step": 767
+    },
+    {
+      "epoch": 1.536,
+      "grad_norm": 7.046347618103027,
+      "learning_rate": 5.661281951285613e-06,
+      "loss": 1.7801,
+      "step": 768
+    },
+    {
+      "epoch": 1.538,
+      "grad_norm": 6.897932052612305,
+      "learning_rate": 5.649746854619814e-06,
+      "loss": 1.5833,
+      "step": 769
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 8.469090461730957,
+      "learning_rate": 5.638208239302975e-06,
+      "loss": 1.6203,
+      "step": 770
+    },
+    {
+      "epoch": 1.542,
+      "grad_norm": 10.167533874511719,
+      "learning_rate": 5.626666167821522e-06,
+      "loss": 1.5861,
+      "step": 771
+    },
+    {
+      "epoch": 1.544,
+      "grad_norm": 7.769595146179199,
+      "learning_rate": 5.615120702680604e-06,
+      "loss": 1.5514,
+      "step": 772
+    },
+    {
+      "epoch": 1.546,
+      "grad_norm": 7.826175689697266,
+      "learning_rate": 5.6035719064037446e-06,
+      "loss": 1.6232,
+      "step": 773
+    },
+    {
+      "epoch": 1.548,
+      "grad_norm": 7.963818550109863,
+      "learning_rate": 5.592019841532507e-06,
+      "loss": 1.5703,
+      "step": 774
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 8.253701210021973,
+      "learning_rate": 5.5804645706261515e-06,
+      "loss": 1.5996,
+      "step": 775
+    },
+    {
+      "epoch": 1.552,
+      "grad_norm": 10.267999649047852,
+      "learning_rate": 5.568906156261309e-06,
+      "loss": 1.5394,
+      "step": 776
+    },
+    {
+      "epoch": 1.554,
+      "grad_norm": 9.601211547851562,
+      "learning_rate": 5.557344661031628e-06,
+      "loss": 1.511,
+      "step": 777
+    },
+    {
+      "epoch": 1.556,
+      "grad_norm": 9.681934356689453,
+      "learning_rate": 5.54578014754744e-06,
+      "loss": 1.5503,
+      "step": 778
+    },
+    {
+      "epoch": 1.558,
+      "grad_norm": 8.93424129486084,
+      "learning_rate": 5.5342126784354265e-06,
+      "loss": 1.561,
+      "step": 779
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 9.094952583312988,
+      "learning_rate": 5.522642316338268e-06,
+      "loss": 1.5146,
+      "step": 780
+    },
+    {
+      "epoch": 1.562,
+      "grad_norm": 8.412158012390137,
+      "learning_rate": 5.511069123914319e-06,
+      "loss": 1.6634,
+      "step": 781
+    },
+    {
+      "epoch": 1.564,
+      "grad_norm": 9.692502975463867,
+      "learning_rate": 5.499493163837258e-06,
+      "loss": 1.5547,
+      "step": 782
+    },
+    {
+      "epoch": 1.5659999999999998,
+      "grad_norm": 6.936313152313232,
+      "learning_rate": 5.487914498795748e-06,
+      "loss": 1.5515,
+      "step": 783
+    },
+    {
+      "epoch": 1.568,
+      "grad_norm": 7.49924373626709,
+      "learning_rate": 5.476333191493108e-06,
+      "loss": 1.5958,
+      "step": 784
+    },
+    {
+      "epoch": 1.5699999999999998,
+      "grad_norm": 7.130555152893066,
+      "learning_rate": 5.464749304646963e-06,
+      "loss": 1.607,
+      "step": 785
+    },
+    {
+      "epoch": 1.572,
+      "grad_norm": 8.909390449523926,
+      "learning_rate": 5.453162900988902e-06,
+      "loss": 1.5895,
+      "step": 786
+    },
+    {
+      "epoch": 1.5739999999999998,
+      "grad_norm": 9.79128360748291,
+      "learning_rate": 5.44157404326415e-06,
+      "loss": 1.6367,
+      "step": 787
+    },
+    {
+      "epoch": 1.576,
+      "grad_norm": 8.873651504516602,
+      "learning_rate": 5.429982794231221e-06,
+      "loss": 1.5858,
+      "step": 788
+    },
+    {
+      "epoch": 1.5779999999999998,
+      "grad_norm": 9.033736228942871,
+      "learning_rate": 5.41838921666158e-06,
+      "loss": 1.6132,
+      "step": 789
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 8.515612602233887,
+      "learning_rate": 5.406793373339292e-06,
+      "loss": 1.5178,
+      "step": 790
+    },
+    {
+      "epoch": 1.5819999999999999,
+      "grad_norm": 8.608535766601562,
+      "learning_rate": 5.395195327060707e-06,
+      "loss": 1.5546,
+      "step": 791
+    },
+    {
+      "epoch": 1.584,
+      "grad_norm": 9.189397811889648,
+      "learning_rate": 5.383595140634093e-06,
+      "loss": 1.64,
+      "step": 792
+    },
+    {
+      "epoch": 1.5859999999999999,
+      "grad_norm": 9.675806045532227,
+      "learning_rate": 5.371992876879318e-06,
+      "loss": 1.6044,
+      "step": 793
+    },
+    {
+      "epoch": 1.588,
+      "grad_norm": 10.594379425048828,
+      "learning_rate": 5.360388598627487e-06,
+      "loss": 1.6525,
+      "step": 794
+    },
+    {
+      "epoch": 1.5899999999999999,
+      "grad_norm": 7.640488624572754,
+      "learning_rate": 5.348782368720627e-06,
+      "loss": 1.5487,
+      "step": 795
+    },
+    {
+      "epoch": 1.592,
+      "grad_norm": 8.902362823486328,
+      "learning_rate": 5.337174250011326e-06,
+      "loss": 1.5257,
+      "step": 796
+    },
+    {
+      "epoch": 1.5939999999999999,
+      "grad_norm": 9.681800842285156,
+      "learning_rate": 5.325564305362404e-06,
+      "loss": 1.5424,
+      "step": 797
+    },
+    {
+      "epoch": 1.596,
+      "grad_norm": 7.919555187225342,
+      "learning_rate": 5.3139525976465675e-06,
+      "loss": 1.4885,
+      "step": 798
+    },
+    {
+      "epoch": 1.5979999999999999,
+      "grad_norm": 11.47796630859375,
+      "learning_rate": 5.3023391897460715e-06,
+      "loss": 1.6287,
+      "step": 799
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 8.65668773651123,
+      "learning_rate": 5.290724144552379e-06,
+      "loss": 1.6287,
+      "step": 800
+    },
+    {
+      "epoch": 1.6019999999999999,
+      "grad_norm": 8.334498405456543,
+      "learning_rate": 5.27910752496582e-06,
+      "loss": 1.5546,
+      "step": 801
+    },
+    {
+      "epoch": 1.604,
+      "grad_norm": 9.96323013305664,
+      "learning_rate": 5.267489393895247e-06,
+      "loss": 1.658,
+      "step": 802
+    },
+    {
+      "epoch": 1.6059999999999999,
+      "grad_norm": 11.58132266998291,
+      "learning_rate": 5.255869814257701e-06,
+      "loss": 1.5074,
+      "step": 803
+    },
+    {
+      "epoch": 1.608,
+      "grad_norm": 10.45251178741455,
+      "learning_rate": 5.244248848978067e-06,
+      "loss": 1.6304,
+      "step": 804
+    },
+    {
+      "epoch": 1.6099999999999999,
+      "grad_norm": 9.885043144226074,
+      "learning_rate": 5.232626560988735e-06,
+      "loss": 1.6131,
+      "step": 805
+    },
+    {
+      "epoch": 1.612,
+      "grad_norm": 6.945827960968018,
+      "learning_rate": 5.221003013229253e-06,
+      "loss": 1.5726,
+      "step": 806
+    },
+    {
+      "epoch": 1.6139999999999999,
+      "grad_norm": 10.760123252868652,
+      "learning_rate": 5.209378268645998e-06,
+      "loss": 1.6255,
+      "step": 807
+    },
+    {
+      "epoch": 1.616,
+      "grad_norm": 8.588194847106934,
+      "learning_rate": 5.197752390191827e-06,
+      "loss": 1.6538,
+      "step": 808
+    },
+    {
+      "epoch": 1.6179999999999999,
+      "grad_norm": 7.07314920425415,
+      "learning_rate": 5.18612544082573e-06,
+      "loss": 1.5529,
+      "step": 809
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 9.155885696411133,
+      "learning_rate": 5.174497483512506e-06,
+      "loss": 1.7452,
+      "step": 810
+    },
+    {
+      "epoch": 1.6219999999999999,
+      "grad_norm": 5.848752498626709,
+      "learning_rate": 5.162868581222407e-06,
+      "loss": 1.6038,
+      "step": 811
+    },
+    {
+      "epoch": 1.624,
+      "grad_norm": 6.948676109313965,
+      "learning_rate": 5.151238796930804e-06,
+      "loss": 1.5384,
+      "step": 812
+    },
+    {
+      "epoch": 1.626,
+      "grad_norm": 7.449524402618408,
+      "learning_rate": 5.139608193617846e-06,
+      "loss": 1.7695,
+      "step": 813
+    },
+    {
+      "epoch": 1.6280000000000001,
+      "grad_norm": 9.431687355041504,
+      "learning_rate": 5.127976834268112e-06,
+      "loss": 1.6164,
+      "step": 814
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 11.5729341506958,
+      "learning_rate": 5.116344781870282e-06,
+      "loss": 1.5993,
+      "step": 815
+    },
+    {
+      "epoch": 1.6320000000000001,
+      "grad_norm": 10.78165054321289,
+      "learning_rate": 5.1047120994167855e-06,
+      "loss": 1.483,
+      "step": 816
+    },
+    {
+      "epoch": 1.634,
+      "grad_norm": 7.827059745788574,
+      "learning_rate": 5.093078849903464e-06,
+      "loss": 1.675,
+      "step": 817
+    },
+    {
+      "epoch": 1.6360000000000001,
+      "grad_norm": 8.432703971862793,
+      "learning_rate": 5.081445096329229e-06,
+      "loss": 1.5087,
+      "step": 818
+    },
+    {
+      "epoch": 1.638,
+      "grad_norm": 8.069623947143555,
+      "learning_rate": 5.069810901695727e-06,
+      "loss": 1.6049,
+      "step": 819
+    },
+    {
+      "epoch": 1.6400000000000001,
+      "grad_norm": 8.646689414978027,
+      "learning_rate": 5.0581763290069865e-06,
+      "loss": 1.5509,
+      "step": 820
+    },
+    {
+      "epoch": 1.642,
+      "grad_norm": 7.84726619720459,
+      "learning_rate": 5.046541441269085e-06,
+      "loss": 1.5971,
+      "step": 821
+    },
+    {
+      "epoch": 1.6440000000000001,
+      "grad_norm": 9.143599510192871,
+      "learning_rate": 5.034906301489808e-06,
+      "loss": 1.5909,
+      "step": 822
+    },
+    {
+      "epoch": 1.646,
+      "grad_norm": 7.073722839355469,
+      "learning_rate": 5.0232709726783065e-06,
+      "loss": 1.4608,
+      "step": 823
+    },
+    {
+      "epoch": 1.6480000000000001,
+      "grad_norm": 7.9401397705078125,
+      "learning_rate": 5.011635517844753e-06,
+      "loss": 1.6247,
+      "step": 824
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 7.164168834686279,
+      "learning_rate": 5e-06,
+      "loss": 1.5878,
+      "step": 825
+    },
+    {
+      "epoch": 1.6520000000000001,
+      "grad_norm": 8.36746883392334,
+      "learning_rate": 4.988364482155249e-06,
+      "loss": 1.5431,
+      "step": 826
+    },
+    {
+      "epoch": 1.654,
+      "grad_norm": 7.337077617645264,
+      "learning_rate": 4.976729027321694e-06,
+      "loss": 1.574,
+      "step": 827
+    },
+    {
+      "epoch": 1.6560000000000001,
+      "grad_norm": 8.89696979522705,
+      "learning_rate": 4.965093698510192e-06,
+      "loss": 1.6283,
+      "step": 828
+    },
+    {
+      "epoch": 1.658,
+      "grad_norm": 10.577298164367676,
+      "learning_rate": 4.953458558730917e-06,
+      "loss": 1.6708,
+      "step": 829
+    },
+    {
+      "epoch": 1.6600000000000001,
+      "grad_norm": 9.475666046142578,
+      "learning_rate": 4.941823670993016e-06,
+      "loss": 1.5338,
+      "step": 830
+    },
+    {
+      "epoch": 1.662,
+      "grad_norm": 9.556079864501953,
+      "learning_rate": 4.9301890983042744e-06,
+      "loss": 1.5048,
+      "step": 831
+    },
+    {
+      "epoch": 1.6640000000000001,
+      "grad_norm": 6.210651397705078,
+      "learning_rate": 4.9185549036707715e-06,
+      "loss": 1.5286,
+      "step": 832
+    },
+    {
+      "epoch": 1.666,
+      "grad_norm": 8.869709968566895,
+      "learning_rate": 4.906921150096538e-06,
+      "loss": 1.6674,
+      "step": 833
+    },
+    {
+      "epoch": 1.6680000000000001,
+      "grad_norm": 7.895263671875,
+      "learning_rate": 4.895287900583216e-06,
+      "loss": 1.5605,
+      "step": 834
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 12.988919258117676,
+      "learning_rate": 4.883655218129719e-06,
+      "loss": 1.7044,
+      "step": 835
+    },
+    {
+      "epoch": 1.6720000000000002,
+      "grad_norm": 7.571807861328125,
+      "learning_rate": 4.87202316573189e-06,
+      "loss": 1.5871,
+      "step": 836
+    },
+    {
+      "epoch": 1.674,
+      "grad_norm": 10.251978874206543,
+      "learning_rate": 4.860391806382157e-06,
+      "loss": 1.4593,
+      "step": 837
+    },
+    {
+      "epoch": 1.6760000000000002,
+      "grad_norm": 7.005547046661377,
+      "learning_rate": 4.8487612030691975e-06,
+      "loss": 1.5433,
+      "step": 838
+    },
+    {
+      "epoch": 1.678,
+      "grad_norm": 10.263413429260254,
+      "learning_rate": 4.837131418777595e-06,
+      "loss": 1.617,
+      "step": 839
+    },
+    {
+      "epoch": 1.6800000000000002,
+      "grad_norm": 6.690262794494629,
+      "learning_rate": 4.825502516487497e-06,
+      "loss": 1.6738,
+      "step": 840
+    },
+    {
+      "epoch": 1.682,
+      "grad_norm": 9.945344924926758,
+      "learning_rate": 4.813874559174271e-06,
+      "loss": 1.5897,
+      "step": 841
+    },
+    {
+      "epoch": 1.6840000000000002,
+      "grad_norm": 6.344468116760254,
+      "learning_rate": 4.802247609808175e-06,
+      "loss": 1.5544,
+      "step": 842
+    },
+    {
+      "epoch": 1.686,
+      "grad_norm": 6.898435592651367,
+      "learning_rate": 4.7906217313540035e-06,
+      "loss": 1.667,
+      "step": 843
+    },
+    {
+      "epoch": 1.688,
+      "grad_norm": 7.620174407958984,
+      "learning_rate": 4.778996986770747e-06,
+      "loss": 1.6153,
+      "step": 844
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 8.348420143127441,
+      "learning_rate": 4.767373439011267e-06,
+      "loss": 1.5119,
+      "step": 845
+    },
+    {
+      "epoch": 1.692,
+      "grad_norm": 6.85077428817749,
+      "learning_rate": 4.755751151021934e-06,
+      "loss": 1.6303,
+      "step": 846
+    },
+    {
+      "epoch": 1.694,
+      "grad_norm": 6.931562900543213,
+      "learning_rate": 4.744130185742301e-06,
+      "loss": 1.535,
+      "step": 847
+    },
+    {
+      "epoch": 1.696,
+      "grad_norm": 7.309017658233643,
+      "learning_rate": 4.732510606104754e-06,
+      "loss": 1.6259,
+      "step": 848
+    },
+    {
+      "epoch": 1.698,
+      "grad_norm": 8.722718238830566,
+      "learning_rate": 4.720892475034181e-06,
+      "loss": 1.5336,
+      "step": 849
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.0813627243042,
+      "learning_rate": 4.7092758554476215e-06,
+      "loss": 1.5323,
+      "step": 850
+    },
+    {
+      "epoch": 1.702,
+      "grad_norm": 8.41715145111084,
+      "learning_rate": 4.6976608102539285e-06,
+      "loss": 1.5765,
+      "step": 851
+    },
+    {
+      "epoch": 1.704,
+      "grad_norm": 9.829058647155762,
+      "learning_rate": 4.686047402353433e-06,
+      "loss": 1.4992,
+      "step": 852
+    },
+    {
+      "epoch": 1.706,
+      "grad_norm": 8.533679962158203,
+      "learning_rate": 4.674435694637597e-06,
+      "loss": 1.6102,
+      "step": 853
+    },
+    {
+      "epoch": 1.708,
+      "grad_norm": 8.357962608337402,
+      "learning_rate": 4.662825749988675e-06,
+      "loss": 1.5547,
+      "step": 854
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 6.73824405670166,
+      "learning_rate": 4.651217631279374e-06,
+      "loss": 1.5051,
+      "step": 855
+    },
+    {
+      "epoch": 1.712,
+      "grad_norm": 13.358663558959961,
+      "learning_rate": 4.639611401372514e-06,
+      "loss": 1.6093,
+      "step": 856
+    },
+    {
+      "epoch": 1.714,
+      "grad_norm": 9.384882926940918,
+      "learning_rate": 4.6280071231206845e-06,
+      "loss": 1.5455,
+      "step": 857
+    },
+    {
+      "epoch": 1.716,
+      "grad_norm": 10.82388687133789,
+      "learning_rate": 4.6164048593659076e-06,
+      "loss": 1.5374,
+      "step": 858
+    },
+    {
+      "epoch": 1.718,
+      "grad_norm": 6.8409743309021,
