diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,42028 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.21271221366941864,
+  "eval_steps": 1000,
+  "global_step": 12000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.7726017805784885e-05,
+      "grad_norm": 636.0,
+      "learning_rate": 2.9411764705882356e-07,
+      "loss": 14.7402,
+      "step": 1
+    },
+    {
+      "epoch": 3.545203561156977e-05,
+      "grad_norm": 620.0,
+      "learning_rate": 5.882352941176471e-07,
+      "loss": 13.8226,
+      "step": 2
+    },
+    {
+      "epoch": 7.090407122313954e-05,
+      "grad_norm": 648.0,
+      "learning_rate": 1.1764705882352942e-06,
+      "loss": 14.3719,
+      "step": 4
+    },
+    {
+      "epoch": 0.00010635610683470932,
+      "grad_norm": 556.0,
+      "learning_rate": 1.7647058823529412e-06,
+      "loss": 13.7975,
+      "step": 6
+    },
+    {
+      "epoch": 0.00014180814244627908,
+      "grad_norm": 672.0,
+      "learning_rate": 2.3529411764705885e-06,
+      "loss": 14.2378,
+      "step": 8
+    },
+    {
+      "epoch": 0.00017726017805784887,
+      "grad_norm": 552.0,
+      "learning_rate": 2.9411764705882355e-06,
+      "loss": 13.8208,
+      "step": 10
+    },
+    {
+      "epoch": 0.00021271221366941863,
+      "grad_norm": 492.0,
+      "learning_rate": 3.5294117647058825e-06,
+      "loss": 12.2904,
+      "step": 12
+    },
+    {
+      "epoch": 0.0002481642492809884,
+      "grad_norm": 450.0,
+      "learning_rate": 4.11764705882353e-06,
+      "loss": 12.2658,
+      "step": 14
+    },
+    {
+      "epoch": 0.00028361628489255816,
+      "grad_norm": 338.0,
+      "learning_rate": 4.705882352941177e-06,
+      "loss": 10.4614,
+      "step": 16
+    },
+    {
+      "epoch": 0.00031906832050412795,
+      "grad_norm": 209.0,
+      "learning_rate": 5.294117647058824e-06,
+      "loss": 8.4051,
+      "step": 18
+    },
+    {
+      "epoch": 0.00035452035611569774,
+      "grad_norm": 125.5,
+      "learning_rate": 5.882352941176471e-06,
+      "loss": 7.5553,
+      "step": 20
+    },
+    {
+      "epoch": 0.00038997239172726747,
+      "grad_norm": 90.0,
+      "learning_rate": 6.470588235294119e-06,
+      "loss": 7.013,
+      "step": 22
+    },
+    {
+      "epoch": 0.00042542442733883726,
+      "grad_norm": 81.0,
+      "learning_rate": 7.058823529411765e-06,
+      "loss": 6.7415,
+      "step": 24
+    },
+    {
+      "epoch": 0.00046087646295040705,
+      "grad_norm": 103.0,
+      "learning_rate": 7.647058823529413e-06,
+      "loss": 6.0122,
+      "step": 26
+    },
+    {
+      "epoch": 0.0004963284985619768,
+      "grad_norm": 108.0,
+      "learning_rate": 8.23529411764706e-06,
+      "loss": 5.1164,
+      "step": 28
+    },
+    {
+      "epoch": 0.0005317805341735466,
+      "grad_norm": 68.5,
+      "learning_rate": 8.823529411764707e-06,
+      "loss": 4.0865,
+      "step": 30
+    },
+    {
+      "epoch": 0.0005672325697851163,
+      "grad_norm": 70.0,
+      "learning_rate": 9.411764705882354e-06,
+      "loss": 3.4705,
+      "step": 32
+    },
+    {
+      "epoch": 0.0006026846053966862,
+      "grad_norm": 56.0,
+      "learning_rate": 1e-05,
+      "loss": 2.6308,
+      "step": 34
+    },
+    {
+      "epoch": 0.0006381366410082559,
+      "grad_norm": 19.5,
+      "learning_rate": 1.0588235294117648e-05,
+      "loss": 2.1415,
+      "step": 36
+    },
+    {
+      "epoch": 0.0006735886766198256,
+      "grad_norm": 13.4375,
+      "learning_rate": 1.1176470588235295e-05,
+      "loss": 1.9987,
+      "step": 38
+    },
+    {
+      "epoch": 0.0007090407122313955,
+      "grad_norm": 13.625,
+      "learning_rate": 1.1764705882352942e-05,
+      "loss": 1.8561,
+      "step": 40
+    },
+    {
+      "epoch": 0.0007444927478429652,
+      "grad_norm": 10.875,
+      "learning_rate": 1.2352941176470589e-05,
+      "loss": 1.8186,
+      "step": 42
+    },
+    {
+      "epoch": 0.0007799447834545349,
+      "grad_norm": 6.8125,
+      "learning_rate": 1.2941176470588238e-05,
+      "loss": 1.7927,
+      "step": 44
+    },
+    {
+      "epoch": 0.0008153968190661048,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.3529411764705883e-05,
+      "loss": 1.6379,
+      "step": 46
+    },
+    {
+      "epoch": 0.0008508488546776745,
+      "grad_norm": 5.5,
+      "learning_rate": 1.411764705882353e-05,
+      "loss": 1.5964,
+      "step": 48
+    },
+    {
+      "epoch": 0.0008863008902892443,
+      "grad_norm": 5.3125,
+      "learning_rate": 1.4705882352941177e-05,
+      "loss": 1.4906,
+      "step": 50
+    },
+    {
+      "epoch": 0.0009217529259008141,
+      "grad_norm": 5.59375,
+      "learning_rate": 1.5294117647058826e-05,
+      "loss": 1.4529,
+      "step": 52
+    },
+    {
+      "epoch": 0.0009572049615123838,
+      "grad_norm": 4.46875,
+      "learning_rate": 1.588235294117647e-05,
+      "loss": 1.4715,
+      "step": 54
+    },
+    {
+      "epoch": 0.0009926569971239537,
+      "grad_norm": 4.46875,
+      "learning_rate": 1.647058823529412e-05,
+      "loss": 1.3749,
+      "step": 56
+    },
+    {
+      "epoch": 0.0010281090327355234,
+      "grad_norm": 3.0625,
+      "learning_rate": 1.7058823529411767e-05,
+      "loss": 1.3041,
+      "step": 58
+    },
+    {
+      "epoch": 0.0010635610683470932,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.7647058823529414e-05,
+      "loss": 1.2544,
+      "step": 60
+    },
+    {
+      "epoch": 0.0010990131039586629,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.8235294117647057e-05,
+      "loss": 1.3137,
+      "step": 62
+    },
+    {
+      "epoch": 0.0011344651395702326,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.8823529411764708e-05,
+      "loss": 1.2781,
+      "step": 64
+    },
+    {
+      "epoch": 0.0011699171751818026,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.9411764705882355e-05,
+      "loss": 1.2188,
+      "step": 66
+    },
+    {
+      "epoch": 0.0012053692107933723,
+      "grad_norm": 3.984375,
+      "learning_rate": 2e-05,
+      "loss": 1.2637,
+      "step": 68
+    },
+    {
+      "epoch": 0.001240821246404942,
+      "grad_norm": 3.625,
+      "learning_rate": 2.058823529411765e-05,
+      "loss": 1.2316,
+      "step": 70
+    },
+    {
+      "epoch": 0.0012762732820165118,
+      "grad_norm": 4.0625,
+      "learning_rate": 2.1176470588235296e-05,
+      "loss": 1.1779,
+      "step": 72
+    },
+    {
+      "epoch": 0.0013117253176280815,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.1764705882352943e-05,
+      "loss": 1.214,
+      "step": 74
+    },
+    {
+      "epoch": 0.0013471773532396513,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.235294117647059e-05,
+      "loss": 1.3211,
+      "step": 76
+    },
+    {
+      "epoch": 0.0013826293888512212,
+      "grad_norm": 3.765625,
+      "learning_rate": 2.2941176470588237e-05,
+      "loss": 1.1991,
+      "step": 78
+    },
+    {
+      "epoch": 0.001418081424462791,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.3529411764705884e-05,
+      "loss": 1.234,
+      "step": 80
+    },
+    {
+      "epoch": 0.0014535334600743607,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.411764705882353e-05,
+      "loss": 1.2325,
+      "step": 82
+    },
+    {
+      "epoch": 0.0014889854956859304,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.4705882352941178e-05,
+      "loss": 1.2462,
+      "step": 84
+    },
+    {
+      "epoch": 0.0015244375312975001,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.5294117647058825e-05,
+      "loss": 1.1721,
+      "step": 86
+    },
+    {
+      "epoch": 0.0015598895669090699,
+      "grad_norm": 5.9375,
+      "learning_rate": 2.5882352941176475e-05,
+      "loss": 1.2239,
+      "step": 88
+    },
+    {
+      "epoch": 0.0015953416025206398,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.647058823529412e-05,
+      "loss": 1.1905,
+      "step": 90
+    },
+    {
+      "epoch": 0.0016307936381322096,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.7058823529411766e-05,
+      "loss": 1.1956,
+      "step": 92
+    },
+    {
+      "epoch": 0.0016662456737437793,
+      "grad_norm": 3.71875,
+      "learning_rate": 2.7647058823529416e-05,
+      "loss": 1.1939,
+      "step": 94
+    },
+    {
+      "epoch": 0.001701697709355349,
+      "grad_norm": 3.625,
+      "learning_rate": 2.823529411764706e-05,
+      "loss": 1.1435,
+      "step": 96
+    },
+    {
+      "epoch": 0.0017371497449669188,
+      "grad_norm": 4.4375,
+      "learning_rate": 2.8823529411764703e-05,
+      "loss": 1.1985,
+      "step": 98
+    },
+    {
+      "epoch": 0.0017726017805784885,
+      "grad_norm": 3.890625,
+      "learning_rate": 2.9411764705882354e-05,
+      "loss": 1.1163,
+      "step": 100
+    },
+    {
+      "epoch": 0.0018080538161900585,
+      "grad_norm": 3.921875,
+      "learning_rate": 3e-05,
+      "loss": 1.1596,
+      "step": 102
+    },
+    {
+      "epoch": 0.0018435058518016282,
+      "grad_norm": 3.828125,
+      "learning_rate": 3.058823529411765e-05,
+      "loss": 1.2044,
+      "step": 104
+    },
+    {
+      "epoch": 0.001878957887413198,
+      "grad_norm": 3.390625,
+      "learning_rate": 3.11764705882353e-05,
+      "loss": 1.177,
+      "step": 106
+    },
+    {
+      "epoch": 0.0019144099230247677,
+      "grad_norm": 4.40625,
+      "learning_rate": 3.176470588235294e-05,
+      "loss": 1.099,
+      "step": 108
+    },
+    {
+      "epoch": 0.0019498619586363374,
+      "grad_norm": 3.703125,
+      "learning_rate": 3.235294117647059e-05,
+      "loss": 1.0945,
+      "step": 110
+    },
+    {
+      "epoch": 0.0019853139942479074,
+      "grad_norm": 3.84375,
+      "learning_rate": 3.294117647058824e-05,
+      "loss": 1.1515,
+      "step": 112
+    },
+    {
+      "epoch": 0.002020766029859477,
+      "grad_norm": 4.0,
+      "learning_rate": 3.352941176470588e-05,
+      "loss": 1.1436,
+      "step": 114
+    },
+    {
+      "epoch": 0.002056218065471047,
+      "grad_norm": 4.5,
+      "learning_rate": 3.411764705882353e-05,
+      "loss": 1.1481,
+      "step": 116
+    },
+    {
+      "epoch": 0.0020916701010826166,
+      "grad_norm": 3.6875,
+      "learning_rate": 3.470588235294118e-05,
+      "loss": 1.1345,
+      "step": 118
+    },
+    {
+      "epoch": 0.0021271221366941863,
+      "grad_norm": 4.75,
+      "learning_rate": 3.529411764705883e-05,
+      "loss": 1.1136,
+      "step": 120
+    },
+    {
+      "epoch": 0.002162574172305756,
+      "grad_norm": 3.203125,
+      "learning_rate": 3.5882352941176474e-05,
+      "loss": 1.1009,
+      "step": 122
+    },
+    {
+      "epoch": 0.0021980262079173258,
+      "grad_norm": 3.640625,
+      "learning_rate": 3.6470588235294114e-05,
+      "loss": 1.1125,
+      "step": 124
+    },
+    {
+      "epoch": 0.0022334782435288955,
+      "grad_norm": 3.5625,
+      "learning_rate": 3.705882352941177e-05,
+      "loss": 1.1209,
+      "step": 126
+    },
+    {
+      "epoch": 0.0022689302791404652,
+      "grad_norm": 4.5,
+      "learning_rate": 3.7647058823529415e-05,
+      "loss": 1.1285,
+      "step": 128
+    },
+    {
+      "epoch": 0.002304382314752035,
+      "grad_norm": 4.0,
+      "learning_rate": 3.8235294117647055e-05,
+      "loss": 1.1407,
+      "step": 130
+    },
+    {
+      "epoch": 0.002339834350363605,
+      "grad_norm": 3.78125,
+      "learning_rate": 3.882352941176471e-05,
+      "loss": 1.1664,
+      "step": 132
+    },
+    {
+      "epoch": 0.002375286385975175,
+      "grad_norm": 3.53125,
+      "learning_rate": 3.9411764705882356e-05,
+      "loss": 1.1043,
+      "step": 134
+    },
+    {
+      "epoch": 0.0024107384215867446,
+      "grad_norm": 3.71875,
+      "learning_rate": 4e-05,
+      "loss": 1.1204,
+      "step": 136
+    },
+    {
+      "epoch": 0.0024461904571983144,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.058823529411765e-05,
+      "loss": 1.0994,
+      "step": 138
+    },
+    {
+      "epoch": 0.002481642492809884,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.11764705882353e-05,
+      "loss": 1.0434,
+      "step": 140
+    },
+    {
+      "epoch": 0.002517094528421454,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.1764705882352944e-05,
+      "loss": 1.0621,
+      "step": 142
+    },
+    {
+      "epoch": 0.0025525465640330236,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.235294117647059e-05,
+      "loss": 1.1062,
+      "step": 144
+    },
+    {
+      "epoch": 0.0025879985996445933,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.294117647058823e-05,
+      "loss": 1.0814,
+      "step": 146
+    },
+    {
+      "epoch": 0.002623450635256163,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.3529411764705885e-05,
+      "loss": 1.0785,
+      "step": 148
+    },
+    {
+      "epoch": 0.0026589026708677328,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.411764705882353e-05,
+      "loss": 1.0745,
+      "step": 150
+    },
+    {
+      "epoch": 0.0026943547064793025,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.470588235294118e-05,
+      "loss": 1.1046,
+      "step": 152
+    },
+    {
+      "epoch": 0.0027298067420908722,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.5294117647058826e-05,
+      "loss": 1.0708,
+      "step": 154
+    },
+    {
+      "epoch": 0.0027652587777024424,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.588235294117647e-05,
+      "loss": 1.1357,
+      "step": 156
+    },
+    {
+      "epoch": 0.002800710813314012,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.647058823529412e-05,
+      "loss": 1.1327,
+      "step": 158
+    },
+    {
+      "epoch": 0.002836162848925582,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.705882352941177e-05,
+      "loss": 1.06,
+      "step": 160
+    },
+    {
+      "epoch": 0.0028716148845371516,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.7647058823529414e-05,
+      "loss": 1.0476,
+      "step": 162
+    },
+    {
+      "epoch": 0.0029070669201487214,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.823529411764706e-05,
+      "loss": 1.0745,
+      "step": 164
+    },
+    {
+      "epoch": 0.002942518955760291,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.882352941176471e-05,
+      "loss": 1.0779,
+      "step": 166
+    },
+    {
+      "epoch": 0.002977970991371861,
+      "grad_norm": 3.90625,
+      "learning_rate": 4.9411764705882355e-05,
+      "loss": 1.1239,
+      "step": 168
+    },
+    {
+      "epoch": 0.0030134230269834306,
+      "grad_norm": 3.59375,
+      "learning_rate": 5e-05,
+      "loss": 1.0765,
+      "step": 170
+    },
+    {
+      "epoch": 0.0030488750625950003,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.999999984400261e-05,
+      "loss": 1.0344,
+      "step": 172
+    },
+    {
+      "epoch": 0.00308432709820657,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.999999937601042e-05,
+      "loss": 1.1249,
+      "step": 174
+    },
+    {
+      "epoch": 0.0031197791338181398,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.9999998596023444e-05,
+      "loss": 1.0789,
+      "step": 176
+    },
+    {
+      "epoch": 0.0031552311694297095,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.9999997504041694e-05,
+      "loss": 1.0495,
+      "step": 178
+    },
+    {
+      "epoch": 0.0031906832050412797,
+      "grad_norm": 4.28125,
+      "learning_rate": 4.999999610006519e-05,
+      "loss": 1.0348,
+      "step": 180
+    },
+    {
+      "epoch": 0.0032261352406528494,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.999999438409393e-05,
+      "loss": 1.0471,
+      "step": 182
+    },
+    {
+      "epoch": 0.003261587276264419,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.9999992356127956e-05,
+      "loss": 1.0496,
+      "step": 184
+    },
+    {
+      "epoch": 0.003297039311875989,
+      "grad_norm": 4.125,
+      "learning_rate": 4.9999990016167286e-05,
+      "loss": 1.0611,
+      "step": 186
+    },
+    {
+      "epoch": 0.0033324913474875586,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.999998736421194e-05,
+      "loss": 1.0915,
+      "step": 188
+    },
+    {
+      "epoch": 0.0033679433830991284,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.999998440026197e-05,
+      "loss": 1.0573,
+      "step": 190
+    },
+    {
+      "epoch": 0.003403395418710698,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.999998112431738e-05,
+      "loss": 1.0416,
+      "step": 192
+    },
+    {
+      "epoch": 0.003438847454322268,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.999997753637825e-05,
+      "loss": 1.0162,
+      "step": 194
+    },
+    {
+      "epoch": 0.0034742994899338376,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.999997363644461e-05,
+      "loss": 1.004,
+      "step": 196
+    },
+    {
+      "epoch": 0.0035097515255454073,
+      "grad_norm": 4.65625,
+      "learning_rate": 4.99999694245165e-05,
+      "loss": 1.0658,
+      "step": 198
+    },
+    {
+      "epoch": 0.003545203561156977,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.9999964900593975e-05,
+      "loss": 0.9414,
+      "step": 200
+    },
+    {
+      "epoch": 0.0035806555967685468,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.9999960064677104e-05,
+      "loss": 1.0537,
+      "step": 202
+    },
+    {
+      "epoch": 0.003616107632380117,
+      "grad_norm": 4.8125,
+      "learning_rate": 4.9999954916765934e-05,
+      "loss": 1.0669,
+      "step": 204
+    },
+    {
+      "epoch": 0.0036515596679916867,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.999994945686053e-05,
+      "loss": 0.9318,
+      "step": 206
+    },
+    {
+      "epoch": 0.0036870117036032564,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.999994368496097e-05,
+      "loss": 1.0402,
+      "step": 208
+    },
+    {
+      "epoch": 0.003722463739214826,
+      "grad_norm": 4.40625,
+      "learning_rate": 4.999993760106732e-05,
+      "loss": 1.0629,
+      "step": 210
+    },
+    {
+      "epoch": 0.003757915774826396,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.999993120517965e-05,
+      "loss": 1.0532,
+      "step": 212
+    },
+    {
+      "epoch": 0.0037933678104379656,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.999992449729806e-05,
+      "loss": 1.0498,
+      "step": 214
+    },
+    {
+      "epoch": 0.0038288198460495354,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.9999917477422594e-05,
+      "loss": 0.9999,
+      "step": 216
+    },
+    {
+      "epoch": 0.003864271881661105,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.9999910145553386e-05,
+      "loss": 1.0786,
+      "step": 218
+    },
+    {
+      "epoch": 0.003899723917272675,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.99999025016905e-05,
+      "loss": 1.0175,
+      "step": 220
+    },
+    {
+      "epoch": 0.0039351759528842446,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.9999894545834034e-05,
+      "loss": 1.0597,
+      "step": 222
+    },
+    {
+      "epoch": 0.003970627988495815,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.99998862779841e-05,
+      "loss": 0.9901,
+      "step": 224
+    },
+    {
+      "epoch": 0.004006080024107384,
+      "grad_norm": 3.625,
+      "learning_rate": 4.9999877698140783e-05,
+      "loss": 1.053,
+      "step": 226
+    },
+    {
+      "epoch": 0.004041532059718954,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.999986880630421e-05,
+      "loss": 1.0277,
+      "step": 228
+    },
+    {
+      "epoch": 0.0040769840953305235,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.9999859602474474e-05,
+      "loss": 0.9884,
+      "step": 230
+    },
+    {
+      "epoch": 0.004112436130942094,
+      "grad_norm": 3.375,
+      "learning_rate": 4.999985008665169e-05,
+      "loss": 0.9866,
+      "step": 232
+    },
+    {
+      "epoch": 0.004147888166553663,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.9999840258835994e-05,
+      "loss": 0.976,
+      "step": 234
+    },
+    {
+      "epoch": 0.004183340202165233,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.99998301190275e-05,
+      "loss": 1.0003,
+      "step": 236
+    },
+    {
+      "epoch": 0.0042187922377768024,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.999981966722633e-05,
+      "loss": 1.0255,
+      "step": 238
+    },
+    {
+      "epoch": 0.004254244273388373,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.999980890343262e-05,
+      "loss": 1.0533,
+      "step": 240
+    },
+    {
+      "epoch": 0.004289696308999943,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.99997978276465e-05,
+      "loss": 1.0061,
+      "step": 242
+    },
+    {
+      "epoch": 0.004325148344611512,
+      "grad_norm": 3.625,
+      "learning_rate": 4.999978643986811e-05,
+      "loss": 1.0097,
+      "step": 244
+    },
+    {
+      "epoch": 0.004360600380223082,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.9999774740097597e-05,
+      "loss": 0.996,
+      "step": 246
+    },
+    {
+      "epoch": 0.0043960524158346516,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.9999762728335094e-05,
+      "loss": 0.9974,
+      "step": 248
+    },
+    {
+      "epoch": 0.004431504451446222,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.999975040458076e-05,
+      "loss": 0.9845,
+      "step": 250
+    },
+    {
+      "epoch": 0.004466956487057791,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.999973776883475e-05,
+      "loss": 1.0087,
+      "step": 252
+    },
+    {
+      "epoch": 0.004502408522669361,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.9999724821097226e-05,
+      "loss": 1.0533,
+      "step": 254
+    },
+    {
+      "epoch": 0.0045378605582809305,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.999971156136833e-05,
+      "loss": 1.0019,
+      "step": 256
+    },
+    {
+      "epoch": 0.004573312593892501,
+      "grad_norm": 4.40625,
+      "learning_rate": 4.999969798964825e-05,
+      "loss": 1.0433,
+      "step": 258
+    },
+    {
+      "epoch": 0.00460876462950407,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.999968410593715e-05,
+      "loss": 0.9849,
+      "step": 260
+    },
+    {
+      "epoch": 0.00464421666511564,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.999966991023519e-05,
+      "loss": 0.9957,
+      "step": 262
+    },
+    {
+      "epoch": 0.00467966870072721,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.999965540254257e-05,
+      "loss": 0.9741,
+      "step": 264
+    },
+    {
+      "epoch": 0.00471512073633878,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.999964058285944e-05,
+      "loss": 1.0172,
+      "step": 266
+    },
+    {
+      "epoch": 0.00475057277195035,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.9999625451186014e-05,
+      "loss": 1.0139,
+      "step": 268
+    },
+    {
+      "epoch": 0.004786024807561919,
+      "grad_norm": 3.875,
+      "learning_rate": 4.999961000752247e-05,
+      "loss": 1.0714,
+      "step": 270
+    },
+    {
+      "epoch": 0.004821476843173489,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.999959425186899e-05,
+      "loss": 1.044,
+      "step": 272
+    },
+    {
+      "epoch": 0.0048569288787850586,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.999957818422579e-05,
+      "loss": 0.9591,
+      "step": 274
+    },
+    {
+      "epoch": 0.004892380914396629,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.999956180459306e-05,
+      "loss": 0.9922,
+      "step": 276
+    },
+    {
+      "epoch": 0.004927832950008198,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.9999545112971004e-05,
+      "loss": 1.0256,
+      "step": 278
+    },
+    {
+      "epoch": 0.004963284985619768,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.9999528109359825e-05,
+      "loss": 1.0298,
+      "step": 280
+    },
+    {
+      "epoch": 0.0049987370212313375,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.999951079375976e-05,
+      "loss": 1.0022,
+      "step": 282
+    },
+    {
+      "epoch": 0.005034189056842908,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.9999493166170994e-05,
+      "loss": 1.0502,
+      "step": 284
+    },
+    {
+      "epoch": 0.005069641092454477,
+      "grad_norm": 3.625,
+      "learning_rate": 4.999947522659376e-05,
+      "loss": 1.0071,
+      "step": 286
+    },
+    {
+      "epoch": 0.005105093128066047,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.999945697502829e-05,
+      "loss": 1.0428,
+      "step": 288
+    },
+    {
+      "epoch": 0.005140545163677617,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.9999438411474794e-05,
+      "loss": 1.0481,
+      "step": 290
+    },
+    {
+      "epoch": 0.005175997199289187,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.999941953593352e-05,
+      "loss": 1.0106,
+      "step": 292
+    },
+    {
+      "epoch": 0.005211449234900757,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.99994003484047e-05,
+      "loss": 0.9525,
+      "step": 294
+    },
+    {
+      "epoch": 0.005246901270512326,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.9999380848888566e-05,
+      "loss": 0.9956,
+      "step": 296
+    },
+    {
+      "epoch": 0.005282353306123896,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.9999361037385366e-05,
+      "loss": 1.002,
+      "step": 298
+    },
+    {
+      "epoch": 0.0053178053417354655,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.9999340913895347e-05,
+      "loss": 1.0014,
+      "step": 300
+    },
+    {
+      "epoch": 0.005353257377347036,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.9999320478418766e-05,
+      "loss": 1.0703,
+      "step": 302
+    },
+    {
+      "epoch": 0.005388709412958605,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.999929973095586e-05,
+      "loss": 0.9922,
+      "step": 304
+    },
+    {
+      "epoch": 0.005424161448570175,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.999927867150691e-05,
+      "loss": 1.0016,
+      "step": 306
+    },
+    {
+      "epoch": 0.0054596134841817445,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.999925730007217e-05,
+      "loss": 1.0145,
+      "step": 308
+    },
+    {
+      "epoch": 0.005495065519793315,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.999923561665191e-05,
+      "loss": 1.0087,
+      "step": 310
+    },
+    {
+      "epoch": 0.005530517555404885,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.999921362124639e-05,
+      "loss": 0.9935,
+      "step": 312
+    },
+    {
+      "epoch": 0.005565969591016454,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.9999191313855884e-05,
+      "loss": 1.013,
+      "step": 314
+    },
+    {
+      "epoch": 0.005601421626628024,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.999916869448069e-05,
+      "loss": 0.9903,
+      "step": 316
+    },
+    {
+      "epoch": 0.005636873662239594,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.999914576312107e-05,
+      "loss": 0.9984,
+      "step": 318
+    },
+    {
+      "epoch": 0.005672325697851164,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.999912251977732e-05,
+      "loss": 1.0084,
+      "step": 320
+    },
+    {
+      "epoch": 0.005707777733462733,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.999909896444972e-05,
+      "loss": 0.9749,
+      "step": 322
+    },
+    {
+      "epoch": 0.005743229769074303,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.9999075097138584e-05,
+      "loss": 1.0102,
+      "step": 324
+    },
+    {
+      "epoch": 0.0057786818046858725,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.9999050917844194e-05,
+      "loss": 0.9708,
+      "step": 326
+    },
+    {
+      "epoch": 0.005814133840297443,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.9999026426566854e-05,
+      "loss": 1.0564,
+      "step": 328
+    },
+    {
+      "epoch": 0.005849585875909012,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.9999001623306876e-05,
+      "loss": 0.9852,
+      "step": 330
+    },
+    {
+      "epoch": 0.005885037911520582,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.999897650806455e-05,
+      "loss": 0.9472,
+      "step": 332
+    },
+    {
+      "epoch": 0.0059204899471321515,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.9998951080840214e-05,
+      "loss": 0.9868,
+      "step": 334
+    },
+    {
+      "epoch": 0.005955941982743722,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.9998925341634187e-05,
+      "loss": 1.0065,
+      "step": 336
+    },
+    {
+      "epoch": 0.005991394018355292,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.999889929044676e-05,
+      "loss": 1.008,
+      "step": 338
+    },
+    {
+      "epoch": 0.006026846053966861,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.9998872927278284e-05,
+      "loss": 1.0051,
+      "step": 340
+    },
+    {
+      "epoch": 0.006062298089578431,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.9998846252129086e-05,
+      "loss": 0.9748,
+      "step": 342
+    },
+    {
+      "epoch": 0.006097750125190001,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.999881926499949e-05,
+      "loss": 0.9838,
+      "step": 344
+    },
+    {
+      "epoch": 0.006133202160801571,
+      "grad_norm": 3.25,
+      "learning_rate": 4.999879196588983e-05,
+      "loss": 0.9938,
+      "step": 346
+    },
+    {
+      "epoch": 0.00616865419641314,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.999876435480045e-05,
+      "loss": 0.9663,
+      "step": 348
+    },
+    {
+      "epoch": 0.00620410623202471,
+      "grad_norm": 3.75,
+      "learning_rate": 4.9998736431731715e-05,
+      "loss": 0.963,
+      "step": 350
+    },
+    {
+      "epoch": 0.0062395582676362795,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.9998708196683945e-05,
+      "loss": 0.9871,
+      "step": 352
+    },
+    {
+      "epoch": 0.00627501030324785,
+      "grad_norm": 3.625,
+      "learning_rate": 4.999867964965751e-05,
+      "loss": 0.9627,
+      "step": 354
+    },
+    {
+      "epoch": 0.006310462338859419,
+      "grad_norm": 3.421875,
+      "learning_rate": 4.999865079065275e-05,
+      "loss": 1.0292,
+      "step": 356
+    },
+    {
+      "epoch": 0.006345914374470989,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.9998621619670036e-05,
+      "loss": 0.9694,
+      "step": 358
+    },
+    {
+      "epoch": 0.006381366410082559,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.9998592136709746e-05,
+      "loss": 0.9795,
+      "step": 360
+    },
+    {
+      "epoch": 0.006416818445694129,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.999856234177221e-05,
+      "loss": 0.9889,
+      "step": 362
+    },
+    {
+      "epoch": 0.006452270481305699,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.999853223485784e-05,
+      "loss": 0.9322,
+      "step": 364
+    },
+    {
+      "epoch": 0.006487722516917268,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.999850181596699e-05,
+      "loss": 1.0098,
+      "step": 366
+    },
+    {
+      "epoch": 0.006523174552528838,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.9998471085100045e-05,
+      "loss": 0.9977,
+      "step": 368
+    },
+    {
+      "epoch": 0.006558626588140408,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.999844004225739e-05,
+      "loss": 0.9989,
+      "step": 370
+    },
+    {
+      "epoch": 0.006594078623751978,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.9998408687439405e-05,
+      "loss": 0.9484,
+      "step": 372
+    },
+    {
+      "epoch": 0.006629530659363547,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.999837702064649e-05,
+      "loss": 0.9547,
+      "step": 374
+    },
+    {
+      "epoch": 0.006664982694975117,
+      "grad_norm": 3.5,
+      "learning_rate": 4.999834504187904e-05,
+      "loss": 1.0299,
+      "step": 376
+    },
+    {
+      "epoch": 0.0067004347305866865,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.999831275113744e-05,
+      "loss": 1.0,
+      "step": 378
+    },
+    {
+      "epoch": 0.006735886766198257,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.9998280148422104e-05,
+      "loss": 1.0383,
+      "step": 380
+    },
+    {
+      "epoch": 0.006771338801809826,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.999824723373345e-05,
+      "loss": 0.9557,
+      "step": 382
+    },
+    {
+      "epoch": 0.006806790837421396,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.999821400707186e-05,
+      "loss": 0.9847,
+      "step": 384
+    },
+    {
+      "epoch": 0.006842242873032966,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.9998180468437786e-05,
+      "loss": 0.9431,
+      "step": 386
+    },
+    {
+      "epoch": 0.006877694908644536,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.9998146617831616e-05,
+      "loss": 1.0244,
+      "step": 388
+    },
+    {
+      "epoch": 0.006913146944256106,
+      "grad_norm": 2.75,
+      "learning_rate": 4.999811245525378e-05,
+      "loss": 0.9302,
+      "step": 390
+    },
+    {
+      "epoch": 0.006948598979867675,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.9998077980704716e-05,
+      "loss": 0.9856,
+      "step": 392
+    },
+    {
+      "epoch": 0.006984051015479245,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.999804319418484e-05,
+      "loss": 0.978,
+      "step": 394
+    },
+    {
+      "epoch": 0.007019503051090815,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.999800809569459e-05,
+      "loss": 1.0063,
+      "step": 396
+    },
+    {
+      "epoch": 0.007054955086702385,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.999797268523441e-05,
+      "loss": 1.0083,
+      "step": 398
+    },
+    {
+      "epoch": 0.007090407122313954,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.9997936962804736e-05,
+      "loss": 1.0052,
+      "step": 400
+    },
+    {
+      "epoch": 0.007125859157925524,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.999790092840602e-05,
+      "loss": 0.9596,
+      "step": 402
+    },
+    {
+      "epoch": 0.0071613111935370935,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.99978645820387e-05,
+      "loss": 0.899,
+      "step": 404
+    },
+    {
+      "epoch": 0.007196763229148664,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.999782792370323e-05,
+      "loss": 0.9701,
+      "step": 406
+    },
+    {
+      "epoch": 0.007232215264760234,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.999779095340009e-05,
+      "loss": 0.9746,
+      "step": 408
+    },
+    {
+      "epoch": 0.007267667300371803,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.999775367112972e-05,
+      "loss": 1.0319,
+      "step": 410
+    },
+    {
+      "epoch": 0.007303119335983373,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.999771607689259e-05,
+      "loss": 0.9785,
+      "step": 412
+    },
+    {
+      "epoch": 0.007338571371594943,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.9997678170689164e-05,
+      "loss": 0.968,
+      "step": 414
+    },
+    {
+      "epoch": 0.007374023407206513,
+      "grad_norm": 3.671875,
+      "learning_rate": 4.999763995251993e-05,
+      "loss": 0.9797,
+      "step": 416
+    },
+    {
+      "epoch": 0.007409475442818082,
+      "grad_norm": 3.125,
+      "learning_rate": 4.999760142238535e-05,
+      "loss": 1.0253,
+      "step": 418
+    },
+    {
+      "epoch": 0.007444927478429652,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.9997562580285914e-05,
+      "loss": 0.9627,
+      "step": 420
+    },
+    {
+      "epoch": 0.007480379514041222,
+      "grad_norm": 3.125,
+      "learning_rate": 4.999752342622211e-05,
+      "loss": 0.9922,
+      "step": 422
+    },
+    {
+      "epoch": 0.007515831549652792,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.9997483960194405e-05,
+      "loss": 0.973,
+      "step": 424
+    },
+    {
+      "epoch": 0.007551283585264361,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.9997444182203316e-05,
+      "loss": 0.9356,
+      "step": 426
+    },
+    {
+      "epoch": 0.007586735620875931,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.9997404092249336e-05,
+      "loss": 0.9873,
+      "step": 428
+    },
+    {
+      "epoch": 0.0076221876564875005,
+      "grad_norm": 3.25,
+      "learning_rate": 4.999736369033295e-05,
+      "loss": 0.952,
+      "step": 430
+    },
+    {
+      "epoch": 0.007657639692099071,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.999732297645467e-05,
+      "loss": 0.9977,
+      "step": 432
+    },
+    {
+      "epoch": 0.007693091727710641,
+      "grad_norm": 3.578125,
+      "learning_rate": 4.999728195061502e-05,
+      "loss": 1.0054,
+      "step": 434
+    },
+    {
+      "epoch": 0.00772854376332221,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.999724061281448e-05,
+      "loss": 0.9507,
+      "step": 436
+    },
+    {
+      "epoch": 0.00776399579893378,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.9997198963053595e-05,
+      "loss": 0.9529,
+      "step": 438
+    },
+    {
+      "epoch": 0.00779944783454535,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.999715700133287e-05,
+      "loss": 1.0338,
+      "step": 440
+    },
+    {
+      "epoch": 0.00783489987015692,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.999711472765283e-05,
+      "loss": 0.9645,
+      "step": 442
+    },
+    {
+      "epoch": 0.007870351905768489,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.999707214201401e-05,
+      "loss": 0.9345,
+      "step": 444
+    },
+    {
+      "epoch": 0.007905803941380058,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.999702924441693e-05,
+      "loss": 0.9372,
+      "step": 446
+    },
+    {
+      "epoch": 0.00794125597699163,
+      "grad_norm": 3.578125,
+      "learning_rate": 4.999698603486214e-05,
+      "loss": 0.9688,
+      "step": 448
+    },
+    {
+      "epoch": 0.007976708012603199,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.999694251335016e-05,
+      "loss": 0.9641,
+      "step": 450
+    },
+    {
+      "epoch": 0.008012160048214768,
+      "grad_norm": 3.125,
+      "learning_rate": 4.9996898679881556e-05,
+      "loss": 0.9635,
+      "step": 452
+    },
+    {
+      "epoch": 0.008047612083826337,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.999685453445685e-05,
+      "loss": 0.9986,
+      "step": 454
+    },
+    {
+      "epoch": 0.008083064119437908,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.999681007707662e-05,
+      "loss": 0.9567,
+      "step": 456
+    },
+    {
+      "epoch": 0.008118516155049478,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.9996765307741394e-05,
+      "loss": 0.8916,
+      "step": 458
+    },
+    {
+      "epoch": 0.008153968190661047,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.999672022645175e-05,
+      "loss": 0.9889,
+      "step": 460
+    },
+    {
+      "epoch": 0.008189420226272618,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.999667483320825e-05,
+      "loss": 0.9913,
+      "step": 462
+    },
+    {
+      "epoch": 0.008224872261884187,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.999662912801144e-05,
+      "loss": 0.9699,
+      "step": 464
+    },
+    {
+      "epoch": 0.008260324297495757,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.999658311086191e-05,
+      "loss": 0.9943,
+      "step": 466
+    },
+    {
+      "epoch": 0.008295776333107326,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.9996536781760236e-05,
+      "loss": 0.9744,
+      "step": 468
+    },
+    {
+      "epoch": 0.008331228368718897,
+      "grad_norm": 3.625,
+      "learning_rate": 4.999649014070698e-05,
+      "loss": 0.9958,
+      "step": 470
+    },
+    {
+      "epoch": 0.008366680404330466,
+      "grad_norm": 2.875,
+      "learning_rate": 4.999644318770274e-05,
+      "loss": 0.9638,
+      "step": 472
+    },
+    {
+      "epoch": 0.008402132439942036,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.99963959227481e-05,
+      "loss": 0.9533,
+      "step": 474
+    },
+    {
+      "epoch": 0.008437584475553605,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.999634834584363e-05,
+      "loss": 0.9881,
+      "step": 476
+    },
+    {
+      "epoch": 0.008473036511165176,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.999630045698995e-05,
+      "loss": 0.9871,
+      "step": 478
+    },
+    {
+      "epoch": 0.008508488546776745,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.9996252256187646e-05,
+      "loss": 0.9489,
+      "step": 480
+    },
+    {
+      "epoch": 0.008543940582388315,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.999620374343732e-05,
+      "loss": 0.9673,
+      "step": 482
+    },
+    {
+      "epoch": 0.008579392617999886,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.999615491873957e-05,
+      "loss": 0.9671,
+      "step": 484
+    },
+    {
+      "epoch": 0.008614844653611455,
+      "grad_norm": 3.5,
+      "learning_rate": 4.999610578209502e-05,
+      "loss": 0.9086,
+      "step": 486
+    },
+    {
+      "epoch": 0.008650296689223024,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.9996056333504275e-05,
+      "loss": 0.9864,
+      "step": 488
+    },
+    {
+      "epoch": 0.008685748724834593,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.999600657296796e-05,
+      "loss": 0.9359,
+      "step": 490
+    },
+    {
+      "epoch": 0.008721200760446165,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.9995956500486676e-05,
+      "loss": 0.9322,
+      "step": 492
+    },
+    {
+      "epoch": 0.008756652796057734,
+      "grad_norm": 3.375,
+      "learning_rate": 4.999590611606107e-05,
+      "loss": 0.9402,
+      "step": 494
+    },
+    {
+      "epoch": 0.008792104831669303,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.9995855419691765e-05,
+      "loss": 0.9617,
+      "step": 496
+    },
+    {
+      "epoch": 0.008827556867280872,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.999580441137938e-05,
+      "loss": 0.9414,
+      "step": 498
+    },
+    {
+      "epoch": 0.008863008902892443,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.999575309112456e-05,
+      "loss": 0.9602,
+      "step": 500
+    },
+    {
+      "epoch": 0.008898460938504013,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.999570145892796e-05,
+      "loss": 0.9457,
+      "step": 502
+    },
+    {
+      "epoch": 0.008933912974115582,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.999564951479021e-05,
+      "loss": 1.0017,
+      "step": 504
+    },
+    {
+      "epoch": 0.008969365009727153,
+      "grad_norm": 3.421875,
+      "learning_rate": 4.9995597258711954e-05,
+      "loss": 0.9839,
+      "step": 506
+    },
+    {
+      "epoch": 0.009004817045338722,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.999554469069386e-05,
+      "loss": 0.9671,
+      "step": 508
+    },
+    {
+      "epoch": 0.009040269080950292,
+      "grad_norm": 3.125,
+      "learning_rate": 4.9995491810736564e-05,
+      "loss": 0.9471,
+      "step": 510
+    },
+    {
+      "epoch": 0.009075721116561861,
+      "grad_norm": 3.5,
+      "learning_rate": 4.999543861884074e-05,
+      "loss": 0.9322,
+      "step": 512
+    },
+    {
+      "epoch": 0.009111173152173432,
+      "grad_norm": 3.5,
+      "learning_rate": 4.9995385115007055e-05,
+      "loss": 0.9606,
+      "step": 514
+    },
+    {
+      "epoch": 0.009146625187785001,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.999533129923616e-05,
+      "loss": 0.965,
+      "step": 516
+    },
+    {
+      "epoch": 0.00918207722339657,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.999527717152874e-05,
+      "loss": 1.0181,
+      "step": 518
+    },
+    {
+      "epoch": 0.00921752925900814,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.999522273188547e-05,
+      "loss": 0.9654,
+      "step": 520
+    },
+    {
+      "epoch": 0.009252981294619711,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.9995167980307024e-05,
+      "loss": 1.0039,
+      "step": 522
+    },
+    {
+      "epoch": 0.00928843333023128,
+      "grad_norm": 3.0,
+      "learning_rate": 4.9995112916794084e-05,
+      "loss": 0.9443,
+      "step": 524
+    },
+    {
+      "epoch": 0.00932388536584285,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.999505754134734e-05,
+      "loss": 0.9385,
+      "step": 526
+    },
+    {
+      "epoch": 0.00935933740145442,
+      "grad_norm": 3.125,
+      "learning_rate": 4.999500185396749e-05,
+      "loss": 0.9158,
+      "step": 528
+    },
+    {
+      "epoch": 0.00939478943706599,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.999494585465523e-05,
+      "loss": 0.9583,
+      "step": 530
+    },
+    {
+      "epoch": 0.00943024147267756,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.999488954341124e-05,
+      "loss": 0.9829,
+      "step": 532
+    },
+    {
+      "epoch": 0.009465693508289129,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.999483292023624e-05,
+      "loss": 0.9174,
+      "step": 534
+    },
+    {
+      "epoch": 0.0095011455439007,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.9994775985130925e-05,
+      "loss": 0.9316,
+      "step": 536
+    },
+    {
+      "epoch": 0.009536597579512269,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.999471873809602e-05,
+      "loss": 1.006,
+      "step": 538
+    },
+    {
+      "epoch": 0.009572049615123838,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.999466117913223e-05,
+      "loss": 1.008,
+      "step": 540
+    },
+    {
+      "epoch": 0.009607501650735407,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.999460330824027e-05,
+      "loss": 0.9878,
+      "step": 542
+    },
+    {
+      "epoch": 0.009642953686346978,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.999454512542087e-05,
+      "loss": 0.9953,
+      "step": 544
+    },
+    {
+      "epoch": 0.009678405721958548,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.9994486630674744e-05,
+      "loss": 0.9758,
+      "step": 546
+    },
+    {
+      "epoch": 0.009713857757570117,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.999442782400264e-05,
+      "loss": 0.9526,
+      "step": 548
+    },
+    {
+      "epoch": 0.009749309793181686,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.999436870540528e-05,
+      "loss": 0.9034,
+      "step": 550
+    },
+    {
+      "epoch": 0.009784761828793257,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.999430927488341e-05,
+      "loss": 0.9539,
+      "step": 552
+    },
+    {
+      "epoch": 0.009820213864404827,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.999424953243776e-05,
+      "loss": 0.9947,
+      "step": 554
+    },
+    {
+      "epoch": 0.009855665900016396,
+      "grad_norm": 3.625,
+      "learning_rate": 4.999418947806908e-05,
+      "loss": 0.9475,
+      "step": 556
+    },
+    {
+      "epoch": 0.009891117935627967,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.999412911177813e-05,
+      "loss": 0.9541,
+      "step": 558
+    },
+    {
+      "epoch": 0.009926569971239536,
+      "grad_norm": 3.75,
+      "learning_rate": 4.999406843356564e-05,
+      "loss": 0.9163,
+      "step": 560
+    },
+    {
+      "epoch": 0.009962022006851106,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.99940074434324e-05,
+      "loss": 0.932,
+      "step": 562
+    },
+    {
+      "epoch": 0.009997474042462675,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.9993946141379145e-05,
+      "loss": 0.9848,
+      "step": 564
+    },
+    {
+      "epoch": 0.010032926078074246,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.999388452740664e-05,
+      "loss": 0.9937,
+      "step": 566
+    },
+    {
+      "epoch": 0.010068378113685815,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.999382260151567e-05,
+      "loss": 0.9689,
+      "step": 568
+    },
+    {
+      "epoch": 0.010103830149297385,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.9993760363706996e-05,
+      "loss": 0.9488,
+      "step": 570
+    },
+    {
+      "epoch": 0.010139282184908954,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.9993697813981404e-05,
+      "loss": 0.9778,
+      "step": 572
+    },
+    {
+      "epoch": 0.010174734220520525,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.999363495233966e-05,
+      "loss": 0.9597,
+      "step": 574
+    },
+    {
+      "epoch": 0.010210186256132094,
+      "grad_norm": 3.125,
+      "learning_rate": 4.9993571778782565e-05,
+      "loss": 0.968,
+      "step": 576
+    },
+    {
+      "epoch": 0.010245638291743664,
+      "grad_norm": 3.25,
+      "learning_rate": 4.9993508293310905e-05,
+      "loss": 0.9588,
+      "step": 578
+    },
+    {
+      "epoch": 0.010281090327355235,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.999344449592546e-05,
+      "loss": 0.9334,
+      "step": 580
+    },
+    {
+      "epoch": 0.010316542362966804,
+      "grad_norm": 3.25,
+      "learning_rate": 4.999338038662703e-05,
+      "loss": 0.975,
+      "step": 582
+    },
+    {
+      "epoch": 0.010351994398578373,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.999331596541643e-05,
+      "loss": 0.9033,
+      "step": 584
+    },
+    {
+      "epoch": 0.010387446434189943,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.999325123229444e-05,
+      "loss": 0.9773,
+      "step": 586
+    },
+    {
+      "epoch": 0.010422898469801514,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.9993186187261884e-05,
+      "loss": 0.9611,
+      "step": 588
+    },
+    {
+      "epoch": 0.010458350505413083,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.999312083031957e-05,
+      "loss": 0.9743,
+      "step": 590
+    },
+    {
+      "epoch": 0.010493802541024652,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.999305516146832e-05,
+      "loss": 0.9143,
+      "step": 592
+    },
+    {
+      "epoch": 0.010529254576636221,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.999298918070894e-05,
+      "loss": 0.9472,
+      "step": 594
+    },
+    {
+      "epoch": 0.010564706612247792,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.999292288804227e-05,
+      "loss": 0.934,
+      "step": 596
+    },
+    {
+      "epoch": 0.010600158647859362,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.999285628346912e-05,
+      "loss": 0.9666,
+      "step": 598
+    },
+    {
+      "epoch": 0.010635610683470931,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.999278936699033e-05,
+      "loss": 0.9855,
+      "step": 600
+    },
+    {
+      "epoch": 0.010671062719082502,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.999272213860674e-05,
+      "loss": 0.9246,
+      "step": 602
+    },
+    {
+      "epoch": 0.010706514754694071,
+      "grad_norm": 3.578125,
+      "learning_rate": 4.999265459831917e-05,
+      "loss": 0.9739,
+      "step": 604
+    },
+    {
+      "epoch": 0.01074196679030564,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.999258674612849e-05,
+      "loss": 0.993,
+      "step": 606
+    },
+    {
+      "epoch": 0.01077741882591721,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.999251858203553e-05,
+      "loss": 0.9682,
+      "step": 608
+    },
+    {
+      "epoch": 0.010812870861528781,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.9992450106041135e-05,
+      "loss": 0.9582,
+      "step": 610
+    },
+    {
+      "epoch": 0.01084832289714035,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.999238131814617e-05,
+      "loss": 0.9464,
+      "step": 612
+    },
+    {
+      "epoch": 0.01088377493275192,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.99923122183515e-05,
+      "loss": 0.908,
+      "step": 614
+    },
+    {
+      "epoch": 0.010919226968363489,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.999224280665798e-05,
+      "loss": 0.9571,
+      "step": 616
+    },
+    {
+      "epoch": 0.01095467900397506,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.9992173083066466e-05,
+      "loss": 0.9992,
+      "step": 618
+    },
+    {
+      "epoch": 0.01099013103958663,
+      "grad_norm": 3.0,
+      "learning_rate": 4.999210304757784e-05,
+      "loss": 0.9529,
+      "step": 620
+    },
+    {
+      "epoch": 0.011025583075198199,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.999203270019298e-05,
+      "loss": 0.9182,
+      "step": 622
+    },
+    {
+      "epoch": 0.01106103511080977,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.9991962040912756e-05,
+      "loss": 0.9034,
+      "step": 624
+    },
+    {
+      "epoch": 0.011096487146421339,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.999189106973804e-05,
+      "loss": 0.9601,
+      "step": 626
+    },
+    {
+      "epoch": 0.011131939182032908,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.999181978666974e-05,
+      "loss": 0.9431,
+      "step": 628
+    },
+    {
+      "epoch": 0.011167391217644478,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.999174819170873e-05,
+      "loss": 0.8976,
+      "step": 630
+    },
+    {
+      "epoch": 0.011202843253256049,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.999167628485591e-05,
+      "loss": 0.9586,
+      "step": 632
+    },
+    {
+      "epoch": 0.011238295288867618,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.999160406611218e-05,
+      "loss": 0.933,
+      "step": 634
+    },
+    {
+      "epoch": 0.011273747324479187,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.999153153547842e-05,
+      "loss": 0.9944,
+      "step": 636
+    },
+    {
+      "epoch": 0.011309199360090757,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.999145869295557e-05,
+      "loss": 0.9273,
+      "step": 638
+    },
+    {
+      "epoch": 0.011344651395702328,
+      "grad_norm": 3.25,
+      "learning_rate": 4.999138553854451e-05,
+      "loss": 0.8649,
+      "step": 640
+    },
+    {
+      "epoch": 0.011380103431313897,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.999131207224617e-05,
+      "loss": 0.9654,
+      "step": 642
+    },
+    {
+      "epoch": 0.011415555466925466,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.999123829406146e-05,
+      "loss": 0.9769,
+      "step": 644
+    },
+    {
+      "epoch": 0.011451007502537035,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.9991164203991295e-05,
+      "loss": 0.9016,
+      "step": 646
+    },
+    {
+      "epoch": 0.011486459538148606,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.999108980203662e-05,
+      "loss": 0.9064,
+      "step": 648
+    },
+    {
+      "epoch": 0.011521911573760176,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.999101508819833e-05,
+      "loss": 0.8923,
+      "step": 650
+    },
+    {
+      "epoch": 0.011557363609371745,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.9990940062477386e-05,
+      "loss": 0.9613,
+      "step": 652
+    },
+    {
+      "epoch": 0.011592815644983316,
+      "grad_norm": 3.421875,
+      "learning_rate": 4.999086472487472e-05,
+      "loss": 1.0074,
+      "step": 654
+    },
+    {
+      "epoch": 0.011628267680594885,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.9990789075391264e-05,
+      "loss": 0.9456,
+      "step": 656
+    },
+    {
+      "epoch": 0.011663719716206455,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.9990713114027966e-05,
+      "loss": 0.9861,
+      "step": 658
+    },
+    {
+      "epoch": 0.011699171751818024,
+      "grad_norm": 3.25,
+      "learning_rate": 4.9990636840785775e-05,
+      "loss": 0.9855,
+      "step": 660
+    },
+    {
+      "epoch": 0.011734623787429595,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.999056025566564e-05,
+      "loss": 0.9234,
+      "step": 662
+    },
+    {
+      "epoch": 0.011770075823041164,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.9990483358668514e-05,
+      "loss": 0.9737,
+      "step": 664
+    },
+    {
+      "epoch": 0.011805527858652734,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.999040614979536e-05,
+      "loss": 0.9562,
+      "step": 666
+    },
+    {
+      "epoch": 0.011840979894264303,
+      "grad_norm": 3.125,
+      "learning_rate": 4.999032862904715e-05,
+      "loss": 0.9459,
+      "step": 668
+    },
+    {
+      "epoch": 0.011876431929875874,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.999025079642484e-05,
+      "loss": 0.9167,
+      "step": 670
+    },
+    {
+      "epoch": 0.011911883965487443,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.999017265192941e-05,
+      "loss": 0.9068,
+      "step": 672
+    },
+    {
+      "epoch": 0.011947336001099013,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.999009419556182e-05,
+      "loss": 0.9414,
+      "step": 674
+    },
+    {
+      "epoch": 0.011982788036710584,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.999001542732307e-05,
+      "loss": 0.9365,
+      "step": 676
+    },
+    {
+      "epoch": 0.012018240072322153,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.9989936347214125e-05,
+      "loss": 0.9686,
+      "step": 678
+    },
+    {
+      "epoch": 0.012053692107933722,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.9989856955235985e-05,
+      "loss": 0.9942,
+      "step": 680
+    },
+    {
+      "epoch": 0.012089144143545292,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.998977725138964e-05,
+      "loss": 0.9682,
+      "step": 682
+    },
+    {
+      "epoch": 0.012124596179156863,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.998969723567607e-05,
+      "loss": 0.9719,
+      "step": 684
+    },
+    {
+      "epoch": 0.012160048214768432,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.998961690809628e-05,
+      "loss": 0.901,
+      "step": 686
+    },
+    {
+      "epoch": 0.012195500250380001,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.998953626865128e-05,
+      "loss": 0.9027,
+      "step": 688
+    },
+    {
+      "epoch": 0.01223095228599157,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.9989455317342076e-05,
+      "loss": 0.943,
+      "step": 690
+    },
+    {
+      "epoch": 0.012266404321603142,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.9989374054169676e-05,
+      "loss": 0.9789,
+      "step": 692
+    },
+    {
+      "epoch": 0.01230185635721471,
+      "grad_norm": 3.125,
+      "learning_rate": 4.998929247913509e-05,
+      "loss": 0.9397,
+      "step": 694
+    },
+    {
+      "epoch": 0.01233730839282628,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.998921059223933e-05,
+      "loss": 0.9447,
+      "step": 696
+    },
+    {
+      "epoch": 0.012372760428437851,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.9989128393483445e-05,
+      "loss": 0.946,
+      "step": 698
+    },
+    {
+      "epoch": 0.01240821246404942,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.9989045882868426e-05,
+      "loss": 0.9301,
+      "step": 700
+    },
+    {
+      "epoch": 0.01244366449966099,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.998896306039533e-05,
+      "loss": 0.9113,
+      "step": 702
+    },
+    {
+      "epoch": 0.012479116535272559,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.9988879926065174e-05,
+      "loss": 0.9572,
+      "step": 704
+    },
+    {
+      "epoch": 0.01251456857088413,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.998879647987901e-05,
+      "loss": 0.9825,
+      "step": 706
+    },
+    {
+      "epoch": 0.0125500206064957,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.998871272183786e-05,
+      "loss": 0.9766,
+      "step": 708
+    },
+    {
+      "epoch": 0.012585472642107269,
+      "grad_norm": 2.875,
+      "learning_rate": 4.998862865194278e-05,
+      "loss": 0.9208,
+      "step": 710
+    },
+    {
+      "epoch": 0.012620924677718838,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.998854427019483e-05,
+      "loss": 0.9375,
+      "step": 712
+    },
+    {
+      "epoch": 0.012656376713330409,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.998845957659504e-05,
+      "loss": 0.9579,
+      "step": 714
+    },
+    {
+      "epoch": 0.012691828748941978,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.9988374571144484e-05,
+      "loss": 0.9579,
+      "step": 716
+    },
+    {
+      "epoch": 0.012727280784553548,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.9988289253844214e-05,
+      "loss": 0.8884,
+      "step": 718
+    },
+    {
+      "epoch": 0.012762732820165119,
+      "grad_norm": 3.0,
+      "learning_rate": 4.998820362469531e-05,
+      "loss": 0.9132,
+      "step": 720
+    },
+    {
+      "epoch": 0.012798184855776688,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.998811768369882e-05,
+      "loss": 0.958,
+      "step": 722
+    },
+    {
+      "epoch": 0.012833636891388257,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.998803143085583e-05,
+      "loss": 0.9573,
+      "step": 724
+    },
+    {
+      "epoch": 0.012869088926999827,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.9987944866167405e-05,
+      "loss": 0.9226,
+      "step": 726
+    },
+    {
+      "epoch": 0.012904540962611398,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.998785798963464e-05,
+      "loss": 0.9214,
+      "step": 728
+    },
+    {
+      "epoch": 0.012939992998222967,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.9987770801258617e-05,
+      "loss": 0.9307,
+      "step": 730
+    },
+    {
+      "epoch": 0.012975445033834536,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.998768330104041e-05,
+      "loss": 0.9398,
+      "step": 732
+    },
+    {
+      "epoch": 0.013010897069446106,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.998759548898112e-05,
+      "loss": 0.9196,
+      "step": 734
+    },
+    {
+      "epoch": 0.013046349105057677,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.998750736508184e-05,
+      "loss": 0.9937,
+      "step": 736
+    },
+    {
+      "epoch": 0.013081801140669246,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.998741892934368e-05,
+      "loss": 0.963,
+      "step": 738
+    },
+    {
+      "epoch": 0.013117253176280815,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.998733018176774e-05,
+      "loss": 0.9119,
+      "step": 740
+    },
+    {
+      "epoch": 0.013152705211892384,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.9987241122355116e-05,
+      "loss": 0.861,
+      "step": 742
+    },
+    {
+      "epoch": 0.013188157247503956,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.9987151751106934e-05,
+      "loss": 0.9322,
+      "step": 744
+    },
+    {
+      "epoch": 0.013223609283115525,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.998706206802429e-05,
+      "loss": 0.9604,
+      "step": 746
+    },
+    {
+      "epoch": 0.013259061318727094,
+      "grad_norm": 3.125,
+      "learning_rate": 4.9986972073108326e-05,
+      "loss": 0.9062,
+      "step": 748
+    },
+    {
+      "epoch": 0.013294513354338665,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.998688176636015e-05,
+      "loss": 0.9155,
+      "step": 750
+    },
+    {
+      "epoch": 0.013329965389950234,
+      "grad_norm": 3.375,
+      "learning_rate": 4.99867911477809e-05,
+      "loss": 0.897,
+      "step": 752
+    },
+    {
+      "epoch": 0.013365417425561804,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.9986700217371694e-05,
+      "loss": 0.9338,
+      "step": 754
+    },
+    {
+      "epoch": 0.013400869461173373,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.998660897513367e-05,
+      "loss": 0.9416,
+      "step": 756
+    },
+    {
+      "epoch": 0.013436321496784944,
+      "grad_norm": 3.125,
+      "learning_rate": 4.9986517421067986e-05,
+      "loss": 0.9024,
+      "step": 758
+    },
+    {
+      "epoch": 0.013471773532396513,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.9986425555175766e-05,
+      "loss": 0.9484,
+      "step": 760
+    },
+    {
+      "epoch": 0.013507225568008083,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.998633337745815e-05,
+      "loss": 0.8972,
+      "step": 762
+    },
+    {
+      "epoch": 0.013542677603619652,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.99862408879163e-05,
+      "loss": 0.9608,
+      "step": 764
+    },
+    {
+      "epoch": 0.013578129639231223,
+      "grad_norm": 3.375,
+      "learning_rate": 4.998614808655137e-05,
+      "loss": 0.9107,
+      "step": 766
+    },
+    {
+      "epoch": 0.013613581674842792,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.9986054973364516e-05,
+      "loss": 0.9088,
+      "step": 768
+    },
+    {
+      "epoch": 0.013649033710454362,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.99859615483569e-05,
+      "loss": 0.9374,
+      "step": 770
+    },
+    {
+      "epoch": 0.013684485746065933,
+      "grad_norm": 3.125,
+      "learning_rate": 4.9985867811529685e-05,
+      "loss": 0.9202,
+      "step": 772
+    },
+    {
+      "epoch": 0.013719937781677502,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.998577376288405e-05,
+      "loss": 0.9546,
+      "step": 774
+    },
+    {
+      "epoch": 0.013755389817289071,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.998567940242116e-05,
+      "loss": 0.9569,
+      "step": 776
+    },
+    {
+      "epoch": 0.01379084185290064,
+      "grad_norm": 3.0,
+      "learning_rate": 4.9985584730142185e-05,
+      "loss": 0.9551,
+      "step": 778
+    },
+    {
+      "epoch": 0.013826293888512212,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.998548974604833e-05,
+      "loss": 0.9098,
+      "step": 780
+    },
+    {
+      "epoch": 0.013861745924123781,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.998539445014077e-05,
+      "loss": 0.9948,
+      "step": 782
+    },
+    {
+      "epoch": 0.01389719795973535,
+      "grad_norm": 3.125,
+      "learning_rate": 4.998529884242068e-05,
+      "loss": 0.9789,
+      "step": 784
+    },
+    {
+      "epoch": 0.01393264999534692,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.998520292288927e-05,
+      "loss": 0.9182,
+      "step": 786
+    },
+    {
+      "epoch": 0.01396810203095849,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.998510669154773e-05,
+      "loss": 0.9248,
+      "step": 788
+    },
+    {
+      "epoch": 0.01400355406657006,
+      "grad_norm": 3.125,
+      "learning_rate": 4.998501014839726e-05,
+      "loss": 0.9428,
+      "step": 790
+    },
+    {
+      "epoch": 0.01403900610218163,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.998491329343907e-05,
+      "loss": 0.9488,
+      "step": 792
+    },
+    {
+      "epoch": 0.0140744581377932,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.998481612667437e-05,
+      "loss": 0.9789,
+      "step": 794
+    },
+    {
+      "epoch": 0.01410991017340477,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.9984718648104365e-05,
+      "loss": 0.9332,
+      "step": 796
+    },
+    {
+      "epoch": 0.014145362209016339,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.998462085773027e-05,
+      "loss": 0.9086,
+      "step": 798
+    },
+    {
+      "epoch": 0.014180814244627908,
+      "grad_norm": 3.375,
+      "learning_rate": 4.998452275555332e-05,
+      "loss": 0.9826,
+      "step": 800
+    },
+    {
+      "epoch": 0.01421626628023948,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.9984424341574724e-05,
+      "loss": 0.903,
+      "step": 802
+    },
+    {
+      "epoch": 0.014251718315851048,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.998432561579572e-05,
+      "loss": 0.9578,
+      "step": 804
+    },
+    {
+      "epoch": 0.014287170351462618,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.998422657821753e-05,
+      "loss": 0.9535,
+      "step": 806
+    },
+    {
+      "epoch": 0.014322622387074187,
+      "grad_norm": 3.375,
+      "learning_rate": 4.99841272288414e-05,
+      "loss": 0.945,
+      "step": 808
+    },
+    {
+      "epoch": 0.014358074422685758,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.998402756766857e-05,
+      "loss": 0.9676,
+      "step": 810
+    },
+    {
+      "epoch": 0.014393526458297327,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.9983927594700275e-05,
+      "loss": 0.874,
+      "step": 812
+    },
+    {
+      "epoch": 0.014428978493908897,
+      "grad_norm": 3.125,
+      "learning_rate": 4.998382730993777e-05,
+      "loss": 0.935,
+      "step": 814
+    },
+    {
+      "epoch": 0.014464430529520468,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.99837267133823e-05,
+      "loss": 0.9059,
+      "step": 816
+    },
+    {
+      "epoch": 0.014499882565132037,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.998362580503513e-05,
+      "loss": 0.9293,
+      "step": 818
+    },
+    {
+      "epoch": 0.014535334600743606,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.998352458489751e-05,
+      "loss": 1.0035,
+      "step": 820
+    },
+    {
+      "epoch": 0.014570786636355176,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.99834230529707e-05,
+      "loss": 0.9548,
+      "step": 822
+    },
+    {
+      "epoch": 0.014606238671966747,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.998332120925598e-05,
+      "loss": 0.9474,
+      "step": 824
+    },
+    {
+      "epoch": 0.014641690707578316,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.9983219053754627e-05,
+      "loss": 0.9112,
+      "step": 826
+    },
+    {
+      "epoch": 0.014677142743189885,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.9983116586467896e-05,
+      "loss": 0.9462,
+      "step": 828
+    },
+    {
+      "epoch": 0.014712594778801455,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.998301380739706e-05,
+      "loss": 0.9302,
+      "step": 830
+    },
+    {
+      "epoch": 0.014748046814413026,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.998291071654343e-05,
+      "loss": 0.9114,
+      "step": 832
+    },
+    {
+      "epoch": 0.014783498850024595,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.9982807313908273e-05,
+      "loss": 0.9189,
+      "step": 834
+    },
+    {
+      "epoch": 0.014818950885636164,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.998270359949289e-05,
+      "loss": 0.9278,
+      "step": 836
+    },
+    {
+      "epoch": 0.014854402921247734,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.998259957329856e-05,
+      "loss": 0.9242,
+      "step": 838
+    },
+    {
+      "epoch": 0.014889854956859305,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.99824952353266e-05,
+      "loss": 0.9375,
+      "step": 840
+    },
+    {
+      "epoch": 0.014925306992470874,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.9982390585578295e-05,
+      "loss": 0.9858,
+      "step": 842
+    },
+    {
+      "epoch": 0.014960759028082443,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.9982285624054956e-05,
+      "loss": 0.9225,
+      "step": 844
+    },
+    {
+      "epoch": 0.014996211063694014,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.9982180350757905e-05,
+      "loss": 0.9209,
+      "step": 846
+    },
+    {
+      "epoch": 0.015031663099305584,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.998207476568845e-05,
+      "loss": 0.9432,
+      "step": 848
+    },
+    {
+      "epoch": 0.015067115134917153,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.99819688688479e-05,
+      "loss": 0.9355,
+      "step": 850
+    },
+    {
+      "epoch": 0.015102567170528722,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.998186266023758e-05,
+      "loss": 0.9108,
+      "step": 852
+    },
+    {
+      "epoch": 0.015138019206140293,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.998175613985882e-05,
+      "loss": 0.9086,
+      "step": 854
+    },
+    {
+      "epoch": 0.015173471241751862,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.998164930771294e-05,
+      "loss": 0.8956,
+      "step": 856
+    },
+    {
+      "epoch": 0.015208923277363432,
+      "grad_norm": 3.125,
+      "learning_rate": 4.998154216380129e-05,
+      "loss": 0.9252,
+      "step": 858
+    },
+    {
+      "epoch": 0.015244375312975001,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.9981434708125184e-05,
+      "loss": 0.9134,
+      "step": 860
+    },
+    {
+      "epoch": 0.015279827348586572,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.9981326940685985e-05,
+      "loss": 0.9403,
+      "step": 862
+    },
+    {
+      "epoch": 0.015315279384198141,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.998121886148503e-05,
+      "loss": 0.8918,
+      "step": 864
+    },
+    {
+      "epoch": 0.01535073141980971,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.998111047052366e-05,
+      "loss": 0.9087,
+      "step": 866
+    },
+    {
+      "epoch": 0.015386183455421282,
+      "grad_norm": 3.625,
+      "learning_rate": 4.9981001767803245e-05,
+      "loss": 0.9514,
+      "step": 868
+    },
+    {
+      "epoch": 0.015421635491032851,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.998089275332513e-05,
+      "loss": 0.9317,
+      "step": 870
+    },
+    {
+      "epoch": 0.01545708752664442,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.998078342709067e-05,
+      "loss": 0.874,
+      "step": 872
+    },
+    {
+      "epoch": 0.01549253956225599,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.9980673789101234e-05,
+      "loss": 0.9233,
+      "step": 874
+    },
+    {
+      "epoch": 0.01552799159786756,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.99805638393582e-05,
+      "loss": 0.9601,
+      "step": 876
+    },
+    {
+      "epoch": 0.01556344363347913,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.998045357786293e-05,
+      "loss": 0.9901,
+      "step": 878
+    },
+    {
+      "epoch": 0.0155988956690907,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.9980343004616795e-05,
+      "loss": 0.9001,
+      "step": 880
+    },
+    {
+      "epoch": 0.01563434770470227,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.998023211962119e-05,
+      "loss": 0.9135,
+      "step": 882
+    },
+    {
+      "epoch": 0.01566979974031384,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.998012092287749e-05,
+      "loss": 0.9025,
+      "step": 884
+    },
+    {
+      "epoch": 0.015705251775925407,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.998000941438709e-05,
+      "loss": 0.9301,
+      "step": 886
+    },
+    {
+      "epoch": 0.015740703811536978,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.997989759415136e-05,
+      "loss": 0.9281,
+      "step": 888
+    },
+    {
+      "epoch": 0.01577615584714855,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.997978546217172e-05,
+      "loss": 0.8869,
+      "step": 890
+    },
+    {
+      "epoch": 0.015811607882760117,
+      "grad_norm": 3.375,
+      "learning_rate": 4.9979673018449555e-05,
+      "loss": 0.9427,
+      "step": 892
+    },
+    {
+      "epoch": 0.015847059918371688,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.9979560262986284e-05,
+      "loss": 0.9285,
+      "step": 894
+    },
+    {
+      "epoch": 0.01588251195398326,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.9979447195783304e-05,
+      "loss": 0.9488,
+      "step": 896
+    },
+    {
+      "epoch": 0.015917963989594826,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.997933381684202e-05,
+      "loss": 0.9196,
+      "step": 898
+    },
+    {
+      "epoch": 0.015953416025206398,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.997922012616385e-05,
+      "loss": 0.9429,
+      "step": 900
+    },
+    {
+      "epoch": 0.01598886806081797,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.997910612375022e-05,
+      "loss": 0.9284,
+      "step": 902
+    },
+    {
+      "epoch": 0.016024320096429536,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.997899180960255e-05,
+      "loss": 0.9539,
+      "step": 904
+    },
+    {
+      "epoch": 0.016059772132041107,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.997887718372226e-05,
+      "loss": 0.9394,
+      "step": 906
+    },
+    {
+      "epoch": 0.016095224167652675,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.997876224611079e-05,
+      "loss": 0.9208,
+      "step": 908
+    },
+    {
+      "epoch": 0.016130676203264246,
+      "grad_norm": 3.5,
+      "learning_rate": 4.9978646996769563e-05,
+      "loss": 0.9476,
+      "step": 910
+    },
+    {
+      "epoch": 0.016166128238875817,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.997853143570003e-05,
+      "loss": 0.9158,
+      "step": 912
+    },
+    {
+      "epoch": 0.016201580274487384,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.997841556290362e-05,
+      "loss": 0.9036,
+      "step": 914
+    },
+    {
+      "epoch": 0.016237032310098955,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.99782993783818e-05,
+      "loss": 0.8826,
+      "step": 916
+    },
+    {
+      "epoch": 0.016272484345710526,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.997818288213599e-05,
+      "loss": 0.906,
+      "step": 918
+    },
+    {
+      "epoch": 0.016307936381322094,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.997806607416767e-05,
+      "loss": 0.9132,
+      "step": 920
+    },
+    {
+      "epoch": 0.016343388416933665,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.997794895447829e-05,
+      "loss": 0.9249,
+      "step": 922
+    },
+    {
+      "epoch": 0.016378840452545236,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.9977831523069305e-05,
+      "loss": 0.942,
+      "step": 924
+    },
+    {
+      "epoch": 0.016414292488156804,
+      "grad_norm": 3.25,
+      "learning_rate": 4.9977713779942195e-05,
+      "loss": 0.9376,
+      "step": 926
+    },
+    {
+      "epoch": 0.016449744523768375,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.99775957250984e-05,
+      "loss": 0.8812,
+      "step": 928
+    },
+    {
+      "epoch": 0.016485196559379942,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.997747735853943e-05,
+      "loss": 0.9415,
+      "step": 930
+    },
+    {
+      "epoch": 0.016520648594991513,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.997735868026674e-05,
+      "loss": 0.9229,
+      "step": 932
+    },
+    {
+      "epoch": 0.016556100630603084,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.9977239690281816e-05,
+      "loss": 0.9193,
+      "step": 934
+    },
+    {
+      "epoch": 0.016591552666214652,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.9977120388586144e-05,
+      "loss": 0.9391,
+      "step": 936
+    },
+    {
+      "epoch": 0.016627004701826223,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.99770007751812e-05,
+      "loss": 0.8891,
+      "step": 938
+    },
+    {
+      "epoch": 0.016662456737437794,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.99768808500685e-05,
+      "loss": 0.9539,
+      "step": 940
+    },
+    {
+      "epoch": 0.01669790877304936,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.997676061324953e-05,
+      "loss": 0.9205,
+      "step": 942
+    },
+    {
+      "epoch": 0.016733360808660933,
+      "grad_norm": 3.25,
+      "learning_rate": 4.997664006472579e-05,
+      "loss": 0.9477,
+      "step": 944
+    },
+    {
+      "epoch": 0.016768812844272504,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.997651920449878e-05,
+      "loss": 0.9004,
+      "step": 946
+    },
+    {
+      "epoch": 0.01680426487988407,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.9976398032570015e-05,
+      "loss": 0.8733,
+      "step": 948
+    },
+    {
+      "epoch": 0.016839716915495642,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.9976276548941005e-05,
+      "loss": 0.9192,
+      "step": 950
+    },
+    {
+      "epoch": 0.01687516895110721,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.997615475361327e-05,
+      "loss": 0.9484,
+      "step": 952
+    },
+    {
+      "epoch": 0.01691062098671878,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.997603264658832e-05,
+      "loss": 0.8731,
+      "step": 954
+    },
+    {
+      "epoch": 0.016946073022330352,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.997591022786768e-05,
+      "loss": 0.9159,
+      "step": 956
+    },
+    {
+      "epoch": 0.01698152505794192,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.997578749745288e-05,
+      "loss": 0.9644,
+      "step": 958
+    },
+    {
+      "epoch": 0.01701697709355349,
+      "grad_norm": 3.0,
+      "learning_rate": 4.997566445534547e-05,
+      "loss": 0.9076,
+      "step": 960
+    },
+    {
+      "epoch": 0.01705242912916506,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.997554110154696e-05,
+      "loss": 0.9178,
+      "step": 962
+    },
+    {
+      "epoch": 0.01708788116477663,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.99754174360589e-05,
+      "loss": 0.935,
+      "step": 964
+    },
+    {
+      "epoch": 0.0171233332003882,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.997529345888284e-05,
+      "loss": 0.9035,
+      "step": 966
+    },
+    {
+      "epoch": 0.01715878523599977,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.9975169170020306e-05,
+      "loss": 0.8921,
+      "step": 968
+    },
+    {
+      "epoch": 0.01719423727161134,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.997504456947287e-05,
+      "loss": 0.8597,
+      "step": 970
+    },
+    {
+      "epoch": 0.01722968930722291,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.997491965724208e-05,
+      "loss": 0.9268,
+      "step": 972
+    },
+    {
+      "epoch": 0.017265141342834477,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.99747944333295e-05,
+      "loss": 0.9444,
+      "step": 974
+    },
+    {
+      "epoch": 0.01730059337844605,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.997466889773668e-05,
+      "loss": 0.9268,
+      "step": 976
+    },
+    {
+      "epoch": 0.01733604541405762,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.997454305046519e-05,
+      "loss": 0.9425,
+      "step": 978
+    },
+    {
+      "epoch": 0.017371497449669187,
+      "grad_norm": 3.0,
+      "learning_rate": 4.9974416891516615e-05,
+      "loss": 0.9128,
+      "step": 980
+    },
+    {
+      "epoch": 0.017406949485280758,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.9974290420892514e-05,
+      "loss": 0.9078,
+      "step": 982
+    },
+    {
+      "epoch": 0.01744240152089233,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.9974163638594475e-05,
+      "loss": 0.9496,
+      "step": 984
+    },
+    {
+      "epoch": 0.017477853556503897,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.9974036544624063e-05,
+      "loss": 0.9128,
+      "step": 986
+    },
+    {
+      "epoch": 0.017513305592115468,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.997390913898289e-05,
+      "loss": 0.8829,
+      "step": 988
+    },
+    {
+      "epoch": 0.01754875762772704,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.997378142167253e-05,
+      "loss": 0.8951,
+      "step": 990
+    },
+    {
+      "epoch": 0.017584209663338606,
+      "grad_norm": 2.875,
+      "learning_rate": 4.997365339269457e-05,
+      "loss": 0.9845,
+      "step": 992
+    },
+    {
+      "epoch": 0.017619661698950177,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.997352505205062e-05,
+      "loss": 0.8814,
+      "step": 994
+    },
+    {
+      "epoch": 0.017655113734561745,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.997339639974229e-05,
+      "loss": 0.9382,
+      "step": 996
+    },
+    {
+      "epoch": 0.017690565770173316,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.9973267435771165e-05,
+      "loss": 0.9112,
+      "step": 998
+    },
+    {
+      "epoch": 0.017726017805784887,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.9973138160138865e-05,
+      "loss": 0.886,
+      "step": 1000
+    },
+    {
+      "epoch": 0.017761469841396454,
+      "grad_norm": 2.875,
+      "learning_rate": 4.9973008572847e-05,
+      "loss": 0.9003,
+      "step": 1002
+    },
+    {
+      "epoch": 0.017796921877008025,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.9972878673897194e-05,
+      "loss": 0.9066,
+      "step": 1004
+    },
+    {
+      "epoch": 0.017832373912619597,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.997274846329106e-05,
+      "loss": 0.8871,
+      "step": 1006
+    },
+    {
+      "epoch": 0.017867825948231164,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.9972617941030236e-05,
+      "loss": 0.8963,
+      "step": 1008
+    },
+    {
+      "epoch": 0.017903277983842735,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.9972487107116336e-05,
+      "loss": 0.8911,
+      "step": 1010
+    },
+    {
+      "epoch": 0.017938730019454306,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.9972355961550995e-05,
+      "loss": 0.9564,
+      "step": 1012
+    },
+    {
+      "epoch": 0.017974182055065874,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.997222450433585e-05,
+      "loss": 0.9578,
+      "step": 1014
+    },
+    {
+      "epoch": 0.018009634090677445,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.9972092735472556e-05,
+      "loss": 0.9335,
+      "step": 1016
+    },
+    {
+      "epoch": 0.018045086126289012,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.997196065496274e-05,
+      "loss": 0.9272,
+      "step": 1018
+    },
+    {
+      "epoch": 0.018080538161900583,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.997182826280805e-05,
+      "loss": 0.9273,
+      "step": 1020
+    },
+    {
+      "epoch": 0.018115990197512154,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.9971695559010155e-05,
+      "loss": 0.8782,
+      "step": 1022
+    },
+    {
+      "epoch": 0.018151442233123722,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.997156254357069e-05,
+      "loss": 0.935,
+      "step": 1024
+    },
+    {
+      "epoch": 0.018186894268735293,
+      "grad_norm": 3.671875,
+      "learning_rate": 4.997142921649134e-05,
+      "loss": 0.9054,
+      "step": 1026
+    },
+    {
+      "epoch": 0.018222346304346864,
+      "grad_norm": 3.25,
+      "learning_rate": 4.997129557777375e-05,
+      "loss": 0.911,
+      "step": 1028
+    },
+    {
+      "epoch": 0.01825779833995843,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.9971161627419585e-05,
+      "loss": 0.9384,
+      "step": 1030
+    },
+    {
+      "epoch": 0.018293250375570003,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.9971027365430526e-05,
+      "loss": 0.9084,
+      "step": 1032
+    },
+    {
+      "epoch": 0.018328702411181574,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.997089279180825e-05,
+      "loss": 0.8903,
+      "step": 1034
+    },
+    {
+      "epoch": 0.01836415444679314,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.997075790655443e-05,
+      "loss": 0.9273,
+      "step": 1036
+    },
+    {
+      "epoch": 0.018399606482404712,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.997062270967075e-05,
+      "loss": 0.9002,
+      "step": 1038
+    },
+    {
+      "epoch": 0.01843505851801628,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.9970487201158903e-05,
+      "loss": 0.887,
+      "step": 1040
+    },
+    {
+      "epoch": 0.01847051055362785,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.997035138102058e-05,
+      "loss": 0.9246,
+      "step": 1042
+    },
+    {
+      "epoch": 0.018505962589239422,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.997021524925747e-05,
+      "loss": 0.9138,
+      "step": 1044
+    },
+    {
+      "epoch": 0.01854141462485099,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.997007880587127e-05,
+      "loss": 0.9084,
+      "step": 1046
+    },
+    {
+      "epoch": 0.01857686666046256,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.99699420508637e-05,
+      "loss": 0.9483,
+      "step": 1048
+    },
+    {
+      "epoch": 0.01861231869607413,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.996980498423644e-05,
+      "loss": 0.9521,
+      "step": 1050
+    },
+    {
+      "epoch": 0.0186477707316857,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.996966760599122e-05,
+      "loss": 0.9273,
+      "step": 1052
+    },
+    {
+      "epoch": 0.01868322276729727,
+      "grad_norm": 2.875,
+      "learning_rate": 4.996952991612975e-05,
+      "loss": 0.8884,
+      "step": 1054
+    },
+    {
+      "epoch": 0.01871867480290884,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.996939191465375e-05,
+      "loss": 0.9442,
+      "step": 1056
+    },
+    {
+      "epoch": 0.01875412683852041,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.9969253601564935e-05,
+      "loss": 0.9197,
+      "step": 1058
+    },
+    {
+      "epoch": 0.01878957887413198,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.996911497686503e-05,
+      "loss": 0.9621,
+      "step": 1060
+    },
+    {
+      "epoch": 0.018825030909743547,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.9968976040555785e-05,
+      "loss": 0.9149,
+      "step": 1062
+    },
+    {
+      "epoch": 0.01886048294535512,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.9968836792638904e-05,
+      "loss": 0.9447,
+      "step": 1064
+    },
+    {
+      "epoch": 0.01889593498096669,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.996869723311616e-05,
+      "loss": 0.9461,
+      "step": 1066
+    },
+    {
+      "epoch": 0.018931387016578257,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.996855736198926e-05,
+      "loss": 0.9095,
+      "step": 1068
+    },
+    {
+      "epoch": 0.018966839052189828,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.996841717925997e-05,
+      "loss": 0.9161,
+      "step": 1070
+    },
+    {
+      "epoch": 0.0190022910878014,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.996827668493003e-05,
+      "loss": 0.9484,
+      "step": 1072
+    },
+    {
+      "epoch": 0.019037743123412967,
+      "grad_norm": 3.421875,
+      "learning_rate": 4.99681358790012e-05,
+      "loss": 0.9372,
+      "step": 1074
+    },
+    {
+      "epoch": 0.019073195159024538,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.996799476147524e-05,
+      "loss": 0.9262,
+      "step": 1076
+    },
+    {
+      "epoch": 0.019108647194636105,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.99678533323539e-05,
+      "loss": 0.9125,
+      "step": 1078
+    },
+    {
+      "epoch": 0.019144099230247676,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.996771159163895e-05,
+      "loss": 0.8768,
+      "step": 1080
+    },
+    {
+      "epoch": 0.019179551265859247,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.996756953933216e-05,
+      "loss": 0.9753,
+      "step": 1082
+    },
+    {
+      "epoch": 0.019215003301470815,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.996742717543531e-05,
+      "loss": 0.8453,
+      "step": 1084
+    },
+    {
+      "epoch": 0.019250455337082386,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.996728449995016e-05,
+      "loss": 0.96,
+      "step": 1086
+    },
+    {
+      "epoch": 0.019285907372693957,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.99671415128785e-05,
+      "loss": 0.9682,
+      "step": 1088
+    },
+    {
+      "epoch": 0.019321359408305525,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.996699821422212e-05,
+      "loss": 0.922,
+      "step": 1090
+    },
+    {
+      "epoch": 0.019356811443917096,
+      "grad_norm": 3.125,
+      "learning_rate": 4.99668546039828e-05,
+      "loss": 0.8806,
+      "step": 1092
+    },
+    {
+      "epoch": 0.019392263479528667,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.996671068216233e-05,
+      "loss": 0.8965,
+      "step": 1094
+    },
+    {
+      "epoch": 0.019427715515140234,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.996656644876252e-05,
+      "loss": 0.9415,
+      "step": 1096
+    },
+    {
+      "epoch": 0.019463167550751805,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.996642190378515e-05,
+      "loss": 0.9455,
+      "step": 1098
+    },
+    {
+      "epoch": 0.019498619586363373,
+      "grad_norm": 3.375,
+      "learning_rate": 4.996627704723205e-05,
+      "loss": 0.9358,
+      "step": 1100
+    },
+    {
+      "epoch": 0.019534071621974944,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.9966131879105003e-05,
+      "loss": 0.9367,
+      "step": 1102
+    },
+    {
+      "epoch": 0.019569523657586515,
+      "grad_norm": 3.0,
+      "learning_rate": 4.996598639940583e-05,
+      "loss": 0.9078,
+      "step": 1104
+    },
+    {
+      "epoch": 0.019604975693198082,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.996584060813635e-05,
+      "loss": 0.9562,
+      "step": 1106
+    },
+    {
+      "epoch": 0.019640427728809653,
+      "grad_norm": 3.421875,
+      "learning_rate": 4.996569450529838e-05,
+      "loss": 0.9156,
+      "step": 1108
+    },
+    {
+      "epoch": 0.019675879764421225,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.996554809089373e-05,
+      "loss": 0.9195,
+      "step": 1110
+    },
+    {
+      "epoch": 0.019711331800032792,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.9965401364924254e-05,
+      "loss": 0.8973,
+      "step": 1112
+    },
+    {
+      "epoch": 0.019746783835644363,
+      "grad_norm": 3.375,
+      "learning_rate": 4.996525432739176e-05,
+      "loss": 0.9351,
+      "step": 1114
+    },
+    {
+      "epoch": 0.019782235871255934,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.99651069782981e-05,
+      "loss": 0.9017,
+      "step": 1116
+    },
+    {
+      "epoch": 0.0198176879068675,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.99649593176451e-05,
+      "loss": 0.9318,
+      "step": 1118
+    },
+    {
+      "epoch": 0.019853139942479073,
+      "grad_norm": 3.25,
+      "learning_rate": 4.996481134543461e-05,
+      "loss": 0.8764,
+      "step": 1120
+    },
+    {
+      "epoch": 0.01988859197809064,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.996466306166847e-05,
+      "loss": 0.9343,
+      "step": 1122
+    },
+    {
+      "epoch": 0.01992404401370221,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.996451446634854e-05,
+      "loss": 0.9393,
+      "step": 1124
+    },
+    {
+      "epoch": 0.019959496049313782,
+      "grad_norm": 2.75,
+      "learning_rate": 4.996436555947667e-05,
+      "loss": 0.9263,
+      "step": 1126
+    },
+    {
+      "epoch": 0.01999494808492535,
+      "grad_norm": 3.25,
+      "learning_rate": 4.996421634105471e-05,
+      "loss": 0.9262,
+      "step": 1128
+    },
+    {
+      "epoch": 0.02003040012053692,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.996406681108453e-05,
+      "loss": 0.9101,
+      "step": 1130
+    },
+    {
+      "epoch": 0.020065852156148492,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.9963916969568e-05,
+      "loss": 0.9742,
+      "step": 1132
+    },
+    {
+      "epoch": 0.02010130419176006,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.996376681650698e-05,
+      "loss": 0.9335,
+      "step": 1134
+    },
+    {
+      "epoch": 0.02013675622737163,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.996361635190336e-05,
+      "loss": 0.9756,
+      "step": 1136
+    },
+    {
+      "epoch": 0.0201722082629832,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.9963465575759006e-05,
+      "loss": 0.9531,
+      "step": 1138
+    },
+    {
+      "epoch": 0.02020766029859477,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.9963314488075795e-05,
+      "loss": 0.8895,
+      "step": 1140
+    },
+    {
+      "epoch": 0.02024311233420634,
+      "grad_norm": 2.75,
+      "learning_rate": 4.996316308885562e-05,
+      "loss": 0.8957,
+      "step": 1142
+    },
+    {
+      "epoch": 0.020278564369817908,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.9963011378100376e-05,
+      "loss": 0.9058,
+      "step": 1144
+    },
+    {
+      "epoch": 0.02031401640542948,
+      "grad_norm": 3.0,
+      "learning_rate": 4.996285935581194e-05,
+      "loss": 0.9273,
+      "step": 1146
+    },
+    {
+      "epoch": 0.02034946844104105,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.996270702199223e-05,
+      "loss": 0.9218,
+      "step": 1148
+    },
+    {
+      "epoch": 0.020384920476652617,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.996255437664312e-05,
+      "loss": 0.9251,
+      "step": 1150
+    },
+    {
+      "epoch": 0.02042037251226419,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.996240141976654e-05,
+      "loss": 0.892,
+      "step": 1152
+    },
+    {
+      "epoch": 0.02045582454787576,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.996224815136439e-05,
+      "loss": 0.9465,
+      "step": 1154
+    },
+    {
+      "epoch": 0.020491276583487327,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.996209457143858e-05,
+      "loss": 0.9364,
+      "step": 1156
+    },
+    {
+      "epoch": 0.020526728619098898,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.996194067999103e-05,
+      "loss": 0.9111,
+      "step": 1158
+    },
+    {
+      "epoch": 0.02056218065471047,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.996178647702366e-05,
+      "loss": 0.8519,
+      "step": 1160