+      "learning_rate": 4.604804672939295e-06,
+      "loss": 1.6312,
+      "step": 859
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 7.881139755249023,
+      "learning_rate": 4.59320662666071e-06,
+      "loss": 1.5693,
+      "step": 860
+    },
+    {
+      "epoch": 1.722,
+      "grad_norm": 9.830620765686035,
+      "learning_rate": 4.581610783338424e-06,
+      "loss": 1.7035,
+      "step": 861
+    },
+    {
+      "epoch": 1.724,
+      "grad_norm": 7.074033737182617,
+      "learning_rate": 4.570017205768779e-06,
+      "loss": 1.5734,
+      "step": 862
+    },
+    {
+      "epoch": 1.726,
+      "grad_norm": 10.345975875854492,
+      "learning_rate": 4.5584259567358505e-06,
+      "loss": 1.474,
+      "step": 863
+    },
+    {
+      "epoch": 1.728,
+      "grad_norm": 8.930947303771973,
+      "learning_rate": 4.546837099011101e-06,
+      "loss": 1.5575,
+      "step": 864
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 8.436446189880371,
+      "learning_rate": 4.53525069535304e-06,
+      "loss": 1.6057,
+      "step": 865
+    },
+    {
+      "epoch": 1.732,
+      "grad_norm": 8.274133682250977,
+      "learning_rate": 4.523666808506893e-06,
+      "loss": 1.6484,
+      "step": 866
+    },
+    {
+      "epoch": 1.734,
+      "grad_norm": 8.548066139221191,
+      "learning_rate": 4.512085501204254e-06,
+      "loss": 1.5368,
+      "step": 867
+    },
+    {
+      "epoch": 1.736,
+      "grad_norm": 8.803610801696777,
+      "learning_rate": 4.500506836162746e-06,
+      "loss": 1.6525,
+      "step": 868
+    },
+    {
+      "epoch": 1.738,
+      "grad_norm": 8.110183715820312,
+      "learning_rate": 4.4889308760856826e-06,
+      "loss": 1.5937,
+      "step": 869
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 11.329172134399414,
+      "learning_rate": 4.477357683661734e-06,
+      "loss": 1.6,
+      "step": 870
+    },
+    {
+      "epoch": 1.742,
+      "grad_norm": 7.598846435546875,
+      "learning_rate": 4.465787321564576e-06,
+      "loss": 1.546,
+      "step": 871
+    },
+    {
+      "epoch": 1.744,
+      "grad_norm": 6.989120960235596,
+      "learning_rate": 4.45421985245256e-06,
+      "loss": 1.6186,
+      "step": 872
+    },
+    {
+      "epoch": 1.746,
+      "grad_norm": 9.312042236328125,
+      "learning_rate": 4.442655338968373e-06,
+      "loss": 1.6592,
+      "step": 873
+    },
+    {
+      "epoch": 1.748,
+      "grad_norm": 5.902655124664307,
+      "learning_rate": 4.431093843738693e-06,
+      "loss": 1.5185,
+      "step": 874
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 6.853744029998779,
+      "learning_rate": 4.4195354293738484e-06,
+      "loss": 1.5383,
+      "step": 875
+    },
+    {
+      "epoch": 1.752,
+      "grad_norm": 8.076836585998535,
+      "learning_rate": 4.4079801584674955e-06,
+      "loss": 1.4652,
+      "step": 876
+    },
+    {
+      "epoch": 1.754,
+      "grad_norm": 6.988975524902344,
+      "learning_rate": 4.396428093596258e-06,
+      "loss": 1.5636,
+      "step": 877
+    },
+    {
+      "epoch": 1.756,
+      "grad_norm": 7.838133335113525,
+      "learning_rate": 4.384879297319398e-06,
+      "loss": 1.5891,
+      "step": 878
+    },
+    {
+      "epoch": 1.758,
+      "grad_norm": 6.956199645996094,
+      "learning_rate": 4.373333832178478e-06,
+      "loss": 1.5573,
+      "step": 879
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 7.702134132385254,
+      "learning_rate": 4.361791760697027e-06,
+      "loss": 1.6047,
+      "step": 880
+    },
+    {
+      "epoch": 1.762,
+      "grad_norm": 8.466774940490723,
+      "learning_rate": 4.3502531453801885e-06,
+      "loss": 1.6113,
+      "step": 881
+    },
+    {
+      "epoch": 1.764,
+      "grad_norm": 7.680281639099121,
+      "learning_rate": 4.3387180487143875e-06,
+      "loss": 1.5202,
+      "step": 882
+    },
+    {
+      "epoch": 1.766,
+      "grad_norm": 8.012104988098145,
+      "learning_rate": 4.3271865331670036e-06,
+      "loss": 1.6169,
+      "step": 883
+    },
+    {
+      "epoch": 1.768,
+      "grad_norm": 9.084745407104492,
+      "learning_rate": 4.315658661186016e-06,
+      "loss": 1.4818,
+      "step": 884
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 8.11100959777832,
+      "learning_rate": 4.304134495199675e-06,
+      "loss": 1.6068,
+      "step": 885
+    },
+    {
+      "epoch": 1.772,
+      "grad_norm": 7.566526412963867,
+      "learning_rate": 4.2926140976161555e-06,
+      "loss": 1.5804,
+      "step": 886
+    },
+    {
+      "epoch": 1.774,
+      "grad_norm": 6.527312755584717,
+      "learning_rate": 4.281097530823237e-06,
+      "loss": 1.5348,
+      "step": 887
+    },
+    {
+      "epoch": 1.776,
+      "grad_norm": 8.60621166229248,
+      "learning_rate": 4.269584857187942e-06,
+      "loss": 1.6018,
+      "step": 888
+    },
+    {
+      "epoch": 1.778,
+      "grad_norm": 6.590750217437744,
+      "learning_rate": 4.258076139056217e-06,
+      "loss": 1.5499,
+      "step": 889
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 8.905508995056152,
+      "learning_rate": 4.246571438752585e-06,
+      "loss": 1.7064,
+      "step": 890
+    },
+    {
+      "epoch": 1.782,
+      "grad_norm": 8.529908180236816,
+      "learning_rate": 4.23507081857981e-06,
+      "loss": 1.534,
+      "step": 891
+    },
+    {
+      "epoch": 1.784,
+      "grad_norm": 9.130575180053711,
+      "learning_rate": 4.2235743408185635e-06,
+      "loss": 1.6427,
+      "step": 892
+    },
+    {
+      "epoch": 1.786,
+      "grad_norm": 8.667200088500977,
+      "learning_rate": 4.212082067727079e-06,
+      "loss": 1.4933,
+      "step": 893
+    },
+    {
+      "epoch": 1.788,
+      "grad_norm": 8.957341194152832,
+      "learning_rate": 4.200594061540827e-06,
+      "loss": 1.591,
+      "step": 894
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 7.048657417297363,
+      "learning_rate": 4.189110384472164e-06,
+      "loss": 1.5583,
+      "step": 895
+    },
+    {
+      "epoch": 1.792,
+      "grad_norm": 8.028057098388672,
+      "learning_rate": 4.1776310987100054e-06,
+      "loss": 1.4866,
+      "step": 896
+    },
+    {
+      "epoch": 1.794,
+      "grad_norm": 8.738648414611816,
+      "learning_rate": 4.166156266419489e-06,
+      "loss": 1.6125,
+      "step": 897
+    },
+    {
+      "epoch": 1.796,
+      "grad_norm": 8.001562118530273,
+      "learning_rate": 4.154685949741631e-06,
+      "loss": 1.4699,
+      "step": 898
+    },
+    {
+      "epoch": 1.798,
+      "grad_norm": 8.324260711669922,
+      "learning_rate": 4.143220210792993e-06,
+      "loss": 1.5949,
+      "step": 899
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 11.599272727966309,
+      "learning_rate": 4.131759111665349e-06,
+      "loss": 1.6686,
+      "step": 900
+    },
+    {
+      "epoch": 1.802,
+      "grad_norm": 7.560389518737793,
+      "learning_rate": 4.1203027144253466e-06,
+      "loss": 1.5041,
+      "step": 901
+    },
+    {
+      "epoch": 1.804,
+      "grad_norm": 7.603427886962891,
+      "learning_rate": 4.108851081114169e-06,
+      "loss": 1.4909,
+      "step": 902
+    },
+    {
+      "epoch": 1.806,
+      "grad_norm": 6.763309955596924,
+      "learning_rate": 4.0974042737472005e-06,
+      "loss": 1.5467,
+      "step": 903
+    },
+    {
+      "epoch": 1.808,
+      "grad_norm": 9.465215682983398,
+      "learning_rate": 4.0859623543136935e-06,
+      "loss": 1.5266,
+      "step": 904
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 7.232966423034668,
+      "learning_rate": 4.074525384776428e-06,
+      "loss": 1.5645,
+      "step": 905
+    },
+    {
+      "epoch": 1.812,
+      "grad_norm": 7.509275436401367,
+      "learning_rate": 4.063093427071376e-06,
+      "loss": 1.5804,
+      "step": 906
+    },
+    {
+      "epoch": 1.814,
+      "grad_norm": 8.851505279541016,
+      "learning_rate": 4.051666543107377e-06,
+      "loss": 1.5903,
+      "step": 907
+    },
+    {
+      "epoch": 1.8159999999999998,
+      "grad_norm": 7.193111419677734,
+      "learning_rate": 4.040244794765783e-06,
+      "loss": 1.6361,
+      "step": 908
+    },
+    {
+      "epoch": 1.818,
+      "grad_norm": 13.53274154663086,
+      "learning_rate": 4.028828243900141e-06,
+      "loss": 1.5629,
+      "step": 909
+    },
+    {
+      "epoch": 1.8199999999999998,
+      "grad_norm": 8.000975608825684,
+      "learning_rate": 4.017416952335849e-06,
+      "loss": 1.55,
+      "step": 910
+    },
+    {
+      "epoch": 1.822,
+      "grad_norm": 9.145268440246582,
+      "learning_rate": 4.006010981869829e-06,
+      "loss": 1.5949,
+      "step": 911
+    },
+    {
+      "epoch": 1.8239999999999998,
+      "grad_norm": 9.474929809570312,
+      "learning_rate": 3.994610394270178e-06,
+      "loss": 1.5537,
+      "step": 912
+    },
+    {
+      "epoch": 1.826,
+      "grad_norm": 8.653277397155762,
+      "learning_rate": 3.983215251275847e-06,
+      "loss": 1.6259,
+      "step": 913
+    },
+    {
+      "epoch": 1.8279999999999998,
+      "grad_norm": 7.7228617668151855,
+      "learning_rate": 3.971825614596308e-06,
+      "loss": 1.5946,
+      "step": 914
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.129724502563477,
+      "learning_rate": 3.960441545911205e-06,
+      "loss": 1.5925,
+      "step": 915
+    },
+    {
+      "epoch": 1.8319999999999999,
+      "grad_norm": 9.415990829467773,
+      "learning_rate": 3.949063106870031e-06,
+      "loss": 1.5227,
+      "step": 916
+    },
+    {
+      "epoch": 1.834,
+      "grad_norm": 9.796669006347656,
+      "learning_rate": 3.9376903590917945e-06,
+      "loss": 1.5435,
+      "step": 917
+    },
+    {
+      "epoch": 1.8359999999999999,
+      "grad_norm": 9.218023300170898,
+      "learning_rate": 3.926323364164684e-06,
+      "loss": 1.4847,
+      "step": 918
+    },
+    {
+      "epoch": 1.838,
+      "grad_norm": 6.569913387298584,
+      "learning_rate": 3.914962183645733e-06,
+      "loss": 1.598,
+      "step": 919
+    },
+    {
+      "epoch": 1.8399999999999999,
+      "grad_norm": 8.879981994628906,
+      "learning_rate": 3.903606879060483e-06,
+      "loss": 1.625,
+      "step": 920
+    },
+    {
+      "epoch": 1.842,
+      "grad_norm": 6.769296169281006,
+      "learning_rate": 3.892257511902664e-06,
+      "loss": 1.5214,
+      "step": 921
+    },
+    {
+      "epoch": 1.8439999999999999,
+      "grad_norm": 7.948380470275879,
+      "learning_rate": 3.880914143633844e-06,
+      "loss": 1.5543,
+      "step": 922
+    },
+    {
+      "epoch": 1.846,
+      "grad_norm": 6.776271820068359,
+      "learning_rate": 3.869576835683109e-06,
+      "loss": 1.5748,
+      "step": 923
+    },
+    {
+      "epoch": 1.8479999999999999,
+      "grad_norm": 8.43204402923584,
+      "learning_rate": 3.8582456494467214e-06,
+      "loss": 1.5795,
+      "step": 924
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 9.641144752502441,
+      "learning_rate": 3.8469206462878e-06,
+      "loss": 1.5457,
+      "step": 925
+    },
+    {
+      "epoch": 1.8519999999999999,
+      "grad_norm": 9.725312232971191,
+      "learning_rate": 3.835601887535971e-06,
+      "loss": 1.5125,
+      "step": 926
+    },
+    {
+      "epoch": 1.854,
+      "grad_norm": 10.626245498657227,
+      "learning_rate": 3.82428943448705e-06,
+      "loss": 1.6646,
+      "step": 927
+    },
+    {
+      "epoch": 1.8559999999999999,
+      "grad_norm": 8.524333000183105,
+      "learning_rate": 3.812983348402703e-06,
+      "loss": 1.6704,
+      "step": 928
+    },
+    {
+      "epoch": 1.858,
+      "grad_norm": 8.207454681396484,
+      "learning_rate": 3.8016836905101157e-06,
+      "loss": 1.6347,
+      "step": 929
+    },
+    {
+      "epoch": 1.8599999999999999,
+      "grad_norm": 9.849774360656738,
+      "learning_rate": 3.790390522001662e-06,
+      "loss": 1.5782,
+      "step": 930
+    },
+    {
+      "epoch": 1.862,
+      "grad_norm": 8.06809139251709,
+      "learning_rate": 3.7791039040345743e-06,
+      "loss": 1.6321,
+      "step": 931
+    },
+    {
+      "epoch": 1.8639999999999999,
+      "grad_norm": 8.71696949005127,
+      "learning_rate": 3.767823897730612e-06,
+      "loss": 1.66,
+      "step": 932
+    },
+    {
+      "epoch": 1.866,
+      "grad_norm": 7.418612480163574,
+      "learning_rate": 3.756550564175727e-06,
+      "loss": 1.5837,
+      "step": 933
+    },
+    {
+      "epoch": 1.8679999999999999,
+      "grad_norm": 7.238306999206543,
+      "learning_rate": 3.745283964419736e-06,
+      "loss": 1.5742,
+      "step": 934
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 7.4198899269104,
+      "learning_rate": 3.7340241594759917e-06,
+      "loss": 1.7463,
+      "step": 935
+    },
+    {
+      "epoch": 1.8719999999999999,
+      "grad_norm": 6.619482517242432,
+      "learning_rate": 3.7227712103210485e-06,
+      "loss": 1.738,
+      "step": 936
+    },
+    {
+      "epoch": 1.874,
+      "grad_norm": 7.6474223136901855,
+      "learning_rate": 3.7115251778943314e-06,
+      "loss": 1.5902,
+      "step": 937
+    },
+    {
+      "epoch": 1.876,
+      "grad_norm": 9.21036434173584,
+      "learning_rate": 3.700286123097814e-06,
+      "loss": 1.5674,
+      "step": 938
+    },
+    {
+      "epoch": 1.8780000000000001,
+      "grad_norm": 7.860700607299805,
+      "learning_rate": 3.6890541067956775e-06,
+      "loss": 1.6321,
+      "step": 939
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 7.944121837615967,
+      "learning_rate": 3.6778291898139907e-06,
+      "loss": 1.6441,
+      "step": 940
+    },
+    {
+      "epoch": 1.8820000000000001,
+      "grad_norm": 9.204889297485352,
+      "learning_rate": 3.6666114329403723e-06,
+      "loss": 1.5388,
+      "step": 941
+    },
+    {
+      "epoch": 1.884,
+      "grad_norm": 6.569108009338379,
+      "learning_rate": 3.655400896923672e-06,
+      "loss": 1.5072,
+      "step": 942
+    },
+    {
+      "epoch": 1.8860000000000001,
+      "grad_norm": 6.905144691467285,
+      "learning_rate": 3.6441976424736315e-06,
+      "loss": 1.591,
+      "step": 943
+    },
+    {
+      "epoch": 1.888,
+      "grad_norm": 9.775949478149414,
+      "learning_rate": 3.633001730260558e-06,
+      "loss": 1.5154,
+      "step": 944
+    },
+    {
+      "epoch": 1.8900000000000001,
+      "grad_norm": 8.357000350952148,
+      "learning_rate": 3.6218132209150047e-06,
+      "loss": 1.7292,
+      "step": 945
+    },
+    {
+      "epoch": 1.892,
+      "grad_norm": 10.853804588317871,
+      "learning_rate": 3.6106321750274275e-06,
+      "loss": 1.6071,
+      "step": 946
+    },
+    {
+      "epoch": 1.8940000000000001,
+      "grad_norm": 7.702592849731445,
+      "learning_rate": 3.5994586531478672e-06,
+      "loss": 1.5247,
+      "step": 947
+    },
+    {
+      "epoch": 1.896,
+      "grad_norm": 8.18514633178711,
+      "learning_rate": 3.5882927157856175e-06,
+      "loss": 1.5674,
+      "step": 948
+    },
+    {
+      "epoch": 1.8980000000000001,
+      "grad_norm": 6.387351989746094,
+      "learning_rate": 3.577134423408906e-06,
+      "loss": 1.5946,
+      "step": 949
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 7.652561187744141,
+      "learning_rate": 3.5659838364445505e-06,
+      "loss": 1.5211,
+      "step": 950
+    },
+    {
+      "epoch": 1.9020000000000001,
+      "grad_norm": 6.697483062744141,
+      "learning_rate": 3.5548410152776414e-06,
+      "loss": 1.6415,
+      "step": 951
+    },
+    {
+      "epoch": 1.904,
+      "grad_norm": 7.114959716796875,
+      "learning_rate": 3.543706020251223e-06,
+      "loss": 1.5515,
+      "step": 952
+    },
+    {
+      "epoch": 1.9060000000000001,
+      "grad_norm": 7.068447113037109,
+      "learning_rate": 3.5325789116659493e-06,
+      "loss": 1.6781,
+      "step": 953
+    },
+    {
+      "epoch": 1.908,
+      "grad_norm": 6.634418487548828,
+      "learning_rate": 3.521459749779769e-06,
+      "loss": 1.5848,
+      "step": 954
+    },
+    {
+      "epoch": 1.9100000000000001,
+      "grad_norm": 7.493570327758789,
+      "learning_rate": 3.51034859480759e-06,
+      "loss": 1.6289,
+      "step": 955
+    },
+    {
+      "epoch": 1.912,
+      "grad_norm": 9.925775527954102,
+      "learning_rate": 3.4992455069209717e-06,
+      "loss": 1.5641,
+      "step": 956
+    },
+    {
+      "epoch": 1.9140000000000001,
+      "grad_norm": 10.061575889587402,
+      "learning_rate": 3.488150546247778e-06,
+      "loss": 1.5989,
+      "step": 957
+    },
+    {
+      "epoch": 1.916,
+      "grad_norm": 8.778186798095703,
+      "learning_rate": 3.4770637728718608e-06,
+      "loss": 1.7018,
+      "step": 958
+    },
+    {
+      "epoch": 1.9180000000000001,
+      "grad_norm": 8.304412841796875,
+      "learning_rate": 3.465985246832739e-06,
+      "loss": 1.6029,
+      "step": 959
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 7.178003311157227,
+      "learning_rate": 3.4549150281252635e-06,
+      "loss": 1.5976,
+      "step": 960
+    },
+    {
+      "epoch": 1.9220000000000002,
+      "grad_norm": 9.42418384552002,
+      "learning_rate": 3.4438531766993012e-06,
+      "loss": 1.5237,
+      "step": 961
+    },
+    {
+      "epoch": 1.924,
+      "grad_norm": 8.352529525756836,
+      "learning_rate": 3.4327997524594026e-06,
+      "loss": 1.6348,
+      "step": 962
+    },
+    {
+      "epoch": 1.9260000000000002,
+      "grad_norm": 10.59170913696289,
+      "learning_rate": 3.4217548152644887e-06,
+      "loss": 1.5547,
+      "step": 963
+    },
+    {
+      "epoch": 1.928,
+      "grad_norm": 7.5832343101501465,
+      "learning_rate": 3.4107184249275114e-06,
+      "loss": 1.5294,
+      "step": 964
+    },
+    {
+      "epoch": 1.9300000000000002,
+      "grad_norm": 9.453595161437988,
+      "learning_rate": 3.399690641215142e-06,
+      "loss": 1.6929,
+      "step": 965
+    },
+    {
+      "epoch": 1.932,
+      "grad_norm": 8.6815185546875,
+      "learning_rate": 3.3886715238474454e-06,
+      "loss": 1.6084,
+      "step": 966
+    },
+    {
+      "epoch": 1.9340000000000002,
+      "grad_norm": 6.6434149742126465,
+      "learning_rate": 3.3776611324975496e-06,
+      "loss": 1.4945,
+      "step": 967
+    },
+    {
+      "epoch": 1.936,
+      "grad_norm": 9.94503402709961,
+      "learning_rate": 3.3666595267913293e-06,
+      "loss": 1.587,
+      "step": 968
+    },
+    {
+      "epoch": 1.938,
+      "grad_norm": 7.814234256744385,
+      "learning_rate": 3.355666766307084e-06,
+      "loss": 1.6181,
+      "step": 969
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.182234764099121,
+      "learning_rate": 3.3446829105752103e-06,
+      "loss": 1.5493,
+      "step": 970
+    },
+    {
+      "epoch": 1.942,
+      "grad_norm": 7.502097129821777,
+      "learning_rate": 3.3337080190778816e-06,
+      "loss": 1.4921,
+      "step": 971
+    },
+    {
+      "epoch": 1.944,
+      "grad_norm": 6.4139604568481445,
+      "learning_rate": 3.322742151248726e-06,
+      "loss": 1.5533,
+      "step": 972
+    },
+    {
+      "epoch": 1.946,
+      "grad_norm": 7.537562370300293,
+      "learning_rate": 3.311785366472506e-06,
+      "loss": 1.5851,
+      "step": 973
+    },
+    {
+      "epoch": 1.948,
+      "grad_norm": 8.741056442260742,
+      "learning_rate": 3.3008377240847955e-06,
+      "loss": 1.5118,
+      "step": 974
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 7.846621513366699,
+      "learning_rate": 3.289899283371657e-06,
+      "loss": 1.5772,
+      "step": 975
+    },
+    {
+      "epoch": 1.952,
+      "grad_norm": 9.379910469055176,
+      "learning_rate": 3.2789701035693242e-06,
+      "loss": 1.5339,
+      "step": 976
+    },
+    {
+      "epoch": 1.954,
+      "grad_norm": 9.359630584716797,
+      "learning_rate": 3.268050243863877e-06,
+      "loss": 1.6382,
+      "step": 977
+    },
+    {
+      "epoch": 1.956,
+      "grad_norm": 11.636152267456055,
+      "learning_rate": 3.2571397633909252e-06,
+      "loss": 1.5506,
+      "step": 978
+    },
+    {
+      "epoch": 1.958,
+      "grad_norm": 8.888190269470215,
+      "learning_rate": 3.246238721235283e-06,
+      "loss": 1.4885,
+      "step": 979
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.042271614074707,
+      "learning_rate": 3.2353471764306567e-06,
+      "loss": 1.5741,
+      "step": 980
+    },
+    {
+      "epoch": 1.962,
+      "grad_norm": 7.534533500671387,
+      "learning_rate": 3.224465187959316e-06,
+      "loss": 1.5664,
+      "step": 981
+    },
+    {
+      "epoch": 1.964,
+      "grad_norm": 6.7709479331970215,
+      "learning_rate": 3.2135928147517803e-06,
+      "loss": 1.5943,
+      "step": 982
+    },
+    {
+      "epoch": 1.966,
+      "grad_norm": 8.381340026855469,
+      "learning_rate": 3.2027301156865015e-06,
+      "loss": 1.6603,
+      "step": 983
+    },
+    {
+      "epoch": 1.968,
+      "grad_norm": 9.999662399291992,
+      "learning_rate": 3.1918771495895395e-06,
+      "loss": 1.7147,
+      "step": 984
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 11.495597839355469,
+      "learning_rate": 3.1810339752342446e-06,
+      "loss": 1.6706,
+      "step": 985
+    },
+    {
+      "epoch": 1.972,
+      "grad_norm": 9.23818588256836,
+      "learning_rate": 3.1702006513409393e-06,
+      "loss": 1.6012,
+      "step": 986
+    },
+    {
+      "epoch": 1.974,
+      "grad_norm": 10.354619026184082,
+      "learning_rate": 3.1593772365766107e-06,
+      "loss": 1.6126,
+      "step": 987
+    },
+    {
+      "epoch": 1.976,
+      "grad_norm": 9.37294864654541,
+      "learning_rate": 3.148563789554575e-06,
+      "loss": 1.6156,
+      "step": 988
+    },
+    {
+      "epoch": 1.978,
+      "grad_norm": 10.598193168640137,
+      "learning_rate": 3.137760368834169e-06,
+      "loss": 1.7388,
+      "step": 989
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 9.898024559020996,
+      "learning_rate": 3.12696703292044e-06,
+      "loss": 1.6421,
+      "step": 990
+    },
+    {
+      "epoch": 1.982,
+      "grad_norm": 9.265511512756348,
+      "learning_rate": 3.1161838402638158e-06,
+      "loss": 1.5518,
+      "step": 991
+    },
+    {
+      "epoch": 1.984,
+      "grad_norm": 7.578424453735352,
+      "learning_rate": 3.105410849259796e-06,
+      "loss": 1.5866,
+      "step": 992
+    },
+    {
+      "epoch": 1.986,
+      "grad_norm": 10.20102310180664,
+      "learning_rate": 3.09464811824863e-06,
+      "loss": 1.6485,
+      "step": 993
+    },
+    {
+      "epoch": 1.988,
+      "grad_norm": 7.367624759674072,
+      "learning_rate": 3.0838957055150136e-06,
+      "loss": 1.6409,
+      "step": 994
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 7.860062599182129,
+      "learning_rate": 3.0731536692877596e-06,
+      "loss": 1.535,
+      "step": 995
+    },
+    {
+      "epoch": 1.992,
+      "grad_norm": 10.079544067382812,
+      "learning_rate": 3.0624220677394854e-06,
+      "loss": 1.5819,
+      "step": 996
+    },
+    {
+      "epoch": 1.994,
+      "grad_norm": 6.722997665405273,
+      "learning_rate": 3.0517009589863057e-06,
+      "loss": 1.6481,
+      "step": 997
+    },
+    {
+      "epoch": 1.996,
+      "grad_norm": 11.341468811035156,
+      "learning_rate": 3.040990401087508e-06,
+      "loss": 1.695,
+      "step": 998
+    },
+    {
+      "epoch": 1.998,
+      "grad_norm": 6.343021392822266,
+      "learning_rate": 3.030290452045245e-06,
+      "loss": 1.6707,
+      "step": 999
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 11.605053901672363,
+      "learning_rate": 3.019601169804216e-06,
+      "loss": 1.5376,
+      "step": 1000
+    },
+    {
+      "epoch": 2.002,
+      "grad_norm": 9.129085540771484,
+      "learning_rate": 3.0089226122513583e-06,
+      "loss": 1.3932,
+      "step": 1001
+    },
+    {
+      "epoch": 2.004,
+      "grad_norm": 7.879745006561279,
+      "learning_rate": 2.9982548372155264e-06,
+      "loss": 1.4429,
+      "step": 1002
+    },
+    {
+      "epoch": 2.006,
+      "grad_norm": 7.70796537399292,
+      "learning_rate": 2.9875979024671846e-06,
+      "loss": 1.4238,
+      "step": 1003
+    },
+    {
+      "epoch": 2.008,
+      "grad_norm": 8.425646781921387,
+      "learning_rate": 2.9769518657180953e-06,
+      "loss": 1.3642,
+      "step": 1004
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 7.608778953552246,
+      "learning_rate": 2.966316784621e-06,
+      "loss": 1.3483,
+      "step": 1005
+    },
+    {
+      "epoch": 2.012,
+      "grad_norm": 9.824869155883789,
+      "learning_rate": 2.9556927167693107e-06,
+      "loss": 1.4165,
+      "step": 1006
+    },
+    {
+      "epoch": 2.014,
+      "grad_norm": 7.270105838775635,
+      "learning_rate": 2.945079719696802e-06,
+      "loss": 1.3235,
+      "step": 1007
+    },
+    {
+      "epoch": 2.016,
+      "grad_norm": 10.12110424041748,
+      "learning_rate": 2.934477850877292e-06,
+      "loss": 1.3102,
+      "step": 1008
+    },
+    {
+      "epoch": 2.018,
+      "grad_norm": 8.959726333618164,
+      "learning_rate": 2.9238871677243354e-06,
+      "loss": 1.2703,
+      "step": 1009
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 8.507242202758789,
+      "learning_rate": 2.9133077275909112e-06,
+      "loss": 1.1562,
+      "step": 1010
+    },
+    {
+      "epoch": 2.022,
+      "grad_norm": 11.486827850341797,
+      "learning_rate": 2.9027395877691143e-06,
+      "loss": 1.2312,
+      "step": 1011
+    },
+    {
+      "epoch": 2.024,
+      "grad_norm": 9.935345649719238,
+      "learning_rate": 2.892182805489846e-06,
+      "loss": 1.1142,
+      "step": 1012
+    },
+    {
+      "epoch": 2.026,
+      "grad_norm": 9.927567481994629,
+      "learning_rate": 2.8816374379224932e-06,
+      "loss": 1.0935,
+      "step": 1013
+    },
+    {
+      "epoch": 2.028,
+      "grad_norm": 11.30349349975586,
+      "learning_rate": 2.871103542174637e-06,
+      "loss": 1.0115,
+      "step": 1014
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 11.31701946258545,
+      "learning_rate": 2.86058117529173e-06,
+      "loss": 1.0013,
+      "step": 1015
+    },
+    {
+      "epoch": 2.032,
+      "grad_norm": 11.658658027648926,
+      "learning_rate": 2.8500703942567874e-06,
+      "loss": 1.0056,
+      "step": 1016
+    },
+    {
+      "epoch": 2.034,
+      "grad_norm": 15.923133850097656,
+      "learning_rate": 2.839571255990088e-06,
+      "loss": 1.1577,
+      "step": 1017
+    },
+    {
+      "epoch": 2.036,
+      "grad_norm": 16.112945556640625,
+      "learning_rate": 2.82908381734886e-06,
+      "loss": 1.0646,
+      "step": 1018
+    },
+    {
+      "epoch": 2.038,
+      "grad_norm": 16.52297019958496,
+      "learning_rate": 2.818608135126967e-06,
+      "loss": 1.1384,
+      "step": 1019
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 17.75345802307129,
+      "learning_rate": 2.8081442660546126e-06,
+      "loss": 1.0859,
+      "step": 1020
+    },
+    {
+      "epoch": 2.042,
+      "grad_norm": 18.585851669311523,
+      "learning_rate": 2.797692266798027e-06,
+      "loss": 1.0863,
+      "step": 1021
+    },
+    {
+      "epoch": 2.044,
+      "grad_norm": 17.183208465576172,
+      "learning_rate": 2.7872521939591556e-06,
+      "loss": 1.0416,
+      "step": 1022
+    },
+    {
+      "epoch": 2.046,
+      "grad_norm": 18.7218074798584,
+      "learning_rate": 2.776824104075364e-06,
+      "loss": 0.8856,
+      "step": 1023
+    },
+    {
+      "epoch": 2.048,
+      "grad_norm": 20.450927734375,
+      "learning_rate": 2.7664080536191178e-06,
+      "loss": 1.1884,
+      "step": 1024
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 17.043212890625,
+      "learning_rate": 2.7560040989976894e-06,
+      "loss": 1.0495,
+      "step": 1025
+    },
+    {
+      "epoch": 2.052,
+      "grad_norm": 15.958205223083496,
+      "learning_rate": 2.7456122965528475e-06,
+      "loss": 0.9814,
+      "step": 1026
+    },
+    {
+      "epoch": 2.054,
+      "grad_norm": 15.604181289672852,
+      "learning_rate": 2.7352327025605464e-06,
+      "loss": 0.9974,
+      "step": 1027
+    },
+    {
+      "epoch": 2.056,
+      "grad_norm": 14.632473945617676,
+      "learning_rate": 2.724865373230632e-06,
+      "loss": 1.002,
+      "step": 1028
+    },
+    {
+      "epoch": 2.058,
+      "grad_norm": 14.43281364440918,
+      "learning_rate": 2.714510364706531e-06,
+      "loss": 0.9437,
+      "step": 1029
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 17.825098037719727,
+      "learning_rate": 2.7041677330649408e-06,
+      "loss": 1.0608,
+      "step": 1030
+    },
+    {
+      "epoch": 2.062,
+      "grad_norm": 15.265949249267578,
+      "learning_rate": 2.6938375343155464e-06,
+      "loss": 1.1083,
+      "step": 1031
+    },
+    {
+      "epoch": 2.064,
+      "grad_norm": 13.251029014587402,
+      "learning_rate": 2.683519824400693e-06,
+      "loss": 1.0742,
+      "step": 1032
+    },
+    {
+      "epoch": 2.066,
+      "grad_norm": 13.636655807495117,
+      "learning_rate": 2.6732146591950924e-06,
+      "loss": 1.0812,
+      "step": 1033
+    },
+    {
+      "epoch": 2.068,
+      "grad_norm": 14.354068756103516,
+      "learning_rate": 2.662922094505529e-06,
+      "loss": 1.1726,
+      "step": 1034
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 13.713906288146973,
+      "learning_rate": 2.6526421860705474e-06,
+      "loss": 1.1794,
+      "step": 1035
+    },
+    {
+      "epoch": 2.072,
+      "grad_norm": 15.267590522766113,
+      "learning_rate": 2.6423749895601494e-06,
+      "loss": 1.1732,
+      "step": 1036
+    },
+    {
+      "epoch": 2.074,
+      "grad_norm": 13.415190696716309,
+      "learning_rate": 2.6321205605755002e-06,
+      "loss": 1.1165,
+      "step": 1037
+    },
+    {
+      "epoch": 2.076,
+      "grad_norm": 12.924043655395508,
+      "learning_rate": 2.6218789546486235e-06,
+      "loss": 1.1036,
+      "step": 1038
+    },
+    {
+      "epoch": 2.078,
+      "grad_norm": 11.99133014678955,
+      "learning_rate": 2.611650227242102e-06,
+      "loss": 1.0163,
+      "step": 1039
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 11.737472534179688,
+      "learning_rate": 2.601434433748771e-06,
+      "loss": 0.9887,
+      "step": 1040
+    },
+    {
+      "epoch": 2.082,
+      "grad_norm": 14.03266429901123,
+      "learning_rate": 2.5912316294914232e-06,
+      "loss": 1.0226,
+      "step": 1041
+    },
+    {
+      "epoch": 2.084,
+      "grad_norm": 13.012822151184082,
+      "learning_rate": 2.581041869722519e-06,
+      "loss": 1.1165,
+      "step": 1042
+    },
+    {
+      "epoch": 2.086,
+      "grad_norm": 13.519615173339844,
+      "learning_rate": 2.5708652096238674e-06,
+      "loss": 1.1066,