+    },
+    {
+      "epoch": 0.020597632690322037,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.996163196253839e-05,
+      "loss": 0.976,
+      "step": 1162
+    },
+    {
+      "epoch": 0.020633084725933608,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.996147713653716e-05,
+      "loss": 0.8858,
+      "step": 1164
+    },
+    {
+      "epoch": 0.020668536761545175,
+      "grad_norm": 2.875,
+      "learning_rate": 4.9961321999021886e-05,
+      "loss": 0.926,
+      "step": 1166
+    },
+    {
+      "epoch": 0.020703988797156746,
+      "grad_norm": 3.0,
+      "learning_rate": 4.996116654999452e-05,
+      "loss": 0.9177,
+      "step": 1168
+    },
+    {
+      "epoch": 0.020739440832768317,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.996101078945699e-05,
+      "loss": 0.9003,
+      "step": 1170
+    },
+    {
+      "epoch": 0.020774892868379885,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.9960854717411243e-05,
+      "loss": 0.9368,
+      "step": 1172
+    },
+    {
+      "epoch": 0.020810344903991456,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.9960698333859234e-05,
+      "loss": 0.9016,
+      "step": 1174
+    },
+    {
+      "epoch": 0.020845796939603027,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.9960541638802903e-05,
+      "loss": 0.8804,
+      "step": 1176
+    },
+    {
+      "epoch": 0.020881248975214595,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.9960384632244216e-05,
+      "loss": 0.9707,
+      "step": 1178
+    },
+    {
+      "epoch": 0.020916701010826166,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.9960227314185124e-05,
+      "loss": 0.8634,
+      "step": 1180
+    },
+    {
+      "epoch": 0.020952153046437737,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.99600696846276e-05,
+      "loss": 0.9332,
+      "step": 1182
+    },
+    {
+      "epoch": 0.020987605082049304,
+      "grad_norm": 3.125,
+      "learning_rate": 4.99599117435736e-05,
+      "loss": 0.9643,
+      "step": 1184
+    },
+    {
+      "epoch": 0.021023057117660875,
+      "grad_norm": 3.421875,
+      "learning_rate": 4.9959753491025095e-05,
+      "loss": 0.888,
+      "step": 1186
+    },
+    {
+      "epoch": 0.021058509153272443,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.9959594926984074e-05,
+      "loss": 0.9664,
+      "step": 1188
+    },
+    {
+      "epoch": 0.021093961188884014,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.99594360514525e-05,
+      "loss": 0.9423,
+      "step": 1190
+    },
+    {
+      "epoch": 0.021129413224495585,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.995927686443237e-05,
+      "loss": 0.8912,
+      "step": 1192
+    },
+    {
+      "epoch": 0.021164865260107153,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.9959117365925654e-05,
+      "loss": 0.9002,
+      "step": 1194
+    },
+    {
+      "epoch": 0.021200317295718724,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.995895755593436e-05,
+      "loss": 0.9246,
+      "step": 1196
+    },
+    {
+      "epoch": 0.021235769331330295,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.995879743446047e-05,
+      "loss": 0.8685,
+      "step": 1198
+    },
+    {
+      "epoch": 0.021271221366941862,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.995863700150599e-05,
+      "loss": 0.9362,
+      "step": 1200
+    },
+    {
+      "epoch": 0.021306673402553433,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.9958476257072914e-05,
+      "loss": 0.9406,
+      "step": 1202
+    },
+    {
+      "epoch": 0.021342125438165004,
+      "grad_norm": 3.0,
+      "learning_rate": 4.995831520116326e-05,
+      "loss": 0.9096,
+      "step": 1204
+    },
+    {
+      "epoch": 0.021377577473776572,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.9958153833779027e-05,
+      "loss": 0.8959,
+      "step": 1206
+    },
+    {
+      "epoch": 0.021413029509388143,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.995799215492223e-05,
+      "loss": 0.8913,
+      "step": 1208
+    },
+    {
+      "epoch": 0.02144848154499971,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.9957830164594893e-05,
+      "loss": 0.9452,
+      "step": 1210
+    },
+    {
+      "epoch": 0.02148393358061128,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.995766786279903e-05,
+      "loss": 0.9203,
+      "step": 1212
+    },
+    {
+      "epoch": 0.021519385616222853,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.9957505249536676e-05,
+      "loss": 0.9122,
+      "step": 1214
+    },
+    {
+      "epoch": 0.02155483765183442,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.995734232480985e-05,
+      "loss": 0.8905,
+      "step": 1216
+    },
+    {
+      "epoch": 0.02159028968744599,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.995717908862059e-05,
+      "loss": 0.9156,
+      "step": 1218
+    },
+    {
+      "epoch": 0.021625741723057562,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.995701554097094e-05,
+      "loss": 0.9158,
+      "step": 1220
+    },
+    {
+      "epoch": 0.02166119375866913,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.995685168186293e-05,
+      "loss": 0.8903,
+      "step": 1222
+    },
+    {
+      "epoch": 0.0216966457942807,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.9956687511298604e-05,
+      "loss": 0.9174,
+      "step": 1224
+    },
+    {
+      "epoch": 0.021732097829892272,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.995652302928002e-05,
+      "loss": 0.9353,
+      "step": 1226
+    },
+    {
+      "epoch": 0.02176754986550384,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.995635823580922e-05,
+      "loss": 0.9103,
+      "step": 1228
+    },
+    {
+      "epoch": 0.02180300190111541,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.9956193130888276e-05,
+      "loss": 0.9258,
+      "step": 1230
+    },
+    {
+      "epoch": 0.021838453936726978,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.995602771451924e-05,
+      "loss": 0.8866,
+      "step": 1232
+    },
+    {
+      "epoch": 0.02187390597233855,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.9955861986704175e-05,
+      "loss": 0.9572,
+      "step": 1234
+    },
+    {
+      "epoch": 0.02190935800795012,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.9955695947445145e-05,
+      "loss": 0.9017,
+      "step": 1236
+    },
+    {
+      "epoch": 0.021944810043561688,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.995552959674423e-05,
+      "loss": 0.9137,
+      "step": 1238
+    },
+    {
+      "epoch": 0.02198026207917326,
+      "grad_norm": 2.875,
+      "learning_rate": 4.995536293460351e-05,
+      "loss": 0.8681,
+      "step": 1240
+    },
+    {
+      "epoch": 0.02201571411478483,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.995519596102506e-05,
+      "loss": 0.9348,
+      "step": 1242
+    },
+    {
+      "epoch": 0.022051166150396397,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.995502867601095e-05,
+      "loss": 0.9005,
+      "step": 1244
+    },
+    {
+      "epoch": 0.02208661818600797,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.995486107956329e-05,
+      "loss": 0.9338,
+      "step": 1246
+    },
+    {
+      "epoch": 0.02212207022161954,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.995469317168415e-05,
+      "loss": 0.8978,
+      "step": 1248
+    },
+    {
+      "epoch": 0.022157522257231107,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.9954524952375646e-05,
+      "loss": 0.8685,
+      "step": 1250
+    },
+    {
+      "epoch": 0.022192974292842678,
+      "grad_norm": 2.75,
+      "learning_rate": 4.995435642163987e-05,
+      "loss": 0.9427,
+      "step": 1252
+    },
+    {
+      "epoch": 0.022228426328454245,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.995418757947892e-05,
+      "loss": 0.915,
+      "step": 1254
+    },
+    {
+      "epoch": 0.022263878364065817,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.99540184258949e-05,
+      "loss": 0.8801,
+      "step": 1256
+    },
+    {
+      "epoch": 0.022299330399677388,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.995384896088994e-05,
+      "loss": 0.9515,
+      "step": 1258
+    },
+    {
+      "epoch": 0.022334782435288955,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.995367918446613e-05,
+      "loss": 0.9193,
+      "step": 1260
+    },
+    {
+      "epoch": 0.022370234470900526,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.995350909662561e-05,
+      "loss": 0.9507,
+      "step": 1262
+    },
+    {
+      "epoch": 0.022405686506512097,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.995333869737049e-05,
+      "loss": 0.8967,
+      "step": 1264
+    },
+    {
+      "epoch": 0.022441138542123665,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.9953167986702905e-05,
+      "loss": 0.9278,
+      "step": 1266
+    },
+    {
+      "epoch": 0.022476590577735236,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.9952996964624976e-05,
+      "loss": 0.8715,
+      "step": 1268
+    },
+    {
+      "epoch": 0.022512042613346807,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.995282563113885e-05,
+      "loss": 0.8882,
+      "step": 1270
+    },
+    {
+      "epoch": 0.022547494648958374,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.9952653986246646e-05,
+      "loss": 0.9051,
+      "step": 1272
+    },
+    {
+      "epoch": 0.022582946684569945,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.995248202995052e-05,
+      "loss": 0.9401,
+      "step": 1274
+    },
+    {
+      "epoch": 0.022618398720181513,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.9952309762252624e-05,
+      "loss": 0.8906,
+      "step": 1276
+    },
+    {
+      "epoch": 0.022653850755793084,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.995213718315509e-05,
+      "loss": 0.9692,
+      "step": 1278
+    },
+    {
+      "epoch": 0.022689302791404655,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.995196429266009e-05,
+      "loss": 0.8744,
+      "step": 1280
+    },
+    {
+      "epoch": 0.022724754827016223,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.995179109076976e-05,
+      "loss": 0.8542,
+      "step": 1282
+    },
+    {
+      "epoch": 0.022760206862627794,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.9951617577486285e-05,
+      "loss": 0.8996,
+      "step": 1284
+    },
+    {
+      "epoch": 0.022795658898239365,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.995144375281182e-05,
+      "loss": 0.9078,
+      "step": 1286
+    },
+    {
+      "epoch": 0.022831110933850932,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.9951269616748534e-05,
+      "loss": 0.9386,
+      "step": 1288
+    },
+    {
+      "epoch": 0.022866562969462503,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.995109516929859e-05,
+      "loss": 0.8791,
+      "step": 1290
+    },
+    {
+      "epoch": 0.02290201500507407,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.995092041046419e-05,
+      "loss": 0.8965,
+      "step": 1292
+    },
+    {
+      "epoch": 0.022937467040685642,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.995074534024748e-05,
+      "loss": 0.9523,
+      "step": 1294
+    },
+    {
+      "epoch": 0.022972919076297213,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.9950569958650684e-05,
+      "loss": 0.9555,
+      "step": 1296
+    },
+    {
+      "epoch": 0.02300837111190878,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.995039426567596e-05,
+      "loss": 0.8951,
+      "step": 1298
+    },
+    {
+      "epoch": 0.02304382314752035,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.995021826132552e-05,
+      "loss": 0.8788,
+      "step": 1300
+    },
+    {
+      "epoch": 0.023079275183131923,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.995004194560155e-05,
+      "loss": 0.9304,
+      "step": 1302
+    },
+    {
+      "epoch": 0.02311472721874349,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.9949865318506254e-05,
+      "loss": 0.9267,
+      "step": 1304
+    },
+    {
+      "epoch": 0.02315017925435506,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.994968838004184e-05,
+      "loss": 0.8883,
+      "step": 1306
+    },
+    {
+      "epoch": 0.023185631289966632,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.99495111302105e-05,
+      "loss": 0.9201,
+      "step": 1308
+    },
+    {
+      "epoch": 0.0232210833255782,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.9949333569014464e-05,
+      "loss": 0.9868,
+      "step": 1310
+    },
+    {
+      "epoch": 0.02325653536118977,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.994915569645594e-05,
+      "loss": 0.8611,
+      "step": 1312
+    },
+    {
+      "epoch": 0.02329198739680134,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.994897751253715e-05,
+      "loss": 0.9253,
+      "step": 1314
+    },
+    {
+      "epoch": 0.02332743943241291,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.9948799017260325e-05,
+      "loss": 0.9014,
+      "step": 1316
+    },
+    {
+      "epoch": 0.02336289146802448,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.994862021062767e-05,
+      "loss": 0.9189,
+      "step": 1318
+    },
+    {
+      "epoch": 0.023398343503636048,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.994844109264145e-05,
+      "loss": 0.8979,
+      "step": 1320
+    },
+    {
+      "epoch": 0.02343379553924762,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.994826166330386e-05,
+      "loss": 0.8681,
+      "step": 1322
+    },
+    {
+      "epoch": 0.02346924757485919,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.994808192261718e-05,
+      "loss": 0.8864,
+      "step": 1324
+    },
+    {
+      "epoch": 0.023504699610470758,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.994790187058363e-05,
+      "loss": 0.9201,
+      "step": 1326
+    },
+    {
+      "epoch": 0.02354015164608233,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.994772150720545e-05,
+      "loss": 0.8995,
+      "step": 1328
+    },
+    {
+      "epoch": 0.0235756036816939,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.9947540832484904e-05,
+      "loss": 0.8978,
+      "step": 1330
+    },
+    {
+      "epoch": 0.023611055717305467,
+      "grad_norm": 2.875,
+      "learning_rate": 4.994735984642426e-05,
+      "loss": 0.903,
+      "step": 1332
+    },
+    {
+      "epoch": 0.02364650775291704,
+      "grad_norm": 2.875,
+      "learning_rate": 4.9947178549025745e-05,
+      "loss": 0.907,
+      "step": 1334
+    },
+    {
+      "epoch": 0.023681959788528606,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.9946996940291644e-05,
+      "loss": 0.8918,
+      "step": 1336
+    },
+    {
+      "epoch": 0.023717411824140177,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.9946815020224215e-05,
+      "loss": 0.8664,
+      "step": 1338
+    },
+    {
+      "epoch": 0.023752863859751748,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.994663278882573e-05,
+      "loss": 0.9018,
+      "step": 1340
+    },
+    {
+      "epoch": 0.023788315895363316,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.994645024609847e-05,
+      "loss": 0.875,
+      "step": 1342
+    },
+    {
+      "epoch": 0.023823767930974887,
+      "grad_norm": 3.125,
+      "learning_rate": 4.9946267392044696e-05,
+      "loss": 0.9201,
+      "step": 1344
+    },
+    {
+      "epoch": 0.023859219966586458,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.994608422666671e-05,
+      "loss": 0.929,
+      "step": 1346
+    },
+    {
+      "epoch": 0.023894672002198025,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.994590074996679e-05,
+      "loss": 0.929,
+      "step": 1348
+    },
+    {
+      "epoch": 0.023930124037809596,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.994571696194722e-05,
+      "loss": 0.9022,
+      "step": 1350
+    },
+    {
+      "epoch": 0.023965576073421167,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.99455328626103e-05,
+      "loss": 0.8865,
+      "step": 1352
+    },
+    {
+      "epoch": 0.024001028109032735,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.994534845195832e-05,
+      "loss": 0.8835,
+      "step": 1354
+    },
+    {
+      "epoch": 0.024036480144644306,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.99451637299936e-05,
+      "loss": 0.9338,
+      "step": 1356
+    },
+    {
+      "epoch": 0.024071932180255873,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.9944978696718416e-05,
+      "loss": 0.9045,
+      "step": 1358
+    },
+    {
+      "epoch": 0.024107384215867445,
+      "grad_norm": 2.625,
+      "learning_rate": 4.99447933521351e-05,
+      "loss": 0.9218,
+      "step": 1360
+    },
+    {
+      "epoch": 0.024142836251479016,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.994460769624596e-05,
+      "loss": 0.9163,
+      "step": 1362
+    },
+    {
+      "epoch": 0.024178288287090583,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.994442172905331e-05,
+      "loss": 0.8869,
+      "step": 1364
+    },
+    {
+      "epoch": 0.024213740322702154,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.994423545055948e-05,
+      "loss": 0.9001,
+      "step": 1366
+    },
+    {
+      "epoch": 0.024249192358313725,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.994404886076678e-05,
+      "loss": 0.9176,
+      "step": 1368
+    },
+    {
+      "epoch": 0.024284644393925293,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.994386195967754e-05,
+      "loss": 0.9031,
+      "step": 1370
+    },
+    {
+      "epoch": 0.024320096429536864,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.994367474729411e-05,
+      "loss": 0.8798,
+      "step": 1372
+    },
+    {
+      "epoch": 0.024355548465148435,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.994348722361881e-05,
+      "loss": 0.8774,
+      "step": 1374
+    },
+    {
+      "epoch": 0.024391000500760002,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.9943299388653984e-05,
+      "loss": 0.9432,
+      "step": 1376
+    },
+    {
+      "epoch": 0.024426452536371573,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.9943111242401974e-05,
+      "loss": 0.9392,
+      "step": 1378
+    },
+    {
+      "epoch": 0.02446190457198314,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.994292278486514e-05,
+      "loss": 0.8717,
+      "step": 1380
+    },
+    {
+      "epoch": 0.024497356607594712,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.994273401604582e-05,
+      "loss": 0.8945,
+      "step": 1382
+    },
+    {
+      "epoch": 0.024532808643206283,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.994254493594637e-05,
+      "loss": 0.9413,
+      "step": 1384
+    },
+    {
+      "epoch": 0.02456826067881785,
+      "grad_norm": 2.625,
+      "learning_rate": 4.994235554456916e-05,
+      "loss": 0.889,
+      "step": 1386
+    },
+    {
+      "epoch": 0.02460371271442942,
+      "grad_norm": 3.125,
+      "learning_rate": 4.9942165841916545e-05,
+      "loss": 0.9052,
+      "step": 1388
+    },
+    {
+      "epoch": 0.024639164750040993,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.99419758279909e-05,
+      "loss": 0.9182,
+      "step": 1390
+    },
+    {
+      "epoch": 0.02467461678565256,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.9941785502794586e-05,
+      "loss": 0.8578,
+      "step": 1392
+    },
+    {
+      "epoch": 0.02471006882126413,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.994159486632999e-05,
+      "loss": 0.8931,
+      "step": 1394
+    },
+    {
+      "epoch": 0.024745520856875702,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.994140391859947e-05,
+      "loss": 0.9515,
+      "step": 1396
+    },
+    {
+      "epoch": 0.02478097289248727,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.994121265960544e-05,
+      "loss": 0.9105,
+      "step": 1398
+    },
+    {
+      "epoch": 0.02481642492809884,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.994102108935027e-05,
+      "loss": 0.9351,
+      "step": 1400
+    },
+    {
+      "epoch": 0.02485187696371041,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.9940829207836346e-05,
+      "loss": 0.8385,
+      "step": 1402
+    },
+    {
+      "epoch": 0.02488732899932198,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.994063701506607e-05,
+      "loss": 0.8899,
+      "step": 1404
+    },
+    {
+      "epoch": 0.02492278103493355,
+      "grad_norm": 2.875,
+      "learning_rate": 4.994044451104184e-05,
+      "loss": 0.9206,
+      "step": 1406
+    },
+    {
+      "epoch": 0.024958233070545118,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.994025169576605e-05,
+      "loss": 0.9093,
+      "step": 1408
+    },
+    {
+      "epoch": 0.02499368510615669,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.9940058569241125e-05,
+      "loss": 0.9178,
+      "step": 1410
+    },
+    {
+      "epoch": 0.02502913714176826,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.993986513146945e-05,
+      "loss": 0.8901,
+      "step": 1412
+    },
+    {
+      "epoch": 0.025064589177379828,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.993967138245347e-05,
+      "loss": 0.8781,
+      "step": 1414
+    },
+    {
+      "epoch": 0.0251000412129914,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.9939477322195574e-05,
+      "loss": 0.9092,
+      "step": 1416
+    },
+    {
+      "epoch": 0.02513549324860297,
+      "grad_norm": 3.375,
+      "learning_rate": 4.9939282950698195e-05,
+      "loss": 0.9014,
+      "step": 1418
+    },
+    {
+      "epoch": 0.025170945284214537,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.9939088267963763e-05,
+      "loss": 0.8958,
+      "step": 1420
+    },
+    {
+      "epoch": 0.02520639731982611,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.9938893273994706e-05,
+      "loss": 0.918,
+      "step": 1422
+    },
+    {
+      "epoch": 0.025241849355437676,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.9938697968793454e-05,
+      "loss": 0.8702,
+      "step": 1424
+    },
+    {
+      "epoch": 0.025277301391049247,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.9938502352362435e-05,
+      "loss": 0.9012,
+      "step": 1426
+    },
+    {
+      "epoch": 0.025312753426660818,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.9938306424704114e-05,
+      "loss": 0.8525,
+      "step": 1428
+    },
+    {
+      "epoch": 0.025348205462272386,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.993811018582092e-05,
+      "loss": 0.909,
+      "step": 1430
+    },
+    {
+      "epoch": 0.025383657497883957,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.993791363571531e-05,
+      "loss": 0.8888,
+      "step": 1432
+    },
+    {
+      "epoch": 0.025419109533495528,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.993771677438972e-05,
+      "loss": 0.9036,
+      "step": 1434
+    },
+    {
+      "epoch": 0.025454561569107095,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.993751960184663e-05,
+      "loss": 0.8765,
+      "step": 1436
+    },
+    {
+      "epoch": 0.025490013604718666,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.993732211808848e-05,
+      "loss": 0.9343,
+      "step": 1438
+    },
+    {
+      "epoch": 0.025525465640330237,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.993712432311775e-05,
+      "loss": 0.8502,
+      "step": 1440
+    },
+    {
+      "epoch": 0.025560917675941805,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.9936926216936905e-05,
+      "loss": 0.9446,
+      "step": 1442
+    },
+    {
+      "epoch": 0.025596369711553376,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.993672779954841e-05,
+      "loss": 0.9214,
+      "step": 1444
+    },
+    {
+      "epoch": 0.025631821747164944,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.993652907095475e-05,
+      "loss": 0.8599,
+      "step": 1446
+    },
+    {
+      "epoch": 0.025667273782776515,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.99363300311584e-05,
+      "loss": 0.9155,
+      "step": 1448
+    },
+    {
+      "epoch": 0.025702725818388086,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.993613068016184e-05,
+      "loss": 0.9058,
+      "step": 1450
+    },
+    {
+      "epoch": 0.025738177853999653,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.993593101796756e-05,
+      "loss": 0.9138,
+      "step": 1452
+    },
+    {
+      "epoch": 0.025773629889611224,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.993573104457806e-05,
+      "loss": 0.888,
+      "step": 1454
+    },
+    {
+      "epoch": 0.025809081925222795,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.993553075999584e-05,
+      "loss": 0.9234,
+      "step": 1456
+    },
+    {
+      "epoch": 0.025844533960834363,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.9935330164223376e-05,
+      "loss": 0.91,
+      "step": 1458
+    },
+    {
+      "epoch": 0.025879985996445934,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.993512925726319e-05,
+      "loss": 0.9006,
+      "step": 1460
+    },
+    {
+      "epoch": 0.025915438032057505,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.993492803911778e-05,
+      "loss": 0.8324,
+      "step": 1462
+    },
+    {
+      "epoch": 0.025950890067669072,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.993472650978968e-05,
+      "loss": 0.8797,
+      "step": 1464
+    },
+    {
+      "epoch": 0.025986342103280644,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.993452466928137e-05,
+      "loss": 0.932,
+      "step": 1466
+    },
+    {
+      "epoch": 0.02602179413889221,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.993432251759538e-05,
+      "loss": 0.8812,
+      "step": 1468
+    },
+    {
+      "epoch": 0.026057246174503782,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.993412005473425e-05,
+      "loss": 0.9186,
+      "step": 1470
+    },
+    {
+      "epoch": 0.026092698210115353,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.993391728070049e-05,
+      "loss": 0.9155,
+      "step": 1472
+    },
+    {
+      "epoch": 0.02612815024572692,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.993371419549664e-05,
+      "loss": 0.8935,
+      "step": 1474
+    },
+    {
+      "epoch": 0.026163602281338492,
+      "grad_norm": 2.875,
+      "learning_rate": 4.9933510799125224e-05,
+      "loss": 0.8947,
+      "step": 1476
+    },
+    {
+      "epoch": 0.026199054316950063,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.9933307091588796e-05,
+      "loss": 0.8763,
+      "step": 1478
+    },
+    {
+      "epoch": 0.02623450635256163,
+      "grad_norm": 2.875,
+      "learning_rate": 4.993310307288988e-05,
+      "loss": 0.8889,
+      "step": 1480
+    },
+    {
+      "epoch": 0.0262699583881732,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.993289874303103e-05,
+      "loss": 0.8904,
+      "step": 1482
+    },
+    {
+      "epoch": 0.02630541042378477,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.993269410201481e-05,
+      "loss": 0.9644,
+      "step": 1484
+    },
+    {
+      "epoch": 0.02634086245939634,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.993248914984375e-05,
+      "loss": 0.9051,
+      "step": 1486
+    },
+    {
+      "epoch": 0.02637631449500791,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.9932283886520413e-05,
+      "loss": 0.9146,
+      "step": 1488
+    },
+    {
+      "epoch": 0.02641176653061948,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.993207831204738e-05,
+      "loss": 0.8888,
+      "step": 1490
+    },
+    {
+      "epoch": 0.02644721856623105,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.9931872426427196e-05,
+      "loss": 0.9044,
+      "step": 1492
+    },
+    {
+      "epoch": 0.02648267060184262,
+      "grad_norm": 2.875,
+      "learning_rate": 4.9931666229662435e-05,
+      "loss": 0.955,
+      "step": 1494
+    },
+    {
+      "epoch": 0.02651812263745419,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.993145972175567e-05,
+      "loss": 0.9151,
+      "step": 1496
+    },
+    {
+      "epoch": 0.02655357467306576,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.993125290270949e-05,
+      "loss": 0.917,
+      "step": 1498
+    },
+    {
+      "epoch": 0.02658902670867733,
+      "grad_norm": 3.25,
+      "learning_rate": 4.993104577252646e-05,
+      "loss": 0.9137,
+      "step": 1500
+    },
+    {
+      "epoch": 0.026624478744288898,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.993083833120917e-05,
+      "loss": 0.9038,
+      "step": 1502
+    },
+    {
+      "epoch": 0.02665993077990047,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.993063057876022e-05,
+      "loss": 0.9221,
+      "step": 1504
+    },
+    {
+      "epoch": 0.026695382815512037,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.993042251518218e-05,
+      "loss": 0.8918,
+      "step": 1506
+    },
+    {
+      "epoch": 0.026730834851123608,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.993021414047767e-05,
+      "loss": 0.9025,
+      "step": 1508
+    },
+    {
+      "epoch": 0.02676628688673518,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.9930005454649276e-05,
+      "loss": 0.8823,
+      "step": 1510
+    },
+    {
+      "epoch": 0.026801738922346746,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.9929796457699606e-05,
+      "loss": 0.8694,
+      "step": 1512
+    },
+    {
+      "epoch": 0.026837190957958317,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.9929587149631265e-05,
+      "loss": 0.9034,
+      "step": 1514
+    },
+    {
+      "epoch": 0.026872642993569888,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.9929377530446876e-05,
+      "loss": 0.9044,
+      "step": 1516
+    },
+    {
+      "epoch": 0.026908095029181456,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.992916760014904e-05,
+      "loss": 0.9068,
+      "step": 1518
+    },
+    {
+      "epoch": 0.026943547064793027,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.992895735874039e-05,
+      "loss": 0.8867,
+      "step": 1520
+    },
+    {
+      "epoch": 0.026978999100404598,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.9928746806223545e-05,
+      "loss": 0.8914,
+      "step": 1522
+    },
+    {
+      "epoch": 0.027014451136016165,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.992853594260114e-05,
+      "loss": 0.9118,
+      "step": 1524
+    },
+    {
+      "epoch": 0.027049903171627736,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.992832476787579e-05,
+      "loss": 0.8973,
+      "step": 1526
+    },
+    {
+      "epoch": 0.027085355207239304,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.992811328205013e-05,
+      "loss": 0.9082,
+      "step": 1528
+    },
+    {
+      "epoch": 0.027120807242850875,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.992790148512682e-05,
+      "loss": 0.8552,
+      "step": 1530
+    },
+    {
+      "epoch": 0.027156259278462446,
+      "grad_norm": 3.25,
+      "learning_rate": 4.992768937710849e-05,
+      "loss": 0.9192,
+      "step": 1532
+    },
+    {
+      "epoch": 0.027191711314074014,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.992747695799779e-05,
+      "loss": 0.9207,
+      "step": 1534
+    },
+    {
+      "epoch": 0.027227163349685585,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.992726422779737e-05,
+      "loss": 0.9069,
+      "step": 1536
+    },
+    {
+      "epoch": 0.027262615385297156,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.9927051186509876e-05,
+      "loss": 0.8553,
+      "step": 1538
+    },
+    {
+      "epoch": 0.027298067420908723,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.992683783413798e-05,
+      "loss": 0.9094,
+      "step": 1540
+    },
+    {
+      "epoch": 0.027333519456520294,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.9926624170684345e-05,
+      "loss": 0.9098,
+      "step": 1542
+    },
+    {
+      "epoch": 0.027368971492131865,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.9926410196151625e-05,
+      "loss": 0.8634,
+      "step": 1544
+    },
+    {
+      "epoch": 0.027404423527743433,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.99261959105425e-05,
+      "loss": 0.9041,
+      "step": 1546
+    },
+    {
+      "epoch": 0.027439875563355004,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.992598131385964e-05,
+      "loss": 0.8923,
+      "step": 1548
+    },
+    {
+      "epoch": 0.02747532759896657,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.992576640610572e-05,
+      "loss": 0.9115,
+      "step": 1550
+    },
+    {
+      "epoch": 0.027510779634578143,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.992555118728344e-05,
+      "loss": 0.8475,
+      "step": 1552
+    },
+    {
+      "epoch": 0.027546231670189714,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.992533565739547e-05,
+      "loss": 0.9031,
+      "step": 1554
+    },
+    {
+      "epoch": 0.02758168370580128,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.99251198164445e-05,
+      "loss": 0.8947,
+      "step": 1556
+    },
+    {
+      "epoch": 0.027617135741412852,
+      "grad_norm": 2.625,
+      "learning_rate": 4.992490366443322e-05,
+      "loss": 0.8605,
+      "step": 1558
+    },
+    {
+      "epoch": 0.027652587777024423,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.992468720136434e-05,
+      "loss": 0.8773,
+      "step": 1560
+    },
+    {
+      "epoch": 0.02768803981263599,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.9924470427240556e-05,
+      "loss": 0.9447,
+      "step": 1562
+    },
+    {
+      "epoch": 0.027723491848247562,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.992425334206457e-05,
+      "loss": 0.9087,
+      "step": 1564
+    },
+    {
+      "epoch": 0.027758943883859133,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.992403594583909e-05,
+      "loss": 0.9347,
+      "step": 1566
+    },
+    {
+      "epoch": 0.0277943959194707,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.9923818238566844e-05,
+      "loss": 0.894,
+      "step": 1568
+    },
+    {
+      "epoch": 0.02782984795508227,
+      "grad_norm": 3.25,
+      "learning_rate": 4.9923600220250526e-05,
+      "loss": 0.855,
+      "step": 1570
+    },
+    {
+      "epoch": 0.02786529999069384,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.9923381890892874e-05,
+      "loss": 0.8794,
+      "step": 1572
+    },
+    {
+      "epoch": 0.02790075202630541,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.9923163250496606e-05,
+      "loss": 0.8898,
+      "step": 1574
+    },
+    {
+      "epoch": 0.02793620406191698,
+      "grad_norm": 3.25,
+      "learning_rate": 4.992294429906445e-05,
+      "loss": 0.9033,
+      "step": 1576
+    },
+    {
+      "epoch": 0.02797165609752855,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.9922725036599146e-05,
+      "loss": 0.9061,
+      "step": 1578
+    },
+    {
+      "epoch": 0.02800710813314012,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.992250546310342e-05,
+      "loss": 0.8935,
+      "step": 1580
+    },
+    {
+      "epoch": 0.02804256016875169,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.992228557858002e-05,
+      "loss": 0.8952,
+      "step": 1582
+    },
+    {
+      "epoch": 0.02807801220436326,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.992206538303168e-05,
+      "loss": 0.8433,
+      "step": 1584
+    },
+    {
+      "epoch": 0.02811346423997483,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.992184487646116e-05,
+      "loss": 0.8904,
+      "step": 1586
+    },
+    {
+      "epoch": 0.0281489162755864,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.992162405887121e-05,
+      "loss": 0.8927,
+      "step": 1588
+    },
+    {
+      "epoch": 0.028184368311197968,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.992140293026458e-05,
+      "loss": 0.908,
+      "step": 1590
+    },
+    {
+      "epoch": 0.02821982034680954,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.992118149064403e-05,
+      "loss": 0.8945,
+      "step": 1592
+    },
+    {
+      "epoch": 0.028255272382421107,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.9920959740012326e-05,
+      "loss": 0.9162,
+      "step": 1594
+    },
+    {
+      "epoch": 0.028290724418032678,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.9920737678372234e-05,
+      "loss": 0.8884,
+      "step": 1596
+    },
+    {
+      "epoch": 0.02832617645364425,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.9920515305726526e-05,
+      "loss": 0.8824,
+      "step": 1598
+    },
+    {
+      "epoch": 0.028361628489255816,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.992029262207798e-05,
+      "loss": 0.8814,
+      "step": 1600
+    },
+    {
+      "epoch": 0.028397080524867387,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.9920069627429375e-05,
+      "loss": 0.9012,
+      "step": 1602
+    },
+    {
+      "epoch": 0.02843253256047896,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.991984632178349e-05,
+      "loss": 0.9315,
+      "step": 1604
+    },
+    {
+      "epoch": 0.028467984596090526,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.9919622705143106e-05,
+      "loss": 0.9475,
+      "step": 1606
+    },
+    {
+      "epoch": 0.028503436631702097,
+      "grad_norm": 3.125,
+      "learning_rate": 4.991939877751103e-05,
+      "loss": 0.9051,
+      "step": 1608
+    },
+    {
+      "epoch": 0.028538888667313668,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.991917453889004e-05,
+      "loss": 0.887,
+      "step": 1610
+    },
+    {
+      "epoch": 0.028574340702925236,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.991894998928295e-05,
+      "loss": 0.9247,
+      "step": 1612
+    },
+    {
+      "epoch": 0.028609792738536807,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.991872512869256e-05,
+      "loss": 0.9195,
+      "step": 1614
+    },
+    {
+      "epoch": 0.028645244774148374,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.9918499957121654e-05,
+      "loss": 0.9334,
+      "step": 1616
+    },
+    {
+      "epoch": 0.028680696809759945,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.991827447457307e-05,
+      "loss": 0.8793,
+      "step": 1618
+    },
+    {
+      "epoch": 0.028716148845371516,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.991804868104961e-05,
+      "loss": 0.9034,
+      "step": 1620
+    },
+    {
+      "epoch": 0.028751600880983084,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.991782257655408e-05,
+      "loss": 0.9069,
+      "step": 1622
+    },
+    {
+      "epoch": 0.028787052916594655,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.991759616108933e-05,
+      "loss": 0.8551,
+      "step": 1624
+    },
+    {
+      "epoch": 0.028822504952206226,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.991736943465816e-05,
+      "loss": 0.9543,
+      "step": 1626
+    },
+    {
+      "epoch": 0.028857956987817793,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.991714239726342e-05,
+      "loss": 0.9071,
+      "step": 1628
+    },
+    {
+      "epoch": 0.028893409023429364,
+      "grad_norm": 3.125,
+      "learning_rate": 4.991691504890792e-05,
+      "loss": 0.9125,
+      "step": 1630
+    },
+    {
+      "epoch": 0.028928861059040935,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.991668738959452e-05,
+      "loss": 0.9668,
+      "step": 1632
+    },
+    {
+      "epoch": 0.028964313094652503,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.991645941932604e-05,
+      "loss": 0.8547,
+      "step": 1634
+    },
+    {
+      "epoch": 0.028999765130264074,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.9916231138105354e-05,
+      "loss": 0.8627,
+      "step": 1636
+    },
+    {
+      "epoch": 0.02903521716587564,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.991600254593527e-05,
+      "loss": 0.8693,
+      "step": 1638
+    },
+    {
+      "epoch": 0.029070669201487213,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.9915773642818684e-05,
+      "loss": 0.8523,
+      "step": 1640
+    },
+    {
+      "epoch": 0.029106121237098784,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.991554442875842e-05,
+      "loss": 0.8958,
+      "step": 1642
+    },
+    {
+      "epoch": 0.02914157327271035,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.991531490375736e-05,
+      "loss": 0.8493,
+      "step": 1644
+    },
+    {
+      "epoch": 0.029177025308321922,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.9915085067818355e-05,
+      "loss": 0.9503,
+      "step": 1646
+    },
+    {
+      "epoch": 0.029212477343933493,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.9914854920944276e-05,
+      "loss": 0.9112,
+      "step": 1648
+    },
+    {
+      "epoch": 0.02924792937954506,
+      "grad_norm": 3.125,
+      "learning_rate": 4.9914624463138e-05,
+      "loss": 0.9116,
+      "step": 1650
+    },
+    {
+      "epoch": 0.029283381415156632,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.991439369440239e-05,
+      "loss": 0.899,
+      "step": 1652
+    },
+    {
+      "epoch": 0.029318833450768203,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.9914162614740355e-05,
+      "loss": 0.8528,
+      "step": 1654
+    },
+    {
+      "epoch": 0.02935428548637977,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.991393122415475e-05,
+      "loss": 0.9004,
+      "step": 1656
+    },
+    {
+      "epoch": 0.02938973752199134,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.991369952264847e-05,
+      "loss": 0.915,
+      "step": 1658
+    },
+    {
+      "epoch": 0.02942518955760291,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.991346751022441e-05,
+      "loss": 0.8629,
+      "step": 1660
+    },
+    {
+      "epoch": 0.02946064159321448,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.9913235186885464e-05,
+      "loss": 0.8786,
+      "step": 1662
+    },
+    {
+      "epoch": 0.02949609362882605,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.991300255263454e-05,
+      "loss": 0.8372,
+      "step": 1664
+    },
+    {
+      "epoch": 0.02953154566443762,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.991276960747452e-05,
+      "loss": 0.8547,
+      "step": 1666
+    },
+    {
+      "epoch": 0.02956699770004919,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.9912536351408334e-05,
+      "loss": 0.9204,
+      "step": 1668
+    },
+    {
+      "epoch": 0.02960244973566076,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.991230278443888e-05,
+      "loss": 0.8767,
+      "step": 1670
+    },
+    {
+      "epoch": 0.02963790177127233,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.9912068906569076e-05,
+      "loss": 0.8974,
+      "step": 1672
+    },
+    {
+      "epoch": 0.0296733538068839,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.991183471780184e-05,
+      "loss": 0.8934,
+      "step": 1674
+    },
+    {
+      "epoch": 0.029708805842495467,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.9911600218140107e-05,
+      "loss": 0.9092,
+      "step": 1676
+    },
+    {
+      "epoch": 0.029744257878107038,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.9911365407586774e-05,
+      "loss": 0.8898,
+      "step": 1678
+    },
+    {
+      "epoch": 0.02977970991371861,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.9911130286144805e-05,
+      "loss": 0.8761,
+      "step": 1680
+    },
+    {
+      "epoch": 0.029815161949330177,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.9910894853817106e-05,
+      "loss": 0.8469,
+      "step": 1682
+    },
+    {
+      "epoch": 0.029850613984941748,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.991065911060663e-05,
+      "loss": 0.8584,
+      "step": 1684
+    },
+    {
+      "epoch": 0.02988606602055332,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.991042305651632e-05,
+      "loss": 0.8896,
+      "step": 1686
+    },
+    {
+      "epoch": 0.029921518056164886,
+      "grad_norm": 2.875,
+      "learning_rate": 4.9910186691549123e-05,
+      "loss": 0.8259,
+      "step": 1688
+    },
+    {
+      "epoch": 0.029956970091776457,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.990995001570798e-05,
+      "loss": 0.9028,
+      "step": 1690
+    },
+    {
+      "epoch": 0.02999242212738803,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.9909713028995845e-05,
+      "loss": 0.8824,
+      "step": 1692
+    },
+    {
+      "epoch": 0.030027874162999596,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.9909475731415686e-05,
+      "loss": 0.9096,
+      "step": 1694
+    },
+    {
+      "epoch": 0.030063326198611167,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.990923812297046e-05,
+      "loss": 0.8724,
+      "step": 1696
+    },
+    {
+      "epoch": 0.030098778234222735,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.990900020366313e-05,
+      "loss": 0.8796,
+      "step": 1698
+    },
+    {
+      "epoch": 0.030134230269834306,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.990876197349665e-05,
+      "loss": 0.8311,
+      "step": 1700
+    },
+    {
+      "epoch": 0.030169682305445877,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.9908523432474024e-05,
+      "loss": 0.9301,
+      "step": 1702
+    },
+    {
+      "epoch": 0.030205134341057444,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.9908284580598206e-05,
+      "loss": 0.8491,
+      "step": 1704
+    },
+    {
+      "epoch": 0.030240586376669015,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.99080454178722e-05,
+      "loss": 0.9107,
+      "step": 1706
+    },
+    {
+      "epoch": 0.030276038412280586,
+      "grad_norm": 2.875,
+      "learning_rate": 4.990780594429896e-05,
+      "loss": 0.9217,
+      "step": 1708
+    },
+    {
+      "epoch": 0.030311490447892154,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.990756615988149e-05,
+      "loss": 0.9592,
+      "step": 1710
+    },
+    {
+      "epoch": 0.030346942483503725,
+      "grad_norm": 2.625,
+      "learning_rate": 4.9907326064622786e-05,
+      "loss": 0.8798,
+      "step": 1712
+    },
+    {
+      "epoch": 0.030382394519115296,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.990708565852584e-05,
+      "loss": 0.9163,
+      "step": 1714
+    },
+    {
+      "epoch": 0.030417846554726864,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.9906844941593654e-05,
+      "loss": 0.9099,
+      "step": 1716
+    },
+    {
+      "epoch": 0.030453298590338435,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.990660391382923e-05,
+      "loss": 0.9131,
+      "step": 1718
+    },
+    {
+      "epoch": 0.030488750625950002,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.9906362575235575e-05,
+      "loss": 0.8907,
+      "step": 1720
+    },
+    {
+      "epoch": 0.030524202661561573,
+      "grad_norm": 3.0,
+      "learning_rate": 4.9906120925815706e-05,
+      "loss": 0.8797,
+      "step": 1722
+    },
+    {
+      "epoch": 0.030559654697173144,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.990587896557263e-05,
+      "loss": 0.934,
+      "step": 1724
+    },
+    {
+      "epoch": 0.030595106732784712,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.990563669450938e-05,
+      "loss": 0.8834,
+      "step": 1726
+    },
+    {
+      "epoch": 0.030630558768396283,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.990539411262897e-05,
+      "loss": 0.9206,
+      "step": 1728
+    },
+    {
+      "epoch": 0.030666010804007854,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.990515121993442e-05,
+      "loss": 0.8585,
+      "step": 1730
+    },
+    {
+      "epoch": 0.03070146283961942,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.990490801642878e-05,
+      "loss": 0.8668,
+      "step": 1732
+    },
+    {
+      "epoch": 0.030736914875230992,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.990466450211507e-05,
+      "loss": 0.911,
+      "step": 1734
+    },
+    {
+      "epoch": 0.030772366910842563,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.990442067699634e-05,
+      "loss": 0.8918,
+      "step": 1736
+    },
+    {
+      "epoch": 0.03080781894645413,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.990417654107562e-05,
+      "loss": 0.881,
+      "step": 1738
+    },
+    {
+      "epoch": 0.030843270982065702,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.990393209435596e-05,
+      "loss": 0.9107,
+      "step": 1740
+    },
+    {
+      "epoch": 0.03087872301767727,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.990368733684043e-05,
+      "loss": 0.9067,
+      "step": 1742
+    },
+    {
+      "epoch": 0.03091417505328884,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.9903442268532066e-05,
+      "loss": 0.853,
+      "step": 1744
+    },
+    {
+      "epoch": 0.03094962708890041,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.990319688943392e-05,
+      "loss": 0.8587,
+      "step": 1746
+    },
+    {
+      "epoch": 0.03098507912451198,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.990295119954906e-05,
+      "loss": 0.8913,
+      "step": 1748
+    },
+    {
+      "epoch": 0.03102053116012355,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.990270519888057e-05,
+      "loss": 0.9057,
+      "step": 1750
+    },
+    {
+      "epoch": 0.03105598319573512,
+      "grad_norm": 2.75,
+      "learning_rate": 4.99024588874315e-05,
+      "loss": 0.8899,
+      "step": 1752
+    },
+    {
+      "epoch": 0.03109143523134669,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.990221226520493e-05,
+      "loss": 0.8811,
+      "step": 1754
+    },
+    {
+      "epoch": 0.03112688726695826,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.990196533220394e-05,
+      "loss": 0.8812,
+      "step": 1756
+    },
+    {
+      "epoch": 0.03116233930256983,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.99017180884316e-05,
+      "loss": 0.8764,
+      "step": 1758
+    },
+    {
+      "epoch": 0.0311977913381814,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.9901470533891014e-05,
+      "loss": 0.9297,
+      "step": 1760
+    },
+    {
+      "epoch": 0.03123324337379297,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.9901222668585266e-05,
+      "loss": 0.8929,
+      "step": 1762
+    },
+    {
+      "epoch": 0.03126869540940454,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.9900974492517435e-05,
+      "loss": 0.8731,
+      "step": 1764
+    },
+    {
+      "epoch": 0.03130414744501611,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.990072600569064e-05,
+      "loss": 0.8794,
+      "step": 1766
+    },
+    {
+      "epoch": 0.03133959948062768,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.9900477208107957e-05,
+      "loss": 0.8343,
+      "step": 1768
+    },
+    {
+      "epoch": 0.03137505151623925,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.9900228099772516e-05,
+      "loss": 0.9005,
+      "step": 1770
+    },
+    {
+      "epoch": 0.031410503551850814,
+      "grad_norm": 2.625,
+      "learning_rate": 4.9899978680687406e-05,
+      "loss": 0.879,
+      "step": 1772
+    },
+    {
+      "epoch": 0.03144595558746239,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.9899728950855764e-05,
+      "loss": 0.876,
+      "step": 1774
+    },
+    {
+      "epoch": 0.031481407623073956,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.989947891028067e-05,
+      "loss": 0.9064,
+      "step": 1776
+    },
+    {
+      "epoch": 0.031516859658685524,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.989922855896528e-05,
+      "loss": 0.8831,
+      "step": 1778
+    },
+    {
+      "epoch": 0.0315523116942971,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.98989778969127e-05,
+      "loss": 0.8988,
+      "step": 1780
+    },
+    {
+      "epoch": 0.031587763729908666,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.989872692412606e-05,
+      "loss": 0.8684,
+      "step": 1782
+    },
+    {
+      "epoch": 0.031623215765520234,
+      "grad_norm": 3.578125,
+      "learning_rate": 4.98984756406085e-05,
+      "loss": 0.8947,
+      "step": 1784
+    },
+    {
+      "epoch": 0.03165866780113181,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.989822404636314e-05,
+      "loss": 0.8621,
+      "step": 1786
+    },
+    {
+      "epoch": 0.031694119836743376,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.9897972141393135e-05,
+      "loss": 0.872,
+      "step": 1788
+    },
+    {
+      "epoch": 0.03172957187235494,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.989771992570163e-05,
+      "loss": 0.893,
+      "step": 1790
+    },
+    {
+      "epoch": 0.03176502390796652,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.9897467399291756e-05,
+      "loss": 0.916,
+      "step": 1792
+    },
+    {
+      "epoch": 0.031800475943578085,
+      "grad_norm": 2.375,
+      "learning_rate": 4.989721456216668e-05,
+      "loss": 0.8348,
+      "step": 1794
+    },
+    {
+      "epoch": 0.03183592797918965,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.989696141432955e-05,
+      "loss": 0.8458,
+      "step": 1796
+    },
+    {
+      "epoch": 0.03187138001480123,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.9896707955783526e-05,
+      "loss": 0.9631,
+      "step": 1798
+    },
+    {
+      "epoch": 0.031906832050412795,
+      "grad_norm": 3.0,
+      "learning_rate": 4.989645418653177e-05,
+      "loss": 0.9095,
+      "step": 1800
+    },
+    {
+      "epoch": 0.03194228408602436,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.9896200106577465e-05,
+      "loss": 0.8879,
+      "step": 1802
+    },
+    {
+      "epoch": 0.03197773612163594,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.9895945715923754e-05,
+      "loss": 0.9092,
+      "step": 1804
+    },
+    {
+      "epoch": 0.032013188157247505,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.989569101457383e-05,
+      "loss": 0.843,
+      "step": 1806
+    },
+    {
+      "epoch": 0.03204864019285907,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.989543600253087e-05,
+      "loss": 0.9134,
+      "step": 1808
+    },
+    {
+      "epoch": 0.03208409222847065,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.989518067979805e-05,
+      "loss": 0.8729,
+      "step": 1810
+    },
+    {
+      "epoch": 0.032119544264082214,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.989492504637856e-05,
+      "loss": 0.8983,
+      "step": 1812
+    },
+    {
+      "epoch": 0.03215499629969378,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.989466910227559e-05,
+      "loss": 0.8304,
+      "step": 1814
+    },
+    {
+      "epoch": 0.03219044833530535,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.9894412847492345e-05,
+      "loss": 0.8675,
+      "step": 1816
+    },
+    {
+      "epoch": 0.032225900370916924,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.9894156282032e-05,
+      "loss": 0.9471,
+      "step": 1818
+    },
+    {
+      "epoch": 0.03226135240652849,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.989389940589778e-05,
+      "loss": 0.9131,
+      "step": 1820
+    },
+    {
+      "epoch": 0.03229680444214006,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.989364221909287e-05,
+      "loss": 0.909,
+      "step": 1822
+    },
+    {
+      "epoch": 0.032332256477751634,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.98933847216205e-05,
+      "loss": 0.8465,
+      "step": 1824
+    },
+    {
+      "epoch": 0.0323677085133632,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.989312691348387e-05,
+      "loss": 0.8581,
+      "step": 1826
+    },
+    {
+      "epoch": 0.03240316054897477,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.98928687946862e-05,
+      "loss": 0.9257,
+      "step": 1828
+    },
+    {
+      "epoch": 0.03243861258458634,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.989261036523071e-05,
+      "loss": 0.8431,
+      "step": 1830
+    },
+    {
+      "epoch": 0.03247406462019791,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.989235162512064e-05,
+      "loss": 0.9564,
+      "step": 1832
+    },
+    {
+      "epoch": 0.03250951665580948,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.989209257435919e-05,
+      "loss": 0.8758,
+      "step": 1834
+    },
+    {
+      "epoch": 0.03254496869142105,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.989183321294961e-05,
+      "loss": 0.8783,
+      "step": 1836
+    },
+    {
+      "epoch": 0.03258042072703262,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.989157354089515e-05,
+      "loss": 0.927,
+      "step": 1838
+    },
+    {
+      "epoch": 0.03261587276264419,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.9891313558199025e-05,
+      "loss": 0.9246,
+      "step": 1840
+    },
+    {
+      "epoch": 0.03265132479825576,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.98910532648645e-05,
+      "loss": 0.8639,
+      "step": 1842
+    },
+    {
+      "epoch": 0.03268677683386733,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.9890792660894806e-05,
+      "loss": 0.856,
+      "step": 1844
+    },
+    {
+      "epoch": 0.0327222288694789,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.989053174629321e-05,
+      "loss": 0.8842,
+      "step": 1846
+    },
+    {
+      "epoch": 0.03275768090509047,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.989027052106295e-05,
+      "loss": 0.8997,
+      "step": 1848
+    },
+    {
+      "epoch": 0.03279313294070204,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.989000898520732e-05,
+      "loss": 0.8674,
+      "step": 1850
+    },
+    {
+      "epoch": 0.03282858497631361,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.988974713872955e-05,
+      "loss": 0.8712,
+      "step": 1852
+    },
+    {
+      "epoch": 0.03286403701192518,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.9889484981632913e-05,
+      "loss": 0.9267,
+      "step": 1854
+    },
+    {
+      "epoch": 0.03289948904753675,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.98892225139207e-05,
+      "loss": 0.8959,
+      "step": 1856
+    },
+    {
+      "epoch": 0.03293494108314832,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.9888959735596165e-05,
+      "loss": 0.887,
+      "step": 1858
+    },
+    {
+      "epoch": 0.032970393118759884,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.9888696646662606e-05,
+      "loss": 0.8938,
+      "step": 1860
+    },
+    {
+      "epoch": 0.03300584515437146,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.9888433247123284e-05,
+      "loss": 0.8704,
+      "step": 1862
+    },
+    {
+      "epoch": 0.03304129718998303,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.988816953698151e-05,
+      "loss": 0.926,
+      "step": 1864
+    },
+    {
+      "epoch": 0.033076749225594594,
+      "grad_norm": 2.875,
+      "learning_rate": 4.9887905516240555e-05,
+      "loss": 0.8907,
+      "step": 1866
+    },
+    {
+      "epoch": 0.03311220126120617,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.988764118490373e-05,
+      "loss": 0.9096,
+      "step": 1868
+    },
+    {
+      "epoch": 0.033147653296817736,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.988737654297432e-05,
+      "loss": 0.8758,
+      "step": 1870
+    },
+    {
+      "epoch": 0.033183105332429304,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.988711159045564e-05,
+      "loss": 0.9465,
+      "step": 1872
+    },
+    {
+      "epoch": 0.03321855736804088,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.988684632735099e-05,
+      "loss": 0.8731,
+      "step": 1874
+    },
+    {
+      "epoch": 0.033254009403652446,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.988658075366368e-05,
+      "loss": 0.9215,
+      "step": 1876
+    },
+    {
+      "epoch": 0.03328946143926401,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.9886314869397025e-05,
+      "loss": 0.9032,
+      "step": 1878
+    },
+    {
+      "epoch": 0.03332491347487559,
+      "grad_norm": 2.875,
+      "learning_rate": 4.9886048674554347e-05,
+      "loss": 0.8975,
+      "step": 1880
+    },
+    {
+      "epoch": 0.033360365510487155,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.988578216913896e-05,
+      "loss": 0.9586,
+      "step": 1882
+    },
+    {
+      "epoch": 0.03339581754609872,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.98855153531542e-05,
+      "loss": 0.8882,
+      "step": 1884
+    },
+    {
+      "epoch": 0.0334312695817103,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.9885248226603397e-05,
+      "loss": 0.8855,
+      "step": 1886
+    },
+    {
+      "epoch": 0.033466721617321865,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.9884980789489865e-05,
+      "loss": 0.8807,
+      "step": 1888
+    },
+    {
+      "epoch": 0.03350217365293343,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.988471304181697e-05,
+      "loss": 0.8932,
+      "step": 1890
+    },
+    {
+      "epoch": 0.03353762568854501,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.988444498358803e-05,
+      "loss": 0.8452,
+      "step": 1892
+    },
+    {
+      "epoch": 0.033573077724156575,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.98841766148064e-05,
+      "loss": 0.8801,
+      "step": 1894
+    },
+    {
+      "epoch": 0.03360852975976814,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.9883907935475436e-05,
+      "loss": 0.8889,
+      "step": 1896
+    },
+    {
+      "epoch": 0.03364398179537972,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.988363894559847e-05,
+      "loss": 0.8927,
+      "step": 1898
+    },
+    {
+      "epoch": 0.033679433830991284,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.988336964517889e-05,
+      "loss": 0.8917,
+      "step": 1900
+    },
+    {
+      "epoch": 0.03371488586660285,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.988310003422003e-05,
+      "loss": 0.8618,
+      "step": 1902
+    },
+    {
+      "epoch": 0.03375033790221442,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.9882830112725264e-05,
+      "loss": 0.9061,
+      "step": 1904
+    },
+    {
+      "epoch": 0.033785789937825994,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.9882559880697964e-05,
+      "loss": 0.8871,
+      "step": 1906
+    },
+    {
+      "epoch": 0.03382124197343756,
+      "grad_norm": 2.75,
+      "learning_rate": 4.9882289338141494e-05,
+      "loss": 0.9005,
+      "step": 1908
+    },
+    {
+      "epoch": 0.03385669400904913,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.988201848505925e-05,
+      "loss": 0.8677,
+      "step": 1910
+    },
+    {
+      "epoch": 0.033892146044660704,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.988174732145458e-05,
+      "loss": 0.8617,
+      "step": 1912
+    },
+    {
+      "epoch": 0.03392759808027227,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.988147584733089e-05,
+      "loss": 0.8704,
+      "step": 1914
+    },
+    {
+      "epoch": 0.03396305011588384,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.9881204062691575e-05,
+      "loss": 0.8464,
+      "step": 1916
+    },
+    {
+      "epoch": 0.03399850215149541,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.988093196754001e-05,
+      "loss": 0.9313,
+      "step": 1918
+    },
+    {
+      "epoch": 0.03403395418710698,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.9880659561879596e-05,
+      "loss": 0.8973,
+      "step": 1920
+    },
+    {
+      "epoch": 0.03406940622271855,
+      "grad_norm": 2.75,
+      "learning_rate": 4.988038684571373e-05,
+      "loss": 0.9628,
+      "step": 1922
+    },
+    {
+      "epoch": 0.03410485825833012,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.988011381904581e-05,
+      "loss": 0.87,
+      "step": 1924
+    },
+    {
+      "epoch": 0.03414031029394169,
+      "grad_norm": 3.0,
+      "learning_rate": 4.987984048187927e-05,
+      "loss": 0.9069,
+      "step": 1926
+    },
+    {
+      "epoch": 0.03417576232955326,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.98795668342175e-05,
+      "loss": 0.9091,
+      "step": 1928
+    },
+    {
+      "epoch": 0.03421121436516483,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.987929287606391e-05,
+      "loss": 0.8688,
+      "step": 1930
+    },
+    {
+      "epoch": 0.0342466664007764,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.9879018607421927e-05,
+      "loss": 0.8515,
+      "step": 1932
+    },
+    {
+      "epoch": 0.03428211843638797,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.9878744028294974e-05,
+      "loss": 0.899,
+      "step": 1934
+    },
+    {
+      "epoch": 0.03431757047199954,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.987846913868648e-05,
+      "loss": 0.8567,
+      "step": 1936
+    },
+    {
+      "epoch": 0.03435302250761111,
+      "grad_norm": 3.125,
+      "learning_rate": 4.987819393859987e-05,
+      "loss": 0.8997,
+      "step": 1938
+    },
+    {
+      "epoch": 0.03438847454322268,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.987791842803858e-05,
+      "loss": 0.8767,
+      "step": 1940
+    },
+    {
+      "epoch": 0.03442392657883425,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.9877642607006056e-05,
+      "loss": 0.8827,
+      "step": 1942
+    },
+    {
+      "epoch": 0.03445937861444582,
+      "grad_norm": 3.0,
+      "learning_rate": 4.9877366475505735e-05,
+      "loss": 0.8539,
+      "step": 1944
+    },
+    {
+      "epoch": 0.03449483065005739,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.9877090033541065e-05,
+      "loss": 0.9004,
+      "step": 1946
+    },
+    {
+      "epoch": 0.034530282685668955,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.987681328111548e-05,
+      "loss": 0.8699,
+      "step": 1948
+    },
+    {
+      "epoch": 0.03456573472128053,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.987653621823245e-05,
+      "loss": 0.8887,
+      "step": 1950
+    },
+    {
+      "epoch": 0.0346011867568921,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.987625884489544e-05,
+      "loss": 0.9072,
+      "step": 1952
+    },
+    {
+      "epoch": 0.034636638792503664,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.9875981161107885e-05,
+      "loss": 0.9238,
+      "step": 1954
+    },
+    {
+      "epoch": 0.03467209082811524,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.987570316687328e-05,
+      "loss": 0.9007,
+      "step": 1956
+    },
+    {
+      "epoch": 0.034707542863726806,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.987542486219507e-05,
+      "loss": 0.8812,
+      "step": 1958
+    },
+    {
+      "epoch": 0.034742994899338374,
+      "grad_norm": 3.421875,
+      "learning_rate": 4.987514624707675e-05,
+      "loss": 0.9161,
+      "step": 1960
+    },
+    {
+      "epoch": 0.03477844693494995,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.9874867321521776e-05,
+      "loss": 0.893,
+      "step": 1962
+    },
+    {
+      "epoch": 0.034813898970561516,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.9874588085533644e-05,
+      "loss": 0.8903,
+      "step": 1964
+    },
+    {
+      "epoch": 0.034849351006173084,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.987430853911583e-05,
+      "loss": 0.9052,
+      "step": 1966
+    },
+    {
+      "epoch": 0.03488480304178466,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.987402868227183e-05,
+      "loss": 0.8775,
+      "step": 1968
+    },
+    {
+      "epoch": 0.034920255077396226,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.9873748515005134e-05,
+      "loss": 0.861,
+      "step": 1970
+    },
+    {
+      "epoch": 0.03495570711300779,
+      "grad_norm": 2.75,
+      "learning_rate": 4.987346803731924e-05,
+      "loss": 0.8776,
+      "step": 1972
+    },
+    {
+      "epoch": 0.03499115914861937,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.987318724921764e-05,
+      "loss": 0.8981,
+      "step": 1974
+    },
+    {
+      "epoch": 0.035026611184230935,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.987290615070385e-05,
+      "loss": 0.8984,
+      "step": 1976
+    },
+    {
+      "epoch": 0.0350620632198425,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.987262474178136e-05,
+      "loss": 0.9129,
+      "step": 1978
+    },
+    {
+      "epoch": 0.03509751525545408,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.98723430224537e-05,
+      "loss": 0.9045,
+      "step": 1980
+    },
+    {
+      "epoch": 0.035132967291065645,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.9872060992724387e-05,
+      "loss": 0.8857,
+      "step": 1982
+    },
+    {
+      "epoch": 0.03516841932667721,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.9871778652596926e-05,
+      "loss": 0.8699,
+      "step": 1984
+    },
+    {
+      "epoch": 0.03520387136228878,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.9871496002074845e-05,
+      "loss": 0.8877,
+      "step": 1986
+    },
+    {
+      "epoch": 0.035239323397900355,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.987121304116168e-05,
+      "loss": 0.8973,
+      "step": 1988
+    },
+    {
+      "epoch": 0.03527477543351192,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.987092976986095e-05,
+      "loss": 0.8566,
+      "step": 1990
+    },
+    {
+      "epoch": 0.03531022746912349,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.9870646188176205e-05,
+      "loss": 0.8753,
+      "step": 1992
+    },
+    {
+      "epoch": 0.035345679504735064,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.9870362296110974e-05,
+      "loss": 0.9174,
+      "step": 1994
+    },
+    {
+      "epoch": 0.03538113154034663,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.9870078093668795e-05,
+      "loss": 0.8929,
+      "step": 1996
+    },
+    {
+      "epoch": 0.0354165835759582,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.986979358085323e-05,
+      "loss": 0.8878,
+      "step": 1998
+    },
+    {
+      "epoch": 0.035452035611569774,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.9869508757667816e-05,
+      "loss": 0.912,
+      "step": 2000
+    },
+    {
+      "epoch": 0.03548748764718134,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.986922362411611e-05,
+      "loss": 0.8657,
+      "step": 2002
+    },
+    {
+      "epoch": 0.03552293968279291,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.986893818020168e-05,
+      "loss": 0.8613,
+      "step": 2004
+    },
+    {
+      "epoch": 0.03555839171840448,
+      "grad_norm": 3.0,
+      "learning_rate": 4.986865242592807e-05,
+      "loss": 0.8583,
+      "step": 2006
+    },
+    {
+      "epoch": 0.03559384375401605,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.9868366361298873e-05,
+      "loss": 0.8839,
+      "step": 2008
+    },
+    {
+      "epoch": 0.03562929578962762,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.986807998631764e-05,
+      "loss": 0.8905,
+      "step": 2010
+    },
+    {
+      "epoch": 0.03566474782523919,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.9867793300987945e-05,
+      "loss": 0.8616,
+      "step": 2012
+    },
+    {
+      "epoch": 0.03570019986085076,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.986750630531336e-05,
+      "loss": 0.8757,
+      "step": 2014
+    },
+    {
+      "epoch": 0.03573565189646233,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.9867218999297486e-05,
+      "loss": 0.8668,
+      "step": 2016
+    },
+    {
+      "epoch": 0.0357711039320739,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.98669313829439e-05,
+      "loss": 0.8505,
+      "step": 2018
+    },
+    {
+      "epoch": 0.03580655596768547,
+      "grad_norm": 2.875,
+      "learning_rate": 4.986664345625619e-05,
+      "loss": 0.89,
+      "step": 2020
+    },
+    {
+      "epoch": 0.03584200800329704,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.986635521923794e-05,
+      "loss": 0.8667,
+      "step": 2022
+    },
+    {
+      "epoch": 0.03587746003890861,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.9866066671892765e-05,
+      "loss": 0.954,
+      "step": 2024
+    },
+    {
+      "epoch": 0.03591291207452018,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.9865777814224254e-05,
+      "loss": 0.9229,
+      "step": 2026
+    },
+    {
+      "epoch": 0.03594836411013175,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.9865488646236014e-05,
+      "loss": 0.8738,
+      "step": 2028
+    },
+    {
+      "epoch": 0.035983816145743315,
+      "grad_norm": 3.0,
+      "learning_rate": 4.9865199167931657e-05,
+      "loss": 0.9409,
+      "step": 2030
+    },
+    {
+      "epoch": 0.03601926818135489,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.986490937931479e-05,
+      "loss": 0.8501,
+      "step": 2032
+    },
+    {
+      "epoch": 0.03605472021696646,
+      "grad_norm": 2.625,
+      "learning_rate": 4.9864619280389036e-05,
+      "loss": 0.8757,
+      "step": 2034
+    },
+    {
+      "epoch": 0.036090172252578025,
+      "grad_norm": 2.875,
+      "learning_rate": 4.986432887115801e-05,
+      "loss": 0.8583,
+      "step": 2036
+    },
+    {
+      "epoch": 0.0361256242881896,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.986403815162534e-05,
+      "loss": 0.8564,
+      "step": 2038
+    },
+    {
+      "epoch": 0.03616107632380117,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.9863747121794654e-05,
+      "loss": 0.9092,
+      "step": 2040
+    },
+    {
+      "epoch": 0.036196528359412734,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.986345578166958e-05,
+      "loss": 0.9119,
+      "step": 2042
+    },
+    {
+      "epoch": 0.03623198039502431,
+      "grad_norm": 3.25,
+      "learning_rate": 4.986316413125376e-05,
+      "loss": 0.8834,
+      "step": 2044
+    },
+    {
+      "epoch": 0.036267432430635876,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.9862872170550826e-05,
+      "loss": 0.8866,
+      "step": 2046
+    },
+    {
+      "epoch": 0.036302884466247444,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.9862579899564435e-05,
+      "loss": 0.9096,
+      "step": 2048
+    },
+    {
+      "epoch": 0.03633833650185902,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.9862287318298213e-05,
+      "loss": 0.8794,
+      "step": 2050
+    },
+    {
+      "epoch": 0.036373788537470586,
+      "grad_norm": 2.875,
+      "learning_rate": 4.9861994426755834e-05,
+      "loss": 0.909,
+      "step": 2052
+    },
+    {
+      "epoch": 0.036409240573082154,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.986170122494093e-05,
+      "loss": 0.8997,
+      "step": 2054
+    },
+    {
+      "epoch": 0.03644469260869373,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.9861407712857185e-05,
+      "loss": 0.8905,
+      "step": 2056
+    },
+    {
+      "epoch": 0.036480144644305296,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.9861113890508246e-05,
+      "loss": 0.9056,
+      "step": 2058
+    },
+    {
+      "epoch": 0.03651559667991686,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.986081975789778e-05,
+      "loss": 0.9046,
+      "step": 2060
+    },
+    {
+      "epoch": 0.03655104871552844,
+      "grad_norm": 2.875,
+      "learning_rate": 4.986052531502947e-05,
+      "loss": 0.9152,
+      "step": 2062
+    },
+    {
+      "epoch": 0.036586500751140005,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.9860230561906976e-05,
+      "loss": 0.8837,
+      "step": 2064
+    },
+    {
+      "epoch": 0.03662195278675157,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.985993549853398e-05,
+      "loss": 0.9131,
+      "step": 2066
+    },
+    {
+      "epoch": 0.03665740482236315,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.9859640124914174e-05,
+      "loss": 0.8591,
+      "step": 2068
+    },
+    {
+      "epoch": 0.036692856857974715,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.9859344441051234e-05,
+      "loss": 0.9129,
+      "step": 2070
+    },
+    {
+      "epoch": 0.03672830889358628,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.985904844694885e-05,
+      "loss": 0.8509,
+      "step": 2072
+    },
+    {
+      "epoch": 0.03676376092919785,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.985875214261073e-05,
+      "loss": 0.8665,
+      "step": 2074
+    },
+    {
+      "epoch": 0.036799212964809425,
+      "grad_norm": 2.75,
+      "learning_rate": 4.985845552804055e-05,
+      "loss": 0.8363,
+      "step": 2076
+    },
+    {
+      "epoch": 0.03683466500042099,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.985815860324203e-05,
+      "loss": 0.8305,
+      "step": 2078
+    },
+    {
+      "epoch": 0.03687011703603256,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.985786136821886e-05,
+      "loss": 0.8186,
+      "step": 2080
+    },
+    {
+      "epoch": 0.036905569071644134,
+      "grad_norm": 2.875,
+      "learning_rate": 4.985756382297476e-05,
+      "loss": 0.8541,
+      "step": 2082
+    },
+    {
+      "epoch": 0.0369410211072557,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.985726596751344e-05,
+      "loss": 0.877,
+      "step": 2084
+    },
+    {
+      "epoch": 0.03697647314286727,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.985696780183863e-05,
+      "loss": 0.9137,
+      "step": 2086
+    },
+    {
+      "epoch": 0.037011925178478844,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.985666932595403e-05,
+      "loss": 0.89,
+      "step": 2088
+    },
+    {
+      "epoch": 0.03704737721409041,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.985637053986337e-05,
+      "loss": 0.8867,
+      "step": 2090
+    },
+    {
+      "epoch": 0.03708282924970198,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.985607144357039e-05,
+      "loss": 0.8978,
+      "step": 2092
+    },
+    {
+      "epoch": 0.037118281285313554,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.9855772037078814e-05,
+      "loss": 0.8873,
+      "step": 2094
+    },
+    {
+      "epoch": 0.03715373332092512,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.9855472320392375e-05,
+      "loss": 0.9264,
+      "step": 2096
+    },
+    {
+      "epoch": 0.03718918535653669,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.9855172293514826e-05,
+      "loss": 0.8793,
+      "step": 2098
+    },
+    {
+      "epoch": 0.03722463739214826,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.98548719564499e-05,
+      "loss": 0.8243,
+      "step": 2100
+    },
+    {
+      "epoch": 0.03726008942775983,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.985457130920135e-05,
+      "loss": 0.8314,
+      "step": 2102
+    },
+    {
+      "epoch": 0.0372955414633714,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.9854270351772925e-05,
+      "loss": 0.8263,
+      "step": 2104
+    },
+    {
+      "epoch": 0.03733099349898297,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.9853969084168386e-05,
+      "loss": 0.8607,
+      "step": 2106
+    },
+    {
+      "epoch": 0.03736644553459454,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.985366750639148e-05,
+      "loss": 0.8569,
+      "step": 2108
+    },
+    {
+      "epoch": 0.03740189757020611,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.985336561844599e-05,
+      "loss": 0.8951,
+      "step": 2110
+    },
+    {
+      "epoch": 0.03743734960581768,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.985306342033567e-05,
+      "loss": 0.9011,
+      "step": 2112
+    },
+    {
+      "epoch": 0.03747280164142925,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.98527609120643e-05,
+      "loss": 0.884,
+      "step": 2114
+    },
+    {
+      "epoch": 0.03750825367704082,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.9852458093635646e-05,
+      "loss": 0.8846,
+      "step": 2116
+    },
+    {
+      "epoch": 0.037543705712652385,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.985215496505349e-05,
+      "loss": 0.8899,
+      "step": 2118
+    },
+    {
+      "epoch": 0.03757915774826396,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.985185152632162e-05,
+      "loss": 0.8511,
+      "step": 2120
+    },
+    {
+      "epoch": 0.03761460978387553,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.985154777744382e-05,
+      "loss": 0.8845,
+      "step": 2122
+    },
+    {
+      "epoch": 0.037650061819487095,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.9851243718423875e-05,
+      "loss": 0.8729,
+      "step": 2124
+    },
+    {
+      "epoch": 0.03768551385509867,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.9850939349265587e-05,
+      "loss": 0.8871,
+      "step": 2126
+    },
+    {
+      "epoch": 0.03772096589071024,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.985063466997275e-05,
+      "loss": 0.9002,
+      "step": 2128
+    },
+    {
+      "epoch": 0.037756417926321804,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.985032968054917e-05,
+      "loss": 0.8798,
+      "step": 2130
+    },
+    {
+      "epoch": 0.03779186996193338,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.985002438099865e-05,
+      "loss": 0.9102,
+      "step": 2132
+    },
+    {
+      "epoch": 0.037827321997544947,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.984971877132501e-05,
+      "loss": 0.8697,
+      "step": 2134
+    },
+    {
+      "epoch": 0.037862774033156514,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.984941285153204e-05,
+      "loss": 0.9321,
+      "step": 2136
+    },
+    {
+      "epoch": 0.03789822606876809,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.984910662162359e-05,
+      "loss": 0.8666,
+      "step": 2138
+    },
+    {
+      "epoch": 0.037933678104379656,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.9848800081603455e-05,
+      "loss": 0.8904,
+      "step": 2140
+    },
+    {
+      "epoch": 0.037969130139991224,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.9848493231475466e-05,
+      "loss": 0.9229,
+      "step": 2142
+    },
+    {
+      "epoch": 0.0380045821756028,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.9848186071243463e-05,
+      "loss": 0.8472,
+      "step": 2144
+    },
+    {
+      "epoch": 0.038040034211214366,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.984787860091128e-05,
+      "loss": 0.8923,
+      "step": 2146
+    },
+    {
+      "epoch": 0.03807548624682593,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.984757082048274e-05,
+      "loss": 0.8802,
+      "step": 2148
+    },
+    {
+      "epoch": 0.03811093828243751,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.9847262729961694e-05,
+      "loss": 0.8569,
+      "step": 2150
+    },
+    {
+      "epoch": 0.038146390318049075,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.984695432935198e-05,
+      "loss": 0.9036,
+      "step": 2152
+    },
+    {
+      "epoch": 0.03818184235366064,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.9846645618657463e-05,
+      "loss": 0.8576,
+      "step": 2154
+    },
+    {
+      "epoch": 0.03821729438927221,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.984633659788197e-05,
+      "loss": 0.8861,
+      "step": 2156
+    },
+    {
+      "epoch": 0.038252746424883785,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.984602726702938e-05,
+      "loss": 0.9025,
+      "step": 2158
+    },
+    {
+      "epoch": 0.03828819846049535,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.984571762610354e-05,
+      "loss": 0.8429,
+      "step": 2160
+    },
+    {
+      "epoch": 0.03832365049610692,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.984540767510833e-05,
+      "loss": 0.8757,
+      "step": 2162
+    },
+    {
+      "epoch": 0.038359102531718495,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.984509741404759e-05,
+      "loss": 0.8641,
+      "step": 2164
+    },
+    {
+      "epoch": 0.03839455456733006,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.984478684292522e-05,
+      "loss": 0.8471,
+      "step": 2166
+    },
+    {
+      "epoch": 0.03843000660294163,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.984447596174507e-05,
+      "loss": 0.8905,
+      "step": 2168
+    },
+    {
+      "epoch": 0.038465458638553204,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.984416477051106e-05,
+      "loss": 0.8571,
+      "step": 2170
+    },
+    {
+      "epoch": 0.03850091067416477,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.984385326922703e-05,
+      "loss": 0.8629,
+      "step": 2172
+    },
+    {
+      "epoch": 0.03853636270977634,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.9843541457896894e-05,
+      "loss": 0.8954,
+      "step": 2174
+    },
+    {
+      "epoch": 0.038571814745387914,
+      "grad_norm": 2.75,
+      "learning_rate": 4.9843229336524526e-05,
+      "loss": 0.8768,
+      "step": 2176
+    },
+    {
+      "epoch": 0.03860726678099948,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.984291690511384e-05,
+      "loss": 0.8464,
+      "step": 2178
+    },
+    {
+      "epoch": 0.03864271881661105,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.984260416366872e-05,
+      "loss": 0.9009,
+      "step": 2180
+    },
+    {
+      "epoch": 0.038678170852222624,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.9842291112193076e-05,
+      "loss": 0.8651,
+      "step": 2182
+    },
+    {
+      "epoch": 0.03871362288783419,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.9841977750690815e-05,
+      "loss": 0.8651,
+      "step": 2184
+    },
+    {
+      "epoch": 0.03874907492344576,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.984166407916584e-05,
+      "loss": 0.8658,
+      "step": 2186
+    },
+    {
+      "epoch": 0.03878452695905733,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.984135009762208e-05,
+      "loss": 0.9066,
+      "step": 2188
+    },
+    {
+      "epoch": 0.0388199789946689,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.984103580606344e-05,
+      "loss": 0.9346,
+      "step": 2190
+    },
+    {
+      "epoch": 0.03885543103028047,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.984072120449385e-05,
+      "loss": 0.8453,
+      "step": 2192
+    },
+    {
+      "epoch": 0.03889088306589204,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.984040629291723e-05,
+      "loss": 0.8937,
+      "step": 2194
+    },
+    {
+      "epoch": 0.03892633510150361,
+      "grad_norm": 2.875,
+      "learning_rate": 4.9840091071337514e-05,
+      "loss": 0.844,
+      "step": 2196
+    },
+    {
+      "epoch": 0.03896178713711518,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.9839775539758635e-05,
+      "loss": 0.843,
+      "step": 2198
+    },
+    {
+      "epoch": 0.038997239172726746,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.983945969818453e-05,
+      "loss": 0.8641,
+      "step": 2200
+    },
+    {
+      "epoch": 0.03903269120833832,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.9839143546619146e-05,
+      "loss": 0.9192,
+      "step": 2202
+    },
+    {
+      "epoch": 0.03906814324394989,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.983882708506642e-05,
+      "loss": 0.8461,
+      "step": 2204
+    },
+    {
+      "epoch": 0.039103595279561455,
+      "grad_norm": 2.625,
+      "learning_rate": 4.9838510313530304e-05,
+      "loss": 0.8884,
+      "step": 2206
+    },
+    {
+      "epoch": 0.03913904731517303,
+      "grad_norm": 2.875,
+      "learning_rate": 4.983819323201476e-05,
+      "loss": 0.8661,
+      "step": 2208
+    },
+    {
+      "epoch": 0.0391744993507846,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.9837875840523734e-05,
+      "loss": 0.9119,
+      "step": 2210
+    },
+    {
+      "epoch": 0.039209951386396165,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.983755813906119e-05,
+      "loss": 0.929,
+      "step": 2212
+    },
+    {
+      "epoch": 0.03924540342200774,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.9837240127631094e-05,
+      "loss": 0.8929,
+      "step": 2214
+    },
+    {
+      "epoch": 0.03928085545761931,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.9836921806237416e-05,
+      "loss": 0.8669,
+      "step": 2216
+    },
+    {
+      "epoch": 0.039316307493230875,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.983660317488412e-05,
+      "loss": 0.8984,
+      "step": 2218
+    },
+    {
+      "epoch": 0.03935175952884245,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.98362842335752e-05,
+      "loss": 0.9475,
+      "step": 2220
+    },
+    {
+      "epoch": 0.03938721156445402,
+      "grad_norm": 2.875,
+      "learning_rate": 4.983596498231462e-05,
+      "loss": 0.8664,
+      "step": 2222
+    },
+    {
+      "epoch": 0.039422663600065584,
+      "grad_norm": 2.625,
+      "learning_rate": 4.983564542110637e-05,
+      "loss": 0.89,
+      "step": 2224
+    },
+    {
+      "epoch": 0.03945811563567716,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.9835325549954446e-05,
+      "loss": 0.864,
+      "step": 2226
+    },
+    {
+      "epoch": 0.039493567671288726,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.983500536886282e-05,
+      "loss": 0.8795,
+      "step": 2228
+    },
+    {
+      "epoch": 0.039529019706900294,
+      "grad_norm": 3.125,
+      "learning_rate": 4.9834684877835506e-05,
+      "loss": 0.894,
+      "step": 2230
+    },
+    {
+      "epoch": 0.03956447174251187,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.9834364076876493e-05,
+      "loss": 0.8737,
+      "step": 2232
+    },
+    {
+      "epoch": 0.039599923778123436,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.983404296598979e-05,
+      "loss": 0.866,
+      "step": 2234
+    },
+    {
+      "epoch": 0.039635375813735,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.9833721545179415e-05,
+      "loss": 0.8949,
+      "step": 2236
+    },
+    {
+      "epoch": 0.03967082784934658,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.983339981444936e-05,
+      "loss": 0.8475,
+      "step": 2238
+    },
+    {
+      "epoch": 0.039706279884958146,
+      "grad_norm": 3.0,
+      "learning_rate": 4.9833077773803635e-05,
+      "loss": 0.868,
+      "step": 2240
+    },
+    {
+      "epoch": 0.03974173192056971,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.983275542324629e-05,
+      "loss": 0.8695,
+      "step": 2242
+    },
+    {
+      "epoch": 0.03977718395618128,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.983243276278132e-05,
+      "loss": 0.9194,
+      "step": 2244
+    },
+    {
+      "epoch": 0.039812635991792855,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.9832109792412764e-05,
+      "loss": 0.8362,
+      "step": 2246
+    },
+    {
+      "epoch": 0.03984808802740442,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.983178651214465e-05,
+      "loss": 0.9008,
+      "step": 2248
+    },
+    {
+      "epoch": 0.03988354006301599,
+      "grad_norm": 3.25,
+      "learning_rate": 4.983146292198101e-05,
+      "loss": 0.8607,
+      "step": 2250
+    },
+    {
+      "epoch": 0.039918992098627565,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.9831139021925886e-05,
+      "loss": 0.9423,
+      "step": 2252
+    },
+    {
+      "epoch": 0.03995444413423913,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.9830814811983316e-05,
+      "loss": 0.9248,
+      "step": 2254
+    },
+    {
+      "epoch": 0.0399898961698507,
+      "grad_norm": 3.0,
+      "learning_rate": 4.983049029215735e-05,
+      "loss": 0.9025,
+      "step": 2256
+    },
+    {
+      "epoch": 0.040025348205462274,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.983016546245205e-05,
+      "loss": 0.8686,
+      "step": 2258
+    },
+    {
+      "epoch": 0.04006080024107384,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.982984032287145e-05,
+      "loss": 0.8586,
+      "step": 2260
+    },
+    {
+      "epoch": 0.04009625227668541,
+      "grad_norm": 2.875,
+      "learning_rate": 4.98295148734196e-05,
+      "loss": 0.8336,
+      "step": 2262
+    },
+    {
+      "epoch": 0.040131704312296984,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.982918911410059e-05,
+      "loss": 0.8966,
+      "step": 2264
+    },
+    {
+      "epoch": 0.04016715634790855,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.982886304491847e-05,
+      "loss": 0.8924,
+      "step": 2266
+    },
+    {
+      "epoch": 0.04020260838352012,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.9828536665877304e-05,
+      "loss": 0.8736,
+      "step": 2268
+    },
+    {
+      "epoch": 0.040238060419131694,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.9828209976981174e-05,
+      "loss": 0.864,
+      "step": 2270
+    },
+    {
+      "epoch": 0.04027351245474326,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.9827882978234165e-05,
+      "loss": 0.8624,
+      "step": 2272
+    },
+    {
+      "epoch": 0.04030896449035483,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.9827555669640335e-05,
+      "loss": 0.8635,
+      "step": 2274
+    },
+    {
+      "epoch": 0.0403444165259664,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.982722805120379e-05,
+      "loss": 0.8958,
+      "step": 2276
+    },
+    {
+      "epoch": 0.04037986856157797,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.982690012292861e-05,
+      "loss": 0.9135,
+      "step": 2278
+    },
+    {
+      "epoch": 0.04041532059718954,
+      "grad_norm": 2.875,
+      "learning_rate": 4.9826571884818886e-05,
+      "loss": 0.8516,
+      "step": 2280
+    },
+    {
+      "epoch": 0.04045077263280111,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.982624333687871e-05,
+      "loss": 0.8412,
+      "step": 2282
+    },
+    {
+      "epoch": 0.04048622466841268,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.982591447911219e-05,
+      "loss": 0.9317,
+      "step": 2284
+    },
+    {
+      "epoch": 0.04052167670402425,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.982558531152343e-05,
+      "loss": 0.8704,
+      "step": 2286
+    },
+    {
+      "epoch": 0.040557128739635816,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.982525583411654e-05,
+      "loss": 0.8408,
+      "step": 2288
+    },
+    {
+      "epoch": 0.04059258077524739,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.982492604689562e-05,
+      "loss": 0.8701,
+      "step": 2290
+    },
+    {
+      "epoch": 0.04062803281085896,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.98245959498648e-05,
+      "loss": 0.8691,
+      "step": 2292
+    },
+    {
+      "epoch": 0.040663484846470525,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.9824265543028195e-05,
+      "loss": 0.8744,
+      "step": 2294
+    },
+    {
+      "epoch": 0.0406989368820821,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.9823934826389916e-05,
+      "loss": 0.863,
+      "step": 2296
+    },
+    {
+      "epoch": 0.04073438891769367,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.982360379995411e-05,
+      "loss": 0.8763,
+      "step": 2298
+    },
+    {
+      "epoch": 0.040769840953305235,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.98232724637249e-05,
+      "loss": 0.8696,
+      "step": 2300
+    },
+    {
+      "epoch": 0.04080529298891681,
+      "grad_norm": 2.875,
+      "learning_rate": 4.982294081770641e-05,
+      "loss": 0.9117,
+      "step": 2302
+    },
+    {
+      "epoch": 0.04084074502452838,
+      "grad_norm": 2.75,
+      "learning_rate": 4.9822608861902795e-05,
+      "loss": 0.9322,
+      "step": 2304
+    },
+    {
+      "epoch": 0.040876197060139945,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.9822276596318195e-05,
+      "loss": 0.8368,
+      "step": 2306
+    },
+    {
+      "epoch": 0.04091164909575152,
+      "grad_norm": 3.0,
+      "learning_rate": 4.982194402095675e-05,
+      "loss": 0.8842,
+      "step": 2308
+    },
+    {
+      "epoch": 0.04094710113136309,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.9821611135822615e-05,
+      "loss": 0.8414,
+      "step": 2310
+    },
+    {
+      "epoch": 0.040982553166974654,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.982127794091994e-05,
+      "loss": 0.8953,
+      "step": 2312
+    },
+    {
+      "epoch": 0.04101800520258623,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.982094443625289e-05,
+      "loss": 0.9333,
+      "step": 2314
+    },
+    {
+      "epoch": 0.041053457238197796,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.982061062182562e-05,
+      "loss": 0.876,
+      "step": 2316
+    },
+    {
+      "epoch": 0.041088909273809364,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.98202764976423e-05,
+      "loss": 0.8729,
+      "step": 2318
+    },
+    {
+      "epoch": 0.04112436130942094,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.9819942063707105e-05,
+      "loss": 0.9321,
+      "step": 2320
+    },
+    {
+      "epoch": 0.041159813345032506,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.981960732002419e-05,
+      "loss": 0.867,
+      "step": 2322
+    },
+    {
+      "epoch": 0.041195265380644074,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.9819272266597755e-05,
+      "loss": 0.8381,
+      "step": 2324
+    },
+    {
+      "epoch": 0.04123071741625565,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.981893690343197e-05,
+      "loss": 0.847,
+      "step": 2326
+    },
+    {
+      "epoch": 0.041266169451867216,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.981860123053102e-05,
+      "loss": 0.8989,
+      "step": 2328
+    },
+    {
+      "epoch": 0.04130162148747878,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.98182652478991e-05,
+      "loss": 0.8794,
+      "step": 2330
+    },
+    {
+      "epoch": 0.04133707352309035,
+      "grad_norm": 2.625,
+      "learning_rate": 4.9817928955540395e-05,
+      "loss": 0.8438,
+      "step": 2332
+    },
+    {
+      "epoch": 0.041372525558701925,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.9817592353459104e-05,
+      "loss": 0.8878,
+      "step": 2334
+    },
+    {
+      "epoch": 0.04140797759431349,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.981725544165943e-05,
+      "loss": 0.8578,
+      "step": 2336
+    },
+    {
+      "epoch": 0.04144342962992506,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.981691822014558e-05,
+      "loss": 0.8464,
+      "step": 2338
+    },
+    {
+      "epoch": 0.041478881665536635,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.981658068892176e-05,
+      "loss": 0.8744,
+      "step": 2340
+    },
+    {
+      "epoch": 0.0415143337011482,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.9816242847992176e-05,
+      "loss": 0.8701,
+      "step": 2342
+    },
+    {
+      "epoch": 0.04154978573675977,
+      "grad_norm": 2.75,
+      "learning_rate": 4.981590469736106e-05,
+      "loss": 0.9046,
+      "step": 2344
+    },
+    {
+      "epoch": 0.041585237772371345,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.981556623703262e-05,
+      "loss": 0.841,
+      "step": 2346
+    },
+    {
+      "epoch": 0.04162068980798291,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.981522746701107e-05,
+      "loss": 0.9096,
+      "step": 2348
+    },
+    {
+      "epoch": 0.04165614184359448,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.981488838730066e-05,
+      "loss": 0.8283,
+      "step": 2350
+    },
+    {
+      "epoch": 0.041691593879206054,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.9814548997905616e-05,
+      "loss": 0.8977,
+      "step": 2352
+    },
+    {
+      "epoch": 0.04172704591481762,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.981420929883016e-05,
+      "loss": 0.8914,
+      "step": 2354
+    },
+    {
+      "epoch": 0.04176249795042919,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.981386929007854e-05,
+      "loss": 0.8514,
+      "step": 2356
+    },
+    {
+      "epoch": 0.041797949986040764,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.981352897165501e-05,
+      "loss": 0.8564,
+      "step": 2358
+    },
+    {
+      "epoch": 0.04183340202165233,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.98131883435638e-05,
+      "loss": 0.9608,
+      "step": 2360
+    },
+    {
+      "epoch": 0.0418688540572639,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.981284740580916e-05,
+      "loss": 0.9024,
+      "step": 2362
+    },
+    {
+      "epoch": 0.041904306092875473,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.9812506158395365e-05,
+      "loss": 0.906,
+      "step": 2364
+    },
+    {
+      "epoch": 0.04193975812848704,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.981216460132666e-05,
+      "loss": 0.8399,
+      "step": 2366
+    },
+    {
+      "epoch": 0.04197521016409861,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.9811822734607304e-05,
+      "loss": 0.8214,
+      "step": 2368
+    },
+    {
+      "epoch": 0.042010662199710176,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.981148055824157e-05,
+      "loss": 0.8488,
+      "step": 2370
+    },
+    {
+      "epoch": 0.04204611423532175,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.981113807223372e-05,
+      "loss": 0.9123,
+      "step": 2372
+    },
+    {
+      "epoch": 0.04208156627093332,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.981079527658804e-05,
+      "loss": 0.8889,
+      "step": 2374
+    },
+    {
+      "epoch": 0.042117018306544886,
+      "grad_norm": 2.5,
+      "learning_rate": 4.981045217130881e-05,
+      "loss": 0.871,
+      "step": 2376
+    },
+    {
+      "epoch": 0.04215247034215646,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.9810108756400294e-05,
+      "loss": 0.8747,
+      "step": 2378
+    },
+    {
+      "epoch": 0.04218792237776803,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.980976503186679e-05,
+      "loss": 0.831,
+      "step": 2380
+    },
+    {
+      "epoch": 0.042223374413379595,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.980942099771259e-05,
+      "loss": 0.8865,
+      "step": 2382
+    },
+    {
+      "epoch": 0.04225882644899117,
+      "grad_norm": 3.125,
+      "learning_rate": 4.980907665394198e-05,
+      "loss": 0.8713,
+      "step": 2384
+    },
+    {
+      "epoch": 0.04229427848460274,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.980873200055927e-05,
+      "loss": 0.8802,
+      "step": 2386
+    },
+    {
+      "epoch": 0.042329730520214305,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.980838703756874e-05,
+      "loss": 0.8807,
+      "step": 2388
+    },
+    {
+      "epoch": 0.04236518255582588,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.980804176497471e-05,
+      "loss": 0.8972,
+      "step": 2390
+    },
+    {
+      "epoch": 0.04240063459143745,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.980769618278149e-05,
+      "loss": 0.8419,
+      "step": 2392
+    },
+    {
+      "epoch": 0.042436086627049015,
+      "grad_norm": 2.75,
+      "learning_rate": 4.980735029099338e-05,
+      "loss": 0.8692,
+      "step": 2394
+    },
+    {
+      "epoch": 0.04247153866266059,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.9807004089614714e-05,
+      "loss": 0.8991,
+      "step": 2396
+    },
+    {
+      "epoch": 0.04250699069827216,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.98066575786498e-05,
+      "loss": 0.9319,
+      "step": 2398
+    },
+    {
+      "epoch": 0.042542442733883724,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.980631075810297e-05,
+      "loss": 0.8623,
+      "step": 2400
+    },
+    {
+      "epoch": 0.0425778947694953,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.980596362797855e-05,
+      "loss": 0.8915,
+      "step": 2402
+    },
+    {
+      "epoch": 0.042613346805106866,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.9805616188280855e-05,
+      "loss": 0.843,
+      "step": 2404
+    },
+    {
+      "epoch": 0.042648798840718434,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.980526843901425e-05,
+      "loss": 0.9275,
+      "step": 2406
+    },
+    {
+      "epoch": 0.04268425087633001,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.9804920380183064e-05,
+      "loss": 0.8761,
+      "step": 2408
+    },
+    {
+      "epoch": 0.042719702911941576,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.980457201179163e-05,
+      "loss": 0.8647,
+      "step": 2410
+    },
+    {
+      "epoch": 0.042755154947553144,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.980422333384431e-05,
+      "loss": 0.8317,
+      "step": 2412
+    },
+    {
+      "epoch": 0.04279060698316471,
+      "grad_norm": 2.625,
+      "learning_rate": 4.9803874346345445e-05,
+      "loss": 0.8614,
+      "step": 2414
+    },
+    {
+      "epoch": 0.042826059018776286,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.980352504929939e-05,
+      "loss": 0.8696,
+      "step": 2416
+    },
+    {
+      "epoch": 0.04286151105438785,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.980317544271052e-05,
+      "loss": 0.8983,
+      "step": 2418
+    },
+    {
+      "epoch": 0.04289696308999942,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.980282552658318e-05,
+      "loss": 0.8849,
+      "step": 2420
+    },
+    {
+      "epoch": 0.042932415125610995,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.9802475300921744e-05,
+      "loss": 0.8875,
+      "step": 2422
+    },
+    {
+      "epoch": 0.04296786716122256,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.9802124765730586e-05,
+      "loss": 0.8473,
+      "step": 2424
+    },
+    {
+      "epoch": 0.04300331919683413,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.980177392101407e-05,
+      "loss": 0.8613,
+      "step": 2426
+    },
+    {
+      "epoch": 0.043038771232445705,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.980142276677658e-05,
+      "loss": 0.8858,
+      "step": 2428
+    },
+    {
+      "epoch": 0.04307422326805727,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.9801071303022514e-05,
+      "loss": 0.8649,
+      "step": 2430
+    },
+    {
+      "epoch": 0.04310967530366884,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.9800719529756236e-05,
+      "loss": 0.8845,
+      "step": 2432
+    },
+    {
+      "epoch": 0.043145127339280415,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.980036744698214e-05,
+      "loss": 0.8627,
+      "step": 2434
+    },
+    {
+      "epoch": 0.04318057937489198,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.9800015054704625e-05,
+      "loss": 0.878,
+      "step": 2436
+    },
+    {
+      "epoch": 0.04321603141050355,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.979966235292809e-05,
+      "loss": 0.8916,
+      "step": 2438
+    },
+    {
+      "epoch": 0.043251483446115124,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.979930934165693e-05,
+      "loss": 0.8307,
+      "step": 2440
+    },
+    {
+      "epoch": 0.04328693548172669,
+      "grad_norm": 2.75,
+      "learning_rate": 4.979895602089556e-05,
+      "loss": 0.8751,
+      "step": 2442
+    },
+    {
+      "epoch": 0.04332238751733826,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.979860239064838e-05,
+      "loss": 0.8464,
+      "step": 2444
+    },
+    {
+      "epoch": 0.043357839552949834,
+      "grad_norm": 3.375,
+      "learning_rate": 4.9798248450919804e-05,
+      "loss": 0.8977,
+      "step": 2446
+    },
+    {
+      "epoch": 0.0433932915885614,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.9797894201714266e-05,
+      "loss": 0.8583,
+      "step": 2448
+    },
+    {
+      "epoch": 0.04342874362417297,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.979753964303616e-05,
+      "loss": 0.9041,
+      "step": 2450
+    },
+    {
+      "epoch": 0.043464195659784544,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.9797184774889926e-05,
+      "loss": 0.8823,
+      "step": 2452
+    },
+    {
+      "epoch": 0.04349964769539611,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.979682959728e-05,
+      "loss": 0.8753,
+      "step": 2454
+    },
+    {
+      "epoch": 0.04353509973100768,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.97964741102108e-05,
+      "loss": 0.9116,
+      "step": 2456
+    },
+    {
+      "epoch": 0.043570551766619246,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.9796118313686765e-05,
+      "loss": 0.8646,
+      "step": 2458
+    },
+    {
+      "epoch": 0.04360600380223082,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.979576220771234e-05,
+      "loss": 0.852,
+      "step": 2460
+    },
+    {
+      "epoch": 0.04364145583784239,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.979540579229197e-05,
+      "loss": 0.8815,
+      "step": 2462
+    },
+    {
+      "epoch": 0.043676907873453956,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.979504906743009e-05,
+      "loss": 0.8904,
+      "step": 2464
+    },
+    {
+      "epoch": 0.04371235990906553,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.9794692033131176e-05,
+      "loss": 0.8501,
+      "step": 2466
+    },
+    {
+      "epoch": 0.0437478119446771,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.979433468939966e-05,
+      "loss": 0.8315,
+      "step": 2468
+    },
+    {
+      "epoch": 0.043783263980288666,
+      "grad_norm": 2.875,
+      "learning_rate": 4.9793977036240015e-05,
+      "loss": 0.8595,
+      "step": 2470
+    },
+    {
+      "epoch": 0.04381871601590024,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.979361907365671e-05,
+      "loss": 0.8838,
+      "step": 2472
+    },
+    {
+      "epoch": 0.04385416805151181,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.979326080165419e-05,
+      "loss": 0.8831,
+      "step": 2474
+    },
+    {
+      "epoch": 0.043889620087123375,
+      "grad_norm": 2.5,
+      "learning_rate": 4.9792902220236936e-05,
+      "loss": 0.8185,
+      "step": 2476
+    },
+    {
+      "epoch": 0.04392507212273495,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.979254332940944e-05,
+      "loss": 0.8773,
+      "step": 2478
+    },
+    {
+      "epoch": 0.04396052415834652,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.9792184129176153e-05,
+      "loss": 0.8821,
+      "step": 2480
+    },
+    {
+      "epoch": 0.043995976193958085,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.979182461954158e-05,
+      "loss": 0.8751,
+      "step": 2482
+    },
+    {
+      "epoch": 0.04403142822956966,
+      "grad_norm": 3.25,
+      "learning_rate": 4.97914648005102e-05,
+      "loss": 0.9332,
+      "step": 2484
+    },
+    {
+      "epoch": 0.04406688026518123,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.979110467208651e-05,
+      "loss": 0.8437,
+      "step": 2486
+    },
+    {
+      "epoch": 0.044102332300792794,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.9790744234274986e-05,
+      "loss": 0.8874,
+      "step": 2488
+    },
+    {
+      "epoch": 0.04413778433640437,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.979038348708014e-05,
+      "loss": 0.8942,
+      "step": 2490
+    },
+    {
+      "epoch": 0.04417323637201594,
+      "grad_norm": 3.25,
+      "learning_rate": 4.9790022430506463e-05,
+      "loss": 0.8691,
+      "step": 2492
+    },
+    {
+      "epoch": 0.044208688407627504,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.978966106455848e-05,
+      "loss": 0.9051,
+      "step": 2494
+    },
+    {
+      "epoch": 0.04424414044323908,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.9789299389240694e-05,
+      "loss": 0.8664,
+      "step": 2496
+    },
+    {
+      "epoch": 0.044279592478850646,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.9788937404557615e-05,
+      "loss": 0.8914,
+      "step": 2498
+    },
+    {
+      "epoch": 0.044315044514462214,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.9788575110513755e-05,
+      "loss": 0.8917,
+      "step": 2500
+    },
+    {
+      "epoch": 0.04435049655007378,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.978821250711364e-05,
+      "loss": 0.8945,
+      "step": 2502
+    },
+    {
+      "epoch": 0.044385948585685356,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.978784959436179e-05,
+      "loss": 0.8685,
+      "step": 2504
+    },
+    {
+      "epoch": 0.04442140062129692,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.9787486372262746e-05,
+      "loss": 0.8572,
+      "step": 2506
+    },
+    {
+      "epoch": 0.04445685265690849,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.978712284082104e-05,
+      "loss": 0.8378,
+      "step": 2508
+    },
+    {
+      "epoch": 0.044492304692520065,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.9786759000041197e-05,
+      "loss": 0.8705,
+      "step": 2510
+    },
+    {
+      "epoch": 0.04452775672813163,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.978639484992777e-05,
+      "loss": 0.874,
+      "step": 2512
+    },
+    {
+      "epoch": 0.0445632087637432,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.9786030390485295e-05,
+      "loss": 0.9042,
+      "step": 2514
+    },
+    {
+      "epoch": 0.044598660799354775,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.9785665621718325e-05,
+      "loss": 0.8862,
+      "step": 2516
+    },
+    {
+      "epoch": 0.04463411283496634,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.978530054363141e-05,
+      "loss": 0.8207,
+      "step": 2518
+    },
+    {
+      "epoch": 0.04466956487057791,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.97849351562291e-05,
+      "loss": 0.8353,
+      "step": 2520
+    },
+    {
+      "epoch": 0.044705016906189485,
+      "grad_norm": 2.875,
+      "learning_rate": 4.978456945951597e-05,
+      "loss": 0.8698,
+      "step": 2522
+    },
+    {
+      "epoch": 0.04474046894180105,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.978420345349657e-05,
+      "loss": 0.9051,
+      "step": 2524
+    },
+    {
+      "epoch": 0.04477592097741262,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.978383713817548e-05,
+      "loss": 0.8803,
+      "step": 2526
+    },
+    {
+      "epoch": 0.044811373013024194,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.9783470513557255e-05,
+      "loss": 0.8891,
+      "step": 2528
+    },
+    {
+      "epoch": 0.04484682504863576,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.978310357964648e-05,
+      "loss": 0.8661,
+      "step": 2530
+    },
+    {
+      "epoch": 0.04488227708424733,
+      "grad_norm": 3.25,
+      "learning_rate": 4.978273633644775e-05,
+      "loss": 0.8592,
+      "step": 2532
+    },
+    {
+      "epoch": 0.044917729119858904,
+      "grad_norm": 3.125,
+      "learning_rate": 4.978236878396562e-05,
+      "loss": 0.8831,
+      "step": 2534
+    },
+    {
+      "epoch": 0.04495318115547047,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.978200092220469e-05,
+      "loss": 0.9059,
+      "step": 2536
+    },
+    {
+      "epoch": 0.04498863319108204,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.978163275116955e-05,
+      "loss": 0.9341,
+      "step": 2538
+    },
+    {
+      "epoch": 0.045024085226693614,
+      "grad_norm": 2.625,
+      "learning_rate": 4.978126427086479e-05,
+      "loss": 0.8293,
+      "step": 2540
+    },
+    {
+      "epoch": 0.04505953726230518,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.978089548129502e-05,
+      "loss": 0.8728,
+      "step": 2542
+    },
+    {
+      "epoch": 0.04509498929791675,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.9780526382464846e-05,
+      "loss": 0.8605,
+      "step": 2544
+    },
+    {
+      "epoch": 0.045130441333528316,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.978015697437884e-05,
+      "loss": 0.9002,
+      "step": 2546
+    },
+    {
+      "epoch": 0.04516589336913989,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.977978725704165e-05,
+      "loss": 0.9107,
+      "step": 2548
+    },
+    {
+      "epoch": 0.04520134540475146,
+      "grad_norm": 2.875,
+      "learning_rate": 4.977941723045788e-05,
+      "loss": 0.8659,
+      "step": 2550
+    },
+    {
+      "epoch": 0.045236797440363026,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.9779046894632145e-05,
+      "loss": 0.8158,
+      "step": 2552
+    },
+    {
+      "epoch": 0.0452722494759746,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.9778676249569055e-05,
+      "loss": 0.8675,
+      "step": 2554
+    },
+    {
+      "epoch": 0.04530770151158617,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.977830529527325e-05,
+      "loss": 0.8723,
+      "step": 2556
+    },
+    {
+      "epoch": 0.045343153547197736,
+      "grad_norm": 3.0,
+      "learning_rate": 4.977793403174936e-05,
+      "loss": 0.8759,
+      "step": 2558
+    },
+    {
+      "epoch": 0.04537860558280931,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.9777562459002016e-05,
+      "loss": 0.8769,
+      "step": 2560
+    },
+    {
+      "epoch": 0.04541405761842088,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.977719057703585e-05,
+      "loss": 0.8584,
+      "step": 2562
+    },
+    {
+      "epoch": 0.045449509654032445,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.977681838585551e-05,
+      "loss": 0.8723,
+      "step": 2564
+    },
+    {
+      "epoch": 0.04548496168964402,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.977644588546563e-05,
+      "loss": 0.8643,
+      "step": 2566
+    },
+    {
+      "epoch": 0.04552041372525559,
+      "grad_norm": 3.125,
+      "learning_rate": 4.977607307587087e-05,
+      "loss": 0.9124,
+      "step": 2568
+    },
+    {
+      "epoch": 0.045555865760867155,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.977569995707588e-05,
+      "loss": 0.8738,
+      "step": 2570
+    },
+    {
+      "epoch": 0.04559131779647873,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.977532652908531e-05,
+      "loss": 0.8377,
+      "step": 2572
+    },
+    {
+      "epoch": 0.0456267698320903,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.977495279190383e-05,
+      "loss": 0.8863,
+      "step": 2574
+    },
+    {
+      "epoch": 0.045662221867701865,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.9774578745536095e-05,
+      "loss": 0.8753,
+      "step": 2576
+    },
+    {
+      "epoch": 0.04569767390331344,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.9774204389986776e-05,
+      "loss": 0.8668,
+      "step": 2578
+    },
+    {
+      "epoch": 0.04573312593892501,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.977382972526056e-05,
+      "loss": 0.8861,
+      "step": 2580
+    },
+    {
+      "epoch": 0.045768577974536574,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.9773454751362095e-05,
+      "loss": 0.8983,
+      "step": 2582
+    },
+    {
+      "epoch": 0.04580403001014814,
+      "grad_norm": 3.0,
+      "learning_rate": 4.977307946829608e-05,
+      "loss": 0.8544,
+      "step": 2584
+    },
+    {
+      "epoch": 0.045839482045759716,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.977270387606719e-05,
+      "loss": 0.8897,
+      "step": 2586
+    },
+    {
+      "epoch": 0.045874934081371284,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.977232797468012e-05,
+      "loss": 0.8848,
+      "step": 2588
+    },
+    {
+      "epoch": 0.04591038611698285,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.977195176413955e-05,
+      "loss": 0.9034,
+      "step": 2590
+    },
+    {
+      "epoch": 0.045945838152594426,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.9771575244450186e-05,
+      "loss": 0.8224,
+      "step": 2592
+    },
+    {
+      "epoch": 0.045981290188205994,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.977119841561672e-05,
+      "loss": 0.8883,
+      "step": 2594
+    },
+    {
+      "epoch": 0.04601674222381756,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.9770821277643867e-05,
+      "loss": 0.8868,
+      "step": 2596
+    },
+    {
+      "epoch": 0.046052194259429136,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.977044383053631e-05,
+      "loss": 0.8422,
+      "step": 2598
+    },
+    {
+      "epoch": 0.0460876462950407,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.977006607429878e-05,
+      "loss": 0.8772,
+      "step": 2600
+    },
+    {
+      "epoch": 0.04612309833065227,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.976968800893598e-05,
+      "loss": 0.8761,
+      "step": 2602
+    },
+    {
+      "epoch": 0.046158550366263845,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.9769309634452644e-05,
+      "loss": 0.8523,
+      "step": 2604
+    },
+    {
+      "epoch": 0.04619400240187541,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.976893095085347e-05,
+      "loss": 0.8927,
+      "step": 2606
+    },
+    {
+      "epoch": 0.04622945443748698,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.976855195814321e-05,
+      "loss": 0.8732,
+      "step": 2608
+    },
+    {
+      "epoch": 0.046264906473098555,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.976817265632657e-05,
+      "loss": 0.8469,
+      "step": 2610
+    },
+    {
+      "epoch": 0.04630035850871012,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.9767793045408293e-05,
+      "loss": 0.8568,
+      "step": 2612
+    },
+    {
+      "epoch": 0.04633581054432169,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.976741312539313e-05,
+      "loss": 0.8712,
+      "step": 2614
+    },
+    {
+      "epoch": 0.046371262579933265,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.976703289628579e-05,
+      "loss": 0.8554,
+      "step": 2616
+    },
+    {
+      "epoch": 0.04640671461554483,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.9766652358091054e-05,
+      "loss": 0.8811,
+      "step": 2618
+    },
+    {
+      "epoch": 0.0464421666511564,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.976627151081365e-05,
+      "loss": 0.9108,
+      "step": 2620
+    },
+    {
+      "epoch": 0.046477618686767974,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.9765890354458335e-05,
+      "loss": 0.8402,
+      "step": 2622
+    },
+    {
+      "epoch": 0.04651307072237954,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.976550888902987e-05,
+      "loss": 0.8537,
+      "step": 2624
+    },
+    {
+      "epoch": 0.04654852275799111,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.976512711453301e-05,
+      "loss": 0.8806,
+      "step": 2626
+    },
+    {
+      "epoch": 0.04658397479360268,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.976474503097252e-05,
+      "loss": 0.8581,
+      "step": 2628
+    },
+    {
+      "epoch": 0.04661942682921425,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.976436263835317e-05,
+      "loss": 0.8177,
+      "step": 2630
+    },
+    {
+      "epoch": 0.04665487886482582,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.976397993667974e-05,
+      "loss": 0.8971,
+      "step": 2632
+    },
+    {
+      "epoch": 0.046690330900437386,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.976359692595699e-05,
+      "loss": 0.8278,
+      "step": 2634
+    },
+    {
+      "epoch": 0.04672578293604896,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.976321360618972e-05,
+      "loss": 0.8896,
+      "step": 2636
+    },
+    {
+      "epoch": 0.04676123497166053,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.976282997738269e-05,
+      "loss": 0.8968,
+      "step": 2638
+    },
+    {
+      "epoch": 0.046796687007272096,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.97624460395407e-05,
+      "loss": 0.8867,
+      "step": 2640
+    },
+    {
+      "epoch": 0.04683213904288367,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.9762061792668546e-05,
+      "loss": 0.8456,
+      "step": 2642
+    },
+    {
+      "epoch": 0.04686759107849524,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.976167723677102e-05,
+      "loss": 0.8924,
+      "step": 2644
+    },
+    {
+      "epoch": 0.046903043114106806,
+      "grad_norm": 3.0,
+      "learning_rate": 4.976129237185291e-05,
+      "loss": 0.8889,
+      "step": 2646
+    },
+    {
+      "epoch": 0.04693849514971838,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.976090719791904e-05,
+      "loss": 0.8419,
+      "step": 2648
+    },
+    {
+      "epoch": 0.04697394718532995,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.97605217149742e-05,
+      "loss": 0.8749,
+      "step": 2650
+    },
+    {
+      "epoch": 0.047009399220941515,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.97601359230232e-05,
+      "loss": 0.8336,
+      "step": 2652
+    },
+    {
+      "epoch": 0.04704485125655309,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.9759749822070864e-05,
+      "loss": 0.8508,
+      "step": 2654
+    },
+    {
+      "epoch": 0.04708030329216466,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.975936341212202e-05,
+      "loss": 0.8548,
+      "step": 2656
+    },
+    {
+      "epoch": 0.047115755327776225,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.9758976693181464e-05,
+      "loss": 0.8688,
+      "step": 2658
+    },
+    {
+      "epoch": 0.0471512073633878,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.9758589665254044e-05,
+      "loss": 0.9214,
+      "step": 2660
+    },
+    {
+      "epoch": 0.04718665939899937,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.975820232834457e-05,
+      "loss": 0.8447,
+      "step": 2662
+    },
+    {
+      "epoch": 0.047222111434610935,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.97578146824579e-05,
+      "loss": 0.8716,
+      "step": 2664
+    },
+    {
+      "epoch": 0.04725756347022251,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.975742672759885e-05,
+      "loss": 0.86,
+      "step": 2666
+    },
+    {
+      "epoch": 0.04729301550583408,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.975703846377228e-05,
+      "loss": 0.8478,
+      "step": 2668
+    },
+    {
+      "epoch": 0.047328467541445644,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.975664989098302e-05,
+      "loss": 0.842,
+      "step": 2670
+    },
+    {
+      "epoch": 0.04736391957705721,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.975626100923593e-05,
+      "loss": 0.8574,
+      "step": 2672
+    },
+    {
+      "epoch": 0.047399371612668786,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.975587181853586e-05,
+      "loss": 0.8438,
+      "step": 2674
+    },
+    {
+      "epoch": 0.047434823648280354,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.9755482318887656e-05,
+      "loss": 0.8721,
+      "step": 2676
+    },
+    {
+      "epoch": 0.04747027568389192,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.9755092510296204e-05,
+      "loss": 0.8788,
+      "step": 2678
+    },
+    {
+      "epoch": 0.047505727719503496,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.975470239276634e-05,
+      "loss": 0.8623,
+      "step": 2680
+    },
+    {
+      "epoch": 0.047541179755115064,
+      "grad_norm": 2.75,
+      "learning_rate": 4.975431196630296e-05,
+      "loss": 0.8361,
+      "step": 2682
+    },
+    {
+      "epoch": 0.04757663179072663,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.975392123091091e-05,
+      "loss": 0.8689,
+      "step": 2684
+    },
+    {
+      "epoch": 0.047612083826338206,
+      "grad_norm": 2.75,
+      "learning_rate": 4.975353018659508e-05,
+      "loss": 0.8697,
+      "step": 2686
+    },
+    {
+      "epoch": 0.04764753586194977,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.975313883336036e-05,
+      "loss": 0.8277,
+      "step": 2688
+    },
+    {
+      "epoch": 0.04768298789756134,
+      "grad_norm": 3.0,
+      "learning_rate": 4.9752747171211614e-05,
+      "loss": 0.8991,
+      "step": 2690
+    },
+    {
+      "epoch": 0.047718439933172915,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.9752355200153735e-05,
+      "loss": 0.8643,
+      "step": 2692
+    },
+    {
+      "epoch": 0.04775389196878448,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.975196292019163e-05,
+      "loss": 0.8467,
+      "step": 2694
+    },
+    {
+      "epoch": 0.04778934400439605,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.9751570331330176e-05,
+      "loss": 0.8837,
+      "step": 2696
+    },
+    {
+      "epoch": 0.047824796040007625,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.975117743357428e-05,
+      "loss": 0.8534,
+      "step": 2698
+    },
+    {
+      "epoch": 0.04786024807561919,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.975078422692884e-05,
+      "loss": 0.8852,
+      "step": 2700
+    },
+    {
+      "epoch": 0.04789570011123076,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.975039071139878e-05,
+      "loss": 0.886,
+      "step": 2702
+    },
+    {
+      "epoch": 0.047931152146842335,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.9749996886988994e-05,
+      "loss": 0.88,
+      "step": 2704
+    },
+    {
+      "epoch": 0.0479666041824539,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.974960275370439e-05,
+      "loss": 0.8631,
+      "step": 2706
+    },
+    {
+      "epoch": 0.04800205621806547,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.9749208311549916e-05,
+      "loss": 0.9029,
+      "step": 2708
+    },
+    {
+      "epoch": 0.048037508253677044,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.974881356053047e-05,
+      "loss": 0.8651,
+      "step": 2710
+    },
+    {
+      "epoch": 0.04807296028928861,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.974841850065098e-05,
+      "loss": 0.8155,
+      "step": 2712
+    },
+    {
+      "epoch": 0.04810841232490018,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.97480231319164e-05,
+      "loss": 0.8669,
+      "step": 2714
+    },
+    {
+      "epoch": 0.04814386436051175,
+      "grad_norm": 3.125,
+      "learning_rate": 4.974762745433163e-05,
+      "loss": 0.8886,
+      "step": 2716
+    },
+    {
+      "epoch": 0.04817931639612332,
+      "grad_norm": 3.0,
+      "learning_rate": 4.9747231467901625e-05,
+      "loss": 0.868,
+      "step": 2718
+    },
+    {
+      "epoch": 0.04821476843173489,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.974683517263134e-05,
+      "loss": 0.8811,
+      "step": 2720
+    },
+    {
+      "epoch": 0.04825022046734646,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.974643856852569e-05,
+      "loss": 0.8448,
+      "step": 2722
+    },
+    {
+      "epoch": 0.04828567250295803,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.9746041655589644e-05,
+      "loss": 0.8653,
+      "step": 2724
+    },
+    {
+      "epoch": 0.0483211245385696,
+      "grad_norm": 2.875,
+      "learning_rate": 4.974564443382815e-05,
+      "loss": 0.8171,
+      "step": 2726
+    },
+    {
+      "epoch": 0.048356576574181166,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.974524690324618e-05,
+      "loss": 0.8275,
+      "step": 2728
+    },
+    {
+      "epoch": 0.04839202860979274,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.974484906384867e-05,
+      "loss": 0.8725,
+      "step": 2730
+    },
+    {
+      "epoch": 0.04842748064540431,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.9744450915640605e-05,
+      "loss": 0.8374,
+      "step": 2732
+    },
+    {
+      "epoch": 0.048462932681015876,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.9744052458626944e-05,
+      "loss": 0.9077,
+      "step": 2734
+    },
+    {
+      "epoch": 0.04849838471662745,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.9743653692812657e-05,
+      "loss": 0.8366,
+      "step": 2736
+    },
+    {
+      "epoch": 0.04853383675223902,
+      "grad_norm": 3.0,
+      "learning_rate": 4.974325461820273e-05,
+      "loss": 0.8669,
+      "step": 2738
+    },
+    {
+      "epoch": 0.048569288787850586,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.9742855234802146e-05,
+      "loss": 0.904,
+      "step": 2740
+    },
+    {
+      "epoch": 0.04860474082346216,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.9742455542615876e-05,
+      "loss": 0.854,
+      "step": 2742
+    },
+    {
+      "epoch": 0.04864019285907373,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.9742055541648916e-05,
+      "loss": 0.8762,
+      "step": 2744
+    },
+    {
+      "epoch": 0.048675644894685295,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.9741655231906246e-05,
+      "loss": 0.8533,
+      "step": 2746
+    },
+    {
+      "epoch": 0.04871109693029687,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.9741254613392887e-05,
+      "loss": 0.8636,
+      "step": 2748
+    },
+    {
+      "epoch": 0.04874654896590844,
+      "grad_norm": 2.75,
+      "learning_rate": 4.974085368611381e-05,
+      "loss": 0.8573,
+      "step": 2750
+    },
+    {
+      "epoch": 0.048782001001520005,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.9740452450074044e-05,
+      "loss": 0.9177,
+      "step": 2752
+    },
+    {
+      "epoch": 0.04881745303713157,
+      "grad_norm": 2.75,
+      "learning_rate": 4.9740050905278577e-05,
+      "loss": 0.8821,
+      "step": 2754
+    },
+    {
+      "epoch": 0.04885290507274315,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.973964905173243e-05,
+      "loss": 0.8554,
+      "step": 2756
+    },
+    {
+      "epoch": 0.048888357108354714,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.973924688944061e-05,
+      "loss": 0.8452,
+      "step": 2758
+    },
+    {
+      "epoch": 0.04892380914396628,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.973884441840816e-05,
+      "loss": 0.8852,
+      "step": 2760
+    },
+    {
+      "epoch": 0.048959261179577857,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.9738441638640064e-05,
+      "loss": 0.8495,
+      "step": 2762
+    },
+    {
+      "epoch": 0.048994713215189424,
+      "grad_norm": 2.875,
+      "learning_rate": 4.973803855014138e-05,
+      "loss": 0.883,
+      "step": 2764
+    },
+    {
+      "epoch": 0.04903016525080099,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.973763515291713e-05,
+      "loss": 0.8985,
+      "step": 2766
+    },
+    {
+      "epoch": 0.049065617286412566,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.9737231446972334e-05,
+      "loss": 0.8508,
+      "step": 2768
+    },
+    {
+      "epoch": 0.049101069322024134,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.973682743231205e-05,
+      "loss": 0.8939,
+      "step": 2770
+    },
+    {
+      "epoch": 0.0491365213576357,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.973642310894131e-05,
+      "loss": 0.8824,
+      "step": 2772
+    },
+    {
+      "epoch": 0.049171973393247276,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.9736018476865165e-05,
+      "loss": 0.9063,
+      "step": 2774
+    },
+    {
+      "epoch": 0.04920742542885884,
+      "grad_norm": 2.75,
+      "learning_rate": 4.973561353608866e-05,
+      "loss": 0.8409,
+      "step": 2776
+    },
+    {
+      "epoch": 0.04924287746447041,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.973520828661684e-05,
+      "loss": 0.8812,
+      "step": 2778
+    },
+    {
+      "epoch": 0.049278329500081985,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.973480272845479e-05,
+      "loss": 0.8498,
+      "step": 2780
+    },
+    {
+      "epoch": 0.04931378153569355,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.9734396861607543e-05,
+      "loss": 0.8471,
+      "step": 2782
+    },
+    {
+      "epoch": 0.04934923357130512,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.973399068608018e-05,
+      "loss": 0.8652,
+      "step": 2784
+    },
+    {
+      "epoch": 0.049384685606916695,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.973358420187776e-05,
+      "loss": 0.8479,
+      "step": 2786
+    },
+    {
+      "epoch": 0.04942013764252826,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.973317740900536e-05,
+      "loss": 0.8945,
+      "step": 2788
+    },
+    {
+      "epoch": 0.04945558967813983,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.973277030746806e-05,
+      "loss": 0.85,
+      "step": 2790
+    },
+    {
+      "epoch": 0.049491041713751405,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.973236289727094e-05,
+      "loss": 0.8745,
+      "step": 2792
+    },
+    {
+      "epoch": 0.04952649374936297,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.9731955178419075e-05,
+      "loss": 0.842,
+      "step": 2794
+    },
+    {
+      "epoch": 0.04956194578497454,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.9731547150917566e-05,
+      "loss": 0.9042,
+      "step": 2796
+    },
+    {
+      "epoch": 0.04959739782058611,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.973113881477151e-05,
+      "loss": 0.885,
+      "step": 2798
+    },
+    {
+      "epoch": 0.04963284985619768,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.973073016998598e-05,
+      "loss": 0.8983,
+      "step": 2800
+    },
+    {
+      "epoch": 0.04966830189180925,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.973032121656609e-05,
+      "loss": 0.8418,
+      "step": 2802
+    },
+    {
+      "epoch": 0.04970375392742082,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.9729911954516946e-05,
+      "loss": 0.8868,
+      "step": 2804
+    },
+    {
+      "epoch": 0.04973920596303239,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.9729502383843643e-05,
+      "loss": 0.838,
+      "step": 2806
+    },
+    {
+      "epoch": 0.04977465799864396,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.972909250455131e-05,
+      "loss": 0.8609,
+      "step": 2808
+    },
+    {
+      "epoch": 0.04981011003425553,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.972868231664505e-05,
+      "loss": 0.8558,
+      "step": 2810
+    },
+    {
+      "epoch": 0.0498455620698671,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.972827182012999e-05,
+      "loss": 0.858,
+      "step": 2812
+    },
+    {
+      "epoch": 0.04988101410547867,
+      "grad_norm": 2.875,
+      "learning_rate": 4.972786101501125e-05,
+      "loss": 0.8594,
+      "step": 2814
+    },
+    {
+      "epoch": 0.049916466141090236,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.972744990129394e-05,
+      "loss": 0.9087,
+      "step": 2816
+    },
+    {
+      "epoch": 0.04995191817670181,
+      "grad_norm": 2.75,
+      "learning_rate": 4.972703847898321e-05,
+      "loss": 0.8705,
+      "step": 2818
+    },
+    {
+      "epoch": 0.04998737021231338,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.97266267480842e-05,
+      "loss": 0.8763,
+      "step": 2820
+    },
+    {
+      "epoch": 0.050022822247924946,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.972621470860204e-05,
+      "loss": 0.8277,
+      "step": 2822
+    },
+    {
+      "epoch": 0.05005827428353652,
+      "grad_norm": 2.75,
+      "learning_rate": 4.9725802360541854e-05,
+      "loss": 0.8925,
+      "step": 2824
+    },
+    {
+      "epoch": 0.05009372631914809,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.972538970390882e-05,
+      "loss": 0.8644,
+      "step": 2826
+    },
+    {
+      "epoch": 0.050129178354759656,
+      "grad_norm": 3.375,
+      "learning_rate": 4.972497673870806e-05,
+      "loss": 0.895,
+      "step": 2828
+    },
+    {
+      "epoch": 0.05016463039037123,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.9724563464944754e-05,
+      "loss": 0.8475,
+      "step": 2830
+    },
+    {
+      "epoch": 0.0502000824259828,
+      "grad_norm": 2.875,
+      "learning_rate": 4.972414988262404e-05,
+      "loss": 0.8468,
+      "step": 2832
+    },
+    {
+      "epoch": 0.050235534461594365,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.9723735991751076e-05,
+      "loss": 0.9013,
+      "step": 2834
+    },
+    {
+      "epoch": 0.05027098649720594,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.972332179233105e-05,
+      "loss": 0.9036,
+      "step": 2836
+    },
+    {
+      "epoch": 0.05030643853281751,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.972290728436911e-05,
+      "loss": 0.8596,
+      "step": 2838
+    },
+    {
+      "epoch": 0.050341890568429075,
+      "grad_norm": 2.875,
+      "learning_rate": 4.9722492467870434e-05,
+      "loss": 0.8977,
+      "step": 2840
+    },
+    {
+      "epoch": 0.05037734260404064,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.9722077342840204e-05,
+      "loss": 0.8696,
+      "step": 2842
+    },
+    {
+      "epoch": 0.05041279463965222,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.97216619092836e-05,
+      "loss": 0.8877,
+      "step": 2844
+    },
+    {
+      "epoch": 0.050448246675263785,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.97212461672058e-05,
+      "loss": 0.8741,
+      "step": 2846
+    },
+    {
+      "epoch": 0.05048369871087535,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.9720830116612004e-05,
+      "loss": 0.877,
+      "step": 2848
+    },
+    {
+      "epoch": 0.05051915074648693,
+      "grad_norm": 2.875,
+      "learning_rate": 4.97204137575074e-05,
+      "loss": 0.8109,
+      "step": 2850
+    },
+    {
+      "epoch": 0.050554602782098494,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.9719997089897176e-05,
+      "loss": 0.8455,
+      "step": 2852
+    },
+    {
+      "epoch": 0.05059005481771006,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.971958011378653e-05,
+      "loss": 0.8942,
+      "step": 2854
+    },
+    {
+      "epoch": 0.050625506853321636,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.9719162829180684e-05,
+      "loss": 0.8563,
+      "step": 2856
+    },
+    {
+      "epoch": 0.050660958888933204,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.971874523608483e-05,
+      "loss": 0.8732,
+      "step": 2858
+    },
+    {
+      "epoch": 0.05069641092454477,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.9718327334504175e-05,
+      "loss": 0.8292,
+      "step": 2860
+    },
+    {
+      "epoch": 0.050731862960156346,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.971790912444396e-05,
+      "loss": 0.9058,
+      "step": 2862
+    },
+    {
+      "epoch": 0.05076731499576791,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.971749060590938e-05,
+      "loss": 0.883,
+      "step": 2864
+    },
+    {
+      "epoch": 0.05080276703137948,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.9717071778905667e-05,
+      "loss": 0.8377,
+      "step": 2866
+    },
+    {
+      "epoch": 0.050838219066991056,
+      "grad_norm": 3.125,
+      "learning_rate": 4.971665264343804e-05,
+      "loss": 0.8734,
+      "step": 2868
+    },
+    {
+      "epoch": 0.05087367110260262,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.971623319951174e-05,
+      "loss": 0.8644,
+      "step": 2870
+    },
+    {
+      "epoch": 0.05090912313821419,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.9715813447132e-05,
+      "loss": 0.8551,
+      "step": 2872
+    },
+    {
+      "epoch": 0.050944575173825765,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.9715393386304056e-05,
+      "loss": 0.8708,
+      "step": 2874
+    },
+    {
+      "epoch": 0.05098002720943733,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.971497301703315e-05,
+      "loss": 0.8547,
+      "step": 2876
+    },
+    {
+      "epoch": 0.0510154792450489,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.9714552339324525e-05,
+      "loss": 0.8751,
+      "step": 2878
+    },
+    {
+      "epoch": 0.051050931280660475,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.9714131353183435e-05,
+      "loss": 0.8485,
+      "step": 2880
+    },
+    {
+      "epoch": 0.05108638331627204,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.971371005861514e-05,
+      "loss": 0.8416,
+      "step": 2882
+    },
+    {
+      "epoch": 0.05112183535188361,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.971328845562488e-05,
+      "loss": 0.9008,
+      "step": 2884
+    },
+    {
+      "epoch": 0.05115728738749518,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.971286654421793e-05,
+      "loss": 0.822,
+      "step": 2886
+    },
+    {
+      "epoch": 0.05119273942310675,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.971244432439956e-05,
+      "loss": 0.8887,
+      "step": 2888
+    },
+    {
+      "epoch": 0.05122819145871832,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.971202179617502e-05,
+      "loss": 0.8879,
+      "step": 2890
+    },
+    {
+      "epoch": 0.05126364349432989,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.971159895954961e-05,
+      "loss": 0.8847,
+      "step": 2892
+    },
+    {
+      "epoch": 0.05129909552994146,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.971117581452859e-05,
+      "loss": 0.872,
+      "step": 2894
+    },
+    {
+      "epoch": 0.05133454756555303,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.971075236111724e-05,
+      "loss": 0.861,
+      "step": 2896
+    },
+    {
+      "epoch": 0.0513699996011646,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.9710328599320846e-05,
+      "loss": 0.8889,
+      "step": 2898
+    },
+    {
+      "epoch": 0.05140545163677617,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.970990452914469e-05,
+      "loss": 0.8878,
+      "step": 2900
+    },
+    {
+      "epoch": 0.05144090367238774,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.970948015059408e-05,
+      "loss": 0.8651,
+      "step": 2902
+    },
+    {
+      "epoch": 0.051476355707999306,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.9709055463674304e-05,
+      "loss": 0.8522,
+      "step": 2904
+    },
+    {
+      "epoch": 0.05151180774361088,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.970863046839066e-05,
+      "loss": 0.8409,
+      "step": 2906
+    },
+    {
+      "epoch": 0.05154725977922245,
+      "grad_norm": 3.0,
+      "learning_rate": 4.970820516474846e-05,
+      "loss": 0.8737,
+      "step": 2908
+    },
+    {
+      "epoch": 0.051582711814834016,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.9707779552753e-05,
+      "loss": 0.87,
+      "step": 2910
+    },
+    {
+      "epoch": 0.05161816385044559,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.97073536324096e-05,
+      "loss": 0.8184,
+      "step": 2912
+    },
+    {
+      "epoch": 0.05165361588605716,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.9706927403723574e-05,
+      "loss": 0.9204,
+      "step": 2914
+    },
+    {
+      "epoch": 0.051689067921668726,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.970650086670024e-05,
+      "loss": 0.8298,
+      "step": 2916
+    },
+    {
+      "epoch": 0.0517245199572803,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.9706074021344916e-05,
+      "loss": 0.8427,
+      "step": 2918
+    },
+    {
+      "epoch": 0.05175997199289187,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.970564686766294e-05,
+      "loss": 0.8631,
+      "step": 2920
+    },
+    {
+      "epoch": 0.051795424028503435,
+      "grad_norm": 2.875,
+      "learning_rate": 4.9705219405659635e-05,
+      "loss": 0.8462,
+      "step": 2922
+    },
+    {
+      "epoch": 0.05183087606411501,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.970479163534034e-05,
+      "loss": 0.8948,
+      "step": 2924
+    },
+    {
+      "epoch": 0.05186632809972658,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.970436355671039e-05,
+      "loss": 0.8498,
+      "step": 2926
+    },
+    {
+      "epoch": 0.051901780135338145,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.970393516977513e-05,
+      "loss": 0.885,
+      "step": 2928
+    },
+    {
+      "epoch": 0.05193723217094971,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.9703506474539894e-05,
+      "loss": 0.8458,
+      "step": 2930
+    },
+    {
+      "epoch": 0.05197268420656129,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.970307747101005e-05,
+      "loss": 0.8869,
+      "step": 2932
+    },
+    {
+      "epoch": 0.052008136242172855,
+      "grad_norm": 2.625,
+      "learning_rate": 4.9702648159190944e-05,
+      "loss": 0.8571,
+      "step": 2934
+    },
+    {
+      "epoch": 0.05204358827778442,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.970221853908794e-05,
+      "loss": 0.8827,
+      "step": 2936
+    },
+    {
+      "epoch": 0.052079040313396,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.9701788610706384e-05,
+      "loss": 0.8422,
+      "step": 2938
+    },
+    {
+      "epoch": 0.052114492349007564,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.970135837405166e-05,
+      "loss": 0.8821,
+      "step": 2940
+    },
+    {
+      "epoch": 0.05214994438461913,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.970092782912912e-05,
+      "loss": 0.8914,
+      "step": 2942
+    },
+    {
+      "epoch": 0.052185396420230706,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.970049697594415e-05,
+      "loss": 0.8454,
+      "step": 2944
+    },
+    {
+      "epoch": 0.052220848455842274,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.9700065814502125e-05,
+      "loss": 0.8119,
+      "step": 2946
+    },
+    {
+      "epoch": 0.05225630049145384,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.9699634344808425e-05,
+      "loss": 0.904,
+      "step": 2948
+    },
+    {
+      "epoch": 0.052291752527065416,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.969920256686842e-05,
+      "loss": 0.8324,
+      "step": 2950
+    },
+    {
+      "epoch": 0.052327204562676984,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.9698770480687515e-05,
+      "loss": 0.8968,
+      "step": 2952
+    },
+    {
+      "epoch": 0.05236265659828855,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.9698338086271114e-05,
+      "loss": 0.8728,
+      "step": 2954
+    },
+    {
+      "epoch": 0.052398108633900126,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.969790538362458e-05,
+      "loss": 0.8537,
+      "step": 2956
+    },
+    {
+      "epoch": 0.05243356066951169,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.969747237275334e-05,
+      "loss": 0.8752,
+      "step": 2958
+    },
+    {
+      "epoch": 0.05246901270512326,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.9697039053662785e-05,
+      "loss": 0.8249,
+      "step": 2960
+    },
+    {
+      "epoch": 0.052504464740734835,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.969660542635833e-05,
+      "loss": 0.8757,
+      "step": 2962
+    },
+    {
+      "epoch": 0.0525399167763464,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.969617149084538e-05,
+      "loss": 0.8594,
+      "step": 2964
+    },
+    {
+      "epoch": 0.05257536881195797,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.969573724712936e-05,
+      "loss": 0.8425,
+      "step": 2966
+    },
+    {
+      "epoch": 0.05261082084756954,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.969530269521568e-05,
+      "loss": 0.8623,
+      "step": 2968
+    },
+    {
+      "epoch": 0.05264627288318111,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.969486783510976e-05,
+      "loss": 0.8533,
+      "step": 2970
+    },
+    {
+      "epoch": 0.05268172491879268,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.9694432666817036e-05,
+      "loss": 0.8297,
+      "step": 2972
+    },
+    {
+      "epoch": 0.05271717695440425,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.969399719034295e-05,
+      "loss": 0.8356,
+      "step": 2974
+    },
+    {
+      "epoch": 0.05275262899001582,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.96935614056929e-05,
+      "loss": 0.8644,
+      "step": 2976
+    },
+    {
+      "epoch": 0.05278808102562739,
+      "grad_norm": 3.375,
+      "learning_rate": 4.9693125312872356e-05,
+      "loss": 0.8769,
+      "step": 2978
+    },
+    {
+      "epoch": 0.05282353306123896,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.969268891188676e-05,
+      "loss": 0.8532,
+      "step": 2980
+    },
+    {
+      "epoch": 0.05285898509685053,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.969225220274154e-05,
+      "loss": 0.8512,
+      "step": 2982
+    },
+    {
+      "epoch": 0.0528944371324621,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.9691815185442155e-05,
+      "loss": 0.869,
+      "step": 2984
+    },
+    {
+      "epoch": 0.05292988916807367,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.9691377859994056e-05,
+      "loss": 0.8594,
+      "step": 2986
+    },
+    {
+      "epoch": 0.05296534120368524,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.9690940226402716e-05,
+      "loss": 0.8524,
+      "step": 2988
+    },
+    {
+      "epoch": 0.05300079323929681,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.969050228467358e-05,
+      "loss": 0.8984,
+      "step": 2990
+    },
+    {
+      "epoch": 0.05303624527490838,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.9690064034812114e-05,
+      "loss": 0.8926,
+      "step": 2992
+    },
+    {
+      "epoch": 0.05307169731051995,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.9689625476823795e-05,
+      "loss": 0.8046,
+      "step": 2994
+    },
+    {
+      "epoch": 0.05310714934613152,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.968918661071409e-05,
+      "loss": 0.8651,
+      "step": 2996
+    },
+    {
+      "epoch": 0.053142601381743086,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.968874743648848e-05,
+      "loss": 0.913,
+      "step": 2998
+    },
+    {
+      "epoch": 0.05317805341735466,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.968830795415245e-05,
+      "loss": 0.8606,
+      "step": 3000
+    },
+    {
+      "epoch": 0.05321350545296623,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.9687868163711474e-05,
+      "loss": 0.8876,
+      "step": 3002
+    },
+    {
+      "epoch": 0.053248957488577796,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.968742806517104e-05,
+      "loss": 0.8648,
+      "step": 3004
+    },
+    {
+      "epoch": 0.05328440952418937,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.9686987658536646e-05,
+      "loss": 0.8341,
+      "step": 3006
+    },
+    {
+      "epoch": 0.05331986155980094,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.968654694381379e-05,
+      "loss": 0.8511,
+      "step": 3008
+    },
+    {
+      "epoch": 0.053355313595412505,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.968610592100797e-05,
+      "loss": 0.8549,
+      "step": 3010
+    },
+    {
+      "epoch": 0.05339076563102407,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.96856645901247e-05,
+      "loss": 0.8929,
+      "step": 3012
+    },
+    {
+      "epoch": 0.05342621766663565,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.968522295116947e-05,
+      "loss": 0.8663,
+      "step": 3014
+    },
+    {
+      "epoch": 0.053461669702247215,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.9684781004147795e-05,
+      "loss": 0.8973,
+      "step": 3016
+    },
+    {
+      "epoch": 0.05349712173785878,
+      "grad_norm": 2.625,
+      "learning_rate": 4.96843387490652e-05,
+      "loss": 0.8569,
+      "step": 3018
+    },
+    {
+      "epoch": 0.05353257377347036,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.96838961859272e-05,
+      "loss": 0.8066,
+      "step": 3020
+    },
+    {
+      "epoch": 0.053568025809081925,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.9683453314739314e-05,
+      "loss": 0.9235,
+      "step": 3022
+    },
+    {
+      "epoch": 0.05360347784469349,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.968301013550707e-05,
+      "loss": 0.8878,
+      "step": 3024
+    },
+    {
+      "epoch": 0.05363892988030507,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.9682566648236007e-05,
+      "loss": 0.894,
+      "step": 3026
+    },
+    {
+      "epoch": 0.053674381915916634,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.968212285293165e-05,
+      "loss": 0.8596,
+      "step": 3028
+    },
+    {
+      "epoch": 0.0537098339515282,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.9681678749599536e-05,
+      "loss": 0.8688,
+      "step": 3030
+    },
+    {
+      "epoch": 0.053745285987139776,
+      "grad_norm": 2.75,
+      "learning_rate": 4.9681234338245214e-05,
+      "loss": 0.872,
+      "step": 3032
+    },
+    {
+      "epoch": 0.053780738022751344,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.968078961887423e-05,
+      "loss": 0.859,
+      "step": 3034
+    },
+    {
+      "epoch": 0.05381619005836291,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.968034459149213e-05,
+      "loss": 0.8821,
+      "step": 3036
+    },
+    {
+      "epoch": 0.053851642093974486,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.967989925610447e-05,
+      "loss": 0.8507,
+      "step": 3038
+    },
+    {
+      "epoch": 0.053887094129586054,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.9679453612716816e-05,
+      "loss": 0.8604,
+      "step": 3040
+    },
+    {
+      "epoch": 0.05392254616519762,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.967900766133471e-05,
+      "loss": 0.8521,
+      "step": 3042
+    },
+    {
+      "epoch": 0.053957998200809196,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.9678561401963736e-05,
+      "loss": 0.9058,
+      "step": 3044
+    },
+    {
+      "epoch": 0.05399345023642076,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.967811483460946e-05,
+      "loss": 0.8202,
+      "step": 3046
+    },
+    {
+      "epoch": 0.05402890227203233,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.967766795927744e-05,
+      "loss": 0.8616,
+      "step": 3048
+    },
+    {
+      "epoch": 0.054064354307643905,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.967722077597327e-05,
+      "loss": 0.8462,
+      "step": 3050
+    },
+    {
+      "epoch": 0.05409980634325547,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.967677328470252e-05,
+      "loss": 0.8751,
+      "step": 3052
+    },
+    {
+      "epoch": 0.05413525837886704,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.967632548547078e-05,
+      "loss": 0.9041,
+      "step": 3054
+    },
+    {
+      "epoch": 0.05417071041447861,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.967587737828364e-05,
+      "loss": 0.8535,
+      "step": 3056
+    },
+    {
+      "epoch": 0.05420616245009018,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.967542896314669e-05,
+      "loss": 0.823,
+      "step": 3058
+    },
+    {
+      "epoch": 0.05424161448570175,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.967498024006553e-05,
+      "loss": 0.847,
+      "step": 3060
+    },
+    {
+      "epoch": 0.05427706652131332,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.967453120904575e-05,
+      "loss": 0.8793,
+      "step": 3062
+    },
+    {
+      "epoch": 0.05431251855692489,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.967408187009296e-05,
+      "loss": 0.87,
+      "step": 3064
+    },
+    {
+      "epoch": 0.05434797059253646,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.967363222321277e-05,
+      "loss": 0.8522,
+      "step": 3066
+    },
+    {
+      "epoch": 0.05438342262814803,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.967318226841079e-05,
+      "loss": 0.8765,
+      "step": 3068
+    },
+    {
+      "epoch": 0.0544188746637596,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.967273200569263e-05,
+      "loss": 0.8881,
+      "step": 3070
+    },
+    {
+      "epoch": 0.05445432669937117,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.9672281435063915e-05,
+      "loss": 0.8436,
+      "step": 3072
+    },
+    {
+      "epoch": 0.05448977873498274,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.967183055653027e-05,
+      "loss": 0.8846,
+      "step": 3074
+    },
+    {
+      "epoch": 0.05452523077059431,
+      "grad_norm": 2.75,
+      "learning_rate": 4.9671379370097314e-05,
+      "loss": 0.8671,
+      "step": 3076
+    },
+    {
+      "epoch": 0.05456068280620588,
+      "grad_norm": 3.25,
+      "learning_rate": 4.967092787577068e-05,
+      "loss": 0.8623,
+      "step": 3078
+    },
+    {
+      "epoch": 0.05459613484181745,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.9670476073556015e-05,
+      "loss": 0.8926,
+      "step": 3080
+    },
+    {
+      "epoch": 0.05463158687742902,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.967002396345894e-05,
+      "loss": 0.8009,
+      "step": 3082
+    },
+    {
+      "epoch": 0.05466703891304059,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.966957154548511e-05,
+      "loss": 0.8722,
+      "step": 3084
+    },
+    {
+      "epoch": 0.054702490948652156,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.966911881964016e-05,
+      "loss": 0.8625,
+      "step": 3086
+    },
+    {
+      "epoch": 0.05473794298426373,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.9668665785929744e-05,
+      "loss": 0.8557,
+      "step": 3088
+    },
+    {
+      "epoch": 0.0547733950198753,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.966821244435952e-05,
+      "loss": 0.8461,
+      "step": 3090
+    },
+    {
+      "epoch": 0.054808847055486866,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.966775879493514e-05,
+      "loss": 0.8348,
+      "step": 3092
+    },
+    {
+      "epoch": 0.05484429909109844,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.9667304837662265e-05,
+      "loss": 0.8429,
+      "step": 3094
+    },
+    {
+      "epoch": 0.05487975112671001,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.9666850572546575e-05,
+      "loss": 0.8614,
+      "step": 3096
+    },
+    {
+      "epoch": 0.054915203162321576,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.966639599959372e-05,
+      "loss": 0.9079,
+      "step": 3098
+    },
+    {
+      "epoch": 0.05495065519793314,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.9665941118809375e-05,
+      "loss": 0.8368,
+      "step": 3100
+    },
+    {
+      "epoch": 0.05498610723354472,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.966548593019923e-05,
+      "loss": 0.8453,
+      "step": 3102
+    },
+    {
+      "epoch": 0.055021559269156285,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.966503043376896e-05,
+      "loss": 0.9079,
+      "step": 3104
+    },
+    {
+      "epoch": 0.05505701130476785,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.966457462952424e-05,
+      "loss": 0.8182,
+      "step": 3106
+    },
+    {
+      "epoch": 0.05509246334037943,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.966411851747078e-05,
+      "loss": 0.8326,
+      "step": 3108
+    },
+    {
+      "epoch": 0.055127915375990995,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.9663662097614245e-05,
+      "loss": 0.8604,
+      "step": 3110
+    },
+    {
+      "epoch": 0.05516336741160256,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.9663205369960345e-05,
+      "loss": 0.8305,
+      "step": 3112
+    },
+    {
+      "epoch": 0.05519881944721414,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.9662748334514784e-05,
+      "loss": 0.8346,
+      "step": 3114
+    },
+    {
+      "epoch": 0.055234271482825704,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.9662290991283254e-05,
+      "loss": 0.914,
+      "step": 3116
+    },
+    {
+      "epoch": 0.05526972351843727,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.966183334027148e-05,
+      "loss": 0.8923,
+      "step": 3118
+    },
+    {
+      "epoch": 0.05530517555404885,
+      "grad_norm": 3.125,
+      "learning_rate": 4.966137538148515e-05,
+      "loss": 0.878,
+      "step": 3120
+    },
+    {
+      "epoch": 0.055340627589660414,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.966091711493e-05,
+      "loss": 0.8524,
+      "step": 3122
+    },
+    {
+      "epoch": 0.05537607962527198,
+      "grad_norm": 2.5,
+      "learning_rate": 4.966045854061174e-05,
+      "loss": 0.8129,
+      "step": 3124
+    },
+    {
+      "epoch": 0.055411531660883556,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.965999965853609e-05,
+      "loss": 0.8476,
+      "step": 3126
+    },
+    {
+      "epoch": 0.055446983696495124,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.965954046870879e-05,
+      "loss": 0.8393,
+      "step": 3128
+    },
+    {
+      "epoch": 0.05548243573210669,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.965908097113555e-05,
+      "loss": 0.8137,
+      "step": 3130
+    },
+    {
+      "epoch": 0.055517887767718266,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.965862116582212e-05,
+      "loss": 0.8436,
+      "step": 3132
+    },
+    {
+      "epoch": 0.05555333980332983,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.965816105277423e-05,
+      "loss": 0.8907,
+      "step": 3134
+    },
+    {
+      "epoch": 0.0555887918389414,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.965770063199763e-05,
+      "loss": 0.8394,
+      "step": 3136
+    },
+    {
+      "epoch": 0.05562424387455297,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.9657239903498064e-05,
+      "loss": 0.848,
+      "step": 3138
+    },
+    {
+      "epoch": 0.05565969591016454,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.965677886728128e-05,
+      "loss": 0.8919,
+      "step": 3140
+    },
+    {
+      "epoch": 0.05569514794577611,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.965631752335302e-05,
+      "loss": 0.8882,
+      "step": 3142
+    },
+    {
+      "epoch": 0.05573059998138768,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.965585587171907e-05,
+      "loss": 0.8359,
+      "step": 3144
+    },
+    {
+      "epoch": 0.05576605201699925,
+      "grad_norm": 2.625,
+      "learning_rate": 4.9655393912385164e-05,
+      "loss": 0.8831,
+      "step": 3146
+    },
+    {
+      "epoch": 0.05580150405261082,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.9654931645357075e-05,
+      "loss": 0.8672,
+      "step": 3148
+    },
+    {
+      "epoch": 0.05583695608822239,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.965446907064059e-05,
+      "loss": 0.8515,
+      "step": 3150
+    },
+    {
+      "epoch": 0.05587240812383396,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.965400618824145e-05,
+      "loss": 0.9172,
+      "step": 3152
+    },
+    {
+      "epoch": 0.05590786015944553,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.965354299816545e-05,
+      "loss": 0.8372,
+      "step": 3154
+    },
+    {
+      "epoch": 0.0559433121950571,
+      "grad_norm": 2.625,
+      "learning_rate": 4.965307950041837e-05,
+      "loss": 0.8744,
+      "step": 3156
+    },
+    {
+      "epoch": 0.05597876423066867,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.9652615695006e-05,
+      "loss": 0.8757,
+      "step": 3158
+    },
+    {
+      "epoch": 0.05601421626628024,
+      "grad_norm": 3.25,
+      "learning_rate": 4.965215158193411e-05,
+      "loss": 0.8774,
+      "step": 3160
+    },
+    {
+      "epoch": 0.05604966830189181,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.9651687161208505e-05,
+      "loss": 0.8498,
+      "step": 3162
+    },
+    {
+      "epoch": 0.05608512033750338,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.9651222432834985e-05,
+      "loss": 0.9136,
+      "step": 3164
+    },
+    {
+      "epoch": 0.05612057237311495,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.9650757396819335e-05,
+      "loss": 0.8381,
+      "step": 3166
+    },
+    {
+      "epoch": 0.05615602440872652,
+      "grad_norm": 3.0,
+      "learning_rate": 4.965029205316737e-05,
+      "loss": 0.8501,
+      "step": 3168
+    },
+    {
+      "epoch": 0.05619147644433809,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.9649826401884904e-05,
+      "loss": 0.8755,
+      "step": 3170
+    },
+    {
+      "epoch": 0.05622692847994966,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.964936044297773e-05,
+      "loss": 0.9018,
+      "step": 3172
+    },
+    {
+      "epoch": 0.056262380515561226,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.964889417645166e-05,
+      "loss": 0.8797,
+      "step": 3174
+    },
+    {
+      "epoch": 0.0562978325511728,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.964842760231254e-05,
+      "loss": 0.8833,
+      "step": 3176
+    },
+    {
+      "epoch": 0.05633328458678437,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.964796072056618e-05,
+      "loss": 0.8543,
+      "step": 3178
+    },
+    {
+      "epoch": 0.056368736622395936,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.96474935312184e-05,
+      "loss": 0.9254,
+      "step": 3180
+    },
+    {
+      "epoch": 0.056404188658007504,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.964702603427504e-05,
+      "loss": 0.8885,
+      "step": 3182
+    },
+    {
+      "epoch": 0.05643964069361908,
+      "grad_norm": 3.0,
+      "learning_rate": 4.964655822974191e-05,
+      "loss": 0.8349,
+      "step": 3184
+    },
+    {
+      "epoch": 0.056475092729230646,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.964609011762488e-05,
+      "loss": 0.8539,
+      "step": 3186
+    },
+    {
+      "epoch": 0.05651054476484221,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.964562169792978e-05,
+      "loss": 0.8605,
+      "step": 3188
+    },
+    {
+      "epoch": 0.05654599680045379,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.964515297066245e-05,
+      "loss": 0.8885,
+      "step": 3190
+    },
+    {
+      "epoch": 0.056581448836065355,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.964468393582875e-05,
+      "loss": 0.812,
+      "step": 3192
+    },
+    {
+      "epoch": 0.05661690087167692,
+      "grad_norm": 2.5,
+      "learning_rate": 4.964421459343452e-05,
+      "loss": 0.8265,
+      "step": 3194
+    },
+    {
+      "epoch": 0.0566523529072885,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.9643744943485626e-05,
+      "loss": 0.866,
+      "step": 3196
+    },
+    {
+      "epoch": 0.056687804942900065,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.964327498598793e-05,
+      "loss": 0.8846,
+      "step": 3198
+    },
+    {
+      "epoch": 0.05672325697851163,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.96428047209473e-05,
+      "loss": 0.8741,
+      "step": 3200
+    },
+    {
+      "epoch": 0.05675870901412321,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.9642334148369595e-05,
+      "loss": 0.8554,
+      "step": 3202
+    },
+    {
+      "epoch": 0.056794161049734775,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.964186326826069e-05,
+      "loss": 0.8448,
+      "step": 3204
+    },
+    {
+      "epoch": 0.05682961308534634,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.964139208062647e-05,
+      "loss": 0.8586,
+      "step": 3206
+    },
+    {
+      "epoch": 0.05686506512095792,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.964092058547281e-05,
+      "loss": 0.9298,
+      "step": 3208
+    },
+    {
+      "epoch": 0.056900517156569484,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.964044878280558e-05,
+      "loss": 0.8457,
+      "step": 3210
+    },
+    {
+      "epoch": 0.05693596919218105,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.963997667263069e-05,
+      "loss": 0.9313,
+      "step": 3212
+    },
+    {
+      "epoch": 0.056971421227792626,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.9639504254954026e-05,
+      "loss": 0.8426,
+      "step": 3214
+    },
+    {
+      "epoch": 0.057006873263404194,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.963903152978148e-05,
+      "loss": 0.8391,
+      "step": 3216
+    },
+    {
+      "epoch": 0.05704232529901576,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.9638558497118956e-05,
+      "loss": 0.8089,
+      "step": 3218
+    },
+    {
+      "epoch": 0.057077777334627336,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.963808515697235e-05,
+      "loss": 0.8812,
+      "step": 3220
+    },
+    {
+      "epoch": 0.057113229370238904,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.963761150934757e-05,
+      "loss": 0.831,
+      "step": 3222
+    },
+    {
+      "epoch": 0.05714868140585047,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.9637137554250535e-05,
+      "loss": 0.9013,
+      "step": 3224
+    },
+    {
+      "epoch": 0.05718413344146204,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.963666329168715e-05,
+      "loss": 0.8752,
+      "step": 3226
+    },
+    {
+      "epoch": 0.05721958547707361,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.963618872166334e-05,
+      "loss": 0.882,
+      "step": 3228
+    },
+    {
+      "epoch": 0.05725503751268518,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.9635713844185025e-05,
+      "loss": 0.9392,
+      "step": 3230
+    },
+    {
+      "epoch": 0.05729048954829675,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.9635238659258136e-05,
+      "loss": 0.7893,
+      "step": 3232
+    },
+    {
+      "epoch": 0.05732594158390832,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.96347631668886e-05,
+      "loss": 0.8776,
+      "step": 3234
+    },
+    {
+      "epoch": 0.05736139361951989,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.9634287367082346e-05,
+      "loss": 0.8508,
+      "step": 3236
+    },
+    {
+      "epoch": 0.05739684565513146,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.963381125984532e-05,
+      "loss": 0.8219,
+      "step": 3238
+    },
+    {
+      "epoch": 0.05743229769074303,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.963333484518346e-05,
+      "loss": 0.8711,
+      "step": 3240
+    },
+    {
+      "epoch": 0.0574677497263546,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.963285812310271e-05,
+      "loss": 0.892,
+      "step": 3242
+    },
+    {
+      "epoch": 0.05750320176196617,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.9632381093609024e-05,
+      "loss": 0.8696,
+      "step": 3244
+    },
+    {
+      "epoch": 0.05753865379757774,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.963190375670835e-05,
+      "loss": 0.8536,
+      "step": 3246
+    },
+    {
+      "epoch": 0.05757410583318931,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.963142611240665e-05,
+      "loss": 0.883,
+      "step": 3248
+    },
+    {
+      "epoch": 0.05760955786880088,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.963094816070988e-05,
+      "loss": 0.9322,
+      "step": 3250
+    },
+    {
+      "epoch": 0.05764500990441245,
+      "grad_norm": 2.75,
+      "learning_rate": 4.963046990162401e-05,
+      "loss": 0.8714,
+      "step": 3252
+    },
+    {
+      "epoch": 0.05768046194002402,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.9629991335155e-05,
+      "loss": 0.8551,
+      "step": 3254
+    },
+    {
+      "epoch": 0.05771591397563559,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.962951246130884e-05,
+      "loss": 0.9051,
+      "step": 3256
+    },
+    {
+      "epoch": 0.05775136601124716,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.962903328009149e-05,
+      "loss": 0.8693,
+      "step": 3258
+    },
+    {
+      "epoch": 0.05778681804685873,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.962855379150893e-05,
+      "loss": 0.8398,
+      "step": 3260
+    },
+    {
+      "epoch": 0.057822270082470296,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.962807399556715e-05,
+      "loss": 0.8445,
+      "step": 3262
+    },
+    {
+      "epoch": 0.05785772211808187,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.962759389227213e-05,
+      "loss": 0.8498,
+      "step": 3264
+    },
+    {
+      "epoch": 0.05789317415369344,
+      "grad_norm": 3.0,
+      "learning_rate": 4.962711348162987e-05,
+      "loss": 0.8548,
+      "step": 3266
+    },
+    {
+      "epoch": 0.057928626189305006,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.962663276364637e-05,
+      "loss": 0.8758,
+      "step": 3268
+    },
+    {
+      "epoch": 0.057964078224916574,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.962615173832762e-05,
+      "loss": 0.8511,
+      "step": 3270
+    },
+    {
+      "epoch": 0.05799953026052815,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.9625670405679626e-05,
+      "loss": 0.8197,
+      "step": 3272
+    },
+    {
+      "epoch": 0.058034982296139716,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.96251887657084e-05,
+      "loss": 0.8447,
+      "step": 3274
+    },
+    {
+      "epoch": 0.05807043433175128,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.962470681841993e-05,
+      "loss": 0.8452,
+      "step": 3276
+    },
+    {
+      "epoch": 0.05810588636736286,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.962422456382026e-05,
+      "loss": 0.8332,
+      "step": 3278
+    },
+    {
+      "epoch": 0.058141338402974425,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.96237420019154e-05,
+      "loss": 0.8803,
+      "step": 3280
+    },
+    {
+      "epoch": 0.05817679043858599,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.9623259132711365e-05,
+      "loss": 0.7956,
+      "step": 3282
+    },
+    {
+      "epoch": 0.05821224247419757,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.9622775956214187e-05,
+      "loss": 0.8399,
+      "step": 3284
+    },
+    {
+      "epoch": 0.058247694509809135,
+      "grad_norm": 2.75,
+      "learning_rate": 4.962229247242989e-05,
+      "loss": 0.8275,
+      "step": 3286
+    },
+    {
+      "epoch": 0.0582831465454207,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.9621808681364506e-05,
+      "loss": 0.8737,
+      "step": 3288
+    },
+    {
+      "epoch": 0.05831859858103228,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.9621324583024085e-05,
+      "loss": 0.8652,
+      "step": 3290
+    },
+    {
+      "epoch": 0.058354050616643845,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.962084017741466e-05,
+      "loss": 0.8329,
+      "step": 3292
+    },
+    {
+      "epoch": 0.05838950265225541,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.962035546454228e-05,
+      "loss": 0.8641,
+      "step": 3294
+    },
+    {
+      "epoch": 0.05842495468786699,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.961987044441299e-05,
+      "loss": 0.8403,
+      "step": 3296
+    },
+    {
+      "epoch": 0.058460406723478554,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.961938511703284e-05,
+      "loss": 0.854,
+      "step": 3298
+    },
+    {
+      "epoch": 0.05849585875909012,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.96188994824079e-05,
+      "loss": 0.8576,
+      "step": 3300
+    },
+    {
+      "epoch": 0.058531310794701696,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.961841354054422e-05,
+      "loss": 0.8808,
+      "step": 3302
+    },
+    {
+      "epoch": 0.058566762830313264,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.961792729144786e-05,
+      "loss": 0.8476,
+      "step": 3304
+    },
+    {
+      "epoch": 0.05860221486592483,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.96174407351249e-05,
+      "loss": 0.8627,
+      "step": 3306
+    },
+    {
+      "epoch": 0.058637666901536406,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.9616953871581406e-05,
+      "loss": 0.8171,
+      "step": 3308
+    },
+    {
+      "epoch": 0.058673118937147974,
+      "grad_norm": 2.875,
+      "learning_rate": 4.9616466700823455e-05,
+      "loss": 0.8769,
+      "step": 3310
+    },
+    {
+      "epoch": 0.05870857097275954,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.961597922285712e-05,
+      "loss": 0.8673,
+      "step": 3312
+    },
+    {
+      "epoch": 0.05874402300837111,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.96154914376885e-05,
+      "loss": 0.8796,
+      "step": 3314
+    },
+    {
+      "epoch": 0.05877947504398268,
+      "grad_norm": 2.875,
+      "learning_rate": 4.961500334532368e-05,
+      "loss": 0.8608,
+      "step": 3316
+    },
+    {
+      "epoch": 0.05881492707959425,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.9614514945768734e-05,
+      "loss": 0.8851,
+      "step": 3318
+    },
+    {
+      "epoch": 0.05885037911520582,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.961402623902978e-05,
+      "loss": 0.8761,
+      "step": 3320
+    },
+    {
+      "epoch": 0.05888583115081739,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.9613537225112893e-05,
+      "loss": 0.8832,
+      "step": 3322
+    },
+    {
+      "epoch": 0.05892128318642896,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.961304790402419e-05,
+      "loss": 0.8012,
+      "step": 3324
+    },
+    {
+      "epoch": 0.05895673522204053,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.961255827576978e-05,
+      "loss": 0.8719,
+      "step": 3326
+    },
+    {
+      "epoch": 0.0589921872576521,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.9612068340355766e-05,
+      "loss": 0.9218,
+      "step": 3328
+    },
+    {
+      "epoch": 0.05902763929326367,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.961157809778827e-05,
+      "loss": 0.862,
+      "step": 3330
+    },
+    {
+      "epoch": 0.05906309132887524,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.96110875480734e-05,
+      "loss": 0.8544,
+      "step": 3332
+    },
+    {
+      "epoch": 0.05909854336448681,
+      "grad_norm": 2.875,
+      "learning_rate": 4.9610596691217284e-05,
+      "loss": 0.8294,
+      "step": 3334
+    },
+    {
+      "epoch": 0.05913399540009838,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.961010552722605e-05,
+      "loss": 0.8704,
+      "step": 3336
+    },
+    {
+      "epoch": 0.05916944743570995,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.960961405610582e-05,
+      "loss": 0.7835,
+      "step": 3338
+    },
+    {
+      "epoch": 0.05920489947132152,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.960912227786274e-05,
+      "loss": 0.8676,
+      "step": 3340
+    },
+    {
+      "epoch": 0.05924035150693309,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.9608630192502935e-05,
+      "loss": 0.86,
+      "step": 3342
+    },
+    {
+      "epoch": 0.05927580354254466,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.960813780003256e-05,
+      "loss": 0.7854,
+      "step": 3344
+    },
+    {
+      "epoch": 0.05931125557815623,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.960764510045774e-05,
+      "loss": 0.8139,
+      "step": 3346
+    },
+    {
+      "epoch": 0.0593467076137678,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.960715209378464e-05,
+      "loss": 0.8455,
+      "step": 3348
+    },
+    {
+      "epoch": 0.05938215964937937,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.96066587800194e-05,
+      "loss": 0.9059,
+      "step": 3350
+    },
+    {
+      "epoch": 0.059417611684990934,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.960616515916819e-05,
+      "loss": 0.8427,
+      "step": 3352
+    },
+    {
+      "epoch": 0.05945306372060251,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.960567123123716e-05,
+      "loss": 0.8606,
+      "step": 3354
+    },
+    {
+      "epoch": 0.059488515756214076,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.960517699623248e-05,
+      "loss": 0.9067,
+      "step": 3356
+    },
+    {
+      "epoch": 0.059523967791825644,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.960468245416032e-05,
+      "loss": 0.873,
+      "step": 3358
+    },
+    {
+      "epoch": 0.05955941982743722,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.9604187605026845e-05,
+      "loss": 0.8933,
+      "step": 3360
+    },
+    {
+      "epoch": 0.059594871863048786,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.960369244883823e-05,
+      "loss": 0.8775,
+      "step": 3362
+    },
+    {
+      "epoch": 0.05963032389866035,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.960319698560066e-05,
+      "loss": 0.8348,
+      "step": 3364
+    },
+    {
+      "epoch": 0.05966577593427193,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.960270121532031e-05,
+      "loss": 0.8505,
+      "step": 3366
+    },
+    {
+      "epoch": 0.059701227969883496,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.960220513800339e-05,
+      "loss": 0.8609,
+      "step": 3368
+    },
+    {
+      "epoch": 0.05973668000549506,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.960170875365606e-05,
+      "loss": 0.8479,
+      "step": 3370
+    },
+    {
+      "epoch": 0.05977213204110664,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.960121206228453e-05,
+      "loss": 0.8407,
+      "step": 3372
+    },
+    {
+      "epoch": 0.059807584076718205,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.9600715063895e-05,
+      "loss": 0.8408,
+      "step": 3374
+    },
+    {
+      "epoch": 0.05984303611232977,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.960021775849367e-05,
+      "loss": 0.8333,
+      "step": 3376
+    },
+    {
+      "epoch": 0.05987848814794135,
+      "grad_norm": 2.875,
+      "learning_rate": 4.959972014608675e-05,
+      "loss": 0.8421,
+      "step": 3378
+    },
+    {
+      "epoch": 0.059913940183552915,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.959922222668044e-05,
+      "loss": 0.8293,
+      "step": 3380
+    },
+    {
+      "epoch": 0.05994939221916448,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.959872400028096e-05,
+      "loss": 0.8708,
+      "step": 3382
+    },
+    {
+      "epoch": 0.05998484425477606,
+      "grad_norm": 2.625,
+      "learning_rate": 4.959822546689453e-05,
+      "loss": 0.8583,
+      "step": 3384
+    },
+    {
+      "epoch": 0.060020296290387624,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.959772662652737e-05,
+      "loss": 0.8624,
+      "step": 3386
+    },
+    {
+      "epoch": 0.06005574832599919,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.959722747918571e-05,
+      "loss": 0.879,
+      "step": 3388
+    },
+    {
+      "epoch": 0.060091200361610767,
+      "grad_norm": 2.75,
+      "learning_rate": 4.9596728024875774e-05,
+      "loss": 0.8485,
+      "step": 3390
+    },
+    {
+      "epoch": 0.060126652397222334,
+      "grad_norm": 2.875,
+      "learning_rate": 4.959622826360378e-05,
+      "loss": 0.8543,
+      "step": 3392
+    },
+    {
+      "epoch": 0.0601621044328339,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.9595728195375996e-05,
+      "loss": 0.8681,
+      "step": 3394
+    },
+    {
+      "epoch": 0.06019755646844547,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.959522782019864e-05,
+      "loss": 0.8927,
+      "step": 3396
+    },
+    {
+      "epoch": 0.060233008504057044,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.9594727138077967e-05,
+      "loss": 0.8581,
+      "step": 3398
+    },
+    {
+      "epoch": 0.06026846053966861,
+      "grad_norm": 2.625,
+      "learning_rate": 4.9594226149020226e-05,
+      "loss": 0.8652,
+      "step": 3400
+    },
+    {
+      "epoch": 0.06030391257528018,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.959372485303165e-05,
+      "loss": 0.8713,
+      "step": 3402
+    },
+    {
+      "epoch": 0.06033936461089175,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.9593223250118524e-05,
+      "loss": 0.8531,
+      "step": 3404
+    },
+    {
+      "epoch": 0.06037481664650332,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.95927213402871e-05,
+      "loss": 0.869,
+      "step": 3406
+    },
+    {
+      "epoch": 0.06041026868211489,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.959221912354362e-05,
+      "loss": 0.8642,
+      "step": 3408
+    },
+    {
+      "epoch": 0.06044572071772646,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.959171659989438e-05,
+      "loss": 0.8605,
+      "step": 3410
+    },
+    {
+      "epoch": 0.06048117275333803,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.959121376934563e-05,
+      "loss": 0.858,
+      "step": 3412
+    },
+    {
+      "epoch": 0.0605166247889496,
+      "grad_norm": 2.75,
+      "learning_rate": 4.959071063190366e-05,
+      "loss": 0.8314,
+      "step": 3414
+    },
+    {
+      "epoch": 0.06055207682456117,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.959020718757474e-05,
+      "loss": 0.8443,
+      "step": 3416
+    },
+    {
+      "epoch": 0.06058752886017274,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.9589703436365156e-05,
+      "loss": 0.8542,
+      "step": 3418
+    },
+    {
+      "epoch": 0.06062298089578431,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.9589199378281194e-05,
+      "loss": 0.844,
+      "step": 3420
+    },
+    {
+      "epoch": 0.06065843293139588,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.958869501332914e-05,
+      "loss": 0.8895,
+      "step": 3422
+    },
+    {
+      "epoch": 0.06069388496700745,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.958819034151531e-05,
+      "loss": 0.8338,
+      "step": 3424
+    },
+    {
+      "epoch": 0.06072933700261902,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.958768536284597e-05,
+      "loss": 0.8165,
+      "step": 3426
+    },
+    {
+      "epoch": 0.06076478903823059,
+      "grad_norm": 2.625,
+      "learning_rate": 4.9587180077327444e-05,
+      "loss": 0.7967,
+      "step": 3428
+    },
+    {
+      "epoch": 0.06080024107384216,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.958667448496604e-05,
+      "loss": 0.8665,
+      "step": 3430
+    },
+    {
+      "epoch": 0.06083569310945373,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.958616858576804e-05,
+      "loss": 0.8432,
+      "step": 3432
+    },
+    {
+      "epoch": 0.0608711451450653,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.9585662379739796e-05,
+      "loss": 0.8507,
+      "step": 3434
+    },
+    {
+      "epoch": 0.06090659718067687,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.95851558668876e-05,
+      "loss": 0.8548,
+      "step": 3436
+    },
+    {
+      "epoch": 0.06094204921628844,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.958464904721778e-05,
+      "loss": 0.8712,
+      "step": 3438
+    },
+    {
+      "epoch": 0.060977501251900004,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.9584141920736656e-05,
+      "loss": 0.8419,
+      "step": 3440
+    },
+    {
+      "epoch": 0.06101295328751158,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.9583634487450565e-05,
+      "loss": 0.8738,
+      "step": 3442
+    },
+    {
+      "epoch": 0.061048405323123146,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.9583126747365834e-05,
+      "loss": 0.8648,
+      "step": 3444
+    },
+    {
+      "epoch": 0.061083857358734714,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.9582618700488805e-05,
+      "loss": 0.8645,
+      "step": 3446
+    },
+    {
+      "epoch": 0.06111930939434629,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.9582110346825814e-05,
+      "loss": 0.8694,
+      "step": 3448
+    },
+    {
+      "epoch": 0.061154761429957856,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.95816016863832e-05,
+      "loss": 0.8641,
+      "step": 3450
+    },
+    {
+      "epoch": 0.061190213465569424,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.9581092719167324e-05,
+      "loss": 0.8866,
+      "step": 3452
+    },
+    {
+      "epoch": 0.061225665501181,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.958058344518452e-05,
+      "loss": 0.895,
+      "step": 3454
+    },
+    {
+      "epoch": 0.061261117536792566,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.958007386444117e-05,
+      "loss": 0.8642,
+      "step": 3456
+    },
+    {
+      "epoch": 0.06129656957240413,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.957956397694361e-05,
+      "loss": 0.8568,
+      "step": 3458
+    },
+    {
+      "epoch": 0.06133202160801571,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.957905378269821e-05,
+      "loss": 0.8839,
+      "step": 3460
+    },
+    {
+      "epoch": 0.061367473643627275,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.9578543281711345e-05,
+      "loss": 0.842,
+      "step": 3462
+    },
+    {
+      "epoch": 0.06140292567923884,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.9578032473989364e-05,
+      "loss": 0.8433,
+      "step": 3464
+    },
+    {
+      "epoch": 0.06143837771485042,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.957752135953867e-05,
+      "loss": 0.8188,
+      "step": 3466
+    },
+    {
+      "epoch": 0.061473829750461985,
+      "grad_norm": 3.0,
+      "learning_rate": 4.9577009938365624e-05,
+      "loss": 0.8758,
+      "step": 3468
+    },
+    {
+      "epoch": 0.06150928178607355,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.957649821047662e-05,
+      "loss": 0.8724,
+      "step": 3470
+    },
+    {
+      "epoch": 0.06154473382168513,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.957598617587803e-05,
+      "loss": 0.8746,
+      "step": 3472
+    },
+    {
+      "epoch": 0.061580185857296695,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.957547383457625e-05,
+      "loss": 0.8618,
+      "step": 3474
+    },
+    {
+      "epoch": 0.06161563789290826,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.957496118657768e-05,
+      "loss": 0.8622,
+      "step": 3476
+    },
+    {
+      "epoch": 0.06165108992851984,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.957444823188871e-05,
+      "loss": 0.8822,
+      "step": 3478
+    },
+    {
+      "epoch": 0.061686541964131404,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.9573934970515744e-05,
+      "loss": 0.8869,
+      "step": 3480
+    },
+    {
+      "epoch": 0.06172199399974297,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.957342140246519e-05,
+      "loss": 0.8593,
+      "step": 3482
+    },
+    {
+      "epoch": 0.06175744603535454,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.957290752774346e-05,
+      "loss": 0.8421,
+      "step": 3484
+    },
+    {
+      "epoch": 0.061792898070966114,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.957239334635696e-05,