+      "step": 1043
+    },
+    {
+      "epoch": 2.088,
+      "grad_norm": 11.919333457946777,
+      "learning_rate": 2.560701704306336e-06,
+      "loss": 0.9055,
+      "step": 1044
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 11.374495506286621,
+      "learning_rate": 2.550551408809566e-06,
+      "loss": 1.024,
+      "step": 1045
+    },
+    {
+      "epoch": 2.092,
+      "grad_norm": 12.566267013549805,
+      "learning_rate": 2.540414378101647e-06,
+      "loss": 1.1746,
+      "step": 1046
+    },
+    {
+      "epoch": 2.094,
+      "grad_norm": 13.454764366149902,
+      "learning_rate": 2.5302906670788463e-06,
+      "loss": 0.9756,
+      "step": 1047
+    },
+    {
+      "epoch": 2.096,
+      "grad_norm": 12.35405445098877,
+      "learning_rate": 2.52018033056529e-06,
+      "loss": 1.0644,
+      "step": 1048
+    },
+    {
+      "epoch": 2.098,
+      "grad_norm": 12.588571548461914,
+      "learning_rate": 2.5100834233126827e-06,
+      "loss": 1.1582,
+      "step": 1049
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 13.59836196899414,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 1.0943,
+      "step": 1050
+    },
+    {
+      "epoch": 2.102,
+      "grad_norm": 12.1016206741333,
+      "learning_rate": 2.489930115233199e-06,
+      "loss": 1.0855,
+      "step": 1051
+    },
+    {
+      "epoch": 2.104,
+      "grad_norm": 13.028971672058105,
+      "learning_rate": 2.4798738235449164e-06,
+      "loss": 1.1361,
+      "step": 1052
+    },
+    {
+      "epoch": 2.106,
+      "grad_norm": 13.81318187713623,
+      "learning_rate": 2.469831179394182e-06,
+      "loss": 1.0715,
+      "step": 1053
+    },
+    {
+      "epoch": 2.108,
+      "grad_norm": 12.744100570678711,
+      "learning_rate": 2.4598022371661113e-06,
+      "loss": 1.1322,
+      "step": 1054
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 13.312298774719238,
+      "learning_rate": 2.4497870511716237e-06,
+      "loss": 1.0601,
+      "step": 1055
+    },
+    {
+      "epoch": 2.112,
+      "grad_norm": 12.346517562866211,
+      "learning_rate": 2.4397856756471435e-06,
+      "loss": 1.0386,
+      "step": 1056
+    },
+    {
+      "epoch": 2.114,
+      "grad_norm": 11.398303985595703,
+      "learning_rate": 2.429798164754299e-06,
+      "loss": 1.1074,
+      "step": 1057
+    },
+    {
+      "epoch": 2.116,
+      "grad_norm": 14.021239280700684,
+      "learning_rate": 2.4198245725796427e-06,
+      "loss": 1.1536,
+      "step": 1058
+    },
+    {
+      "epoch": 2.118,
+      "grad_norm": 13.355626106262207,
+      "learning_rate": 2.40986495313435e-06,
+      "loss": 1.1139,
+      "step": 1059
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 11.496376037597656,
+      "learning_rate": 2.3999193603539234e-06,
+      "loss": 1.1207,
+      "step": 1060
+    },
+    {
+      "epoch": 2.122,
+      "grad_norm": 12.229351043701172,
+      "learning_rate": 2.3899878480979098e-06,
+      "loss": 1.0847,
+      "step": 1061
+    },
+    {
+      "epoch": 2.124,
+      "grad_norm": 12.624899864196777,
+      "learning_rate": 2.380070470149605e-06,
+      "loss": 1.1526,
+      "step": 1062
+    },
+    {
+      "epoch": 2.126,
+      "grad_norm": 12.152226448059082,
+      "learning_rate": 2.3701672802157567e-06,
+      "loss": 1.0407,
+      "step": 1063
+    },
+    {
+      "epoch": 2.128,
+      "grad_norm": 12.016361236572266,
+      "learning_rate": 2.3602783319262847e-06,
+      "loss": 1.0879,
+      "step": 1064
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 11.808967590332031,
+      "learning_rate": 2.3504036788339763e-06,
+      "loss": 1.1118,
+      "step": 1065
+    },
+    {
+      "epoch": 2.132,
+      "grad_norm": 11.515680313110352,
+      "learning_rate": 2.340543374414212e-06,
+      "loss": 1.0536,
+      "step": 1066
+    },
+    {
+      "epoch": 2.134,
+      "grad_norm": 10.819012641906738,
+      "learning_rate": 2.330697472064667e-06,
+      "loss": 1.0542,
+      "step": 1067
+    },
+    {
+      "epoch": 2.136,
+      "grad_norm": 12.405117988586426,
+      "learning_rate": 2.320866025105016e-06,
+      "loss": 1.0506,
+      "step": 1068
+    },
+    {
+      "epoch": 2.138,
+      "grad_norm": 11.67018985748291,
+      "learning_rate": 2.3110490867766644e-06,
+      "loss": 1.101,
+      "step": 1069
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 12.757566452026367,
+      "learning_rate": 2.3012467102424373e-06,
+      "loss": 1.093,
+      "step": 1070
+    },
+    {
+      "epoch": 2.142,
+      "grad_norm": 12.47193717956543,
+      "learning_rate": 2.2914589485863015e-06,
+      "loss": 1.1508,
+      "step": 1071
+    },
+    {
+      "epoch": 2.144,
+      "grad_norm": 11.597710609436035,
+      "learning_rate": 2.2816858548130837e-06,
+      "loss": 1.1538,
+      "step": 1072
+    },
+    {
+      "epoch": 2.146,
+      "grad_norm": 12.810328483581543,
+      "learning_rate": 2.2719274818481767e-06,
+      "loss": 1.0238,
+      "step": 1073
+    },
+    {
+      "epoch": 2.148,
+      "grad_norm": 13.841683387756348,
+      "learning_rate": 2.2621838825372496e-06,
+      "loss": 1.202,
+      "step": 1074
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 12.471292495727539,
+      "learning_rate": 2.2524551096459703e-06,
+      "loss": 1.0969,
+      "step": 1075
+    },
+    {
+      "epoch": 2.152,
+      "grad_norm": 11.695208549499512,
+      "learning_rate": 2.2427412158597133e-06,
+      "loss": 1.0346,
+      "step": 1076
+    },
+    {
+      "epoch": 2.154,
+      "grad_norm": 12.01142406463623,
+      "learning_rate": 2.23304225378328e-06,
+      "loss": 1.091,
+      "step": 1077
+    },
+    {
+      "epoch": 2.156,
+      "grad_norm": 14.317254066467285,
+      "learning_rate": 2.2233582759406065e-06,
+      "loss": 1.1849,
+      "step": 1078
+    },
+    {
+      "epoch": 2.158,
+      "grad_norm": 11.322208404541016,
+      "learning_rate": 2.213689334774479e-06,
+      "loss": 1.0879,
+      "step": 1079
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 14.340149879455566,
+      "learning_rate": 2.204035482646267e-06,
+      "loss": 1.1532,
+      "step": 1080
+    },
+    {
+      "epoch": 2.162,
+      "grad_norm": 12.387587547302246,
+      "learning_rate": 2.1943967718356123e-06,
+      "loss": 1.0726,
+      "step": 1081
+    },
+    {
+      "epoch": 2.164,
+      "grad_norm": 12.853300094604492,
+      "learning_rate": 2.184773254540169e-06,
+      "loss": 1.1011,
+      "step": 1082
+    },
+    {
+      "epoch": 2.166,
+      "grad_norm": 12.014087677001953,
+      "learning_rate": 2.175164982875311e-06,
+      "loss": 1.1168,
+      "step": 1083
+    },
+    {
+      "epoch": 2.168,
+      "grad_norm": 13.224556922912598,
+      "learning_rate": 2.165572008873845e-06,
+      "loss": 1.1332,
+      "step": 1084
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 10.886467933654785,
+      "learning_rate": 2.155994384485742e-06,
+      "loss": 0.9585,
+      "step": 1085
+    },
+    {
+      "epoch": 2.172,
+      "grad_norm": 13.911365509033203,
+      "learning_rate": 2.146432161577842e-06,
+      "loss": 1.1193,
+      "step": 1086
+    },
+    {
+      "epoch": 2.174,
+      "grad_norm": 11.698256492614746,
+      "learning_rate": 2.1368853919335835e-06,
+      "loss": 1.1248,
+      "step": 1087
+    },
+    {
+      "epoch": 2.176,
+      "grad_norm": 11.102900505065918,
+      "learning_rate": 2.12735412725272e-06,
+      "loss": 1.1625,
+      "step": 1088
+    },
+    {
+      "epoch": 2.178,
+      "grad_norm": 11.995986938476562,
+      "learning_rate": 2.1178384191510344e-06,
+      "loss": 1.1379,
+      "step": 1089
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 10.211990356445312,
+      "learning_rate": 2.1083383191600676e-06,
+      "loss": 1.0004,
+      "step": 1090
+    },
+    {
+      "epoch": 2.182,
+      "grad_norm": 11.571313858032227,
+      "learning_rate": 2.0988538787268374e-06,
+      "loss": 1.1064,
+      "step": 1091
+    },
+    {
+      "epoch": 2.184,
+      "grad_norm": 12.789441108703613,
+      "learning_rate": 2.0893851492135536e-06,
+      "loss": 1.1492,
+      "step": 1092
+    },
+    {
+      "epoch": 2.186,
+      "grad_norm": 11.842870712280273,
+      "learning_rate": 2.0799321818973488e-06,
+      "loss": 1.1277,
+      "step": 1093
+    },
+    {
+      "epoch": 2.188,
+      "grad_norm": 12.860237121582031,
+      "learning_rate": 2.0704950279699986e-06,
+      "loss": 1.112,
+      "step": 1094
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 12.648483276367188,
+      "learning_rate": 2.061073738537635e-06,
+      "loss": 0.9933,
+      "step": 1095
+    },
+    {
+      "epoch": 2.192,
+      "grad_norm": 12.022543907165527,
+      "learning_rate": 2.0516683646204836e-06,
+      "loss": 1.0655,
+      "step": 1096
+    },
+    {
+      "epoch": 2.194,
+      "grad_norm": 12.181710243225098,
+      "learning_rate": 2.0422789571525813e-06,
+      "loss": 1.1167,
+      "step": 1097
+    },
+    {
+      "epoch": 2.196,
+      "grad_norm": 12.39856243133545,
+      "learning_rate": 2.0329055669814936e-06,
+      "loss": 1.1543,
+      "step": 1098
+    },
+    {
+      "epoch": 2.198,
+      "grad_norm": 10.813960075378418,
+      "learning_rate": 2.023548244868051e-06,
+      "loss": 1.031,
+      "step": 1099
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 9.889363288879395,
+      "learning_rate": 2.0142070414860704e-06,
+      "loss": 1.0438,
+      "step": 1100
+    },
+    {
+      "epoch": 2.202,
+      "grad_norm": 11.947040557861328,
+      "learning_rate": 2.0048820074220716e-06,
+      "loss": 1.0579,
+      "step": 1101
+    },
+    {
+      "epoch": 2.204,
+      "grad_norm": 10.942092895507812,
+      "learning_rate": 1.9955731931750182e-06,
+      "loss": 1.1079,
+      "step": 1102
+    },
+    {
+      "epoch": 2.206,
+      "grad_norm": 10.924947738647461,
+      "learning_rate": 1.9862806491560315e-06,
+      "loss": 1.1075,
+      "step": 1103
+    },
+    {
+      "epoch": 2.208,
+      "grad_norm": 11.159443855285645,
+      "learning_rate": 1.977004425688126e-06,
+      "loss": 1.0548,
+      "step": 1104
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 11.316449165344238,
+      "learning_rate": 1.9677445730059348e-06,
+      "loss": 1.1579,
+      "step": 1105
+    },
+    {
+      "epoch": 2.212,
+      "grad_norm": 12.146157264709473,
+      "learning_rate": 1.958501141255427e-06,
+      "loss": 1.136,
+      "step": 1106
+    },
+    {
+      "epoch": 2.214,
+      "grad_norm": 10.939238548278809,
+      "learning_rate": 1.9492741804936623e-06,
+      "loss": 1.0623,
+      "step": 1107
+    },
+    {
+      "epoch": 2.216,
+      "grad_norm": 12.354981422424316,
+      "learning_rate": 1.9400637406884875e-06,
+      "loss": 1.1298,
+      "step": 1108
+    },
+    {
+      "epoch": 2.218,
+      "grad_norm": 10.884641647338867,
+      "learning_rate": 1.9308698717182874e-06,
+      "loss": 0.9976,
+      "step": 1109
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 12.507721900939941,
+      "learning_rate": 1.9216926233717087e-06,
+      "loss": 1.1589,
+      "step": 1110
+    },
+    {
+      "epoch": 2.222,
+      "grad_norm": 12.828377723693848,
+      "learning_rate": 1.9125320453473923e-06,
+      "loss": 1.0951,
+      "step": 1111
+    },
+    {
+      "epoch": 2.224,
+      "grad_norm": 12.240373611450195,
+      "learning_rate": 1.9033881872537009e-06,
+      "loss": 1.1288,
+      "step": 1112
+    },
+    {
+      "epoch": 2.226,
+      "grad_norm": 11.162793159484863,
+      "learning_rate": 1.8942610986084487e-06,
+      "loss": 1.1229,
+      "step": 1113
+    },
+    {
+      "epoch": 2.228,
+      "grad_norm": 12.57970905303955,
+      "learning_rate": 1.88515082883864e-06,
+      "loss": 1.0853,
+      "step": 1114
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 12.0857515335083,
+      "learning_rate": 1.8760574272802002e-06,
+      "loss": 1.1373,
+      "step": 1115
+    },
+    {
+      "epoch": 2.232,
+      "grad_norm": 10.519421577453613,
+      "learning_rate": 1.8669809431776991e-06,
+      "loss": 1.1059,
+      "step": 1116
+    },
+    {
+      "epoch": 2.234,
+      "grad_norm": 11.660649299621582,
+      "learning_rate": 1.8579214256840938e-06,
+      "loss": 1.2105,
+      "step": 1117
+    },
+    {
+      "epoch": 2.2359999999999998,
+      "grad_norm": 11.070806503295898,
+      "learning_rate": 1.8488789238604676e-06,
+      "loss": 1.0679,
+      "step": 1118
+    },
+    {
+      "epoch": 2.238,
+      "grad_norm": 11.053348541259766,
+      "learning_rate": 1.8398534866757455e-06,
+      "loss": 1.0581,
+      "step": 1119
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 12.008465766906738,
+      "learning_rate": 1.8308451630064484e-06,
+      "loss": 1.1217,
+      "step": 1120
+    },
+    {
+      "epoch": 2.242,
+      "grad_norm": 12.64566707611084,
+      "learning_rate": 1.8218540016364178e-06,
+      "loss": 1.0691,
+      "step": 1121
+    },
+    {
+      "epoch": 2.2439999999999998,
+      "grad_norm": 12.815990447998047,
+      "learning_rate": 1.8128800512565514e-06,
+      "loss": 1.0881,
+      "step": 1122
+    },
+    {
+      "epoch": 2.246,
+      "grad_norm": 12.682955741882324,
+      "learning_rate": 1.8039233604645468e-06,
+      "loss": 1.0893,
+      "step": 1123
+    },
+    {
+      "epoch": 2.248,
+      "grad_norm": 11.445165634155273,
+      "learning_rate": 1.7949839777646327e-06,
+      "loss": 1.1113,
+      "step": 1124
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 11.555493354797363,
+      "learning_rate": 1.7860619515673034e-06,
+      "loss": 1.1268,
+      "step": 1125
+    },
+    {
+      "epoch": 2.252,
+      "grad_norm": 12.906476974487305,
+      "learning_rate": 1.7771573301890666e-06,
+      "loss": 1.1886,
+      "step": 1126
+    },
+    {
+      "epoch": 2.254,
+      "grad_norm": 11.322837829589844,
+      "learning_rate": 1.7682701618521687e-06,
+      "loss": 1.0748,
+      "step": 1127
+    },
+    {
+      "epoch": 2.2560000000000002,
+      "grad_norm": 14.397950172424316,
+      "learning_rate": 1.7594004946843458e-06,
+      "loss": 1.174,
+      "step": 1128
+    },
+    {
+      "epoch": 2.258,
+      "grad_norm": 12.00009536743164,
+      "learning_rate": 1.7505483767185583e-06,
+      "loss": 1.0512,
+      "step": 1129
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 13.125904083251953,
+      "learning_rate": 1.7417138558927244e-06,
+      "loss": 1.1748,
+      "step": 1130
+    },
+    {
+      "epoch": 2.262,
+      "grad_norm": 11.499984741210938,
+      "learning_rate": 1.7328969800494727e-06,
+      "loss": 0.9966,
+      "step": 1131
+    },
+    {
+      "epoch": 2.2640000000000002,
+      "grad_norm": 11.893850326538086,
+      "learning_rate": 1.7240977969358757e-06,
+      "loss": 1.1099,
+      "step": 1132
+    },
+    {
+      "epoch": 2.266,
+      "grad_norm": 13.56734848022461,
+      "learning_rate": 1.7153163542031881e-06,
+      "loss": 1.1765,
+      "step": 1133
+    },
+    {
+      "epoch": 2.268,
+      "grad_norm": 11.519514083862305,
+      "learning_rate": 1.7065526994065973e-06,
+      "loss": 1.0598,
+      "step": 1134
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 13.245732307434082,
+      "learning_rate": 1.6978068800049624e-06,
+      "loss": 0.9934,
+      "step": 1135
+    },
+    {
+      "epoch": 2.2720000000000002,
+      "grad_norm": 12.127825736999512,
+      "learning_rate": 1.6890789433605508e-06,
+      "loss": 1.1238,
+      "step": 1136
+    },
+    {
+      "epoch": 2.274,
+      "grad_norm": 12.772904396057129,
+      "learning_rate": 1.680368936738792e-06,
+      "loss": 1.1137,
+      "step": 1137
+    },
+    {
+      "epoch": 2.276,
+      "grad_norm": 13.146944046020508,
+      "learning_rate": 1.671676907308018e-06,
+      "loss": 1.1266,
+      "step": 1138
+    },
+    {
+      "epoch": 2.278,
+      "grad_norm": 11.267017364501953,
+      "learning_rate": 1.6630029021392007e-06,
+      "loss": 1.1258,
+      "step": 1139
+    },
+    {
+      "epoch": 2.2800000000000002,
+      "grad_norm": 11.609925270080566,
+      "learning_rate": 1.6543469682057105e-06,
+      "loss": 1.097,
+      "step": 1140
+    },
+    {
+      "epoch": 2.282,
+      "grad_norm": 12.532341957092285,
+      "learning_rate": 1.645709152383046e-06,
+      "loss": 1.193,
+      "step": 1141
+    },
+    {
+      "epoch": 2.284,
+      "grad_norm": 11.958260536193848,
+      "learning_rate": 1.6370895014486e-06,
+      "loss": 1.0575,
+      "step": 1142
+    },
+    {
+      "epoch": 2.286,
+      "grad_norm": 12.34417724609375,
+      "learning_rate": 1.6284880620813847e-06,
+      "loss": 1.0509,
+      "step": 1143
+    },
+    {
+      "epoch": 2.288,
+      "grad_norm": 12.214522361755371,
+      "learning_rate": 1.6199048808617896e-06,
+      "loss": 1.0467,
+      "step": 1144
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 12.764805793762207,
+      "learning_rate": 1.611340004271339e-06,
+      "loss": 1.0441,
+      "step": 1145
+    },
+    {
+      "epoch": 2.292,
+      "grad_norm": 11.342555046081543,
+      "learning_rate": 1.6027934786924187e-06,
+      "loss": 1.1076,
+      "step": 1146
+    },
+    {
+      "epoch": 2.294,
+      "grad_norm": 13.207151412963867,
+      "learning_rate": 1.594265350408039e-06,
+      "loss": 1.2069,
+      "step": 1147
+    },
+    {
+      "epoch": 2.296,
+      "grad_norm": 11.775276184082031,
+      "learning_rate": 1.5857556656015837e-06,
+      "loss": 1.0898,
+      "step": 1148
+    },
+    {
+      "epoch": 2.298,
+      "grad_norm": 11.530139923095703,
+      "learning_rate": 1.5772644703565564e-06,
+      "loss": 1.098,
+      "step": 1149
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 11.4733247756958,
+      "learning_rate": 1.5687918106563326e-06,
+      "loss": 1.0295,
+      "step": 1150
+    },
+    {
+      "epoch": 2.302,
+      "grad_norm": 11.146039962768555,
+      "learning_rate": 1.5603377323839069e-06,
+      "loss": 1.0014,
+      "step": 1151
+    },
+    {
+      "epoch": 2.304,
+      "grad_norm": 13.669384002685547,
+      "learning_rate": 1.551902281321651e-06,
+      "loss": 1.2271,
+      "step": 1152
+    },
+    {
+      "epoch": 2.306,
+      "grad_norm": 12.079588890075684,
+      "learning_rate": 1.5434855031510626e-06,
+      "loss": 1.0904,
+      "step": 1153
+    },
+    {
+      "epoch": 2.308,
+      "grad_norm": 11.25638198852539,
+      "learning_rate": 1.5350874434525142e-06,
+      "loss": 1.0801,
+      "step": 1154
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 11.719844818115234,
+      "learning_rate": 1.5267081477050132e-06,
+      "loss": 1.0488,
+      "step": 1155
+    },
+    {
+      "epoch": 2.312,
+      "grad_norm": 11.61817455291748,
+      "learning_rate": 1.5183476612859538e-06,
+      "loss": 1.0197,
+      "step": 1156
+    },
+    {
+      "epoch": 2.314,
+      "grad_norm": 12.426447868347168,
+      "learning_rate": 1.5100060294708647e-06,
+      "loss": 1.155,
+      "step": 1157
+    },
+    {
+      "epoch": 2.316,
+      "grad_norm": 11.935876846313477,
+      "learning_rate": 1.5016832974331725e-06,
+      "loss": 1.0953,
+      "step": 1158
+    },
+    {
+      "epoch": 2.318,
+      "grad_norm": 15.38991641998291,
+      "learning_rate": 1.4933795102439558e-06,
+      "loss": 1.212,
+      "step": 1159
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 11.428476333618164,
+      "learning_rate": 1.4850947128716914e-06,
+      "loss": 1.1117,
+      "step": 1160
+    },
+    {
+      "epoch": 2.322,
+      "grad_norm": 11.7312593460083,
+      "learning_rate": 1.4768289501820265e-06,
+      "loss": 1.0994,
+      "step": 1161
+    },
+    {
+      "epoch": 2.324,
+      "grad_norm": 11.2421236038208,
+      "learning_rate": 1.4685822669375239e-06,
+      "loss": 1.1205,
+      "step": 1162
+    },
+    {
+      "epoch": 2.326,
+      "grad_norm": 10.058606147766113,
+      "learning_rate": 1.4603547077974217e-06,
+      "loss": 1.0785,
+      "step": 1163
+    },
+    {
+      "epoch": 2.328,
+      "grad_norm": 11.7360258102417,
+      "learning_rate": 1.4521463173173966e-06,
+      "loss": 1.097,
+      "step": 1164
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 13.815256118774414,
+      "learning_rate": 1.4439571399493146e-06,
+      "loss": 1.0887,
+      "step": 1165
+    },
+    {
+      "epoch": 2.332,
+      "grad_norm": 12.337494850158691,
+      "learning_rate": 1.4357872200409988e-06,
+      "loss": 1.1101,
+      "step": 1166
+    },
+    {
+      "epoch": 2.334,
+      "grad_norm": 11.706501007080078,
+      "learning_rate": 1.4276366018359845e-06,
+      "loss": 1.0746,
+      "step": 1167
+    },
+    {
+      "epoch": 2.336,
+      "grad_norm": 12.499887466430664,
+      "learning_rate": 1.4195053294732757e-06,
+      "loss": 1.1403,
+      "step": 1168
+    },
+    {
+      "epoch": 2.338,
+      "grad_norm": 12.3418550491333,
+      "learning_rate": 1.4113934469871166e-06,
+      "loss": 1.1632,
+      "step": 1169
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 12.219500541687012,
+      "learning_rate": 1.4033009983067454e-06,
+      "loss": 1.1484,
+      "step": 1170
+    },
+    {
+      "epoch": 2.342,
+      "grad_norm": 11.66283893585205,
+      "learning_rate": 1.3952280272561541e-06,
+      "loss": 1.0012,
+      "step": 1171
+    },
+    {
+      "epoch": 2.344,
+      "grad_norm": 13.044709205627441,
+      "learning_rate": 1.3871745775538598e-06,
+      "loss": 1.2462,
+      "step": 1172
+    },
+    {
+      "epoch": 2.346,
+      "grad_norm": 11.258277893066406,
+      "learning_rate": 1.3791406928126638e-06,
+      "loss": 1.1165,
+      "step": 1173
+    },
+    {
+      "epoch": 2.348,
+      "grad_norm": 11.9083833694458,
+      "learning_rate": 1.371126416539409e-06,
+      "loss": 1.0926,
+      "step": 1174
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 12.25865364074707,
+      "learning_rate": 1.3631317921347564e-06,
+      "loss": 1.1481,
+      "step": 1175
+    },
+    {
+      "epoch": 2.352,
+      "grad_norm": 11.566573143005371,
+      "learning_rate": 1.3551568628929434e-06,
+      "loss": 1.0938,
+      "step": 1176
+    },
+    {
+      "epoch": 2.354,
+      "grad_norm": 12.630902290344238,
+      "learning_rate": 1.3472016720015447e-06,
+      "loss": 1.1382,
+      "step": 1177
+    },
+    {
+      "epoch": 2.356,
+      "grad_norm": 12.472233772277832,
+      "learning_rate": 1.339266262541249e-06,
+      "loss": 1.149,
+      "step": 1178
+    },
+    {
+      "epoch": 2.358,
+      "grad_norm": 11.453776359558105,
+      "learning_rate": 1.3313506774856177e-06,
+      "loss": 1.0327,
+      "step": 1179
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 10.934307098388672,
+      "learning_rate": 1.3234549597008572e-06,
+      "loss": 1.1564,
+      "step": 1180
+    },
+    {
+      "epoch": 2.362,
+      "grad_norm": 13.961017608642578,
+      "learning_rate": 1.3155791519455812e-06,
+      "loss": 1.0754,
+      "step": 1181
+    },
+    {
+      "epoch": 2.364,
+      "grad_norm": 12.650055885314941,
+      "learning_rate": 1.3077232968705805e-06,
+      "loss": 1.1337,
+      "step": 1182
+    },
+    {
+      "epoch": 2.366,
+      "grad_norm": 12.112220764160156,
+      "learning_rate": 1.2998874370186026e-06,
+      "loss": 1.0891,
+      "step": 1183
+    },
+    {
+      "epoch": 2.368,
+      "grad_norm": 11.407106399536133,
+      "learning_rate": 1.2920716148241036e-06,
+      "loss": 1.1488,
+      "step": 1184
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 11.474730491638184,
+      "learning_rate": 1.2842758726130283e-06,
+      "loss": 1.1326,
+      "step": 1185
+    },
+    {
+      "epoch": 2.372,
+      "grad_norm": 12.02614974975586,
+      "learning_rate": 1.2765002526025871e-06,
+      "loss": 1.0903,
+      "step": 1186
+    },
+    {
+      "epoch": 2.374,
+      "grad_norm": 13.217632293701172,
+      "learning_rate": 1.2687447969010113e-06,
+      "loss": 1.1562,
+      "step": 1187
+    },
+    {
+      "epoch": 2.376,
+      "grad_norm": 13.802258491516113,
+      "learning_rate": 1.2610095475073415e-06,
+      "loss": 1.1038,
+      "step": 1188
+    },
+    {
+      "epoch": 2.378,
+      "grad_norm": 11.107643127441406,
+      "learning_rate": 1.2532945463111856e-06,
+      "loss": 1.0914,
+      "step": 1189
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 12.443587303161621,
+      "learning_rate": 1.2455998350925042e-06,
+      "loss": 1.1518,
+      "step": 1190
+    },
+    {
+      "epoch": 2.382,
+      "grad_norm": 12.572113990783691,
+      "learning_rate": 1.2379254555213788e-06,
+      "loss": 1.1528,
+      "step": 1191
+    },
+    {
+      "epoch": 2.384,
+      "grad_norm": 11.02535629272461,
+      "learning_rate": 1.2302714491577834e-06,
+      "loss": 1.1726,
+      "step": 1192
+    },
+    {
+      "epoch": 2.386,
+      "grad_norm": 12.016326904296875,
+      "learning_rate": 1.2226378574513654e-06,
+      "loss": 1.1441,
+      "step": 1193
+    },
+    {
+      "epoch": 2.388,
+      "grad_norm": 11.50503921508789,
+      "learning_rate": 1.2150247217412186e-06,
+      "loss": 1.2099,
+      "step": 1194
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 11.974237442016602,
+      "learning_rate": 1.2074320832556558e-06,
+      "loss": 1.1982,
+      "step": 1195
+    },
+    {
+      "epoch": 2.392,
+      "grad_norm": 11.554012298583984,
+      "learning_rate": 1.1998599831119912e-06,
+      "loss": 1.0467,
+      "step": 1196
+    },
+    {
+      "epoch": 2.394,
+      "grad_norm": 13.131747245788574,
+      "learning_rate": 1.1923084623163172e-06,
+      "loss": 1.143,
+      "step": 1197
+    },
+    {
+      "epoch": 2.396,
+      "grad_norm": 10.756204605102539,
+      "learning_rate": 1.1847775617632746e-06,
+      "loss": 1.1161,
+      "step": 1198
+    },
+    {
+      "epoch": 2.398,
+      "grad_norm": 12.911304473876953,
+      "learning_rate": 1.1772673222358421e-06,
+      "loss": 1.1532,
+      "step": 1199
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 10.806382179260254,
+      "learning_rate": 1.1697777844051105e-06,
+      "loss": 1.0573,
+      "step": 1200
+    },
+    {
+      "epoch": 2.402,
+      "grad_norm": 11.249881744384766,
+      "learning_rate": 1.162308988830057e-06,
+      "loss": 1.1033,
+      "step": 1201
+    },
+    {
+      "epoch": 2.404,
+      "grad_norm": 11.274486541748047,
+      "learning_rate": 1.1548609759573375e-06,
+      "loss": 1.2207,
+      "step": 1202
+    },
+    {
+      "epoch": 2.406,
+      "grad_norm": 10.787302017211914,
+      "learning_rate": 1.1474337861210543e-06,
+      "loss": 1.1259,
+      "step": 1203
+    },
+    {
+      "epoch": 2.408,
+      "grad_norm": 10.975671768188477,
+      "learning_rate": 1.1400274595425499e-06,
+      "loss": 1.0418,
+      "step": 1204
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 10.228716850280762,
+      "learning_rate": 1.132642036330181e-06,
+      "loss": 1.087,
+      "step": 1205
+    },
+    {
+      "epoch": 2.412,
+      "grad_norm": 13.028627395629883,
+      "learning_rate": 1.1252775564791023e-06,
+      "loss": 1.1651,
+      "step": 1206
+    },
+    {
+      "epoch": 2.414,
+      "grad_norm": 10.35562801361084,
+      "learning_rate": 1.1179340598710547e-06,
+      "loss": 1.1121,
+      "step": 1207
+    },
+    {
+      "epoch": 2.416,
+      "grad_norm": 11.16613483428955,
+      "learning_rate": 1.1106115862741457e-06,
+      "loss": 1.1242,
+      "step": 1208
+    },
+    {
+      "epoch": 2.418,
+      "grad_norm": 11.81697940826416,
+      "learning_rate": 1.1033101753426285e-06,
+      "loss": 1.167,
+      "step": 1209
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 13.364669799804688,
+      "learning_rate": 1.096029866616704e-06,
+      "loss": 1.0961,
+      "step": 1210
+    },
+    {
+      "epoch": 2.422,
+      "grad_norm": 11.796897888183594,
+      "learning_rate": 1.0887706995222864e-06,
+      "loss": 1.137,
+      "step": 1211
+    },
+    {
+      "epoch": 2.424,
+      "grad_norm": 11.452189445495605,
+      "learning_rate": 1.0815327133708015e-06,
+      "loss": 1.1626,
+      "step": 1212
+    },
+    {
+      "epoch": 2.426,
+      "grad_norm": 11.605690956115723,
+      "learning_rate": 1.0743159473589738e-06,
+      "loss": 1.1818,
+      "step": 1213
+    },
+    {
+      "epoch": 2.428,
+      "grad_norm": 12.917741775512695,
+      "learning_rate": 1.0671204405686108e-06,
+      "loss": 1.0901,
+      "step": 1214
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 11.108345985412598,
+      "learning_rate": 1.0599462319663906e-06,
+      "loss": 1.0953,
+      "step": 1215
+    },
+    {
+      "epoch": 2.432,
+      "grad_norm": 11.12553882598877,
+      "learning_rate": 1.052793360403655e-06,
+      "loss": 1.1915,
+      "step": 1216
+    },
+    {
+      "epoch": 2.434,
+      "grad_norm": 12.073833465576172,
+      "learning_rate": 1.0456618646161954e-06,
+      "loss": 1.2316,
+      "step": 1217
+    },
+    {
+      "epoch": 2.436,
+      "grad_norm": 11.272497177124023,
+      "learning_rate": 1.0385517832240472e-06,
+      "loss": 1.1353,
+      "step": 1218
+    },
+    {
+      "epoch": 2.438,
+      "grad_norm": 11.970922470092773,
+      "learning_rate": 1.0314631547312738e-06,
+      "loss": 1.1071,
+      "step": 1219
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 11.06673812866211,
+      "learning_rate": 1.0243960175257605e-06,
+      "loss": 1.1197,
+      "step": 1220
+    },
+    {
+      "epoch": 2.442,
+      "grad_norm": 11.756420135498047,
+      "learning_rate": 1.0173504098790188e-06,
+      "loss": 1.0385,
+      "step": 1221
+    },
+    {
+      "epoch": 2.444,
+      "grad_norm": 11.263325691223145,
+      "learning_rate": 1.010326369945957e-06,
+      "loss": 1.0892,
+      "step": 1222
+    },
+    {
+      "epoch": 2.446,
+      "grad_norm": 11.817276954650879,
+      "learning_rate": 1.0033239357646913e-06,
+      "loss": 1.138,
+      "step": 1223
+    },
+    {
+      "epoch": 2.448,
+      "grad_norm": 11.443320274353027,
+      "learning_rate": 9.963431452563331e-07,
+      "loss": 1.0683,
+      "step": 1224
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 11.445216178894043,
+      "learning_rate": 9.893840362247809e-07,
+      "loss": 1.1286,
+      "step": 1225
+    },
+    {
+      "epoch": 2.452,
+      "grad_norm": 12.109580039978027,
+      "learning_rate": 9.824466463565246e-07,
+      "loss": 1.2696,
+      "step": 1226
+    },
+    {
+      "epoch": 2.454,
+      "grad_norm": 12.882740020751953,
+      "learning_rate": 9.7553101322043e-07,
+      "loss": 1.1698,
+      "step": 1227
+    },
+    {
+      "epoch": 2.456,
+      "grad_norm": 12.377523422241211,