+      "loss": 0.8195,
+      "step": 3486
+    },
+    {
+      "epoch": 0.06182835010657768,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.95718788583121e-05,
+      "loss": 0.8544,
+      "step": 3488
+    },
+    {
+      "epoch": 0.06186380214218925,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.957136406361532e-05,
+      "loss": 0.8591,
+      "step": 3490
+    },
+    {
+      "epoch": 0.06189925417780082,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.957084896227303e-05,
+      "loss": 0.8311,
+      "step": 3492
+    },
+    {
+      "epoch": 0.06193470621341239,
+      "grad_norm": 3.125,
+      "learning_rate": 4.957033355429166e-05,
+      "loss": 0.8216,
+      "step": 3494
+    },
+    {
+      "epoch": 0.06197015824902396,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.9569817839677646e-05,
+      "loss": 0.8608,
+      "step": 3496
+    },
+    {
+      "epoch": 0.06200561028463553,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.956930181843742e-05,
+      "loss": 0.8534,
+      "step": 3498
+    },
+    {
+      "epoch": 0.0620410623202471,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.956878549057743e-05,
+      "loss": 0.8135,
+      "step": 3500
+    },
+    {
+      "epoch": 0.06207651435585867,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.956826885610412e-05,
+      "loss": 0.8465,
+      "step": 3502
+    },
+    {
+      "epoch": 0.06211196639147024,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.9567751915023925e-05,
+      "loss": 0.8692,
+      "step": 3504
+    },
+    {
+      "epoch": 0.06214741842708181,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.9567234667343305e-05,
+      "loss": 0.8903,
+      "step": 3506
+    },
+    {
+      "epoch": 0.06218287046269338,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.9566717113068715e-05,
+      "loss": 0.8463,
+      "step": 3508
+    },
+    {
+      "epoch": 0.06221832249830495,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.9566199252206605e-05,
+      "loss": 0.8344,
+      "step": 3510
+    },
+    {
+      "epoch": 0.06225377453391652,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.956568108476345e-05,
+      "loss": 0.8576,
+      "step": 3512
+    },
+    {
+      "epoch": 0.06228922656952809,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.956516261074571e-05,
+      "loss": 0.8552,
+      "step": 3514
+    },
+    {
+      "epoch": 0.06232467860513966,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.956464383015986e-05,
+      "loss": 0.8446,
+      "step": 3516
+    },
+    {
+      "epoch": 0.06236013064075123,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.956412474301237e-05,
+      "loss": 0.8456,
+      "step": 3518
+    },
+    {
+      "epoch": 0.0623955826763628,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.9563605349309714e-05,
+      "loss": 0.8691,
+      "step": 3520
+    },
+    {
+      "epoch": 0.062431034711974365,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.9563085649058395e-05,
+      "loss": 0.8493,
+      "step": 3522
+    },
+    {
+      "epoch": 0.06246648674758594,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.956256564226487e-05,
+      "loss": 0.8059,
+      "step": 3524
+    },
+    {
+      "epoch": 0.06250193878319751,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.9562045328935644e-05,
+      "loss": 0.8029,
+      "step": 3526
+    },
+    {
+      "epoch": 0.06253739081880907,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.9561524709077215e-05,
+      "loss": 0.8365,
+      "step": 3528
+    },
+    {
+      "epoch": 0.06257284285442065,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.9561003782696055e-05,
+      "loss": 0.8374,
+      "step": 3530
+    },
+    {
+      "epoch": 0.06260829489003222,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.95604825497987e-05,
+      "loss": 0.8928,
+      "step": 3532
+    },
+    {
+      "epoch": 0.06264374692564378,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.955996101039164e-05,
+      "loss": 0.8294,
+      "step": 3534
+    },
+    {
+      "epoch": 0.06267919896125536,
+      "grad_norm": 3.0,
+      "learning_rate": 4.955943916448137e-05,
+      "loss": 0.85,
+      "step": 3536
+    },
+    {
+      "epoch": 0.06271465099686693,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.9558917012074425e-05,
+      "loss": 0.8115,
+      "step": 3538
+    },
+    {
+      "epoch": 0.0627501030324785,
+      "grad_norm": 3.0,
+      "learning_rate": 4.955839455317731e-05,
+      "loss": 0.8667,
+      "step": 3540
+    },
+    {
+      "epoch": 0.06278555506809007,
+      "grad_norm": 2.625,
+      "learning_rate": 4.955787178779654e-05,
+      "loss": 0.8485,
+      "step": 3542
+    },
+    {
+      "epoch": 0.06282100710370163,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.9557348715938646e-05,
+      "loss": 0.8938,
+      "step": 3544
+    },
+    {
+      "epoch": 0.0628564591393132,
+      "grad_norm": 2.625,
+      "learning_rate": 4.9556825337610156e-05,
+      "loss": 0.8341,
+      "step": 3546
+    },
+    {
+      "epoch": 0.06289191117492478,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.9556301652817604e-05,
+      "loss": 0.8989,
+      "step": 3548
+    },
+    {
+      "epoch": 0.06292736321053634,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.955577766156752e-05,
+      "loss": 0.853,
+      "step": 3550
+    },
+    {
+      "epoch": 0.06296281524614791,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.955525336386645e-05,
+      "loss": 0.8993,
+      "step": 3552
+    },
+    {
+      "epoch": 0.06299826728175949,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.9554728759720925e-05,
+      "loss": 0.8373,
+      "step": 3554
+    },
+    {
+      "epoch": 0.06303371931737105,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.95542038491375e-05,
+      "loss": 0.8574,
+      "step": 3556
+    },
+    {
+      "epoch": 0.06306917135298262,
+      "grad_norm": 3.0,
+      "learning_rate": 4.9553678632122724e-05,
+      "loss": 0.9052,
+      "step": 3558
+    },
+    {
+      "epoch": 0.0631046233885942,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.955315310868316e-05,
+      "loss": 0.8823,
+      "step": 3560
+    },
+    {
+      "epoch": 0.06314007542420576,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.955262727882536e-05,
+      "loss": 0.8426,
+      "step": 3562
+    },
+    {
+      "epoch": 0.06317552745981733,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.9552101142555874e-05,
+      "loss": 0.8356,
+      "step": 3564
+    },
+    {
+      "epoch": 0.0632109794954289,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.9551574699881285e-05,
+      "loss": 0.8112,
+      "step": 3566
+    },
+    {
+      "epoch": 0.06324643153104047,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.955104795080816e-05,
+      "loss": 0.8955,
+      "step": 3568
+    },
+    {
+      "epoch": 0.06328188356665204,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.955052089534308e-05,
+      "loss": 0.8834,
+      "step": 3570
+    },
+    {
+      "epoch": 0.06331733560226362,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.9549993533492595e-05,
+      "loss": 0.8221,
+      "step": 3572
+    },
+    {
+      "epoch": 0.06335278763787518,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.9549465865263314e-05,
+      "loss": 0.9124,
+      "step": 3574
+    },
+    {
+      "epoch": 0.06338823967348675,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.954893789066181e-05,
+      "loss": 0.8662,
+      "step": 3576
+    },
+    {
+      "epoch": 0.06342369170909833,
+      "grad_norm": 3.0,
+      "learning_rate": 4.954840960969467e-05,
+      "loss": 0.8301,
+      "step": 3578
+    },
+    {
+      "epoch": 0.06345914374470989,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.9547881022368495e-05,
+      "loss": 0.8467,
+      "step": 3580
+    },
+    {
+      "epoch": 0.06349459578032146,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.954735212868988e-05,
+      "loss": 0.8899,
+      "step": 3582
+    },
+    {
+      "epoch": 0.06353004781593304,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.954682292866542e-05,
+      "loss": 0.8437,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0635654998515446,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.9546293422301724e-05,
+      "loss": 0.8412,
+      "step": 3586
+    },
+    {
+      "epoch": 0.06360095188715617,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.95457636096054e-05,
+      "loss": 0.7977,
+      "step": 3588
+    },
+    {
+      "epoch": 0.06363640392276775,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.9545233490583057e-05,
+      "loss": 0.8696,
+      "step": 3590
+    },
+    {
+      "epoch": 0.0636718559583793,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.954470306524131e-05,
+      "loss": 0.8215,
+      "step": 3592
+    },
+    {
+      "epoch": 0.06370730799399088,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.954417233358678e-05,
+      "loss": 0.8238,
+      "step": 3594
+    },
+    {
+      "epoch": 0.06374276002960245,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.9543641295626096e-05,
+      "loss": 0.8648,
+      "step": 3596
+    },
+    {
+      "epoch": 0.06377821206521402,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.9543109951365886e-05,
+      "loss": 0.8195,
+      "step": 3598
+    },
+    {
+      "epoch": 0.06381366410082559,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.954257830081276e-05,
+      "loss": 0.8593,
+      "step": 3600
+    },
+    {
+      "epoch": 0.06384911613643716,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.954204634397338e-05,
+      "loss": 0.8781,
+      "step": 3602
+    },
+    {
+      "epoch": 0.06388456817204873,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.9541514080854375e-05,
+      "loss": 0.8603,
+      "step": 3604
+    },
+    {
+      "epoch": 0.0639200202076603,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.954098151146238e-05,
+      "loss": 0.8483,
+      "step": 3606
+    },
+    {
+      "epoch": 0.06395547224327187,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.954044863580405e-05,
+      "loss": 0.8716,
+      "step": 3608
+    },
+    {
+      "epoch": 0.06399092427888343,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.953991545388603e-05,
+      "loss": 0.8673,
+      "step": 3610
+    },
+    {
+      "epoch": 0.06402637631449501,
+      "grad_norm": 3.125,
+      "learning_rate": 4.953938196571498e-05,
+      "loss": 0.8617,
+      "step": 3612
+    },
+    {
+      "epoch": 0.06406182835010658,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.953884817129755e-05,
+      "loss": 0.857,
+      "step": 3614
+    },
+    {
+      "epoch": 0.06409728038571814,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.953831407064041e-05,
+      "loss": 0.853,
+      "step": 3616
+    },
+    {
+      "epoch": 0.06413273242132972,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.9537779663750225e-05,
+      "loss": 0.8622,
+      "step": 3618
+    },
+    {
+      "epoch": 0.0641681844569413,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.953724495063365e-05,
+      "loss": 0.8242,
+      "step": 3620
+    },
+    {
+      "epoch": 0.06420363649255285,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.953670993129738e-05,
+      "loss": 0.8159,
+      "step": 3622
+    },
+    {
+      "epoch": 0.06423908852816443,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.953617460574807e-05,
+      "loss": 0.8065,
+      "step": 3624
+    },
+    {
+      "epoch": 0.064274540563776,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.9535638973992416e-05,
+      "loss": 0.8516,
+      "step": 3626
+    },
+    {
+      "epoch": 0.06430999259938756,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.95351030360371e-05,
+      "loss": 0.8336,
+      "step": 3628
+    },
+    {
+      "epoch": 0.06434544463499914,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.9534566791888804e-05,
+      "loss": 0.8426,
+      "step": 3630
+    },
+    {
+      "epoch": 0.0643808966706107,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.953403024155423e-05,
+      "loss": 0.849,
+      "step": 3632
+    },
+    {
+      "epoch": 0.06441634870622227,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.9533493385040067e-05,
+      "loss": 0.8788,
+      "step": 3634
+    },
+    {
+      "epoch": 0.06445180074183385,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.9532956222353014e-05,
+      "loss": 0.8311,
+      "step": 3636
+    },
+    {
+      "epoch": 0.06448725277744541,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.953241875349978e-05,
+      "loss": 0.8983,
+      "step": 3638
+    },
+    {
+      "epoch": 0.06452270481305698,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.9531880978487065e-05,
+      "loss": 0.8764,
+      "step": 3640
+    },
+    {
+      "epoch": 0.06455815684866856,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.953134289732159e-05,
+      "loss": 0.835,
+      "step": 3642
+    },
+    {
+      "epoch": 0.06459360888428012,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.9530804510010065e-05,
+      "loss": 0.8383,
+      "step": 3644
+    },
+    {
+      "epoch": 0.06462906091989169,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.953026581655921e-05,
+      "loss": 0.856,
+      "step": 3646
+    },
+    {
+      "epoch": 0.06466451295550327,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.952972681697574e-05,
+      "loss": 0.8197,
+      "step": 3648
+    },
+    {
+      "epoch": 0.06469996499111483,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.9529187511266395e-05,
+      "loss": 0.8752,
+      "step": 3650
+    },
+    {
+      "epoch": 0.0647354170267264,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.9528647899437894e-05,
+      "loss": 0.8281,
+      "step": 3652
+    },
+    {
+      "epoch": 0.06477086906233798,
+      "grad_norm": 2.875,
+      "learning_rate": 4.9528107981496985e-05,
+      "loss": 0.8465,
+      "step": 3654
+    },
+    {
+      "epoch": 0.06480632109794954,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.952756775745039e-05,
+      "loss": 0.8236,
+      "step": 3656
+    },
+    {
+      "epoch": 0.06484177313356111,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.952702722730486e-05,
+      "loss": 0.8802,
+      "step": 3658
+    },
+    {
+      "epoch": 0.06487722516917269,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.952648639106714e-05,
+      "loss": 0.8799,
+      "step": 3660
+    },
+    {
+      "epoch": 0.06491267720478425,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.9525945248743974e-05,
+      "loss": 0.8126,
+      "step": 3662
+    },
+    {
+      "epoch": 0.06494812924039582,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.952540380034212e-05,
+      "loss": 0.842,
+      "step": 3664
+    },
+    {
+      "epoch": 0.0649835812760074,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.952486204586834e-05,
+      "loss": 0.8368,
+      "step": 3666
+    },
+    {
+      "epoch": 0.06501903331161896,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.952431998532939e-05,
+      "loss": 0.8623,
+      "step": 3668
+    },
+    {
+      "epoch": 0.06505448534723053,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.952377761873203e-05,
+      "loss": 0.8137,
+      "step": 3670
+    },
+    {
+      "epoch": 0.0650899373828421,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.952323494608303e-05,
+      "loss": 0.8041,
+      "step": 3672
+    },
+    {
+      "epoch": 0.06512538941845367,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.9522691967389175e-05,
+      "loss": 0.8597,
+      "step": 3674
+    },
+    {
+      "epoch": 0.06516084145406524,
+      "grad_norm": 2.875,
+      "learning_rate": 4.952214868265723e-05,
+      "loss": 0.8554,
+      "step": 3676
+    },
+    {
+      "epoch": 0.06519629348967682,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.952160509189397e-05,
+      "loss": 0.8432,
+      "step": 3678
+    },
+    {
+      "epoch": 0.06523174552528838,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.952106119510619e-05,
+      "loss": 0.8919,
+      "step": 3680
+    },
+    {
+      "epoch": 0.06526719756089995,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.9520516992300675e-05,
+      "loss": 0.862,
+      "step": 3682
+    },
+    {
+      "epoch": 0.06530264959651153,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.9519972483484214e-05,
+      "loss": 0.8214,
+      "step": 3684
+    },
+    {
+      "epoch": 0.06533810163212309,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.9519427668663603e-05,
+      "loss": 0.8427,
+      "step": 3686
+    },
+    {
+      "epoch": 0.06537355366773466,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.9518882547845645e-05,
+      "loss": 0.8962,
+      "step": 3688
+    },
+    {
+      "epoch": 0.06540900570334623,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.951833712103714e-05,
+      "loss": 0.8666,
+      "step": 3690
+    },
+    {
+      "epoch": 0.0654444577389578,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.951779138824489e-05,
+      "loss": 0.8367,
+      "step": 3692
+    },
+    {
+      "epoch": 0.06547990977456937,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.951724534947571e-05,
+      "loss": 0.8496,
+      "step": 3694
+    },
+    {
+      "epoch": 0.06551536181018094,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.9516699004736415e-05,
+      "loss": 0.878,
+      "step": 3696
+    },
+    {
+      "epoch": 0.0655508138457925,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.9516152354033826e-05,
+      "loss": 0.8568,
+      "step": 3698
+    },
+    {
+      "epoch": 0.06558626588140408,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.9515605397374765e-05,
+      "loss": 0.8316,
+      "step": 3700
+    },
+    {
+      "epoch": 0.06562171791701565,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.951505813476605e-05,
+      "loss": 0.8079,
+      "step": 3702
+    },
+    {
+      "epoch": 0.06565716995262721,
+      "grad_norm": 2.5,
+      "learning_rate": 4.951451056621451e-05,
+      "loss": 0.9063,
+      "step": 3704
+    },
+    {
+      "epoch": 0.06569262198823879,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.9513962691726986e-05,
+      "loss": 0.8466,
+      "step": 3706
+    },
+    {
+      "epoch": 0.06572807402385036,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.9513414511310325e-05,
+      "loss": 0.917,
+      "step": 3708
+    },
+    {
+      "epoch": 0.06576352605946192,
+      "grad_norm": 2.75,
+      "learning_rate": 4.951286602497135e-05,
+      "loss": 0.8531,
+      "step": 3710
+    },
+    {
+      "epoch": 0.0657989780950735,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.951231723271691e-05,
+      "loss": 0.8474,
+      "step": 3712
+    },
+    {
+      "epoch": 0.06583443013068506,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.951176813455386e-05,
+      "loss": 0.839,
+      "step": 3714
+    },
+    {
+      "epoch": 0.06586988216629663,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.951121873048905e-05,
+      "loss": 0.8619,
+      "step": 3716
+    },
+    {
+      "epoch": 0.06590533420190821,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.951066902052933e-05,
+      "loss": 0.8007,
+      "step": 3718
+    },
+    {
+      "epoch": 0.06594078623751977,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.951011900468157e-05,
+      "loss": 0.9208,
+      "step": 3720
+    },
+    {
+      "epoch": 0.06597623827313134,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.9509568682952627e-05,
+      "loss": 0.8499,
+      "step": 3722
+    },
+    {
+      "epoch": 0.06601169030874292,
+      "grad_norm": 2.875,
+      "learning_rate": 4.9509018055349374e-05,
+      "loss": 0.8741,
+      "step": 3724
+    },
+    {
+      "epoch": 0.06604714234435448,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.950846712187868e-05,
+      "loss": 0.8479,
+      "step": 3726
+    },
+    {
+      "epoch": 0.06608259437996605,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.950791588254742e-05,
+      "loss": 0.8415,
+      "step": 3728
+    },
+    {
+      "epoch": 0.06611804641557763,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.950736433736248e-05,
+      "loss": 0.9057,
+      "step": 3730
+    },
+    {
+      "epoch": 0.06615349845118919,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.9506812486330734e-05,
+      "loss": 0.8189,
+      "step": 3732
+    },
+    {
+      "epoch": 0.06618895048680076,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.950626032945907e-05,
+      "loss": 0.884,
+      "step": 3734
+    },
+    {
+      "epoch": 0.06622440252241234,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.950570786675438e-05,
+      "loss": 0.8709,
+      "step": 3736
+    },
+    {
+      "epoch": 0.0662598545580239,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.9505155098223565e-05,
+      "loss": 0.845,
+      "step": 3738
+    },
+    {
+      "epoch": 0.06629530659363547,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.9504602023873514e-05,
+      "loss": 0.8461,
+      "step": 3740
+    },
+    {
+      "epoch": 0.06633075862924705,
+      "grad_norm": 3.125,
+      "learning_rate": 4.950404864371114e-05,
+      "loss": 0.8842,
+      "step": 3742
+    },
+    {
+      "epoch": 0.06636621066485861,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.950349495774333e-05,
+      "loss": 0.8245,
+      "step": 3744
+    },
+    {
+      "epoch": 0.06640166270047018,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.9502940965977026e-05,
+      "loss": 0.8499,
+      "step": 3746
+    },
+    {
+      "epoch": 0.06643711473608176,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.950238666841911e-05,
+      "loss": 0.8319,
+      "step": 3748
+    },
+    {
+      "epoch": 0.06647256677169332,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.950183206507651e-05,
+      "loss": 0.8837,
+      "step": 3750
+    },
+    {
+      "epoch": 0.06650801880730489,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.9501277155956164e-05,
+      "loss": 0.8457,
+      "step": 3752
+    },
+    {
+      "epoch": 0.06654347084291647,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.9500721941064964e-05,
+      "loss": 0.8314,
+      "step": 3754
+    },
+    {
+      "epoch": 0.06657892287852803,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.9500166420409866e-05,
+      "loss": 0.8802,
+      "step": 3756
+    },
+    {
+      "epoch": 0.0666143749141396,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.949961059399779e-05,
+      "loss": 0.8275,
+      "step": 3758
+    },
+    {
+      "epoch": 0.06664982694975118,
+      "grad_norm": 2.75,
+      "learning_rate": 4.9499054461835684e-05,
+      "loss": 0.8344,
+      "step": 3760
+    },
+    {
+      "epoch": 0.06668527898536274,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.949849802393047e-05,
+      "loss": 0.8534,
+      "step": 3762
+    },
+    {
+      "epoch": 0.06672073102097431,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.9497941280289116e-05,
+      "loss": 0.8407,
+      "step": 3764
+    },
+    {
+      "epoch": 0.06675618305658589,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.949738423091855e-05,
+      "loss": 0.8525,
+      "step": 3766
+    },
+    {
+      "epoch": 0.06679163509219745,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.949682687582573e-05,
+      "loss": 0.8505,
+      "step": 3768
+    },
+    {
+      "epoch": 0.06682708712780902,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.9496269215017624e-05,
+      "loss": 0.829,
+      "step": 3770
+    },
+    {
+      "epoch": 0.0668625391634206,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.949571124850116e-05,
+      "loss": 0.8904,
+      "step": 3772
+    },
+    {
+      "epoch": 0.06689799119903216,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.949515297628334e-05,
+      "loss": 0.8788,
+      "step": 3774
+    },
+    {
+      "epoch": 0.06693344323464373,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.949459439837111e-05,
+      "loss": 0.817,
+      "step": 3776
+    },
+    {
+      "epoch": 0.0669688952702553,
+      "grad_norm": 2.625,
+      "learning_rate": 4.949403551477144e-05,
+      "loss": 0.8703,
+      "step": 3778
+    },
+    {
+      "epoch": 0.06700434730586687,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.9493476325491306e-05,
+      "loss": 0.8258,
+      "step": 3780
+    },
+    {
+      "epoch": 0.06703979934147844,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.949291683053769e-05,
+      "loss": 0.859,
+      "step": 3782
+    },
+    {
+      "epoch": 0.06707525137709001,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.949235702991757e-05,
+      "loss": 0.8668,
+      "step": 3784
+    },
+    {
+      "epoch": 0.06711070341270158,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.9491796923637945e-05,
+      "loss": 0.8673,
+      "step": 3786
+    },
+    {
+      "epoch": 0.06714615544831315,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.949123651170579e-05,
+      "loss": 0.8853,
+      "step": 3788
+    },
+    {
+      "epoch": 0.06718160748392472,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.9490675794128105e-05,
+      "loss": 0.8029,
+      "step": 3790
+    },
+    {
+      "epoch": 0.06721705951953628,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.9490114770911886e-05,
+      "loss": 0.858,
+      "step": 3792
+    },
+    {
+      "epoch": 0.06725251155514786,
+      "grad_norm": 2.75,
+      "learning_rate": 4.948955344206414e-05,
+      "loss": 0.8789,
+      "step": 3794
+    },
+    {
+      "epoch": 0.06728796359075943,
+      "grad_norm": 2.75,
+      "learning_rate": 4.948899180759187e-05,
+      "loss": 0.8451,
+      "step": 3796
+    },
+    {
+      "epoch": 0.067323415626371,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.948842986750207e-05,
+      "loss": 0.8629,
+      "step": 3798
+    },
+    {
+      "epoch": 0.06735886766198257,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.948786762180178e-05,
+      "loss": 0.8486,
+      "step": 3800
+    },
+    {
+      "epoch": 0.06739431969759413,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.9487305070498e-05,
+      "loss": 0.8094,
+      "step": 3802
+    },
+    {
+      "epoch": 0.0674297717332057,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.9486742213597745e-05,
+      "loss": 0.8793,
+      "step": 3804
+    },
+    {
+      "epoch": 0.06746522376881728,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.9486179051108054e-05,
+      "loss": 0.8378,
+      "step": 3806
+    },
+    {
+      "epoch": 0.06750067580442884,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.9485615583035946e-05,
+      "loss": 0.8348,
+      "step": 3808
+    },
+    {
+      "epoch": 0.06753612784004041,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.948505180938846e-05,
+      "loss": 0.83,
+      "step": 3810
+    },
+    {
+      "epoch": 0.06757157987565199,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.9484487730172624e-05,
+      "loss": 0.8492,
+      "step": 3812
+    },
+    {
+      "epoch": 0.06760703191126355,
+      "grad_norm": 2.875,
+      "learning_rate": 4.948392334539548e-05,
+      "loss": 0.9011,
+      "step": 3814
+    },
+    {
+      "epoch": 0.06764248394687512,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.948335865506407e-05,
+      "loss": 0.8411,
+      "step": 3816
+    },
+    {
+      "epoch": 0.0676779359824867,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.948279365918544e-05,
+      "loss": 0.8459,
+      "step": 3818
+    },
+    {
+      "epoch": 0.06771338801809826,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.948222835776666e-05,
+      "loss": 0.8186,
+      "step": 3820
+    },
+    {
+      "epoch": 0.06774884005370983,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.948166275081476e-05,
+      "loss": 0.8643,
+      "step": 3822
+    },
+    {
+      "epoch": 0.06778429208932141,
+      "grad_norm": 3.125,
+      "learning_rate": 4.9481096838336804e-05,
+      "loss": 0.8403,
+      "step": 3824
+    },
+    {
+      "epoch": 0.06781974412493297,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.948053062033986e-05,
+      "loss": 0.8526,
+      "step": 3826
+    },
+    {
+      "epoch": 0.06785519616054454,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.9479964096831e-05,
+      "loss": 0.8543,
+      "step": 3828
+    },
+    {
+      "epoch": 0.06789064819615612,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.947939726781729e-05,
+      "loss": 0.8389,
+      "step": 3830
+    },
+    {
+      "epoch": 0.06792610023176768,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.947883013330579e-05,
+      "loss": 0.8496,
+      "step": 3832
+    },
+    {
+      "epoch": 0.06796155226737925,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.947826269330359e-05,
+      "loss": 0.8223,
+      "step": 3834
+    },
+    {
+      "epoch": 0.06799700430299083,
+      "grad_norm": 2.625,
+      "learning_rate": 4.947769494781777e-05,
+      "loss": 0.8621,
+      "step": 3836
+    },
+    {
+      "epoch": 0.06803245633860239,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.947712689685542e-05,
+      "loss": 0.903,
+      "step": 3838
+    },
+    {
+      "epoch": 0.06806790837421396,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.947655854042362e-05,
+      "loss": 0.845,
+      "step": 3840
+    },
+    {
+      "epoch": 0.06810336040982554,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.947598987852947e-05,
+      "loss": 0.858,
+      "step": 3842
+    },
+    {
+      "epoch": 0.0681388124454371,
+      "grad_norm": 2.625,
+      "learning_rate": 4.947542091118006e-05,
+      "loss": 0.8404,
+      "step": 3844
+    },
+    {
+      "epoch": 0.06817426448104867,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.9474851638382504e-05,
+      "loss": 0.8334,
+      "step": 3846
+    },
+    {
+      "epoch": 0.06820971651666025,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.9474282060143885e-05,
+      "loss": 0.8613,
+      "step": 3848
+    },
+    {
+      "epoch": 0.0682451685522718,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.947371217647133e-05,
+      "loss": 0.8584,
+      "step": 3850
+    },
+    {
+      "epoch": 0.06828062058788338,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.947314198737195e-05,
+      "loss": 0.8718,
+      "step": 3852
+    },
+    {
+      "epoch": 0.06831607262349496,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.947257149285285e-05,
+      "loss": 0.8363,
+      "step": 3854
+    },
+    {
+      "epoch": 0.06835152465910652,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.947200069292115e-05,
+      "loss": 0.8623,
+      "step": 3856
+    },
+    {
+      "epoch": 0.06838697669471809,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.9471429587583985e-05,
+      "loss": 0.8581,
+      "step": 3858
+    },
+    {
+      "epoch": 0.06842242873032967,
+      "grad_norm": 2.625,
+      "learning_rate": 4.947085817684848e-05,
+      "loss": 0.85,
+      "step": 3860
+    },
+    {
+      "epoch": 0.06845788076594123,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.947028646072175e-05,
+      "loss": 0.8351,
+      "step": 3862
+    },
+    {
+      "epoch": 0.0684933328015528,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.9469714439210954e-05,
+      "loss": 0.8576,
+      "step": 3864
+    },
+    {
+      "epoch": 0.06852878483716437,
+      "grad_norm": 2.625,
+      "learning_rate": 4.946914211232321e-05,
+      "loss": 0.8174,
+      "step": 3866
+    },
+    {
+      "epoch": 0.06856423687277594,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.946856948006567e-05,
+      "loss": 0.8789,
+      "step": 3868
+    },
+    {
+      "epoch": 0.06859968890838751,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.946799654244548e-05,
+      "loss": 0.8533,
+      "step": 3870
+    },
+    {
+      "epoch": 0.06863514094399908,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.9467423299469796e-05,
+      "loss": 0.8777,
+      "step": 3872
+    },
+    {
+      "epoch": 0.06867059297961065,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.9466849751145754e-05,
+      "loss": 0.8299,
+      "step": 3874
+    },
+    {
+      "epoch": 0.06870604501522222,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.946627589748053e-05,
+      "loss": 0.8746,
+      "step": 3876
+    },
+    {
+      "epoch": 0.0687414970508338,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.9465701738481276e-05,
+      "loss": 0.8488,
+      "step": 3878
+    },
+    {
+      "epoch": 0.06877694908644535,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.9465127274155165e-05,
+      "loss": 0.8468,
+      "step": 3880
+    },
+    {
+      "epoch": 0.06881240112205693,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.9464552504509353e-05,
+      "loss": 0.8292,
+      "step": 3882
+    },
+    {
+      "epoch": 0.0688478531576685,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.946397742955103e-05,
+      "loss": 0.8555,
+      "step": 3884
+    },
+    {
+      "epoch": 0.06888330519328006,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.946340204928736e-05,
+      "loss": 0.857,
+      "step": 3886
+    },
+    {
+      "epoch": 0.06891875722889164,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.946282636372553e-05,
+      "loss": 0.8918,
+      "step": 3888
+    },
+    {
+      "epoch": 0.0689542092645032,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.946225037287272e-05,
+      "loss": 0.8305,
+      "step": 3890
+    },
+    {
+      "epoch": 0.06898966130011477,
+      "grad_norm": 2.875,
+      "learning_rate": 4.946167407673612e-05,
+      "loss": 0.9015,
+      "step": 3892
+    },
+    {
+      "epoch": 0.06902511333572635,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.9461097475322925e-05,
+      "loss": 0.8525,
+      "step": 3894
+    },
+    {
+      "epoch": 0.06906056537133791,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.946052056864032e-05,
+      "loss": 0.8384,
+      "step": 3896
+    },
+    {
+      "epoch": 0.06909601740694948,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.945994335669552e-05,
+      "loss": 0.7948,
+      "step": 3898
+    },
+    {
+      "epoch": 0.06913146944256106,
+      "grad_norm": 3.25,
+      "learning_rate": 4.945936583949573e-05,
+      "loss": 0.8752,
+      "step": 3900
+    },
+    {
+      "epoch": 0.06916692147817262,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.945878801704814e-05,
+      "loss": 0.8423,
+      "step": 3902
+    },
+    {
+      "epoch": 0.0692023735137842,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.945820988935997e-05,
+      "loss": 0.7475,
+      "step": 3904
+    },
+    {
+      "epoch": 0.06923782554939577,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.945763145643844e-05,
+      "loss": 0.8427,
+      "step": 3906
+    },
+    {
+      "epoch": 0.06927327758500733,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.9457052718290756e-05,
+      "loss": 0.8373,
+      "step": 3908
+    },
+    {
+      "epoch": 0.0693087296206189,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.945647367492415e-05,
+      "loss": 0.8304,
+      "step": 3910
+    },
+    {
+      "epoch": 0.06934418165623048,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.945589432634584e-05,
+      "loss": 0.8281,
+      "step": 3912
+    },
+    {
+      "epoch": 0.06937963369184204,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.945531467256307e-05,
+      "loss": 0.8817,
+      "step": 3914
+    },
+    {
+      "epoch": 0.06941508572745361,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.9454734713583075e-05,
+      "loss": 0.8514,
+      "step": 3916
+    },
+    {
+      "epoch": 0.06945053776306519,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.945415444941307e-05,
+      "loss": 0.8647,
+      "step": 3918
+    },
+    {
+      "epoch": 0.06948598979867675,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.9453573880060324e-05,
+      "loss": 0.8591,
+      "step": 3920
+    },
+    {
+      "epoch": 0.06952144183428832,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.945299300553206e-05,
+      "loss": 0.8554,
+      "step": 3922
+    },
+    {
+      "epoch": 0.0695568938698999,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.945241182583554e-05,
+      "loss": 0.8347,
+      "step": 3924
+    },
+    {
+      "epoch": 0.06959234590551146,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.9451830340978014e-05,
+      "loss": 0.8306,
+      "step": 3926
+    },
+    {
+      "epoch": 0.06962779794112303,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.945124855096673e-05,
+      "loss": 0.8073,
+      "step": 3928
+    },
+    {
+      "epoch": 0.0696632499767346,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.9450666455808965e-05,
+      "loss": 0.8574,
+      "step": 3930
+    },
+    {
+      "epoch": 0.06969870201234617,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.945008405551197e-05,
+      "loss": 0.8346,
+      "step": 3932
+    },
+    {
+      "epoch": 0.06973415404795774,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.9449501350083024e-05,
+      "loss": 0.8432,
+      "step": 3934
+    },
+    {
+      "epoch": 0.06976960608356932,
+      "grad_norm": 2.875,
+      "learning_rate": 4.944891833952939e-05,
+      "loss": 0.8358,
+      "step": 3936
+    },
+    {
+      "epoch": 0.06980505811918088,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.944833502385835e-05,
+      "loss": 0.8209,
+      "step": 3938
+    },
+    {
+      "epoch": 0.06984051015479245,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.944775140307718e-05,
+      "loss": 0.8509,
+      "step": 3940
+    },
+    {
+      "epoch": 0.06987596219040403,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.944716747719317e-05,
+      "loss": 0.8755,
+      "step": 3942
+    },
+    {
+      "epoch": 0.06991141422601559,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.9446583246213594e-05,
+      "loss": 0.8793,
+      "step": 3944
+    },
+    {
+      "epoch": 0.06994686626162716,
+      "grad_norm": 2.625,
+      "learning_rate": 4.944599871014576e-05,
+      "loss": 0.8635,
+      "step": 3946
+    },
+    {
+      "epoch": 0.06998231829723874,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.944541386899694e-05,
+      "loss": 0.8418,
+      "step": 3948
+    },
+    {
+      "epoch": 0.0700177703328503,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.9444828722774455e-05,
+      "loss": 0.8428,
+      "step": 3950
+    },
+    {
+      "epoch": 0.07005322236846187,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.94442432714856e-05,
+      "loss": 0.9022,
+      "step": 3952
+    },
+    {
+      "epoch": 0.07008867440407344,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.9443657515137674e-05,
+      "loss": 0.836,
+      "step": 3954
+    },
+    {
+      "epoch": 0.070124126439685,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.9443071453738e-05,
+      "loss": 0.8242,
+      "step": 3956
+    },
+    {
+      "epoch": 0.07015957847529658,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.9442485087293886e-05,
+      "loss": 0.8566,
+      "step": 3958
+    },
+    {
+      "epoch": 0.07019503051090815,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.944189841581265e-05,
+      "loss": 0.8319,
+      "step": 3960
+    },
+    {
+      "epoch": 0.07023048254651972,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.944131143930161e-05,
+      "loss": 0.8564,
+      "step": 3962
+    },
+    {
+      "epoch": 0.07026593458213129,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.944072415776809e-05,
+      "loss": 0.862,
+      "step": 3964
+    },
+    {
+      "epoch": 0.07030138661774286,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.944013657121942e-05,
+      "loss": 0.8417,
+      "step": 3966
+    },
+    {
+      "epoch": 0.07033683865335442,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.943954867966295e-05,
+      "loss": 0.845,
+      "step": 3968
+    },
+    {
+      "epoch": 0.070372290688966,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.943896048310599e-05,
+      "loss": 0.8121,
+      "step": 3970
+    },
+    {
+      "epoch": 0.07040774272457756,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.94383719815559e-05,
+      "loss": 0.8761,
+      "step": 3972
+    },
+    {
+      "epoch": 0.07044319476018913,
+      "grad_norm": 3.421875,
+      "learning_rate": 4.9437783175020015e-05,
+      "loss": 0.8621,
+      "step": 3974
+    },
+    {
+      "epoch": 0.07047864679580071,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.943719406350569e-05,
+      "loss": 0.8443,
+      "step": 3976
+    },
+    {
+      "epoch": 0.07051409883141227,
+      "grad_norm": 2.75,
+      "learning_rate": 4.943660464702027e-05,
+      "loss": 0.8103,
+      "step": 3978
+    },
+    {
+      "epoch": 0.07054955086702384,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.943601492557112e-05,
+      "loss": 0.8744,
+      "step": 3980
+    },
+    {
+      "epoch": 0.07058500290263542,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.9435424899165586e-05,
+      "loss": 0.8371,
+      "step": 3982
+    },
+    {
+      "epoch": 0.07062045493824698,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.943483456781104e-05,
+      "loss": 0.8318,
+      "step": 3984
+    },
+    {
+      "epoch": 0.07065590697385855,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.943424393151485e-05,
+      "loss": 0.8621,
+      "step": 3986
+    },
+    {
+      "epoch": 0.07069135900947013,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.9433652990284375e-05,
+      "loss": 0.8897,
+      "step": 3988
+    },
+    {
+      "epoch": 0.07072681104508169,
+      "grad_norm": 2.875,
+      "learning_rate": 4.943306174412701e-05,
+      "loss": 0.8527,
+      "step": 3990
+    },
+    {
+      "epoch": 0.07076226308069326,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.943247019305012e-05,
+      "loss": 0.831,
+      "step": 3992
+    },
+    {
+      "epoch": 0.07079771511630484,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.943187833706109e-05,
+      "loss": 0.8465,
+      "step": 3994
+    },
+    {
+      "epoch": 0.0708331671519164,
+      "grad_norm": 2.875,
+      "learning_rate": 4.943128617616731e-05,
+      "loss": 0.865,
+      "step": 3996
+    },
+    {
+      "epoch": 0.07086861918752797,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.943069371037618e-05,
+      "loss": 0.8672,
+      "step": 3998
+    },
+    {
+      "epoch": 0.07090407122313955,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.943010093969506e-05,
+      "loss": 0.833,
+      "step": 4000
+    },
+    {
+      "epoch": 0.07093952325875111,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.9429507864131375e-05,
+      "loss": 0.8364,
+      "step": 4002
+    },
+    {
+      "epoch": 0.07097497529436268,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.942891448369252e-05,
+      "loss": 0.8505,
+      "step": 4004
+    },
+    {
+      "epoch": 0.07101042732997426,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.942832079838591e-05,
+      "loss": 0.8213,
+      "step": 4006
+    },
+    {
+      "epoch": 0.07104587936558582,
+      "grad_norm": 3.0,
+      "learning_rate": 4.9427726808218935e-05,
+      "loss": 0.8397,
+      "step": 4008
+    },
+    {
+      "epoch": 0.07108133140119739,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.9427132513199015e-05,
+      "loss": 0.8628,
+      "step": 4010
+    },
+    {
+      "epoch": 0.07111678343680897,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.942653791333357e-05,
+      "loss": 0.8477,
+      "step": 4012
+    },
+    {
+      "epoch": 0.07115223547242053,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.942594300863003e-05,
+      "loss": 0.9068,
+      "step": 4014
+    },
+    {
+      "epoch": 0.0711876875080321,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.94253477990958e-05,
+      "loss": 0.8385,
+      "step": 4016
+    },
+    {
+      "epoch": 0.07122313954364368,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.942475228473832e-05,
+      "loss": 0.8585,
+      "step": 4018
+    },
+    {
+      "epoch": 0.07125859157925524,
+      "grad_norm": 2.875,
+      "learning_rate": 4.942415646556501e-05,
+      "loss": 0.8674,
+      "step": 4020
+    },
+    {
+      "epoch": 0.07129404361486681,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.9423560341583325e-05,
+      "loss": 0.841,
+      "step": 4022
+    },
+    {
+      "epoch": 0.07132949565047839,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.942296391280069e-05,
+      "loss": 0.8507,
+      "step": 4024
+    },
+    {
+      "epoch": 0.07136494768608995,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.9422367179224555e-05,
+      "loss": 0.8421,
+      "step": 4026
+    },
+    {
+      "epoch": 0.07140039972170152,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.942177014086236e-05,
+      "loss": 0.8639,
+      "step": 4028
+    },
+    {
+      "epoch": 0.0714358517573131,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.9421172797721566e-05,
+      "loss": 0.8645,
+      "step": 4030
+    },
+    {
+      "epoch": 0.07147130379292466,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.942057514980962e-05,
+      "loss": 0.8383,
+      "step": 4032
+    },
+    {
+      "epoch": 0.07150675582853623,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.9419977197133984e-05,
+      "loss": 0.8567,
+      "step": 4034
+    },
+    {
+      "epoch": 0.0715422078641478,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.941937893970211e-05,
+      "loss": 0.8575,
+      "step": 4036
+    },
+    {
+      "epoch": 0.07157765989975937,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.941878037752148e-05,
+      "loss": 0.8374,
+      "step": 4038
+    },
+    {
+      "epoch": 0.07161311193537094,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.941818151059956e-05,
+      "loss": 0.8745,
+      "step": 4040
+    },
+    {
+      "epoch": 0.07164856397098252,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.9417582338943815e-05,
+      "loss": 0.8833,
+      "step": 4042
+    },
+    {
+      "epoch": 0.07168401600659408,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.9416982862561726e-05,
+      "loss": 0.8612,
+      "step": 4044
+    },
+    {
+      "epoch": 0.07171946804220565,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.941638308146078e-05,
+      "loss": 0.819,
+      "step": 4046
+    },
+    {
+      "epoch": 0.07175492007781722,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.941578299564846e-05,
+      "loss": 0.837,
+      "step": 4048
+    },
+    {
+      "epoch": 0.07179037211342879,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.9415182605132255e-05,
+      "loss": 0.895,
+      "step": 4050
+    },
+    {
+      "epoch": 0.07182582414904036,
+      "grad_norm": 2.625,
+      "learning_rate": 4.9414581909919656e-05,
+      "loss": 0.8425,
+      "step": 4052
+    },
+    {
+      "epoch": 0.07186127618465193,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.941398091001815e-05,
+      "loss": 0.8123,
+      "step": 4054
+    },
+    {
+      "epoch": 0.0718967282202635,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.9413379605435264e-05,
+      "loss": 0.8288,
+      "step": 4056
+    },
+    {
+      "epoch": 0.07193218025587507,
+      "grad_norm": 2.625,
+      "learning_rate": 4.9412777996178474e-05,
+      "loss": 0.8381,
+      "step": 4058
+    },
+    {
+      "epoch": 0.07196763229148663,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.9412176082255304e-05,
+      "loss": 0.9139,
+      "step": 4060
+    },
+    {
+      "epoch": 0.0720030843270982,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.941157386367326e-05,
+      "loss": 0.8468,
+      "step": 4062
+    },
+    {
+      "epoch": 0.07203853636270978,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.941097134043986e-05,
+      "loss": 0.8892,
+      "step": 4064
+    },
+    {
+      "epoch": 0.07207398839832134,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.9410368512562624e-05,
+      "loss": 0.8498,
+      "step": 4066
+    },
+    {
+      "epoch": 0.07210944043393291,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.940976538004907e-05,
+      "loss": 0.8358,
+      "step": 4068
+    },
+    {
+      "epoch": 0.07214489246954449,
+      "grad_norm": 2.875,
+      "learning_rate": 4.9409161942906724e-05,
+      "loss": 0.8611,
+      "step": 4070
+    },
+    {
+      "epoch": 0.07218034450515605,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.940855820114312e-05,
+      "loss": 0.8264,
+      "step": 4072
+    },
+    {
+      "epoch": 0.07221579654076762,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.94079541547658e-05,
+      "loss": 0.9031,
+      "step": 4074
+    },
+    {
+      "epoch": 0.0722512485763792,
+      "grad_norm": 2.5,
+      "learning_rate": 4.94073498037823e-05,
+      "loss": 0.7923,
+      "step": 4076
+    },
+    {
+      "epoch": 0.07228670061199076,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.940674514820015e-05,
+      "loss": 0.8308,
+      "step": 4078
+    },
+    {
+      "epoch": 0.07232215264760233,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.9406140188026905e-05,
+      "loss": 0.8739,
+      "step": 4080
+    },
+    {
+      "epoch": 0.07235760468321391,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.940553492327012e-05,
+      "loss": 0.8497,
+      "step": 4082
+    },
+    {
+      "epoch": 0.07239305671882547,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.9404929353937336e-05,
+      "loss": 0.8896,
+      "step": 4084
+    },
+    {
+      "epoch": 0.07242850875443704,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.9404323480036116e-05,
+      "loss": 0.8515,
+      "step": 4086
+    },
+    {
+      "epoch": 0.07246396079004862,
+      "grad_norm": 2.625,
+      "learning_rate": 4.940371730157403e-05,
+      "loss": 0.831,
+      "step": 4088
+    },
+    {
+      "epoch": 0.07249941282566018,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.940311081855863e-05,
+      "loss": 0.8463,
+      "step": 4090
+    },
+    {
+      "epoch": 0.07253486486127175,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.940250403099749e-05,
+      "loss": 0.8772,
+      "step": 4092
+    },
+    {
+      "epoch": 0.07257031689688333,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.9401896938898185e-05,
+      "loss": 0.837,
+      "step": 4094
+    },
+    {
+      "epoch": 0.07260576893249489,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.940128954226828e-05,
+      "loss": 0.863,
+      "step": 4096
+    },
+    {
+      "epoch": 0.07264122096810646,
+      "grad_norm": 2.75,
+      "learning_rate": 4.940068184111537e-05,
+      "loss": 0.8407,
+      "step": 4098
+    },
+    {
+      "epoch": 0.07267667300371804,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.9400073835447035e-05,
+      "loss": 0.8389,
+      "step": 4100
+    },
+    {
+      "epoch": 0.0727121250393296,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.939946552527086e-05,
+      "loss": 0.8972,
+      "step": 4102
+    },
+    {
+      "epoch": 0.07274757707494117,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.939885691059444e-05,
+      "loss": 0.8153,
+      "step": 4104
+    },
+    {
+      "epoch": 0.07278302911055275,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.939824799142536e-05,
+      "loss": 0.8642,
+      "step": 4106
+    },
+    {
+      "epoch": 0.07281848114616431,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.939763876777122e-05,
+      "loss": 0.8415,
+      "step": 4108
+    },
+    {
+      "epoch": 0.07285393318177588,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.939702923963965e-05,
+      "loss": 0.819,
+      "step": 4110
+    },
+    {
+      "epoch": 0.07288938521738746,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.9396419407038226e-05,
+      "loss": 0.863,
+      "step": 4112
+    },
+    {
+      "epoch": 0.07292483725299902,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.939580926997457e-05,
+      "loss": 0.8575,
+      "step": 4114
+    },
+    {
+      "epoch": 0.07296028928861059,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.9395198828456294e-05,
+      "loss": 0.8643,
+      "step": 4116
+    },
+    {
+      "epoch": 0.07299574132422217,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.9394588082491024e-05,
+      "loss": 0.8993,
+      "step": 4118
+    },
+    {
+      "epoch": 0.07303119335983373,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.939397703208637e-05,
+      "loss": 0.8818,
+      "step": 4120
+    },
+    {
+      "epoch": 0.0730666453954453,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.939336567724996e-05,
+      "loss": 0.8545,
+      "step": 4122
+    },
+    {
+      "epoch": 0.07310209743105688,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.9392754017989435e-05,
+      "loss": 0.8493,
+      "step": 4124
+    },
+    {
+      "epoch": 0.07313754946666844,
+      "grad_norm": 2.625,
+      "learning_rate": 4.9392142054312416e-05,
+      "loss": 0.83,
+      "step": 4126
+    },
+    {
+      "epoch": 0.07317300150228001,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.939152978622655e-05,
+      "loss": 0.9047,
+      "step": 4128
+    },
+    {
+      "epoch": 0.07320845353789159,
+      "grad_norm": 2.375,
+      "learning_rate": 4.939091721373946e-05,
+      "loss": 0.8372,
+      "step": 4130
+    },
+    {
+      "epoch": 0.07324390557350315,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.9390304336858814e-05,
+      "loss": 0.8822,
+      "step": 4132
+    },
+    {
+      "epoch": 0.07327935760911472,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.9389691155592256e-05,
+      "loss": 0.7915,
+      "step": 4134
+    },
+    {
+      "epoch": 0.0733148096447263,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.938907766994742e-05,
+      "loss": 0.8189,
+      "step": 4136
+    },
+    {
+      "epoch": 0.07335026168033786,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.938846387993198e-05,
+      "loss": 0.8671,
+      "step": 4138
+    },
+    {
+      "epoch": 0.07338571371594943,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.9387849785553584e-05,
+      "loss": 0.8696,
+      "step": 4140
+    },
+    {
+      "epoch": 0.07342116575156099,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.938723538681991e-05,
+      "loss": 0.7995,
+      "step": 4142
+    },
+    {
+      "epoch": 0.07345661778717257,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.9386620683738616e-05,
+      "loss": 0.8268,
+      "step": 4144
+    },
+    {
+      "epoch": 0.07349206982278414,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.9386005676317385e-05,
+      "loss": 0.7975,
+      "step": 4146
+    },
+    {
+      "epoch": 0.0735275218583957,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.9385390364563864e-05,
+      "loss": 0.8414,
+      "step": 4148
+    },
+    {
+      "epoch": 0.07356297389400727,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.9384774748485764e-05,
+      "loss": 0.8708,
+      "step": 4150
+    },
+    {
+      "epoch": 0.07359842592961885,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.938415882809074e-05,
+      "loss": 0.9021,
+      "step": 4152
+    },
+    {
+      "epoch": 0.07363387796523041,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.938354260338651e-05,
+      "loss": 0.8506,
+      "step": 4154
+    },
+    {
+      "epoch": 0.07366933000084198,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.938292607438074e-05,
+      "loss": 0.8132,
+      "step": 4156
+    },
+    {
+      "epoch": 0.07370478203645356,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.938230924108113e-05,
+      "loss": 0.8767,
+      "step": 4158
+    },
+    {
+      "epoch": 0.07374023407206512,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.938169210349538e-05,
+      "loss": 0.8324,
+      "step": 4160
+    },
+    {
+      "epoch": 0.0737756861076767,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.938107466163119e-05,
+      "loss": 0.8516,
+      "step": 4162
+    },
+    {
+      "epoch": 0.07381113814328827,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.938045691549626e-05,
+      "loss": 0.8218,
+      "step": 4164
+    },
+    {
+      "epoch": 0.07384659017889983,
+      "grad_norm": 2.875,
+      "learning_rate": 4.937983886509832e-05,
+      "loss": 0.8524,
+      "step": 4166
+    },
+    {
+      "epoch": 0.0738820422145114,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.937922051044506e-05,
+      "loss": 0.8436,
+      "step": 4168
+    },
+    {
+      "epoch": 0.07391749425012298,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.93786018515442e-05,
+      "loss": 0.8616,
+      "step": 4170
+    },
+    {
+      "epoch": 0.07395294628573454,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.937798288840347e-05,
+      "loss": 0.8628,
+      "step": 4172
+    },
+    {
+      "epoch": 0.07398839832134611,
+      "grad_norm": 2.875,
+      "learning_rate": 4.9377363621030596e-05,
+      "loss": 0.8646,
+      "step": 4174
+    },
+    {
+      "epoch": 0.07402385035695769,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.93767440494333e-05,
+      "loss": 0.8985,
+      "step": 4176
+    },
+    {
+      "epoch": 0.07405930239256925,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.937612417361932e-05,
+      "loss": 0.8776,
+      "step": 4178
+    },
+    {
+      "epoch": 0.07409475442818082,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.937550399359638e-05,
+      "loss": 0.8236,
+      "step": 4180
+    },
+    {
+      "epoch": 0.0741302064637924,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.937488350937223e-05,
+      "loss": 0.8464,
+      "step": 4182
+    },
+    {
+      "epoch": 0.07416565849940396,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.937426272095461e-05,
+      "loss": 0.8683,
+      "step": 4184
+    },
+    {
+      "epoch": 0.07420111053501553,
+      "grad_norm": 2.875,
+      "learning_rate": 4.937364162835127e-05,
+      "loss": 0.8014,
+      "step": 4186
+    },
+    {
+      "epoch": 0.07423656257062711,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.9373020231569956e-05,
+      "loss": 0.8531,
+      "step": 4188
+    },
+    {
+      "epoch": 0.07427201460623867,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.937239853061843e-05,
+      "loss": 0.8679,
+      "step": 4190
+    },
+    {
+      "epoch": 0.07430746664185024,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.9371776525504446e-05,
+      "loss": 0.8294,
+      "step": 4192
+    },
+    {
+      "epoch": 0.07434291867746182,
+      "grad_norm": 2.75,
+      "learning_rate": 4.937115421623577e-05,
+      "loss": 0.8347,
+      "step": 4194
+    },
+    {
+      "epoch": 0.07437837071307338,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.937053160282016e-05,
+      "loss": 0.843,
+      "step": 4196
+    },
+    {
+      "epoch": 0.07441382274868495,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.936990868526539e-05,
+      "loss": 0.8282,
+      "step": 4198
+    },
+    {
+      "epoch": 0.07444927478429653,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.936928546357924e-05,
+      "loss": 0.8278,
+      "step": 4200
+    },
+    {
+      "epoch": 0.07448472681990809,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.9368661937769475e-05,
+      "loss": 0.8537,
+      "step": 4202
+    },
+    {
+      "epoch": 0.07452017885551966,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.936803810784389e-05,
+      "loss": 0.8322,
+      "step": 4204
+    },
+    {
+      "epoch": 0.07455563089113124,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.936741397381027e-05,
+      "loss": 0.8855,
+      "step": 4206
+    },
+    {
+      "epoch": 0.0745910829267428,
+      "grad_norm": 3.0,
+      "learning_rate": 4.93667895356764e-05,
+      "loss": 0.851,
+      "step": 4208
+    },
+    {
+      "epoch": 0.07462653496235437,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.9366164793450066e-05,
+      "loss": 0.8417,
+      "step": 4210
+    },
+    {
+      "epoch": 0.07466198699796595,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.936553974713907e-05,
+      "loss": 0.8607,
+      "step": 4212
+    },
+    {
+      "epoch": 0.0746974390335775,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.936491439675122e-05,
+      "loss": 0.816,
+      "step": 4214
+    },
+    {
+      "epoch": 0.07473289106918908,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.9364288742294306e-05,
+      "loss": 0.8346,
+      "step": 4216
+    },
+    {
+      "epoch": 0.07476834310480066,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.9363662783776146e-05,
+      "loss": 0.8284,
+      "step": 4218
+    },
+    {
+      "epoch": 0.07480379514041222,
+      "grad_norm": 2.75,
+      "learning_rate": 4.9363036521204546e-05,
+      "loss": 0.8666,
+      "step": 4220
+    },
+    {
+      "epoch": 0.07483924717602379,
+      "grad_norm": 2.5,
+      "learning_rate": 4.936240995458733e-05,
+      "loss": 0.8194,
+      "step": 4222
+    },
+    {
+      "epoch": 0.07487469921163536,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.9361783083932304e-05,
+      "loss": 0.8879,
+      "step": 4224
+    },
+    {
+      "epoch": 0.07491015124724693,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.93611559092473e-05,
+      "loss": 0.8462,
+      "step": 4226
+    },
+    {
+      "epoch": 0.0749456032828585,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.936052843054015e-05,
+      "loss": 0.8267,
+      "step": 4228
+    },
+    {
+      "epoch": 0.07498105531847006,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.935990064781868e-05,
+      "loss": 0.897,
+      "step": 4230
+    },
+    {
+      "epoch": 0.07501650735408164,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.935927256109072e-05,
+      "loss": 0.8587,
+      "step": 4232
+    },
+    {
+      "epoch": 0.07505195938969321,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.935864417036412e-05,
+      "loss": 0.8527,
+      "step": 4234
+    },
+    {
+      "epoch": 0.07508741142530477,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.93580154756467e-05,
+      "loss": 0.8459,
+      "step": 4236
+    },
+    {
+      "epoch": 0.07512286346091634,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.9357386476946334e-05,
+      "loss": 0.8307,
+      "step": 4238
+    },
+    {
+      "epoch": 0.07515831549652792,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.935675717427085e-05,
+      "loss": 0.8156,
+      "step": 4240
+    },
+    {
+      "epoch": 0.07519376753213948,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.935612756762811e-05,
+      "loss": 0.8588,
+      "step": 4242
+    },
+    {
+      "epoch": 0.07522921956775105,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.935549765702597e-05,
+      "loss": 0.8664,
+      "step": 4244
+    },
+    {
+      "epoch": 0.07526467160336263,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.93548674424723e-05,
+      "loss": 0.8555,
+      "step": 4246
+    },
+    {
+      "epoch": 0.07530012363897419,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.935423692397495e-05,
+      "loss": 0.8634,
+      "step": 4248
+    },
+    {
+      "epoch": 0.07533557567458576,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.93536061015418e-05,
+      "loss": 0.835,
+      "step": 4250
+    },
+    {
+      "epoch": 0.07537102771019734,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.935297497518071e-05,
+      "loss": 0.8366,
+      "step": 4252
+    },
+    {
+      "epoch": 0.0754064797458089,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.935234354489958e-05,
+      "loss": 0.7994,
+      "step": 4254
+    },
+    {
+      "epoch": 0.07544193178142047,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.935171181070626e-05,
+      "loss": 0.8496,
+      "step": 4256
+    },
+    {
+      "epoch": 0.07547738381703205,
+      "grad_norm": 3.0,
+      "learning_rate": 4.935107977260865e-05,
+      "loss": 0.8336,
+      "step": 4258
+    },
+    {
+      "epoch": 0.07551283585264361,
+      "grad_norm": 2.75,
+      "learning_rate": 4.9350447430614647e-05,
+      "loss": 0.8468,
+      "step": 4260
+    },
+    {
+      "epoch": 0.07554828788825518,
+      "grad_norm": 3.125,
+      "learning_rate": 4.934981478473213e-05,
+      "loss": 0.8597,
+      "step": 4262
+    },
+    {
+      "epoch": 0.07558373992386676,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.934918183496898e-05,
+      "loss": 0.82,
+      "step": 4264
+    },
+    {
+      "epoch": 0.07561919195947832,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.934854858133313e-05,
+      "loss": 0.8744,
+      "step": 4266
+    },
+    {
+      "epoch": 0.07565464399508989,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.934791502383246e-05,
+      "loss": 0.8545,
+      "step": 4268
+    },
+    {
+      "epoch": 0.07569009603070147,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.934728116247488e-05,
+      "loss": 0.8635,
+      "step": 4270
+    },
+    {
+      "epoch": 0.07572554806631303,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.93466469972683e-05,
+      "loss": 0.8592,
+      "step": 4272
+    },
+    {
+      "epoch": 0.0757610001019246,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.934601252822064e-05,
+      "loss": 0.8474,
+      "step": 4274
+    },
+    {
+      "epoch": 0.07579645213753618,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.9345377755339815e-05,
+      "loss": 0.9021,
+      "step": 4276
+    },
+    {
+      "epoch": 0.07583190417314774,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.934474267863375e-05,
+      "loss": 0.8184,
+      "step": 4278
+    },
+    {
+      "epoch": 0.07586735620875931,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.934410729811036e-05,
+      "loss": 0.8187,
+      "step": 4280
+    },
+    {
+      "epoch": 0.07590280824437089,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.9343471613777584e-05,
+      "loss": 0.8448,
+      "step": 4282
+    },
+    {
+      "epoch": 0.07593826027998245,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.934283562564335e-05,
+      "loss": 0.8228,
+      "step": 4284
+    },
+    {
+      "epoch": 0.07597371231559402,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.93421993337156e-05,
+      "loss": 0.8537,
+      "step": 4286
+    },
+    {
+      "epoch": 0.0760091643512056,
+      "grad_norm": 2.625,
+      "learning_rate": 4.934156273800228e-05,
+      "loss": 0.8121,
+      "step": 4288
+    },
+    {
+      "epoch": 0.07604461638681716,
+      "grad_norm": 2.625,
+      "learning_rate": 4.934092583851132e-05,
+      "loss": 0.8435,
+      "step": 4290
+    },
+    {
+      "epoch": 0.07608006842242873,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.934028863525067e-05,
+      "loss": 0.7778,
+      "step": 4292
+    },
+    {
+      "epoch": 0.0761155204580403,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.933965112822829e-05,
+      "loss": 0.8711,
+      "step": 4294
+    },
+    {
+      "epoch": 0.07615097249365187,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.9339013317452145e-05,
+      "loss": 0.8524,
+      "step": 4296
+    },
+    {
+      "epoch": 0.07618642452926344,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.933837520293017e-05,
+      "loss": 0.842,
+      "step": 4298
+    },
+    {
+      "epoch": 0.07622187656487502,
+      "grad_norm": 2.75,
+      "learning_rate": 4.933773678467035e-05,
+      "loss": 0.8656,
+      "step": 4300
+    },
+    {
+      "epoch": 0.07625732860048658,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.9337098062680635e-05,
+      "loss": 0.8061,
+      "step": 4302
+    },
+    {
+      "epoch": 0.07629278063609815,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.933645903696901e-05,
+      "loss": 0.8413,
+      "step": 4304
+    },
+    {
+      "epoch": 0.07632823267170973,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.933581970754345e-05,
+      "loss": 0.8499,
+      "step": 4306
+    },
+    {
+      "epoch": 0.07636368470732129,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.9335180074411926e-05,
+      "loss": 0.8621,
+      "step": 4308
+    },
+    {
+      "epoch": 0.07639913674293286,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.933454013758242e-05,
+      "loss": 0.8514,
+      "step": 4310
+    },
+    {
+      "epoch": 0.07643458877854442,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.933389989706292e-05,
+      "loss": 0.8217,
+      "step": 4312
+    },
+    {
+      "epoch": 0.076470040814156,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.933325935286142e-05,
+      "loss": 0.8471,
+      "step": 4314
+    },
+    {
+      "epoch": 0.07650549284976757,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.933261850498592e-05,
+      "loss": 0.8507,
+      "step": 4316
+    },
+    {
+      "epoch": 0.07654094488537913,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.93319773534444e-05,
+      "loss": 0.873,
+      "step": 4318
+    },
+    {
+      "epoch": 0.0765763969209907,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.9331335898244866e-05,
+      "loss": 0.8346,
+      "step": 4320
+    },
+    {
+      "epoch": 0.07661184895660228,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.933069413939534e-05,
+      "loss": 0.8501,
+      "step": 4322
+    },
+    {
+      "epoch": 0.07664730099221384,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.933005207690381e-05,
+      "loss": 0.8429,
+      "step": 4324
+    },
+    {
+      "epoch": 0.07668275302782541,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.93294097107783e-05,
+      "loss": 0.811,
+      "step": 4326
+    },
+    {
+      "epoch": 0.07671820506343699,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.9328767041026824e-05,
+      "loss": 0.8589,
+      "step": 4328
+    },
+    {
+      "epoch": 0.07675365709904855,
+      "grad_norm": 2.875,
+      "learning_rate": 4.9328124067657406e-05,
+      "loss": 0.8443,
+      "step": 4330
+    },
+    {
+      "epoch": 0.07678910913466012,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.932748079067806e-05,
+      "loss": 0.8548,
+      "step": 4332
+    },
+    {
+      "epoch": 0.0768245611702717,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.932683721009683e-05,
+      "loss": 0.8519,
+      "step": 4334
+    },
+    {
+      "epoch": 0.07686001320588326,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.9326193325921734e-05,
+      "loss": 0.8667,
+      "step": 4336
+    },
+    {
+      "epoch": 0.07689546524149483,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.932554913816081e-05,
+      "loss": 0.815,
+      "step": 4338
+    },
+    {
+      "epoch": 0.07693091727710641,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.9324904646822104e-05,
+      "loss": 0.8384,
+      "step": 4340
+    },
+    {
+      "epoch": 0.07696636931271797,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.932425985191365e-05,
+      "loss": 0.8983,
+      "step": 4342
+    },
+    {
+      "epoch": 0.07700182134832954,
+      "grad_norm": 2.5,
+      "learning_rate": 4.9323614753443506e-05,
+      "loss": 0.8064,
+      "step": 4344
+    },
+    {
+      "epoch": 0.07703727338394112,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.932296935141971e-05,
+      "loss": 0.8356,
+      "step": 4346
+    },
+    {
+      "epoch": 0.07707272541955268,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.932232364585032e-05,
+      "loss": 0.8279,
+      "step": 4348
+    },
+    {
+      "epoch": 0.07710817745516425,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.93216776367434e-05,
+      "loss": 0.8228,
+      "step": 4350
+    },
+    {
+      "epoch": 0.07714362949077583,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.9321031324107016e-05,
+      "loss": 0.8484,
+      "step": 4352
+    },
+    {
+      "epoch": 0.07717908152638739,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.932038470794922e-05,
+      "loss": 0.833,
+      "step": 4354
+    },
+    {
+      "epoch": 0.07721453356199896,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.931973778827809e-05,
+      "loss": 0.8317,
+      "step": 4356
+    },
+    {
+      "epoch": 0.07724998559761054,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.931909056510169e-05,
+      "loss": 0.8346,
+      "step": 4358
+    },
+    {
+      "epoch": 0.0772854376332221,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.931844303842811e-05,
+      "loss": 0.8479,
+      "step": 4360
+    },
+    {
+      "epoch": 0.07732088966883367,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.931779520826543e-05,
+      "loss": 0.8641,
+      "step": 4362
+    },
+    {
+      "epoch": 0.07735634170444525,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.931714707462173e-05,
+      "loss": 0.8309,
+      "step": 4364
+    },
+    {
+      "epoch": 0.07739179374005681,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.93164986375051e-05,
+      "loss": 0.9095,
+      "step": 4366
+    },
+    {
+      "epoch": 0.07742724577566838,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.931584989692363e-05,
+      "loss": 0.8253,
+      "step": 4368
+    },
+    {
+      "epoch": 0.07746269781127996,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.9315200852885415e-05,
+      "loss": 0.8257,
+      "step": 4370
+    },
+    {
+      "epoch": 0.07749814984689152,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.931455150539856e-05,
+      "loss": 0.8824,
+      "step": 4372
+    },
+    {
+      "epoch": 0.07753360188250309,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.931390185447117e-05,
+      "loss": 0.845,
+      "step": 4374
+    },
+    {
+      "epoch": 0.07756905391811467,
+      "grad_norm": 3.0,
+      "learning_rate": 4.9313251900111346e-05,
+      "loss": 0.8355,
+      "step": 4376
+    },
+    {
+      "epoch": 0.07760450595372623,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.93126016423272e-05,
+      "loss": 0.8556,
+      "step": 4378
+    },
+    {
+      "epoch": 0.0776399579893378,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.931195108112685e-05,
+      "loss": 0.8141,
+      "step": 4380
+    },
+    {
+      "epoch": 0.07767541002494938,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.9311300216518416e-05,
+      "loss": 0.8393,
+      "step": 4382
+    },
+    {
+      "epoch": 0.07771086206056094,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.931064904851003e-05,
+      "loss": 0.8447,
+      "step": 4384
+    },
+    {
+      "epoch": 0.07774631409617251,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.930999757710979e-05,
+      "loss": 0.8567,
+      "step": 4386
+    },
+    {
+      "epoch": 0.07778176613178409,
+      "grad_norm": 2.75,
+      "learning_rate": 4.930934580232585e-05,
+      "loss": 0.8612,
+      "step": 4388
+    },
+    {
+      "epoch": 0.07781721816739565,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.930869372416634e-05,
+      "loss": 0.8599,
+      "step": 4390
+    },
+    {
+      "epoch": 0.07785267020300722,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.930804134263939e-05,
+      "loss": 0.861,
+      "step": 4392
+    },
+    {
+      "epoch": 0.0778881222386188,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.930738865775315e-05,
+      "loss": 0.8451,
+      "step": 4394
+    },
+    {
+      "epoch": 0.07792357427423036,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.930673566951577e-05,
+      "loss": 0.8911,
+      "step": 4396
+    },
+    {
+      "epoch": 0.07795902630984193,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.9306082377935384e-05,
+      "loss": 0.8285,
+      "step": 4398
+    },
+    {
+      "epoch": 0.07799447834545349,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.930542878302015e-05,
+      "loss": 0.8523,
+      "step": 4400
+    },
+    {
+      "epoch": 0.07802993038106507,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.9304774884778224e-05,
+      "loss": 0.866,
+      "step": 4402
+    },
+    {
+      "epoch": 0.07806538241667664,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.930412068321778e-05,
+      "loss": 0.8531,
+      "step": 4404
+    },
+    {
+      "epoch": 0.0781008344522882,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.930346617834697e-05,
+      "loss": 0.8139,
+      "step": 4406
+    },
+    {
+      "epoch": 0.07813628648789978,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.9302811370173966e-05,
+      "loss": 0.836,
+      "step": 4408
+    },
+    {
+      "epoch": 0.07817173852351135,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.930215625870693e-05,
+      "loss": 0.8246,
+      "step": 4410
+    },
+    {
+      "epoch": 0.07820719055912291,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.930150084395405e-05,
+      "loss": 0.7908,
+      "step": 4412
+    },
+    {
+      "epoch": 0.07824264259473449,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.93008451259235e-05,
+      "loss": 0.8515,
+      "step": 4414
+    },
+    {
+      "epoch": 0.07827809463034606,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.9300189104623466e-05,
+      "loss": 0.8549,
+      "step": 4416
+    },
+    {
+      "epoch": 0.07831354666595762,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.929953278006213e-05,
+      "loss": 0.842,
+      "step": 4418
+    },
+    {
+      "epoch": 0.0783489987015692,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.929887615224769e-05,
+      "loss": 0.8326,
+      "step": 4420
+    },
+    {
+      "epoch": 0.07838445073718077,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.9298219221188336e-05,
+      "loss": 0.844,
+      "step": 4422
+    },
+    {
+      "epoch": 0.07841990277279233,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.929756198689227e-05,
+      "loss": 0.8625,
+      "step": 4424
+    },
+    {
+      "epoch": 0.0784553548084039,
+      "grad_norm": 2.375,
+      "learning_rate": 4.9296904449367685e-05,
+      "loss": 0.8317,
+      "step": 4426
+    },
+    {
+      "epoch": 0.07849080684401548,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.9296246608622795e-05,
+      "loss": 0.795,
+      "step": 4428
+    },
+    {
+      "epoch": 0.07852625887962704,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.929558846466581e-05,
+      "loss": 0.8097,
+      "step": 4430
+    },
+    {
+      "epoch": 0.07856171091523861,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.929493001750494e-05,
+      "loss": 0.834,
+      "step": 4432
+    },
+    {
+      "epoch": 0.07859716295085019,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.9294271267148405e-05,
+      "loss": 0.8514,
+      "step": 4434
+    },
+    {
+      "epoch": 0.07863261498646175,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.929361221360442e-05,
+      "loss": 0.8629,
+      "step": 4436
+    },
+    {
+      "epoch": 0.07866806702207332,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.929295285688122e-05,
+      "loss": 0.854,
+      "step": 4438
+    },
+    {
+      "epoch": 0.0787035190576849,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.929229319698703e-05,
+      "loss": 0.8652,
+      "step": 4440
+    },
+    {
+      "epoch": 0.07873897109329646,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.929163323393008e-05,
+      "loss": 0.8475,
+      "step": 4442
+    },
+    {
+      "epoch": 0.07877442312890803,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.92909729677186e-05,
+      "loss": 0.8511,
+      "step": 4444
+    },
+    {
+      "epoch": 0.07880987516451961,
+      "grad_norm": 2.5,
+      "learning_rate": 4.929031239836084e-05,
+      "loss": 0.7948,
+      "step": 4446
+    },
+    {
+      "epoch": 0.07884532720013117,
+      "grad_norm": 3.25,
+      "learning_rate": 4.9289651525865046e-05,
+      "loss": 0.8679,
+      "step": 4448
+    },
+    {
+      "epoch": 0.07888077923574274,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.928899035023945e-05,
+      "loss": 0.8649,
+      "step": 4450
+    },
+    {
+      "epoch": 0.07891623127135432,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.9288328871492315e-05,
+      "loss": 0.8551,
+      "step": 4452
+    },
+    {
+      "epoch": 0.07895168330696588,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.9287667089631904e-05,
+      "loss": 0.7862,
+      "step": 4454
+    },
+    {
+      "epoch": 0.07898713534257745,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.9287005004666465e-05,
+      "loss": 0.8378,
+      "step": 4456
+    },
+    {
+      "epoch": 0.07902258737818903,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.928634261660425e-05,
+      "loss": 0.8194,
+      "step": 4458
+    },
+    {
+      "epoch": 0.07905803941380059,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.9285679925453545e-05,
+      "loss": 0.8504,
+      "step": 4460
+    },
+    {
+      "epoch": 0.07909349144941216,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.928501693122262e-05,
+      "loss": 0.877,
+      "step": 4462
+    },
+    {
+      "epoch": 0.07912894348502374,
+      "grad_norm": 2.625,
+      "learning_rate": 4.928435363391973e-05,
+      "loss": 0.8103,
+      "step": 4464
+    },
+    {
+      "epoch": 0.0791643955206353,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.9283690033553174e-05,
+      "loss": 0.8439,
+      "step": 4466
+    },
+    {
+      "epoch": 0.07919984755624687,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.928302613013122e-05,
+      "loss": 0.7667,
+      "step": 4468
+    },
+    {
+      "epoch": 0.07923529959185845,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.9282361923662156e-05,
+      "loss": 0.8139,
+      "step": 4470
+    },
+    {
+      "epoch": 0.07927075162747,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.928169741415428e-05,
+      "loss": 0.8062,
+      "step": 4472
+    },
+    {
+      "epoch": 0.07930620366308158,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.928103260161587e-05,
+      "loss": 0.8375,
+      "step": 4474
+    },
+    {
+      "epoch": 0.07934165569869316,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.928036748605523e-05,
+      "loss": 0.874,
+      "step": 4476
+    },
+    {
+      "epoch": 0.07937710773430472,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.927970206748067e-05,
+      "loss": 0.8288,
+      "step": 4478
+    },
+    {
+      "epoch": 0.07941255976991629,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.927903634590048e-05,
+      "loss": 0.8458,
+      "step": 4480
+    },
+    {
+      "epoch": 0.07944801180552787,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.927837032132297e-05,
+      "loss": 0.8423,
+      "step": 4482
+    },
+    {
+      "epoch": 0.07948346384113943,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.927770399375646e-05,
+      "loss": 0.8363,
+      "step": 4484
+    },
+    {
+      "epoch": 0.079518915876751,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.9277037363209256e-05,
+      "loss": 0.8574,
+      "step": 4486
+    },
+    {
+      "epoch": 0.07955436791236256,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.927637042968969e-05,
+      "loss": 0.8135,
+      "step": 4488
+    },
+    {
+      "epoch": 0.07958981994797414,
+      "grad_norm": 2.75,
+      "learning_rate": 4.927570319320607e-05,
+      "loss": 0.8335,
+      "step": 4490
+    },
+    {
+      "epoch": 0.07962527198358571,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.9275035653766735e-05,
+      "loss": 0.848,
+      "step": 4492
+    },
+    {
+      "epoch": 0.07966072401919727,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.927436781138001e-05,
+      "loss": 0.8307,
+      "step": 4494
+    },
+    {
+      "epoch": 0.07969617605480885,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.927369966605423e-05,
+      "loss": 0.8645,
+      "step": 4496
+    },
+    {
+      "epoch": 0.07973162809042042,
+      "grad_norm": 2.875,
+      "learning_rate": 4.927303121779773e-05,
+      "loss": 0.8851,
+      "step": 4498
+    },
+    {
+      "epoch": 0.07976708012603198,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.927236246661886e-05,
+      "loss": 0.8013,
+      "step": 4500
+    },
+    {
+      "epoch": 0.07980253216164356,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.927169341252596e-05,
+      "loss": 0.8538,
+      "step": 4502
+    },
+    {
+      "epoch": 0.07983798419725513,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.927102405552738e-05,
+      "loss": 0.828,
+      "step": 4504
+    },
+    {
+      "epoch": 0.07987343623286669,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.927035439563149e-05,
+      "loss": 0.8303,
+      "step": 4506
+    },
+    {
+      "epoch": 0.07990888826847826,
+      "grad_norm": 2.625,
+      "learning_rate": 4.926968443284662e-05,
+      "loss": 0.8525,
+      "step": 4508
+    },
+    {
+      "epoch": 0.07994434030408984,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.926901416718114e-05,
+      "loss": 0.8281,
+      "step": 4510
+    },
+    {
+      "epoch": 0.0799797923397014,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.926834359864342e-05,
+      "loss": 0.8465,
+      "step": 4512
+    },
+    {
+      "epoch": 0.08001524437531297,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.9267672727241834e-05,
+      "loss": 0.8424,
+      "step": 4514
+    },
+    {
+      "epoch": 0.08005069641092455,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.926700155298474e-05,
+      "loss": 0.792,
+      "step": 4516
+    },
+    {
+      "epoch": 0.08008614844653611,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.926633007588053e-05,
+      "loss": 0.8201,
+      "step": 4518
+    },
+    {
+      "epoch": 0.08012160048214768,
+      "grad_norm": 2.75,
+      "learning_rate": 4.926565829593756e-05,
+      "loss": 0.8458,
+      "step": 4520
+    },
+    {
+      "epoch": 0.08015705251775926,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.9264986213164235e-05,
+      "loss": 0.8366,
+      "step": 4522
+    },
+    {
+      "epoch": 0.08019250455337082,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.926431382756894e-05,
+      "loss": 0.8485,
+      "step": 4524
+    },
+    {
+      "epoch": 0.0802279565889824,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.926364113916006e-05,
+      "loss": 0.8815,
+      "step": 4526
+    },
+    {
+      "epoch": 0.08026340862459397,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.926296814794599e-05,
+      "loss": 0.844,
+      "step": 4528
+    },
+    {
+      "epoch": 0.08029886066020553,
+      "grad_norm": 2.875,
+      "learning_rate": 4.926229485393513e-05,
+      "loss": 0.7989,
+      "step": 4530
+    },
+    {
+      "epoch": 0.0803343126958171,
+      "grad_norm": 2.625,
+      "learning_rate": 4.926162125713589e-05,
+      "loss": 0.8728,
+      "step": 4532
+    },
+    {
+      "epoch": 0.08036976473142868,
+      "grad_norm": 2.625,
+      "learning_rate": 4.9260947357556666e-05,
+      "loss": 0.8449,
+      "step": 4534
+    },
+    {
+      "epoch": 0.08040521676704024,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.926027315520588e-05,
+      "loss": 0.8506,
+      "step": 4536
+    },
+    {
+      "epoch": 0.08044066880265181,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.925959865009193e-05,
+      "loss": 0.8575,
+      "step": 4538
+    },
+    {
+      "epoch": 0.08047612083826339,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.925892384222324e-05,
+      "loss": 0.8483,
+      "step": 4540
+    },
+    {
+      "epoch": 0.08051157287387495,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.9258248731608235e-05,
+      "loss": 0.8541,
+      "step": 4542
+    },
+    {
+      "epoch": 0.08054702490948652,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.9257573318255344e-05,
+      "loss": 0.8851,
+      "step": 4544
+    },
+    {
+      "epoch": 0.0805824769450981,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.9256897602172986e-05,
+      "loss": 0.8394,
+      "step": 4546
+    },
+    {
+      "epoch": 0.08061792898070966,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.92562215833696e-05,
+      "loss": 0.8504,
+      "step": 4548
+    },
+    {
+      "epoch": 0.08065338101632123,
+      "grad_norm": 3.125,
+      "learning_rate": 4.925554526185362e-05,
+      "loss": 0.8731,
+      "step": 4550
+    },
+    {
+      "epoch": 0.0806888330519328,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.925486863763349e-05,
+      "loss": 0.8335,
+      "step": 4552
+    },
+    {
+      "epoch": 0.08072428508754437,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.925419171071765e-05,
+      "loss": 0.8089,
+      "step": 4554
+    },
+    {
+      "epoch": 0.08075973712315594,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.9253514481114535e-05,
+      "loss": 0.8675,
+      "step": 4556
+    },
+    {
+      "epoch": 0.08079518915876752,
+      "grad_norm": 3.0,
+      "learning_rate": 4.925283694883263e-05,
+      "loss": 0.8822,
+      "step": 4558
+    },
+    {
+      "epoch": 0.08083064119437908,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.9252159113880365e-05,
+      "loss": 0.8938,
+      "step": 4560
+    },
+    {
+      "epoch": 0.08086609322999065,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.925148097626621e-05,
+      "loss": 0.8671,
+      "step": 4562
+    },
+    {
+      "epoch": 0.08090154526560223,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.9250802535998605e-05,
+      "loss": 0.8473,
+      "step": 4564
+    },
+    {
+      "epoch": 0.08093699730121379,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.925012379308606e-05,
+      "loss": 0.7768,
+      "step": 4566
+    },
+    {
+      "epoch": 0.08097244933682536,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.924944474753701e-05,
+      "loss": 0.8361,
+      "step": 4568
+    },
+    {
+      "epoch": 0.08100790137243692,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.9248765399359934e-05,
+      "loss": 0.9217,
+      "step": 4570
+    },
+    {
+      "epoch": 0.0810433534080485,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.924808574856332e-05,
+      "loss": 0.8699,
+      "step": 4572
+    },
+    {
+      "epoch": 0.08107880544366007,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.9247405795155655e-05,
+      "loss": 0.8332,
+      "step": 4574
+    },
+    {
+      "epoch": 0.08111425747927163,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.924672553914541e-05,
+      "loss": 0.8051,
+      "step": 4576
+    },
+    {
+      "epoch": 0.0811497095148832,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.9246044980541084e-05,
+      "loss": 0.8353,
+      "step": 4578
+    },
+    {
+      "epoch": 0.08118516155049478,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.924536411935116e-05,
+      "loss": 0.8489,
+      "step": 4580
+    },
+    {
+      "epoch": 0.08122061358610634,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.924468295558415e-05,
+      "loss": 0.8225,
+      "step": 4582
+    },
+    {
+      "epoch": 0.08125606562171792,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.9244001489248535e-05,
+      "loss": 0.8798,
+      "step": 4584
+    },
+    {
+      "epoch": 0.08129151765732949,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.924331972035284e-05,
+      "loss": 0.8132,
+      "step": 4586
+    },
+    {
+      "epoch": 0.08132696969294105,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.924263764890557e-05,
+      "loss": 0.8357,
+      "step": 4588
+    },
+    {
+      "epoch": 0.08136242172855263,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.924195527491522e-05,
+      "loss": 0.8486,
+      "step": 4590
+    },
+    {
+      "epoch": 0.0813978737641642,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.924127259839032e-05,
+      "loss": 0.7952,
+      "step": 4592
+    },
+    {
+      "epoch": 0.08143332579977576,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.924058961933939e-05,
+      "loss": 0.8301,
+      "step": 4594
+    },
+    {
+      "epoch": 0.08146877783538733,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.923990633777095e-05,
+      "loss": 0.8903,
+      "step": 4596
+    },
+    {
+      "epoch": 0.08150422987099891,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.923922275369352e-05,
+      "loss": 0.8317,
+      "step": 4598
+    },
+    {
+      "epoch": 0.08153968190661047,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.923853886711565e-05,
+      "loss": 0.8319,
+      "step": 4600
+    },
+    {
+      "epoch": 0.08157513394222204,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.9237854678045855e-05,
+      "loss": 0.8296,
+      "step": 4602
+    },
+    {
+      "epoch": 0.08161058597783362,
+      "grad_norm": 2.625,
+      "learning_rate": 4.923717018649269e-05,
+      "loss": 0.8253,
+      "step": 4604
+    },
+    {
+      "epoch": 0.08164603801344518,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.923648539246468e-05,
+      "loss": 0.8293,
+      "step": 4606
+    },
+    {
+      "epoch": 0.08168149004905675,
+      "grad_norm": 3.421875,
+      "learning_rate": 4.923580029597039e-05,
+      "loss": 0.8476,
+      "step": 4608
+    },
+    {
+      "epoch": 0.08171694208466833,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.923511489701835e-05,
+      "loss": 0.8445,
+      "step": 4610
+    },
+    {
+      "epoch": 0.08175239412027989,
+      "grad_norm": 2.75,
+      "learning_rate": 4.923442919561714e-05,
+      "loss": 0.8342,
+      "step": 4612
+    },
+    {
+      "epoch": 0.08178784615589146,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.9233743191775286e-05,
+      "loss": 0.8761,
+      "step": 4614
+    },
+    {
+      "epoch": 0.08182329819150304,
+      "grad_norm": 3.25,
+      "learning_rate": 4.923305688550137e-05,
+      "loss": 0.8501,
+      "step": 4616
+    },
+    {
+      "epoch": 0.0818587502271146,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.9232370276803955e-05,
+      "loss": 0.8057,
+      "step": 4618
+    },
+    {
+      "epoch": 0.08189420226272617,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.923168336569159e-05,
+      "loss": 0.8029,
+      "step": 4620
+    },
+    {
+      "epoch": 0.08192965429833775,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.923099615217288e-05,
+      "loss": 0.8876,
+      "step": 4622
+    },
+    {
+      "epoch": 0.08196510633394931,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.9230308636256385e-05,
+      "loss": 0.8402,
+      "step": 4624
+    },
+    {
+      "epoch": 0.08200055836956088,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.922962081795068e-05,
+      "loss": 0.8426,
+      "step": 4626
+    },
+    {
+      "epoch": 0.08203601040517246,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.922893269726435e-05,
+      "loss": 0.8373,
+      "step": 4628
+    },
+    {
+      "epoch": 0.08207146244078402,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.922824427420599e-05,
+      "loss": 0.8424,
+      "step": 4630
+    },
+    {
+      "epoch": 0.08210691447639559,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.9227555548784196e-05,
+      "loss": 0.8827,
+      "step": 4632
+    },
+    {
+      "epoch": 0.08214236651200717,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.922686652100754e-05,
+      "loss": 0.8642,
+      "step": 4634
+    },
+    {
+      "epoch": 0.08217781854761873,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.9226177190884645e-05,
+      "loss": 0.8704,
+      "step": 4636
+    },
+    {
+      "epoch": 0.0822132705832303,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.922548755842411e-05,
+      "loss": 0.8341,
+      "step": 4638
+    },
+    {
+      "epoch": 0.08224872261884188,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.922479762363453e-05,
+      "loss": 0.8625,
+      "step": 4640
+    },
+    {
+      "epoch": 0.08228417465445344,
+      "grad_norm": 2.75,
+      "learning_rate": 4.922410738652452e-05,
+      "loss": 0.813,
+      "step": 4642
+    },
+    {
+      "epoch": 0.08231962669006501,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.92234168471027e-05,
+      "loss": 0.8251,
+      "step": 4644
+    },
+    {
+      "epoch": 0.08235507872567659,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.922272600537767e-05,
+      "loss": 0.8559,
+      "step": 4646
+    },
+    {
+      "epoch": 0.08239053076128815,
+      "grad_norm": 2.75,
+      "learning_rate": 4.922203486135808e-05,
+      "loss": 0.8383,
+      "step": 4648
+    },
+    {
+      "epoch": 0.08242598279689972,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.9221343415052534e-05,
+      "loss": 0.8475,
+      "step": 4650
+    },
+    {
+      "epoch": 0.0824614348325113,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.922065166646966e-05,