+      "learning_rate": 9.686371742675443e-07,
+      "loss": 1.2098,
+      "step": 1228
+    },
+    {
+      "epoch": 2.458,
+      "grad_norm": 11.193984985351562,
+      "learning_rate": 9.617651668308914e-07,
+      "loss": 1.1167,
+      "step": 1229
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 11.221639633178711,
+      "learning_rate": 9.549150281252633e-07,
+      "loss": 1.1706,
+      "step": 1230
+    },
+    {
+      "epoch": 2.462,
+      "grad_norm": 11.720982551574707,
+      "learning_rate": 9.480867952470285e-07,
+      "loss": 1.1991,
+      "step": 1231
+    },
+    {
+      "epoch": 2.464,
+      "grad_norm": 9.958009719848633,
+      "learning_rate": 9.412805051739266e-07,
+      "loss": 1.0561,
+      "step": 1232
+    },
+    {
+      "epoch": 2.466,
+      "grad_norm": 12.225953102111816,
+      "learning_rate": 9.344961947648624e-07,
+      "loss": 1.0812,
+      "step": 1233
+    },
+    {
+      "epoch": 2.468,
+      "grad_norm": 11.826346397399902,
+      "learning_rate": 9.277339007597158e-07,
+      "loss": 1.1611,
+      "step": 1234
+    },
+    {
+      "epoch": 2.4699999999999998,
+      "grad_norm": 11.533407211303711,
+      "learning_rate": 9.209936597791407e-07,
+      "loss": 1.1837,
+      "step": 1235
+    },
+    {
+      "epoch": 2.472,
+      "grad_norm": 10.118829727172852,
+      "learning_rate": 9.142755083243577e-07,
+      "loss": 1.0882,
+      "step": 1236
+    },
+    {
+      "epoch": 2.474,
+      "grad_norm": 13.343694686889648,
+      "learning_rate": 9.075794827769696e-07,
+      "loss": 1.1914,
+      "step": 1237
+    },
+    {
+      "epoch": 2.476,
+      "grad_norm": 11.293721199035645,
+      "learning_rate": 9.009056193987569e-07,
+      "loss": 1.0637,
+      "step": 1238
+    },
+    {
+      "epoch": 2.4779999999999998,
+      "grad_norm": 10.591233253479004,
+      "learning_rate": 8.942539543314799e-07,
+      "loss": 1.1345,
+      "step": 1239
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 13.138537406921387,
+      "learning_rate": 8.876245235966884e-07,
+      "loss": 1.2053,
+      "step": 1240
+    },
+    {
+      "epoch": 2.482,
+      "grad_norm": 10.92805290222168,
+      "learning_rate": 8.810173630955249e-07,
+      "loss": 1.1372,
+      "step": 1241
+    },
+    {
+      "epoch": 2.484,
+      "grad_norm": 11.212373733520508,
+      "learning_rate": 8.744325086085248e-07,
+      "loss": 1.0706,
+      "step": 1242
+    },
+    {
+      "epoch": 2.4859999999999998,
+      "grad_norm": 11.673620223999023,
+      "learning_rate": 8.678699957954323e-07,
+      "loss": 1.1367,
+      "step": 1243
+    },
+    {
+      "epoch": 2.488,
+      "grad_norm": 11.204456329345703,
+      "learning_rate": 8.613298601949971e-07,
+      "loss": 1.0237,
+      "step": 1244
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 12.066752433776855,
+      "learning_rate": 8.54812137224792e-07,
+      "loss": 1.2087,
+      "step": 1245
+    },
+    {
+      "epoch": 2.492,
+      "grad_norm": 10.22867202758789,
+      "learning_rate": 8.483168621810133e-07,
+      "loss": 1.1078,
+      "step": 1246
+    },
+    {
+      "epoch": 2.4939999999999998,
+      "grad_norm": 12.135293960571289,
+      "learning_rate": 8.418440702382897e-07,
+      "loss": 1.0668,
+      "step": 1247
+    },
+    {
+      "epoch": 2.496,
+      "grad_norm": 9.851129531860352,
+      "learning_rate": 8.353937964495029e-07,
+      "loss": 1.0521,
+      "step": 1248
+    },
+    {
+      "epoch": 2.498,
+      "grad_norm": 10.662277221679688,
+      "learning_rate": 8.289660757455803e-07,
+      "loss": 1.0326,
+      "step": 1249
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 13.312153816223145,
+      "learning_rate": 8.225609429353187e-07,
+      "loss": 1.0838,
+      "step": 1250
+    },
+    {
+      "epoch": 2.502,
+      "grad_norm": 10.633809089660645,
+      "learning_rate": 8.161784327051919e-07,
+      "loss": 1.0238,
+      "step": 1251
+    },
+    {
+      "epoch": 2.504,
+      "grad_norm": 11.557962417602539,
+      "learning_rate": 8.098185796191632e-07,
+      "loss": 1.1297,
+      "step": 1252
+    },
+    {
+      "epoch": 2.5060000000000002,
+      "grad_norm": 10.614686965942383,
+      "learning_rate": 8.034814181184996e-07,
+      "loss": 1.1224,
+      "step": 1253
+    },
+    {
+      "epoch": 2.508,
+      "grad_norm": 11.647783279418945,
+      "learning_rate": 7.971669825215789e-07,
+      "loss": 1.191,
+      "step": 1254
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 12.704024314880371,
+      "learning_rate": 7.908753070237124e-07,
+      "loss": 1.0431,
+      "step": 1255
+    },
+    {
+      "epoch": 2.512,
+      "grad_norm": 11.132079124450684,
+      "learning_rate": 7.846064256969571e-07,
+      "loss": 1.1417,
+      "step": 1256
+    },
+    {
+      "epoch": 2.5140000000000002,
+      "grad_norm": 12.66719913482666,
+      "learning_rate": 7.783603724899258e-07,
+      "loss": 1.1726,
+      "step": 1257
+    },
+    {
+      "epoch": 2.516,
+      "grad_norm": 13.010668754577637,
+      "learning_rate": 7.72137181227608e-07,
+      "loss": 1.2599,
+      "step": 1258
+    },
+    {
+      "epoch": 2.518,
+      "grad_norm": 13.327101707458496,
+      "learning_rate": 7.659368856111926e-07,
+      "loss": 1.0874,
+      "step": 1259
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 12.53348445892334,
+      "learning_rate": 7.597595192178702e-07,
+      "loss": 1.0867,
+      "step": 1260
+    },
+    {
+      "epoch": 2.5220000000000002,
+      "grad_norm": 11.78579044342041,
+      "learning_rate": 7.536051155006657e-07,
+      "loss": 1.1224,
+      "step": 1261
+    },
+    {
+      "epoch": 2.524,
+      "grad_norm": 12.989802360534668,
+      "learning_rate": 7.47473707788251e-07,
+      "loss": 1.0848,
+      "step": 1262
+    },
+    {
+      "epoch": 2.526,
+      "grad_norm": 10.429177284240723,
+      "learning_rate": 7.413653292847617e-07,
+      "loss": 1.0456,
+      "step": 1263
+    },
+    {
+      "epoch": 2.528,
+      "grad_norm": 12.047698020935059,
+      "learning_rate": 7.352800130696253e-07,
+      "loss": 1.1725,
+      "step": 1264
+    },
+    {
+      "epoch": 2.5300000000000002,
+      "grad_norm": 13.212532043457031,
+      "learning_rate": 7.292177920973726e-07,
+      "loss": 1.0755,
+      "step": 1265
+    },
+    {
+      "epoch": 2.532,
+      "grad_norm": 11.261322021484375,
+      "learning_rate": 7.23178699197467e-07,
+      "loss": 1.0894,
+      "step": 1266
+    },
+    {
+      "epoch": 2.534,
+      "grad_norm": 12.5214262008667,
+      "learning_rate": 7.171627670741243e-07,
+      "loss": 1.0554,
+      "step": 1267
+    },
+    {
+      "epoch": 2.536,
+      "grad_norm": 11.23154067993164,
+      "learning_rate": 7.111700283061318e-07,
+      "loss": 1.1117,
+      "step": 1268
+    },
+    {
+      "epoch": 2.5380000000000003,
+      "grad_norm": 14.136534690856934,
+      "learning_rate": 7.052005153466779e-07,
+      "loss": 1.1256,
+      "step": 1269
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 11.402527809143066,
+      "learning_rate": 6.992542605231739e-07,
+      "loss": 1.1365,
+      "step": 1270
+    },
+    {
+      "epoch": 2.542,
+      "grad_norm": 10.729368209838867,
+      "learning_rate": 6.933312960370748e-07,
+      "loss": 1.0958,
+      "step": 1271
+    },
+    {
+      "epoch": 2.544,
+      "grad_norm": 12.732880592346191,
+      "learning_rate": 6.874316539637127e-07,
+      "loss": 1.1066,
+      "step": 1272
+    },
+    {
+      "epoch": 2.5460000000000003,
+      "grad_norm": 11.761198043823242,
+      "learning_rate": 6.815553662521185e-07,
+      "loss": 1.0512,
+      "step": 1273
+    },
+    {
+      "epoch": 2.548,
+      "grad_norm": 12.321219444274902,
+      "learning_rate": 6.757024647248456e-07,
+      "loss": 1.1105,
+      "step": 1274
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 10.499734878540039,
+      "learning_rate": 6.698729810778065e-07,
+      "loss": 1.1287,
+      "step": 1275
+    },
+    {
+      "epoch": 2.552,
+      "grad_norm": 11.47294807434082,
+      "learning_rate": 6.640669468800947e-07,
+      "loss": 1.1906,
+      "step": 1276
+    },
+    {
+      "epoch": 2.5540000000000003,
+      "grad_norm": 10.670211791992188,
+      "learning_rate": 6.58284393573812e-07,
+      "loss": 1.0593,
+      "step": 1277
+    },
+    {
+      "epoch": 2.556,
+      "grad_norm": 11.801868438720703,
+      "learning_rate": 6.52525352473905e-07,
+      "loss": 1.0933,
+      "step": 1278
+    },
+    {
+      "epoch": 2.558,
+      "grad_norm": 12.327805519104004,
+      "learning_rate": 6.467898547679913e-07,
+      "loss": 1.0864,
+      "step": 1279
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 12.2786226272583,
+      "learning_rate": 6.410779315161885e-07,
+      "loss": 1.1561,
+      "step": 1280
+    },
+    {
+      "epoch": 2.5620000000000003,
+      "grad_norm": 11.15182876586914,
+      "learning_rate": 6.353896136509524e-07,
+      "loss": 1.1301,
+      "step": 1281
+    },
+    {
+      "epoch": 2.564,
+      "grad_norm": 11.805049896240234,
+      "learning_rate": 6.297249319769016e-07,
+      "loss": 1.0785,
+      "step": 1282
+    },
+    {
+      "epoch": 2.566,
+      "grad_norm": 11.21564769744873,
+      "learning_rate": 6.240839171706608e-07,
+      "loss": 1.2017,
+      "step": 1283
+    },
+    {
+      "epoch": 2.568,
+      "grad_norm": 12.291101455688477,
+      "learning_rate": 6.184665997806832e-07,
+      "loss": 1.0942,
+      "step": 1284
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 11.54765510559082,
+      "learning_rate": 6.128730102270897e-07,
+      "loss": 1.1542,
+      "step": 1285
+    },
+    {
+      "epoch": 2.572,
+      "grad_norm": 10.28836441040039,
+      "learning_rate": 6.073031788015133e-07,
+      "loss": 1.0565,
+      "step": 1286
+    },
+    {
+      "epoch": 2.574,
+      "grad_norm": 12.515631675720215,
+      "learning_rate": 6.017571356669183e-07,
+      "loss": 1.0679,
+      "step": 1287
+    },
+    {
+      "epoch": 2.576,
+      "grad_norm": 12.502131462097168,
+      "learning_rate": 5.962349108574478e-07,
+      "loss": 1.0424,
+      "step": 1288
+    },
+    {
+      "epoch": 2.578,
+      "grad_norm": 11.3397855758667,
+      "learning_rate": 5.9073653427826e-07,
+      "loss": 1.0948,
+      "step": 1289
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 10.07397174835205,
+      "learning_rate": 5.852620357053651e-07,
+      "loss": 1.0421,
+      "step": 1290
+    },
+    {
+      "epoch": 2.582,
+      "grad_norm": 11.410602569580078,
+      "learning_rate": 5.798114447854636e-07,
+      "loss": 1.1638,
+      "step": 1291
+    },
+    {
+      "epoch": 2.584,
+      "grad_norm": 11.01577091217041,
+      "learning_rate": 5.743847910357836e-07,
+      "loss": 1.0889,
+      "step": 1292
+    },
+    {
+      "epoch": 2.586,
+      "grad_norm": 12.289595603942871,
+      "learning_rate": 5.689821038439264e-07,
+      "loss": 1.1365,
+      "step": 1293
+    },
+    {
+      "epoch": 2.588,
+      "grad_norm": 12.653968811035156,
+      "learning_rate": 5.636034124677043e-07,
+      "loss": 1.208,
+      "step": 1294
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 11.91626262664795,
+      "learning_rate": 5.582487460349806e-07,
+      "loss": 1.2131,
+      "step": 1295
+    },
+    {
+      "epoch": 2.592,
+      "grad_norm": 12.287956237792969,
+      "learning_rate": 5.529181335435124e-07,
+      "loss": 1.096,
+      "step": 1296
+    },
+    {
+      "epoch": 2.594,
+      "grad_norm": 14.091611862182617,
+      "learning_rate": 5.476116038607993e-07,
+      "loss": 1.0923,
+      "step": 1297
+    },
+    {
+      "epoch": 2.596,
+      "grad_norm": 11.291535377502441,
+      "learning_rate": 5.423291857239177e-07,
+      "loss": 1.0751,
+      "step": 1298
+    },
+    {
+      "epoch": 2.598,
+      "grad_norm": 10.707423210144043,
+      "learning_rate": 5.370709077393721e-07,
+      "loss": 1.1536,
+      "step": 1299
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 10.066755294799805,
+      "learning_rate": 5.318367983829393e-07,
+      "loss": 1.1073,
+      "step": 1300
+    },
+    {
+      "epoch": 2.602,
+      "grad_norm": 12.727775573730469,
+      "learning_rate": 5.266268859995083e-07,
+      "loss": 1.0421,
+      "step": 1301
+    },
+    {
+      "epoch": 2.604,
+      "grad_norm": 10.819839477539062,
+      "learning_rate": 5.214411988029355e-07,
+      "loss": 1.0953,
+      "step": 1302
+    },
+    {
+      "epoch": 2.606,
+      "grad_norm": 11.471372604370117,
+      "learning_rate": 5.162797648758877e-07,
+      "loss": 1.1238,
+      "step": 1303
+    },
+    {
+      "epoch": 2.608,
+      "grad_norm": 11.147175788879395,
+      "learning_rate": 5.111426121696866e-07,
+      "loss": 1.1459,
+      "step": 1304
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 10.736647605895996,
+      "learning_rate": 5.06029768504166e-07,
+      "loss": 1.0977,
+      "step": 1305
+    },
+    {
+      "epoch": 2.612,
+      "grad_norm": 10.199406623840332,
+      "learning_rate": 5.009412615675102e-07,
+      "loss": 1.0806,
+      "step": 1306
+    },
+    {
+      "epoch": 2.614,
+      "grad_norm": 10.652769088745117,
+      "learning_rate": 4.958771189161149e-07,
+      "loss": 1.1096,
+      "step": 1307
+    },
+    {
+      "epoch": 2.616,
+      "grad_norm": 10.275524139404297,
+      "learning_rate": 4.908373679744316e-07,
+      "loss": 1.1095,
+      "step": 1308
+    },
+    {
+      "epoch": 2.618,
+      "grad_norm": 11.423064231872559,
+      "learning_rate": 4.858220360348187e-07,
+      "loss": 1.1176,
+      "step": 1309
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 12.112739562988281,
+      "learning_rate": 4.808311502573976e-07,
+      "loss": 1.1444,
+      "step": 1310
+    },
+    {
+      "epoch": 2.622,
+      "grad_norm": 10.57923412322998,
+      "learning_rate": 4.758647376699033e-07,
+      "loss": 1.1199,
+      "step": 1311
+    },
+    {
+      "epoch": 2.624,
+      "grad_norm": 11.439313888549805,
+      "learning_rate": 4.709228251675357e-07,
+      "loss": 1.0995,
+      "step": 1312
+    },
+    {
+      "epoch": 2.626,
+      "grad_norm": 11.128246307373047,
+      "learning_rate": 4.6600543951281995e-07,
+      "loss": 1.1438,
+      "step": 1313
+    },
+    {
+      "epoch": 2.628,
+      "grad_norm": 12.107537269592285,
+      "learning_rate": 4.6111260733545714e-07,
+      "loss": 1.0989,
+      "step": 1314
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 11.53816032409668,
+      "learning_rate": 4.562443551321788e-07,
+      "loss": 1.1305,
+      "step": 1315
+    },
+    {
+      "epoch": 2.632,
+      "grad_norm": 10.835124015808105,
+      "learning_rate": 4.514007092666084e-07,
+      "loss": 1.1451,
+      "step": 1316
+    },
+    {
+      "epoch": 2.634,
+      "grad_norm": 11.95799446105957,
+      "learning_rate": 4.4658169596911493e-07,
+      "loss": 1.1823,
+      "step": 1317
+    },
+    {
+      "epoch": 2.636,
+      "grad_norm": 10.443716049194336,
+      "learning_rate": 4.417873413366702e-07,
+      "loss": 1.0862,
+      "step": 1318
+    },
+    {
+      "epoch": 2.638,
+      "grad_norm": 12.407042503356934,
+      "learning_rate": 4.370176713327118e-07,
+      "loss": 1.0715,
+      "step": 1319
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 12.52112102508545,
+      "learning_rate": 4.322727117869951e-07,
+      "loss": 1.1408,
+      "step": 1320
+    },
+    {
+      "epoch": 2.642,
+      "grad_norm": 12.11574935913086,
+      "learning_rate": 4.275524883954657e-07,
+      "loss": 1.1602,
+      "step": 1321
+    },
+    {
+      "epoch": 2.644,
+      "grad_norm": 11.21030330657959,
+      "learning_rate": 4.228570267201049e-07,
+      "loss": 1.0917,
+      "step": 1322
+    },
+    {
+      "epoch": 2.646,
+      "grad_norm": 10.748331069946289,
+      "learning_rate": 4.1818635218880186e-07,
+      "loss": 1.0612,
+      "step": 1323
+    },
+    {
+      "epoch": 2.648,
+      "grad_norm": 12.968511581420898,
+      "learning_rate": 4.1354049009521504e-07,
+      "loss": 1.1723,
+      "step": 1324
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 13.352898597717285,
+      "learning_rate": 4.089194655986306e-07,
+      "loss": 1.127,
+      "step": 1325
+    },
+    {
+      "epoch": 2.652,
+      "grad_norm": 10.655840873718262,
+      "learning_rate": 4.043233037238281e-07,
+      "loss": 1.0859,
+      "step": 1326
+    },
+    {
+      "epoch": 2.654,
+      "grad_norm": 10.203536033630371,
+      "learning_rate": 3.99752029360948e-07,
+      "loss": 1.1438,
+      "step": 1327
+    },
+    {
+      "epoch": 2.656,
+      "grad_norm": 12.646231651306152,
+      "learning_rate": 3.9520566726535367e-07,
+      "loss": 1.2127,
+      "step": 1328
+    },
+    {
+      "epoch": 2.658,
+      "grad_norm": 12.770830154418945,
+      "learning_rate": 3.90684242057498e-07,
+      "loss": 1.0962,
+      "step": 1329
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 11.451815605163574,
+      "learning_rate": 3.8618777822278854e-07,
+      "loss": 1.0883,
+      "step": 1330
+    },
+    {
+      "epoch": 2.662,
+      "grad_norm": 11.632091522216797,
+      "learning_rate": 3.8171630011145877e-07,
+      "loss": 1.1129,
+      "step": 1331
+    },
+    {
+      "epoch": 2.664,
+      "grad_norm": 11.543761253356934,
+      "learning_rate": 3.772698319384349e-07,
+      "loss": 1.0953,
+      "step": 1332
+    },
+    {
+      "epoch": 2.666,
+      "grad_norm": 12.997357368469238,
+      "learning_rate": 3.728483977831998e-07,
+      "loss": 1.1444,
+      "step": 1333
+    },
+    {
+      "epoch": 2.668,
+      "grad_norm": 11.470213890075684,
+      "learning_rate": 3.684520215896703e-07,
+      "loss": 1.1423,
+      "step": 1334
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 10.749590873718262,
+      "learning_rate": 3.6408072716606346e-07,
+      "loss": 1.1258,
+      "step": 1335
+    },
+    {
+      "epoch": 2.672,
+      "grad_norm": 10.158390045166016,
+      "learning_rate": 3.597345381847656e-07,
+      "loss": 1.032,
+      "step": 1336
+    },
+    {
+      "epoch": 2.674,
+      "grad_norm": 10.492109298706055,
+      "learning_rate": 3.554134781822094e-07,
+      "loss": 1.1678,
+      "step": 1337
+    },
+    {
+      "epoch": 2.676,
+      "grad_norm": 12.122913360595703,
+      "learning_rate": 3.511175705587433e-07,
+      "loss": 1.1777,
+      "step": 1338
+    },
+    {
+      "epoch": 2.678,
+      "grad_norm": 10.95935344696045,
+      "learning_rate": 3.468468385785023e-07,
+      "loss": 1.1732,
+      "step": 1339
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 12.425519943237305,
+      "learning_rate": 3.426013053692878e-07,
+      "loss": 1.1649,
+      "step": 1340
+    },
+    {
+      "epoch": 2.682,
+      "grad_norm": 11.766524314880371,
+      "learning_rate": 3.3838099392243915e-07,
+      "loss": 1.0,
+      "step": 1341
+    },
+    {
+      "epoch": 2.684,
+      "grad_norm": 12.33736515045166,
+      "learning_rate": 3.341859270927067e-07,
+      "loss": 1.148,
+      "step": 1342
+    },
+    {
+      "epoch": 2.686,
+      "grad_norm": 10.662920951843262,
+      "learning_rate": 3.30016127598134e-07,
+      "loss": 1.0533,
+      "step": 1343
+    },
+    {
+      "epoch": 2.6879999999999997,
+      "grad_norm": 11.27861213684082,
+      "learning_rate": 3.258716180199278e-07,
+      "loss": 1.114,
+      "step": 1344
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 11.482233047485352,
+      "learning_rate": 3.2175242080234314e-07,
+      "loss": 1.1744,
+      "step": 1345
+    },
+    {
+      "epoch": 2.692,
+      "grad_norm": 11.544189453125,
+      "learning_rate": 3.1765855825255543e-07,
+      "loss": 1.0746,
+      "step": 1346
+    },
+    {
+      "epoch": 2.694,
+      "grad_norm": 13.263728141784668,
+      "learning_rate": 3.135900525405428e-07,
+      "loss": 1.1174,
+      "step": 1347
+    },
+    {
+      "epoch": 2.6959999999999997,
+      "grad_norm": 11.041204452514648,
+      "learning_rate": 3.0954692569896585e-07,
+      "loss": 1.1725,
+      "step": 1348
+    },
+    {
+      "epoch": 2.698,
+      "grad_norm": 11.216727256774902,
+      "learning_rate": 3.055291996230492e-07,
+      "loss": 1.1185,
+      "step": 1349
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 11.795129776000977,
+      "learning_rate": 3.015368960704584e-07,
+      "loss": 1.1468,
+      "step": 1350
+    },
+    {
+      "epoch": 2.702,
+      "grad_norm": 10.980323791503906,
+      "learning_rate": 2.975700366611883e-07,
+      "loss": 1.0928,
+      "step": 1351
+    },
+    {
+      "epoch": 2.7039999999999997,
+      "grad_norm": 12.678966522216797,
+      "learning_rate": 2.9362864287744266e-07,
+      "loss": 1.1624,
+      "step": 1352
+    },
+    {
+      "epoch": 2.706,
+      "grad_norm": 12.718985557556152,
+      "learning_rate": 2.8971273606351656e-07,
+      "loss": 1.141,
+      "step": 1353
+    },
+    {
+      "epoch": 2.708,
+      "grad_norm": 11.222257614135742,
+      "learning_rate": 2.858223374256841e-07,
+      "loss": 1.1179,
+      "step": 1354
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 10.49862003326416,
+      "learning_rate": 2.819574680320825e-07,
+      "loss": 1.1192,
+      "step": 1355
+    },
+    {
+      "epoch": 2.7119999999999997,
+      "grad_norm": 13.092155456542969,
+      "learning_rate": 2.7811814881259503e-07,
+      "loss": 1.1709,
+      "step": 1356
+    },
+    {
+      "epoch": 2.714,
+      "grad_norm": 10.232337951660156,
+      "learning_rate": 2.743044005587425e-07,
+      "loss": 1.0475,
+      "step": 1357
+    },
+    {
+      "epoch": 2.716,
+      "grad_norm": 12.870683670043945,
+      "learning_rate": 2.705162439235648e-07,
+      "loss": 1.0407,
+      "step": 1358
+    },
+    {
+      "epoch": 2.718,
+      "grad_norm": 11.157443046569824,
+      "learning_rate": 2.6675369942151864e-07,
+      "loss": 1.1319,
+      "step": 1359
+    },
+    {
+      "epoch": 2.7199999999999998,
+      "grad_norm": 10.935739517211914,
+      "learning_rate": 2.63016787428354e-07,
+      "loss": 1.1432,
+      "step": 1360
+    },
+    {
+      "epoch": 2.722,
+      "grad_norm": 12.044547080993652,
+      "learning_rate": 2.593055281810125e-07,
+      "loss": 1.1435,
+      "step": 1361
+    },
+    {
+      "epoch": 2.724,
+      "grad_norm": 10.721463203430176,
+      "learning_rate": 2.556199417775174e-07,
+      "loss": 1.1649,
+      "step": 1362
+    },
+    {
+      "epoch": 2.726,
+      "grad_norm": 10.794017791748047,
+      "learning_rate": 2.519600481768597e-07,
+      "loss": 1.1491,
+      "step": 1363
+    },
+    {
+      "epoch": 2.7279999999999998,
+      "grad_norm": 11.180771827697754,
+      "learning_rate": 2.483258671988942e-07,
+      "loss": 1.0898,
+      "step": 1364
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 11.11221694946289,
+      "learning_rate": 2.447174185242324e-07,
+      "loss": 1.1323,
+      "step": 1365
+    },
+    {
+      "epoch": 2.732,
+      "grad_norm": 11.723511695861816,
+      "learning_rate": 2.4113472169413176e-07,
+      "loss": 1.2018,
+      "step": 1366
+    },
+    {
+      "epoch": 2.734,
+      "grad_norm": 12.620880126953125,
+      "learning_rate": 2.37577796110397e-07,
+      "loss": 1.1411,
+      "step": 1367
+    },
+    {
+      "epoch": 2.7359999999999998,
+      "grad_norm": 11.710165023803711,
+      "learning_rate": 2.3404666103526542e-07,
+      "loss": 1.0482,
+      "step": 1368
+    },
+    {
+      "epoch": 2.738,
+      "grad_norm": 11.673789024353027,
+      "learning_rate": 2.3054133559131163e-07,
+      "loss": 1.2405,
+      "step": 1369
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 11.153477668762207,
+      "learning_rate": 2.2706183876134047e-07,
+      "loss": 1.1506,
+      "step": 1370
+    },
+    {
+      "epoch": 2.742,
+      "grad_norm": 11.211380958557129,
+      "learning_rate": 2.2360818938828189e-07,
+      "loss": 1.046,
+      "step": 1371
+    },
+    {
+      "epoch": 2.7439999999999998,
+      "grad_norm": 10.854473114013672,
+      "learning_rate": 2.2018040617509174e-07,
+      "loss": 1.0861,
+      "step": 1372
+    },
+    {
+      "epoch": 2.746,
+      "grad_norm": 11.484323501586914,
+      "learning_rate": 2.167785076846518e-07,
+      "loss": 1.0931,
+      "step": 1373
+    },
+    {
+      "epoch": 2.748,
+      "grad_norm": 12.862896919250488,
+      "learning_rate": 2.134025123396638e-07,
+      "loss": 1.215,
+      "step": 1374
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 12.15230941772461,
+      "learning_rate": 2.1005243842255552e-07,
+      "loss": 1.167,
+      "step": 1375
+    },
+    {
+      "epoch": 2.752,
+      "grad_norm": 10.922401428222656,
+      "learning_rate": 2.0672830407537925e-07,
+      "loss": 1.1432,
+      "step": 1376
+    },
+    {
+      "epoch": 2.754,
+      "grad_norm": 12.105512619018555,
+      "learning_rate": 2.0343012729971244e-07,
+      "loss": 1.1478,
+      "step": 1377
+    },
+    {
+      "epoch": 2.7560000000000002,
+      "grad_norm": 10.566096305847168,
+      "learning_rate": 2.0015792595656225e-07,
+      "loss": 1.1038,
+      "step": 1378
+    },
+    {
+      "epoch": 2.758,
+      "grad_norm": 12.602991104125977,
+      "learning_rate": 1.9691171776626882e-07,
+      "loss": 1.2035,
+      "step": 1379
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 10.561416625976562,
+      "learning_rate": 1.9369152030840553e-07,
+      "loss": 1.1399,
+      "step": 1380
+    },
+    {
+      "epoch": 2.762,
+      "grad_norm": 12.064254760742188,
+      "learning_rate": 1.904973510216912e-07,
+      "loss": 1.2584,
+      "step": 1381
+    },
+    {
+      "epoch": 2.7640000000000002,
+      "grad_norm": 10.637829780578613,
+      "learning_rate": 1.873292272038868e-07,
+      "loss": 1.0919,
+      "step": 1382
+    },
+    {
+      "epoch": 2.766,
+      "grad_norm": 11.943845748901367,
+      "learning_rate": 1.841871660117095e-07,
+      "loss": 1.1898,
+      "step": 1383
+    },
+    {
+      "epoch": 2.768,
+      "grad_norm": 11.168000221252441,
+      "learning_rate": 1.8107118446073492e-07,
+      "loss": 1.1557,
+      "step": 1384
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 10.773613929748535,
+      "learning_rate": 1.779812994253055e-07,
+      "loss": 1.1016,
+      "step": 1385
+    },
+    {
+      "epoch": 2.7720000000000002,
+      "grad_norm": 10.496236801147461,
+      "learning_rate": 1.7491752763844294e-07,
+      "loss": 1.0975,
+      "step": 1386
+    },
+    {
+      "epoch": 2.774,
+      "grad_norm": 11.72801399230957,
+      "learning_rate": 1.7187988569175307e-07,
+      "loss": 1.0915,
+      "step": 1387
+    },
+    {
+      "epoch": 2.776,
+      "grad_norm": 11.977339744567871,
+      "learning_rate": 1.688683900353366e-07,
+      "loss": 1.1441,
+      "step": 1388
+    },
+    {
+      "epoch": 2.778,
+      "grad_norm": 10.9048490524292,
+      "learning_rate": 1.6588305697770313e-07,
+      "loss": 1.0953,
+      "step": 1389
+    },
+    {
+      "epoch": 2.7800000000000002,
+      "grad_norm": 10.485379219055176,
+      "learning_rate": 1.6292390268568103e-07,
+      "loss": 1.0981,
+      "step": 1390
+    },
+    {
+      "epoch": 2.782,
+      "grad_norm": 10.757002830505371,
+      "learning_rate": 1.5999094318432662e-07,
+      "loss": 1.1695,
+      "step": 1391
+    },
+    {
+      "epoch": 2.784,
+      "grad_norm": 11.401092529296875,
+      "learning_rate": 1.5708419435684463e-07,
+      "loss": 1.0509,
+      "step": 1392
+    },
+    {
+      "epoch": 2.786,
+      "grad_norm": 11.38297176361084,
+      "learning_rate": 1.5420367194449448e-07,
+      "loss": 1.0742,
+      "step": 1393
+    },
+    {
+      "epoch": 2.7880000000000003,
+      "grad_norm": 10.606424331665039,
+      "learning_rate": 1.5134939154651196e-07,
+      "loss": 1.1533,
+      "step": 1394
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 11.047826766967773,
+      "learning_rate": 1.4852136862001766e-07,
+      "loss": 1.15,
+      "step": 1395
+    },
+    {
+      "epoch": 2.792,
+      "grad_norm": 11.345439910888672,
+      "learning_rate": 1.4571961847993977e-07,
+      "loss": 1.2146,
+      "step": 1396
+    },
+    {
+      "epoch": 2.794,
+      "grad_norm": 11.464273452758789,
+      "learning_rate": 1.4294415629892756e-07,
+      "loss": 1.0851,
+      "step": 1397
+    },
+    {
+      "epoch": 2.7960000000000003,
+      "grad_norm": 11.990900039672852,
+      "learning_rate": 1.4019499710726913e-07,
+      "loss": 1.1197,
+      "step": 1398
+    },
+    {
+      "epoch": 2.798,
+      "grad_norm": 11.901836395263672,
+      "learning_rate": 1.374721557928116e-07,
+      "loss": 1.2454,
+      "step": 1399
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 10.661705017089844,
+      "learning_rate": 1.3477564710088097e-07,
+      "loss": 1.0545,
+      "step": 1400
+    },
+    {
+      "epoch": 2.802,
+      "grad_norm": 11.410598754882812,
+      "learning_rate": 1.3210548563419857e-07,
+      "loss": 1.0326,
+      "step": 1401
+    },
+    {
+      "epoch": 2.8040000000000003,
+      "grad_norm": 11.102482795715332,
+      "learning_rate": 1.294616858528064e-07,
+      "loss": 1.1085,
+      "step": 1402
+    },
+    {
+      "epoch": 2.806,
+      "grad_norm": 10.520735740661621,
+      "learning_rate": 1.268442620739868e-07,
+      "loss": 1.1905,
+      "step": 1403
+    },
+    {
+      "epoch": 2.808,
+      "grad_norm": 11.018781661987305,
+      "learning_rate": 1.2425322847218368e-07,
+      "loss": 1.1735,
+      "step": 1404
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 12.908723831176758,
+      "learning_rate": 1.2168859907892904e-07,
+      "loss": 1.1788,
+      "step": 1405
+    },
+    {
+      "epoch": 2.8120000000000003,
+      "grad_norm": 11.897324562072754,
+      "learning_rate": 1.1915038778276212e-07,
+      "loss": 1.093,
+      "step": 1406
+    },
+    {
+      "epoch": 2.814,
+      "grad_norm": 11.710148811340332,
+      "learning_rate": 1.166386083291604e-07,
+      "loss": 1.1192,
+      "step": 1407
+    },
+    {
+      "epoch": 2.816,
+      "grad_norm": 11.686257362365723,
+      "learning_rate": 1.1415327432046041e-07,
+      "loss": 1.0824,
+      "step": 1408
+    },
+    {
+      "epoch": 2.818,
+      "grad_norm": 11.252178192138672,
+      "learning_rate": 1.1169439921578485e-07,
+      "loss": 1.0743,
+      "step": 1409
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 10.60265064239502,
+      "learning_rate": 1.0926199633097156e-07,
+      "loss": 1.0421,
+      "step": 1410
+    },
+    {
+      "epoch": 2.822,
+      "grad_norm": 12.824352264404297,
+      "learning_rate": 1.0685607883850035e-07,