+      "loss": 0.8242,
+      "step": 4652
+    },
+    {
+      "epoch": 0.08249688686812286,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.921995961561811e-05,
+      "loss": 0.8398,
+      "step": 4654
+    },
+    {
+      "epoch": 0.08253233890373443,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.921926726250651e-05,
+      "loss": 0.8347,
+      "step": 4656
+    },
+    {
+      "epoch": 0.08256779093934599,
+      "grad_norm": 2.875,
+      "learning_rate": 4.9218574607143485e-05,
+      "loss": 0.8158,
+      "step": 4658
+    },
+    {
+      "epoch": 0.08260324297495757,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.92178816495377e-05,
+      "loss": 0.8181,
+      "step": 4660
+    },
+    {
+      "epoch": 0.08263869501056914,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.9217188389697796e-05,
+      "loss": 0.8476,
+      "step": 4662
+    },
+    {
+      "epoch": 0.0826741470461807,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.921649482763243e-05,
+      "loss": 0.8557,
+      "step": 4664
+    },
+    {
+      "epoch": 0.08270959908179228,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.921580096335025e-05,
+      "loss": 0.8052,
+      "step": 4666
+    },
+    {
+      "epoch": 0.08274505111740385,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.921510679685992e-05,
+      "loss": 0.8763,
+      "step": 4668
+    },
+    {
+      "epoch": 0.08278050315301541,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.921441232817009e-05,
+      "loss": 0.8867,
+      "step": 4670
+    },
+    {
+      "epoch": 0.08281595518862699,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.921371755728945e-05,
+      "loss": 0.8104,
+      "step": 4672
+    },
+    {
+      "epoch": 0.08285140722423856,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.921302248422665e-05,
+      "loss": 0.8524,
+      "step": 4674
+    },
+    {
+      "epoch": 0.08288685925985012,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.9212327108990376e-05,
+      "loss": 0.837,
+      "step": 4676
+    },
+    {
+      "epoch": 0.0829223112954617,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.92116314315893e-05,
+      "loss": 0.8133,
+      "step": 4678
+    },
+    {
+      "epoch": 0.08295776333107327,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.921093545203211e-05,
+      "loss": 0.8389,
+      "step": 4680
+    },
+    {
+      "epoch": 0.08299321536668483,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.921023917032749e-05,
+      "loss": 0.8538,
+      "step": 4682
+    },
+    {
+      "epoch": 0.0830286674022964,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.920954258648413e-05,
+      "loss": 0.8015,
+      "step": 4684
+    },
+    {
+      "epoch": 0.08306411943790798,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.9208845700510707e-05,
+      "loss": 0.8108,
+      "step": 4686
+    },
+    {
+      "epoch": 0.08309957147351954,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.920814851241595e-05,
+      "loss": 0.8173,
+      "step": 4688
+    },
+    {
+      "epoch": 0.08313502350913111,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.9207451022208525e-05,
+      "loss": 0.8193,
+      "step": 4690
+    },
+    {
+      "epoch": 0.08317047554474269,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.9206753229897165e-05,
+      "loss": 0.8298,
+      "step": 4692
+    },
+    {
+      "epoch": 0.08320592758035425,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.9206055135490563e-05,
+      "loss": 0.8196,
+      "step": 4694
+    },
+    {
+      "epoch": 0.08324137961596582,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.920535673899743e-05,
+      "loss": 0.8335,
+      "step": 4696
+    },
+    {
+      "epoch": 0.0832768316515774,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.9204658040426496e-05,
+      "loss": 0.8865,
+      "step": 4698
+    },
+    {
+      "epoch": 0.08331228368718896,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.9203959039786465e-05,
+      "loss": 0.82,
+      "step": 4700
+    },
+    {
+      "epoch": 0.08334773572280053,
+      "grad_norm": 2.875,
+      "learning_rate": 4.920325973708607e-05,
+      "loss": 0.8465,
+      "step": 4702
+    },
+    {
+      "epoch": 0.08338318775841211,
+      "grad_norm": 2.5,
+      "learning_rate": 4.920256013233403e-05,
+      "loss": 0.8028,
+      "step": 4704
+    },
+    {
+      "epoch": 0.08341863979402367,
+      "grad_norm": 2.75,
+      "learning_rate": 4.920186022553909e-05,
+      "loss": 0.838,
+      "step": 4706
+    },
+    {
+      "epoch": 0.08345409182963524,
+      "grad_norm": 2.75,
+      "learning_rate": 4.9201160016709964e-05,
+      "loss": 0.8395,
+      "step": 4708
+    },
+    {
+      "epoch": 0.08348954386524682,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.920045950585541e-05,
+      "loss": 0.8705,
+      "step": 4710
+    },
+    {
+      "epoch": 0.08352499590085838,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.919975869298416e-05,
+      "loss": 0.8168,
+      "step": 4712
+    },
+    {
+      "epoch": 0.08356044793646995,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.919905757810496e-05,
+      "loss": 0.8021,
+      "step": 4714
+    },
+    {
+      "epoch": 0.08359589997208153,
+      "grad_norm": 3.125,
+      "learning_rate": 4.9198356161226555e-05,
+      "loss": 0.8569,
+      "step": 4716
+    },
+    {
+      "epoch": 0.08363135200769309,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.919765444235771e-05,
+      "loss": 0.8266,
+      "step": 4718
+    },
+    {
+      "epoch": 0.08366680404330466,
+      "grad_norm": 2.75,
+      "learning_rate": 4.919695242150718e-05,
+      "loss": 0.8239,
+      "step": 4720
+    },
+    {
+      "epoch": 0.08370225607891624,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.919625009868373e-05,
+      "loss": 0.8637,
+      "step": 4722
+    },
+    {
+      "epoch": 0.0837377081145278,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.919554747389611e-05,
+      "loss": 0.8091,
+      "step": 4724
+    },
+    {
+      "epoch": 0.08377316015013937,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.9194844547153095e-05,
+      "loss": 0.8482,
+      "step": 4726
+    },
+    {
+      "epoch": 0.08380861218575095,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.919414131846346e-05,
+      "loss": 0.8627,
+      "step": 4728
+    },
+    {
+      "epoch": 0.08384406422136251,
+      "grad_norm": 2.75,
+      "learning_rate": 4.9193437787835987e-05,
+      "loss": 0.8841,
+      "step": 4730
+    },
+    {
+      "epoch": 0.08387951625697408,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.9192733955279446e-05,
+      "loss": 0.8262,
+      "step": 4732
+    },
+    {
+      "epoch": 0.08391496829258566,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.919202982080262e-05,
+      "loss": 0.8459,
+      "step": 4734
+    },
+    {
+      "epoch": 0.08395042032819722,
+      "grad_norm": 2.75,
+      "learning_rate": 4.919132538441431e-05,
+      "loss": 0.8286,
+      "step": 4736
+    },
+    {
+      "epoch": 0.08398587236380879,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.919062064612329e-05,
+      "loss": 0.834,
+      "step": 4738
+    },
+    {
+      "epoch": 0.08402132439942035,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.918991560593837e-05,
+      "loss": 0.8834,
+      "step": 4740
+    },
+    {
+      "epoch": 0.08405677643503193,
+      "grad_norm": 2.625,
+      "learning_rate": 4.9189210263868335e-05,
+      "loss": 0.7786,
+      "step": 4742
+    },
+    {
+      "epoch": 0.0840922284706435,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.9188504619922e-05,
+      "loss": 0.8179,
+      "step": 4744
+    },
+    {
+      "epoch": 0.08412768050625506,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.918779867410817e-05,
+      "loss": 0.8539,
+      "step": 4746
+    },
+    {
+      "epoch": 0.08416313254186664,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.9187092426435634e-05,
+      "loss": 0.8234,
+      "step": 4748
+    },
+    {
+      "epoch": 0.08419858457747821,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.918638587691323e-05,
+      "loss": 0.841,
+      "step": 4750
+    },
+    {
+      "epoch": 0.08423403661308977,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.918567902554977e-05,
+      "loss": 0.8399,
+      "step": 4752
+    },
+    {
+      "epoch": 0.08426948864870135,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.918497187235407e-05,
+      "loss": 0.8286,
+      "step": 4754
+    },
+    {
+      "epoch": 0.08430494068431292,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.918426441733496e-05,
+      "loss": 0.8949,
+      "step": 4756
+    },
+    {
+      "epoch": 0.08434039271992448,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.918355666050127e-05,
+      "loss": 0.8858,
+      "step": 4758
+    },
+    {
+      "epoch": 0.08437584475553606,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.918284860186183e-05,
+      "loss": 0.8442,
+      "step": 4760
+    },
+    {
+      "epoch": 0.08441129679114763,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.918214024142547e-05,
+      "loss": 0.859,
+      "step": 4762
+    },
+    {
+      "epoch": 0.08444674882675919,
+      "grad_norm": 2.875,
+      "learning_rate": 4.918143157920104e-05,
+      "loss": 0.8918,
+      "step": 4764
+    },
+    {
+      "epoch": 0.08448220086237077,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.918072261519738e-05,
+      "loss": 0.8376,
+      "step": 4766
+    },
+    {
+      "epoch": 0.08451765289798234,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.9180013349423346e-05,
+      "loss": 0.8575,
+      "step": 4768
+    },
+    {
+      "epoch": 0.0845531049335939,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.917930378188778e-05,
+      "loss": 0.8454,
+      "step": 4770
+    },
+    {
+      "epoch": 0.08458855696920548,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.917859391259952e-05,
+      "loss": 0.8589,
+      "step": 4772
+    },
+    {
+      "epoch": 0.08462400900481705,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.917788374156747e-05,
+      "loss": 0.8413,
+      "step": 4774
+    },
+    {
+      "epoch": 0.08465946104042861,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.917717326880045e-05,
+      "loss": 0.8128,
+      "step": 4776
+    },
+    {
+      "epoch": 0.08469491307604018,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.917646249430735e-05,
+      "loss": 0.8709,
+      "step": 4778
+    },
+    {
+      "epoch": 0.08473036511165176,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.917575141809703e-05,
+      "loss": 0.8053,
+      "step": 4780
+    },
+    {
+      "epoch": 0.08476581714726332,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.917504004017837e-05,
+      "loss": 0.8852,
+      "step": 4782
+    },
+    {
+      "epoch": 0.0848012691828749,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.917432836056025e-05,
+      "loss": 0.8599,
+      "step": 4784
+    },
+    {
+      "epoch": 0.08483672121848647,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.917361637925154e-05,
+      "loss": 0.8266,
+      "step": 4786
+    },
+    {
+      "epoch": 0.08487217325409803,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.9172904096261136e-05,
+      "loss": 0.8163,
+      "step": 4788
+    },
+    {
+      "epoch": 0.0849076252897096,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.917219151159792e-05,
+      "loss": 0.856,
+      "step": 4790
+    },
+    {
+      "epoch": 0.08494307732532118,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.917147862527079e-05,
+      "loss": 0.8377,
+      "step": 4792
+    },
+    {
+      "epoch": 0.08497852936093274,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.9170765437288644e-05,
+      "loss": 0.8261,
+      "step": 4794
+    },
+    {
+      "epoch": 0.08501398139654431,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.917005194766038e-05,
+      "loss": 0.8279,
+      "step": 4796
+    },
+    {
+      "epoch": 0.08504943343215589,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.9169338156394904e-05,
+      "loss": 0.8655,
+      "step": 4798
+    },
+    {
+      "epoch": 0.08508488546776745,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.916862406350112e-05,
+      "loss": 0.8507,
+      "step": 4800
+    },
+    {
+      "epoch": 0.08512033750337902,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.9167909668987935e-05,
+      "loss": 0.8584,
+      "step": 4802
+    },
+    {
+      "epoch": 0.0851557895389906,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.9167194972864275e-05,
+      "loss": 0.8287,
+      "step": 4804
+    },
+    {
+      "epoch": 0.08519124157460216,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.916647997513906e-05,
+      "loss": 0.8062,
+      "step": 4806
+    },
+    {
+      "epoch": 0.08522669361021373,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.91657646758212e-05,
+      "loss": 0.8694,
+      "step": 4808
+    },
+    {
+      "epoch": 0.08526214564582531,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.9165049074919646e-05,
+      "loss": 0.8329,
+      "step": 4810
+    },
+    {
+      "epoch": 0.08529759768143687,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.91643331724433e-05,
+      "loss": 0.8391,
+      "step": 4812
+    },
+    {
+      "epoch": 0.08533304971704844,
+      "grad_norm": 2.625,
+      "learning_rate": 4.91636169684011e-05,
+      "loss": 0.8347,
+      "step": 4814
+    },
+    {
+      "epoch": 0.08536850175266002,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.916290046280201e-05,
+      "loss": 0.8175,
+      "step": 4816
+    },
+    {
+      "epoch": 0.08540395378827158,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.916218365565495e-05,
+      "loss": 0.8785,
+      "step": 4818
+    },
+    {
+      "epoch": 0.08543940582388315,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.916146654696887e-05,
+      "loss": 0.8315,
+      "step": 4820
+    },
+    {
+      "epoch": 0.08547485785949473,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.916074913675272e-05,
+      "loss": 0.8088,
+      "step": 4822
+    },
+    {
+      "epoch": 0.08551030989510629,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.916003142501545e-05,
+      "loss": 0.818,
+      "step": 4824
+    },
+    {
+      "epoch": 0.08554576193071786,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.915931341176603e-05,
+      "loss": 0.8604,
+      "step": 4826
+    },
+    {
+      "epoch": 0.08558121396632942,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.915859509701341e-05,
+      "loss": 0.8509,
+      "step": 4828
+    },
+    {
+      "epoch": 0.085616666001941,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.915787648076654e-05,
+      "loss": 0.8354,
+      "step": 4830
+    },
+    {
+      "epoch": 0.08565211803755257,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.9157157563034414e-05,
+      "loss": 0.858,
+      "step": 4832
+    },
+    {
+      "epoch": 0.08568757007316413,
+      "grad_norm": 2.875,
+      "learning_rate": 4.915643834382599e-05,
+      "loss": 0.8217,
+      "step": 4834
+    },
+    {
+      "epoch": 0.0857230221087757,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.915571882315024e-05,
+      "loss": 0.839,
+      "step": 4836
+    },
+    {
+      "epoch": 0.08575847414438728,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.9154999001016165e-05,
+      "loss": 0.8276,
+      "step": 4838
+    },
+    {
+      "epoch": 0.08579392617999884,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.915427887743273e-05,
+      "loss": 0.8278,
+      "step": 4840
+    },
+    {
+      "epoch": 0.08582937821561042,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.915355845240892e-05,
+      "loss": 0.8404,
+      "step": 4842
+    },
+    {
+      "epoch": 0.08586483025122199,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.915283772595373e-05,
+      "loss": 0.8471,
+      "step": 4844
+    },
+    {
+      "epoch": 0.08590028228683355,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.915211669807616e-05,
+      "loss": 0.8204,
+      "step": 4846
+    },
+    {
+      "epoch": 0.08593573432244513,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.915139536878521e-05,
+      "loss": 0.8382,
+      "step": 4848
+    },
+    {
+      "epoch": 0.0859711863580567,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.915067373808987e-05,
+      "loss": 0.8489,
+      "step": 4850
+    },
+    {
+      "epoch": 0.08600663839366826,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.914995180599915e-05,
+      "loss": 0.8831,
+      "step": 4852
+    },
+    {
+      "epoch": 0.08604209042927984,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.914922957252206e-05,
+      "loss": 0.8454,
+      "step": 4854
+    },
+    {
+      "epoch": 0.08607754246489141,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.914850703766762e-05,
+      "loss": 0.8631,
+      "step": 4856
+    },
+    {
+      "epoch": 0.08611299450050297,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.914778420144484e-05,
+      "loss": 0.8594,
+      "step": 4858
+    },
+    {
+      "epoch": 0.08614844653611455,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.914706106386274e-05,
+      "loss": 0.8227,
+      "step": 4860
+    },
+    {
+      "epoch": 0.08618389857172612,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.914633762493035e-05,
+      "loss": 0.8217,
+      "step": 4862
+    },
+    {
+      "epoch": 0.08621935060733768,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.91456138846567e-05,
+      "loss": 0.8461,
+      "step": 4864
+    },
+    {
+      "epoch": 0.08625480264294925,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.914488984305081e-05,
+      "loss": 0.8445,
+      "step": 4866
+    },
+    {
+      "epoch": 0.08629025467856083,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.914416550012173e-05,
+      "loss": 0.8095,
+      "step": 4868
+    },
+    {
+      "epoch": 0.08632570671417239,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.9143440855878496e-05,
+      "loss": 0.8082,
+      "step": 4870
+    },
+    {
+      "epoch": 0.08636115874978396,
+      "grad_norm": 2.625,
+      "learning_rate": 4.914271591033014e-05,
+      "loss": 0.8297,
+      "step": 4872
+    },
+    {
+      "epoch": 0.08639661078539554,
+      "grad_norm": 2.625,
+      "learning_rate": 4.914199066348573e-05,
+      "loss": 0.8541,
+      "step": 4874
+    },
+    {
+      "epoch": 0.0864320628210071,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.9141265115354294e-05,
+      "loss": 0.804,
+      "step": 4876
+    },
+    {
+      "epoch": 0.08646751485661867,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.9140539265944894e-05,
+      "loss": 0.8892,
+      "step": 4878
+    },
+    {
+      "epoch": 0.08650296689223025,
+      "grad_norm": 2.75,
+      "learning_rate": 4.9139813115266604e-05,
+      "loss": 0.8561,
+      "step": 4880
+    },
+    {
+      "epoch": 0.08653841892784181,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.913908666332847e-05,
+      "loss": 0.8416,
+      "step": 4882
+    },
+    {
+      "epoch": 0.08657387096345338,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.9138359910139556e-05,
+      "loss": 0.8247,
+      "step": 4884
+    },
+    {
+      "epoch": 0.08660932299906496,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.9137632855708946e-05,
+      "loss": 0.8197,
+      "step": 4886
+    },
+    {
+      "epoch": 0.08664477503467652,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.9136905500045705e-05,
+      "loss": 0.8617,
+      "step": 4888
+    },
+    {
+      "epoch": 0.0866802270702881,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.9136177843158906e-05,
+      "loss": 0.8368,
+      "step": 4890
+    },
+    {
+      "epoch": 0.08671567910589967,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.9135449885057636e-05,
+      "loss": 0.8302,
+      "step": 4892
+    },
+    {
+      "epoch": 0.08675113114151123,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.913472162575098e-05,
+      "loss": 0.8489,
+      "step": 4894
+    },
+    {
+      "epoch": 0.0867865831771228,
+      "grad_norm": 2.875,
+      "learning_rate": 4.913399306524803e-05,
+      "loss": 0.8218,
+      "step": 4896
+    },
+    {
+      "epoch": 0.08682203521273438,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.913326420355787e-05,
+      "loss": 0.8156,
+      "step": 4898
+    },
+    {
+      "epoch": 0.08685748724834594,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.913253504068959e-05,
+      "loss": 0.8496,
+      "step": 4900
+    },
+    {
+      "epoch": 0.08689293928395751,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.9131805576652315e-05,
+      "loss": 0.8293,
+      "step": 4902
+    },
+    {
+      "epoch": 0.08692839131956909,
+      "grad_norm": 2.75,
+      "learning_rate": 4.913107581145512e-05,
+      "loss": 0.8324,
+      "step": 4904
+    },
+    {
+      "epoch": 0.08696384335518065,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.913034574510713e-05,
+      "loss": 0.8057,
+      "step": 4906
+    },
+    {
+      "epoch": 0.08699929539079222,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.9129615377617445e-05,
+      "loss": 0.8315,
+      "step": 4908
+    },
+    {
+      "epoch": 0.08703474742640378,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.912888470899519e-05,
+      "loss": 0.8528,
+      "step": 4910
+    },
+    {
+      "epoch": 0.08707019946201536,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.912815373924948e-05,
+      "loss": 0.8459,
+      "step": 4912
+    },
+    {
+      "epoch": 0.08710565149762693,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.912742246838944e-05,
+      "loss": 0.8431,
+      "step": 4914
+    },
+    {
+      "epoch": 0.08714110353323849,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.9126690896424195e-05,
+      "loss": 0.8502,
+      "step": 4916
+    },
+    {
+      "epoch": 0.08717655556885007,
+      "grad_norm": 2.875,
+      "learning_rate": 4.912595902336287e-05,
+      "loss": 0.8689,
+      "step": 4918
+    },
+    {
+      "epoch": 0.08721200760446164,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.912522684921459e-05,
+      "loss": 0.8495,
+      "step": 4920
+    },
+    {
+      "epoch": 0.0872474596400732,
+      "grad_norm": 2.875,
+      "learning_rate": 4.9124494373988516e-05,
+      "loss": 0.8241,
+      "step": 4922
+    },
+    {
+      "epoch": 0.08728291167568478,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.912376159769378e-05,
+      "loss": 0.8331,
+      "step": 4924
+    },
+    {
+      "epoch": 0.08731836371129635,
+      "grad_norm": 2.625,
+      "learning_rate": 4.912302852033952e-05,
+      "loss": 0.8012,
+      "step": 4926
+    },
+    {
+      "epoch": 0.08735381574690791,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.912229514193488e-05,
+      "loss": 0.8198,
+      "step": 4928
+    },
+    {
+      "epoch": 0.08738926778251949,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.912156146248903e-05,
+      "loss": 0.812,
+      "step": 4930
+    },
+    {
+      "epoch": 0.08742471981813106,
+      "grad_norm": 2.875,
+      "learning_rate": 4.912082748201112e-05,
+      "loss": 0.8546,
+      "step": 4932
+    },
+    {
+      "epoch": 0.08746017185374262,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.9120093200510295e-05,
+      "loss": 0.8525,
+      "step": 4934
+    },
+    {
+      "epoch": 0.0874956238893542,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.911935861799574e-05,
+      "loss": 0.8227,
+      "step": 4936
+    },
+    {
+      "epoch": 0.08753107592496577,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.911862373447661e-05,
+      "loss": 0.7985,
+      "step": 4938
+    },
+    {
+      "epoch": 0.08756652796057733,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.911788854996209e-05,
+      "loss": 0.8347,
+      "step": 4940
+    },
+    {
+      "epoch": 0.0876019799961889,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.911715306446133e-05,
+      "loss": 0.8388,
+      "step": 4942
+    },
+    {
+      "epoch": 0.08763743203180048,
+      "grad_norm": 3.0,
+      "learning_rate": 4.911641727798353e-05,
+      "loss": 0.8764,
+      "step": 4944
+    },
+    {
+      "epoch": 0.08767288406741204,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.911568119053786e-05,
+      "loss": 0.8823,
+      "step": 4946
+    },
+    {
+      "epoch": 0.08770833610302362,
+      "grad_norm": 3.125,
+      "learning_rate": 4.911494480213351e-05,
+      "loss": 0.8387,
+      "step": 4948
+    },
+    {
+      "epoch": 0.08774378813863519,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.911420811277968e-05,
+      "loss": 0.7991,
+      "step": 4950
+    },
+    {
+      "epoch": 0.08777924017424675,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.9113471122485556e-05,
+      "loss": 0.805,
+      "step": 4952
+    },
+    {
+      "epoch": 0.08781469220985832,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.9112733831260325e-05,
+      "loss": 0.8836,
+      "step": 4954
+    },
+    {
+      "epoch": 0.0878501442454699,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.911199623911321e-05,
+      "loss": 0.8398,
+      "step": 4956
+    },
+    {
+      "epoch": 0.08788559628108146,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.911125834605339e-05,
+      "loss": 0.837,
+      "step": 4958
+    },
+    {
+      "epoch": 0.08792104831669303,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.9110520152090104e-05,
+      "loss": 0.7892,
+      "step": 4960
+    },
+    {
+      "epoch": 0.08795650035230461,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.910978165723253e-05,
+      "loss": 0.8841,
+      "step": 4962
+    },
+    {
+      "epoch": 0.08799195238791617,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.910904286148992e-05,
+      "loss": 0.8603,
+      "step": 4964
+    },
+    {
+      "epoch": 0.08802740442352774,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.910830376487146e-05,
+      "loss": 0.8726,
+      "step": 4966
+    },
+    {
+      "epoch": 0.08806285645913932,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.9107564367386404e-05,
+      "loss": 0.8372,
+      "step": 4968
+    },
+    {
+      "epoch": 0.08809830849475088,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.9106824669043964e-05,
+      "loss": 0.8159,
+      "step": 4970
+    },
+    {
+      "epoch": 0.08813376053036245,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.910608466985337e-05,
+      "loss": 0.8235,
+      "step": 4972
+    },
+    {
+      "epoch": 0.08816921256597403,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.910534436982386e-05,
+      "loss": 0.8469,
+      "step": 4974
+    },
+    {
+      "epoch": 0.08820466460158559,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.910460376896468e-05,
+      "loss": 0.8395,
+      "step": 4976
+    },
+    {
+      "epoch": 0.08824011663719716,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.9103862867285056e-05,
+      "loss": 0.8427,
+      "step": 4978
+    },
+    {
+      "epoch": 0.08827556867280874,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.910312166479425e-05,
+      "loss": 0.8158,
+      "step": 4980
+    },
+    {
+      "epoch": 0.0883110207084203,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.910238016150151e-05,
+      "loss": 0.8614,
+      "step": 4982
+    },
+    {
+      "epoch": 0.08834647274403187,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.910163835741607e-05,
+      "loss": 0.8538,
+      "step": 4984
+    },
+    {
+      "epoch": 0.08838192477964345,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.9100896252547215e-05,
+      "loss": 0.8273,
+      "step": 4986
+    },
+    {
+      "epoch": 0.08841737681525501,
+      "grad_norm": 2.625,
+      "learning_rate": 4.91001538469042e-05,
+      "loss": 0.792,
+      "step": 4988
+    },
+    {
+      "epoch": 0.08845282885086658,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.909941114049627e-05,
+      "loss": 0.8621,
+      "step": 4990
+    },
+    {
+      "epoch": 0.08848828088647816,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.9098668133332714e-05,
+      "loss": 0.8404,
+      "step": 4992
+    },
+    {
+      "epoch": 0.08852373292208972,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.90979248254228e-05,
+      "loss": 0.8261,
+      "step": 4994
+    },
+    {
+      "epoch": 0.08855918495770129,
+      "grad_norm": 2.75,
+      "learning_rate": 4.9097181216775805e-05,
+      "loss": 0.8467,
+      "step": 4996
+    },
+    {
+      "epoch": 0.08859463699331285,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.9096437307401004e-05,
+      "loss": 0.8357,
+      "step": 4998
+    },
+    {
+      "epoch": 0.08863008902892443,
+      "grad_norm": 2.625,
+      "learning_rate": 4.909569309730769e-05,
+      "loss": 0.8499,
+      "step": 5000
+    },
+    {
+      "epoch": 0.088665541064536,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.909494858650514e-05,
+      "loss": 0.8532,
+      "step": 5002
+    },
+    {
+      "epoch": 0.08870099310014756,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.909420377500266e-05,
+      "loss": 0.8132,
+      "step": 5004
+    },
+    {
+      "epoch": 0.08873644513575914,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.909345866280952e-05,
+      "loss": 0.8429,
+      "step": 5006
+    },
+    {
+      "epoch": 0.08877189717137071,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.909271324993504e-05,
+      "loss": 0.8266,
+      "step": 5008
+    },
+    {
+      "epoch": 0.08880734920698227,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.909196753638852e-05,
+      "loss": 0.8255,
+      "step": 5010
+    },
+    {
+      "epoch": 0.08884280124259385,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.9091221522179264e-05,
+      "loss": 0.8205,
+      "step": 5012
+    },
+    {
+      "epoch": 0.08887825327820542,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.909047520731658e-05,
+      "loss": 0.8872,
+      "step": 5014
+    },
+    {
+      "epoch": 0.08891370531381698,
+      "grad_norm": 2.625,
+      "learning_rate": 4.908972859180978e-05,
+      "loss": 0.8599,
+      "step": 5016
+    },
+    {
+      "epoch": 0.08894915734942856,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.9088981675668185e-05,
+      "loss": 0.8741,
+      "step": 5018
+    },
+    {
+      "epoch": 0.08898460938504013,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.9088234458901114e-05,
+      "loss": 0.839,
+      "step": 5020
+    },
+    {
+      "epoch": 0.08902006142065169,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.90874869415179e-05,
+      "loss": 0.8378,
+      "step": 5022
+    },
+    {
+      "epoch": 0.08905551345626327,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.908673912352786e-05,
+      "loss": 0.8216,
+      "step": 5024
+    },
+    {
+      "epoch": 0.08909096549187484,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.908599100494034e-05,
+      "loss": 0.8167,
+      "step": 5026
+    },
+    {
+      "epoch": 0.0891264175274864,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.908524258576467e-05,
+      "loss": 0.8109,
+      "step": 5028
+    },
+    {
+      "epoch": 0.08916186956309798,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.908449386601019e-05,
+      "loss": 0.8242,
+      "step": 5030
+    },
+    {
+      "epoch": 0.08919732159870955,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.908374484568623e-05,
+      "loss": 0.8618,
+      "step": 5032
+    },
+    {
+      "epoch": 0.08923277363432111,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.9082995524802157e-05,
+      "loss": 0.8933,
+      "step": 5034
+    },
+    {
+      "epoch": 0.08926822566993269,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.908224590336732e-05,
+      "loss": 0.9062,
+      "step": 5036
+    },
+    {
+      "epoch": 0.08930367770554426,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.908149598139107e-05,
+      "loss": 0.81,
+      "step": 5038
+    },
+    {
+      "epoch": 0.08933912974115582,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.908074575888276e-05,
+      "loss": 0.8303,
+      "step": 5040
+    },
+    {
+      "epoch": 0.0893745817767674,
+      "grad_norm": 2.5,
+      "learning_rate": 4.907999523585176e-05,
+      "loss": 0.836,
+      "step": 5042
+    },
+    {
+      "epoch": 0.08941003381237897,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.907924441230743e-05,
+      "loss": 0.7865,
+      "step": 5044
+    },
+    {
+      "epoch": 0.08944548584799053,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.907849328825916e-05,
+      "loss": 0.8685,
+      "step": 5046
+    },
+    {
+      "epoch": 0.0894809378836021,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.9077741863716296e-05,
+      "loss": 0.8481,
+      "step": 5048
+    },
+    {
+      "epoch": 0.08951638991921368,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.907699013868824e-05,
+      "loss": 0.8459,
+      "step": 5050
+    },
+    {
+      "epoch": 0.08955184195482524,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.9076238113184344e-05,
+      "loss": 0.8189,
+      "step": 5052
+    },
+    {
+      "epoch": 0.08958729399043681,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.9075485787214024e-05,
+      "loss": 0.818,
+      "step": 5054
+    },
+    {
+      "epoch": 0.08962274602604839,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.907473316078666e-05,
+      "loss": 0.8193,
+      "step": 5056
+    },
+    {
+      "epoch": 0.08965819806165995,
+      "grad_norm": 3.0,
+      "learning_rate": 4.9073980233911635e-05,
+      "loss": 0.8221,
+      "step": 5058
+    },
+    {
+      "epoch": 0.08969365009727152,
+      "grad_norm": 2.75,
+      "learning_rate": 4.907322700659835e-05,
+      "loss": 0.8591,
+      "step": 5060
+    },
+    {
+      "epoch": 0.0897291021328831,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.907247347885621e-05,
+      "loss": 0.8252,
+      "step": 5062
+    },
+    {
+      "epoch": 0.08976455416849466,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.907171965069461e-05,
+      "loss": 0.7966,
+      "step": 5064
+    },
+    {
+      "epoch": 0.08980000620410623,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.907096552212296e-05,
+      "loss": 0.82,
+      "step": 5066
+    },
+    {
+      "epoch": 0.08983545823971781,
+      "grad_norm": 3.0,
+      "learning_rate": 4.907021109315068e-05,
+      "loss": 0.8558,
+      "step": 5068
+    },
+    {
+      "epoch": 0.08987091027532937,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.906945636378718e-05,
+      "loss": 0.8481,
+      "step": 5070
+    },
+    {
+      "epoch": 0.08990636231094094,
+      "grad_norm": 2.875,
+      "learning_rate": 4.906870133404187e-05,
+      "loss": 0.8009,
+      "step": 5072
+    },
+    {
+      "epoch": 0.08994181434655252,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.906794600392419e-05,
+      "loss": 0.834,
+      "step": 5074
+    },
+    {
+      "epoch": 0.08997726638216408,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.906719037344355e-05,
+      "loss": 0.8621,
+      "step": 5076
+    },
+    {
+      "epoch": 0.09001271841777565,
+      "grad_norm": 3.0,
+      "learning_rate": 4.906643444260939e-05,
+      "loss": 0.8749,
+      "step": 5078
+    },
+    {
+      "epoch": 0.09004817045338723,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.906567821143114e-05,
+      "loss": 0.7919,
+      "step": 5080
+    },
+    {
+      "epoch": 0.09008362248899879,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.906492167991824e-05,
+      "loss": 0.8311,
+      "step": 5082
+    },
+    {
+      "epoch": 0.09011907452461036,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.906416484808013e-05,
+      "loss": 0.8423,
+      "step": 5084
+    },
+    {
+      "epoch": 0.09015452656022192,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.9063407715926255e-05,
+      "loss": 0.8415,
+      "step": 5086
+    },
+    {
+      "epoch": 0.0901899785958335,
+      "grad_norm": 2.625,
+      "learning_rate": 4.906265028346606e-05,
+      "loss": 0.8695,
+      "step": 5088
+    },
+    {
+      "epoch": 0.09022543063144507,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.9061892550709e-05,
+      "loss": 0.8689,
+      "step": 5090
+    },
+    {
+      "epoch": 0.09026088266705663,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.906113451766454e-05,
+      "loss": 0.8314,
+      "step": 5092
+    },
+    {
+      "epoch": 0.09029633470266821,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.906037618434213e-05,
+      "loss": 0.834,
+      "step": 5094
+    },
+    {
+      "epoch": 0.09033178673827978,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.905961755075124e-05,
+      "loss": 0.8216,
+      "step": 5096
+    },
+    {
+      "epoch": 0.09036723877389134,
+      "grad_norm": 2.375,
+      "learning_rate": 4.905885861690133e-05,
+      "loss": 0.7918,
+      "step": 5098
+    },
+    {
+      "epoch": 0.09040269080950292,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.905809938280187e-05,
+      "loss": 0.8244,
+      "step": 5100
+    },
+    {
+      "epoch": 0.09043814284511449,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.905733984846235e-05,
+      "loss": 0.8276,
+      "step": 5102
+    },
+    {
+      "epoch": 0.09047359488072605,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.905658001389223e-05,
+      "loss": 0.8623,
+      "step": 5104
+    },
+    {
+      "epoch": 0.09050904691633763,
+      "grad_norm": 2.875,
+      "learning_rate": 4.905581987910101e-05,
+      "loss": 0.812,
+      "step": 5106
+    },
+    {
+      "epoch": 0.0905444989519492,
+      "grad_norm": 2.875,
+      "learning_rate": 4.905505944409816e-05,
+      "loss": 0.8412,
+      "step": 5108
+    },
+    {
+      "epoch": 0.09057995098756076,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.905429870889319e-05,
+      "loss": 0.8368,
+      "step": 5110
+    },
+    {
+      "epoch": 0.09061540302317234,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.9053537673495576e-05,
+      "loss": 0.8673,
+      "step": 5112
+    },
+    {
+      "epoch": 0.09065085505878391,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.9052776337914816e-05,
+      "loss": 0.8744,
+      "step": 5114
+    },
+    {
+      "epoch": 0.09068630709439547,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.905201470216043e-05,
+      "loss": 0.7991,
+      "step": 5116
+    },
+    {
+      "epoch": 0.09072175913000705,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.90512527662419e-05,
+      "loss": 0.8332,
+      "step": 5118
+    },
+    {
+      "epoch": 0.09075721116561862,
+      "grad_norm": 3.0,
+      "learning_rate": 4.9050490530168743e-05,
+      "loss": 0.8392,
+      "step": 5120
+    },
+    {
+      "epoch": 0.09079266320123018,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.9049727993950486e-05,
+      "loss": 0.8683,
+      "step": 5122
+    },
+    {
+      "epoch": 0.09082811523684176,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.904896515759663e-05,
+      "loss": 0.8318,
+      "step": 5124
+    },
+    {
+      "epoch": 0.09086356727245333,
+      "grad_norm": 2.75,
+      "learning_rate": 4.904820202111669e-05,
+      "loss": 0.8412,
+      "step": 5126
+    },
+    {
+      "epoch": 0.09089901930806489,
+      "grad_norm": 3.125,
+      "learning_rate": 4.90474385845202e-05,
+      "loss": 0.8313,
+      "step": 5128
+    },
+    {
+      "epoch": 0.09093447134367647,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.9046674847816685e-05,
+      "loss": 0.8337,
+      "step": 5130
+    },
+    {
+      "epoch": 0.09096992337928804,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.904591081101568e-05,
+      "loss": 0.8189,
+      "step": 5132
+    },
+    {
+      "epoch": 0.0910053754148996,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.904514647412672e-05,
+      "loss": 0.8416,
+      "step": 5134
+    },
+    {
+      "epoch": 0.09104082745051117,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.904438183715933e-05,
+      "loss": 0.8311,
+      "step": 5136
+    },
+    {
+      "epoch": 0.09107627948612275,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.904361690012307e-05,
+      "loss": 0.8309,
+      "step": 5138
+    },
+    {
+      "epoch": 0.09111173152173431,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.904285166302748e-05,
+      "loss": 0.8497,
+      "step": 5140
+    },
+    {
+      "epoch": 0.09114718355734588,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.9042086125882104e-05,
+      "loss": 0.869,
+      "step": 5142
+    },
+    {
+      "epoch": 0.09118263559295746,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.9041320288696505e-05,
+      "loss": 0.8205,
+      "step": 5144
+    },
+    {
+      "epoch": 0.09121808762856902,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.9040554151480235e-05,
+      "loss": 0.825,
+      "step": 5146
+    },
+    {
+      "epoch": 0.0912535396641806,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.903978771424285e-05,
+      "loss": 0.8514,
+      "step": 5148
+    },
+    {
+      "epoch": 0.09128899169979217,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.903902097699393e-05,
+      "loss": 0.8235,
+      "step": 5150
+    },
+    {
+      "epoch": 0.09132444373540373,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.903825393974303e-05,
+      "loss": 0.869,
+      "step": 5152
+    },
+    {
+      "epoch": 0.0913598957710153,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.903748660249973e-05,
+      "loss": 0.8642,
+      "step": 5154
+    },
+    {
+      "epoch": 0.09139534780662688,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.90367189652736e-05,
+      "loss": 0.8293,
+      "step": 5156
+    },
+    {
+      "epoch": 0.09143079984223844,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.903595102807423e-05,
+      "loss": 0.8324,
+      "step": 5158
+    },
+    {
+      "epoch": 0.09146625187785001,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.903518279091119e-05,
+      "loss": 0.8141,
+      "step": 5160
+    },
+    {
+      "epoch": 0.09150170391346159,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.903441425379408e-05,
+      "loss": 0.8715,
+      "step": 5162
+    },
+    {
+      "epoch": 0.09153715594907315,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.9033645416732486e-05,
+      "loss": 0.8393,
+      "step": 5164
+    },
+    {
+      "epoch": 0.09157260798468472,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.903287627973601e-05,
+      "loss": 0.8631,
+      "step": 5166
+    },
+    {
+      "epoch": 0.09160806002029628,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.903210684281423e-05,
+      "loss": 0.8291,
+      "step": 5168
+    },
+    {
+      "epoch": 0.09164351205590786,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.903133710597677e-05,
+      "loss": 0.8772,
+      "step": 5170
+    },
+    {
+      "epoch": 0.09167896409151943,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.903056706923322e-05,
+      "loss": 0.851,
+      "step": 5172
+    },
+    {
+      "epoch": 0.091714416127131,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.902979673259321e-05,
+      "loss": 0.8378,
+      "step": 5174
+    },
+    {
+      "epoch": 0.09174986816274257,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.902902609606634e-05,
+      "loss": 0.8006,
+      "step": 5176
+    },
+    {
+      "epoch": 0.09178532019835414,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.902825515966223e-05,
+      "loss": 0.8601,
+      "step": 5178
+    },
+    {
+      "epoch": 0.0918207722339657,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.902748392339049e-05,
+      "loss": 0.8244,
+      "step": 5180
+    },
+    {
+      "epoch": 0.09185622426957728,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.902671238726076e-05,
+      "loss": 0.8326,
+      "step": 5182
+    },
+    {
+      "epoch": 0.09189167630518885,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.902594055128267e-05,
+      "loss": 0.8521,
+      "step": 5184
+    },
+    {
+      "epoch": 0.09192712834080041,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.902516841546584e-05,
+      "loss": 0.8358,
+      "step": 5186
+    },
+    {
+      "epoch": 0.09196258037641199,
+      "grad_norm": 2.875,
+      "learning_rate": 4.9024395979819915e-05,
+      "loss": 0.8225,
+      "step": 5188
+    },
+    {
+      "epoch": 0.09199803241202356,
+      "grad_norm": 2.625,
+      "learning_rate": 4.902362324435453e-05,
+      "loss": 0.8264,
+      "step": 5190
+    },
+    {
+      "epoch": 0.09203348444763512,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.902285020907933e-05,
+      "loss": 0.8246,
+      "step": 5192
+    },
+    {
+      "epoch": 0.0920689364832467,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.902207687400396e-05,
+      "loss": 0.8338,
+      "step": 5194
+    },
+    {
+      "epoch": 0.09210438851885827,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.902130323913808e-05,
+      "loss": 0.8269,
+      "step": 5196
+    },
+    {
+      "epoch": 0.09213984055446983,
+      "grad_norm": 2.75,
+      "learning_rate": 4.9020529304491345e-05,
+      "loss": 0.8291,
+      "step": 5198
+    },
+    {
+      "epoch": 0.0921752925900814,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.90197550700734e-05,
+      "loss": 0.8452,
+      "step": 5200
+    },
+    {
+      "epoch": 0.09221074462569298,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.901898053589391e-05,
+      "loss": 0.8634,
+      "step": 5202
+    },
+    {
+      "epoch": 0.09224619666130454,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.901820570196255e-05,
+      "loss": 0.8451,
+      "step": 5204
+    },
+    {
+      "epoch": 0.09228164869691612,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.901743056828899e-05,
+      "loss": 0.9183,
+      "step": 5206
+    },
+    {
+      "epoch": 0.09231710073252769,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.901665513488289e-05,
+      "loss": 0.8343,
+      "step": 5208
+    },
+    {
+      "epoch": 0.09235255276813925,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.901587940175394e-05,
+      "loss": 0.8573,
+      "step": 5210
+    },
+    {
+      "epoch": 0.09238800480375083,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.9015103368911816e-05,
+      "loss": 0.8048,
+      "step": 5212
+    },
+    {
+      "epoch": 0.0924234568393624,
+      "grad_norm": 2.75,
+      "learning_rate": 4.901432703636622e-05,
+      "loss": 0.8261,
+      "step": 5214
+    },
+    {
+      "epoch": 0.09245890887497396,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.901355040412681e-05,
+      "loss": 0.8197,
+      "step": 5216
+    },
+    {
+      "epoch": 0.09249436091058554,
+      "grad_norm": 2.875,
+      "learning_rate": 4.901277347220329e-05,
+      "loss": 0.8323,
+      "step": 5218
+    },
+    {
+      "epoch": 0.09252981294619711,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.901199624060536e-05,
+      "loss": 0.7934,
+      "step": 5220
+    },
+    {
+      "epoch": 0.09256526498180867,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.901121870934272e-05,
+      "loss": 0.8816,
+      "step": 5222
+    },
+    {
+      "epoch": 0.09260071701742024,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.901044087842507e-05,
+      "loss": 0.8536,
+      "step": 5224
+    },
+    {
+      "epoch": 0.09263616905303182,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.9009662747862115e-05,
+      "loss": 0.837,
+      "step": 5226
+    },
+    {
+      "epoch": 0.09267162108864338,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.900888431766359e-05,
+      "loss": 0.8126,
+      "step": 5228
+    },
+    {
+      "epoch": 0.09270707312425495,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.900810558783917e-05,
+      "loss": 0.8512,
+      "step": 5230
+    },
+    {
+      "epoch": 0.09274252515986653,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.9007326558398595e-05,
+      "loss": 0.8265,
+      "step": 5232
+    },
+    {
+      "epoch": 0.09277797719547809,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.9006547229351587e-05,
+      "loss": 0.8007,
+      "step": 5234
+    },
+    {
+      "epoch": 0.09281342923108966,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.9005767600707866e-05,
+      "loss": 0.8372,
+      "step": 5236
+    },
+    {
+      "epoch": 0.09284888126670124,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.900498767247717e-05,
+      "loss": 0.8367,
+      "step": 5238
+    },
+    {
+      "epoch": 0.0928843333023128,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.9004207444669224e-05,
+      "loss": 0.8489,
+      "step": 5240
+    },
+    {
+      "epoch": 0.09291978533792437,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.900342691729378e-05,
+      "loss": 0.828,
+      "step": 5242
+    },
+    {
+      "epoch": 0.09295523737353595,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.900264609036056e-05,
+      "loss": 0.7999,
+      "step": 5244
+    },
+    {
+      "epoch": 0.09299068940914751,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.900186496387931e-05,
+      "loss": 0.831,
+      "step": 5246
+    },
+    {
+      "epoch": 0.09302614144475908,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.900108353785979e-05,
+      "loss": 0.8824,
+      "step": 5248
+    },
+    {
+      "epoch": 0.09306159348037066,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.900030181231175e-05,
+      "loss": 0.826,
+      "step": 5250
+    },
+    {
+      "epoch": 0.09309704551598222,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.899951978724494e-05,
+      "loss": 0.8538,
+      "step": 5252
+    },
+    {
+      "epoch": 0.0931324975515938,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.899873746266912e-05,
+      "loss": 0.8773,
+      "step": 5254
+    },
+    {
+      "epoch": 0.09316794958720535,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.8997954838594055e-05,
+      "loss": 0.8843,
+      "step": 5256
+    },
+    {
+      "epoch": 0.09320340162281693,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.899717191502951e-05,
+      "loss": 0.7928,
+      "step": 5258
+    },
+    {
+      "epoch": 0.0932388536584285,
+      "grad_norm": 2.875,
+      "learning_rate": 4.8996388691985265e-05,
+      "loss": 0.8471,
+      "step": 5260
+    },
+    {
+      "epoch": 0.09327430569404006,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.899560516947108e-05,
+      "loss": 0.8468,
+      "step": 5262
+    },
+    {
+      "epoch": 0.09330975772965164,
+      "grad_norm": 2.5,
+      "learning_rate": 4.899482134749674e-05,
+      "loss": 0.8251,
+      "step": 5264
+    },
+    {
+      "epoch": 0.09334520976526321,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.8994037226072037e-05,
+      "loss": 0.7813,
+      "step": 5266
+    },
+    {
+      "epoch": 0.09338066180087477,
+      "grad_norm": 2.75,
+      "learning_rate": 4.899325280520674e-05,
+      "loss": 0.8214,
+      "step": 5268
+    },
+    {
+      "epoch": 0.09341611383648635,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.899246808491065e-05,
+      "loss": 0.853,
+      "step": 5270
+    },
+    {
+      "epoch": 0.09345156587209792,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.899168306519355e-05,
+      "loss": 0.8129,
+      "step": 5272
+    },
+    {
+      "epoch": 0.09348701790770948,
+      "grad_norm": 2.75,
+      "learning_rate": 4.899089774606525e-05,
+      "loss": 0.8524,
+      "step": 5274
+    },
+    {
+      "epoch": 0.09352246994332106,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.8990112127535525e-05,
+      "loss": 0.7781,
+      "step": 5276
+    },
+    {
+      "epoch": 0.09355792197893263,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.898932620961422e-05,
+      "loss": 0.8256,
+      "step": 5278
+    },
+    {
+      "epoch": 0.09359337401454419,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.89885399923111e-05,
+      "loss": 0.8313,
+      "step": 5280
+    },
+    {
+      "epoch": 0.09362882605015577,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.8987753475636014e-05,
+      "loss": 0.8584,
+      "step": 5282
+    },
+    {
+      "epoch": 0.09366427808576734,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.8986966659598756e-05,
+      "loss": 0.8305,
+      "step": 5284
+    },
+    {
+      "epoch": 0.0936997301213789,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.8986179544209146e-05,
+      "loss": 0.7863,
+      "step": 5286
+    },
+    {
+      "epoch": 0.09373518215699048,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.8985392129477014e-05,
+      "loss": 0.8056,
+      "step": 5288
+    },
+    {
+      "epoch": 0.09377063419260205,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.898460441541218e-05,
+      "loss": 0.8324,
+      "step": 5290
+    },
+    {
+      "epoch": 0.09380608622821361,
+      "grad_norm": 2.875,
+      "learning_rate": 4.8983816402024484e-05,
+      "loss": 0.8534,
+      "step": 5292
+    },
+    {
+      "epoch": 0.09384153826382519,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.8983028089323756e-05,
+      "loss": 0.8487,
+      "step": 5294
+    },
+    {
+      "epoch": 0.09387699029943676,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.8982239477319824e-05,
+      "loss": 0.8318,
+      "step": 5296
+    },
+    {
+      "epoch": 0.09391244233504832,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.8981450566022545e-05,
+      "loss": 0.8097,
+      "step": 5298
+    },
+    {
+      "epoch": 0.0939478943706599,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.898066135544176e-05,
+      "loss": 0.8471,
+      "step": 5300
+    },
+    {
+      "epoch": 0.09398334640627147,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.897987184558731e-05,
+      "loss": 0.8206,
+      "step": 5302
+    },
+    {
+      "epoch": 0.09401879844188303,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.897908203646906e-05,
+      "loss": 0.8759,
+      "step": 5304
+    },
+    {
+      "epoch": 0.0940542504774946,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.897829192809686e-05,
+      "loss": 0.7891,
+      "step": 5306
+    },
+    {
+      "epoch": 0.09408970251310618,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.897750152048057e-05,
+      "loss": 0.8524,
+      "step": 5308
+    },
+    {
+      "epoch": 0.09412515454871774,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.8976710813630056e-05,
+      "loss": 0.8392,
+      "step": 5310
+    },
+    {
+      "epoch": 0.09416060658432931,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.897591980755518e-05,
+      "loss": 0.84,
+      "step": 5312
+    },
+    {
+      "epoch": 0.09419605861994089,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.897512850226582e-05,
+      "loss": 0.7987,
+      "step": 5314
+    },
+    {
+      "epoch": 0.09423151065555245,
+      "grad_norm": 2.875,
+      "learning_rate": 4.8974336897771855e-05,
+      "loss": 0.853,
+      "step": 5316
+    },
+    {
+      "epoch": 0.09426696269116402,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.8973544994083154e-05,
+      "loss": 0.8278,
+      "step": 5318
+    },
+    {
+      "epoch": 0.0943024147267756,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.897275279120961e-05,
+      "loss": 0.8601,
+      "step": 5320
+    },
+    {
+      "epoch": 0.09433786676238716,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.89719602891611e-05,
+      "loss": 0.8513,
+      "step": 5322
+    },
+    {
+      "epoch": 0.09437331879799873,
+      "grad_norm": 2.625,
+      "learning_rate": 4.8971167487947525e-05,
+      "loss": 0.8074,
+      "step": 5324
+    },
+    {
+      "epoch": 0.09440877083361031,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.897037438757876e-05,
+      "loss": 0.8192,
+      "step": 5326
+    },
+    {
+      "epoch": 0.09444422286922187,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.8969580988064724e-05,
+      "loss": 0.8364,
+      "step": 5328
+    },
+    {
+      "epoch": 0.09447967490483344,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.896878728941531e-05,
+      "loss": 0.832,
+      "step": 5330
+    },
+    {
+      "epoch": 0.09451512694044502,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.896799329164043e-05,
+      "loss": 0.8042,
+      "step": 5332
+    },
+    {
+      "epoch": 0.09455057897605658,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.8967198994749966e-05,
+      "loss": 0.8357,
+      "step": 5334
+    },
+    {
+      "epoch": 0.09458603101166815,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.896640439875386e-05,
+      "loss": 0.8761,
+      "step": 5336
+    },
+    {
+      "epoch": 0.09462148304727971,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.896560950366202e-05,
+      "loss": 0.8303,
+      "step": 5338
+    },
+    {
+      "epoch": 0.09465693508289129,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.896481430948437e-05,
+      "loss": 0.8626,
+      "step": 5340
+    },
+    {
+      "epoch": 0.09469238711850286,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.896401881623083e-05,
+      "loss": 0.8361,
+      "step": 5342
+    },
+    {
+      "epoch": 0.09472783915411442,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.8963223023911315e-05,
+      "loss": 0.8183,
+      "step": 5344
+    },
+    {
+      "epoch": 0.094763291189726,
+      "grad_norm": 3.0,
+      "learning_rate": 4.8962426932535775e-05,
+      "loss": 0.8441,
+      "step": 5346
+    },
+    {
+      "epoch": 0.09479874322533757,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.8961630542114135e-05,
+      "loss": 0.8247,
+      "step": 5348
+    },
+    {
+      "epoch": 0.09483419526094913,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.896083385265634e-05,
+      "loss": 0.8313,
+      "step": 5350
+    },
+    {
+      "epoch": 0.09486964729656071,
+      "grad_norm": 2.75,
+      "learning_rate": 4.896003686417233e-05,
+      "loss": 0.8515,
+      "step": 5352
+    },
+    {
+      "epoch": 0.09490509933217228,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.895923957667204e-05,
+      "loss": 0.8214,
+      "step": 5354
+    },
+    {
+      "epoch": 0.09494055136778384,
+      "grad_norm": 2.875,
+      "learning_rate": 4.895844199016544e-05,
+      "loss": 0.8345,
+      "step": 5356
+    },
+    {
+      "epoch": 0.09497600340339542,
+      "grad_norm": 2.875,
+      "learning_rate": 4.895764410466248e-05,
+      "loss": 0.8197,
+      "step": 5358
+    },
+    {
+      "epoch": 0.09501145543900699,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.89568459201731e-05,
+      "loss": 0.86,
+      "step": 5360
+    },
+    {
+      "epoch": 0.09504690747461855,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.8956047436707276e-05,
+      "loss": 0.8541,
+      "step": 5362
+    },
+    {
+      "epoch": 0.09508235951023013,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.8955248654274974e-05,
+      "loss": 0.8416,
+      "step": 5364
+    },
+    {
+      "epoch": 0.0951178115458417,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.8954449572886154e-05,
+      "loss": 0.8919,
+      "step": 5366
+    },
+    {
+      "epoch": 0.09515326358145326,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.895365019255079e-05,
+      "loss": 0.8213,
+      "step": 5368
+    },
+    {
+      "epoch": 0.09518871561706484,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.895285051327887e-05,
+      "loss": 0.816,
+      "step": 5370
+    },
+    {
+      "epoch": 0.09522416765267641,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.895205053508036e-05,
+      "loss": 0.8613,
+      "step": 5372
+    },
+    {
+      "epoch": 0.09525961968828797,
+      "grad_norm": 2.75,
+      "learning_rate": 4.895125025796525e-05,
+      "loss": 0.8261,
+      "step": 5374
+    },
+    {
+      "epoch": 0.09529507172389955,
+      "grad_norm": 2.5,
+      "learning_rate": 4.895044968194352e-05,
+      "loss": 0.81,
+      "step": 5376
+    },
+    {
+      "epoch": 0.09533052375951112,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.894964880702517e-05,
+      "loss": 0.8326,
+      "step": 5378
+    },
+    {
+      "epoch": 0.09536597579512268,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.894884763322019e-05,
+      "loss": 0.8157,
+      "step": 5380
+    },
+    {
+      "epoch": 0.09540142783073426,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.894804616053858e-05,
+      "loss": 0.8442,
+      "step": 5382
+    },
+    {
+      "epoch": 0.09543687986634583,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.8947244388990345e-05,
+      "loss": 0.8449,
+      "step": 5384
+    },
+    {
+      "epoch": 0.09547233190195739,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.894644231858548e-05,
+      "loss": 0.8198,
+      "step": 5386
+    },
+    {
+      "epoch": 0.09550778393756897,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.894563994933401e-05,
+      "loss": 0.8566,
+      "step": 5388
+    },
+    {
+      "epoch": 0.09554323597318054,
+      "grad_norm": 2.75,
+      "learning_rate": 4.8944837281245934e-05,
+      "loss": 0.8476,
+      "step": 5390
+    },
+    {
+      "epoch": 0.0955786880087921,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.894403431433129e-05,
+      "loss": 0.8254,
+      "step": 5392
+    },
+    {
+      "epoch": 0.09561414004440368,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.894323104860007e-05,
+      "loss": 0.8202,
+      "step": 5394
+    },
+    {
+      "epoch": 0.09564959208001525,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.894242748406232e-05,
+      "loss": 0.8245,
+      "step": 5396
+    },
+    {
+      "epoch": 0.09568504411562681,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.894162362072806e-05,
+      "loss": 0.8343,
+      "step": 5398
+    },
+    {
+      "epoch": 0.09572049615123839,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.8940819458607323e-05,
+      "loss": 0.8675,
+      "step": 5400
+    },
+    {
+      "epoch": 0.09575594818684996,
+      "grad_norm": 2.75,
+      "learning_rate": 4.894001499771015e-05,
+      "loss": 0.8272,
+      "step": 5402
+    },
+    {
+      "epoch": 0.09579140022246152,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.8939210238046577e-05,
+      "loss": 0.7998,
+      "step": 5404
+    },
+    {
+      "epoch": 0.0958268522580731,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.8938405179626644e-05,
+      "loss": 0.8126,
+      "step": 5406
+    },
+    {
+      "epoch": 0.09586230429368467,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.89375998224604e-05,
+      "loss": 0.8555,
+      "step": 5408
+    },
+    {
+      "epoch": 0.09589775632929623,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.8936794166557895e-05,
+      "loss": 0.8274,
+      "step": 5410
+    },
+    {
+      "epoch": 0.0959332083649078,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.893598821192918e-05,
+      "loss": 0.8555,
+      "step": 5412
+    },
+    {
+      "epoch": 0.09596866040051938,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.893518195858433e-05,
+      "loss": 0.776,
+      "step": 5414
+    },
+    {
+      "epoch": 0.09600411243613094,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.8934375406533384e-05,
+      "loss": 0.8853,
+      "step": 5416
+    },
+    {
+      "epoch": 0.09603956447174251,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.8933568555786416e-05,
+      "loss": 0.8464,
+      "step": 5418
+    },
+    {
+      "epoch": 0.09607501650735409,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.8932761406353506e-05,
+      "loss": 0.847,
+      "step": 5420
+    },
+    {
+      "epoch": 0.09611046854296565,
+      "grad_norm": 2.625,
+      "learning_rate": 4.893195395824472e-05,
+      "loss": 0.8339,
+      "step": 5422
+    },
+    {
+      "epoch": 0.09614592057857722,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.8931146211470126e-05,
+      "loss": 0.8212,
+      "step": 5424
+    },
+    {
+      "epoch": 0.09618137261418878,
+      "grad_norm": 2.875,
+      "learning_rate": 4.8930338166039815e-05,
+      "loss": 0.812,
+      "step": 5426
+    },
+    {
+      "epoch": 0.09621682464980036,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.892952982196387e-05,
+      "loss": 0.846,
+      "step": 5428
+    },
+    {
+      "epoch": 0.09625227668541193,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.892872117925237e-05,
+      "loss": 0.8056,
+      "step": 5430
+    },
+    {
+      "epoch": 0.0962877287210235,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.8927912237915416e-05,
+      "loss": 0.8156,
+      "step": 5432
+    },
+    {
+      "epoch": 0.09632318075663507,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.8927102997963105e-05,
+      "loss": 0.7889,
+      "step": 5434
+    },
+    {
+      "epoch": 0.09635863279224664,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.8926293459405524e-05,
+      "loss": 0.8546,
+      "step": 5436
+    },
+    {
+      "epoch": 0.0963940848278582,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.8925483622252796e-05,
+      "loss": 0.8052,
+      "step": 5438
+    },
+    {
+      "epoch": 0.09642953686346978,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.892467348651501e-05,
+      "loss": 0.8228,
+      "step": 5440
+    },
+    {
+      "epoch": 0.09646498889908135,
+      "grad_norm": 2.5,
+      "learning_rate": 4.892386305220228e-05,
+      "loss": 0.8544,
+      "step": 5442
+    },
+    {
+      "epoch": 0.09650044093469291,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.892305231932473e-05,
+      "loss": 0.8331,
+      "step": 5444
+    },
+    {
+      "epoch": 0.09653589297030449,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.892224128789246e-05,
+      "loss": 0.7923,
+      "step": 5446
+    },
+    {
+      "epoch": 0.09657134500591606,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.8921429957915606e-05,
+      "loss": 0.853,
+      "step": 5448
+    },
+    {
+      "epoch": 0.09660679704152762,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.8920618329404286e-05,
+      "loss": 0.8734,
+      "step": 5450
+    },
+    {
+      "epoch": 0.0966422490771392,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.891980640236864e-05,
+      "loss": 0.8263,
+      "step": 5452
+    },
+    {
+      "epoch": 0.09667770111275077,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.891899417681878e-05,
+      "loss": 0.8675,
+      "step": 5454
+    },
+    {
+      "epoch": 0.09671315314836233,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.891818165276486e-05,
+      "loss": 0.7996,
+      "step": 5456
+    },
+    {
+      "epoch": 0.09674860518397391,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.891736883021701e-05,
+      "loss": 0.8463,
+      "step": 5458
+    },
+    {
+      "epoch": 0.09678405721958548,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.891655570918539e-05,
+      "loss": 0.8306,
+      "step": 5460
+    },
+    {
+      "epoch": 0.09681950925519704,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.8915742289680136e-05,
+      "loss": 0.8614,
+      "step": 5462
+    },
+    {
+      "epoch": 0.09685496129080862,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.891492857171139e-05,
+      "loss": 0.8186,
+      "step": 5464
+    },
+    {
+      "epoch": 0.09689041332642019,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.891411455528932e-05,
+      "loss": 0.7825,
+      "step": 5466
+    },
+    {
+      "epoch": 0.09692586536203175,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.891330024042408e-05,
+      "loss": 0.8715,
+      "step": 5468
+    },
+    {
+      "epoch": 0.09696131739764333,
+      "grad_norm": 2.5,
+      "learning_rate": 4.8912485627125835e-05,
+      "loss": 0.823,
+      "step": 5470
+    },
+    {
+      "epoch": 0.0969967694332549,
+      "grad_norm": 2.875,
+      "learning_rate": 4.891167071540475e-05,
+      "loss": 0.8334,
+      "step": 5472
+    },
+    {
+      "epoch": 0.09703222146886646,
+      "grad_norm": 3.0,
+      "learning_rate": 4.8910855505271e-05,
+      "loss": 0.8304,
+      "step": 5474
+    },
+    {
+      "epoch": 0.09706767350447804,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.891003999673475e-05,
+      "loss": 0.8579,
+      "step": 5476
+    },
+    {
+      "epoch": 0.09710312554008961,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.890922418980617e-05,
+      "loss": 0.8458,
+      "step": 5478
+    },
+    {
+      "epoch": 0.09713857757570117,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.890840808449547e-05,
+      "loss": 0.7615,
+      "step": 5480
+    },
+    {
+      "epoch": 0.09717402961131275,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.89075916808128e-05,
+      "loss": 0.8277,
+      "step": 5482
+    },
+    {
+      "epoch": 0.09720948164692432,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.8906774978768376e-05,
+      "loss": 0.8035,
+      "step": 5484
+    },
+    {
+      "epoch": 0.09724493368253588,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.8905957978372377e-05,
+      "loss": 0.8383,
+      "step": 5486
+    },
+    {
+      "epoch": 0.09728038571814746,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.890514067963501e-05,
+      "loss": 0.8408,
+      "step": 5488
+    },
+    {
+      "epoch": 0.09731583775375903,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.8904323082566456e-05,
+      "loss": 0.8368,
+      "step": 5490
+    },
+    {
+      "epoch": 0.09735128978937059,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.890350518717693e-05,
+      "loss": 0.8349,
+      "step": 5492
+    },
+    {
+      "epoch": 0.09738674182498216,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.890268699347664e-05,
+      "loss": 0.8517,
+      "step": 5494
+    },
+    {
+      "epoch": 0.09742219386059374,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.890186850147579e-05,
+      "loss": 0.8121,
+      "step": 5496
+    },
+    {
+      "epoch": 0.0974576458962053,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.890104971118461e-05,
+      "loss": 0.8557,
+      "step": 5498
+    },
+    {
+      "epoch": 0.09749309793181687,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.8900230622613294e-05,
+      "loss": 0.8497,
+      "step": 5500
+    },
+    {
+      "epoch": 0.09752854996742845,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.889941123577209e-05,
+      "loss": 0.8452,
+      "step": 5502
+    },
+    {
+      "epoch": 0.09756400200304001,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.8898591550671205e-05,
+      "loss": 0.825,
+      "step": 5504
+    },
+    {
+      "epoch": 0.09759945403865158,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.889777156732088e-05,
+      "loss": 0.8256,
+      "step": 5506
+    },
+    {
+      "epoch": 0.09763490607426314,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.889695128573134e-05,
+      "loss": 0.8618,
+      "step": 5508
+    },
+    {
+      "epoch": 0.09767035810987472,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.889613070591283e-05,
+      "loss": 0.8529,
+      "step": 5510
+    },
+    {
+      "epoch": 0.0977058101454863,
+      "grad_norm": 2.625,
+      "learning_rate": 4.889530982787558e-05,
+      "loss": 0.825,
+      "step": 5512
+    },
+    {
+      "epoch": 0.09774126218109785,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.8894488651629844e-05,
+      "loss": 0.8353,
+      "step": 5514
+    },
+    {
+      "epoch": 0.09777671421670943,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.889366717718587e-05,
+      "loss": 0.8779,
+      "step": 5516
+    },
+    {
+      "epoch": 0.097812166252321,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.8892845404553897e-05,
+      "loss": 0.7985,
+      "step": 5518
+    },
+    {
+      "epoch": 0.09784761828793256,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.88920233337442e-05,
+      "loss": 0.8459,
+      "step": 5520
+    },
+    {
+      "epoch": 0.09788307032354414,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.8891200964767014e-05,
+      "loss": 0.7836,
+      "step": 5522
+    },
+    {
+      "epoch": 0.09791852235915571,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.889037829763262e-05,
+      "loss": 0.8491,
+      "step": 5524
+    },
+    {
+      "epoch": 0.09795397439476727,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.888955533235129e-05,
+      "loss": 0.844,
+      "step": 5526
+    },
+    {
+      "epoch": 0.09798942643037885,
+      "grad_norm": 2.625,
+      "learning_rate": 4.888873206893328e-05,
+      "loss": 0.8221,
+      "step": 5528
+    },
+    {
+      "epoch": 0.09802487846599042,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.888790850738887e-05,
+      "loss": 0.8318,
+      "step": 5530
+    },
+    {
+      "epoch": 0.09806033050160198,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.888708464772834e-05,
+      "loss": 0.8489,
+      "step": 5532
+    },
+    {
+      "epoch": 0.09809578253721356,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.8886260489961963e-05,
+      "loss": 0.8574,
+      "step": 5534
+    },
+    {
+      "epoch": 0.09813123457282513,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.888543603410004e-05,
+      "loss": 0.8249,
+      "step": 5536
+    },
+    {
+      "epoch": 0.09816668660843669,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.888461128015283e-05,
+      "loss": 0.8662,
+      "step": 5538
+    },
+    {
+      "epoch": 0.09820213864404827,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.888378622813067e-05,
+      "loss": 0.8338,
+      "step": 5540
+    },
+    {
+      "epoch": 0.09823759067965984,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.888296087804383e-05,
+      "loss": 0.8295,
+      "step": 5542
+    },
+    {
+      "epoch": 0.0982730427152714,
+      "grad_norm": 2.75,
+      "learning_rate": 4.88821352299026e-05,
+      "loss": 0.8539,
+      "step": 5544
+    },
+    {
+      "epoch": 0.09830849475088298,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.8881309283717305e-05,
+      "loss": 0.8205,
+      "step": 5546
+    },
+    {
+      "epoch": 0.09834394678649455,
+      "grad_norm": 2.375,
+      "learning_rate": 4.888048303949824e-05,
+      "loss": 0.8314,
+      "step": 5548
+    },
+    {
+      "epoch": 0.09837939882210611,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.887965649725572e-05,
+      "loss": 0.878,
+      "step": 5550
+    },
+    {
+      "epoch": 0.09841485085771769,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.887882965700006e-05,
+      "loss": 0.8663,
+      "step": 5552
+    },
+    {
+      "epoch": 0.09845030289332926,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.8878002518741585e-05,
+      "loss": 0.785,
+      "step": 5554
+    },
+    {
+      "epoch": 0.09848575492894082,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.8877175082490606e-05,
+      "loss": 0.7979,
+      "step": 5556
+    },
+    {
+      "epoch": 0.0985212069645524,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.8876347348257454e-05,
+      "loss": 0.8108,
+      "step": 5558
+    },
+    {
+      "epoch": 0.09855665900016397,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.8875519316052464e-05,
+      "loss": 0.8471,
+      "step": 5560
+    },
+    {
+      "epoch": 0.09859211103577553,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.8874690985885975e-05,
+      "loss": 0.8127,
+      "step": 5562
+    },
+    {
+      "epoch": 0.0986275630713871,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.88738623577683e-05,
+      "loss": 0.8511,
+      "step": 5564
+    },
+    {
+      "epoch": 0.09866301510699868,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.8873033431709804e-05,
+      "loss": 0.8387,
+      "step": 5566
+    },
+    {
+      "epoch": 0.09869846714261024,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.887220420772082e-05,
+      "loss": 0.8615,
+      "step": 5568
+    },
+    {
+      "epoch": 0.09873391917822182,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.88713746858117e-05,
+      "loss": 0.8746,
+      "step": 5570
+    },
+    {
+      "epoch": 0.09876937121383339,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.88705448659928e-05,
+      "loss": 0.8258,
+      "step": 5572
+    },
+    {
+      "epoch": 0.09880482324944495,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.8869714748274464e-05,
+      "loss": 0.8558,
+      "step": 5574
+    },
+    {
+      "epoch": 0.09884027528505653,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.8868884332667066e-05,
+      "loss": 0.8201,
+      "step": 5576
+    },
+    {
+      "epoch": 0.0988757273206681,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.886805361918096e-05,
+      "loss": 0.8155,
+      "step": 5578
+    },
+    {
+      "epoch": 0.09891117935627966,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.886722260782652e-05,
+      "loss": 0.8609,
+      "step": 5580
+    },
+    {
+      "epoch": 0.09894663139189123,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.886639129861411e-05,
+      "loss": 0.833,
+      "step": 5582
+    },
+    {
+      "epoch": 0.09898208342750281,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.886555969155411e-05,
+      "loss": 0.8938,
+      "step": 5584
+    },
+    {
+      "epoch": 0.09901753546311437,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.8864727786656886e-05,
+      "loss": 0.8474,
+      "step": 5586
+    },
+    {
+      "epoch": 0.09905298749872594,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.8863895583932836e-05,
+      "loss": 0.7775,
+      "step": 5588
+    },
+    {
+      "epoch": 0.09908843953433752,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.886306308339235e-05,
+      "loss": 0.853,
+      "step": 5590
+    },
+    {
+      "epoch": 0.09912389156994908,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.8862230285045794e-05,
+      "loss": 0.8057,
+      "step": 5592
+    },
+    {
+      "epoch": 0.09915934360556065,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.886139718890358e-05,
+      "loss": 0.8473,
+      "step": 5594
+    },
+    {
+      "epoch": 0.09919479564117221,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.88605637949761e-05,
+      "loss": 0.829,
+      "step": 5596
+    },
+    {
+      "epoch": 0.09923024767678379,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.885973010327375e-05,
+      "loss": 0.8445,
+      "step": 5598
+    },
+    {
+      "epoch": 0.09926569971239536,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.885889611380694e-05,
+      "loss": 0.8265,
+      "step": 5600
+    },
+    {
+      "epoch": 0.09930115174800692,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.885806182658607e-05,
+      "loss": 0.8717,
+      "step": 5602
+    },
+    {
+      "epoch": 0.0993366037836185,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.885722724162156e-05,
+      "loss": 0.8219,
+      "step": 5604
+    },
+    {
+      "epoch": 0.09937205581923007,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.885639235892383e-05,
+      "loss": 0.8432,
+      "step": 5606
+    },
+    {
+      "epoch": 0.09940750785484163,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.8855557178503286e-05,
+      "loss": 0.8082,
+      "step": 5608
+    },
+    {
+      "epoch": 0.09944295989045321,
+      "grad_norm": 2.875,
+      "learning_rate": 4.885472170037037e-05,
+      "loss": 0.836,
+      "step": 5610
+    },
+    {
+      "epoch": 0.09947841192606478,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.885388592453548e-05,
+      "loss": 0.8404,
+      "step": 5612
+    },
+    {
+      "epoch": 0.09951386396167634,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.8853049851009064e-05,
+      "loss": 0.8383,
+      "step": 5614
+    },
+    {
+      "epoch": 0.09954931599728792,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.885221347980156e-05,
+      "loss": 0.8697,
+      "step": 5616
+    },
+    {
+      "epoch": 0.09958476803289949,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.8851376810923396e-05,
+      "loss": 0.8104,
+      "step": 5618
+    },
+    {
+      "epoch": 0.09962022006851105,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.8850539844385017e-05,
+      "loss": 0.8173,
+      "step": 5620
+    },
+    {
+      "epoch": 0.09965567210412263,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.8849702580196874e-05,
+      "loss": 0.8496,
+      "step": 5622
+    },
+    {
+      "epoch": 0.0996911241397342,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.8848865018369404e-05,
+      "loss": 0.8188,
+      "step": 5624
+    },
+    {
+      "epoch": 0.09972657617534576,
+      "grad_norm": 2.875,
+      "learning_rate": 4.884802715891307e-05,
+      "loss": 0.8529,
+      "step": 5626
+    },
+    {
+      "epoch": 0.09976202821095734,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.8847189001838324e-05,
+      "loss": 0.8529,
+      "step": 5628
+    },
+    {
+      "epoch": 0.09979748024656891,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.884635054715562e-05,
+      "loss": 0.8271,
+      "step": 5630
+    },
+    {
+      "epoch": 0.09983293228218047,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.884551179487543e-05,
+      "loss": 0.8768,
+      "step": 5632
+    },
+    {
+      "epoch": 0.09986838431779205,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.884467274500822e-05,
+      "loss": 0.8548,
+      "step": 5634
+    },
+    {
+      "epoch": 0.09990383635340362,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.884383339756447e-05,
+      "loss": 0.8354,
+      "step": 5636
+    },
+    {
+      "epoch": 0.09993928838901518,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.8842993752554635e-05,
+      "loss": 0.8216,
+      "step": 5638
+    },
+    {
+      "epoch": 0.09997474042462676,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.884215380998921e-05,
+      "loss": 0.8471,
+      "step": 5640
+    },
+    {
+      "epoch": 0.10001019246023833,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.884131356987867e-05,
+      "loss": 0.8059,
+      "step": 5642
+    },
+    {
+      "epoch": 0.10004564449584989,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.8840473032233494e-05,
+      "loss": 0.8661,
+      "step": 5644
+    },
+    {
+      "epoch": 0.10008109653146147,
+      "grad_norm": 2.75,
+      "learning_rate": 4.883963219706419e-05,
+      "loss": 0.7908,
+      "step": 5646
+    },
+    {
+      "epoch": 0.10011654856707304,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.883879106438124e-05,
+      "loss": 0.8687,
+      "step": 5648
+    },
+    {
+      "epoch": 0.1001520006026846,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.883794963419514e-05,
+      "loss": 0.8066,
+      "step": 5650
+    },
+    {
+      "epoch": 0.10018745263829618,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.88371079065164e-05,
+      "loss": 0.827,
+      "step": 5652
+    },
+    {
+      "epoch": 0.10022290467390775,
+      "grad_norm": 2.625,
+      "learning_rate": 4.883626588135551e-05,
+      "loss": 0.8052,
+      "step": 5654
+    },
+    {
+      "epoch": 0.10025835670951931,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.883542355872299e-05,
+      "loss": 0.8144,
+      "step": 5656
+    },
+    {
+      "epoch": 0.10029380874513089,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.883458093862935e-05,
+      "loss": 0.8118,
+      "step": 5658
+    },
+    {
+      "epoch": 0.10032926078074246,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.88337380210851e-05,
+      "loss": 0.8369,
+      "step": 5660
+    },
+    {
+      "epoch": 0.10036471281635402,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.883289480610077e-05,
+      "loss": 0.8589,
+      "step": 5662
+    },
+    {
+      "epoch": 0.1004001648519656,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.883205129368688e-05,
+      "loss": 0.8265,
+      "step": 5664
+    },
+    {
+      "epoch": 0.10043561688757717,
+      "grad_norm": 3.125,
+      "learning_rate": 4.883120748385394e-05,
+      "loss": 0.8489,
+      "step": 5666
+    },
+    {
+      "epoch": 0.10047106892318873,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.8830363376612497e-05,
+      "loss": 0.8463,
+      "step": 5668
+    },
+    {
+      "epoch": 0.1005065209588003,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.8829518971973085e-05,
+      "loss": 0.8248,
+      "step": 5670
+    },
+    {
+      "epoch": 0.10054197299441188,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.8828674269946254e-05,
+      "loss": 0.8391,
+      "step": 5672
+    },
+    {
+      "epoch": 0.10057742503002344,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.882782927054251e-05,
+      "loss": 0.8326,
+      "step": 5674
+    },
+    {
+      "epoch": 0.10061287706563501,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.882698397377243e-05,
+      "loss": 0.8218,
+      "step": 5676
+    },
+    {
+      "epoch": 0.10064832910124658,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.882613837964655e-05,
+      "loss": 0.8307,
+      "step": 5678
+    },
+    {
+      "epoch": 0.10068378113685815,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.882529248817543e-05,
+      "loss": 0.8427,
+      "step": 5680
+    },
+    {
+      "epoch": 0.10071923317246972,
+      "grad_norm": 2.875,
+      "learning_rate": 4.882444629936962e-05,
+      "loss": 0.8637,
+      "step": 5682
+    },
+    {
+      "epoch": 0.10075468520808128,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.882359981323968e-05,
+      "loss": 0.8579,
+      "step": 5684
+    },
+    {
+      "epoch": 0.10079013724369286,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.8822753029796174e-05,
+      "loss": 0.8361,
+      "step": 5686
+    },
+    {
+      "epoch": 0.10082558927930443,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.8821905949049685e-05,
+      "loss": 0.8359,
+      "step": 5688
+    },
+    {
+      "epoch": 0.100861041314916,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.882105857101076e-05,
+      "loss": 0.8288,
+      "step": 5690
+    },
+    {
+      "epoch": 0.10089649335052757,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.882021089568999e-05,
+      "loss": 0.8332,
+      "step": 5692
+    },
+    {
+      "epoch": 0.10093194538613914,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.881936292309795e-05,
+      "loss": 0.8297,
+      "step": 5694
+    },
+    {
+      "epoch": 0.1009673974217507,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.881851465324522e-05,
+      "loss": 0.8713,
+      "step": 5696
+    },
+    {
+      "epoch": 0.10100284945736228,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.881766608614238e-05,
+      "loss": 0.8066,
+      "step": 5698
+    },
+    {
+      "epoch": 0.10103830149297385,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.881681722180004e-05,
+      "loss": 0.8149,
+      "step": 5700
+    },
+    {
+      "epoch": 0.10107375352858541,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.881596806022878e-05,
+      "loss": 0.7993,
+      "step": 5702
+    },
+    {
+      "epoch": 0.10110920556419699,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.881511860143919e-05,
+      "loss": 0.7886,
+      "step": 5704
+    },
+    {
+      "epoch": 0.10114465759980856,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.881426884544189e-05,
+      "loss": 0.8257,
+      "step": 5706
+    },
+    {
+      "epoch": 0.10118010963542012,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.881341879224747e-05,
+      "loss": 0.8037,
+      "step": 5708
+    },
+    {
+      "epoch": 0.1012155616710317,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.881256844186655e-05,
+      "loss": 0.8028,
+      "step": 5710
+    },
+    {
+      "epoch": 0.10125101370664327,
+      "grad_norm": 3.25,
+      "learning_rate": 4.881171779430973e-05,
+      "loss": 0.8512,
+      "step": 5712
+    },
+    {
+      "epoch": 0.10128646574225483,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.881086684958763e-05,
+      "loss": 0.7972,
+      "step": 5714
+    },
+    {
+      "epoch": 0.10132191777786641,
+      "grad_norm": 3.25,
+      "learning_rate": 4.881001560771087e-05,
+      "loss": 0.8728,
+      "step": 5716
+    },
+    {
+      "epoch": 0.10135736981347798,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.8809164068690084e-05,
+      "loss": 0.8728,
+      "step": 5718
+    },
+    {
+      "epoch": 0.10139282184908954,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.880831223253588e-05,
+      "loss": 0.8824,
+      "step": 5720
+    },
+    {
+      "epoch": 0.10142827388470112,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.8807460099258906e-05,
+      "loss": 0.8078,
+      "step": 5722
+    },
+    {
+      "epoch": 0.10146372592031269,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.8806607668869786e-05,
+      "loss": 0.8533,
+      "step": 5724
+    },
+    {
+      "epoch": 0.10149917795592425,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.880575494137916e-05,
+      "loss": 0.8264,
+      "step": 5726
+    },
+    {
+      "epoch": 0.10153462999153583,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.880490191679767e-05,
+      "loss": 0.8368,
+      "step": 5728
+    },
+    {
+      "epoch": 0.1015700820271474,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.880404859513596e-05,
+      "loss": 0.8025,
+      "step": 5730
+    },
+    {
+      "epoch": 0.10160553406275896,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.8803194976404685e-05,
+      "loss": 0.7964,
+      "step": 5732
+    },
+    {
+      "epoch": 0.10164098609837054,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.8802341060614495e-05,
+      "loss": 0.8358,
+      "step": 5734
+    },
+    {
+      "epoch": 0.10167643813398211,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.8801486847776044e-05,
+      "loss": 0.8685,
+      "step": 5736
+    },
+    {
+      "epoch": 0.10171189016959367,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.880063233789999e-05,
+      "loss": 0.8414,
+      "step": 5738
+    },
+    {
+      "epoch": 0.10174734220520525,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.8799777530997017e-05,
+      "loss": 0.8109,
+      "step": 5740
+    },
+    {
+      "epoch": 0.10178279424081682,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.8798922427077764e-05,
+      "loss": 0.8377,
+      "step": 5742
+    },
+    {
+      "epoch": 0.10181824627642838,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.8798067026152914e-05,
+      "loss": 0.8486,
+      "step": 5744
+    },
+    {
+      "epoch": 0.10185369831203996,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.879721132823315e-05,
+      "loss": 0.8192,
+      "step": 5746
+    },
+    {
+      "epoch": 0.10188915034765153,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.8796355333329145e-05,
+      "loss": 0.8028,
+      "step": 5748
+    },
+    {
+      "epoch": 0.10192460238326309,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.8795499041451585e-05,
+      "loss": 0.8343,
+      "step": 5750
+    },
+    {
+      "epoch": 0.10196005441887467,
+      "grad_norm": 2.875,
+      "learning_rate": 4.879464245261115e-05,
+      "loss": 0.8285,
+      "step": 5752
+    },
+    {
+      "epoch": 0.10199550645448624,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.879378556681854e-05,
+      "loss": 0.8591,
+      "step": 5754
+    },
+    {
+      "epoch": 0.1020309584900978,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.8792928384084435e-05,
+      "loss": 0.8476,
+      "step": 5756
+    },
+    {
+      "epoch": 0.10206641052570938,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.879207090441954e-05,
+      "loss": 0.8165,
+      "step": 5758
+    },
+    {
+      "epoch": 0.10210186256132095,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.8791213127834555e-05,
+      "loss": 0.8621,
+      "step": 5760
+    },
+    {
+      "epoch": 0.10213731459693251,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.879035505434019e-05,
+      "loss": 0.8944,
+      "step": 5762
+    },
+    {
+      "epoch": 0.10217276663254408,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.878949668394714e-05,
+      "loss": 0.815,
+      "step": 5764
+    },
+    {
+      "epoch": 0.10220821866815565,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.878863801666613e-05,
+      "loss": 0.8699,
+      "step": 5766
+    },
+    {
+      "epoch": 0.10224367070376722,
+      "grad_norm": 2.625,
+      "learning_rate": 4.8787779052507874e-05,
+      "loss": 0.8312,
+      "step": 5768
+    },
+    {
+      "epoch": 0.1022791227393788,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.8786919791483094e-05,
+      "loss": 0.8142,
+      "step": 5770
+    },
+    {
+      "epoch": 0.10231457477499036,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.8786060233602506e-05,
+      "loss": 0.8463,
+      "step": 5772
+    },
+    {
+      "epoch": 0.10235002681060193,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.8785200378876836e-05,
+      "loss": 0.8769,
+      "step": 5774
+    },
+    {
+      "epoch": 0.1023854788462135,
+      "grad_norm": 2.875,
+      "learning_rate": 4.8784340227316825e-05,
+      "loss": 0.8068,
+      "step": 5776
+    },
+    {
+      "epoch": 0.10242093088182506,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.8783479778933207e-05,
+      "loss": 0.8446,
+      "step": 5778
+    },
+    {
+      "epoch": 0.10245638291743664,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.878261903373671e-05,
+      "loss": 0.8235,
+      "step": 5780
+    },
+    {
+      "epoch": 0.10249183495304821,
+      "grad_norm": 2.75,
+      "learning_rate": 4.8781757991738074e-05,
+      "loss": 0.8487,
+      "step": 5782
+    },
+    {
+      "epoch": 0.10252728698865977,
+      "grad_norm": 2.875,
+      "learning_rate": 4.8780896652948054e-05,
+      "loss": 0.8151,
+      "step": 5784
+    },
+    {
+      "epoch": 0.10256273902427135,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.87800350173774e-05,
+      "loss": 0.8187,
+      "step": 5786
+    },
+    {
+      "epoch": 0.10259819105988292,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.8779173085036865e-05,
+      "loss": 0.8441,
+      "step": 5788
+    },
+    {
+      "epoch": 0.10263364309549448,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.87783108559372e-05,
+      "loss": 0.8383,
+      "step": 5790
+    },
+    {
+      "epoch": 0.10266909513110606,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.877744833008917e-05,
+      "loss": 0.8315,
+      "step": 5792
+    },
+    {
+      "epoch": 0.10270454716671763,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.877658550750353e-05,
+      "loss": 0.8346,
+      "step": 5794
+    },
+    {
+      "epoch": 0.1027399992023292,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.877572238819106e-05,
+      "loss": 0.8366,
+      "step": 5796
+    },
+    {
+      "epoch": 0.10277545123794077,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.877485897216253e-05,
+      "loss": 0.8602,
+      "step": 5798
+    },
+    {
+      "epoch": 0.10281090327355234,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.877399525942871e-05,
+      "loss": 0.8087,
+      "step": 5800
+    },
+    {
+      "epoch": 0.1028463553091639,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.877313125000038e-05,
+      "loss": 0.8278,
+      "step": 5802
+    },
+    {
+      "epoch": 0.10288180734477548,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.877226694388832e-05,
+      "loss": 0.8435,
+      "step": 5804
+    },
+    {
+      "epoch": 0.10291725938038705,
+      "grad_norm": 3.0,
+      "learning_rate": 4.877140234110333e-05,
+      "loss": 0.8759,
+      "step": 5806
+    },
+    {
+      "epoch": 0.10295271141599861,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.8770537441656184e-05,
+      "loss": 0.8405,
+      "step": 5808
+    },
+    {
+      "epoch": 0.10298816345161019,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.876967224555768e-05,
+      "loss": 0.8133,
+      "step": 5810
+    },
+    {
+      "epoch": 0.10302361548722176,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.876880675281862e-05,
+      "loss": 0.8224,
+      "step": 5812
+    },
+    {
+      "epoch": 0.10305906752283332,
+      "grad_norm": 2.625,
+      "learning_rate": 4.87679409634498e-05,
+      "loss": 0.8069,
+      "step": 5814
+    },
+    {
+      "epoch": 0.1030945195584449,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.876707487746203e-05,
+      "loss": 0.8132,
+      "step": 5816
+    },
+    {
+      "epoch": 0.10312997159405647,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.8766208494866114e-05,
+      "loss": 0.8216,
+      "step": 5818
+    },
+    {
+      "epoch": 0.10316542362966803,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.8765341815672865e-05,
+      "loss": 0.8498,
+      "step": 5820
+    },
+    {
+      "epoch": 0.1032008756652796,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.87644748398931e-05,
+      "loss": 0.8971,
+      "step": 5822
+    },
+    {
+      "epoch": 0.10323632770089118,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.876360756753764e-05,
+      "loss": 0.8124,
+      "step": 5824
+    },
+    {
+      "epoch": 0.10327177973650274,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.876273999861731e-05,
+      "loss": 0.8198,
+      "step": 5826
+    },
+    {
+      "epoch": 0.10330723177211432,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.8761872133142935e-05,
+      "loss": 0.8254,
+      "step": 5828
+    },
+    {
+      "epoch": 0.10334268380772589,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.876100397112534e-05,
+      "loss": 0.8737,
+      "step": 5830
+    },
+    {
+      "epoch": 0.10337813584333745,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.8760135512575364e-05,
+      "loss": 0.8547,
+      "step": 5832
+    },
+    {
+      "epoch": 0.10341358787894903,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.8759266757503855e-05,
+      "loss": 0.8565,
+      "step": 5834
+    },
+    {
+      "epoch": 0.1034490399145606,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.875839770592163e-05,
+      "loss": 0.8471,
+      "step": 5836
+    },
+    {
+      "epoch": 0.10348449195017216,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.8757528357839564e-05,
+      "loss": 0.8266,
+      "step": 5838
+    },
+    {
+      "epoch": 0.10351994398578374,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.8756658713268486e-05,
+      "loss": 0.8161,
+      "step": 5840
+    },
+    {
+      "epoch": 0.10355539602139531,
+      "grad_norm": 2.5,
+      "learning_rate": 4.8755788772219256e-05,
+      "loss": 0.8541,
+      "step": 5842
+    },
+    {
+      "epoch": 0.10359084805700687,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.8754918534702733e-05,
+      "loss": 0.8992,
+      "step": 5844
+    },
+    {
+      "epoch": 0.10362630009261845,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.875404800072977e-05,
+      "loss": 0.8513,
+      "step": 5846
+    },
+    {
+      "epoch": 0.10366175212823002,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.875317717031124e-05,
+      "loss": 0.8752,
+      "step": 5848
+    },
+    {
+      "epoch": 0.10369720416384158,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.8752306043458e-05,
+      "loss": 0.8375,
+      "step": 5850
+    },
+    {
+      "epoch": 0.10373265619945315,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.875143462018094e-05,
+      "loss": 0.8408,
+      "step": 5852
+    },
+    {
+      "epoch": 0.10376810823506472,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.875056290049091e-05,
+      "loss": 0.8534,
+      "step": 5854
+    },
+    {
+      "epoch": 0.10380356027067629,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.87496908843988e-05,
+      "loss": 0.7775,
+      "step": 5856
+    },
+    {
+      "epoch": 0.10383901230628786,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.874881857191551e-05,
+      "loss": 0.8615,
+      "step": 5858
+    },
+    {
+      "epoch": 0.10387446434189943,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.87479459630519e-05,
+      "loss": 0.8127,
+      "step": 5860
+    },
+    {
+      "epoch": 0.103909916377511,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.874707305781887e-05,
+      "loss": 0.8257,
+      "step": 5862
+    },
+    {
+      "epoch": 0.10394536841312257,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.8746199856227315e-05,
+      "loss": 0.772,
+      "step": 5864
+    },
+    {
+      "epoch": 0.10398082044873413,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.8745326358288133e-05,
+      "loss": 0.8003,
+      "step": 5866
+    },
+    {
+      "epoch": 0.10401627248434571,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.874445256401223e-05,
+      "loss": 0.8276,
+      "step": 5868
+    },
+    {
+      "epoch": 0.10405172451995728,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.87435784734105e-05,
+      "loss": 0.8251,
+      "step": 5870
+    },
+    {
+      "epoch": 0.10408717655556884,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.874270408649385e-05,
+      "loss": 0.7982,
+      "step": 5872
+    },
+    {
+      "epoch": 0.10412262859118042,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.874182940327321e-05,
+      "loss": 0.8599,
+      "step": 5874
+    },
+    {
+      "epoch": 0.104158080626792,
+      "grad_norm": 2.75,
+      "learning_rate": 4.874095442375948e-05,
+      "loss": 0.8527,
+      "step": 5876
+    },
+    {
+      "epoch": 0.10419353266240355,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.874007914796358e-05,
+      "loss": 0.8557,
+      "step": 5878
+    },
+    {
+      "epoch": 0.10422898469801513,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.873920357589644e-05,
+      "loss": 0.8352,
+      "step": 5880
+    },
+    {
+      "epoch": 0.1042644367336267,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.8738327707568974e-05,
+      "loss": 0.8541,
+      "step": 5882
+    },
+    {
+      "epoch": 0.10429988876923826,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.8737451542992136e-05,
+      "loss": 0.8092,
+      "step": 5884
+    },
+    {
+      "epoch": 0.10433534080484984,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.873657508217684e-05,
+      "loss": 0.8368,
+      "step": 5886
+    },
+    {
+      "epoch": 0.10437079284046141,
+      "grad_norm": 2.625,
+      "learning_rate": 4.873569832513403e-05,
+      "loss": 0.8305,
+      "step": 5888
+    },
+    {
+      "epoch": 0.10440624487607297,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.8734821271874656e-05,
+      "loss": 0.833,
+      "step": 5890
+    },
+    {
+      "epoch": 0.10444169691168455,
+      "grad_norm": 2.875,
+      "learning_rate": 4.873394392240965e-05,
+      "loss": 0.8522,
+      "step": 5892
+    },
+    {
+      "epoch": 0.10447714894729612,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.873306627674997e-05,
+      "loss": 0.8828,
+      "step": 5894
+    },
+    {
+      "epoch": 0.10451260098290768,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.873218833490656e-05,
+      "loss": 0.817,
+      "step": 5896
+    },
+    {
+      "epoch": 0.10454805301851926,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.873131009689039e-05,
+      "loss": 0.82,
+      "step": 5898
+    },
+    {
+      "epoch": 0.10458350505413083,
+      "grad_norm": 2.5,
+      "learning_rate": 4.873043156271241e-05,
+      "loss": 0.8608,
+      "step": 5900
+    },
+    {
+      "epoch": 0.10461895708974239,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.8729552732383586e-05,
+      "loss": 0.8599,
+      "step": 5902
+    },
+    {
+      "epoch": 0.10465440912535397,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.872867360591489e-05,
+      "loss": 0.8522,
+      "step": 5904
+    },
+    {
+      "epoch": 0.10468986116096554,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.872779418331729e-05,
+      "loss": 0.8401,
+      "step": 5906
+    },
+    {
+      "epoch": 0.1047253131965771,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.872691446460176e-05,
+      "loss": 0.8433,
+      "step": 5908
+    },
+    {
+      "epoch": 0.10476076523218868,
+      "grad_norm": 3.0,
+      "learning_rate": 4.872603444977927e-05,
+      "loss": 0.7944,
+      "step": 5910
+    },
+    {
+      "epoch": 0.10479621726780025,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.8725154138860826e-05,
+      "loss": 0.839,
+      "step": 5912
+    },
+    {
+      "epoch": 0.10483166930341181,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.872427353185739e-05,
+      "loss": 0.8284,
+      "step": 5914
+    },
+    {
+      "epoch": 0.10486712133902339,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.872339262877996e-05,
+      "loss": 0.8338,
+      "step": 5916
+    },
+    {
+      "epoch": 0.10490257337463496,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.8722511429639536e-05,
+      "loss": 0.826,
+      "step": 5918
+    },
+    {
+      "epoch": 0.10493802541024652,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.872162993444712e-05,
+      "loss": 0.8543,
+      "step": 5920
+    },
+    {
+      "epoch": 0.1049734774458581,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.872074814321369e-05,
+      "loss": 0.8229,
+      "step": 5922
+    },
+    {
+      "epoch": 0.10500892948146967,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.871986605595027e-05,
+      "loss": 0.7872,
+      "step": 5924
+    },
+    {
+      "epoch": 0.10504438151708123,
+      "grad_norm": 2.5,
+      "learning_rate": 4.871898367266785e-05,
+      "loss": 0.7904,
+      "step": 5926
+    },
+    {
+      "epoch": 0.1050798335526928,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.871810099337747e-05,
+      "loss": 0.8301,
+      "step": 5928
+    },
+    {
+      "epoch": 0.10511528558830438,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.871721801809013e-05,
+      "loss": 0.8046,
+      "step": 5930
+    },
+    {
+      "epoch": 0.10515073762391594,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.871633474681684e-05,
+      "loss": 0.8217,
+      "step": 5932
+    },
+    {
+      "epoch": 0.10518618965952752,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.871545117956863e-05,
+      "loss": 0.8079,
+      "step": 5934
+    },
+    {
+      "epoch": 0.10522164169513908,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.871456731635653e-05,
+      "loss": 0.8503,
+      "step": 5936
+    },
+    {
+      "epoch": 0.10525709373075065,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.871368315719158e-05,
+      "loss": 0.8112,
+      "step": 5938
+    },
+    {
+      "epoch": 0.10529254576636222,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.871279870208479e-05,
+      "loss": 0.8458,
+      "step": 5940
+    },
+    {
+      "epoch": 0.10532799780197379,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.8711913951047224e-05,
+      "loss": 0.8309,
+      "step": 5942
+    },
+    {
+      "epoch": 0.10536344983758536,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.8711028904089905e-05,
+      "loss": 0.775,
+      "step": 5944
+    },
+    {
+      "epoch": 0.10539890187319693,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.871014356122388e-05,
+      "loss": 0.8295,
+      "step": 5946
+    },
+    {
+      "epoch": 0.1054343539088085,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.870925792246021e-05,
+      "loss": 0.8225,
+      "step": 5948
+    },
+    {
+      "epoch": 0.10546980594442007,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.870837198780993e-05,
+      "loss": 0.8658,
+      "step": 5950
+    },
+    {
+      "epoch": 0.10550525798003164,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.870748575728411e-05,
+      "loss": 0.8338,
+      "step": 5952
+    },
+    {
+      "epoch": 0.1055407100156432,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.8706599230893805e-05,
+      "loss": 0.8351,
+      "step": 5954
+    },
+    {
+      "epoch": 0.10557616205125478,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.8705712408650086e-05,
+      "loss": 0.8041,
+      "step": 5956
+    },
+    {
+      "epoch": 0.10561161408686635,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.8704825290564004e-05,
+      "loss": 0.8343,
+      "step": 5958
+    },
+    {
+      "epoch": 0.10564706612247791,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.870393787664664e-05,
+      "loss": 0.8447,
+      "step": 5960
+    },
+    {
+      "epoch": 0.10568251815808949,
+      "grad_norm": 2.625,
+      "learning_rate": 4.870305016690908e-05,
+      "loss": 0.852,
+      "step": 5962
+    },
+    {
+      "epoch": 0.10571797019370106,
+      "grad_norm": 2.875,
+      "learning_rate": 4.870216216136238e-05,
+      "loss": 0.8384,
+      "step": 5964
+    },
+    {
+      "epoch": 0.10575342222931262,
+      "grad_norm": 2.875,
+      "learning_rate": 4.870127386001764e-05,
+      "loss": 0.8305,
+      "step": 5966
+    },
+    {
+      "epoch": 0.1057888742649242,
+      "grad_norm": 2.625,
+      "learning_rate": 4.870038526288593e-05,
+      "loss": 0.8172,
+      "step": 5968
+    },
+    {
+      "epoch": 0.10582432630053577,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.8699496369978346e-05,
+      "loss": 0.8577,
+      "step": 5970
+    },
+    {
+      "epoch": 0.10585977833614733,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.8698607181306e-05,
+      "loss": 0.8757,
+      "step": 5972
+    },
+    {
+      "epoch": 0.10589523037175891,
+      "grad_norm": 2.75,
+      "learning_rate": 4.8697717696879965e-05,
+      "loss": 0.8196,
+      "step": 5974
+    },
+    {
+      "epoch": 0.10593068240737048,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.869682791671134e-05,
+      "loss": 0.84,
+      "step": 5976
+    },
+    {
+      "epoch": 0.10596613444298204,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.869593784081124e-05,
+      "loss": 0.82,
+      "step": 5978
+    },
+    {
+      "epoch": 0.10600158647859362,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.869504746919078e-05,
+      "loss": 0.8291,
+      "step": 5980
+    },
+    {
+      "epoch": 0.10603703851420519,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.8694156801861065e-05,
+      "loss": 0.7903,
+      "step": 5982
+    },
+    {
+      "epoch": 0.10607249054981675,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.86932658388332e-05,
+      "loss": 0.7952,
+      "step": 5984
+    },
+    {
+      "epoch": 0.10610794258542833,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.8692374580118314e-05,
+      "loss": 0.8418,
+      "step": 5986
+    },
+    {
+      "epoch": 0.1061433946210399,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.869148302572753e-05,
+      "loss": 0.8257,
+      "step": 5988
+    },
+    {
+      "epoch": 0.10617884665665146,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.869059117567198e-05,
+      "loss": 0.8086,
+      "step": 5990
+    },
+    {
+      "epoch": 0.10621429869226304,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.8689699029962774e-05,
+      "loss": 0.8294,
+      "step": 5992
+    },
+    {
+      "epoch": 0.10624975072787461,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.868880658861106e-05,
+      "loss": 0.8258,
+      "step": 5994
+    },
+    {
+      "epoch": 0.10628520276348617,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.868791385162797e-05,
+      "loss": 0.8312,
+      "step": 5996
+    },
+    {
+      "epoch": 0.10632065479909775,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.868702081902466e-05,
+      "loss": 0.8579,
+      "step": 5998
+    },
+    {
+      "epoch": 0.10635610683470932,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.868612749081226e-05,
+      "loss": 0.8733,
+      "step": 6000
+    },
+    {
+      "epoch": 0.10639155887032088,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.868523386700192e-05,
+      "loss": 0.8503,
+      "step": 6002
+    },
+    {
+      "epoch": 0.10642701090593246,
+      "grad_norm": 2.5,
+      "learning_rate": 4.86843399476048e-05,
+      "loss": 0.7982,
+      "step": 6004
+    },
+    {
+      "epoch": 0.10646246294154403,
+      "grad_norm": 3.0,
+      "learning_rate": 4.8683445732632046e-05,
+      "loss": 0.8144,
+      "step": 6006
+    },
+    {
+      "epoch": 0.10649791497715559,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.868255122209482e-05,
+      "loss": 0.8306,
+      "step": 6008
+    },
+    {
+      "epoch": 0.10653336701276717,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.868165641600429e-05,
+      "loss": 0.7931,
+      "step": 6010
+    },
+    {
+      "epoch": 0.10656881904837874,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.868076131437162e-05,
+      "loss": 0.8033,
+      "step": 6012
+    },
+    {
+      "epoch": 0.1066042710839903,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.8679865917207986e-05,
+      "loss": 0.805,
+      "step": 6014
+    },
+    {
+      "epoch": 0.10663972311960188,
+      "grad_norm": 2.625,
+      "learning_rate": 4.867897022452455e-05,
+      "loss": 0.8566,
+      "step": 6016
+    },
+    {
+      "epoch": 0.10667517515521345,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.8678074236332505e-05,
+      "loss": 0.8544,
+      "step": 6018
+    },
+    {
+      "epoch": 0.10671062719082501,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.867717795264301e-05,
+      "loss": 0.7666,
+      "step": 6020
+    },
+    {
+      "epoch": 0.10674607922643659,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.8676281373467284e-05,
+      "loss": 0.8344,
+      "step": 6022
+    },
+    {
+      "epoch": 0.10678153126204815,
+      "grad_norm": 3.0,
+      "learning_rate": 4.867538449881649e-05,
+      "loss": 0.8397,
+      "step": 6024
+    },
+    {
+      "epoch": 0.10681698329765972,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.8674487328701836e-05,
+      "loss": 0.8351,
+      "step": 6026
+    },
+    {
+      "epoch": 0.1068524353332713,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.86735898631345e-05,
+      "loss": 0.8393,
+      "step": 6028
+    },
+    {
+      "epoch": 0.10688788736888286,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.8672692102125696e-05,
+      "loss": 0.8348,
+      "step": 6030
+    },
+    {
+      "epoch": 0.10692333940449443,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.867179404568663e-05,
+      "loss": 0.8175,
+      "step": 6032
+    },
+    {
+      "epoch": 0.106958791440106,
+      "grad_norm": 2.5,
+      "learning_rate": 4.86708956938285e-05,
+      "loss": 0.8695,
+      "step": 6034
+    },
+    {
+      "epoch": 0.10699424347571757,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.8669997046562524e-05,
+      "loss": 0.8068,
+      "step": 6036
+    },
+    {
+      "epoch": 0.10702969551132914,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.866909810389991e-05,
+      "loss": 0.8816,
+      "step": 6038
+    },
+    {
+      "epoch": 0.10706514754694071,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.866819886585189e-05,
+      "loss": 0.8582,
+      "step": 6040
+    },
+    {
+      "epoch": 0.10710059958255227,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.866729933242968e-05,
+      "loss": 0.7931,
+      "step": 6042
+    },
+    {
+      "epoch": 0.10713605161816385,
+      "grad_norm": 3.0,
+      "learning_rate": 4.866639950364449e-05,
+      "loss": 0.8141,
+      "step": 6044
+    },
+    {
+      "epoch": 0.10717150365377542,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.866549937950757e-05,
+      "loss": 0.8202,
+      "step": 6046
+    },
+    {
+      "epoch": 0.10720695568938698,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.8664598960030154e-05,
+      "loss": 0.8198,
+      "step": 6048
+    },
+    {
+      "epoch": 0.10724240772499856,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.8663698245223466e-05,
+      "loss": 0.7826,
+      "step": 6050
+    },
+    {
+      "epoch": 0.10727785976061013,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.8662797235098754e-05,
+      "loss": 0.8455,
+      "step": 6052
+    },
+    {
+      "epoch": 0.1073133117962217,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.866189592966726e-05,
+      "loss": 0.8529,
+      "step": 6054
+    },
+    {
+      "epoch": 0.10734876383183327,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.8660994328940235e-05,
+      "loss": 0.809,
+      "step": 6056
+    },
+    {
+      "epoch": 0.10738421586744484,
+      "grad_norm": 2.625,
+      "learning_rate": 4.866009243292893e-05,
+      "loss": 0.8287,
+      "step": 6058
+    },
+    {
+      "epoch": 0.1074196679030564,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.86591902416446e-05,
+      "loss": 0.8295,
+      "step": 6060
+    },
+    {
+      "epoch": 0.10745511993866798,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.86582877550985e-05,
+      "loss": 0.8268,
+      "step": 6062
+    },
+    {
+      "epoch": 0.10749057197427955,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.865738497330189e-05,
+      "loss": 0.8194,
+      "step": 6064
+    },
+    {
+      "epoch": 0.10752602400989111,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.865648189626605e-05,
+      "loss": 0.8585,
+      "step": 6066
+    },
+    {
+      "epoch": 0.10756147604550269,
+      "grad_norm": 2.75,
+      "learning_rate": 4.865557852400225e-05,
+      "loss": 0.8076,
+      "step": 6068
+    },
+    {
+      "epoch": 0.10759692808111426,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.8654674856521745e-05,
+      "loss": 0.7988,
+      "step": 6070
+    },
+    {
+      "epoch": 0.10763238011672582,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.865377089383584e-05,
+      "loss": 0.8241,
+      "step": 6072
+    },
+    {
+      "epoch": 0.1076678321523374,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.865286663595578e-05,
+      "loss": 0.8282,
+      "step": 6074
+    },
+    {
+      "epoch": 0.10770328418794897,
+      "grad_norm": 2.375,
+      "learning_rate": 4.865196208289289e-05,
+      "loss": 0.8366,
+      "step": 6076
+    },
+    {
+      "epoch": 0.10773873622356053,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.865105723465843e-05,
+      "loss": 0.8635,
+      "step": 6078
+    },
+    {
+      "epoch": 0.10777418825917211,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.86501520912637e-05,
+      "loss": 0.7817,
+      "step": 6080
+    },
+    {
+      "epoch": 0.10780964029478368,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.8649246652720005e-05,
+      "loss": 0.8349,
+      "step": 6082
+    },
+    {
+      "epoch": 0.10784509233039524,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.864834091903864e-05,
+      "loss": 0.8655,
+      "step": 6084
+    },
+    {
+      "epoch": 0.10788054436600682,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.864743489023089e-05,
+      "loss": 0.7774,
+      "step": 6086
+    },
+    {
+      "epoch": 0.10791599640161839,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.8646528566308094e-05,
+      "loss": 0.8279,
+      "step": 6088
+    },
+    {
+      "epoch": 0.10795144843722995,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.864562194728154e-05,
+      "loss": 0.8073,
+      "step": 6090
+    },
+    {
+      "epoch": 0.10798690047284153,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.8644715033162546e-05,
+      "loss": 0.8185,
+      "step": 6092
+    },
+    {
+      "epoch": 0.1080223525084531,
+      "grad_norm": 2.75,
+      "learning_rate": 4.864380782396244e-05,
+      "loss": 0.8483,
+      "step": 6094
+    },
+    {
+      "epoch": 0.10805780454406466,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.8642900319692536e-05,
+      "loss": 0.8436,
+      "step": 6096
+    },
+    {
+      "epoch": 0.10809325657967624,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.864199252036415e-05,
+      "loss": 0.8808,
+      "step": 6098
+    },
+    {
+      "epoch": 0.10812870861528781,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.864108442598864e-05,
+      "loss": 0.8295,
+      "step": 6100
+    },
+    {
+      "epoch": 0.10816416065089937,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.864017603657731e-05,
+      "loss": 0.839,
+      "step": 6102
+    },
+    {
+      "epoch": 0.10819961268651095,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.86392673521415e-05,
+      "loss": 0.8202,
+      "step": 6104
+    },
+    {
+      "epoch": 0.1082350647221225,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.863835837269257e-05,
+      "loss": 0.8195,
+      "step": 6106
+    },
+    {
+      "epoch": 0.10827051675773408,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.863744909824185e-05,
+      "loss": 0.7983,
+      "step": 6108
+    },
+    {
+      "epoch": 0.10830596879334566,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.8636539528800685e-05,
+      "loss": 0.8431,
+      "step": 6110
+    },
+    {
+      "epoch": 0.10834142082895722,
+      "grad_norm": 2.625,
+      "learning_rate": 4.863562966438042e-05,
+      "loss": 0.8398,
+      "step": 6112
+    },
+    {
+      "epoch": 0.10837687286456879,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.863471950499243e-05,
+      "loss": 0.8235,
+      "step": 6114
+    },
+    {
+      "epoch": 0.10841232490018037,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.8633809050648064e-05,
+      "loss": 0.8191,
+      "step": 6116
+    },
+    {
+      "epoch": 0.10844777693579193,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.8632898301358684e-05,
+      "loss": 0.81,
+      "step": 6118
+    },
+    {
+      "epoch": 0.1084832289714035,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.863198725713565e-05,
+      "loss": 0.7965,
+      "step": 6120
+    },
+    {
+      "epoch": 0.10851868100701507,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.863107591799034e-05,
+      "loss": 0.805,
+      "step": 6122
+    },
+    {
+      "epoch": 0.10855413304262664,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.863016428393413e-05,
+      "loss": 0.8054,
+      "step": 6124
+    },
+    {
+      "epoch": 0.10858958507823821,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.862925235497839e-05,
+      "loss": 0.7906,
+      "step": 6126
+    },
+    {
+      "epoch": 0.10862503711384978,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.8628340131134496e-05,
+      "loss": 0.8641,
+      "step": 6128
+    },
+    {
+      "epoch": 0.10866048914946135,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.862742761241384e-05,
+      "loss": 0.8086,
+      "step": 6130
+    },
+    {
+      "epoch": 0.10869594118507292,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.862651479882782e-05,
+      "loss": 0.8464,
+      "step": 6132
+    },
+    {
+      "epoch": 0.1087313932206845,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.8625601690387804e-05,
+      "loss": 0.8145,
+      "step": 6134
+    },
+    {
+      "epoch": 0.10876684525629605,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.8624688287105195e-05,
+      "loss": 0.8007,
+      "step": 6136
+    },
+    {
+      "epoch": 0.10880229729190763,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.86237745889914e-05,
+      "loss": 0.8283,
+      "step": 6138
+    },
+    {
+      "epoch": 0.1088377493275192,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.8622860596057826e-05,
+      "loss": 0.7944,
+      "step": 6140
+    },
+    {
+      "epoch": 0.10887320136313076,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.862194630831587e-05,
+      "loss": 0.8282,
+      "step": 6142
+    },
+    {
+      "epoch": 0.10890865339874234,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.862103172577695e-05,
+      "loss": 0.8247,
+      "step": 6144
+    },
+    {
+      "epoch": 0.10894410543435391,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.862011684845246e-05,
+      "loss": 0.8667,
+      "step": 6146
+    },
+    {
+      "epoch": 0.10897955746996547,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.861920167635384e-05,
+      "loss": 0.8452,
+      "step": 6148
+    },
+    {
+      "epoch": 0.10901500950557705,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.86182862094925e-05,
+      "loss": 0.7993,
+      "step": 6150
+    },
+    {
+      "epoch": 0.10905046154118862,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.861737044787987e-05,
+      "loss": 0.8357,
+      "step": 6152
+    },
+    {
+      "epoch": 0.10908591357680018,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.861645439152738e-05,
+      "loss": 0.8498,
+      "step": 6154
+    },
+    {
+      "epoch": 0.10912136561241176,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.8615538040446446e-05,
+      "loss": 0.8273,
+      "step": 6156
+    },
+    {
+      "epoch": 0.10915681764802333,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.8614621394648525e-05,
+      "loss": 0.8296,
+      "step": 6158
+    },
+    {
+      "epoch": 0.1091922696836349,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.8613704454145045e-05,
+      "loss": 0.8243,
+      "step": 6160
+    },
+    {
+      "epoch": 0.10922772171924647,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.8612787218947454e-05,
+      "loss": 0.8467,
+      "step": 6162
+    },
+    {
+      "epoch": 0.10926317375485804,
+      "grad_norm": 2.625,
+      "learning_rate": 4.8611869689067194e-05,
+      "loss": 0.8519,
+      "step": 6164
+    },
+    {
+      "epoch": 0.1092986257904696,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.861095186451572e-05,
+      "loss": 0.8219,
+      "step": 6166
+    },
+    {
+      "epoch": 0.10933407782608118,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.861003374530448e-05,
+      "loss": 0.8044,
+      "step": 6168
+    },
+    {
+      "epoch": 0.10936952986169275,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.860911533144494e-05,
+      "loss": 0.7915,
+      "step": 6170
+    },
+    {
+      "epoch": 0.10940498189730431,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.8608196622948554e-05,
+      "loss": 0.8269,
+      "step": 6172
+    },
+    {
+      "epoch": 0.10944043393291589,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.8607277619826796e-05,
+      "loss": 0.838,
+      "step": 6174
+    },
+    {
+      "epoch": 0.10947588596852746,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.860635832209113e-05,
+      "loss": 0.803,
+      "step": 6176
+    },
+    {
+      "epoch": 0.10951133800413902,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.860543872975303e-05,
+      "loss": 0.8131,
+      "step": 6178
+    },
+    {
+      "epoch": 0.1095467900397506,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.8604518842823974e-05,
+      "loss": 0.8283,
+      "step": 6180
+    },
+    {
+      "epoch": 0.10958224207536217,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.860359866131543e-05,
+      "loss": 0.8426,
+      "step": 6182
+    },
+    {
+      "epoch": 0.10961769411097373,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.86026781852389e-05,
+      "loss": 0.8239,
+      "step": 6184
+    },
+    {
+      "epoch": 0.1096531461465853,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.860175741460585e-05,
+      "loss": 0.8929,
+      "step": 6186
+    },
+    {
+      "epoch": 0.10968859818219688,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.8600836349427805e-05,
+      "loss": 0.8083,
+      "step": 6188
+    },
+    {
+      "epoch": 0.10972405021780844,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.8599914989716223e-05,
+      "loss": 0.8372,
+      "step": 6190
+    },
+    {
+      "epoch": 0.10975950225342002,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.859899333548261e-05,
+      "loss": 0.8588,
+      "step": 6192
+    },
+    {
+      "epoch": 0.10979495428903158,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.8598071386738485e-05,
+      "loss": 0.7989,
+      "step": 6194
+    },
+    {
+      "epoch": 0.10983040632464315,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.859714914349535e-05,
+      "loss": 0.8067,
+      "step": 6196
+    },
+    {
+      "epoch": 0.10986585836025473,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.8596226605764704e-05,
+      "loss": 0.7671,
+      "step": 6198
+    },
+    {
+      "epoch": 0.10990131039586629,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.8595303773558064e-05,
+      "loss": 0.8118,
+      "step": 6200
+    },
+    {
+      "epoch": 0.10993676243147786,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.8594380646886945e-05,
+      "loss": 0.8167,
+      "step": 6202
+    },
+    {
+      "epoch": 0.10997221446708944,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.8593457225762873e-05,
+      "loss": 0.8439,
+      "step": 6204
+    },
+    {
+      "epoch": 0.110007666502701,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.859253351019737e-05,
+      "loss": 0.8135,
+      "step": 6206
+    },
+    {
+      "epoch": 0.11004311853831257,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.859160950020196e-05,
+      "loss": 0.8585,
+      "step": 6208
+    },
+    {
+      "epoch": 0.11007857057392414,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.859068519578818e-05,
+      "loss": 0.8055,
+      "step": 6210
+    },
+    {
+      "epoch": 0.1101140226095357,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.8589760596967555e-05,
+      "loss": 0.8373,
+      "step": 6212
+    },
+    {
+      "epoch": 0.11014947464514728,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.858883570375163e-05,
+      "loss": 0.805,
+      "step": 6214
+    },
+    {
+      "epoch": 0.11018492668075885,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.858791051615196e-05,
+      "loss": 0.8253,
+      "step": 6216
+    },
+    {
+      "epoch": 0.11022037871637042,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.8586985034180076e-05,
+      "loss": 0.8373,
+      "step": 6218
+    },
+    {
+      "epoch": 0.11025583075198199,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.858605925784753e-05,
+      "loss": 0.833,
+      "step": 6220
+    },
+    {
+      "epoch": 0.11029128278759356,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.8585133187165876e-05,
+      "loss": 0.8138,
+      "step": 6222
+    },
+    {
+      "epoch": 0.11032673482320512,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.858420682214667e-05,
+      "loss": 0.8439,
+      "step": 6224
+    },
+    {
+      "epoch": 0.1103621868588167,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.858328016280148e-05,
+      "loss": 0.8632,
+      "step": 6226
+    },
+    {
+      "epoch": 0.11039763889442827,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.858235320914187e-05,
+      "loss": 0.7924,
+      "step": 6228
+    },
+    {
+      "epoch": 0.11043309093003983,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.8581425961179396e-05,
+      "loss": 0.7989,
+      "step": 6230
+    },
+    {
+      "epoch": 0.11046854296565141,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.858049841892564e-05,
+      "loss": 0.796,
+      "step": 6232
+    },
+    {
+      "epoch": 0.11050399500126298,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.8579570582392176e-05,
+      "loss": 0.7959,
+      "step": 6234
+    },
+    {
+      "epoch": 0.11053944703687454,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.8578642451590585e-05,
+      "loss": 0.7785,
+      "step": 6236
+    },
+    {
+      "epoch": 0.11057489907248612,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.857771402653244e-05,
+      "loss": 0.824,
+      "step": 6238
+    },
+    {
+      "epoch": 0.1106103511080977,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.857678530722933e-05,
+      "loss": 0.7858,
+      "step": 6240
+    },
+    {
+      "epoch": 0.11064580314370925,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.857585629369287e-05,
+      "loss": 0.8481,
+      "step": 6242
+    },
+    {
+      "epoch": 0.11068125517932083,
+      "grad_norm": 2.625,
+      "learning_rate": 4.857492698593462e-05,
+      "loss": 0.8059,
+      "step": 6244
+    },
+    {
+      "epoch": 0.1107167072149324,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.85739973839662e-05,
+      "loss": 0.7678,
+      "step": 6246
+    },
+    {
+      "epoch": 0.11075215925054396,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.857306748779919e-05,
+      "loss": 0.814,
+      "step": 6248
+    },
+    {
+      "epoch": 0.11078761128615554,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.857213729744521e-05,
+      "loss": 0.8241,
+      "step": 6250
+    },
+    {
+      "epoch": 0.11082306332176711,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.857120681291587e-05,
+      "loss": 0.8479,
+      "step": 6252
+    },
+    {
+      "epoch": 0.11085851535737867,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.8570276034222787e-05,
+      "loss": 0.7965,
+      "step": 6254
+    },
+    {
+      "epoch": 0.11089396739299025,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.856934496137756e-05,
+      "loss": 0.8589,
+      "step": 6256
+    },
+    {
+      "epoch": 0.11092941942860182,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.8568413594391814e-05,
+      "loss": 0.7966,
+      "step": 6258
+    },
+    {
+      "epoch": 0.11096487146421338,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.856748193327718e-05,
+      "loss": 0.8463,
+      "step": 6260
+    },
+    {
+      "epoch": 0.11100032349982496,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.856654997804528e-05,
+      "loss": 0.8121,
+      "step": 6262
+    },
+    {
+      "epoch": 0.11103577553543653,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.8565617728707745e-05,
+      "loss": 0.838,
+      "step": 6264
+    },
+    {
+      "epoch": 0.11107122757104809,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.8564685185276204e-05,
+      "loss": 0.8355,
+      "step": 6266
+    },
+    {
+      "epoch": 0.11110667960665967,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.8563752347762305e-05,
+      "loss": 0.8608,
+      "step": 6268
+    },
+    {
+      "epoch": 0.11114213164227124,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.856281921617768e-05,
+      "loss": 0.7789,
+      "step": 6270
+    },
+    {
+      "epoch": 0.1111775836778828,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.856188579053399e-05,
+      "loss": 0.8424,
+      "step": 6272
+    },
+    {
+      "epoch": 0.11121303571349438,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.856095207084286e-05,
+      "loss": 0.8807,
+      "step": 6274
+    },
+    {
+      "epoch": 0.11124848774910594,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.8560018057115965e-05,
+      "loss": 0.83,
+      "step": 6276
+    },
+    {
+      "epoch": 0.11128393978471751,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.855908374936495e-05,
+      "loss": 0.806,
+      "step": 6278
+    },
+    {
+      "epoch": 0.11131939182032909,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.8558149147601474e-05,
+      "loss": 0.855,
+      "step": 6280
+    },
+    {
+      "epoch": 0.11135484385594065,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.85572142518372e-05,
+      "loss": 0.7635,
+      "step": 6282
+    },
+    {
+      "epoch": 0.11139029589155222,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.855627906208381e-05,
+      "loss": 0.7634,
+      "step": 6284
+    },
+    {
+      "epoch": 0.1114257479271638,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.8555343578352955e-05,
+      "loss": 0.8574,
+      "step": 6286
+    },
+    {
+      "epoch": 0.11146119996277536,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.855440780065632e-05,
+      "loss": 0.8141,
+      "step": 6288
+    },
+    {
+      "epoch": 0.11149665199838693,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.8553471729005577e-05,
+      "loss": 0.8272,
+      "step": 6290
+    },
+    {
+      "epoch": 0.1115321040339985,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.855253536341242e-05,
+      "loss": 0.8545,
+      "step": 6292
+    },
+    {
+      "epoch": 0.11156755606961007,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.8551598703888525e-05,
+      "loss": 0.8552,
+      "step": 6294
+    },
+    {
+      "epoch": 0.11160300810522164,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.855066175044558e-05,
+      "loss": 0.8481,
+      "step": 6296
+    },
+    {
+      "epoch": 0.11163846014083322,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.8549724503095286e-05,
+      "loss": 0.8124,
+      "step": 6298
+    },
+    {
+      "epoch": 0.11167391217644478,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.8548786961849334e-05,
+      "loss": 0.8208,
+      "step": 6300
+    },
+    {
+      "epoch": 0.11170936421205635,
+      "grad_norm": 2.75,
+      "learning_rate": 4.8547849126719426e-05,
+      "loss": 0.8274,
+      "step": 6302
+    },
+    {
+      "epoch": 0.11174481624766792,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.8546910997717264e-05,
+      "loss": 0.8817,
+      "step": 6304
+    },
+    {
+      "epoch": 0.11178026828327949,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.854597257485456e-05,
+      "loss": 0.8826,
+      "step": 6306
+    },
+    {
+      "epoch": 0.11181572031889106,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.8545033858143025e-05,
+      "loss": 0.8136,
+      "step": 6308
+    },
+    {
+      "epoch": 0.11185117235450263,
+      "grad_norm": 2.875,
+      "learning_rate": 4.8544094847594366e-05,
+      "loss": 0.8205,
+      "step": 6310
+    },
+    {
+      "epoch": 0.1118866243901142,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.854315554322031e-05,
+      "loss": 0.8596,
+      "step": 6312
+    },
+    {
+      "epoch": 0.11192207642572577,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.854221594503258e-05,
+      "loss": 0.8556,
+      "step": 6314
+    },
+    {
+      "epoch": 0.11195752846133734,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.854127605304289e-05,
+      "loss": 0.8152,
+      "step": 6316
+    },
+    {
+      "epoch": 0.1119929804969489,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.8540335867262976e-05,
+      "loss": 0.8278,
+      "step": 6318
+    },
+    {
+      "epoch": 0.11202843253256048,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.853939538770458e-05,
+      "loss": 0.8356,
+      "step": 6320
+    },
+    {
+      "epoch": 0.11206388456817205,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.8538454614379435e-05,
+      "loss": 0.809,
+      "step": 6322
+    },
+    {
+      "epoch": 0.11209933660378361,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.8537513547299276e-05,
+      "loss": 0.8496,
+      "step": 6324
+    },
+    {
+      "epoch": 0.11213478863939519,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.853657218647585e-05,
+      "loss": 0.8314,
+      "step": 6326
+    },
+    {
+      "epoch": 0.11217024067500676,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.853563053192091e-05,
+      "loss": 0.8568,
+      "step": 6328
+    },
+    {
+      "epoch": 0.11220569271061832,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.853468858364619e-05,
+      "loss": 0.8321,
+      "step": 6330
+    },
+    {
+      "epoch": 0.1122411447462299,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.853374634166347e-05,
+      "loss": 0.854,
+      "step": 6332
+    },
+    {
+      "epoch": 0.11227659678184147,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.8532803805984496e-05,
+      "loss": 0.8576,
+      "step": 6334
+    },
+    {
+      "epoch": 0.11231204881745303,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.853186097662103e-05,
+      "loss": 0.8006,
+      "step": 6336
+    },
+    {
+      "epoch": 0.11234750085306461,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.853091785358485e-05,
+      "loss": 0.8828,
+      "step": 6338
+    },
+    {
+      "epoch": 0.11238295288867618,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.85299744368877e-05,
+      "loss": 0.8404,
+      "step": 6340
+    },
+    {
+      "epoch": 0.11241840492428774,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.8529030726541383e-05,
+      "loss": 0.8326,
+      "step": 6342
+    },
+    {
+      "epoch": 0.11245385695989932,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.8528086722557656e-05,
+      "loss": 0.8589,
+      "step": 6344
+    },
+    {
+      "epoch": 0.11248930899551089,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.8527142424948305e-05,
+      "loss": 0.8304,
+      "step": 6346
+    },
+    {
+      "epoch": 0.11252476103112245,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.8526197833725126e-05,
+      "loss": 0.8005,
+      "step": 6348
+    },
+    {
+      "epoch": 0.11256021306673403,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.852525294889989e-05,
+      "loss": 0.8631,
+      "step": 6350
+    },
+    {
+      "epoch": 0.1125956651023456,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.85243077704844e-05,
+      "loss": 0.8739,
+      "step": 6352
+    },
+    {
+      "epoch": 0.11263111713795716,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.8523362298490446e-05,
+      "loss": 0.8413,
+      "step": 6354
+    },
+    {
+      "epoch": 0.11266656917356874,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.852241653292984e-05,
+      "loss": 0.8242,
+      "step": 6356
+    },
+    {
+      "epoch": 0.11270202120918031,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.8521470473814365e-05,
+      "loss": 0.8035,
+      "step": 6358
+    },
+    {
+      "epoch": 0.11273747324479187,
+      "grad_norm": 2.75,
+      "learning_rate": 4.852052412115584e-05,
+      "loss": 0.7923,
+      "step": 6360
+    },
+    {
+      "epoch": 0.11277292528040345,
+      "grad_norm": 2.875,
+      "learning_rate": 4.8519577474966074e-05,
+      "loss": 0.8228,
+      "step": 6362
+    },
+    {
+      "epoch": 0.11280837731601501,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.851863053525688e-05,
+      "loss": 0.7987,
+      "step": 6364
+    },
+    {
+      "epoch": 0.11284382935162658,
+      "grad_norm": 2.75,
+      "learning_rate": 4.8517683302040075e-05,
+      "loss": 0.8717,
+      "step": 6366
+    },
+    {
+      "epoch": 0.11287928138723816,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.851673577532748e-05,
+      "loss": 0.8595,
+      "step": 6368
+    },
+    {
+      "epoch": 0.11291473342284972,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.851578795513092e-05,
+      "loss": 0.8714,
+      "step": 6370
+    },
+    {
+      "epoch": 0.11295018545846129,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.851483984146223e-05,
+      "loss": 0.8443,
+      "step": 6372
+    },
+    {
+      "epoch": 0.11298563749407287,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.851389143433323e-05,
+      "loss": 0.8583,
+      "step": 6374
+    },
+    {
+      "epoch": 0.11302108952968443,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.8512942733755764e-05,
+      "loss": 0.8792,
+      "step": 6376
+    },
+    {
+      "epoch": 0.113056541565296,
+      "grad_norm": 2.875,
+      "learning_rate": 4.8511993739741676e-05,
+      "loss": 0.8085,
+      "step": 6378
+    },
+    {
+      "epoch": 0.11309199360090758,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.85110444523028e-05,
+      "loss": 0.8323,
+      "step": 6380
+    },
+    {
+      "epoch": 0.11312744563651914,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.851009487145098e-05,
+      "loss": 0.8491,
+      "step": 6382
+    },
+    {
+      "epoch": 0.11316289767213071,
+      "grad_norm": 2.875,
+      "learning_rate": 4.850914499719807e-05,
+      "loss": 0.842,
+      "step": 6384
+    },
+    {
+      "epoch": 0.11319834970774229,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.850819482955594e-05,
+      "loss": 0.8243,
+      "step": 6386
+    },
+    {
+      "epoch": 0.11323380174335385,
+      "grad_norm": 2.75,
+      "learning_rate": 4.850724436853643e-05,
+      "loss": 0.8169,
+      "step": 6388
+    },
+    {
+      "epoch": 0.11326925377896542,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.8506293614151404e-05,
+      "loss": 0.83,
+      "step": 6390
+    },
+    {
+      "epoch": 0.113304705814577,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.8505342566412734e-05,
+      "loss": 0.8608,
+      "step": 6392
+    },
+    {
+      "epoch": 0.11334015785018856,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.8504391225332277e-05,
+      "loss": 0.8466,
+      "step": 6394
+    },
+    {
+      "epoch": 0.11337560988580013,
+      "grad_norm": 2.875,
+      "learning_rate": 4.8503439590921925e-05,
+      "loss": 0.8541,
+      "step": 6396
+    },
+    {
+      "epoch": 0.1134110619214117,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.850248766319353e-05,
+      "loss": 0.8495,
+      "step": 6398
+    },
+    {
+      "epoch": 0.11344651395702327,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.8501535442159e-05,
+      "loss": 0.8642,
+      "step": 6400
+    },
+    {
+      "epoch": 0.11348196599263484,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.8500582927830185e-05,
+      "loss": 0.8203,
+      "step": 6402
+    },
+    {
+      "epoch": 0.11351741802824641,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.8499630120218994e-05,
+      "loss": 0.8405,
+      "step": 6404
+    },
+    {
+      "epoch": 0.11355287006385797,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.849867701933732e-05,
+      "loss": 0.8076,
+      "step": 6406
+    },
+    {
+      "epoch": 0.11358832209946955,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.849772362519704e-05,
+      "loss": 0.8343,
+      "step": 6408
+    },
+    {
+      "epoch": 0.11362377413508112,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.849676993781008e-05,
+      "loss": 0.8003,
+      "step": 6410
+    },
+    {
+      "epoch": 0.11365922617069268,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.849581595718832e-05,
+      "loss": 0.8195,
+      "step": 6412
+    },
+    {
+      "epoch": 0.11369467820630426,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.849486168334366e-05,
+      "loss": 0.8063,
+      "step": 6414
+    },
+    {
+      "epoch": 0.11373013024191583,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.849390711628803e-05,
+      "loss": 0.8344,
+      "step": 6416
+    },
+    {
+      "epoch": 0.1137655822775274,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.8492952256033333e-05,
+      "loss": 0.852,
+      "step": 6418
+    },
+    {
+      "epoch": 0.11380103431313897,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.849199710259148e-05,
+      "loss": 0.8153,
+      "step": 6420
+    },
+    {
+      "epoch": 0.11383648634875054,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.84910416559744e-05,
+      "loss": 0.8212,
+      "step": 6422
+    },
+    {
+      "epoch": 0.1138719383843621,
+      "grad_norm": 2.75,
+      "learning_rate": 4.8490085916194005e-05,
+      "loss": 0.8511,
+      "step": 6424
+    },
+    {
+      "epoch": 0.11390739041997368,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.848912988326224e-05,
+      "loss": 0.8078,
+      "step": 6426
+    },
+    {
+      "epoch": 0.11394284245558525,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.8488173557191026e-05,
+      "loss": 0.8286,
+      "step": 6428
+    },
+    {
+      "epoch": 0.11397829449119681,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.8487216937992294e-05,
+      "loss": 0.863,
+      "step": 6430
+    },
+    {
+      "epoch": 0.11401374652680839,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.848626002567799e-05,
+      "loss": 0.8603,
+      "step": 6432
+    },
+    {
+      "epoch": 0.11404919856241996,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.8485302820260045e-05,
+      "loss": 0.847,
+      "step": 6434
+    },
+    {
+      "epoch": 0.11408465059803152,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.848434532175042e-05,
+      "loss": 0.8007,
+      "step": 6436
+    },
+    {
+      "epoch": 0.1141201026336431,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.8483387530161054e-05,
+      "loss": 0.8277,
+      "step": 6438
+    },
+    {
+      "epoch": 0.11415555466925467,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.8482429445503905e-05,
+      "loss": 0.7848,
+      "step": 6440
+    },
+    {
+      "epoch": 0.11419100670486623,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.848147106779093e-05,
+      "loss": 0.804,
+      "step": 6442
+    },
+    {
+      "epoch": 0.11422645874047781,
+      "grad_norm": 3.0,
+      "learning_rate": 4.848051239703408e-05,
+      "loss": 0.8579,
+      "step": 6444
+    },
+    {
+      "epoch": 0.11426191077608938,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.8479553433245325e-05,
+      "loss": 0.8028,
+      "step": 6446
+    },
+    {
+      "epoch": 0.11429736281170094,
+      "grad_norm": 2.75,
+      "learning_rate": 4.847859417643664e-05,
+      "loss": 0.8015,
+      "step": 6448
+    },
+    {
+      "epoch": 0.11433281484731252,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.847763462661999e-05,
+      "loss": 0.8323,
+      "step": 6450
+    },
+    {
+      "epoch": 0.11436826688292408,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.8476674783807344e-05,
+      "loss": 0.8275,
+      "step": 6452
+    },
+    {
+      "epoch": 0.11440371891853565,
+      "grad_norm": 2.625,
+      "learning_rate": 4.84757146480107e-05,
+      "loss": 0.9091,
+      "step": 6454
+    },
+    {
+      "epoch": 0.11443917095414723,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.8474754219242016e-05,
+      "loss": 0.8291,
+      "step": 6456
+    },
+    {
+      "epoch": 0.11447462298975879,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.847379349751329e-05,
+      "loss": 0.8048,
+      "step": 6458
+    },
+    {
+      "epoch": 0.11451007502537036,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.8472832482836504e-05,
+      "loss": 0.8122,
+      "step": 6460
+    },
+    {
+      "epoch": 0.11454552706098194,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.847187117522366e-05,
+      "loss": 0.8276,
+      "step": 6462
+    },
+    {
+      "epoch": 0.1145809790965935,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.8470909574686764e-05,
+      "loss": 0.8586,
+      "step": 6464
+    },
+    {
+      "epoch": 0.11461643113220507,
+      "grad_norm": 2.75,
+      "learning_rate": 4.846994768123779e-05,
+      "loss": 0.8073,
+      "step": 6466
+    },
+    {
+      "epoch": 0.11465188316781665,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.846898549488877e-05,
+      "loss": 0.8771,
+      "step": 6468
+    },
+    {
+      "epoch": 0.1146873352034282,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.846802301565169e-05,
+      "loss": 0.8059,
+      "step": 6470
+    },
+    {
+      "epoch": 0.11472278723903978,
+      "grad_norm": 2.625,
+      "learning_rate": 4.8467060243538574e-05,
+      "loss": 0.8012,
+      "step": 6472
+    },
+    {
+      "epoch": 0.11475823927465136,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.8466097178561435e-05,
+      "loss": 0.865,
+      "step": 6474
+    },
+    {
+      "epoch": 0.11479369131026292,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.84651338207323e-05,
+      "loss": 0.8429,
+      "step": 6476
+    },
+    {
+      "epoch": 0.11482914334587449,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.846417017006317e-05,
+      "loss": 0.8308,
+      "step": 6478
+    },
+    {
+      "epoch": 0.11486459538148606,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.8463206226566084e-05,
+      "loss": 0.8076,
+      "step": 6480
+    },
+    {
+      "epoch": 0.11490004741709763,
+      "grad_norm": 2.625,
+      "learning_rate": 4.8462241990253077e-05,
+      "loss": 0.8155,
+      "step": 6482
+    },
+    {
+      "epoch": 0.1149354994527092,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.846127746113617e-05,
+      "loss": 0.8628,
+      "step": 6484
+    },
+    {
+      "epoch": 0.11497095148832077,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.8460312639227414e-05,
+      "loss": 0.8624,
+      "step": 6486
+    },
+    {
+      "epoch": 0.11500640352393234,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.8459347524538834e-05,
+      "loss": 0.8101,
+      "step": 6488
+    },
+    {
+      "epoch": 0.11504185555954391,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.845838211708249e-05,
+      "loss": 0.8185,
+      "step": 6490
+    },
+    {
+      "epoch": 0.11507730759515548,
+      "grad_norm": 2.5,
+      "learning_rate": 4.845741641687042e-05,
+      "loss": 0.8343,
+      "step": 6492
+    },
+    {
+      "epoch": 0.11511275963076704,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.8456450423914677e-05,
+      "loss": 0.8335,
+      "step": 6494
+    },
+    {
+      "epoch": 0.11514821166637862,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.845548413822733e-05,
+      "loss": 0.8088,
+      "step": 6496
+    },
+    {
+      "epoch": 0.1151836637019902,
+      "grad_norm": 3.0,
+      "learning_rate": 4.8454517559820414e-05,
+      "loss": 0.8173,
+      "step": 6498
+    },
+    {
+      "epoch": 0.11521911573760175,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.8453550688706006e-05,
+      "loss": 0.8532,
+      "step": 6500
+    },
+    {
+      "epoch": 0.11525456777321333,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.845258352489618e-05,
+      "loss": 0.7932,
+      "step": 6502
+    },
+    {
+      "epoch": 0.1152900198088249,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.8451616068402985e-05,
+      "loss": 0.8306,
+      "step": 6504
+    },
+    {
+      "epoch": 0.11532547184443646,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.845064831923851e-05,
+      "loss": 0.8666,
+      "step": 6506
+    },
+    {
+      "epoch": 0.11536092388004804,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.8449680277414834e-05,
+      "loss": 0.8321,
+      "step": 6508
+    },
+    {
+      "epoch": 0.11539637591565961,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.8448711942944025e-05,
+      "loss": 0.8143,
+      "step": 6510
+    },
+    {
+      "epoch": 0.11543182795127117,
+      "grad_norm": 2.625,
+      "learning_rate": 4.844774331583818e-05,
+      "loss": 0.8275,
+      "step": 6512
+    },
+    {
+      "epoch": 0.11546727998688275,
+      "grad_norm": 2.5,
+      "learning_rate": 4.8446774396109375e-05,
+      "loss": 0.8232,
+      "step": 6514
+    },
+    {
+      "epoch": 0.11550273202249432,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.844580518376971e-05,
+      "loss": 0.8541,
+      "step": 6516
+    },
+    {
+      "epoch": 0.11553818405810588,
+      "grad_norm": 2.875,
+      "learning_rate": 4.844483567883128e-05,
+      "loss": 0.857,
+      "step": 6518
+    },
+    {
+      "epoch": 0.11557363609371746,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.8443865881306194e-05,
+      "loss": 0.8258,
+      "step": 6520
+    },
+    {
+      "epoch": 0.11560908812932903,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.844289579120653e-05,
+      "loss": 0.8415,
+      "step": 6522
+    },
+    {
+      "epoch": 0.11564454016494059,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.844192540854442e-05,
+      "loss": 0.8205,
+      "step": 6524
+    },
+    {
+      "epoch": 0.11567999220055217,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.844095473333197e-05,
+      "loss": 0.7767,
+      "step": 6526
+    },
+    {
+      "epoch": 0.11571544423616374,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.8439983765581274e-05,
+      "loss": 0.7606,
+      "step": 6528
+    },
+    {
+      "epoch": 0.1157508962717753,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.8439012505304465e-05,
+      "loss": 0.7859,
+      "step": 6530
+    },
+    {
+      "epoch": 0.11578634830738688,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.8438040952513664e-05,
+      "loss": 0.7816,
+      "step": 6532
+    },
+    {
+      "epoch": 0.11582180034299844,
+      "grad_norm": 2.625,
+      "learning_rate": 4.8437069107220994e-05,
+      "loss": 0.8353,
+      "step": 6534
+    },
+    {
+      "epoch": 0.11585725237861001,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.8436096969438584e-05,
+      "loss": 0.8546,
+      "step": 6536
+    },
+    {
+      "epoch": 0.11589270441422159,
+      "grad_norm": 2.625,
+      "learning_rate": 4.8435124539178564e-05,
+      "loss": 0.7771,
+      "step": 6538
+    },
+    {
+      "epoch": 0.11592815644983315,
+      "grad_norm": 2.875,
+      "learning_rate": 4.843415181645306e-05,
+      "loss": 0.8452,
+      "step": 6540
+    },
+    {
+      "epoch": 0.11596360848544472,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.8433178801274244e-05,
+      "loss": 0.8423,
+      "step": 6542
+    },
+    {
+      "epoch": 0.1159990605210563,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.8432205493654224e-05,
+      "loss": 0.8103,
+      "step": 6544
+    },
+    {
+      "epoch": 0.11603451255666786,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.843123189360516e-05,
+      "loss": 0.8496,
+      "step": 6546
+    },
+    {
+      "epoch": 0.11606996459227943,
+      "grad_norm": 3.0,
+      "learning_rate": 4.8430258001139206e-05,
+      "loss": 0.857,
+      "step": 6548
+    },
+    {
+      "epoch": 0.116105416627891,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.842928381626851e-05,
+      "loss": 0.8203,
+      "step": 6550
+    },
+    {
+      "epoch": 0.11614086866350257,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.8428309339005235e-05,
+      "loss": 0.8282,
+      "step": 6552
+    },
+    {
+      "epoch": 0.11617632069911414,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.8427334569361537e-05,
+      "loss": 0.8114,
+      "step": 6554
+    },
+    {
+      "epoch": 0.11621177273472572,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.842635950734958e-05,
+      "loss": 0.8155,
+      "step": 6556
+    },
+    {
+      "epoch": 0.11624722477033728,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.8425384152981545e-05,
+      "loss": 0.8523,
+      "step": 6558
+    },
+    {
+      "epoch": 0.11628267680594885,
+      "grad_norm": 2.625,
+      "learning_rate": 4.8424408506269585e-05,
+      "loss": 0.8089,
+      "step": 6560
+    },
+    {
+      "epoch": 0.11631812884156043,
+      "grad_norm": 3.578125,
+      "learning_rate": 4.842343256722589e-05,
+      "loss": 0.8448,
+      "step": 6562
+    },
+    {
+      "epoch": 0.11635358087717199,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.842245633586264e-05,
+      "loss": 0.8025,
+      "step": 6564
+    },
+    {
+      "epoch": 0.11638903291278356,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.842147981219201e-05,
+      "loss": 0.8363,
+      "step": 6566
+    },
+    {
+      "epoch": 0.11642448494839513,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.842050299622618e-05,
+      "loss": 0.8789,
+      "step": 6568
+    },
+    {
+      "epoch": 0.1164599369840067,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.841952588797736e-05,
+      "loss": 0.7869,
+      "step": 6570
+    },
+    {
+      "epoch": 0.11649538901961827,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.841854848745774e-05,
+      "loss": 0.8549,
+      "step": 6572
+    },
+    {
+      "epoch": 0.11653084105522984,
+      "grad_norm": 2.625,
+      "learning_rate": 4.84175707946795e-05,
+      "loss": 0.829,
+      "step": 6574
+    },
+    {
+      "epoch": 0.1165662930908414,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.8416592809654865e-05,
+      "loss": 0.8299,
+      "step": 6576
+    },
+    {
+      "epoch": 0.11660174512645298,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.841561453239602e-05,
+      "loss": 0.8258,
+      "step": 6578
+    },
+    {
+      "epoch": 0.11663719716206455,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.84146359629152e-05,
+      "loss": 0.7927,
+      "step": 6580
+    },
+    {
+      "epoch": 0.11667264919767611,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.841365710122458e-05,
+      "loss": 0.8182,
+      "step": 6582
+    },
+    {
+      "epoch": 0.11670810123328769,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.84126779473364e-05,
+      "loss": 0.8024,
+      "step": 6584
+    },
+    {
+      "epoch": 0.11674355326889926,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.841169850126288e-05,
+      "loss": 0.8211,
+      "step": 6586
+    },
+    {
+      "epoch": 0.11677900530451082,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.841071876301625e-05,
+      "loss": 0.8367,
+      "step": 6588
+    },
+    {
+      "epoch": 0.1168144573401224,
+      "grad_norm": 2.75,
+      "learning_rate": 4.840973873260871e-05,
+      "loss": 0.8627,
+      "step": 6590
+    },
+    {
+      "epoch": 0.11684990937573397,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.8408758410052514e-05,
+      "loss": 0.8322,
+      "step": 6592
+    },
+    {
+      "epoch": 0.11688536141134553,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.840777779535988e-05,
+      "loss": 0.8435,
+      "step": 6594
+    },
+    {
+      "epoch": 0.11692081344695711,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.840679688854306e-05,
+      "loss": 0.8276,
+      "step": 6596
+    },
+    {
+      "epoch": 0.11695626548256868,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.840581568961429e-05,
+      "loss": 0.8271,
+      "step": 6598
+    },
+    {
+      "epoch": 0.11699171751818024,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.840483419858582e-05,
+      "loss": 0.8112,
+      "step": 6600
+    },
+    {
+      "epoch": 0.11702716955379182,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.8403852415469885e-05,
+      "loss": 0.8333,
+      "step": 6602
+    },
+    {
+      "epoch": 0.11706262158940339,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.840287034027876e-05,
+      "loss": 0.8411,
+      "step": 6604
+    },
+    {
+      "epoch": 0.11709807362501495,
+      "grad_norm": 2.375,
+      "learning_rate": 4.840188797302467e-05,
+      "loss": 0.824,
+      "step": 6606
+    },
+    {
+      "epoch": 0.11713352566062653,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.84009053137199e-05,
+      "loss": 0.8263,
+      "step": 6608
+    },
+    {
+      "epoch": 0.1171689776962381,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.839992236237672e-05,
+      "loss": 0.8372,
+      "step": 6610
+    },
+    {
+      "epoch": 0.11720442973184966,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.8398939119007365e-05,
+      "loss": 0.8426,
+      "step": 6612
+    },
+    {
+      "epoch": 0.11723988176746124,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.839795558362413e-05,
+      "loss": 0.812,
+      "step": 6614
+    },
+    {
+      "epoch": 0.11727533380307281,
+      "grad_norm": 2.75,
+      "learning_rate": 4.839697175623928e-05,
+      "loss": 0.8206,
+      "step": 6616
+    },
+    {
+      "epoch": 0.11731078583868437,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.839598763686509e-05,
+      "loss": 0.8273,
+      "step": 6618
+    },
+    {
+      "epoch": 0.11734623787429595,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.839500322551386e-05,
+      "loss": 0.8424,
+      "step": 6620
+    },
+    {
+      "epoch": 0.11738168990990751,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.839401852219786e-05,
+      "loss": 0.8293,
+      "step": 6622
+    },
+    {
+      "epoch": 0.11741714194551908,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.839303352692938e-05,
+      "loss": 0.8094,
+      "step": 6624
+    },
+    {
+      "epoch": 0.11745259398113066,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.8392048239720703e-05,
+      "loss": 0.7789,
+      "step": 6626
+    },
+    {
+      "epoch": 0.11748804601674222,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.839106266058415e-05,
+      "loss": 0.8413,
+      "step": 6628
+    },
+    {
+      "epoch": 0.11752349805235379,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.8390076789532004e-05,
+      "loss": 0.7869,
+      "step": 6630
+    },
+    {
+      "epoch": 0.11755895008796537,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.838909062657657e-05,
+      "loss": 0.833,
+      "step": 6632
+    },
+    {
+      "epoch": 0.11759440212357693,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.838810417173015e-05,
+      "loss": 0.8326,
+      "step": 6634
+    },
+    {
+      "epoch": 0.1176298541591885,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.8387117425005066e-05,
+      "loss": 0.8743,
+      "step": 6636
+    },
+    {
+      "epoch": 0.11766530619480008,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.8386130386413635e-05,
+      "loss": 0.8445,
+      "step": 6638
+    },
+    {
+      "epoch": 0.11770075823041164,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.8385143055968166e-05,
+      "loss": 0.8199,
+      "step": 6640
+    },
+    {
+      "epoch": 0.11773621026602321,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.838415543368098e-05,
+      "loss": 0.8384,
+      "step": 6642
+    },
+    {
+      "epoch": 0.11777166230163479,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.83831675195644e-05,
+      "loss": 0.8122,
+      "step": 6644
+    },
+    {
+      "epoch": 0.11780711433724635,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.838217931363076e-05,
+      "loss": 0.8082,
+      "step": 6646
+    },
+    {
+      "epoch": 0.11784256637285792,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.8381190815892394e-05,
+      "loss": 0.829,
+      "step": 6648
+    },
+    {
+      "epoch": 0.1178780184084695,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.8380202026361644e-05,
+      "loss": 0.8042,
+      "step": 6650
+    },
+    {
+      "epoch": 0.11791347044408106,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.837921294505083e-05,
+      "loss": 0.8213,
+      "step": 6652
+    },
+    {
+      "epoch": 0.11794892247969263,
+      "grad_norm": 2.625,
+      "learning_rate": 4.837822357197232e-05,
+      "loss": 0.7967,
+      "step": 6654
+    },
+    {
+      "epoch": 0.1179843745153042,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.8377233907138444e-05,
+      "loss": 0.8582,
+      "step": 6656
+    },
+    {
+      "epoch": 0.11801982655091577,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.837624395056155e-05,
+      "loss": 0.8131,
+      "step": 6658
+    },
+    {
+      "epoch": 0.11805527858652734,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.837525370225401e-05,
+      "loss": 0.8288,
+      "step": 6660
+    },
+    {
+      "epoch": 0.11809073062213891,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.8374263162228176e-05,
+      "loss": 0.8583,
+      "step": 6662
+    },
+    {
+      "epoch": 0.11812618265775048,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.83732723304964e-05,
+      "loss": 0.8188,
+      "step": 6664
+    },
+    {
+      "epoch": 0.11816163469336205,
+      "grad_norm": 3.0,
+      "learning_rate": 4.837228120707106e-05,
+      "loss": 0.8512,
+      "step": 6666
+    },
+    {
+      "epoch": 0.11819708672897362,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.837128979196451e-05,
+      "loss": 0.815,
+      "step": 6668
+    },
+    {
+      "epoch": 0.11823253876458518,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.8370298085189134e-05,
+      "loss": 0.8627,
+      "step": 6670
+    },
+    {
+      "epoch": 0.11826799080019676,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.83693060867573e-05,
+      "loss": 0.8659,
+      "step": 6672
+    },
+    {
+      "epoch": 0.11830344283580833,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.8368313796681404e-05,
+      "loss": 0.8722,
+      "step": 6674
+    },
+    {
+      "epoch": 0.1183388948714199,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.8367321214973815e-05,
+      "loss": 0.8038,
+      "step": 6676
+    },
+    {
+      "epoch": 0.11837434690703147,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.836632834164692e-05,
+      "loss": 0.8188,
+      "step": 6678
+    },
+    {
+      "epoch": 0.11840979894264304,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.836533517671312e-05,
+      "loss": 0.8267,
+      "step": 6680
+    },
+    {
+      "epoch": 0.1184452509782546,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.83643417201848e-05,
+      "loss": 0.8017,
+      "step": 6682
+    },
+    {
+      "epoch": 0.11848070301386618,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.836334797207437e-05,
+      "loss": 0.8188,
+      "step": 6684
+    },
+    {
+      "epoch": 0.11851615504947775,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.836235393239421e-05,
+      "loss": 0.7921,
+      "step": 6686
+    },
+    {
+      "epoch": 0.11855160708508931,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.836135960115675e-05,
+      "loss": 0.8295,
+      "step": 6688
+    },
+    {
+      "epoch": 0.11858705912070089,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.8360364978374384e-05,
+      "loss": 0.8566,
+      "step": 6690
+    },
+    {
+      "epoch": 0.11862251115631246,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.835937006405953e-05,
+      "loss": 0.8308,
+      "step": 6692
+    },
+    {
+      "epoch": 0.11865796319192402,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.83583748582246e-05,
+      "loss": 0.8197,
+      "step": 6694
+    },
+    {
+      "epoch": 0.1186934152275356,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.8357379360882014e-05,
+      "loss": 0.8059,
+      "step": 6696
+    },
+    {
+      "epoch": 0.11872886726314717,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.8356383572044206e-05,
+      "loss": 0.829,
+      "step": 6698
+    },
+    {
+      "epoch": 0.11876431929875873,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.835538749172359e-05,
+      "loss": 0.8027,
+      "step": 6700
+    },
+    {
+      "epoch": 0.11879977133437031,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.835439111993261e-05,
+      "loss": 0.8291,
+      "step": 6702
+    },
+    {
+      "epoch": 0.11883522336998187,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.835339445668369e-05,
+      "loss": 0.8159,
+      "step": 6704
+    },
+    {
+      "epoch": 0.11887067540559344,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.8352397501989265e-05,
+      "loss": 0.8232,
+      "step": 6706
+    },
+    {
+      "epoch": 0.11890612744120502,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.835140025586179e-05,
+      "loss": 0.8433,
+      "step": 6708
+    },
+    {
+      "epoch": 0.11894157947681658,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.83504027183137e-05,
+      "loss": 0.811,
+      "step": 6710
+    },
+    {
+      "epoch": 0.11897703151242815,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.8349404889357455e-05,
+      "loss": 0.8117,
+      "step": 6712
+    },
+    {
+      "epoch": 0.11901248354803973,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.8348406769005494e-05,
+      "loss": 0.8082,
+      "step": 6714
+    },
+    {
+      "epoch": 0.11904793558365129,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.834740835727028e-05,
+      "loss": 0.8391,
+      "step": 6716
+    },
+    {
+      "epoch": 0.11908338761926286,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.834640965416427e-05,
+      "loss": 0.7962,
+      "step": 6718
+    },
+    {
+      "epoch": 0.11911883965487444,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.834541065969993e-05,
+      "loss": 0.8261,
+      "step": 6720
+    },
+    {
+      "epoch": 0.119154291690486,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.834441137388973e-05,
+      "loss": 0.7957,
+      "step": 6722
+    },
+    {
+      "epoch": 0.11918974372609757,
+      "grad_norm": 2.75,
+      "learning_rate": 4.834341179674614e-05,
+      "loss": 0.8004,
+      "step": 6724
+    },
+    {
+      "epoch": 0.11922519576170915,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.834241192828164e-05,
+      "loss": 0.8264,
+      "step": 6726
+    },
+    {
+      "epoch": 0.1192606477973207,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.8341411768508684e-05,
+      "loss": 0.7818,
+      "step": 6728
+    },
+    {
+      "epoch": 0.11929609983293228,
+      "grad_norm": 2.75,
+      "learning_rate": 4.8340411317439785e-05,
+      "loss": 0.8018,
+      "step": 6730
+    },
+    {
+      "epoch": 0.11933155186854386,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.833941057508741e-05,
+      "loss": 0.8283,
+      "step": 6732
+    },
+    {
+      "epoch": 0.11936700390415542,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.8338409541464045e-05,
+      "loss": 0.8298,
+      "step": 6734
+    },
+    {
+      "epoch": 0.11940245593976699,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.83374082165822e-05,
+      "loss": 0.8132,
+      "step": 6736
+    },
+    {
+      "epoch": 0.11943790797537857,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.833640660045436e-05,
+      "loss": 0.8407,
+      "step": 6738
+    },
+    {
+      "epoch": 0.11947336001099013,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.8335404693093026e-05,
+      "loss": 0.8115,
+      "step": 6740
+    },
+    {
+      "epoch": 0.1195088120466017,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.83344024945107e-05,
+      "loss": 0.8323,
+      "step": 6742
+    },
+    {
+      "epoch": 0.11954426408221328,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.8333400004719885e-05,
+      "loss": 0.8535,
+      "step": 6744
+    },
+    {
+      "epoch": 0.11957971611782484,
+      "grad_norm": 2.625,
+      "learning_rate": 4.8332397223733104e-05,
+      "loss": 0.8649,
+      "step": 6746
+    },
+    {
+      "epoch": 0.11961516815343641,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.8331394151562864e-05,
+      "loss": 0.8297,
+      "step": 6748
+    },
+    {
+      "epoch": 0.11965062018904798,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.833039078822169e-05,
+      "loss": 0.8305,
+      "step": 6750
+    },
+    {
+      "epoch": 0.11968607222465955,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.832938713372209e-05,
+      "loss": 0.8449,
+      "step": 6752
+    },
+    {
+      "epoch": 0.11972152426027112,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.8328383188076595e-05,
+      "loss": 0.8273,
+      "step": 6754
+    },
+    {
+      "epoch": 0.1197569762958827,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.832737895129775e-05,
+      "loss": 0.8507,
+      "step": 6756
+    },
+    {
+      "epoch": 0.11979242833149426,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.8326374423398066e-05,
+      "loss": 0.7767,
+      "step": 6758
+    },
+    {
+      "epoch": 0.11982788036710583,
+      "grad_norm": 2.875,
+      "learning_rate": 4.832536960439009e-05,
+      "loss": 0.8543,
+      "step": 6760
+    },
+    {
+      "epoch": 0.1198633324027174,
+      "grad_norm": 2.75,
+      "learning_rate": 4.8324364494286364e-05,
+      "loss": 0.8477,
+      "step": 6762
+    },
+    {
+      "epoch": 0.11989878443832896,
+      "grad_norm": 2.75,
+      "learning_rate": 4.832335909309942e-05,
+      "loss": 0.7939,
+      "step": 6764
+    },
+    {
+      "epoch": 0.11993423647394054,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.8322353400841816e-05,
+      "loss": 0.7968,
+      "step": 6766
+    },
+    {
+      "epoch": 0.11996968850955211,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.83213474175261e-05,
+      "loss": 0.8366,
+      "step": 6768
+    },
+    {
+      "epoch": 0.12000514054516367,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.8320341143164815e-05,
+      "loss": 0.8365,
+      "step": 6770
+    },
+    {
+      "epoch": 0.12004059258077525,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.831933457777055e-05,
+      "loss": 0.8163,
+      "step": 6772
+    },
+    {
+      "epoch": 0.12007604461638682,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.8318327721355825e-05,
+      "loss": 0.8055,
+      "step": 6774
+    },
+    {
+      "epoch": 0.12011149665199838,
+      "grad_norm": 2.75,
+      "learning_rate": 4.831732057393324e-05,
+      "loss": 0.7799,
+      "step": 6776
+    },
+    {
+      "epoch": 0.12014694868760996,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.8316313135515343e-05,
+      "loss": 0.8324,
+      "step": 6778
+    },
+    {
+      "epoch": 0.12018240072322153,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.8315305406114726e-05,
+      "loss": 0.8153,
+      "step": 6780
+    },
+    {
+      "epoch": 0.1202178527588331,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.8314297385743945e-05,
+      "loss": 0.8353,
+      "step": 6782
+    },
+    {
+      "epoch": 0.12025330479444467,
+      "grad_norm": 3.125,
+      "learning_rate": 4.831328907441559e-05,
+      "loss": 0.8471,
+      "step": 6784
+    },
+    {
+      "epoch": 0.12028875683005624,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.831228047214224e-05,
+      "loss": 0.8277,
+      "step": 6786
+    },
+    {
+      "epoch": 0.1203242088656678,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.8311271578936496e-05,
+      "loss": 0.8371,
+      "step": 6788
+    },
+    {
+      "epoch": 0.12035966090127938,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.8310262394810934e-05,
+      "loss": 0.7628,
+      "step": 6790
+    },
+    {
+      "epoch": 0.12039511293689094,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.8309252919778146e-05,
+      "loss": 0.8369,
+      "step": 6792
+    },
+    {
+      "epoch": 0.12043056497250251,
+      "grad_norm": 2.875,
+      "learning_rate": 4.830824315385074e-05,
+      "loss": 0.8366,
+      "step": 6794
+    },
+    {
+      "epoch": 0.12046601700811409,
+      "grad_norm": 2.875,
+      "learning_rate": 4.830723309704131e-05,
+      "loss": 0.8113,
+      "step": 6796
+    },
+    {
+      "epoch": 0.12050146904372565,
+      "grad_norm": 2.75,
+      "learning_rate": 4.8306222749362475e-05,
+      "loss": 0.8502,
+      "step": 6798
+    },
+    {
+      "epoch": 0.12053692107933722,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.8305212110826833e-05,
+      "loss": 0.8319,
+      "step": 6800
+    },
+    {
+      "epoch": 0.1205723731149488,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.8304201181446994e-05,
+      "loss": 0.8532,
+      "step": 6802
+    },
+    {
+      "epoch": 0.12060782515056036,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.830318996123557e-05,
+      "loss": 0.8519,
+      "step": 6804
+    },
+    {
+      "epoch": 0.12064327718617193,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.830217845020521e-05,
+      "loss": 0.8027,
+      "step": 6806
+    },
+    {
+      "epoch": 0.1206787292217835,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.83011666483685e-05,
+      "loss": 0.8347,
+      "step": 6808
+    },
+    {
+      "epoch": 0.12071418125739507,
+      "grad_norm": 2.375,
+      "learning_rate": 4.830015455573809e-05,
+      "loss": 0.8236,
+      "step": 6810
+    },
+    {
+      "epoch": 0.12074963329300664,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.82991421723266e-05,
+      "loss": 0.8405,
+      "step": 6812
+    },
+    {
+      "epoch": 0.12078508532861822,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.829812949814667e-05,
+      "loss": 0.8778,
+      "step": 6814
+    },
+    {
+      "epoch": 0.12082053736422978,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.829711653321093e-05,
+      "loss": 0.8311,
+      "step": 6816
+    },
+    {
+      "epoch": 0.12085598939984135,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.829610327753204e-05,
+      "loss": 0.8165,
+      "step": 6818
+    },
+    {
+      "epoch": 0.12089144143545293,
+      "grad_norm": 2.75,
+      "learning_rate": 4.829508973112263e-05,
+      "loss": 0.8344,
+      "step": 6820
+    },
+    {
+      "epoch": 0.12092689347106449,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.829407589399535e-05,
+      "loss": 0.7916,
+      "step": 6822
+    },
+    {
+      "epoch": 0.12096234550667606,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.829306176616285e-05,
+      "loss": 0.7913,
+      "step": 6824
+    },
+    {
+      "epoch": 0.12099779754228764,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.82920473476378e-05,
+      "loss": 0.7872,
+      "step": 6826
+    },
+    {
+      "epoch": 0.1210332495778992,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.8291032638432846e-05,
+      "loss": 0.8254,
+      "step": 6828
+    },
+    {
+      "epoch": 0.12106870161351077,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.8290017638560656e-05,
+      "loss": 0.8217,
+      "step": 6830
+    },
+    {
+      "epoch": 0.12110415364912235,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.8289002348033895e-05,
+      "loss": 0.7936,
+      "step": 6832
+    },
+    {
+      "epoch": 0.1211396056847339,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.828798676686524e-05,
+      "loss": 0.8685,
+      "step": 6834
+    },
+    {
+      "epoch": 0.12117505772034548,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.828697089506736e-05,
+      "loss": 0.8507,
+      "step": 6836
+    },
+    {
+      "epoch": 0.12121050975595705,
+      "grad_norm": 2.875,
+      "learning_rate": 4.8285954732652924e-05,
+      "loss": 0.826,
+      "step": 6838
+    },
+    {
+      "epoch": 0.12124596179156862,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.828493827963464e-05,
+      "loss": 0.8032,
+      "step": 6840
+    },
+    {
+      "epoch": 0.12128141382718019,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.828392153602516e-05,
+      "loss": 0.8642,
+      "step": 6842
+    },
+    {
+      "epoch": 0.12131686586279176,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.82829045018372e-05,
+      "loss": 0.8316,
+      "step": 6844
+    },
+    {
+      "epoch": 0.12135231789840333,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.828188717708343e-05,
+      "loss": 0.81,
+      "step": 6846
+    },
+    {
+      "epoch": 0.1213877699340149,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.828086956177657e-05,
+      "loss": 0.8534,
+      "step": 6848
+    },
+    {
+      "epoch": 0.12142322196962647,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.8279851655929295e-05,
+      "loss": 0.8427,
+      "step": 6850
+    },
+    {
+      "epoch": 0.12145867400523803,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.827883345955433e-05,
+      "loss": 0.8379,
+      "step": 6852
+    },
+    {
+      "epoch": 0.12149412604084961,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.827781497266437e-05,
+      "loss": 0.7829,
+      "step": 6854
+    },
+    {
+      "epoch": 0.12152957807646118,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.827679619527213e-05,
+      "loss": 0.7964,
+      "step": 6856
+    },
+    {
+      "epoch": 0.12156503011207274,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.827577712739031e-05,
+      "loss": 0.8052,
+      "step": 6858
+    },
+    {
+      "epoch": 0.12160048214768432,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.827475776903165e-05,
+      "loss": 0.8155,
+      "step": 6860
+    },
+    {
+      "epoch": 0.1216359341832959,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.827373812020886e-05,
+      "loss": 0.7722,
+      "step": 6862
+    },
+    {
+      "epoch": 0.12167138621890745,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.827271818093466e-05,
+      "loss": 0.8165,
+      "step": 6864
+    },
+    {
+      "epoch": 0.12170683825451903,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.8271697951221794e-05,
+      "loss": 0.7605,
+      "step": 6866
+    },
+    {
+      "epoch": 0.1217422902901306,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.827067743108298e-05,
+      "loss": 0.8532,
+      "step": 6868
+    },
+    {
+      "epoch": 0.12177774232574216,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.826965662053096e-05,
+      "loss": 0.7734,
+      "step": 6870
+    },
+    {
+      "epoch": 0.12181319436135374,
+      "grad_norm": 2.5,
+      "learning_rate": 4.826863551957846e-05,
+      "loss": 0.8054,
+      "step": 6872
+    },
+    {
+      "epoch": 0.1218486463969653,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.826761412823825e-05,
+      "loss": 0.8259,
+      "step": 6874
+    },
+    {
+      "epoch": 0.12188409843257687,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.8266592446523055e-05,
+      "loss": 0.8487,
+      "step": 6876
+    },
+    {
+      "epoch": 0.12191955046818845,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.8265570474445636e-05,
+      "loss": 0.7975,
+      "step": 6878
+    },
+    {
+      "epoch": 0.12195500250380001,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.826454821201875e-05,
+      "loss": 0.7988,
+      "step": 6880
+    },
+    {
+      "epoch": 0.12199045453941158,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.826352565925513e-05,
+      "loss": 0.8201,
+      "step": 6882
+    },
+    {
+      "epoch": 0.12202590657502316,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.826250281616757e-05,
+      "loss": 0.8149,
+      "step": 6884
+    },
+    {
+      "epoch": 0.12206135861063472,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.826147968276881e-05,
+      "loss": 0.7627,
+      "step": 6886
+    },
+    {
+      "epoch": 0.12209681064624629,
+      "grad_norm": 2.875,
+      "learning_rate": 4.826045625907164e-05,
+      "loss": 0.8573,
+      "step": 6888
+    },
+    {
+      "epoch": 0.12213226268185787,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.825943254508881e-05,
+      "loss": 0.8613,
+      "step": 6890
+    },
+    {
+      "epoch": 0.12216771471746943,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.825840854083311e-05,
+      "loss": 0.8319,
+      "step": 6892
+    },
+    {
+      "epoch": 0.122203166753081,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.8257384246317316e-05,
+      "loss": 0.827,
+      "step": 6894
+    },
+    {
+      "epoch": 0.12223861878869258,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.8256359661554215e-05,
+      "loss": 0.8295,
+      "step": 6896
+    },
+    {
+      "epoch": 0.12227407082430414,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.825533478655658e-05,
+      "loss": 0.8233,
+      "step": 6898
+    },
+    {
+      "epoch": 0.12230952285991571,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.825430962133722e-05,
+      "loss": 0.8086,
+      "step": 6900
+    },
+    {
+      "epoch": 0.12234497489552729,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.825328416590891e-05,
+      "loss": 0.7928,
+      "step": 6902
+    },
+    {
+      "epoch": 0.12238042693113885,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.825225842028447e-05,
+      "loss": 0.7984,
+      "step": 6904
+    },
+    {
+      "epoch": 0.12241587896675042,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.8251232384476675e-05,
+      "loss": 0.8349,
+      "step": 6906
+    },
+    {
+      "epoch": 0.122451331002362,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.825020605849835e-05,
+      "loss": 0.8235,
+      "step": 6908
+    },
+    {
+      "epoch": 0.12248678303797356,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.8249179442362294e-05,
+      "loss": 0.8255,
+      "step": 6910
+    },
+    {
+      "epoch": 0.12252223507358513,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.824815253608132e-05,
+      "loss": 0.842,
+      "step": 6912
+    },
+    {
+      "epoch": 0.1225576871091967,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.8247125339668244e-05,
+      "loss": 0.8267,
+      "step": 6914
+    },
+    {
+      "epoch": 0.12259313914480827,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.824609785313589e-05,
+      "loss": 0.8199,
+      "step": 6916
+    },
+    {
+      "epoch": 0.12262859118041984,
+      "grad_norm": 2.875,
+      "learning_rate": 4.824507007649708e-05,
+      "loss": 0.8283,
+      "step": 6918
+    },
+    {
+      "epoch": 0.12266404321603142,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.8244042009764625e-05,
+      "loss": 0.8408,
+      "step": 6920
+    },
+    {
+      "epoch": 0.12269949525164298,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.8243013652951374e-05,
+      "loss": 0.8198,
+      "step": 6922
+    },
+    {
+      "epoch": 0.12273494728725455,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.824198500607016e-05,
+      "loss": 0.8309,
+      "step": 6924
+    },
+    {
+      "epoch": 0.12277039932286613,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.82409560691338e-05,
+      "loss": 0.7655,
+      "step": 6926
+    },
+    {
+      "epoch": 0.12280585135847769,
+      "grad_norm": 3.0,
+      "learning_rate": 4.823992684215516e-05,
+      "loss": 0.8159,
+      "step": 6928
+    },
+    {
+      "epoch": 0.12284130339408926,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.823889732514707e-05,
+      "loss": 0.8199,
+      "step": 6930
+    },
+    {
+      "epoch": 0.12287675542970083,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.823786751812238e-05,
+      "loss": 0.8167,
+      "step": 6932
+    },
+    {
+      "epoch": 0.1229122074653124,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.8236837421093946e-05,
+      "loss": 0.7913,
+      "step": 6934
+    },
+    {
+      "epoch": 0.12294765950092397,
+      "grad_norm": 2.625,
+      "learning_rate": 4.823580703407462e-05,
+      "loss": 0.8211,
+      "step": 6936
+    },
+    {
+      "epoch": 0.12298311153653554,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.823477635707726e-05,
+      "loss": 0.8176,
+      "step": 6938
+    },
+    {
+      "epoch": 0.1230185635721471,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.8233745390114734e-05,
+      "loss": 0.7938,
+      "step": 6940
+    },
+    {
+      "epoch": 0.12305401560775868,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.82327141331999e-05,
+      "loss": 0.8201,
+      "step": 6942
+    },
+    {
+      "epoch": 0.12308946764337025,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.823168258634564e-05,
+      "loss": 0.8104,
+      "step": 6944
+    },
+    {
+      "epoch": 0.12312491967898181,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.823065074956481e-05,
+      "loss": 0.7937,
+      "step": 6946
+    },
+    {
+      "epoch": 0.12316037171459339,
+      "grad_norm": 2.625,
+      "learning_rate": 4.822961862287031e-05,
+      "loss": 0.8213,
+      "step": 6948
+    },
+    {
+      "epoch": 0.12319582375020496,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.8228586206274996e-05,
+      "loss": 0.8671,
+      "step": 6950
+    },
+    {
+      "epoch": 0.12323127578581652,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.8227553499791774e-05,
+      "loss": 0.8332,
+      "step": 6952
+    },
+    {
+      "epoch": 0.1232667278214281,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.8226520503433515e-05,
+      "loss": 0.7827,
+      "step": 6954
+    },
+    {
+      "epoch": 0.12330217985703967,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.8225487217213114e-05,
+      "loss": 0.8011,
+      "step": 6956
+    },
+    {
+      "epoch": 0.12333763189265123,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.822445364114349e-05,
+      "loss": 0.816,
+      "step": 6958
+    },
+    {
+      "epoch": 0.12337308392826281,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.82234197752375e-05,
+      "loss": 0.8316,
+      "step": 6960
+    },
+    {
+      "epoch": 0.12340853596387437,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.822238561950808e-05,
+      "loss": 0.8169,
+      "step": 6962
+    },
+    {
+      "epoch": 0.12344398799948594,
+      "grad_norm": 2.875,
+      "learning_rate": 4.8221351173968124e-05,
+      "loss": 0.8094,
+      "step": 6964
+    },
+    {
+      "epoch": 0.12347944003509752,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.822031643863053e-05,
+      "loss": 0.8237,
+      "step": 6966
+    },
+    {
+      "epoch": 0.12351489207070908,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.821928141350823e-05,
+      "loss": 0.8092,
+      "step": 6968
+    },
+    {
+      "epoch": 0.12355034410632065,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.821824609861414e-05,
+      "loss": 0.8337,
+      "step": 6970
+    },
+    {
+      "epoch": 0.12358579614193223,
+      "grad_norm": 2.75,
+      "learning_rate": 4.821721049396117e-05,
+      "loss": 0.8329,
+      "step": 6972
+    },
+    {
+      "epoch": 0.12362124817754379,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.821617459956225e-05,
+      "loss": 0.823,
+      "step": 6974
+    },
+    {
+      "epoch": 0.12365670021315536,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.8215138415430313e-05,
+      "loss": 0.8266,
+      "step": 6976
+    },
+    {
+      "epoch": 0.12369215224876694,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.821410194157827e-05,
+      "loss": 0.8009,
+      "step": 6978
+    },
+    {
+      "epoch": 0.1237276042843785,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.821306517801908e-05,
+      "loss": 0.8233,
+      "step": 6980
+    },
+    {
+      "epoch": 0.12376305631999007,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.821202812476567e-05,
+      "loss": 0.8174,
+      "step": 6982
+    },
+    {
+      "epoch": 0.12379850835560165,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.821099078183098e-05,