+      "loss": 1.1632,
+      "step": 1411
+    },
+    {
+      "epoch": 2.824,
+      "grad_norm": 11.57745361328125,
+      "learning_rate": 1.044766597674196e-07,
+      "loss": 1.1655,
+      "step": 1412
+    },
+    {
+      "epoch": 2.826,
+      "grad_norm": 11.447077751159668,
+      "learning_rate": 1.0212375200327973e-07,
+      "loss": 1.0834,
+      "step": 1413
+    },
+    {
+      "epoch": 2.828,
+      "grad_norm": 12.608105659484863,
+      "learning_rate": 9.979736828806096e-08,
+      "loss": 1.1674,
+      "step": 1414
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 11.511768341064453,
+      "learning_rate": 9.749752122010347e-08,
+      "loss": 1.2345,
+      "step": 1415
+    },
+    {
+      "epoch": 2.832,
+      "grad_norm": 10.86443042755127,
+      "learning_rate": 9.522422325404234e-08,
+      "loss": 1.0663,
+      "step": 1416
+    },
+    {
+      "epoch": 2.834,
+      "grad_norm": 11.792540550231934,
+      "learning_rate": 9.297748670073658e-08,
+      "loss": 1.0739,
+      "step": 1417
+    },
+    {
+      "epoch": 2.836,
+      "grad_norm": 12.60987377166748,
+      "learning_rate": 9.075732372720414e-08,
+      "loss": 1.121,
+      "step": 1418
+    },
+    {
+      "epoch": 2.838,
+      "grad_norm": 11.70394515991211,
+      "learning_rate": 8.856374635655696e-08,
+      "loss": 1.1565,
+      "step": 1419
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 11.06529426574707,
+      "learning_rate": 8.639676646793382e-08,
+      "loss": 1.0428,
+      "step": 1420
+    },
+    {
+      "epoch": 2.842,
+      "grad_norm": 11.53507137298584,
+      "learning_rate": 8.425639579643763e-08,
+      "loss": 1.094,
+      "step": 1421
+    },
+    {
+      "epoch": 2.844,
+      "grad_norm": 10.740865707397461,
+      "learning_rate": 8.214264593307097e-08,
+      "loss": 1.0513,
+      "step": 1422
+    },
+    {
+      "epoch": 2.846,
+      "grad_norm": 11.995918273925781,
+      "learning_rate": 8.00555283246729e-08,
+      "loss": 1.1279,
+      "step": 1423
+    },
+    {
+      "epoch": 2.848,
+      "grad_norm": 12.916094779968262,
+      "learning_rate": 7.799505427386001e-08,
+      "loss": 1.1227,
+      "step": 1424
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 10.555110931396484,
+      "learning_rate": 7.59612349389599e-08,
+      "loss": 1.0907,
+      "step": 1425
+    },
+    {
+      "epoch": 2.852,
+      "grad_norm": 11.170307159423828,
+      "learning_rate": 7.395408133395509e-08,
+      "loss": 1.1291,
+      "step": 1426
+    },
+    {
+      "epoch": 2.854,
+      "grad_norm": 10.48733901977539,
+      "learning_rate": 7.197360432842359e-08,
+      "loss": 1.1168,
+      "step": 1427
+    },
+    {
+      "epoch": 2.856,
+      "grad_norm": 11.542524337768555,
+      "learning_rate": 7.001981464747565e-08,
+      "loss": 1.1142,
+      "step": 1428
+    },
+    {
+      "epoch": 2.858,
+      "grad_norm": 10.778676986694336,
+      "learning_rate": 6.809272287169988e-08,
+      "loss": 1.1537,
+      "step": 1429
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 10.768854141235352,
+      "learning_rate": 6.61923394371039e-08,
+      "loss": 1.0235,
+      "step": 1430
+    },
+    {
+      "epoch": 2.862,
+      "grad_norm": 10.812145233154297,
+      "learning_rate": 6.431867463506047e-08,
+      "loss": 1.1195,
+      "step": 1431
+    },
+    {
+      "epoch": 2.864,
+      "grad_norm": 10.739622116088867,
+      "learning_rate": 6.247173861224753e-08,
+      "loss": 1.1552,
+      "step": 1432
+    },
+    {
+      "epoch": 2.866,
+      "grad_norm": 10.252985954284668,
+      "learning_rate": 6.065154137059603e-08,
+      "loss": 1.1009,
+      "step": 1433
+    },
+    {
+      "epoch": 2.868,
+      "grad_norm": 11.594807624816895,
+      "learning_rate": 5.8858092767236084e-08,
+      "loss": 1.1601,
+      "step": 1434
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 10.98749828338623,
+      "learning_rate": 5.709140251444201e-08,
+      "loss": 1.1708,
+      "step": 1435
+    },
+    {
+      "epoch": 2.872,
+      "grad_norm": 11.422292709350586,
+      "learning_rate": 5.535148017958014e-08,
+      "loss": 1.0827,
+      "step": 1436
+    },
+    {
+      "epoch": 2.874,
+      "grad_norm": 10.309944152832031,
+      "learning_rate": 5.363833518505834e-08,
+      "loss": 1.1094,
+      "step": 1437
+    },
+    {
+      "epoch": 2.876,
+      "grad_norm": 11.423945426940918,
+      "learning_rate": 5.19519768082738e-08,
+      "loss": 1.2409,
+      "step": 1438
+    },
+    {
+      "epoch": 2.878,
+      "grad_norm": 10.608428001403809,
+      "learning_rate": 5.029241418156139e-08,
+      "loss": 1.1391,
+      "step": 1439
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 13.38857650756836,
+      "learning_rate": 4.865965629214819e-08,
+      "loss": 1.1201,
+      "step": 1440
+    },
+    {
+      "epoch": 2.882,
+      "grad_norm": 11.526823997497559,
+      "learning_rate": 4.7053711982101294e-08,
+      "loss": 1.145,
+      "step": 1441
+    },
+    {
+      "epoch": 2.884,
+      "grad_norm": 11.45752239227295,
+      "learning_rate": 4.5474589948280026e-08,
+      "loss": 1.2218,
+      "step": 1442
+    },
+    {
+      "epoch": 2.886,
+      "grad_norm": 10.46944808959961,
+      "learning_rate": 4.392229874229159e-08,
+      "loss": 1.2481,
+      "step": 1443
+    },
+    {
+      "epoch": 2.888,
+      "grad_norm": 12.171918869018555,
+      "learning_rate": 4.2396846770441644e-08,
+      "loss": 1.1381,
+      "step": 1444
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 11.57470703125,
+      "learning_rate": 4.0898242293691546e-08,
+      "loss": 1.1412,
+      "step": 1445
+    },
+    {
+      "epoch": 2.892,
+      "grad_norm": 11.620773315429688,
+      "learning_rate": 3.9426493427611177e-08,
+      "loss": 1.1664,
+      "step": 1446
+    },
+    {
+      "epoch": 2.894,
+      "grad_norm": 12.99191951751709,
+      "learning_rate": 3.7981608142335644e-08,
+      "loss": 1.2922,
+      "step": 1447
+    },
+    {
+      "epoch": 2.896,
+      "grad_norm": 11.910186767578125,
+      "learning_rate": 3.65635942625242e-08,
+      "loss": 1.285,
+      "step": 1448
+    },
+    {
+      "epoch": 2.898,
+      "grad_norm": 11.08059310913086,
+      "learning_rate": 3.517245946731529e-08,
+      "loss": 1.1577,
+      "step": 1449
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 11.138588905334473,
+      "learning_rate": 3.3808211290284886e-08,
+      "loss": 1.1573,
+      "step": 1450
+    },
+    {
+      "epoch": 2.902,
+      "grad_norm": 10.384156227111816,
+      "learning_rate": 3.247085711940878e-08,
+      "loss": 1.124,
+      "step": 1451
+    },
+    {
+      "epoch": 2.904,
+      "grad_norm": 12.75297737121582,
+      "learning_rate": 3.1160404197018155e-08,
+      "loss": 1.1535,
+      "step": 1452
+    },
+    {
+      "epoch": 2.906,
+      "grad_norm": 12.249570846557617,
+      "learning_rate": 2.9876859619764606e-08,
+      "loss": 1.1892,
+      "step": 1453
+    },
+    {
+      "epoch": 2.908,
+      "grad_norm": 12.755828857421875,
+      "learning_rate": 2.8620230338578526e-08,
+      "loss": 1.0924,
+      "step": 1454
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 13.399191856384277,
+      "learning_rate": 2.7390523158633552e-08,
+      "loss": 1.0451,
+      "step": 1455
+    },
+    {
+      "epoch": 2.912,
+      "grad_norm": 11.16951847076416,
+      "learning_rate": 2.6187744739308297e-08,
+      "loss": 1.0882,
+      "step": 1456
+    },
+    {
+      "epoch": 2.914,
+      "grad_norm": 10.711827278137207,
+      "learning_rate": 2.501190159415079e-08,
+      "loss": 1.164,
+      "step": 1457
+    },
+    {
+      "epoch": 2.916,
+      "grad_norm": 10.825516700744629,
+      "learning_rate": 2.386300009084408e-08,
+      "loss": 1.0708,
+      "step": 1458
+    },
+    {
+      "epoch": 2.918,
+      "grad_norm": 11.371330261230469,
+      "learning_rate": 2.27410464511707e-08,
+      "loss": 1.1395,
+      "step": 1459
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 11.252603530883789,
+      "learning_rate": 2.1646046750978255e-08,
+      "loss": 1.1699,
+      "step": 1460
+    },
+    {
+      "epoch": 2.922,
+      "grad_norm": 11.264942169189453,
+      "learning_rate": 2.057800692014833e-08,
+      "loss": 1.0813,
+      "step": 1461
+    },
+    {
+      "epoch": 2.924,
+      "grad_norm": 11.598838806152344,
+      "learning_rate": 1.953693274256374e-08,
+      "loss": 1.124,
+      "step": 1462
+    },
+    {
+      "epoch": 2.926,
+      "grad_norm": 10.306319236755371,
+      "learning_rate": 1.8522829856076895e-08,
+      "loss": 1.0401,
+      "step": 1463
+    },
+    {
+      "epoch": 2.928,
+      "grad_norm": 11.464588165283203,
+      "learning_rate": 1.753570375247815e-08,
+      "loss": 1.154,
+      "step": 1464
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 11.513689994812012,
+      "learning_rate": 1.657555977746972e-08,
+      "loss": 1.0962,
+      "step": 1465
+    },
+    {
+      "epoch": 2.932,
+      "grad_norm": 11.152294158935547,
+      "learning_rate": 1.5642403130632367e-08,
+      "loss": 1.119,
+      "step": 1466
+    },
+    {
+      "epoch": 2.934,
+      "grad_norm": 11.795382499694824,
+      "learning_rate": 1.4736238865398766e-08,
+      "loss": 1.1594,
+      "step": 1467
+    },
+    {
+      "epoch": 2.936,
+      "grad_norm": 13.539538383483887,
+      "learning_rate": 1.3857071889029073e-08,
+      "loss": 1.1061,
+      "step": 1468
+    },
+    {
+      "epoch": 2.9379999999999997,
+      "grad_norm": 10.830578804016113,
+      "learning_rate": 1.3004906962578723e-08,
+      "loss": 1.114,
+      "step": 1469
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 12.558431625366211,
+      "learning_rate": 1.2179748700879013e-08,
+      "loss": 1.2487,
+      "step": 1470
+    },
+    {
+      "epoch": 2.942,
+      "grad_norm": 11.865883827209473,
+      "learning_rate": 1.1381601572505452e-08,
+      "loss": 1.0191,
+      "step": 1471
+    },
+    {
+      "epoch": 2.944,
+      "grad_norm": 10.461407661437988,
+      "learning_rate": 1.0610469899760001e-08,
+      "loss": 1.1221,
+      "step": 1472
+    },
+    {
+      "epoch": 2.9459999999999997,
+      "grad_norm": 11.201630592346191,
+      "learning_rate": 9.866357858642206e-09,
+      "loss": 0.9862,
+      "step": 1473
+    },
+    {
+      "epoch": 2.948,
+      "grad_norm": 11.596713066101074,
+      "learning_rate": 9.14926947883088e-09,
+      "loss": 1.2098,
+      "step": 1474
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 10.908271789550781,
+      "learning_rate": 8.459208643659122e-09,
+      "loss": 1.1182,
+      "step": 1475
+    },
+    {
+      "epoch": 2.952,
+      "grad_norm": 11.701807975769043,
+      "learning_rate": 7.796179090094891e-09,
+      "loss": 1.2094,
+      "step": 1476
+    },
+    {
+      "epoch": 2.9539999999999997,
+      "grad_norm": 13.19516658782959,
+      "learning_rate": 7.160184408721571e-09,
+      "loss": 1.1626,
+      "step": 1477
+    },
+    {
+      "epoch": 2.956,
+      "grad_norm": 12.156759262084961,
+      "learning_rate": 6.551228043715218e-09,
+      "loss": 1.1171,
+      "step": 1478
+    },
+    {
+      "epoch": 2.958,
+      "grad_norm": 10.075406074523926,
+      "learning_rate": 5.969313292830126e-09,
+      "loss": 0.9876,
+      "step": 1479
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 11.928035736083984,
+      "learning_rate": 5.414443307377171e-09,
+      "loss": 1.1166,
+      "step": 1480
+    },
+    {
+      "epoch": 2.9619999999999997,
+      "grad_norm": 10.68159294128418,
+      "learning_rate": 4.8866210922110525e-09,
+      "loss": 1.058,
+      "step": 1481
+    },
+    {
+      "epoch": 2.964,
+      "grad_norm": 11.216858863830566,
+      "learning_rate": 4.385849505708084e-09,
+      "loss": 1.1111,
+      "step": 1482
+    },
+    {
+      "epoch": 2.966,
+      "grad_norm": 12.130516052246094,
+      "learning_rate": 3.912131259757313e-09,
+      "loss": 1.1494,
+      "step": 1483
+    },
+    {
+      "epoch": 2.968,
+      "grad_norm": 10.646476745605469,
+      "learning_rate": 3.4654689197405335e-09,
+      "loss": 1.0957,
+      "step": 1484
+    },
+    {
+      "epoch": 2.9699999999999998,
+      "grad_norm": 11.538399696350098,
+      "learning_rate": 3.0458649045211897e-09,
+      "loss": 1.1852,
+      "step": 1485
+    },
+    {
+      "epoch": 2.972,
+      "grad_norm": 12.342727661132812,
+      "learning_rate": 2.6533214864310485e-09,
+      "loss": 1.1064,
+      "step": 1486
+    },
+    {
+      "epoch": 2.974,
+      "grad_norm": 11.315055847167969,
+      "learning_rate": 2.287840791256324e-09,
+      "loss": 1.0906,
+      "step": 1487
+    },
+    {
+      "epoch": 2.976,
+      "grad_norm": 12.045461654663086,
+      "learning_rate": 1.9494247982282386e-09,
+      "loss": 1.163,
+      "step": 1488
+    },
+    {
+      "epoch": 2.9779999999999998,
+      "grad_norm": 10.301689147949219,
+      "learning_rate": 1.638075340010814e-09,
+      "loss": 1.1269,
+      "step": 1489
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 11.46061897277832,
+      "learning_rate": 1.3537941026914302e-09,
+      "loss": 1.0705,
+      "step": 1490
+    },
+    {
+      "epoch": 2.982,
+      "grad_norm": 10.900529861450195,
+      "learning_rate": 1.096582625772502e-09,
+      "loss": 1.099,
+      "step": 1491
+    },
+    {
+      "epoch": 2.984,
+      "grad_norm": 10.490682601928711,
+      "learning_rate": 8.664423021614854e-10,
+      "loss": 1.0517,
+      "step": 1492
+    },
+    {
+      "epoch": 2.9859999999999998,
+      "grad_norm": 11.499110221862793,
+      "learning_rate": 6.633743781642166e-10,
+      "loss": 1.177,
+      "step": 1493
+    },
+    {
+      "epoch": 2.988,
+      "grad_norm": 11.795324325561523,
+      "learning_rate": 4.87379953478806e-10,
+      "loss": 1.1048,
+      "step": 1494
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 11.121499061584473,
+      "learning_rate": 3.384599811889766e-10,
+      "loss": 1.1005,
+      "step": 1495
+    },
+    {
+      "epoch": 2.992,
+      "grad_norm": 11.869853973388672,
+      "learning_rate": 2.1661526775795804e-10,
+      "loss": 1.1684,
+      "step": 1496
+    },
+    {
+      "epoch": 2.9939999999999998,
+      "grad_norm": 13.046727180480957,
+      "learning_rate": 1.2184647302626585e-10,
+      "loss": 1.0746,
+      "step": 1497
+    },
+    {
+      "epoch": 2.996,
+      "grad_norm": 12.80275821685791,
+      "learning_rate": 5.4154110206150465e-11,
+      "loss": 1.1424,
+      "step": 1498
+    },
+    {
+      "epoch": 2.998,
+      "grad_norm": 13.787501335144043,
+      "learning_rate": 1.3538545881042198e-11,
+      "loss": 1.172,
+      "step": 1499
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 11.100687026977539,
+      "learning_rate": 0.0,
+      "loss": 1.1813,
+      "step": 1500
+    },
+    {
+      "epoch": 3.0,
+      "step": 1500,
+      "total_flos": 3.656864303572582e+16,
+      "train_loss": 0.0,
+      "train_runtime": 15.8388,
+      "train_samples_per_second": 9091.618,
+      "train_steps_per_second": 94.704
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1500,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "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": 3.656864303572582e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}