+      "loss": 0.8217,
+      "step": 6984
+    },
+    {
+      "epoch": 0.12383396039121321,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.8209953149227966e-05,
+      "loss": 0.7908,
+      "step": 6986
+    },
+    {
+      "epoch": 0.12386941242682478,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.8208915226969566e-05,
+      "loss": 0.8364,
+      "step": 6988
+    },
+    {
+      "epoch": 0.12390486446243636,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.820787701506874e-05,
+      "loss": 0.8028,
+      "step": 6990
+    },
+    {
+      "epoch": 0.12394031649804792,
+      "grad_norm": 3.0,
+      "learning_rate": 4.820683851353844e-05,
+      "loss": 0.8089,
+      "step": 6992
+    },
+    {
+      "epoch": 0.12397576853365949,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.820579972239163e-05,
+      "loss": 0.8408,
+      "step": 6994
+    },
+    {
+      "epoch": 0.12401122056927107,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.8204760641641275e-05,
+      "loss": 0.8123,
+      "step": 6996
+    },
+    {
+      "epoch": 0.12404667260488263,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.820372127130034e-05,
+      "loss": 0.7928,
+      "step": 6998
+    },
+    {
+      "epoch": 0.1240821246404942,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.8202681611381795e-05,
+      "loss": 0.8467,
+      "step": 7000
+    },
+    {
+      "epoch": 0.12411757667610578,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.8201641661898625e-05,
+      "loss": 0.8613,
+      "step": 7002
+    },
+    {
+      "epoch": 0.12415302871171734,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.820060142286379e-05,
+      "loss": 0.8444,
+      "step": 7004
+    },
+    {
+      "epoch": 0.12418848074732891,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.819956089429028e-05,
+      "loss": 0.8008,
+      "step": 7006
+    },
+    {
+      "epoch": 0.12422393278294049,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.8198520076191085e-05,
+      "loss": 0.8038,
+      "step": 7008
+    },
+    {
+      "epoch": 0.12425938481855205,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.819747896857919e-05,
+      "loss": 0.8439,
+      "step": 7010
+    },
+    {
+      "epoch": 0.12429483685416362,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.819643757146759e-05,
+      "loss": 0.8163,
+      "step": 7012
+    },
+    {
+      "epoch": 0.1243302888897752,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.819539588486929e-05,
+      "loss": 0.8019,
+      "step": 7014
+    },
+    {
+      "epoch": 0.12436574092538676,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.819435390879726e-05,
+      "loss": 0.8193,
+      "step": 7016
+    },
+    {
+      "epoch": 0.12440119296099833,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.8193311643264543e-05,
+      "loss": 0.8196,
+      "step": 7018
+    },
+    {
+      "epoch": 0.1244366449966099,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.819226908828412e-05,
+      "loss": 0.8168,
+      "step": 7020
+    },
+    {
+      "epoch": 0.12447209703222147,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.819122624386902e-05,
+      "loss": 0.8314,
+      "step": 7022
+    },
+    {
+      "epoch": 0.12450754906783304,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.819018311003223e-05,
+      "loss": 0.7862,
+      "step": 7024
+    },
+    {
+      "epoch": 0.12454300110344461,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.818913968678679e-05,
+      "loss": 0.8643,
+      "step": 7026
+    },
+    {
+      "epoch": 0.12457845313905618,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.818809597414572e-05,
+      "loss": 0.8531,
+      "step": 7028
+    },
+    {
+      "epoch": 0.12461390517466775,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.818705197212204e-05,
+      "loss": 0.8179,
+      "step": 7030
+    },
+    {
+      "epoch": 0.12464935721027932,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.818600768072878e-05,
+      "loss": 0.786,
+      "step": 7032
+    },
+    {
+      "epoch": 0.12468480924589088,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.818496309997898e-05,
+      "loss": 0.7907,
+      "step": 7034
+    },
+    {
+      "epoch": 0.12472026128150246,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.8183918229885664e-05,
+      "loss": 0.7994,
+      "step": 7036
+    },
+    {
+      "epoch": 0.12475571331711403,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.8182873070461874e-05,
+      "loss": 0.8645,
+      "step": 7038
+    },
+    {
+      "epoch": 0.1247911653527256,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.818182762172066e-05,
+      "loss": 0.8562,
+      "step": 7040
+    },
+    {
+      "epoch": 0.12482661738833717,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.818078188367506e-05,
+      "loss": 0.8527,
+      "step": 7042
+    },
+    {
+      "epoch": 0.12486206942394873,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.8179735856338144e-05,
+      "loss": 0.8573,
+      "step": 7044
+    },
+    {
+      "epoch": 0.1248975214595603,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.8178689539722946e-05,
+      "loss": 0.8259,
+      "step": 7046
+    },
+    {
+      "epoch": 0.12493297349517188,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.817764293384253e-05,
+      "loss": 0.8099,
+      "step": 7048
+    },
+    {
+      "epoch": 0.12496842553078344,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.817659603870995e-05,
+      "loss": 0.8493,
+      "step": 7050
+    },
+    {
+      "epoch": 0.12500387756639503,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.817554885433829e-05,
+      "loss": 0.8322,
+      "step": 7052
+    },
+    {
+      "epoch": 0.1250393296020066,
+      "grad_norm": 2.75,
+      "learning_rate": 4.8174501380740605e-05,
+      "loss": 0.8546,
+      "step": 7054
+    },
+    {
+      "epoch": 0.12507478163761815,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.817345361792996e-05,
+      "loss": 0.864,
+      "step": 7056
+    },
+    {
+      "epoch": 0.12511023367322974,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.8172405565919456e-05,
+      "loss": 0.7856,
+      "step": 7058
+    },
+    {
+      "epoch": 0.1251456857088413,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.8171357224722144e-05,
+      "loss": 0.8015,
+      "step": 7060
+    },
+    {
+      "epoch": 0.12518113774445286,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.817030859435113e-05,
+      "loss": 0.8243,
+      "step": 7062
+    },
+    {
+      "epoch": 0.12521658978006445,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.816925967481949e-05,
+      "loss": 0.796,
+      "step": 7064
+    },
+    {
+      "epoch": 0.125252041815676,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.816821046614031e-05,
+      "loss": 0.8522,
+      "step": 7066
+    },
+    {
+      "epoch": 0.12528749385128757,
+      "grad_norm": 2.875,
+      "learning_rate": 4.816716096832669e-05,
+      "loss": 0.7957,
+      "step": 7068
+    },
+    {
+      "epoch": 0.12532294588689916,
+      "grad_norm": 3.375,
+      "learning_rate": 4.816611118139173e-05,
+      "loss": 0.8612,
+      "step": 7070
+    },
+    {
+      "epoch": 0.12535839792251072,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.816506110534852e-05,
+      "loss": 0.8112,
+      "step": 7072
+    },
+    {
+      "epoch": 0.12539384995812228,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.8164010740210176e-05,
+      "loss": 0.7945,
+      "step": 7074
+    },
+    {
+      "epoch": 0.12542930199373387,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.8162960085989806e-05,
+      "loss": 0.8379,
+      "step": 7076
+    },
+    {
+      "epoch": 0.12546475402934543,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.816190914270051e-05,
+      "loss": 0.7771,
+      "step": 7078
+    },
+    {
+      "epoch": 0.125500206064957,
+      "grad_norm": 2.625,
+      "learning_rate": 4.816085791035543e-05,
+      "loss": 0.7982,
+      "step": 7080
+    },
+    {
+      "epoch": 0.12553565810056858,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.815980638896765e-05,
+      "loss": 0.8187,
+      "step": 7082
+    },
+    {
+      "epoch": 0.12557111013618014,
+      "grad_norm": 3.0,
+      "learning_rate": 4.8158754578550315e-05,
+      "loss": 0.8395,
+      "step": 7084
+    },
+    {
+      "epoch": 0.1256065621717917,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.815770247911655e-05,
+      "loss": 0.8151,
+      "step": 7086
+    },
+    {
+      "epoch": 0.12564201420740326,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.815665009067948e-05,
+      "loss": 0.838,
+      "step": 7088
+    },
+    {
+      "epoch": 0.12567746624301485,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.815559741325223e-05,
+      "loss": 0.8211,
+      "step": 7090
+    },
+    {
+      "epoch": 0.1257129182786264,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.815454444684796e-05,
+      "loss": 0.8336,
+      "step": 7092
+    },
+    {
+      "epoch": 0.12574837031423797,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.8153491191479795e-05,
+      "loss": 0.8184,
+      "step": 7094
+    },
+    {
+      "epoch": 0.12578382234984956,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.8152437647160884e-05,
+      "loss": 0.8088,
+      "step": 7096
+    },
+    {
+      "epoch": 0.12581927438546112,
+      "grad_norm": 2.875,
+      "learning_rate": 4.815138381390437e-05,
+      "loss": 0.8312,
+      "step": 7098
+    },
+    {
+      "epoch": 0.12585472642107268,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.81503296917234e-05,
+      "loss": 0.7815,
+      "step": 7100
+    },
+    {
+      "epoch": 0.12589017845668427,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.814927528063116e-05,
+      "loss": 0.8262,
+      "step": 7102
+    },
+    {
+      "epoch": 0.12592563049229583,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.814822058064077e-05,
+      "loss": 0.8283,
+      "step": 7104
+    },
+    {
+      "epoch": 0.1259610825279074,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.814716559176541e-05,
+      "loss": 0.8609,
+      "step": 7106
+    },
+    {
+      "epoch": 0.12599653456351897,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.8146110314018245e-05,
+      "loss": 0.8504,
+      "step": 7108
+    },
+    {
+      "epoch": 0.12603198659913054,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.814505474741244e-05,
+      "loss": 0.8211,
+      "step": 7110
+    },
+    {
+      "epoch": 0.1260674386347421,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.814399889196119e-05,
+      "loss": 0.8408,
+      "step": 7112
+    },
+    {
+      "epoch": 0.12610289067035368,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.8142942747677634e-05,
+      "loss": 0.8322,
+      "step": 7114
+    },
+    {
+      "epoch": 0.12613834270596525,
+      "grad_norm": 2.75,
+      "learning_rate": 4.814188631457498e-05,
+      "loss": 0.79,
+      "step": 7116
+    },
+    {
+      "epoch": 0.1261737947415768,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.81408295926664e-05,
+      "loss": 0.8469,
+      "step": 7118
+    },
+    {
+      "epoch": 0.1262092467771884,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.81397725819651e-05,
+      "loss": 0.8656,
+      "step": 7120
+    },
+    {
+      "epoch": 0.12624469881279995,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.813871528248425e-05,
+      "loss": 0.8063,
+      "step": 7122
+    },
+    {
+      "epoch": 0.12628015084841152,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.813765769423705e-05,
+      "loss": 0.7747,
+      "step": 7124
+    },
+    {
+      "epoch": 0.1263156028840231,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.8136599817236706e-05,
+      "loss": 0.8538,
+      "step": 7126
+    },
+    {
+      "epoch": 0.12635105491963466,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.8135541651496414e-05,
+      "loss": 0.7959,
+      "step": 7128
+    },
+    {
+      "epoch": 0.12638650695524623,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.8134483197029376e-05,
+      "loss": 0.7977,
+      "step": 7130
+    },
+    {
+      "epoch": 0.1264219589908578,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.813342445384881e-05,
+      "loss": 0.8197,
+      "step": 7132
+    },
+    {
+      "epoch": 0.12645741102646937,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.8132365421967926e-05,
+      "loss": 0.8116,
+      "step": 7134
+    },
+    {
+      "epoch": 0.12649286306208093,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.813130610139994e-05,
+      "loss": 0.8418,
+      "step": 7136
+    },
+    {
+      "epoch": 0.12652831509769252,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.813024649215807e-05,
+      "loss": 0.8424,
+      "step": 7138
+    },
+    {
+      "epoch": 0.12656376713330408,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.812918659425555e-05,
+      "loss": 0.7927,
+      "step": 7140
+    },
+    {
+      "epoch": 0.12659921916891564,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.812812640770559e-05,
+      "loss": 0.8142,
+      "step": 7142
+    },
+    {
+      "epoch": 0.12663467120452723,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.8127065932521434e-05,
+      "loss": 0.8158,
+      "step": 7144
+    },
+    {
+      "epoch": 0.1266701232401388,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.8126005168716305e-05,
+      "loss": 0.853,
+      "step": 7146
+    },
+    {
+      "epoch": 0.12670557527575035,
+      "grad_norm": 2.75,
+      "learning_rate": 4.812494411630345e-05,
+      "loss": 0.844,
+      "step": 7148
+    },
+    {
+      "epoch": 0.12674102731136194,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.8123882775296113e-05,
+      "loss": 0.84,
+      "step": 7150
+    },
+    {
+      "epoch": 0.1267764793469735,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.812282114570753e-05,
+      "loss": 0.8119,
+      "step": 7152
+    },
+    {
+      "epoch": 0.12681193138258506,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.812175922755096e-05,
+      "loss": 0.8149,
+      "step": 7154
+    },
+    {
+      "epoch": 0.12684738341819665,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.812069702083965e-05,
+      "loss": 0.8097,
+      "step": 7156
+    },
+    {
+      "epoch": 0.1268828354538082,
+      "grad_norm": 2.625,
+      "learning_rate": 4.8119634525586856e-05,
+      "loss": 0.8152,
+      "step": 7158
+    },
+    {
+      "epoch": 0.12691828748941977,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.811857174180584e-05,
+      "loss": 0.8176,
+      "step": 7160
+    },
+    {
+      "epoch": 0.12695373952503136,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.811750866950986e-05,
+      "loss": 0.804,
+      "step": 7162
+    },
+    {
+      "epoch": 0.12698919156064292,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.81164453087122e-05,
+      "loss": 0.8052,
+      "step": 7164
+    },
+    {
+      "epoch": 0.12702464359625448,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.8115381659426105e-05,
+      "loss": 0.8712,
+      "step": 7166
+    },
+    {
+      "epoch": 0.12706009563186607,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.811431772166486e-05,
+      "loss": 0.8033,
+      "step": 7168
+    },
+    {
+      "epoch": 0.12709554766747763,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.8113253495441745e-05,
+      "loss": 0.8239,
+      "step": 7170
+    },
+    {
+      "epoch": 0.1271309997030892,
+      "grad_norm": 2.875,
+      "learning_rate": 4.811218898077005e-05,
+      "loss": 0.8461,
+      "step": 7172
+    },
+    {
+      "epoch": 0.12716645173870078,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.811112417766304e-05,
+      "loss": 0.8235,
+      "step": 7174
+    },
+    {
+      "epoch": 0.12720190377431234,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.811005908613402e-05,
+      "loss": 0.8135,
+      "step": 7176
+    },
+    {
+      "epoch": 0.1272373558099239,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.810899370619627e-05,
+      "loss": 0.8182,
+      "step": 7178
+    },
+    {
+      "epoch": 0.1272728078455355,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.81079280378631e-05,
+      "loss": 0.812,
+      "step": 7180
+    },
+    {
+      "epoch": 0.12730825988114705,
+      "grad_norm": 2.875,
+      "learning_rate": 4.81068620811478e-05,
+      "loss": 0.8564,
+      "step": 7182
+    },
+    {
+      "epoch": 0.1273437119167586,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.810579583606367e-05,
+      "loss": 0.8287,
+      "step": 7184
+    },
+    {
+      "epoch": 0.1273791639523702,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.810472930262402e-05,
+      "loss": 0.8113,
+      "step": 7186
+    },
+    {
+      "epoch": 0.12741461598798176,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.810366248084216e-05,
+      "loss": 0.8531,
+      "step": 7188
+    },
+    {
+      "epoch": 0.12745006802359332,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.810259537073141e-05,
+      "loss": 0.7819,
+      "step": 7190
+    },
+    {
+      "epoch": 0.1274855200592049,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.8101527972305075e-05,
+      "loss": 0.8039,
+      "step": 7192
+    },
+    {
+      "epoch": 0.12752097209481647,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.810046028557649e-05,
+      "loss": 0.8555,
+      "step": 7194
+    },
+    {
+      "epoch": 0.12755642413042803,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.8099392310558966e-05,
+      "loss": 0.8321,
+      "step": 7196
+    },
+    {
+      "epoch": 0.12759187616603962,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.809832404726584e-05,
+      "loss": 0.8268,
+      "step": 7198
+    },
+    {
+      "epoch": 0.12762732820165118,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.8097255495710435e-05,
+      "loss": 0.845,
+      "step": 7200
+    },
+    {
+      "epoch": 0.12766278023726274,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.80961866559061e-05,
+      "loss": 0.8416,
+      "step": 7202
+    },
+    {
+      "epoch": 0.12769823227287433,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.809511752786616e-05,
+      "loss": 0.8385,
+      "step": 7204
+    },
+    {
+      "epoch": 0.1277336843084859,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.809404811160397e-05,
+      "loss": 0.7694,
+      "step": 7206
+    },
+    {
+      "epoch": 0.12776913634409745,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.809297840713287e-05,
+      "loss": 0.8489,
+      "step": 7208
+    },
+    {
+      "epoch": 0.12780458837970904,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.8091908414466206e-05,
+      "loss": 0.88,
+      "step": 7210
+    },
+    {
+      "epoch": 0.1278400404153206,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.8090838133617334e-05,
+      "loss": 0.8352,
+      "step": 7212
+    },
+    {
+      "epoch": 0.12787549245093216,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.808976756459961e-05,
+      "loss": 0.8255,
+      "step": 7214
+    },
+    {
+      "epoch": 0.12791094448654375,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.80886967074264e-05,
+      "loss": 0.8162,
+      "step": 7216
+    },
+    {
+      "epoch": 0.1279463965221553,
+      "grad_norm": 2.625,
+      "learning_rate": 4.808762556211106e-05,
+      "loss": 0.823,
+      "step": 7218
+    },
+    {
+      "epoch": 0.12798184855776687,
+      "grad_norm": 2.625,
+      "learning_rate": 4.808655412866697e-05,
+      "loss": 0.7868,
+      "step": 7220
+    },
+    {
+      "epoch": 0.12801730059337846,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.8085482407107483e-05,
+      "loss": 0.8302,
+      "step": 7222
+    },
+    {
+      "epoch": 0.12805275262899002,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.808441039744599e-05,
+      "loss": 0.7836,
+      "step": 7224
+    },
+    {
+      "epoch": 0.12808820466460158,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.8083338099695864e-05,
+      "loss": 0.8161,
+      "step": 7226
+    },
+    {
+      "epoch": 0.12812365670021317,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.8082265513870484e-05,
+      "loss": 0.7852,
+      "step": 7228
+    },
+    {
+      "epoch": 0.12815910873582473,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.808119263998324e-05,
+      "loss": 0.8264,
+      "step": 7230
+    },
+    {
+      "epoch": 0.1281945607714363,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.808011947804751e-05,
+      "loss": 0.7741,
+      "step": 7232
+    },
+    {
+      "epoch": 0.12823001280704788,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.807904602807671e-05,
+      "loss": 0.8024,
+      "step": 7234
+    },
+    {
+      "epoch": 0.12826546484265944,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.807797229008422e-05,
+      "loss": 0.8464,
+      "step": 7236
+    },
+    {
+      "epoch": 0.128300916878271,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.8076898264083435e-05,
+      "loss": 0.8072,
+      "step": 7238
+    },
+    {
+      "epoch": 0.1283363689138826,
+      "grad_norm": 2.875,
+      "learning_rate": 4.8075823950087774e-05,
+      "loss": 0.784,
+      "step": 7240
+    },
+    {
+      "epoch": 0.12837182094949415,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.807474934811063e-05,
+      "loss": 0.8259,
+      "step": 7242
+    },
+    {
+      "epoch": 0.1284072729851057,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.8073674458165416e-05,
+      "loss": 0.7873,
+      "step": 7244
+    },
+    {
+      "epoch": 0.1284427250207173,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.8072599280265565e-05,
+      "loss": 0.8289,
+      "step": 7246
+    },
+    {
+      "epoch": 0.12847817705632886,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.807152381442447e-05,
+      "loss": 0.8145,
+      "step": 7248
+    },
+    {
+      "epoch": 0.12851362909194042,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.807044806065557e-05,
+      "loss": 0.8007,
+      "step": 7250
+    },
+    {
+      "epoch": 0.128549081127552,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.806937201897228e-05,
+      "loss": 0.8343,
+      "step": 7252
+    },
+    {
+      "epoch": 0.12858453316316357,
+      "grad_norm": 2.875,
+      "learning_rate": 4.8068295689388035e-05,
+      "loss": 0.8283,
+      "step": 7254
+    },
+    {
+      "epoch": 0.12861998519877513,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.806721907191626e-05,
+      "loss": 0.8135,
+      "step": 7256
+    },
+    {
+      "epoch": 0.1286554372343867,
+      "grad_norm": 2.625,
+      "learning_rate": 4.8066142166570397e-05,
+      "loss": 0.8934,
+      "step": 7258
+    },
+    {
+      "epoch": 0.12869088926999828,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.806506497336388e-05,
+      "loss": 0.8221,
+      "step": 7260
+    },
+    {
+      "epoch": 0.12872634130560984,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.8063987492310156e-05,
+      "loss": 0.8494,
+      "step": 7262
+    },
+    {
+      "epoch": 0.1287617933412214,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.806290972342268e-05,
+      "loss": 0.7975,
+      "step": 7264
+    },
+    {
+      "epoch": 0.128797245376833,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.806183166671489e-05,
+      "loss": 0.801,
+      "step": 7266
+    },
+    {
+      "epoch": 0.12883269741244455,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.8060753322200244e-05,
+      "loss": 0.8517,
+      "step": 7268
+    },
+    {
+      "epoch": 0.1288681494480561,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.80596746898922e-05,
+      "loss": 0.8417,
+      "step": 7270
+    },
+    {
+      "epoch": 0.1289036014836677,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.8058595769804224e-05,
+      "loss": 0.8524,
+      "step": 7272
+    },
+    {
+      "epoch": 0.12893905351927926,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.805751656194977e-05,
+      "loss": 0.817,
+      "step": 7274
+    },
+    {
+      "epoch": 0.12897450555489082,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.8056437066342315e-05,
+      "loss": 0.8042,
+      "step": 7276
+    },
+    {
+      "epoch": 0.1290099575905024,
+      "grad_norm": 2.625,
+      "learning_rate": 4.805535728299533e-05,
+      "loss": 0.8078,
+      "step": 7278
+    },
+    {
+      "epoch": 0.12904540962611397,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.805427721192228e-05,
+      "loss": 0.8814,
+      "step": 7280
+    },
+    {
+      "epoch": 0.12908086166172553,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.805319685313666e-05,
+      "loss": 0.8336,
+      "step": 7282
+    },
+    {
+      "epoch": 0.12911631369733712,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.805211620665194e-05,
+      "loss": 0.8131,
+      "step": 7284
+    },
+    {
+      "epoch": 0.12915176573294868,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.805103527248161e-05,
+      "loss": 0.8335,
+      "step": 7286
+    },
+    {
+      "epoch": 0.12918721776856024,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.804995405063916e-05,
+      "loss": 0.8351,
+      "step": 7288
+    },
+    {
+      "epoch": 0.12922266980417182,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.804887254113809e-05,
+      "loss": 0.8125,
+      "step": 7290
+    },
+    {
+      "epoch": 0.12925812183978339,
+      "grad_norm": 2.75,
+      "learning_rate": 4.804779074399189e-05,
+      "loss": 0.8221,
+      "step": 7292
+    },
+    {
+      "epoch": 0.12929357387539495,
+      "grad_norm": 3.0,
+      "learning_rate": 4.8046708659214054e-05,
+      "loss": 0.8297,
+      "step": 7294
+    },
+    {
+      "epoch": 0.12932902591100653,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.8045626286818106e-05,
+      "loss": 0.8106,
+      "step": 7296
+    },
+    {
+      "epoch": 0.1293644779466181,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.804454362681754e-05,
+      "loss": 0.8042,
+      "step": 7298
+    },
+    {
+      "epoch": 0.12939992998222966,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.804346067922587e-05,
+      "loss": 0.821,
+      "step": 7300
+    },
+    {
+      "epoch": 0.12943538201784124,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.80423774440566e-05,
+      "loss": 0.7792,
+      "step": 7302
+    },
+    {
+      "epoch": 0.1294708340534528,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.804129392132327e-05,
+      "loss": 0.8335,
+      "step": 7304
+    },
+    {
+      "epoch": 0.12950628608906437,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.804021011103939e-05,
+      "loss": 0.8099,
+      "step": 7306
+    },
+    {
+      "epoch": 0.12954173812467595,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.8039126013218474e-05,
+      "loss": 0.8458,
+      "step": 7308
+    },
+    {
+      "epoch": 0.12957719016028751,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.803804162787408e-05,
+      "loss": 0.8252,
+      "step": 7310
+    },
+    {
+      "epoch": 0.12961264219589907,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.803695695501972e-05,
+      "loss": 0.8265,
+      "step": 7312
+    },
+    {
+      "epoch": 0.12964809423151066,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.803587199466893e-05,
+      "loss": 0.8698,
+      "step": 7314
+    },
+    {
+      "epoch": 0.12968354626712222,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.8034786746835256e-05,
+      "loss": 0.8338,
+      "step": 7316
+    },
+    {
+      "epoch": 0.12971899830273378,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.803370121153225e-05,
+      "loss": 0.861,
+      "step": 7318
+    },
+    {
+      "epoch": 0.12975445033834537,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.8032615388773445e-05,
+      "loss": 0.8264,
+      "step": 7320
+    },
+    {
+      "epoch": 0.12978990237395693,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.8031529278572394e-05,
+      "loss": 0.839,
+      "step": 7322
+    },
+    {
+      "epoch": 0.1298253544095685,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.803044288094266e-05,
+      "loss": 0.8518,
+      "step": 7324
+    },
+    {
+      "epoch": 0.12986080644518008,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.802935619589779e-05,
+      "loss": 0.8324,
+      "step": 7326
+    },
+    {
+      "epoch": 0.12989625848079164,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.802826922345136e-05,
+      "loss": 0.7853,
+      "step": 7328
+    },
+    {
+      "epoch": 0.1299317105164032,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.802718196361692e-05,
+      "loss": 0.8865,
+      "step": 7330
+    },
+    {
+      "epoch": 0.1299671625520148,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.802609441640805e-05,
+      "loss": 0.8392,
+      "step": 7332
+    },
+    {
+      "epoch": 0.13000261458762635,
+      "grad_norm": 3.0,
+      "learning_rate": 4.8025006581838314e-05,
+      "loss": 0.8358,
+      "step": 7334
+    },
+    {
+      "epoch": 0.1300380666232379,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.802391845992129e-05,
+      "loss": 0.8317,
+      "step": 7336
+    },
+    {
+      "epoch": 0.1300735186588495,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.802283005067057e-05,
+      "loss": 0.8186,
+      "step": 7338
+    },
+    {
+      "epoch": 0.13010897069446106,
+      "grad_norm": 2.75,
+      "learning_rate": 4.802174135409971e-05,
+      "loss": 0.8265,
+      "step": 7340
+    },
+    {
+      "epoch": 0.13014442273007262,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.802065237022233e-05,
+      "loss": 0.822,
+      "step": 7342
+    },
+    {
+      "epoch": 0.1301798747656842,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.801956309905199e-05,
+      "loss": 0.8397,
+      "step": 7344
+    },
+    {
+      "epoch": 0.13021532680129577,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.80184735406023e-05,
+      "loss": 0.7997,
+      "step": 7346
+    },
+    {
+      "epoch": 0.13025077883690733,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.8017383694886855e-05,
+      "loss": 0.7548,
+      "step": 7348
+    },
+    {
+      "epoch": 0.13028623087251892,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.8016293561919256e-05,
+      "loss": 0.85,
+      "step": 7350
+    },
+    {
+      "epoch": 0.13032168290813048,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.8015203141713114e-05,
+      "loss": 0.8169,
+      "step": 7352
+    },
+    {
+      "epoch": 0.13035713494374204,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.801411243428202e-05,
+      "loss": 0.7821,
+      "step": 7354
+    },
+    {
+      "epoch": 0.13039258697935363,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.80130214396396e-05,
+      "loss": 0.819,
+      "step": 7356
+    },
+    {
+      "epoch": 0.1304280390149652,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.801193015779947e-05,
+      "loss": 0.8274,
+      "step": 7358
+    },
+    {
+      "epoch": 0.13046349105057675,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.801083858877524e-05,
+      "loss": 0.8147,
+      "step": 7360
+    },
+    {
+      "epoch": 0.13049894308618834,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.800974673258054e-05,
+      "loss": 0.8461,
+      "step": 7362
+    },
+    {
+      "epoch": 0.1305343951217999,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.8008654589228984e-05,
+      "loss": 0.8115,
+      "step": 7364
+    },
+    {
+      "epoch": 0.13056984715741146,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.800756215873422e-05,
+      "loss": 0.7553,
+      "step": 7366
+    },
+    {
+      "epoch": 0.13060529919302305,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.8006469441109874e-05,
+      "loss": 0.7717,
+      "step": 7368
+    },
+    {
+      "epoch": 0.1306407512286346,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.8005376436369576e-05,
+      "loss": 0.7747,
+      "step": 7370
+    },
+    {
+      "epoch": 0.13067620326424617,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.800428314452697e-05,
+      "loss": 0.7978,
+      "step": 7372
+    },
+    {
+      "epoch": 0.13071165529985776,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.800318956559571e-05,
+      "loss": 0.8016,
+      "step": 7374
+    },
+    {
+      "epoch": 0.13074710733546932,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.800209569958943e-05,
+      "loss": 0.8473,
+      "step": 7376
+    },
+    {
+      "epoch": 0.13078255937108088,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.80010015465218e-05,
+      "loss": 0.8325,
+      "step": 7378
+    },
+    {
+      "epoch": 0.13081801140669247,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.799990710640645e-05,
+      "loss": 0.8087,
+      "step": 7380
+    },
+    {
+      "epoch": 0.13085346344230403,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.799881237925704e-05,
+      "loss": 0.7808,
+      "step": 7382
+    },
+    {
+      "epoch": 0.1308889154779156,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.799771736508725e-05,
+      "loss": 0.8534,
+      "step": 7384
+    },
+    {
+      "epoch": 0.13092436751352718,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.7996622063910744e-05,
+      "loss": 0.7919,
+      "step": 7386
+    },
+    {
+      "epoch": 0.13095981954913874,
+      "grad_norm": 3.25,
+      "learning_rate": 4.7995526475741174e-05,
+      "loss": 0.8613,
+      "step": 7388
+    },
+    {
+      "epoch": 0.1309952715847503,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.799443060059223e-05,
+      "loss": 0.8076,
+      "step": 7390
+    },
+    {
+      "epoch": 0.1310307236203619,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.7993334438477576e-05,
+      "loss": 0.8378,
+      "step": 7392
+    },
+    {
+      "epoch": 0.13106617565597345,
+      "grad_norm": 2.625,
+      "learning_rate": 4.7992237989410904e-05,
+      "loss": 0.7977,
+      "step": 7394
+    },
+    {
+      "epoch": 0.131101627691585,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.799114125340589e-05,
+      "loss": 0.8135,
+      "step": 7396
+    },
+    {
+      "epoch": 0.1311370797271966,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.7990044230476215e-05,
+      "loss": 0.8572,
+      "step": 7398
+    },
+    {
+      "epoch": 0.13117253176280816,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.798894692063559e-05,
+      "loss": 0.8426,
+      "step": 7400
+    },
+    {
+      "epoch": 0.13120798379841972,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.798784932389768e-05,
+      "loss": 0.8142,
+      "step": 7402
+    },
+    {
+      "epoch": 0.1312434358340313,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.798675144027621e-05,
+      "loss": 0.845,
+      "step": 7404
+    },
+    {
+      "epoch": 0.13127888786964287,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.798565326978486e-05,
+      "loss": 0.7994,
+      "step": 7406
+    },
+    {
+      "epoch": 0.13131433990525443,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.798455481243735e-05,
+      "loss": 0.8078,
+      "step": 7408
+    },
+    {
+      "epoch": 0.13134979194086602,
+      "grad_norm": 2.25,
+      "learning_rate": 4.798345606824739e-05,
+      "loss": 0.7955,
+      "step": 7410
+    },
+    {
+      "epoch": 0.13138524397647758,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.7982357037228676e-05,
+      "loss": 0.7976,
+      "step": 7412
+    },
+    {
+      "epoch": 0.13142069601208914,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.798125771939493e-05,
+      "loss": 0.8133,
+      "step": 7414
+    },
+    {
+      "epoch": 0.13145614804770073,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.798015811475989e-05,
+      "loss": 0.7703,
+      "step": 7416
+    },
+    {
+      "epoch": 0.1314916000833123,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.7979058223337246e-05,
+      "loss": 0.8613,
+      "step": 7418
+    },
+    {
+      "epoch": 0.13152705211892385,
+      "grad_norm": 2.75,
+      "learning_rate": 4.797795804514075e-05,
+      "loss": 0.81,
+      "step": 7420
+    },
+    {
+      "epoch": 0.13156250415453544,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.797685758018413e-05,
+      "loss": 0.7829,
+      "step": 7422
+    },
+    {
+      "epoch": 0.131597956190147,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.79757568284811e-05,
+      "loss": 0.8287,
+      "step": 7424
+    },
+    {
+      "epoch": 0.13163340822575856,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.797465579004542e-05,
+      "loss": 0.8734,
+      "step": 7426
+    },
+    {
+      "epoch": 0.13166886026137012,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.797355446489081e-05,
+      "loss": 0.8128,
+      "step": 7428
+    },
+    {
+      "epoch": 0.1317043122969817,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.7972452853031035e-05,
+      "loss": 0.7994,
+      "step": 7430
+    },
+    {
+      "epoch": 0.13173976433259327,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.797135095447983e-05,
+      "loss": 0.7838,
+      "step": 7432
+    },
+    {
+      "epoch": 0.13177521636820483,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.797024876925095e-05,
+      "loss": 0.7476,
+      "step": 7434
+    },
+    {
+      "epoch": 0.13181066840381642,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.796914629735815e-05,
+      "loss": 0.8098,
+      "step": 7436
+    },
+    {
+      "epoch": 0.13184612043942798,
+      "grad_norm": 2.625,
+      "learning_rate": 4.796804353881519e-05,
+      "loss": 0.8402,
+      "step": 7438
+    },
+    {
+      "epoch": 0.13188157247503954,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.7966940493635825e-05,
+      "loss": 0.8256,
+      "step": 7440
+    },
+    {
+      "epoch": 0.13191702451065113,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.796583716183383e-05,
+      "loss": 0.8504,
+      "step": 7442
+    },
+    {
+      "epoch": 0.1319524765462627,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.7964733543422975e-05,
+      "loss": 0.7709,
+      "step": 7444
+    },
+    {
+      "epoch": 0.13198792858187425,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.7963629638417015e-05,
+      "loss": 0.7819,
+      "step": 7446
+    },
+    {
+      "epoch": 0.13202338061748584,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.7962525446829757e-05,
+      "loss": 0.825,
+      "step": 7448
+    },
+    {
+      "epoch": 0.1320588326530974,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.7961420968674955e-05,
+      "loss": 0.8014,
+      "step": 7450
+    },
+    {
+      "epoch": 0.13209428468870896,
+      "grad_norm": 2.5,
+      "learning_rate": 4.796031620396641e-05,
+      "loss": 0.7856,
+      "step": 7452
+    },
+    {
+      "epoch": 0.13212973672432055,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.795921115271789e-05,
+      "loss": 0.8327,
+      "step": 7454
+    },
+    {
+      "epoch": 0.1321651887599321,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.79581058149432e-05,
+      "loss": 0.8167,
+      "step": 7456
+    },
+    {
+      "epoch": 0.13220064079554367,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.7957000190656134e-05,
+      "loss": 0.7662,
+      "step": 7458
+    },
+    {
+      "epoch": 0.13223609283115526,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.7955894279870483e-05,
+      "loss": 0.8741,
+      "step": 7460
+    },
+    {
+      "epoch": 0.13227154486676682,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.7954788082600055e-05,
+      "loss": 0.808,
+      "step": 7462
+    },
+    {
+      "epoch": 0.13230699690237838,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.795368159885866e-05,
+      "loss": 0.7918,
+      "step": 7464
+    },
+    {
+      "epoch": 0.13234244893798996,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.7952574828660086e-05,
+      "loss": 0.8284,
+      "step": 7466
+    },
+    {
+      "epoch": 0.13237790097360153,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.7951467772018164e-05,
+      "loss": 0.8582,
+      "step": 7468
+    },
+    {
+      "epoch": 0.1324133530092131,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.7950360428946705e-05,
+      "loss": 0.8033,
+      "step": 7470
+    },
+    {
+      "epoch": 0.13244880504482467,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.794925279945953e-05,
+      "loss": 0.8375,
+      "step": 7472
+    },
+    {
+      "epoch": 0.13248425708043624,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.794814488357046e-05,
+      "loss": 0.8369,
+      "step": 7474
+    },
+    {
+      "epoch": 0.1325197091160478,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.7947036681293325e-05,
+      "loss": 0.787,
+      "step": 7476
+    },
+    {
+      "epoch": 0.13255516115165938,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.7945928192641944e-05,
+      "loss": 0.8105,
+      "step": 7478
+    },
+    {
+      "epoch": 0.13259061318727094,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.7944819417630165e-05,
+      "loss": 0.8264,
+      "step": 7480
+    },
+    {
+      "epoch": 0.1326260652228825,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.7943710356271816e-05,
+      "loss": 0.8561,
+      "step": 7482
+    },
+    {
+      "epoch": 0.1326615172584941,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.794260100858074e-05,
+      "loss": 0.7978,
+      "step": 7484
+    },
+    {
+      "epoch": 0.13269696929410565,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.794149137457078e-05,
+      "loss": 0.8171,
+      "step": 7486
+    },
+    {
+      "epoch": 0.13273242132971722,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.794038145425579e-05,
+      "loss": 0.8441,
+      "step": 7488
+    },
+    {
+      "epoch": 0.1327678733653288,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.793927124764962e-05,
+      "loss": 0.7825,
+      "step": 7490
+    },
+    {
+      "epoch": 0.13280332540094036,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.7938160754766114e-05,
+      "loss": 0.837,
+      "step": 7492
+    },
+    {
+      "epoch": 0.13283877743655192,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.793704997561915e-05,
+      "loss": 0.8297,
+      "step": 7494
+    },
+    {
+      "epoch": 0.1328742294721635,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.793593891022257e-05,
+      "loss": 0.7936,
+      "step": 7496
+    },
+    {
+      "epoch": 0.13290968150777507,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.793482755859026e-05,
+      "loss": 0.8175,
+      "step": 7498
+    },
+    {
+      "epoch": 0.13294513354338663,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.793371592073607e-05,
+      "loss": 0.826,
+      "step": 7500
+    },
+    {
+      "epoch": 0.13298058557899822,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.793260399667388e-05,
+      "loss": 0.7818,
+      "step": 7502
+    },
+    {
+      "epoch": 0.13301603761460978,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.793149178641758e-05,
+      "loss": 0.8198,
+      "step": 7504
+    },
+    {
+      "epoch": 0.13305148965022134,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.793037928998103e-05,
+      "loss": 0.8036,
+      "step": 7506
+    },
+    {
+      "epoch": 0.13308694168583293,
+      "grad_norm": 2.75,
+      "learning_rate": 4.7929266507378125e-05,
+      "loss": 0.8362,
+      "step": 7508
+    },
+    {
+      "epoch": 0.1331223937214445,
+      "grad_norm": 3.0,
+      "learning_rate": 4.792815343862275e-05,
+      "loss": 0.8377,
+      "step": 7510
+    },
+    {
+      "epoch": 0.13315784575705605,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.792704008372879e-05,
+      "loss": 0.7947,
+      "step": 7512
+    },
+    {
+      "epoch": 0.13319329779266764,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.792592644271015e-05,
+      "loss": 0.7883,
+      "step": 7514
+    },
+    {
+      "epoch": 0.1332287498282792,
+      "grad_norm": 2.75,
+      "learning_rate": 4.792481251558073e-05,
+      "loss": 0.8371,
+      "step": 7516
+    },
+    {
+      "epoch": 0.13326420186389076,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.792369830235441e-05,
+      "loss": 0.8336,
+      "step": 7518
+    },
+    {
+      "epoch": 0.13329965389950235,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.792258380304512e-05,
+      "loss": 0.8234,
+      "step": 7520
+    },
+    {
+      "epoch": 0.1333351059351139,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.7921469017666756e-05,
+      "loss": 0.7981,
+      "step": 7522
+    },
+    {
+      "epoch": 0.13337055797072547,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.792035394623323e-05,
+      "loss": 0.8175,
+      "step": 7524
+    },
+    {
+      "epoch": 0.13340601000633706,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.791923858875847e-05,
+      "loss": 0.8514,
+      "step": 7526
+    },
+    {
+      "epoch": 0.13344146204194862,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.791812294525638e-05,
+      "loss": 0.8058,
+      "step": 7528
+    },
+    {
+      "epoch": 0.13347691407756018,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.791700701574089e-05,
+      "loss": 0.843,
+      "step": 7530
+    },
+    {
+      "epoch": 0.13351236611317177,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.7915890800225926e-05,
+      "loss": 0.8108,
+      "step": 7532
+    },
+    {
+      "epoch": 0.13354781814878333,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.791477429872542e-05,
+      "loss": 0.8333,
+      "step": 7534
+    },
+    {
+      "epoch": 0.1335832701843949,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.7913657511253296e-05,
+      "loss": 0.835,
+      "step": 7536
+    },
+    {
+      "epoch": 0.13361872222000648,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.7912540437823506e-05,
+      "loss": 0.8367,
+      "step": 7538
+    },
+    {
+      "epoch": 0.13365417425561804,
+      "grad_norm": 2.75,
+      "learning_rate": 4.791142307844998e-05,
+      "loss": 0.8028,
+      "step": 7540
+    },
+    {
+      "epoch": 0.1336896262912296,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.7910305433146664e-05,
+      "loss": 0.7917,
+      "step": 7542
+    },
+    {
+      "epoch": 0.1337250783268412,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.790918750192751e-05,
+      "loss": 0.7507,
+      "step": 7544
+    },
+    {
+      "epoch": 0.13376053036245275,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.790806928480647e-05,
+      "loss": 0.8264,
+      "step": 7546
+    },
+    {
+      "epoch": 0.1337959823980643,
+      "grad_norm": 2.625,
+      "learning_rate": 4.79069507817975e-05,
+      "loss": 0.787,
+      "step": 7548
+    },
+    {
+      "epoch": 0.1338314344336759,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.790583199291455e-05,
+      "loss": 0.8439,
+      "step": 7550
+    },
+    {
+      "epoch": 0.13386688646928746,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.790471291817159e-05,
+      "loss": 0.8312,
+      "step": 7552
+    },
+    {
+      "epoch": 0.13390233850489902,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.790359355758258e-05,
+      "loss": 0.8275,
+      "step": 7554
+    },
+    {
+      "epoch": 0.1339377905405106,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.79024739111615e-05,
+      "loss": 0.807,
+      "step": 7556
+    },
+    {
+      "epoch": 0.13397324257612217,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.7901353978922306e-05,
+      "loss": 0.8654,
+      "step": 7558
+    },
+    {
+      "epoch": 0.13400869461173373,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.7900233760878986e-05,
+      "loss": 0.8133,
+      "step": 7560
+    },
+    {
+      "epoch": 0.13404414664734532,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.789911325704552e-05,
+      "loss": 0.7941,
+      "step": 7562
+    },
+    {
+      "epoch": 0.13407959868295688,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.789799246743589e-05,
+      "loss": 0.8172,
+      "step": 7564
+    },
+    {
+      "epoch": 0.13411505071856844,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.789687139206409e-05,
+      "loss": 0.8277,
+      "step": 7566
+    },
+    {
+      "epoch": 0.13415050275418003,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.78957500309441e-05,
+      "loss": 0.8077,
+      "step": 7568
+    },
+    {
+      "epoch": 0.1341859547897916,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.789462838408991e-05,
+      "loss": 0.8342,
+      "step": 7570
+    },
+    {
+      "epoch": 0.13422140682540315,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.789350645151554e-05,
+      "loss": 0.8314,
+      "step": 7572
+    },
+    {
+      "epoch": 0.13425685886101474,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.789238423323497e-05,
+      "loss": 0.829,
+      "step": 7574
+    },
+    {
+      "epoch": 0.1342923108966263,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.789126172926222e-05,
+      "loss": 0.8237,
+      "step": 7576
+    },
+    {
+      "epoch": 0.13432776293223786,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.7890138939611285e-05,
+      "loss": 0.8177,
+      "step": 7578
+    },
+    {
+      "epoch": 0.13436321496784945,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.788901586429618e-05,
+      "loss": 0.8511,
+      "step": 7580
+    },
+    {
+      "epoch": 0.134398667003461,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.7887892503330936e-05,
+      "loss": 0.8095,
+      "step": 7582
+    },
+    {
+      "epoch": 0.13443411903907257,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.7886768856729546e-05,
+      "loss": 0.8151,
+      "step": 7584
+    },
+    {
+      "epoch": 0.13446957107468416,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.788564492450606e-05,
+      "loss": 0.8282,
+      "step": 7586
+    },
+    {
+      "epoch": 0.13450502311029572,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.7884520706674485e-05,
+      "loss": 0.7904,
+      "step": 7588
+    },
+    {
+      "epoch": 0.13454047514590728,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.788339620324887e-05,
+      "loss": 0.8623,
+      "step": 7590
+    },
+    {
+      "epoch": 0.13457592718151887,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.788227141424322e-05,
+      "loss": 0.8727,
+      "step": 7592
+    },
+    {
+      "epoch": 0.13461137921713043,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.78811463396716e-05,
+      "loss": 0.7833,
+      "step": 7594
+    },
+    {
+      "epoch": 0.134646831252742,
+      "grad_norm": 2.75,
+      "learning_rate": 4.788002097954804e-05,
+      "loss": 0.8188,
+      "step": 7596
+    },
+    {
+      "epoch": 0.13468228328835358,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.787889533388658e-05,
+      "loss": 0.8437,
+      "step": 7598
+    },
+    {
+      "epoch": 0.13471773532396514,
+      "grad_norm": 2.75,
+      "learning_rate": 4.787776940270127e-05,
+      "loss": 0.788,
+      "step": 7600
+    },
+    {
+      "epoch": 0.1347531873595767,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.787664318600615e-05,
+      "loss": 0.8419,
+      "step": 7602
+    },
+    {
+      "epoch": 0.13478863939518826,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.787551668381531e-05,
+      "loss": 0.7876,
+      "step": 7604
+    },
+    {
+      "epoch": 0.13482409143079985,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.787438989614278e-05,
+      "loss": 0.8196,
+      "step": 7606
+    },
+    {
+      "epoch": 0.1348595434664114,
+      "grad_norm": 2.875,
+      "learning_rate": 4.7873262823002627e-05,
+      "loss": 0.8218,
+      "step": 7608
+    },
+    {
+      "epoch": 0.13489499550202297,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.787213546440892e-05,
+      "loss": 0.8086,
+      "step": 7610
+    },
+    {
+      "epoch": 0.13493044753763456,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.7871007820375725e-05,
+      "loss": 0.7893,
+      "step": 7612
+    },
+    {
+      "epoch": 0.13496589957324612,
+      "grad_norm": 2.875,
+      "learning_rate": 4.786987989091711e-05,
+      "loss": 0.8129,
+      "step": 7614
+    },
+    {
+      "epoch": 0.13500135160885768,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.786875167604716e-05,
+      "loss": 0.8359,
+      "step": 7616
+    },
+    {
+      "epoch": 0.13503680364446927,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.7867623175779955e-05,
+      "loss": 0.8402,
+      "step": 7618
+    },
+    {
+      "epoch": 0.13507225568008083,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.786649439012958e-05,
+      "loss": 0.8359,
+      "step": 7620
+    },
+    {
+      "epoch": 0.1351077077156924,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.786536531911011e-05,
+      "loss": 0.8427,
+      "step": 7622
+    },
+    {
+      "epoch": 0.13514315975130398,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.7864235962735646e-05,
+      "loss": 0.8282,
+      "step": 7624
+    },
+    {
+      "epoch": 0.13517861178691554,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.7863106321020285e-05,
+      "loss": 0.7969,
+      "step": 7626
+    },
+    {
+      "epoch": 0.1352140638225271,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.7861976393978115e-05,
+      "loss": 0.7827,
+      "step": 7628
+    },
+    {
+      "epoch": 0.13524951585813869,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.7860846181623244e-05,
+      "loss": 0.8338,
+      "step": 7630
+    },
+    {
+      "epoch": 0.13528496789375025,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.785971568396977e-05,
+      "loss": 0.8179,
+      "step": 7632
+    },
+    {
+      "epoch": 0.1353204199293618,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.78585849010318e-05,
+      "loss": 0.8643,
+      "step": 7634
+    },
+    {
+      "epoch": 0.1353558719649734,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.785745383282346e-05,
+      "loss": 0.8289,
+      "step": 7636
+    },
+    {
+      "epoch": 0.13539132400058496,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.785632247935886e-05,
+      "loss": 0.8129,
+      "step": 7638
+    },
+    {
+      "epoch": 0.13542677603619652,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.785519084065211e-05,
+      "loss": 0.8034,
+      "step": 7640
+    },
+    {
+      "epoch": 0.1354622280718081,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.7854058916717336e-05,
+      "loss": 0.8267,
+      "step": 7642
+    },
+    {
+      "epoch": 0.13549768010741967,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.7852926707568676e-05,
+      "loss": 0.7633,
+      "step": 7644
+    },
+    {
+      "epoch": 0.13553313214303123,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.785179421322025e-05,
+      "loss": 0.8112,
+      "step": 7646
+    },
+    {
+      "epoch": 0.13556858417864281,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.785066143368618e-05,
+      "loss": 0.8385,
+      "step": 7648
+    },
+    {
+      "epoch": 0.13560403621425438,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.784952836898062e-05,
+      "loss": 0.8185,
+      "step": 7650
+    },
+    {
+      "epoch": 0.13563948824986594,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.784839501911771e-05,
+      "loss": 0.8011,
+      "step": 7652
+    },
+    {
+      "epoch": 0.13567494028547752,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.7847261384111585e-05,
+      "loss": 0.8231,
+      "step": 7654
+    },
+    {
+      "epoch": 0.13571039232108909,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.7846127463976395e-05,
+      "loss": 0.7639,
+      "step": 7656
+    },
+    {
+      "epoch": 0.13574584435670065,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.78449932587263e-05,
+      "loss": 0.8732,
+      "step": 7658
+    },
+    {
+      "epoch": 0.13578129639231223,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.784385876837545e-05,
+      "loss": 0.7955,
+      "step": 7660
+    },
+    {
+      "epoch": 0.1358167484279238,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.784272399293799e-05,
+      "loss": 0.848,
+      "step": 7662
+    },
+    {
+      "epoch": 0.13585220046353536,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.784158893242809e-05,
+      "loss": 0.8288,
+      "step": 7664
+    },
+    {
+      "epoch": 0.13588765249914694,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.784045358685993e-05,
+      "loss": 0.763,
+      "step": 7666
+    },
+    {
+      "epoch": 0.1359231045347585,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.783931795624766e-05,
+      "loss": 0.7886,
+      "step": 7668
+    },
+    {
+      "epoch": 0.13595855657037006,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.783818204060546e-05,
+      "loss": 0.8637,
+      "step": 7670
+    },
+    {
+      "epoch": 0.13599400860598165,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.78370458399475e-05,
+      "loss": 0.814,
+      "step": 7672
+    },
+    {
+      "epoch": 0.13602946064159321,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.7835909354287975e-05,
+      "loss": 0.8021,
+      "step": 7674
+    },
+    {
+      "epoch": 0.13606491267720477,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.7834772583641054e-05,
+      "loss": 0.8465,
+      "step": 7676
+    },
+    {
+      "epoch": 0.13610036471281636,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.783363552802092e-05,
+      "loss": 0.8472,
+      "step": 7678
+    },
+    {
+      "epoch": 0.13613581674842792,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.783249818744178e-05,
+      "loss": 0.8313,
+      "step": 7680
+    },
+    {
+      "epoch": 0.13617126878403948,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.783136056191781e-05,
+      "loss": 0.8137,
+      "step": 7682
+    },
+    {
+      "epoch": 0.13620672081965107,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.783022265146322e-05,
+      "loss": 0.8334,
+      "step": 7684
+    },
+    {
+      "epoch": 0.13624217285526263,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.782908445609221e-05,
+      "loss": 0.8134,
+      "step": 7686
+    },
+    {
+      "epoch": 0.1362776248908742,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.782794597581898e-05,
+      "loss": 0.835,
+      "step": 7688
+    },
+    {
+      "epoch": 0.13631307692648578,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.782680721065773e-05,
+      "loss": 0.8511,
+      "step": 7690
+    },
+    {
+      "epoch": 0.13634852896209734,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.7825668160622686e-05,
+      "loss": 0.8473,
+      "step": 7692
+    },
+    {
+      "epoch": 0.1363839809977089,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.7824528825728055e-05,
+      "loss": 0.8053,
+      "step": 7694
+    },
+    {
+      "epoch": 0.1364194330333205,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.782338920598807e-05,
+      "loss": 0.8145,
+      "step": 7696
+    },
+    {
+      "epoch": 0.13645488506893205,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.782224930141693e-05,
+      "loss": 0.8105,
+      "step": 7698
+    },
+    {
+      "epoch": 0.1364903371045436,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.7821109112028876e-05,
+      "loss": 0.7639,
+      "step": 7700
+    },
+    {
+      "epoch": 0.1365257891401552,
+      "grad_norm": 3.125,
+      "learning_rate": 4.781996863783813e-05,
+      "loss": 0.8154,
+      "step": 7702
+    },
+    {
+      "epoch": 0.13656124117576676,
+      "grad_norm": 2.875,
+      "learning_rate": 4.781882787885893e-05,
+      "loss": 0.7771,
+      "step": 7704
+    },
+    {
+      "epoch": 0.13659669321137832,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.7817686835105513e-05,
+      "loss": 0.8352,
+      "step": 7706
+    },
+    {
+      "epoch": 0.1366321452469899,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.781654550659211e-05,
+      "loss": 0.8448,
+      "step": 7708
+    },
+    {
+      "epoch": 0.13666759728260147,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.781540389333298e-05,
+      "loss": 0.8205,
+      "step": 7710
+    },
+    {
+      "epoch": 0.13670304931821303,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.781426199534236e-05,
+      "loss": 0.8376,
+      "step": 7712
+    },
+    {
+      "epoch": 0.13673850135382462,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.78131198126345e-05,
+      "loss": 0.8491,
+      "step": 7714
+    },
+    {
+      "epoch": 0.13677395338943618,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.781197734522366e-05,
+      "loss": 0.8213,
+      "step": 7716
+    },
+    {
+      "epoch": 0.13680940542504774,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.781083459312409e-05,
+      "loss": 0.8204,
+      "step": 7718
+    },
+    {
+      "epoch": 0.13684485746065933,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.780969155635006e-05,
+      "loss": 0.7855,
+      "step": 7720
+    },
+    {
+      "epoch": 0.1368803094962709,
+      "grad_norm": 2.75,
+      "learning_rate": 4.7808548234915826e-05,
+      "loss": 0.7865,
+      "step": 7722
+    },
+    {
+      "epoch": 0.13691576153188245,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.7807404628835664e-05,
+      "loss": 0.8088,
+      "step": 7724
+    },
+    {
+      "epoch": 0.13695121356749404,
+      "grad_norm": 2.75,
+      "learning_rate": 4.780626073812383e-05,
+      "loss": 0.8477,
+      "step": 7726
+    },
+    {
+      "epoch": 0.1369866656031056,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.780511656279463e-05,
+      "loss": 0.8656,
+      "step": 7728
+    },
+    {
+      "epoch": 0.13702211763871716,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.7803972102862314e-05,
+      "loss": 0.8328,
+      "step": 7730
+    },
+    {
+      "epoch": 0.13705756967432875,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.780282735834119e-05,
+      "loss": 0.7767,
+      "step": 7732
+    },
+    {
+      "epoch": 0.1370930217099403,
+      "grad_norm": 2.625,
+      "learning_rate": 4.780168232924551e-05,
+      "loss": 0.8444,
+      "step": 7734
+    },
+    {
+      "epoch": 0.13712847374555187,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.78005370155896e-05,
+      "loss": 0.8089,
+      "step": 7736
+    },
+    {
+      "epoch": 0.13716392578116346,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.7799391417387727e-05,
+      "loss": 0.8253,
+      "step": 7738
+    },
+    {
+      "epoch": 0.13719937781677502,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.77982455346542e-05,
+      "loss": 0.8302,
+      "step": 7740
+    },
+    {
+      "epoch": 0.13723482985238658,
+      "grad_norm": 2.75,
+      "learning_rate": 4.779709936740332e-05,
+      "loss": 0.8361,
+      "step": 7742
+    },
+    {
+      "epoch": 0.13727028188799817,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.779595291564939e-05,
+      "loss": 0.8076,
+      "step": 7744
+    },
+    {
+      "epoch": 0.13730573392360973,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.77948061794067e-05,
+      "loss": 0.8185,
+      "step": 7746
+    },
+    {
+      "epoch": 0.1373411859592213,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.7793659158689594e-05,
+      "loss": 0.7821,
+      "step": 7748
+    },
+    {
+      "epoch": 0.13737663799483288,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.779251185351237e-05,
+      "loss": 0.8564,
+      "step": 7750
+    },
+    {
+      "epoch": 0.13741209003044444,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.779136426388934e-05,
+      "loss": 0.8062,
+      "step": 7752
+    },
+    {
+      "epoch": 0.137447542066056,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.779021638983483e-05,
+      "loss": 0.8118,
+      "step": 7754
+    },
+    {
+      "epoch": 0.1374829941016676,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.7789068231363165e-05,
+      "loss": 0.8128,
+      "step": 7756
+    },
+    {
+      "epoch": 0.13751844613727915,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.7787919788488675e-05,
+      "loss": 0.8501,
+      "step": 7758
+    },
+    {
+      "epoch": 0.1375538981728907,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.77867710612257e-05,
+      "loss": 0.7931,
+      "step": 7760
+    },
+    {
+      "epoch": 0.1375893502085023,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.778562204958856e-05,
+      "loss": 0.882,
+      "step": 7762
+    },
+    {
+      "epoch": 0.13762480224411386,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.7784472753591606e-05,
+      "loss": 0.8161,
+      "step": 7764
+    },
+    {
+      "epoch": 0.13766025427972542,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.778332317324918e-05,
+      "loss": 0.8651,
+      "step": 7766
+    },
+    {
+      "epoch": 0.137695706315337,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.7782173308575625e-05,
+      "loss": 0.823,
+      "step": 7768
+    },
+    {
+      "epoch": 0.13773115835094857,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.7781023159585295e-05,
+      "loss": 0.8671,
+      "step": 7770
+    },
+    {
+      "epoch": 0.13776661038656013,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.777987272629253e-05,
+      "loss": 0.8153,
+      "step": 7772
+    },
+    {
+      "epoch": 0.1378020624221717,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.7778722008711704e-05,
+      "loss": 0.9034,
+      "step": 7774
+    },
+    {
+      "epoch": 0.13783751445778328,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.7777571006857174e-05,
+      "loss": 0.8068,
+      "step": 7776
+    },
+    {
+      "epoch": 0.13787296649339484,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.777641972074331e-05,
+      "loss": 0.7982,
+      "step": 7778
+    },
+    {
+      "epoch": 0.1379084185290064,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.7775268150384454e-05,
+      "loss": 0.8481,
+      "step": 7780
+    },
+    {
+      "epoch": 0.137943870564618,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.7774116295795e-05,
+      "loss": 0.8235,
+      "step": 7782
+    },
+    {
+      "epoch": 0.13797932260022955,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.777296415698933e-05,
+      "loss": 0.7464,
+      "step": 7784
+    },
+    {
+      "epoch": 0.1380147746358411,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.7771811733981797e-05,
+      "loss": 0.8036,
+      "step": 7786
+    },
+    {
+      "epoch": 0.1380502266714527,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.777065902678681e-05,
+      "loss": 0.7683,
+      "step": 7788
+    },
+    {
+      "epoch": 0.13808567870706426,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.776950603541873e-05,
+      "loss": 0.8879,
+      "step": 7790
+    },
+    {
+      "epoch": 0.13812113074267582,
+      "grad_norm": 2.875,
+      "learning_rate": 4.776835275989196e-05,
+      "loss": 0.7814,
+      "step": 7792
+    },
+    {
+      "epoch": 0.1381565827782874,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.776719920022089e-05,
+      "loss": 0.8611,
+      "step": 7794
+    },
+    {
+      "epoch": 0.13819203481389897,
+      "grad_norm": 2.625,
+      "learning_rate": 4.776604535641992e-05,
+      "loss": 0.8142,
+      "step": 7796
+    },
+    {
+      "epoch": 0.13822748684951053,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.776489122850344e-05,
+      "loss": 0.8255,
+      "step": 7798
+    },
+    {
+      "epoch": 0.13826293888512212,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.776373681648586e-05,
+      "loss": 0.7773,
+      "step": 7800
+    },
+    {
+      "epoch": 0.13829839092073368,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.776258212038159e-05,
+      "loss": 0.7941,
+      "step": 7802
+    },
+    {
+      "epoch": 0.13833384295634524,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.7761427140205034e-05,
+      "loss": 0.7812,
+      "step": 7804
+    },
+    {
+      "epoch": 0.13836929499195683,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.7760271875970606e-05,
+      "loss": 0.7889,
+      "step": 7806
+    },
+    {
+      "epoch": 0.1384047470275684,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.7759116327692726e-05,
+      "loss": 0.8312,
+      "step": 7808
+    },
+    {
+      "epoch": 0.13844019906317995,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.775796049538582e-05,
+      "loss": 0.846,
+      "step": 7810
+    },
+    {
+      "epoch": 0.13847565109879154,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.77568043790643e-05,
+      "loss": 0.7984,
+      "step": 7812
+    },
+    {
+      "epoch": 0.1385111031344031,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.77556479787426e-05,
+      "loss": 0.7918,
+      "step": 7814
+    },
+    {
+      "epoch": 0.13854655517001466,
+      "grad_norm": 2.75,
+      "learning_rate": 4.7754491294435165e-05,
+      "loss": 0.8044,
+      "step": 7816
+    },
+    {
+      "epoch": 0.13858200720562625,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.775333432615641e-05,
+      "loss": 0.791,
+      "step": 7818
+    },
+    {
+      "epoch": 0.1386174592412378,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.775217707392078e-05,
+      "loss": 0.7833,
+      "step": 7820
+    },
+    {
+      "epoch": 0.13865291127684937,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.7751019537742725e-05,
+      "loss": 0.8009,
+      "step": 7822
+    },
+    {
+      "epoch": 0.13868836331246095,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.774986171763668e-05,
+      "loss": 0.8026,
+      "step": 7824
+    },
+    {
+      "epoch": 0.13872381534807252,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.7748703613617095e-05,
+      "loss": 0.7954,
+      "step": 7826
+    },
+    {
+      "epoch": 0.13875926738368408,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.7747545225698434e-05,
+      "loss": 0.8351,
+      "step": 7828
+    },
+    {
+      "epoch": 0.13879471941929566,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.774638655389514e-05,
+      "loss": 0.8444,
+      "step": 7830
+    },
+    {
+      "epoch": 0.13883017145490723,
+      "grad_norm": 2.75,
+      "learning_rate": 4.7745227598221687e-05,
+      "loss": 0.8925,
+      "step": 7832
+    },
+    {
+      "epoch": 0.13886562349051879,
+      "grad_norm": 2.625,
+      "learning_rate": 4.774406835869253e-05,
+      "loss": 0.7889,
+      "step": 7834
+    },
+    {
+      "epoch": 0.13890107552613037,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.7742908835322136e-05,
+      "loss": 0.8259,
+      "step": 7836
+    },
+    {
+      "epoch": 0.13893652756174193,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.774174902812498e-05,
+      "loss": 0.8031,
+      "step": 7838
+    },
+    {
+      "epoch": 0.1389719795973535,
+      "grad_norm": 2.5,
+      "learning_rate": 4.774058893711553e-05,
+      "loss": 0.7957,
+      "step": 7840
+    },
+    {
+      "epoch": 0.13900743163296508,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.7739428562308266e-05,
+      "loss": 0.7898,
+      "step": 7842
+    },
+    {
+      "epoch": 0.13904288366857664,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.773826790371767e-05,
+      "loss": 0.7799,
+      "step": 7844
+    },
+    {
+      "epoch": 0.1390783357041882,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.773710696135822e-05,
+      "loss": 0.8168,
+      "step": 7846
+    },
+    {
+      "epoch": 0.1391137877397998,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.773594573524442e-05,
+      "loss": 0.8682,
+      "step": 7848
+    },
+    {
+      "epoch": 0.13914923977541135,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.773478422539075e-05,
+      "loss": 0.8098,
+      "step": 7850
+    },
+    {
+      "epoch": 0.13918469181102291,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.773362243181171e-05,
+      "loss": 0.7966,
+      "step": 7852
+    },
+    {
+      "epoch": 0.1392201438466345,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.773246035452179e-05,
+      "loss": 0.828,
+      "step": 7854
+    },
+    {
+      "epoch": 0.13925559588224606,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.773129799353551e-05,
+      "loss": 0.8082,
+      "step": 7856
+    },
+    {
+      "epoch": 0.13929104791785762,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.773013534886735e-05,
+      "loss": 0.8269,
+      "step": 7858
+    },
+    {
+      "epoch": 0.1393264999534692,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.772897242053186e-05,
+      "loss": 0.8295,
+      "step": 7860
+    },
+    {
+      "epoch": 0.13936195198908077,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.772780920854351e-05,
+      "loss": 0.8027,
+      "step": 7862
+    },
+    {
+      "epoch": 0.13939740402469233,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.772664571291684e-05,
+      "loss": 0.8514,
+      "step": 7864
+    },
+    {
+      "epoch": 0.13943285606030392,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.772548193366636e-05,
+      "loss": 0.7971,
+      "step": 7866
+    },
+    {
+      "epoch": 0.13946830809591548,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.772431787080661e-05,
+      "loss": 0.8007,
+      "step": 7868
+    },
+    {
+      "epoch": 0.13950376013152704,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.7723153524352096e-05,
+      "loss": 0.8558,
+      "step": 7870
+    },
+    {
+      "epoch": 0.13953921216713863,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.772198889431736e-05,
+      "loss": 0.8324,
+      "step": 7872
+    },
+    {
+      "epoch": 0.1395746642027502,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.7720823980716934e-05,
+      "loss": 0.8482,
+      "step": 7874
+    },
+    {
+      "epoch": 0.13961011623836175,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.771965878356536e-05,
+      "loss": 0.8482,
+      "step": 7876
+    },
+    {
+      "epoch": 0.13964556827397334,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.771849330287718e-05,
+      "loss": 0.8271,
+      "step": 7878
+    },
+    {
+      "epoch": 0.1396810203095849,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.7717327538666935e-05,
+      "loss": 0.8187,
+      "step": 7880
+    },
+    {
+      "epoch": 0.13971647234519646,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.771616149094917e-05,
+      "loss": 0.817,
+      "step": 7882
+    },
+    {
+      "epoch": 0.13975192438080805,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.771499515973844e-05,
+      "loss": 0.7927,
+      "step": 7884
+    },
+    {
+      "epoch": 0.1397873764164196,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.7713828545049303e-05,
+      "loss": 0.8102,
+      "step": 7886
+    },
+    {
+      "epoch": 0.13982282845203117,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.7712661646896316e-05,
+      "loss": 0.8759,
+      "step": 7888
+    },
+    {
+      "epoch": 0.13985828048764276,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.771149446529405e-05,
+      "loss": 0.847,
+      "step": 7890
+    },
+    {
+      "epoch": 0.13989373252325432,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.771032700025706e-05,
+      "loss": 0.8383,
+      "step": 7892
+    },
+    {
+      "epoch": 0.13992918455886588,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.770915925179991e-05,
+      "loss": 0.8035,
+      "step": 7894
+    },
+    {
+      "epoch": 0.13996463659447747,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.7707991219937194e-05,
+      "loss": 0.8306,
+      "step": 7896
+    },
+    {
+      "epoch": 0.14000008863008903,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.770682290468347e-05,
+      "loss": 0.8405,
+      "step": 7898
+    },
+    {
+      "epoch": 0.1400355406657006,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.7705654306053326e-05,
+      "loss": 0.8105,
+      "step": 7900
+    },
+    {
+      "epoch": 0.14007099270131218,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.770448542406135e-05,
+      "loss": 0.7714,
+      "step": 7902
+    },
+    {
+      "epoch": 0.14010644473692374,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.770331625872212e-05,
+      "loss": 0.7847,
+      "step": 7904
+    },
+    {
+      "epoch": 0.1401418967725353,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.770214681005024e-05,
+      "loss": 0.8186,
+      "step": 7906
+    },
+    {
+      "epoch": 0.1401773488081469,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.7700977078060286e-05,
+      "loss": 0.799,
+      "step": 7908
+    },
+    {
+      "epoch": 0.14021280084375845,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.7699807062766876e-05,
+      "loss": 0.7855,
+      "step": 7910
+    },
+    {
+      "epoch": 0.14024825287937,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.7698636764184597e-05,
+      "loss": 0.848,
+      "step": 7912
+    },
+    {
+      "epoch": 0.1402837049149816,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.769746618232805e-05,
+      "loss": 0.8345,
+      "step": 7914
+    },
+    {
+      "epoch": 0.14031915695059316,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.769629531721187e-05,
+      "loss": 0.7926,
+      "step": 7916
+    },
+    {
+      "epoch": 0.14035460898620472,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.769512416885064e-05,
+      "loss": 0.8247,
+      "step": 7918
+    },
+    {
+      "epoch": 0.1403900610218163,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.7693952737259e-05,
+      "loss": 0.8146,
+      "step": 7920
+    },
+    {
+      "epoch": 0.14042551305742787,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.7692781022451536e-05,
+      "loss": 0.8202,
+      "step": 7922
+    },
+    {
+      "epoch": 0.14046096509303943,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.7691609024442905e-05,
+      "loss": 0.819,
+      "step": 7924
+    },
+    {
+      "epoch": 0.14049641712865102,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.7690436743247727e-05,
+      "loss": 0.8124,
+      "step": 7926
+    },
+    {
+      "epoch": 0.14053186916426258,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.768926417888061e-05,
+      "loss": 0.7961,
+      "step": 7928
+    },
+    {
+      "epoch": 0.14056732119987414,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.76880913313562e-05,
+      "loss": 0.7982,
+      "step": 7930
+    },
+    {
+      "epoch": 0.14060277323548573,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.7686918200689144e-05,
+      "loss": 0.8538,
+      "step": 7932
+    },
+    {
+      "epoch": 0.1406382252710973,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.768574478689408e-05,
+      "loss": 0.7942,
+      "step": 7934
+    },
+    {
+      "epoch": 0.14067367730670885,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.768457108998564e-05,
+      "loss": 0.802,
+      "step": 7936
+    },
+    {
+      "epoch": 0.14070912934232044,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.768339710997847e-05,
+      "loss": 0.8082,
+      "step": 7938
+    },
+    {
+      "epoch": 0.140744581377932,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.768222284688724e-05,
+      "loss": 0.7945,
+      "step": 7940
+    },
+    {
+      "epoch": 0.14078003341354356,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.7681048300726584e-05,
+      "loss": 0.845,
+      "step": 7942
+    },
+    {
+      "epoch": 0.14081548544915512,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.767987347151118e-05,
+      "loss": 0.8232,
+      "step": 7944
+    },
+    {
+      "epoch": 0.1408509374847667,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.767869835925567e-05,
+      "loss": 0.8091,
+      "step": 7946
+    },
+    {
+      "epoch": 0.14088638952037827,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.767752296397473e-05,
+      "loss": 0.7925,
+      "step": 7948
+    },
+    {
+      "epoch": 0.14092184155598983,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.767634728568303e-05,
+      "loss": 0.8356,
+      "step": 7950
+    },
+    {
+      "epoch": 0.14095729359160142,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.7675171324395236e-05,
+      "loss": 0.8526,
+      "step": 7952
+    },
+    {
+      "epoch": 0.14099274562721298,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.767399508012603e-05,
+      "loss": 0.8233,
+      "step": 7954
+    },
+    {
+      "epoch": 0.14102819766282454,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.767281855289009e-05,
+      "loss": 0.8237,
+      "step": 7956
+    },
+    {
+      "epoch": 0.14106364969843613,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.767164174270208e-05,
+      "loss": 0.8088,
+      "step": 7958
+    },
+    {
+      "epoch": 0.1410991017340477,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.767046464957672e-05,
+      "loss": 0.8382,
+      "step": 7960
+    },
+    {
+      "epoch": 0.14113455376965925,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.7669287273528676e-05,
+      "loss": 0.7871,
+      "step": 7962
+    },
+    {
+      "epoch": 0.14117000580527084,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.766810961457265e-05,
+      "loss": 0.8088,
+      "step": 7964
+    },
+    {
+      "epoch": 0.1412054578408824,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.7666931672723346e-05,
+      "loss": 0.8341,
+      "step": 7966
+    },
+    {
+      "epoch": 0.14124090987649396,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.766575344799544e-05,
+      "loss": 0.816,
+      "step": 7968
+    },
+    {
+      "epoch": 0.14127636191210555,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.7664574940403666e-05,
+      "loss": 0.8126,
+      "step": 7970
+    },
+    {
+      "epoch": 0.1413118139477171,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.7663396149962715e-05,
+      "loss": 0.7982,
+      "step": 7972
+    },
+    {
+      "epoch": 0.14134726598332867,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.76622170766873e-05,
+      "loss": 0.7979,
+      "step": 7974
+    },
+    {
+      "epoch": 0.14138271801894026,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.766103772059213e-05,
+      "loss": 0.8553,
+      "step": 7976
+    },
+    {
+      "epoch": 0.14141817005455182,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.7659858081691936e-05,
+      "loss": 0.7936,
+      "step": 7978
+    },
+    {
+      "epoch": 0.14145362209016338,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.7658678160001425e-05,
+      "loss": 0.8434,
+      "step": 7980
+    },
+    {
+      "epoch": 0.14148907412577497,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.7657497955535334e-05,
+      "loss": 0.8208,
+      "step": 7982
+    },
+    {
+      "epoch": 0.14152452616138653,
+      "grad_norm": 3.0,
+      "learning_rate": 4.765631746830839e-05,
+      "loss": 0.8071,
+      "step": 7984
+    },
+    {
+      "epoch": 0.1415599781969981,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.7655136698335326e-05,
+      "loss": 0.8083,
+      "step": 7986
+    },
+    {
+      "epoch": 0.14159543023260968,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.7653955645630866e-05,
+      "loss": 0.7766,
+      "step": 7988
+    },
+    {
+      "epoch": 0.14163088226822124,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.765277431020976e-05,
+      "loss": 0.8276,
+      "step": 7990
+    },
+    {
+      "epoch": 0.1416663343038328,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.7651592692086756e-05,
+      "loss": 0.8471,
+      "step": 7992
+    },
+    {
+      "epoch": 0.14170178633944439,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.7650410791276584e-05,
+      "loss": 0.8432,
+      "step": 7994
+    },
+    {
+      "epoch": 0.14173723837505595,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.764922860779401e-05,
+      "loss": 0.7798,
+      "step": 7996
+    },
+    {
+      "epoch": 0.1417726904106675,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.764804614165377e-05,
+      "loss": 0.763,
+      "step": 7998
+    },
+    {
+      "epoch": 0.1418081424462791,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.7646863392870644e-05,
+      "loss": 0.7884,
+      "step": 8000
+    },
+    {
+      "epoch": 0.14184359448189066,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.764568036145938e-05,
+      "loss": 0.819,
+      "step": 8002
+    },
+    {
+      "epoch": 0.14187904651750222,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.764449704743473e-05,
+      "loss": 0.8361,
+      "step": 8004
+    },
+    {
+      "epoch": 0.1419144985531138,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.764331345081148e-05,
+      "loss": 0.829,
+      "step": 8006
+    },
+    {
+      "epoch": 0.14194995058872537,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.76421295716044e-05,
+      "loss": 0.8125,
+      "step": 8008
+    },
+    {
+      "epoch": 0.14198540262433693,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.7640945409828255e-05,
+      "loss": 0.848,
+      "step": 8010
+    },
+    {
+      "epoch": 0.14202085465994851,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.763976096549782e-05,
+      "loss": 0.8195,
+      "step": 8012
+    },
+    {
+      "epoch": 0.14205630669556008,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.7638576238627886e-05,
+      "loss": 0.7931,
+      "step": 8014
+    },
+    {
+      "epoch": 0.14209175873117164,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.763739122923324e-05,
+      "loss": 0.8273,
+      "step": 8016
+    },
+    {
+      "epoch": 0.14212721076678322,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.763620593732867e-05,
+      "loss": 0.8068,
+      "step": 8018
+    },
+    {
+      "epoch": 0.14216266280239478,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.763502036292896e-05,
+      "loss": 0.7955,
+      "step": 8020
+    },
+    {
+      "epoch": 0.14219811483800635,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.763383450604891e-05,
+      "loss": 0.8171,
+      "step": 8022
+    },
+    {
+      "epoch": 0.14223356687361793,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.763264836670332e-05,
+      "loss": 0.8004,
+      "step": 8024
+    },
+    {
+      "epoch": 0.1422690189092295,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.7631461944906994e-05,
+      "loss": 0.8078,
+      "step": 8026
+    },
+    {
+      "epoch": 0.14230447094484105,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.763027524067473e-05,
+      "loss": 0.8291,
+      "step": 8028
+    },
+    {
+      "epoch": 0.14233992298045264,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.7629088254021354e-05,
+      "loss": 0.8019,
+      "step": 8030
+    },
+    {
+      "epoch": 0.1423753750160642,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.762790098496166e-05,
+      "loss": 0.7968,
+      "step": 8032
+    },
+    {
+      "epoch": 0.14241082705167576,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.7626713433510485e-05,
+      "loss": 0.7821,
+      "step": 8034
+    },
+    {
+      "epoch": 0.14244627908728735,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.762552559968264e-05,
+      "loss": 0.8094,
+      "step": 8036
+    },
+    {
+      "epoch": 0.1424817311228989,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.762433748349294e-05,
+      "loss": 0.8488,
+      "step": 8038
+    },
+    {
+      "epoch": 0.14251718315851047,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.762314908495622e-05,
+      "loss": 0.8179,
+      "step": 8040
+    },
+    {
+      "epoch": 0.14255263519412206,
+      "grad_norm": 3.125,
+      "learning_rate": 4.7621960404087316e-05,
+      "loss": 0.8421,
+      "step": 8042
+    },
+    {
+      "epoch": 0.14258808722973362,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.7620771440901056e-05,
+      "loss": 0.7974,
+      "step": 8044
+    },
+    {
+      "epoch": 0.14262353926534518,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.761958219541228e-05,
+      "loss": 0.8883,
+      "step": 8046
+    },
+    {
+      "epoch": 0.14265899130095677,
+      "grad_norm": 2.875,
+      "learning_rate": 4.761839266763583e-05,
+      "loss": 0.8937,
+      "step": 8048
+    },
+    {
+      "epoch": 0.14269444333656833,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.761720285758655e-05,
+      "loss": 0.7914,
+      "step": 8050
+    },
+    {
+      "epoch": 0.1427298953721799,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.761601276527929e-05,
+      "loss": 0.7956,
+      "step": 8052
+    },
+    {
+      "epoch": 0.14276534740779148,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.76148223907289e-05,
+      "loss": 0.8424,
+      "step": 8054
+    },
+    {
+      "epoch": 0.14280079944340304,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.761363173395024e-05,
+      "loss": 0.8371,
+      "step": 8056
+    },
+    {
+      "epoch": 0.1428362514790146,
+      "grad_norm": 2.75,
+      "learning_rate": 4.761244079495817e-05,
+      "loss": 0.8222,
+      "step": 8058
+    },
+    {
+      "epoch": 0.1428717035146262,
+      "grad_norm": 2.875,
+      "learning_rate": 4.761124957376754e-05,
+      "loss": 0.7927,
+      "step": 8060
+    },
+    {
+      "epoch": 0.14290715555023775,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.761005807039323e-05,
+      "loss": 0.8782,
+      "step": 8062
+    },
+    {
+      "epoch": 0.1429426075858493,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.7608866284850104e-05,
+      "loss": 0.8218,
+      "step": 8064
+    },
+    {
+      "epoch": 0.1429780596214609,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.7607674217153034e-05,
+      "loss": 0.8434,
+      "step": 8066
+    },
+    {
+      "epoch": 0.14301351165707246,
+      "grad_norm": 3.421875,
+      "learning_rate": 4.760648186731689e-05,
+      "loss": 0.849,
+      "step": 8068
+    },
+    {
+      "epoch": 0.14304896369268402,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.7605289235356574e-05,
+      "loss": 0.7891,
+      "step": 8070
+    },
+    {
+      "epoch": 0.1430844157282956,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.760409632128695e-05,
+      "loss": 0.8308,
+      "step": 8072
+    },
+    {
+      "epoch": 0.14311986776390717,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.7602903125122914e-05,
+      "loss": 0.801,
+      "step": 8074
+    },
+    {
+      "epoch": 0.14315531979951873,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.760170964687935e-05,
+      "loss": 0.8058,
+      "step": 8076
+    },
+    {
+      "epoch": 0.14319077183513032,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.760051588657117e-05,
+      "loss": 0.8482,
+      "step": 8078
+    },
+    {
+      "epoch": 0.14322622387074188,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.759932184421325e-05,
+      "loss": 0.78,
+      "step": 8080
+    },
+    {
+      "epoch": 0.14326167590635344,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.75981275198205e-05,
+      "loss": 0.8269,
+      "step": 8082
+    },
+    {
+      "epoch": 0.14329712794196503,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.759693291340783e-05,
+      "loss": 0.8253,
+      "step": 8084
+    },
+    {
+      "epoch": 0.1433325799775766,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.759573802499014e-05,
+      "loss": 0.7828,
+      "step": 8086
+    },
+    {
+      "epoch": 0.14336803201318815,
+      "grad_norm": 3.125,
+      "learning_rate": 4.759454285458235e-05,
+      "loss": 0.8468,
+      "step": 8088
+    },
+    {
+      "epoch": 0.14340348404879974,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.759334740219937e-05,
+      "loss": 0.8235,
+      "step": 8090
+    },
+    {
+      "epoch": 0.1434389360844113,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.7592151667856125e-05,
+      "loss": 0.7874,
+      "step": 8092
+    },
+    {
+      "epoch": 0.14347438812002286,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.759095565156752e-05,
+      "loss": 0.768,
+      "step": 8094
+    },
+    {
+      "epoch": 0.14350984015563445,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.75897593533485e-05,
+      "loss": 0.8154,
+      "step": 8096
+    },
+    {
+      "epoch": 0.143545292191246,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.758856277321398e-05,
+      "loss": 0.8223,
+      "step": 8098
+    },
+    {
+      "epoch": 0.14358074422685757,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.758736591117892e-05,
+      "loss": 0.8145,
+      "step": 8100
+    },
+    {
+      "epoch": 0.14361619626246916,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.7586168767258227e-05,
+      "loss": 0.7969,
+      "step": 8102
+    },
+    {
+      "epoch": 0.14365164829808072,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.758497134146686e-05,
+      "loss": 0.8054,
+      "step": 8104
+    },
+    {
+      "epoch": 0.14368710033369228,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.758377363381974e-05,
+      "loss": 0.8435,
+      "step": 8106
+    },
+    {
+      "epoch": 0.14372255236930387,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.7582575644331836e-05,
+      "loss": 0.8286,
+      "step": 8108
+    },
+    {
+      "epoch": 0.14375800440491543,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.758137737301809e-05,
+      "loss": 0.8173,
+      "step": 8110
+    },
+    {
+      "epoch": 0.143793456440527,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.7580178819893465e-05,
+      "loss": 0.8239,
+      "step": 8112
+    },
+    {
+      "epoch": 0.14382890847613855,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.75789799849729e-05,
+      "loss": 0.8259,
+      "step": 8114
+    },
+    {
+      "epoch": 0.14386436051175014,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.757778086827138e-05,
+      "loss": 0.8239,
+      "step": 8116
+    },
+    {
+      "epoch": 0.1438998125473617,
+      "grad_norm": 2.75,
+      "learning_rate": 4.757658146980385e-05,
+      "loss": 0.8177,
+      "step": 8118
+    },
+    {
+      "epoch": 0.14393526458297326,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.7575381789585296e-05,
+      "loss": 0.809,
+      "step": 8120
+    },
+    {
+      "epoch": 0.14397071661858485,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.7574181827630666e-05,
+      "loss": 0.8368,
+      "step": 8122
+    },
+    {
+      "epoch": 0.1440061686541964,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.757298158395496e-05,
+      "loss": 0.8409,
+      "step": 8124
+    },
+    {
+      "epoch": 0.14404162068980797,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.757178105857313e-05,
+      "loss": 0.8275,
+      "step": 8126
+    },
+    {
+      "epoch": 0.14407707272541956,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.757058025150018e-05,
+      "loss": 0.8308,
+      "step": 8128
+    },
+    {
+      "epoch": 0.14411252476103112,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.7569379162751094e-05,
+      "loss": 0.804,
+      "step": 8130
+    },
+    {
+      "epoch": 0.14414797679664268,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.756817779234086e-05,
+      "loss": 0.8118,
+      "step": 8132
+    },
+    {
+      "epoch": 0.14418342883225427,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.756697614028446e-05,
+      "loss": 0.8067,
+      "step": 8134
+    },
+    {
+      "epoch": 0.14421888086786583,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.75657742065969e-05,
+      "loss": 0.8501,
+      "step": 8136
+    },
+    {
+      "epoch": 0.1442543329034774,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.7564571991293184e-05,
+      "loss": 0.7915,
+      "step": 8138
+    },
+    {
+      "epoch": 0.14428978493908898,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.75633694943883e-05,
+      "loss": 0.7987,
+      "step": 8140
+    },
+    {
+      "epoch": 0.14432523697470054,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.756216671589727e-05,
+      "loss": 0.8148,
+      "step": 8142
+    },
+    {
+      "epoch": 0.1443606890103121,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.75609636558351e-05,
+      "loss": 0.8114,
+      "step": 8144
+    },
+    {
+      "epoch": 0.1443961410459237,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.7559760314216794e-05,
+      "loss": 0.826,
+      "step": 8146
+    },
+    {
+      "epoch": 0.14443159308153525,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.755855669105739e-05,
+      "loss": 0.8442,
+      "step": 8148
+    },
+    {
+      "epoch": 0.1444670451171468,
+      "grad_norm": 3.0,
+      "learning_rate": 4.755735278637189e-05,
+      "loss": 0.834,
+      "step": 8150
+    },
+    {
+      "epoch": 0.1445024971527584,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.755614860017533e-05,
+      "loss": 0.7876,
+      "step": 8152
+    },
+    {
+      "epoch": 0.14453794918836996,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.7554944132482724e-05,
+      "loss": 0.8251,
+      "step": 8154
+    },
+    {
+      "epoch": 0.14457340122398152,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.755373938330912e-05,
+      "loss": 0.813,
+      "step": 8156
+    },
+    {
+      "epoch": 0.1446088532595931,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.755253435266955e-05,
+      "loss": 0.8088,
+      "step": 8158
+    },
+    {
+      "epoch": 0.14464430529520467,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.755132904057904e-05,
+      "loss": 0.8135,
+      "step": 8160
+    },
+    {
+      "epoch": 0.14467975733081623,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.7550123447052646e-05,
+      "loss": 0.7945,
+      "step": 8162
+    },
+    {
+      "epoch": 0.14471520936642782,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.754891757210541e-05,
+      "loss": 0.8046,
+      "step": 8164
+    },
+    {
+      "epoch": 0.14475066140203938,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.754771141575237e-05,
+      "loss": 0.8994,
+      "step": 8166
+    },
+    {
+      "epoch": 0.14478611343765094,
+      "grad_norm": 2.625,
+      "learning_rate": 4.75465049780086e-05,
+      "loss": 0.8353,
+      "step": 8168
+    },
+    {
+      "epoch": 0.14482156547326253,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.754529825888914e-05,
+      "loss": 0.7793,
+      "step": 8170
+    },
+    {
+      "epoch": 0.1448570175088741,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.754409125840905e-05,
+      "loss": 0.8158,
+      "step": 8172
+    },
+    {
+      "epoch": 0.14489246954448565,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.75428839765834e-05,
+      "loss": 0.8244,
+      "step": 8174
+    },
+    {
+      "epoch": 0.14492792158009724,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.754167641342725e-05,
+      "loss": 0.7988,
+      "step": 8176
+    },
+    {
+      "epoch": 0.1449633736157088,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.754046856895568e-05,
+      "loss": 0.8256,
+      "step": 8178
+    },
+    {
+      "epoch": 0.14499882565132036,
+      "grad_norm": 2.75,
+      "learning_rate": 4.753926044318375e-05,
+      "loss": 0.8186,
+      "step": 8180
+    },
+    {
+      "epoch": 0.14503427768693194,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.7538052036126545e-05,
+      "loss": 0.7826,
+      "step": 8182
+    },
+    {
+      "epoch": 0.1450697297225435,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.753684334779914e-05,
+      "loss": 0.8254,
+      "step": 8184
+    },
+    {
+      "epoch": 0.14510518175815507,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.7535634378216636e-05,
+      "loss": 0.8321,
+      "step": 8186
+    },
+    {
+      "epoch": 0.14514063379376665,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.7534425127394106e-05,
+      "loss": 0.7746,
+      "step": 8188
+    },
+    {
+      "epoch": 0.14517608582937822,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.7533215595346636e-05,
+      "loss": 0.8301,
+      "step": 8190
+    },
+    {
+      "epoch": 0.14521153786498978,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.753200578208934e-05,
+      "loss": 0.813,
+      "step": 8192
+    },
+    {
+      "epoch": 0.14524698990060136,
+      "grad_norm": 2.875,
+      "learning_rate": 4.75307956876373e-05,
+      "loss": 0.808,
+      "step": 8194
+    },
+    {
+      "epoch": 0.14528244193621292,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.752958531200562e-05,
+      "loss": 0.8109,
+      "step": 8196
+    },
+    {
+      "epoch": 0.14531789397182449,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.7528374655209407e-05,
+      "loss": 0.823,
+      "step": 8198
+    },
+    {
+      "epoch": 0.14535334600743607,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.752716371726378e-05,
+      "loss": 0.8127,
+      "step": 8200
+    },
+    {
+      "epoch": 0.14538879804304763,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.752595249818383e-05,
+      "loss": 0.8062,
+      "step": 8202
+    },
+    {
+      "epoch": 0.1454242500786592,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.752474099798469e-05,
+      "loss": 0.7996,
+      "step": 8204
+    },
+    {
+      "epoch": 0.14545970211427078,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.752352921668147e-05,
+      "loss": 0.8083,
+      "step": 8206
+    },
+    {
+      "epoch": 0.14549515414988234,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.752231715428931e-05,
+      "loss": 0.8477,
+      "step": 8208
+    },
+    {
+      "epoch": 0.1455306061854939,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.752110481082331e-05,
+      "loss": 0.841,
+      "step": 8210
+    },
+    {
+      "epoch": 0.1455660582211055,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.751989218629861e-05,
+      "loss": 0.8154,
+      "step": 8212
+    },
+    {
+      "epoch": 0.14560151025671705,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.751867928073036e-05,
+      "loss": 0.7936,
+      "step": 8214
+    },
+    {
+      "epoch": 0.14563696229232861,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.751746609413367e-05,
+      "loss": 0.8145,
+      "step": 8216
+    },
+    {
+      "epoch": 0.1456724143279402,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.75162526265237e-05,
+      "loss": 0.7813,
+      "step": 8218
+    },
+    {
+      "epoch": 0.14570786636355176,
+      "grad_norm": 3.0,
+      "learning_rate": 4.7515038877915584e-05,
+      "loss": 0.8669,
+      "step": 8220
+    },
+    {
+      "epoch": 0.14574331839916332,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.7513824848324474e-05,
+      "loss": 0.8316,
+      "step": 8222
+    },
+    {
+      "epoch": 0.1457787704347749,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.751261053776552e-05,
+      "loss": 0.8761,
+      "step": 8224
+    },
+    {
+      "epoch": 0.14581422247038647,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.751139594625388e-05,
+      "loss": 0.8457,
+      "step": 8226
+    },
+    {
+      "epoch": 0.14584967450599803,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.751018107380469e-05,
+      "loss": 0.8049,
+      "step": 8228
+    },
+    {
+      "epoch": 0.14588512654160962,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.750896592043315e-05,
+      "loss": 0.82,
+      "step": 8230
+    },
+    {
+      "epoch": 0.14592057857722118,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.7507750486154387e-05,
+      "loss": 0.8268,
+      "step": 8232
+    },
+    {
+      "epoch": 0.14595603061283274,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.7506534770983595e-05,
+      "loss": 0.8836,
+      "step": 8234
+    },
+    {
+      "epoch": 0.14599148264844433,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.750531877493594e-05,
+      "loss": 0.8191,
+      "step": 8236
+    },
+    {
+      "epoch": 0.1460269346840559,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.7504102498026584e-05,
+      "loss": 0.8321,
+      "step": 8238
+    },
+    {
+      "epoch": 0.14606238671966745,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.7502885940270723e-05,
+      "loss": 0.8084,
+      "step": 8240
+    },
+    {
+      "epoch": 0.14609783875527904,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.7501669101683535e-05,
+      "loss": 0.7722,
+      "step": 8242
+    },
+    {
+      "epoch": 0.1461332907908906,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.750045198228019e-05,
+      "loss": 0.8216,
+      "step": 8244
+    },
+    {
+      "epoch": 0.14616874282650216,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.7499234582075905e-05,
+      "loss": 0.813,
+      "step": 8246
+    },
+    {
+      "epoch": 0.14620419486211375,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.749801690108585e-05,
+      "loss": 0.8364,
+      "step": 8248
+    },
+    {
+      "epoch": 0.1462396468977253,
+      "grad_norm": 2.875,
+      "learning_rate": 4.749679893932524e-05,
+      "loss": 0.8664,
+      "step": 8250
+    },
+    {
+      "epoch": 0.14627509893333687,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.7495580696809254e-05,
+      "loss": 0.8306,
+      "step": 8252
+    },
+    {
+      "epoch": 0.14631055096894846,
+      "grad_norm": 2.75,
+      "learning_rate": 4.7494362173553114e-05,
+      "loss": 0.817,
+      "step": 8254
+    },
+    {
+      "epoch": 0.14634600300456002,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.7493143369572013e-05,
+      "loss": 0.8451,
+      "step": 8256
+    },
+    {
+      "epoch": 0.14638145504017158,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.749192428488117e-05,
+      "loss": 0.8197,
+      "step": 8258
+    },
+    {
+      "epoch": 0.14641690707578317,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.7490704919495796e-05,
+      "loss": 0.8499,
+      "step": 8260
+    },
+    {
+      "epoch": 0.14645235911139473,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.748948527343112e-05,
+      "loss": 0.8043,
+      "step": 8262
+    },
+    {
+      "epoch": 0.1464878111470063,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.748826534670234e-05,
+      "loss": 0.8115,
+      "step": 8264
+    },
+    {
+      "epoch": 0.14652326318261788,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.748704513932469e-05,
+      "loss": 0.8248,
+      "step": 8266
+    },
+    {
+      "epoch": 0.14655871521822944,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.748582465131341e-05,
+      "loss": 0.8252,
+      "step": 8268
+    },
+    {
+      "epoch": 0.146594167253841,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.748460388268372e-05,
+      "loss": 0.8104,
+      "step": 8270
+    },
+    {
+      "epoch": 0.1466296192894526,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.748338283345085e-05,
+      "loss": 0.8272,
+      "step": 8272
+    },
+    {
+      "epoch": 0.14666507132506415,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.7482161503630053e-05,
+      "loss": 0.7902,
+      "step": 8274
+    },
+    {
+      "epoch": 0.1467005233606757,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.7480939893236556e-05,
+      "loss": 0.8334,
+      "step": 8276
+    },
+    {
+      "epoch": 0.1467359753962873,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.7479718002285615e-05,
+      "loss": 0.7837,
+      "step": 8278
+    },
+    {
+      "epoch": 0.14677142743189886,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.747849583079248e-05,
+      "loss": 0.7474,
+      "step": 8280
+    },
+    {
+      "epoch": 0.14680687946751042,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.747727337877239e-05,
+      "loss": 0.8168,
+      "step": 8282
+    },
+    {
+      "epoch": 0.14684233150312198,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.747605064624062e-05,
+      "loss": 0.8059,
+      "step": 8284
+    },
+    {
+      "epoch": 0.14687778353873357,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.747482763321241e-05,
+      "loss": 0.8225,
+      "step": 8286
+    },
+    {
+      "epoch": 0.14691323557434513,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.7473604339703034e-05,
+      "loss": 0.8391,
+      "step": 8288
+    },
+    {
+      "epoch": 0.1469486876099567,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.747238076572777e-05,
+      "loss": 0.8136,
+      "step": 8290
+    },
+    {
+      "epoch": 0.14698413964556828,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.747115691130185e-05,
+      "loss": 0.8232,
+      "step": 8292
+    },
+    {
+      "epoch": 0.14701959168117984,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.746993277644059e-05,
+      "loss": 0.8548,
+      "step": 8294
+    },
+    {
+      "epoch": 0.1470550437167914,
+      "grad_norm": 2.75,
+      "learning_rate": 4.746870836115924e-05,
+      "loss": 0.8169,
+      "step": 8296
+    },
+    {
+      "epoch": 0.147090495752403,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.74674836654731e-05,
+      "loss": 0.8498,
+      "step": 8298
+    },
+    {
+      "epoch": 0.14712594778801455,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.7466258689397434e-05,
+      "loss": 0.8027,
+      "step": 8300
+    },
+    {
+      "epoch": 0.1471613998236261,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.7465033432947546e-05,
+      "loss": 0.8488,
+      "step": 8302
+    },
+    {
+      "epoch": 0.1471968518592377,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.746380789613871e-05,
+      "loss": 0.842,
+      "step": 8304
+    },
+    {
+      "epoch": 0.14723230389484926,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.746258207898624e-05,
+      "loss": 0.8198,
+      "step": 8306
+    },
+    {
+      "epoch": 0.14726775593046082,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.746135598150542e-05,
+      "loss": 0.8277,
+      "step": 8308
+    },
+    {
+      "epoch": 0.1473032079660724,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.746012960371156e-05,
+      "loss": 0.8142,
+      "step": 8310
+    },
+    {
+      "epoch": 0.14733866000168397,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.745890294561995e-05,
+      "loss": 0.7867,
+      "step": 8312
+    },
+    {
+      "epoch": 0.14737411203729553,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.745767600724592e-05,
+      "loss": 0.8592,
+      "step": 8314
+    },
+    {
+      "epoch": 0.14740956407290712,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.745644878860478e-05,
+      "loss": 0.8329,
+      "step": 8316
+    },
+    {
+      "epoch": 0.14744501610851868,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.7455221289711814e-05,
+      "loss": 0.799,
+      "step": 8318
+    },
+    {
+      "epoch": 0.14748046814413024,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.745399351058237e-05,
+      "loss": 0.8623,
+      "step": 8320
+    },
+    {
+      "epoch": 0.14751592017974183,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.7452765451231776e-05,
+      "loss": 0.8279,
+      "step": 8322
+    },
+    {
+      "epoch": 0.1475513722153534,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.745153711167534e-05,
+      "loss": 0.8329,
+      "step": 8324
+    },
+    {
+      "epoch": 0.14758682425096495,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.745030849192839e-05,
+      "loss": 0.8266,
+      "step": 8326
+    },
+    {
+      "epoch": 0.14762227628657654,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.744907959200627e-05,
+      "loss": 0.779,
+      "step": 8328
+    },
+    {
+      "epoch": 0.1476577283221881,
+      "grad_norm": 2.75,
+      "learning_rate": 4.744785041192431e-05,
+      "loss": 0.8216,
+      "step": 8330
+    },
+    {
+      "epoch": 0.14769318035779966,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.7446620951697856e-05,
+      "loss": 0.8291,
+      "step": 8332
+    },
+    {
+      "epoch": 0.14772863239341125,
+      "grad_norm": 3.0,
+      "learning_rate": 4.744539121134225e-05,
+      "loss": 0.8277,
+      "step": 8334
+    },
+    {
+      "epoch": 0.1477640844290228,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.744416119087283e-05,
+      "loss": 0.8255,
+      "step": 8336
+    },
+    {
+      "epoch": 0.14779953646463437,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.744293089030496e-05,
+      "loss": 0.8148,
+      "step": 8338
+    },
+    {
+      "epoch": 0.14783498850024596,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.744170030965398e-05,
+      "loss": 0.8207,
+      "step": 8340
+    },
+    {
+      "epoch": 0.14787044053585752,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.7440469448935264e-05,
+      "loss": 0.8454,
+      "step": 8342
+    },
+    {
+      "epoch": 0.14790589257146908,
+      "grad_norm": 2.875,
+      "learning_rate": 4.743923830816416e-05,
+      "loss": 0.8418,
+      "step": 8344
+    },
+    {
+      "epoch": 0.14794134460708067,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.743800688735603e-05,
+      "loss": 0.8134,
+      "step": 8346
+    },
+    {
+      "epoch": 0.14797679664269223,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.743677518652625e-05,
+      "loss": 0.8243,
+      "step": 8348
+    },
+    {
+      "epoch": 0.1480122486783038,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.743554320569019e-05,
+      "loss": 0.7993,
+      "step": 8350
+    },
+    {
+      "epoch": 0.14804770071391538,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.743431094486323e-05,
+      "loss": 0.8062,
+      "step": 8352
+    },
+    {
+      "epoch": 0.14808315274952694,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.743307840406073e-05,
+      "loss": 0.8124,
+      "step": 8354
+    },
+    {
+      "epoch": 0.1481186047851385,
+      "grad_norm": 3.0,
+      "learning_rate": 4.7431845583298084e-05,
+      "loss": 0.8467,
+      "step": 8356
+    },
+    {
+      "epoch": 0.14815405682075009,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.7430612482590685e-05,
+      "loss": 0.7828,
+      "step": 8358
+    },
+    {
+      "epoch": 0.14818950885636165,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.742937910195391e-05,
+      "loss": 0.7977,
+      "step": 8360
+    },
+    {
+      "epoch": 0.1482249608919732,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.742814544140316e-05,
+      "loss": 0.8005,
+      "step": 8362
+    },
+    {
+      "epoch": 0.1482604129275848,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.742691150095383e-05,
+      "loss": 0.8071,
+      "step": 8364
+    },
+    {
+      "epoch": 0.14829586496319636,
+      "grad_norm": 2.875,
+      "learning_rate": 4.74256772806213e-05,
+      "loss": 0.821,
+      "step": 8366
+    },
+    {
+      "epoch": 0.14833131699880792,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.7424442780421003e-05,
+      "loss": 0.835,
+      "step": 8368
+    },
+    {
+      "epoch": 0.1483667690344195,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.742320800036832e-05,
+      "loss": 0.7975,
+      "step": 8370
+    },
+    {
+      "epoch": 0.14840222107003107,
+      "grad_norm": 2.625,
+      "learning_rate": 4.742197294047869e-05,
+      "loss": 0.8017,
+      "step": 8372
+    },
+    {
+      "epoch": 0.14843767310564263,
+      "grad_norm": 2.875,
+      "learning_rate": 4.742073760076749e-05,
+      "loss": 0.863,
+      "step": 8374
+    },
+    {
+      "epoch": 0.14847312514125421,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.741950198125016e-05,
+      "loss": 0.7757,
+      "step": 8376
+    },
+    {
+      "epoch": 0.14850857717686577,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.7418266081942116e-05,
+      "loss": 0.8262,
+      "step": 8378
+    },
+    {
+      "epoch": 0.14854402921247734,
+      "grad_norm": 2.75,
+      "learning_rate": 4.741702990285878e-05,
+      "loss": 0.8187,
+      "step": 8380
+    },
+    {
+      "epoch": 0.14857948124808892,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.7415793444015574e-05,
+      "loss": 0.7889,
+      "step": 8382
+    },
+    {
+      "epoch": 0.14861493328370048,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.741455670542795e-05,
+      "loss": 0.7936,
+      "step": 8384
+    },
+    {
+      "epoch": 0.14865038531931205,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.741331968711131e-05,
+      "loss": 0.8174,
+      "step": 8386
+    },
+    {
+      "epoch": 0.14868583735492363,
+      "grad_norm": 2.75,
+      "learning_rate": 4.741208238908111e-05,
+      "loss": 0.8257,
+      "step": 8388
+    },
+    {
+      "epoch": 0.1487212893905352,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.7410844811352806e-05,
+      "loss": 0.8034,
+      "step": 8390
+    },
+    {
+      "epoch": 0.14875674142614675,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.740960695394181e-05,
+      "loss": 0.8,
+      "step": 8392
+    },
+    {
+      "epoch": 0.14879219346175834,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.7408368816863596e-05,
+      "loss": 0.817,
+      "step": 8394
+    },
+    {
+      "epoch": 0.1488276454973699,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.7407130400133605e-05,
+      "loss": 0.7515,
+      "step": 8396
+    },
+    {
+      "epoch": 0.14886309753298146,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.7405891703767294e-05,
+      "loss": 0.8348,
+      "step": 8398
+    },
+    {
+      "epoch": 0.14889854956859305,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.740465272778012e-05,
+      "loss": 0.851,
+      "step": 8400
+    },
+    {
+      "epoch": 0.1489340016042046,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.740341347218754e-05,
+      "loss": 0.8495,
+      "step": 8402
+    },
+    {
+      "epoch": 0.14896945363981617,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.7402173937005035e-05,
+      "loss": 0.8287,
+      "step": 8404
+    },
+    {
+      "epoch": 0.14900490567542776,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.7400934122248066e-05,
+      "loss": 0.8723,
+      "step": 8406
+    },
+    {
+      "epoch": 0.14904035771103932,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.7399694027932094e-05,
+      "loss": 0.8113,
+      "step": 8408
+    },
+    {
+      "epoch": 0.14907580974665088,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.7398453654072616e-05,
+      "loss": 0.8134,
+      "step": 8410
+    },
+    {
+      "epoch": 0.14911126178226247,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.7397213000685104e-05,
+      "loss": 0.7777,
+      "step": 8412
+    },
+    {
+      "epoch": 0.14914671381787403,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.739597206778503e-05,
+      "loss": 0.8102,
+      "step": 8414
+    },
+    {
+      "epoch": 0.1491821658534856,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.7394730855387895e-05,
+      "loss": 0.8204,
+      "step": 8416
+    },
+    {
+      "epoch": 0.14921761788909718,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.739348936350918e-05,
+      "loss": 0.7826,
+      "step": 8418
+    },
+    {
+      "epoch": 0.14925306992470874,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.7392247592164384e-05,
+      "loss": 0.8538,
+      "step": 8420
+    },
+    {
+      "epoch": 0.1492885219603203,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.739100554136901e-05,
+      "loss": 0.8514,
+      "step": 8422
+    },
+    {
+      "epoch": 0.1493239739959319,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.738976321113854e-05,
+      "loss": 0.8351,
+      "step": 8424
+    },
+    {
+      "epoch": 0.14935942603154345,
+      "grad_norm": 2.625,
+      "learning_rate": 4.738852060148849e-05,
+      "loss": 0.7661,
+      "step": 8426
+    },
+    {
+      "epoch": 0.149394878067155,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.738727771243437e-05,
+      "loss": 0.8071,
+      "step": 8428
+    },
+    {
+      "epoch": 0.1494303301027666,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.7386034543991674e-05,
+      "loss": 0.7708,
+      "step": 8430
+    },
+    {
+      "epoch": 0.14946578213837816,
+      "grad_norm": 2.875,
+      "learning_rate": 4.738479109617594e-05,
+      "loss": 0.8564,
+      "step": 8432
+    },
+    {
+      "epoch": 0.14950123417398972,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.738354736900268e-05,
+      "loss": 0.8329,
+      "step": 8434
+    },
+    {
+      "epoch": 0.1495366862096013,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.73823033624874e-05,
+      "loss": 0.8144,
+      "step": 8436
+    },
+    {
+      "epoch": 0.14957213824521287,
+      "grad_norm": 2.625,
+      "learning_rate": 4.738105907664565e-05,
+      "loss": 0.8032,
+      "step": 8438
+    },
+    {
+      "epoch": 0.14960759028082443,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.737981451149293e-05,
+      "loss": 0.8256,
+      "step": 8440
+    },
+    {
+      "epoch": 0.14964304231643602,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.737856966704479e-05,
+      "loss": 0.7986,
+      "step": 8442
+    },
+    {
+      "epoch": 0.14967849435204758,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.737732454331677e-05,
+      "loss": 0.8465,
+      "step": 8444
+    },
+    {
+      "epoch": 0.14971394638765914,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.737607914032439e-05,
+      "loss": 0.8307,
+      "step": 8446
+    },
+    {
+      "epoch": 0.14974939842327073,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.737483345808321e-05,
+      "loss": 0.8057,
+      "step": 8448
+    },
+    {
+      "epoch": 0.1497848504588823,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.737358749660877e-05,
+      "loss": 0.8325,
+      "step": 8450
+    },
+    {
+      "epoch": 0.14982030249449385,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.737234125591661e-05,
+      "loss": 0.8536,
+      "step": 8452
+    },
+    {
+      "epoch": 0.1498557545301054,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.7371094736022295e-05,
+      "loss": 0.8113,
+      "step": 8454
+    },
+    {
+      "epoch": 0.149891206565717,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.736984793694138e-05,
+      "loss": 0.8725,
+      "step": 8456
+    },
+    {
+      "epoch": 0.14992665860132856,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.736860085868942e-05,
+      "loss": 0.8384,
+      "step": 8458
+    },
+    {
+      "epoch": 0.14996211063694012,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.736735350128199e-05,
+      "loss": 0.8097,
+      "step": 8460
+    },
+    {
+      "epoch": 0.1499975626725517,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.736610586473463e-05,
+      "loss": 0.7781,
+      "step": 8462
+    },
+    {
+      "epoch": 0.15003301470816327,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.736485794906294e-05,
+      "loss": 0.8082,
+      "step": 8464
+    },
+    {
+      "epoch": 0.15006846674377483,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.7363609754282466e-05,
+      "loss": 0.8256,
+      "step": 8466
+    },
+    {
+      "epoch": 0.15010391877938642,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.736236128040882e-05,
+      "loss": 0.8137,
+      "step": 8468
+    },
+    {
+      "epoch": 0.15013937081499798,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.736111252745755e-05,
+      "loss": 0.8189,
+      "step": 8470
+    },
+    {
+      "epoch": 0.15017482285060954,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.7359863495444254e-05,
+      "loss": 0.8005,
+      "step": 8472
+    },
+    {
+      "epoch": 0.15021027488622113,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.735861418438452e-05,
+      "loss": 0.8336,
+      "step": 8474
+    },
+    {
+      "epoch": 0.1502457269218327,
+      "grad_norm": 2.75,
+      "learning_rate": 4.735736459429394e-05,
+      "loss": 0.8146,
+      "step": 8476
+    },
+    {
+      "epoch": 0.15028117895744425,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.735611472518811e-05,
+      "loss": 0.8229,
+      "step": 8478
+    },
+    {
+      "epoch": 0.15031663099305584,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.7354864577082616e-05,
+      "loss": 0.821,
+      "step": 8480
+    },
+    {
+      "epoch": 0.1503520830286674,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.7353614149993074e-05,
+      "loss": 0.7703,
+      "step": 8482
+    },
+    {
+      "epoch": 0.15038753506427896,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.735236344393508e-05,
+      "loss": 0.7681,
+      "step": 8484
+    },
+    {
+      "epoch": 0.15042298709989055,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.735111245892425e-05,
+      "loss": 0.8142,
+      "step": 8486
+    },
+    {
+      "epoch": 0.1504584391355021,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.734986119497619e-05,
+      "loss": 0.8485,
+      "step": 8488
+    },
+    {
+      "epoch": 0.15049389117111367,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.734860965210651e-05,
+      "loss": 0.8056,
+      "step": 8490
+    },
+    {
+      "epoch": 0.15052934320672526,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.734735783033085e-05,
+      "loss": 0.8393,
+      "step": 8492
+    },
+    {
+      "epoch": 0.15056479524233682,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.734610572966481e-05,
+      "loss": 0.8382,
+      "step": 8494
+    },
+    {
+      "epoch": 0.15060024727794838,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.734485335012403e-05,
+      "loss": 0.8161,
+      "step": 8496
+    },
+    {
+      "epoch": 0.15063569931355997,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.734360069172413e-05,
+      "loss": 0.8473,
+      "step": 8498
+    },
+    {
+      "epoch": 0.15067115134917153,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.7342347754480745e-05,
+      "loss": 0.8288,
+      "step": 8500
+    },
+    {
+      "epoch": 0.1507066033847831,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.734109453840952e-05,
+      "loss": 0.8326,
+      "step": 8502
+    },
+    {
+      "epoch": 0.15074205542039468,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.7339841043526085e-05,
+      "loss": 0.8067,
+      "step": 8504
+    },
+    {
+      "epoch": 0.15077750745600624,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.733858726984609e-05,
+      "loss": 0.7916,
+      "step": 8506
+    },
+    {
+      "epoch": 0.1508129594916178,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.7337333217385173e-05,
+      "loss": 0.816,
+      "step": 8508
+    },
+    {
+      "epoch": 0.1508484115272294,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.7336078886158994e-05,
+      "loss": 0.8172,
+      "step": 8510
+    },
+    {
+      "epoch": 0.15088386356284095,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.7334824276183195e-05,
+      "loss": 0.7866,
+      "step": 8512
+    },
+    {
+      "epoch": 0.1509193155984525,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.733356938747345e-05,
+      "loss": 0.7956,
+      "step": 8514
+    },
+    {
+      "epoch": 0.1509547676340641,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.7332314220045417e-05,
+      "loss": 0.7696,
+      "step": 8516
+    },
+    {
+      "epoch": 0.15099021966967566,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.7331058773914736e-05,
+      "loss": 0.8321,
+      "step": 8518
+    },
+    {
+      "epoch": 0.15102567170528722,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.732980304909711e-05,
+      "loss": 0.8247,
+      "step": 8520
+    },
+    {
+      "epoch": 0.1510611237408988,
+      "grad_norm": 2.625,
+      "learning_rate": 4.7328547045608185e-05,
+      "loss": 0.8107,
+      "step": 8522
+    },
+    {
+      "epoch": 0.15109657577651037,
+      "grad_norm": 2.75,
+      "learning_rate": 4.7327290763463636e-05,
+      "loss": 0.8323,
+      "step": 8524
+    },
+    {
+      "epoch": 0.15113202781212193,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.732603420267916e-05,
+      "loss": 0.8015,
+      "step": 8526
+    },
+    {
+      "epoch": 0.15116747984773352,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.7324777363270424e-05,
+      "loss": 0.8537,
+      "step": 8528
+    },
+    {
+      "epoch": 0.15120293188334508,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.7323520245253114e-05,
+      "loss": 0.8273,
+      "step": 8530
+    },
+    {
+      "epoch": 0.15123838391895664,
+      "grad_norm": 2.875,
+      "learning_rate": 4.732226284864293e-05,
+      "loss": 0.8292,
+      "step": 8532
+    },
+    {
+      "epoch": 0.15127383595456823,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.7321005173455546e-05,
+      "loss": 0.8115,
+      "step": 8534
+    },
+    {
+      "epoch": 0.15130928799017979,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.731974721970667e-05,
+      "loss": 0.7805,
+      "step": 8536
+    },
+    {
+      "epoch": 0.15134474002579135,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.7318488987411994e-05,
+      "loss": 0.84,
+      "step": 8538
+    },
+    {
+      "epoch": 0.15138019206140294,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.7317230476587225e-05,
+      "loss": 0.8011,
+      "step": 8540
+    },
+    {
+      "epoch": 0.1514156440970145,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.7315971687248076e-05,
+      "loss": 0.7949,
+      "step": 8542
+    },
+    {
+      "epoch": 0.15145109613262606,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.731471261941024e-05,
+      "loss": 0.7571,
+      "step": 8544
+    },
+    {
+      "epoch": 0.15148654816823764,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.7313453273089445e-05,
+      "loss": 0.8316,
+      "step": 8546
+    },
+    {
+      "epoch": 0.1515220002038492,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.73121936483014e-05,
+      "loss": 0.8164,
+      "step": 8548
+    },
+    {
+      "epoch": 0.15155745223946077,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.7310933745061813e-05,
+      "loss": 0.8136,
+      "step": 8550
+    },
+    {
+      "epoch": 0.15159290427507235,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.7309673563386426e-05,
+      "loss": 0.8086,
+      "step": 8552
+    },
+    {
+      "epoch": 0.15162835631068391,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.730841310329096e-05,
+      "loss": 0.8417,
+      "step": 8554
+    },
+    {
+      "epoch": 0.15166380834629548,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.730715236479115e-05,
+      "loss": 0.8629,
+      "step": 8556
+    },
+    {
+      "epoch": 0.15169926038190706,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.730589134790272e-05,
+      "loss": 0.7892,
+      "step": 8558
+    },
+    {
+      "epoch": 0.15173471241751862,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.730463005264142e-05,
+      "loss": 0.8418,
+      "step": 8560
+    },
+    {
+      "epoch": 0.15177016445313019,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.7303368479022974e-05,
+      "loss": 0.8379,
+      "step": 8562
+    },
+    {
+      "epoch": 0.15180561648874177,
+      "grad_norm": 2.625,
+      "learning_rate": 4.730210662706314e-05,
+      "loss": 0.8355,
+      "step": 8564
+    },
+    {
+      "epoch": 0.15184106852435333,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.730084449677766e-05,
+      "loss": 0.8156,
+      "step": 8566
+    },
+    {
+      "epoch": 0.1518765205599649,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.7299582088182284e-05,
+      "loss": 0.8068,
+      "step": 8568
+    },
+    {
+      "epoch": 0.15191197259557648,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.729831940129277e-05,
+      "loss": 0.7711,
+      "step": 8570
+    },
+    {
+      "epoch": 0.15194742463118804,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.729705643612486e-05,
+      "loss": 0.8524,
+      "step": 8572
+    },
+    {
+      "epoch": 0.1519828766667996,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.729579319269435e-05,
+      "loss": 0.7844,
+      "step": 8574
+    },
+    {
+      "epoch": 0.1520183287024112,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.729452967101697e-05,
+      "loss": 0.8294,
+      "step": 8576
+    },
+    {
+      "epoch": 0.15205378073802275,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.729326587110852e-05,
+      "loss": 0.7946,
+      "step": 8578
+    },
+    {
+      "epoch": 0.15208923277363431,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.729200179298474e-05,
+      "loss": 0.8395,
+      "step": 8580
+    },
+    {
+      "epoch": 0.1521246848092459,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.729073743666143e-05,
+      "loss": 0.8131,
+      "step": 8582
+    },
+    {
+      "epoch": 0.15216013684485746,
+      "grad_norm": 2.875,
+      "learning_rate": 4.728947280215435e-05,
+      "loss": 0.7983,
+      "step": 8584
+    },
+    {
+      "epoch": 0.15219558888046902,
+      "grad_norm": 2.5,
+      "learning_rate": 4.72882078894793e-05,
+      "loss": 0.8259,
+      "step": 8586
+    },
+    {
+      "epoch": 0.1522310409160806,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.728694269865205e-05,
+      "loss": 0.7883,
+      "step": 8588
+    },
+    {
+      "epoch": 0.15226649295169217,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.728567722968841e-05,
+      "loss": 0.7958,
+      "step": 8590
+    },
+    {
+      "epoch": 0.15230194498730373,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.728441148260415e-05,
+      "loss": 0.768,
+      "step": 8592
+    },
+    {
+      "epoch": 0.15233739702291532,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.728314545741508e-05,
+      "loss": 0.8115,
+      "step": 8594
+    },
+    {
+      "epoch": 0.15237284905852688,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.728187915413699e-05,
+      "loss": 0.8365,
+      "step": 8596
+    },
+    {
+      "epoch": 0.15240830109413844,
+      "grad_norm": 2.75,
+      "learning_rate": 4.72806125727857e-05,
+      "loss": 0.8323,
+      "step": 8598
+    },
+    {
+      "epoch": 0.15244375312975003,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.7279345713377e-05,
+      "loss": 0.8174,
+      "step": 8600
+    },
+    {
+      "epoch": 0.1524792051653616,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.72780785759267e-05,
+      "loss": 0.7958,
+      "step": 8602
+    },
+    {
+      "epoch": 0.15251465720097315,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.727681116045063e-05,
+      "loss": 0.8139,
+      "step": 8604
+    },
+    {
+      "epoch": 0.15255010923658474,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.727554346696459e-05,
+      "loss": 0.8109,
+      "step": 8606
+    },
+    {
+      "epoch": 0.1525855612721963,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.727427549548441e-05,
+      "loss": 0.8242,
+      "step": 8608
+    },
+    {
+      "epoch": 0.15262101330780786,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.727300724602591e-05,
+      "loss": 0.8205,
+      "step": 8610
+    },
+    {
+      "epoch": 0.15265646534341945,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.727173871860492e-05,
+      "loss": 0.8197,
+      "step": 8612
+    },
+    {
+      "epoch": 0.152691917379031,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.727046991323726e-05,
+      "loss": 0.7871,
+      "step": 8614
+    },
+    {
+      "epoch": 0.15272736941464257,
+      "grad_norm": 2.75,
+      "learning_rate": 4.7269200829938784e-05,
+      "loss": 0.8023,
+      "step": 8616
+    },
+    {
+      "epoch": 0.15276282145025416,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.7267931468725326e-05,
+      "loss": 0.8246,
+      "step": 8618
+    },
+    {
+      "epoch": 0.15279827348586572,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.726666182961271e-05,
+      "loss": 0.8302,
+      "step": 8620
+    },
+    {
+      "epoch": 0.15283372552147728,
+      "grad_norm": 2.75,
+      "learning_rate": 4.7265391912616796e-05,
+      "loss": 0.8324,
+      "step": 8622
+    },
+    {
+      "epoch": 0.15286917755708884,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.726412171775343e-05,
+      "loss": 0.8154,
+      "step": 8624
+    },
+    {
+      "epoch": 0.15290462959270043,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.7262851245038456e-05,
+      "loss": 0.8289,
+      "step": 8626
+    },
+    {
+      "epoch": 0.152940081628312,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.7261580494487745e-05,
+      "loss": 0.8119,
+      "step": 8628
+    },
+    {
+      "epoch": 0.15297553366392355,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.726030946611714e-05,
+      "loss": 0.7956,
+      "step": 8630
+    },
+    {
+      "epoch": 0.15301098569953514,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.7259038159942514e-05,
+      "loss": 0.7943,
+      "step": 8632
+    },
+    {
+      "epoch": 0.1530464377351467,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.725776657597972e-05,
+      "loss": 0.8546,
+      "step": 8634
+    },
+    {
+      "epoch": 0.15308188977075826,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.725649471424464e-05,
+      "loss": 0.8512,
+      "step": 8636
+    },
+    {
+      "epoch": 0.15311734180636985,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.7255222574753144e-05,
+      "loss": 0.8184,
+      "step": 8638
+    },
+    {
+      "epoch": 0.1531527938419814,
+      "grad_norm": 2.75,
+      "learning_rate": 4.7253950157521106e-05,
+      "loss": 0.8186,
+      "step": 8640
+    },
+    {
+      "epoch": 0.15318824587759297,
+      "grad_norm": 2.875,
+      "learning_rate": 4.72526774625644e-05,
+      "loss": 0.811,
+      "step": 8642
+    },
+    {
+      "epoch": 0.15322369791320456,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.725140448989892e-05,
+      "loss": 0.8125,
+      "step": 8644
+    },
+    {
+      "epoch": 0.15325914994881612,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.725013123954054e-05,
+      "loss": 0.8177,
+      "step": 8646
+    },
+    {
+      "epoch": 0.15329460198442768,
+      "grad_norm": 2.625,
+      "learning_rate": 4.724885771150516e-05,
+      "loss": 0.8118,
+      "step": 8648
+    },
+    {
+      "epoch": 0.15333005402003927,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.724758390580867e-05,
+      "loss": 0.7981,
+      "step": 8650
+    },
+    {
+      "epoch": 0.15336550605565083,
+      "grad_norm": 2.625,
+      "learning_rate": 4.724630982246696e-05,
+      "loss": 0.844,
+      "step": 8652
+    },
+    {
+      "epoch": 0.1534009580912624,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.724503546149595e-05,
+      "loss": 0.8135,
+      "step": 8654
+    },
+    {
+      "epoch": 0.15343641012687398,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.724376082291152e-05,
+      "loss": 0.8015,
+      "step": 8656
+    },
+    {
+      "epoch": 0.15347186216248554,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.724248590672959e-05,
+      "loss": 0.8211,
+      "step": 8658
+    },
+    {
+      "epoch": 0.1535073141980971,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.7241210712966075e-05,
+      "loss": 0.7986,
+      "step": 8660
+    },
+    {
+      "epoch": 0.1535427662337087,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.723993524163688e-05,
+      "loss": 0.8151,
+      "step": 8662
+    },
+    {
+      "epoch": 0.15357821826932025,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.723865949275792e-05,
+      "loss": 0.8132,
+      "step": 8664
+    },
+    {
+      "epoch": 0.1536136703049318,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.723738346634513e-05,
+      "loss": 0.797,
+      "step": 8666
+    },
+    {
+      "epoch": 0.1536491223405434,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.723610716241442e-05,
+      "loss": 0.8153,
+      "step": 8668
+    },
+    {
+      "epoch": 0.15368457437615496,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.723483058098173e-05,
+      "loss": 0.8162,
+      "step": 8670
+    },
+    {
+      "epoch": 0.15372002641176652,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.723355372206297e-05,
+      "loss": 0.8034,
+      "step": 8672
+    },
+    {
+      "epoch": 0.1537554784473781,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.723227658567411e-05,
+      "loss": 0.8138,
+      "step": 8674
+    },
+    {
+      "epoch": 0.15379093048298967,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.723099917183106e-05,
+      "loss": 0.8076,
+      "step": 8676
+    },
+    {
+      "epoch": 0.15382638251860123,
+      "grad_norm": 3.0,
+      "learning_rate": 4.7229721480549774e-05,
+      "loss": 0.8268,
+      "step": 8678
+    },
+    {
+      "epoch": 0.15386183455421282,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.722844351184619e-05,
+      "loss": 0.8274,
+      "step": 8680
+    },
+    {
+      "epoch": 0.15389728658982438,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.722716526573626e-05,
+      "loss": 0.7815,
+      "step": 8682
+    },
+    {
+      "epoch": 0.15393273862543594,
+      "grad_norm": 2.875,
+      "learning_rate": 4.722588674223594e-05,
+      "loss": 0.8101,
+      "step": 8684
+    },
+    {
+      "epoch": 0.15396819066104753,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.722460794136117e-05,
+      "loss": 0.8077,
+      "step": 8686
+    },
+    {
+      "epoch": 0.1540036426966591,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.7223328863127944e-05,
+      "loss": 0.8179,
+      "step": 8688
+    },
+    {
+      "epoch": 0.15403909473227065,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.722204950755219e-05,
+      "loss": 0.8085,
+      "step": 8690
+    },
+    {
+      "epoch": 0.15407454676788224,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.722076987464989e-05,
+      "loss": 0.8197,
+      "step": 8692
+    },
+    {
+      "epoch": 0.1541099988034938,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.721948996443701e-05,
+      "loss": 0.7829,
+      "step": 8694
+    },
+    {
+      "epoch": 0.15414545083910536,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.7218209776929525e-05,
+      "loss": 0.818,
+      "step": 8696
+    },
+    {
+      "epoch": 0.15418090287471695,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.7216929312143396e-05,
+      "loss": 0.8165,
+      "step": 8698
+    },
+    {
+      "epoch": 0.1542163549103285,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.721564857009463e-05,
+      "loss": 0.7966,
+      "step": 8700
+    },
+    {
+      "epoch": 0.15425180694594007,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.7214367550799196e-05,
+      "loss": 0.8127,
+      "step": 8702
+    },
+    {
+      "epoch": 0.15428725898155166,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.721308625427309e-05,
+      "loss": 0.8077,
+      "step": 8704
+    },
+    {
+      "epoch": 0.15432271101716322,
+      "grad_norm": 2.875,
+      "learning_rate": 4.7211804680532276e-05,
+      "loss": 0.81,
+      "step": 8706
+    },
+    {
+      "epoch": 0.15435816305277478,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.7210522829592774e-05,
+      "loss": 0.8512,
+      "step": 8708
+    },
+    {
+      "epoch": 0.15439361508838637,
+      "grad_norm": 2.625,
+      "learning_rate": 4.7209240701470584e-05,
+      "loss": 0.8265,
+      "step": 8710
+    },
+    {
+      "epoch": 0.15442906712399793,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.7207958296181676e-05,
+      "loss": 0.8358,
+      "step": 8712
+    },
+    {
+      "epoch": 0.1544645191596095,
+      "grad_norm": 2.5,
+      "learning_rate": 4.7206675613742084e-05,
+      "loss": 0.7722,
+      "step": 8714
+    },
+    {
+      "epoch": 0.15449997119522108,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.7205392654167806e-05,
+      "loss": 0.8044,
+      "step": 8716
+    },
+    {
+      "epoch": 0.15453542323083264,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.7204109417474854e-05,
+      "loss": 0.8226,
+      "step": 8718
+    },
+    {
+      "epoch": 0.1545708752664442,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.7202825903679234e-05,
+      "loss": 0.8345,
+      "step": 8720
+    },
+    {
+      "epoch": 0.15460632730205578,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.720154211279698e-05,
+      "loss": 0.8133,
+      "step": 8722
+    },
+    {
+      "epoch": 0.15464177933766735,
+      "grad_norm": 2.625,
+      "learning_rate": 4.72002580448441e-05,
+      "loss": 0.7914,
+      "step": 8724
+    },
+    {
+      "epoch": 0.1546772313732789,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.719897369983663e-05,
+      "loss": 0.785,
+      "step": 8726
+    },
+    {
+      "epoch": 0.1547126834088905,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.7197689077790585e-05,
+      "loss": 0.789,
+      "step": 8728
+    },
+    {
+      "epoch": 0.15474813544450206,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.719640417872201e-05,
+      "loss": 0.8307,
+      "step": 8730
+    },
+    {
+      "epoch": 0.15478358748011362,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.719511900264693e-05,
+      "loss": 0.8563,
+      "step": 8732
+    },
+    {
+      "epoch": 0.1548190395157252,
+      "grad_norm": 3.125,
+      "learning_rate": 4.719383354958138e-05,
+      "loss": 0.8509,
+      "step": 8734
+    },
+    {
+      "epoch": 0.15485449155133676,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.7192547819541423e-05,
+      "loss": 0.8402,
+      "step": 8736
+    },
+    {
+      "epoch": 0.15488994358694833,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.7191261812543084e-05,
+      "loss": 0.8446,
+      "step": 8738
+    },
+    {
+      "epoch": 0.1549253956225599,
+      "grad_norm": 3.0,
+      "learning_rate": 4.718997552860243e-05,
+      "loss": 0.8627,
+      "step": 8740
+    },
+    {
+      "epoch": 0.15496084765817147,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.7188688967735486e-05,
+      "loss": 0.8258,
+      "step": 8742
+    },
+    {
+      "epoch": 0.15499629969378304,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.7187402129958334e-05,
+      "loss": 0.8241,
+      "step": 8744
+    },
+    {
+      "epoch": 0.15503175172939462,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.718611501528703e-05,
+      "loss": 0.834,
+      "step": 8746
+    },
+    {
+      "epoch": 0.15506720376500618,
+      "grad_norm": 2.75,
+      "learning_rate": 4.7184827623737623e-05,
+      "loss": 0.8331,
+      "step": 8748
+    },
+    {
+      "epoch": 0.15510265580061774,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.71835399553262e-05,
+      "loss": 0.804,
+      "step": 8750
+    },
+    {
+      "epoch": 0.15513810783622933,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.718225201006881e-05,
+      "loss": 0.8106,
+      "step": 8752
+    },
+    {
+      "epoch": 0.1551735598718409,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.718096378798153e-05,
+      "loss": 0.839,
+      "step": 8754
+    },
+    {
+      "epoch": 0.15520901190745245,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.717967528908045e-05,
+      "loss": 0.8135,
+      "step": 8756
+    },
+    {
+      "epoch": 0.15524446394306404,
+      "grad_norm": 3.0,
+      "learning_rate": 4.717838651338163e-05,
+      "loss": 0.7937,
+      "step": 8758
+    },
+    {
+      "epoch": 0.1552799159786756,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.717709746090118e-05,
+      "loss": 0.8099,
+      "step": 8760
+    },
+    {
+      "epoch": 0.15531536801428716,
+      "grad_norm": 3.125,
+      "learning_rate": 4.717580813165517e-05,
+      "loss": 0.7994,
+      "step": 8762
+    },
+    {
+      "epoch": 0.15535082004989875,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.717451852565969e-05,
+      "loss": 0.7946,
+      "step": 8764
+    },
+    {
+      "epoch": 0.1553862720855103,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.7173228642930846e-05,
+      "loss": 0.8289,
+      "step": 8766
+    },
+    {
+      "epoch": 0.15542172412112187,
+      "grad_norm": 2.625,
+      "learning_rate": 4.717193848348471e-05,
+      "loss": 0.8171,
+      "step": 8768
+    },
+    {
+      "epoch": 0.15545717615673346,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.7170648047337415e-05,
+      "loss": 0.8237,
+      "step": 8770
+    },
+    {
+      "epoch": 0.15549262819234502,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.7169357334505046e-05,
+      "loss": 0.8451,
+      "step": 8772
+    },
+    {
+      "epoch": 0.15552808022795658,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.7168066345003716e-05,
+      "loss": 0.822,
+      "step": 8774
+    },
+    {
+      "epoch": 0.15556353226356817,
+      "grad_norm": 2.625,
+      "learning_rate": 4.716677507884953e-05,
+      "loss": 0.8134,
+      "step": 8776
+    },
+    {
+      "epoch": 0.15559898429917973,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.7165483536058605e-05,
+      "loss": 0.8004,
+      "step": 8778
+    },
+    {
+      "epoch": 0.1556344363347913,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.716419171664708e-05,
+      "loss": 0.8202,
+      "step": 8780
+    },
+    {
+      "epoch": 0.15566988837040288,
+      "grad_norm": 2.625,
+      "learning_rate": 4.716289962063104e-05,
+      "loss": 0.8158,
+      "step": 8782
+    },
+    {
+      "epoch": 0.15570534040601444,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.716160724802664e-05,
+      "loss": 0.797,
+      "step": 8784
+    },
+    {
+      "epoch": 0.155740792441626,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.716031459884999e-05,
+      "loss": 0.7925,
+      "step": 8786
+    },
+    {
+      "epoch": 0.1557762444772376,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.715902167311723e-05,
+      "loss": 0.8325,
+      "step": 8788
+    },
+    {
+      "epoch": 0.15581169651284915,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.71577284708445e-05,
+      "loss": 0.8167,
+      "step": 8790
+    },
+    {
+      "epoch": 0.1558471485484607,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.7156434992047937e-05,
+      "loss": 0.8647,
+      "step": 8792
+    },
+    {
+      "epoch": 0.1558826005840723,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.715514123674367e-05,
+      "loss": 0.8149,
+      "step": 8794
+    },
+    {
+      "epoch": 0.15591805261968386,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.7153847204947866e-05,
+      "loss": 0.7531,
+      "step": 8796
+    },
+    {
+      "epoch": 0.15595350465529542,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.7152552896676656e-05,
+      "loss": 0.8169,
+      "step": 8798
+    },
+    {
+      "epoch": 0.15598895669090698,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.71512583119462e-05,
+      "loss": 0.8065,
+      "step": 8800
+    },
+    {
+      "epoch": 0.15602440872651857,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.714996345077265e-05,
+      "loss": 0.8222,
+      "step": 8802
+    },
+    {
+      "epoch": 0.15605986076213013,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.714866831317218e-05,
+      "loss": 0.8257,
+      "step": 8804
+    },
+    {
+      "epoch": 0.1560953127977417,
+      "grad_norm": 2.875,
+      "learning_rate": 4.714737289916093e-05,
+      "loss": 0.8318,
+      "step": 8806
+    },
+    {
+      "epoch": 0.15613076483335328,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.714607720875509e-05,
+      "loss": 0.7956,
+      "step": 8808
+    },
+    {
+      "epoch": 0.15616621686896484,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.7144781241970815e-05,
+      "loss": 0.868,
+      "step": 8810
+    },
+    {
+      "epoch": 0.1562016689045764,
+      "grad_norm": 2.625,
+      "learning_rate": 4.7143484998824284e-05,
+      "loss": 0.8525,
+      "step": 8812
+    },
+    {
+      "epoch": 0.156237120940188,
+      "grad_norm": 2.5,
+      "learning_rate": 4.7142188479331674e-05,
+      "loss": 0.8132,
+      "step": 8814
+    },
+    {
+      "epoch": 0.15627257297579955,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.714089168350916e-05,
+      "loss": 0.8109,
+      "step": 8816
+    },
+    {
+      "epoch": 0.1563080250114111,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.713959461137293e-05,
+      "loss": 0.8043,
+      "step": 8818
+    },
+    {
+      "epoch": 0.1563434770470227,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.7138297262939173e-05,
+      "loss": 0.8104,
+      "step": 8820
+    },
+    {
+      "epoch": 0.15637892908263426,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.7136999638224076e-05,
+      "loss": 0.8069,
+      "step": 8822
+    },
+    {
+      "epoch": 0.15641438111824582,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.713570173724383e-05,
+      "loss": 0.8431,
+      "step": 8824
+    },
+    {
+      "epoch": 0.1564498331538574,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.713440356001464e-05,
+      "loss": 0.8172,
+      "step": 8826
+    },
+    {
+      "epoch": 0.15648528518946897,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.713310510655271e-05,
+      "loss": 0.8077,
+      "step": 8828
+    },
+    {
+      "epoch": 0.15652073722508053,
+      "grad_norm": 2.875,
+      "learning_rate": 4.713180637687423e-05,
+      "loss": 0.8036,
+      "step": 8830
+    },
+    {
+      "epoch": 0.15655618926069212,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.713050737099542e-05,
+      "loss": 0.8204,
+      "step": 8832
+    },
+    {
+      "epoch": 0.15659164129630368,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.712920808893249e-05,
+      "loss": 0.8062,
+      "step": 8834
+    },
+    {
+      "epoch": 0.15662709333191524,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.712790853070165e-05,
+      "loss": 0.8581,
+      "step": 8836
+    },
+    {
+      "epoch": 0.15666254536752683,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.712660869631912e-05,
+      "loss": 0.8173,
+      "step": 8838
+    },
+    {
+      "epoch": 0.1566979974031384,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.712530858580111e-05,
+      "loss": 0.7951,
+      "step": 8840
+    },
+    {
+      "epoch": 0.15673344943874995,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.712400819916387e-05,
+      "loss": 0.7993,
+      "step": 8842
+    },
+    {
+      "epoch": 0.15676890147436154,
+      "grad_norm": 2.875,
+      "learning_rate": 4.7122707536423615e-05,
+      "loss": 0.787,
+      "step": 8844
+    },
+    {
+      "epoch": 0.1568043535099731,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.712140659759658e-05,
+      "loss": 0.7964,
+      "step": 8846
+    },
+    {
+      "epoch": 0.15683980554558466,
+      "grad_norm": 2.5,
+      "learning_rate": 4.7120105382698996e-05,
+      "loss": 0.8216,
+      "step": 8848
+    },
+    {
+      "epoch": 0.15687525758119625,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.71188038917471e-05,
+      "loss": 0.8313,
+      "step": 8850
+    },
+    {
+      "epoch": 0.1569107096168078,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.711750212475714e-05,
+      "loss": 0.8019,
+      "step": 8852
+    },
+    {
+      "epoch": 0.15694616165241937,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.711620008174536e-05,
+      "loss": 0.8438,
+      "step": 8854
+    },
+    {
+      "epoch": 0.15698161368803096,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.711489776272802e-05,
+      "loss": 0.8425,
+      "step": 8856
+    },
+    {
+      "epoch": 0.15701706572364252,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.711359516772135e-05,
+      "loss": 0.8696,
+      "step": 8858
+    },
+    {
+      "epoch": 0.15705251775925408,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.711229229674162e-05,
+      "loss": 0.816,
+      "step": 8860
+    },
+    {
+      "epoch": 0.15708796979486567,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.7110989149805095e-05,
+      "loss": 0.8127,
+      "step": 8862
+    },
+    {
+      "epoch": 0.15712342183047723,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.710968572692802e-05,
+      "loss": 0.7671,
+      "step": 8864
+    },
+    {
+      "epoch": 0.1571588738660888,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.710838202812668e-05,
+      "loss": 0.8565,
+      "step": 8866
+    },
+    {
+      "epoch": 0.15719432590170038,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.7107078053417335e-05,
+      "loss": 0.8202,
+      "step": 8868
+    },
+    {
+      "epoch": 0.15722977793731194,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.710577380281626e-05,
+      "loss": 0.851,
+      "step": 8870
+    },
+    {
+      "epoch": 0.1572652299729235,
+      "grad_norm": 2.75,
+      "learning_rate": 4.710446927633973e-05,
+      "loss": 0.763,
+      "step": 8872
+    },
+    {
+      "epoch": 0.1573006820085351,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.7103164474004037e-05,
+      "loss": 0.7902,
+      "step": 8874
+    },
+    {
+      "epoch": 0.15733613404414665,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.710185939582544e-05,
+      "loss": 0.8135,
+      "step": 8876
+    },
+    {
+      "epoch": 0.1573715860797582,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.7100554041820255e-05,
+      "loss": 0.8237,
+      "step": 8878
+    },
+    {
+      "epoch": 0.1574070381153698,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.709924841200475e-05,
+      "loss": 0.7817,
+      "step": 8880
+    },
+    {
+      "epoch": 0.15744249015098136,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.709794250639523e-05,
+      "loss": 0.7556,
+      "step": 8882
+    },
+    {
+      "epoch": 0.15747794218659292,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.709663632500799e-05,
+      "loss": 0.8198,
+      "step": 8884
+    },
+    {
+      "epoch": 0.1575133942222045,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.7095329867859335e-05,
+      "loss": 0.7842,
+      "step": 8886
+    },
+    {
+      "epoch": 0.15754884625781607,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.709402313496556e-05,
+      "loss": 0.8254,
+      "step": 8888
+    },
+    {
+      "epoch": 0.15758429829342763,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.709271612634298e-05,
+      "loss": 0.8064,
+      "step": 8890
+    },
+    {
+      "epoch": 0.15761975032903922,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.7091408842007904e-05,
+      "loss": 0.7782,
+      "step": 8892
+    },
+    {
+      "epoch": 0.15765520236465078,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.709010128197665e-05,
+      "loss": 0.8461,
+      "step": 8894
+    },
+    {
+      "epoch": 0.15769065440026234,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.708879344626553e-05,
+      "loss": 0.8172,
+      "step": 8896
+    },
+    {
+      "epoch": 0.15772610643587393,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.7087485334890866e-05,
+      "loss": 0.7925,
+      "step": 8898
+    },
+    {
+      "epoch": 0.15776155847148549,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.708617694786899e-05,
+      "loss": 0.839,
+      "step": 8900
+    },
+    {
+      "epoch": 0.15779701050709705,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.7084868285216234e-05,
+      "loss": 0.818,
+      "step": 8902
+    },
+    {
+      "epoch": 0.15783246254270863,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.70835593469489e-05,
+      "loss": 0.8456,
+      "step": 8904
+    },
+    {
+      "epoch": 0.1578679145783202,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.708225013308336e-05,
+      "loss": 0.8227,
+      "step": 8906
+    },
+    {
+      "epoch": 0.15790336661393176,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.708094064363594e-05,
+      "loss": 0.83,
+      "step": 8908
+    },
+    {
+      "epoch": 0.15793881864954334,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.707963087862297e-05,
+      "loss": 0.8352,
+      "step": 8910
+    },
+    {
+      "epoch": 0.1579742706851549,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.7078320838060816e-05,
+      "loss": 0.8151,
+      "step": 8912
+    },
+    {
+      "epoch": 0.15800972272076647,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.7077010521965816e-05,
+      "loss": 0.8006,
+      "step": 8914
+    },
+    {
+      "epoch": 0.15804517475637805,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.707569993035431e-05,
+      "loss": 0.8574,
+      "step": 8916
+    },
+    {
+      "epoch": 0.15808062679198961,
+      "grad_norm": 2.75,
+      "learning_rate": 4.707438906324267e-05,
+      "loss": 0.7731,
+      "step": 8918
+    },
+    {
+      "epoch": 0.15811607882760118,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.707307792064727e-05,
+      "loss": 0.8476,
+      "step": 8920
+    },
+    {
+      "epoch": 0.15815153086321276,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.707176650258444e-05,
+      "loss": 0.8064,
+      "step": 8922
+    },
+    {
+      "epoch": 0.15818698289882432,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.707045480907056e-05,
+      "loss": 0.8639,
+      "step": 8924
+    },
+    {
+      "epoch": 0.15822243493443588,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.706914284012201e-05,
+      "loss": 0.8088,
+      "step": 8926
+    },
+    {
+      "epoch": 0.15825788697004747,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.706783059575515e-05,
+      "loss": 0.7744,
+      "step": 8928
+    },
+    {
+      "epoch": 0.15829333900565903,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.706651807598635e-05,
+      "loss": 0.8192,
+      "step": 8930
+    },
+    {
+      "epoch": 0.1583287910412706,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.706520528083202e-05,
+      "loss": 0.8281,
+      "step": 8932
+    },
+    {
+      "epoch": 0.15836424307688218,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.706389221030851e-05,
+      "loss": 0.7792,
+      "step": 8934
+    },
+    {
+      "epoch": 0.15839969511249374,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.706257886443222e-05,
+      "loss": 0.7977,
+      "step": 8936
+    },
+    {
+      "epoch": 0.1584351471481053,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.706126524321954e-05,
+      "loss": 0.8396,
+      "step": 8938
+    },
+    {
+      "epoch": 0.1584705991837169,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.705995134668688e-05,
+      "loss": 0.8214,
+      "step": 8940
+    },
+    {
+      "epoch": 0.15850605121932845,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.7058637174850604e-05,
+      "loss": 0.823,
+      "step": 8942
+    },
+    {
+      "epoch": 0.15854150325494,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.7057322727727145e-05,
+      "loss": 0.8337,
+      "step": 8944
+    },
+    {
+      "epoch": 0.1585769552905516,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.7056008005332886e-05,
+      "loss": 0.8593,
+      "step": 8946
+    },
+    {
+      "epoch": 0.15861240732616316,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.7054693007684245e-05,
+      "loss": 0.8306,
+      "step": 8948
+    },
+    {
+      "epoch": 0.15864785936177472,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.705337773479762e-05,
+      "loss": 0.8124,
+      "step": 8950
+    },
+    {
+      "epoch": 0.1586833113973863,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.7052062186689435e-05,
+      "loss": 0.8111,
+      "step": 8952
+    },
+    {
+      "epoch": 0.15871876343299787,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.705074636337612e-05,
+      "loss": 0.808,
+      "step": 8954
+    },
+    {
+      "epoch": 0.15875421546860943,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.704943026487407e-05,
+      "loss": 0.8285,
+      "step": 8956
+    },
+    {
+      "epoch": 0.15878966750422102,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.704811389119973e-05,
+      "loss": 0.8562,
+      "step": 8958
+    },
+    {
+      "epoch": 0.15882511953983258,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.704679724236952e-05,
+      "loss": 0.8075,
+      "step": 8960
+    },
+    {
+      "epoch": 0.15886057157544414,
+      "grad_norm": 2.75,
+      "learning_rate": 4.704548031839987e-05,
+      "loss": 0.8646,
+      "step": 8962
+    },
+    {
+      "epoch": 0.15889602361105573,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.704416311930722e-05,
+      "loss": 0.8135,
+      "step": 8964
+    },
+    {
+      "epoch": 0.1589314756466673,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.7042845645108e-05,
+      "loss": 0.8081,
+      "step": 8966
+    },
+    {
+      "epoch": 0.15896692768227885,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.7041527895818664e-05,
+      "loss": 0.7951,
+      "step": 8968
+    },
+    {
+      "epoch": 0.1590023797178904,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.704020987145565e-05,
+      "loss": 0.7943,
+      "step": 8970
+    },
+    {
+      "epoch": 0.159037831753502,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.70388915720354e-05,
+      "loss": 0.7993,
+      "step": 8972
+    },
+    {
+      "epoch": 0.15907328378911356,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.703757299757439e-05,
+      "loss": 0.8204,
+      "step": 8974
+    },
+    {
+      "epoch": 0.15910873582472512,
+      "grad_norm": 2.625,
+      "learning_rate": 4.703625414808904e-05,
+      "loss": 0.8297,
+      "step": 8976
+    },
+    {
+      "epoch": 0.1591441878603367,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.703493502359584e-05,
+      "loss": 0.8265,
+      "step": 8978
+    },
+    {
+      "epoch": 0.15917963989594827,
+      "grad_norm": 2.75,
+      "learning_rate": 4.703361562411124e-05,
+      "loss": 0.8388,
+      "step": 8980
+    },
+    {
+      "epoch": 0.15921509193155983,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.7032295949651693e-05,
+      "loss": 0.8035,
+      "step": 8982
+    },
+    {
+      "epoch": 0.15925054396717142,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.70309760002337e-05,
+      "loss": 0.8611,
+      "step": 8984
+    },
+    {
+      "epoch": 0.15928599600278298,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.702965577587371e-05,
+      "loss": 0.8407,
+      "step": 8986
+    },
+    {
+      "epoch": 0.15932144803839454,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.7028335276588195e-05,
+      "loss": 0.7786,
+      "step": 8988
+    },
+    {
+      "epoch": 0.15935690007400613,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.702701450239365e-05,
+      "loss": 0.8373,
+      "step": 8990
+    },
+    {
+      "epoch": 0.1593923521096177,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.7025693453306555e-05,
+      "loss": 0.8483,
+      "step": 8992
+    },
+    {
+      "epoch": 0.15942780414522925,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.702437212934339e-05,
+      "loss": 0.8224,
+      "step": 8994
+    },
+    {
+      "epoch": 0.15946325618084084,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.702305053052065e-05,
+      "loss": 0.8397,
+      "step": 8996
+    },
+    {
+      "epoch": 0.1594987082164524,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.702172865685483e-05,
+      "loss": 0.8114,
+      "step": 8998
+    },
+    {
+      "epoch": 0.15953416025206396,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.702040650836241e-05,
+      "loss": 0.802,
+      "step": 9000
+    },
+    {
+      "epoch": 0.15956961228767555,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.701908408505992e-05,
+      "loss": 0.8164,
+      "step": 9002
+    },
+    {
+      "epoch": 0.1596050643232871,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.701776138696383e-05,
+      "loss": 0.7726,
+      "step": 9004
+    },
+    {
+      "epoch": 0.15964051635889867,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.7016438414090674e-05,
+      "loss": 0.8649,
+      "step": 9006
+    },
+    {
+      "epoch": 0.15967596839451026,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.7015115166456954e-05,
+      "loss": 0.7944,
+      "step": 9008
+    },
+    {
+      "epoch": 0.15971142043012182,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.701379164407917e-05,
+      "loss": 0.8098,
+      "step": 9010
+    },
+    {
+      "epoch": 0.15974687246573338,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.701246784697386e-05,
+      "loss": 0.8219,
+      "step": 9012
+    },
+    {
+      "epoch": 0.15978232450134497,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.701114377515754e-05,
+      "loss": 0.8315,
+      "step": 9014
+    },
+    {
+      "epoch": 0.15981777653695653,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.7009819428646726e-05,
+      "loss": 0.7966,
+      "step": 9016
+    },
+    {
+      "epoch": 0.1598532285725681,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.7008494807457954e-05,
+      "loss": 0.7995,
+      "step": 9018
+    },
+    {
+      "epoch": 0.15988868060817968,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.700716991160775e-05,
+      "loss": 0.8324,
+      "step": 9020
+    },
+    {
+      "epoch": 0.15992413264379124,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.7005844741112646e-05,
+      "loss": 0.841,
+      "step": 9022
+    },
+    {
+      "epoch": 0.1599595846794028,
+      "grad_norm": 2.625,
+      "learning_rate": 4.700451929598918e-05,
+      "loss": 0.8093,
+      "step": 9024
+    },
+    {
+      "epoch": 0.1599950367150144,
+      "grad_norm": 3.125,
+      "learning_rate": 4.70031935762539e-05,
+      "loss": 0.8544,
+      "step": 9026
+    },
+    {
+      "epoch": 0.16003048875062595,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.7001867581923355e-05,
+      "loss": 0.8258,
+      "step": 9028
+    },
+    {
+      "epoch": 0.1600659407862375,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.700054131301407e-05,
+      "loss": 0.7826,
+      "step": 9030
+    },
+    {
+      "epoch": 0.1601013928218491,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.699921476954262e-05,
+      "loss": 0.8419,
+      "step": 9032
+    },
+    {
+      "epoch": 0.16013684485746066,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.699788795152555e-05,
+      "loss": 0.7868,
+      "step": 9034
+    },
+    {
+      "epoch": 0.16017229689307222,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.699656085897942e-05,
+      "loss": 0.8306,
+      "step": 9036
+    },
+    {
+      "epoch": 0.1602077489286838,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.69952334919208e-05,
+      "loss": 0.8282,
+      "step": 9038
+    },
+    {
+      "epoch": 0.16024320096429537,
+      "grad_norm": 2.625,
+      "learning_rate": 4.6993905850366237e-05,
+      "loss": 0.825,
+      "step": 9040
+    },
+    {
+      "epoch": 0.16027865299990693,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.6992577934332315e-05,
+      "loss": 0.814,
+      "step": 9042
+    },
+    {
+      "epoch": 0.16031410503551852,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.6991249743835595e-05,
+      "loss": 0.8185,
+      "step": 9044
+    },
+    {
+      "epoch": 0.16034955707113008,
+      "grad_norm": 2.75,
+      "learning_rate": 4.6989921278892665e-05,
+      "loss": 0.8195,
+      "step": 9046
+    },
+    {
+      "epoch": 0.16038500910674164,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.698859253952009e-05,
+      "loss": 0.8143,
+      "step": 9048
+    },
+    {
+      "epoch": 0.16042046114235323,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.6987263525734474e-05,
+      "loss": 0.7709,
+      "step": 9050
+    },
+    {
+      "epoch": 0.1604559131779648,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.698593423755238e-05,
+      "loss": 0.8342,
+      "step": 9052
+    },
+    {
+      "epoch": 0.16049136521357635,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.6984604674990407e-05,
+      "loss": 0.8358,
+      "step": 9054
+    },
+    {
+      "epoch": 0.16052681724918794,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.698327483806515e-05,
+      "loss": 0.7821,
+      "step": 9056
+    },
+    {
+      "epoch": 0.1605622692847995,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.698194472679319e-05,
+      "loss": 0.8044,
+      "step": 9058
+    },
+    {
+      "epoch": 0.16059772132041106,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.698061434119115e-05,
+      "loss": 0.8232,
+      "step": 9060
+    },
+    {
+      "epoch": 0.16063317335602265,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.697928368127562e-05,
+      "loss": 0.8248,
+      "step": 9062
+    },
+    {
+      "epoch": 0.1606686253916342,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.6977952747063204e-05,
+      "loss": 0.8056,
+      "step": 9064
+    },
+    {
+      "epoch": 0.16070407742724577,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.697662153857052e-05,
+      "loss": 0.7992,
+      "step": 9066
+    },
+    {
+      "epoch": 0.16073952946285736,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.697529005581417e-05,
+      "loss": 0.8145,
+      "step": 9068
+    },
+    {
+      "epoch": 0.16077498149846892,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.697395829881078e-05,
+      "loss": 0.8375,
+      "step": 9070
+    },
+    {
+      "epoch": 0.16081043353408048,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.697262626757697e-05,
+      "loss": 0.8664,
+      "step": 9072
+    },
+    {
+      "epoch": 0.16084588556969207,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.697129396212936e-05,
+      "loss": 0.8609,
+      "step": 9074
+    },
+    {
+      "epoch": 0.16088133760530363,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.696996138248457e-05,
+      "loss": 0.7714,
+      "step": 9076
+    },
+    {
+      "epoch": 0.1609167896409152,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.696862852865925e-05,
+      "loss": 0.7998,
+      "step": 9078
+    },
+    {
+      "epoch": 0.16095224167652677,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.6967295400670016e-05,
+      "loss": 0.8001,
+      "step": 9080
+    },
+    {
+      "epoch": 0.16098769371213834,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.696596199853351e-05,
+      "loss": 0.8074,
+      "step": 9082
+    },
+    {
+      "epoch": 0.1610231457477499,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.6964628322266374e-05,
+      "loss": 0.842,
+      "step": 9084
+    },
+    {
+      "epoch": 0.16105859778336148,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.696329437188525e-05,
+      "loss": 0.7974,
+      "step": 9086
+    },
+    {
+      "epoch": 0.16109404981897305,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.696196014740679e-05,
+      "loss": 0.7829,
+      "step": 9088
+    },
+    {
+      "epoch": 0.1611295018545846,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.696062564884764e-05,
+      "loss": 0.8241,
+      "step": 9090
+    },
+    {
+      "epoch": 0.1611649538901962,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.695929087622446e-05,
+      "loss": 0.8322,
+      "step": 9092
+    },
+    {
+      "epoch": 0.16120040592580775,
+      "grad_norm": 2.75,
+      "learning_rate": 4.6957955829553904e-05,
+      "loss": 0.779,
+      "step": 9094
+    },
+    {
+      "epoch": 0.16123585796141932,
+      "grad_norm": 2.625,
+      "learning_rate": 4.695662050885262e-05,
+      "loss": 0.7793,
+      "step": 9096
+    },
+    {
+      "epoch": 0.1612713099970309,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.69552849141373e-05,
+      "loss": 0.8895,
+      "step": 9098
+    },
+    {
+      "epoch": 0.16130676203264246,
+      "grad_norm": 2.625,
+      "learning_rate": 4.6953949045424587e-05,
+      "loss": 0.8151,
+      "step": 9100
+    },
+    {
+      "epoch": 0.16134221406825403,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.6952612902731165e-05,
+      "loss": 0.8219,
+      "step": 9102
+    },
+    {
+      "epoch": 0.1613776661038656,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.6951276486073706e-05,
+      "loss": 0.7823,
+      "step": 9104
+    },
+    {
+      "epoch": 0.16141311813947717,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.69499397954689e-05,
+      "loss": 0.845,
+      "step": 9106
+    },
+    {
+      "epoch": 0.16144857017508873,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.6948602830933404e-05,
+      "loss": 0.838,
+      "step": 9108
+    },
+    {
+      "epoch": 0.16148402221070032,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.694726559248392e-05,
+      "loss": 0.7893,
+      "step": 9110
+    },
+    {
+      "epoch": 0.16151947424631188,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.6945928080137134e-05,
+      "loss": 0.8171,
+      "step": 9112
+    },
+    {
+      "epoch": 0.16155492628192344,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.694459029390973e-05,
+      "loss": 0.7906,
+      "step": 9114
+    },
+    {
+      "epoch": 0.16159037831753503,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.694325223381842e-05,
+      "loss": 0.8033,
+      "step": 9116
+    },
+    {
+      "epoch": 0.1616258303531466,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.694191389987988e-05,
+      "loss": 0.8416,
+      "step": 9118
+    },
+    {
+      "epoch": 0.16166128238875815,
+      "grad_norm": 3.0,
+      "learning_rate": 4.694057529211084e-05,
+      "loss": 0.8006,
+      "step": 9120
+    },
+    {
+      "epoch": 0.16169673442436974,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.693923641052798e-05,
+      "loss": 0.8163,
+      "step": 9122
+    },
+    {
+      "epoch": 0.1617321864599813,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.693789725514802e-05,
+      "loss": 0.8369,
+      "step": 9124
+    },
+    {
+      "epoch": 0.16176763849559286,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.693655782598768e-05,
+      "loss": 0.8142,
+      "step": 9126
+    },
+    {
+      "epoch": 0.16180309053120445,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.693521812306366e-05,
+      "loss": 0.8145,
+      "step": 9128
+    },
+    {
+      "epoch": 0.161838542566816,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.6933878146392685e-05,
+      "loss": 0.8574,
+      "step": 9130
+    },
+    {
+      "epoch": 0.16187399460242757,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.693253789599148e-05,
+      "loss": 0.8351,
+      "step": 9132
+    },
+    {
+      "epoch": 0.16190944663803916,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.693119737187677e-05,
+      "loss": 0.8218,
+      "step": 9134
+    },
+    {
+      "epoch": 0.16194489867365072,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.692985657406529e-05,
+      "loss": 0.8088,
+      "step": 9136
+    },
+    {
+      "epoch": 0.16198035070926228,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.692851550257377e-05,
+      "loss": 0.7974,
+      "step": 9138
+    },
+    {
+      "epoch": 0.16201580274487384,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.6927174157418934e-05,
+      "loss": 0.8186,
+      "step": 9140
+    },
+    {
+      "epoch": 0.16205125478048543,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.6925832538617536e-05,
+      "loss": 0.8193,
+      "step": 9142
+    },
+    {
+      "epoch": 0.162086706816097,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.692449064618631e-05,
+      "loss": 0.7591,
+      "step": 9144
+    },
+    {
+      "epoch": 0.16212215885170855,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.692314848014202e-05,
+      "loss": 0.817,
+      "step": 9146
+    },
+    {
+      "epoch": 0.16215761088732014,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.6921806040501394e-05,
+      "loss": 0.8258,
+      "step": 9148
+    },
+    {
+      "epoch": 0.1621930629229317,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.6920463327281196e-05,
+      "loss": 0.8466,
+      "step": 9150
+    },
+    {
+      "epoch": 0.16222851495854326,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.691912034049818e-05,
+      "loss": 0.8232,
+      "step": 9152
+    },
+    {
+      "epoch": 0.16226396699415485,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.691777708016911e-05,
+      "loss": 0.8537,
+      "step": 9154
+    },
+    {
+      "epoch": 0.1622994190297664,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.6916433546310746e-05,
+      "loss": 0.8061,
+      "step": 9156
+    },
+    {
+      "epoch": 0.16233487106537797,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.691508973893985e-05,
+      "loss": 0.8205,
+      "step": 9158
+    },
+    {
+      "epoch": 0.16237032310098956,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.691374565807321e-05,
+      "loss": 0.8232,
+      "step": 9160
+    },
+    {
+      "epoch": 0.16240577513660112,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.691240130372758e-05,
+      "loss": 0.7885,
+      "step": 9162
+    },
+    {
+      "epoch": 0.16244122717221268,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.6911056675919754e-05,
+      "loss": 0.8036,
+      "step": 9164
+    },
+    {
+      "epoch": 0.16247667920782427,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.69097117746665e-05,
+      "loss": 0.807,
+      "step": 9166
+    },
+    {
+      "epoch": 0.16251213124343583,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.69083665999846e-05,
+      "loss": 0.7837,
+      "step": 9168
+    },
+    {
+      "epoch": 0.1625475832790474,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.690702115189086e-05,
+      "loss": 0.8121,
+      "step": 9170
+    },
+    {
+      "epoch": 0.16258303531465898,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.690567543040205e-05,
+      "loss": 0.8367,
+      "step": 9172
+    },
+    {
+      "epoch": 0.16261848735027054,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.690432943553498e-05,
+      "loss": 0.8392,
+      "step": 9174
+    },
+    {
+      "epoch": 0.1626539393858821,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.690298316730644e-05,
+      "loss": 0.8201,
+      "step": 9176
+    },
+    {
+      "epoch": 0.1626893914214937,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.690163662573323e-05,
+      "loss": 0.7778,
+      "step": 9178
+    },
+    {
+      "epoch": 0.16272484345710525,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.690028981083215e-05,
+      "loss": 0.7806,
+      "step": 9180
+    },
+    {
+      "epoch": 0.1627602954927168,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.6898942722620024e-05,
+      "loss": 0.8281,
+      "step": 9182
+    },
+    {
+      "epoch": 0.1627957475283284,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.689759536111364e-05,
+      "loss": 0.78,
+      "step": 9184
+    },
+    {
+      "epoch": 0.16283119956393996,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.6896247726329846e-05,
+      "loss": 0.7946,
+      "step": 9186
+    },
+    {
+      "epoch": 0.16286665159955152,
+      "grad_norm": 2.875,
+      "learning_rate": 4.689489981828543e-05,
+      "loss": 0.8081,
+      "step": 9188
+    },
+    {
+      "epoch": 0.1629021036351631,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.6893551636997223e-05,
+      "loss": 0.8461,
+      "step": 9190
+    },
+    {
+      "epoch": 0.16293755567077467,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.689220318248207e-05,
+      "loss": 0.7735,
+      "step": 9192
+    },
+    {
+      "epoch": 0.16297300770638623,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.689085445475676e-05,
+      "loss": 0.7966,
+      "step": 9194
+    },
+    {
+      "epoch": 0.16300845974199782,
+      "grad_norm": 2.875,
+      "learning_rate": 4.688950545383815e-05,
+      "loss": 0.7951,
+      "step": 9196
+    },
+    {
+      "epoch": 0.16304391177760938,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.688815617974307e-05,
+      "loss": 0.8245,
+      "step": 9198
+    },
+    {
+      "epoch": 0.16307936381322094,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.688680663248837e-05,
+      "loss": 0.8025,
+      "step": 9200
+    },
+    {
+      "epoch": 0.16311481584883253,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.688545681209087e-05,
+      "loss": 0.8242,
+      "step": 9202
+    },
+    {
+      "epoch": 0.1631502678844441,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.6884106718567435e-05,
+      "loss": 0.7914,
+      "step": 9204
+    },
+    {
+      "epoch": 0.16318571992005565,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.688275635193491e-05,
+      "loss": 0.8505,
+      "step": 9206
+    },
+    {
+      "epoch": 0.16322117195566724,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.688140571221014e-05,
+      "loss": 0.8436,
+      "step": 9208
+    },
+    {
+      "epoch": 0.1632566239912788,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.6880054799409976e-05,
+      "loss": 0.8364,
+      "step": 9210
+    },
+    {
+      "epoch": 0.16329207602689036,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.687870361355129e-05,
+      "loss": 0.8385,
+      "step": 9212
+    },
+    {
+      "epoch": 0.16332752806250195,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.6877352154650945e-05,
+      "loss": 0.8277,
+      "step": 9214
+    },
+    {
+      "epoch": 0.1633629800981135,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.6876000422725795e-05,
+      "loss": 0.8034,
+      "step": 9216
+    },
+    {
+      "epoch": 0.16339843213372507,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.6874648417792724e-05,
+      "loss": 0.7925,
+      "step": 9218
+    },
+    {
+      "epoch": 0.16343388416933666,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.687329613986859e-05,
+      "loss": 0.7897,
+      "step": 9220
+    },
+    {
+      "epoch": 0.16346933620494822,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.687194358897028e-05,
+      "loss": 0.812,
+      "step": 9222
+    },
+    {
+      "epoch": 0.16350478824055978,
+      "grad_norm": 2.75,
+      "learning_rate": 4.687059076511467e-05,
+      "loss": 0.7761,
+      "step": 9224
+    },
+    {
+      "epoch": 0.16354024027617137,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.686923766831864e-05,
+      "loss": 0.7916,
+      "step": 9226
+    },
+    {
+      "epoch": 0.16357569231178293,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.686788429859907e-05,
+      "loss": 0.7843,
+      "step": 9228
+    },
+    {
+      "epoch": 0.1636111443473945,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.686653065597287e-05,
+      "loss": 0.8011,
+      "step": 9230
+    },
+    {
+      "epoch": 0.16364659638300608,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.686517674045693e-05,
+      "loss": 0.8351,
+      "step": 9232
+    },
+    {
+      "epoch": 0.16368204841861764,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.686382255206813e-05,
+      "loss": 0.7966,
+      "step": 9234
+    },
+    {
+      "epoch": 0.1637175004542292,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.686246809082337e-05,
+      "loss": 0.8251,
+      "step": 9236
+    },
+    {
+      "epoch": 0.1637529524898408,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.6861113356739574e-05,
+      "loss": 0.807,
+      "step": 9238
+    },
+    {
+      "epoch": 0.16378840452545235,
+      "grad_norm": 3.0,
+      "learning_rate": 4.6859758349833626e-05,
+      "loss": 0.8265,
+      "step": 9240
+    },
+    {
+      "epoch": 0.1638238565610639,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.6858403070122456e-05,
+      "loss": 0.8123,
+      "step": 9242
+    },
+    {
+      "epoch": 0.1638593085966755,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.685704751762296e-05,
+      "loss": 0.8313,
+      "step": 9244
+    },
+    {
+      "epoch": 0.16389476063228706,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.6855691692352074e-05,
+      "loss": 0.7875,
+      "step": 9246
+    },
+    {
+      "epoch": 0.16393021266789862,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.6854335594326704e-05,
+      "loss": 0.8297,
+      "step": 9248
+    },
+    {
+      "epoch": 0.1639656647035102,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.685297922356378e-05,
+      "loss": 0.8047,
+      "step": 9250
+    },
+    {
+      "epoch": 0.16400111673912177,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.685162258008022e-05,
+      "loss": 0.8121,
+      "step": 9252
+    },
+    {
+      "epoch": 0.16403656877473333,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.6850265663892964e-05,
+      "loss": 0.8477,
+      "step": 9254
+    },
+    {
+      "epoch": 0.16407202081034492,
+      "grad_norm": 2.625,
+      "learning_rate": 4.684890847501894e-05,
+      "loss": 0.8041,
+      "step": 9256
+    },
+    {
+      "epoch": 0.16410747284595648,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.68475510134751e-05,
+      "loss": 0.7815,
+      "step": 9258
+    },
+    {
+      "epoch": 0.16414292488156804,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.684619327927836e-05,
+      "loss": 0.8229,
+      "step": 9260
+    },
+    {
+      "epoch": 0.16417837691717962,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.6844835272445686e-05,
+      "loss": 0.8021,
+      "step": 9262
+    },
+    {
+      "epoch": 0.16421382895279119,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.684347699299402e-05,
+      "loss": 0.8477,
+      "step": 9264
+    },
+    {
+      "epoch": 0.16424928098840275,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.6842118440940306e-05,
+      "loss": 0.7934,
+      "step": 9266
+    },
+    {
+      "epoch": 0.16428473302401433,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.684075961630151e-05,
+      "loss": 0.7917,
+      "step": 9268
+    },
+    {
+      "epoch": 0.1643201850596259,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.683940051909458e-05,
+      "loss": 0.7748,
+      "step": 9270
+    },
+    {
+      "epoch": 0.16435563709523746,
+      "grad_norm": 2.875,
+      "learning_rate": 4.6838041149336476e-05,
+      "loss": 0.8254,
+      "step": 9272
+    },
+    {
+      "epoch": 0.16439108913084904,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.683668150704417e-05,
+      "loss": 0.8116,
+      "step": 9274
+    },
+    {
+      "epoch": 0.1644265411664606,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.683532159223463e-05,
+      "loss": 0.8229,
+      "step": 9276
+    },
+    {
+      "epoch": 0.16446199320207217,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.683396140492481e-05,
+      "loss": 0.7687,
+      "step": 9278
+    },
+    {
+      "epoch": 0.16449744523768375,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.683260094513171e-05,
+      "loss": 0.7867,
+      "step": 9280
+    },
+    {
+      "epoch": 0.16453289727329531,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.6831240212872305e-05,
+      "loss": 0.8179,
+      "step": 9282
+    },
+    {
+      "epoch": 0.16456834930890687,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.6829879208163564e-05,
+      "loss": 0.7967,
+      "step": 9284
+    },
+    {
+      "epoch": 0.16460380134451846,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.682851793102248e-05,
+      "loss": 0.7793,
+      "step": 9286
+    },
+    {
+      "epoch": 0.16463925338013002,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.682715638146603e-05,
+      "loss": 0.8315,
+      "step": 9288
+    },
+    {
+      "epoch": 0.16467470541574158,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.682579455951122e-05,
+      "loss": 0.8027,
+      "step": 9290
+    },
+    {
+      "epoch": 0.16471015745135317,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.682443246517503e-05,
+      "loss": 0.8081,
+      "step": 9292
+    },
+    {
+      "epoch": 0.16474560948696473,
+      "grad_norm": 2.625,
+      "learning_rate": 4.682307009847448e-05,
+      "loss": 0.8358,
+      "step": 9294
+    },
+    {
+      "epoch": 0.1647810615225763,
+      "grad_norm": 2.625,
+      "learning_rate": 4.6821707459426556e-05,
+      "loss": 0.8409,
+      "step": 9296
+    },
+    {
+      "epoch": 0.16481651355818788,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.682034454804827e-05,
+      "loss": 0.841,
+      "step": 9298
+    },
+    {
+      "epoch": 0.16485196559379944,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.681898136435663e-05,
+      "loss": 0.736,
+      "step": 9300
+    },
+    {
+      "epoch": 0.164887417629411,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.6817617908368646e-05,
+      "loss": 0.8464,
+      "step": 9302
+    },
+    {
+      "epoch": 0.1649228696650226,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.681625418010134e-05,
+      "loss": 0.8265,
+      "step": 9304
+    },
+    {
+      "epoch": 0.16495832170063415,
+      "grad_norm": 2.875,
+      "learning_rate": 4.6814890179571714e-05,
+      "loss": 0.7931,
+      "step": 9306
+    },
+    {
+      "epoch": 0.1649937737362457,
+      "grad_norm": 3.0,
+      "learning_rate": 4.681352590679681e-05,
+      "loss": 0.8446,
+      "step": 9308
+    },
+    {
+      "epoch": 0.16502922577185727,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.681216136179365e-05,
+      "loss": 0.7961,
+      "step": 9310
+    },
+    {
+      "epoch": 0.16506467780746886,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.681079654457925e-05,
+      "loss": 0.795,
+      "step": 9312
+    },
+    {
+      "epoch": 0.16510012984308042,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.680943145517066e-05,
+      "loss": 0.7826,
+      "step": 9314
+    },
+    {
+      "epoch": 0.16513558187869198,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.68080660935849e-05,
+      "loss": 0.7961,
+      "step": 9316
+    },
+    {
+      "epoch": 0.16517103391430357,
+      "grad_norm": 2.625,
+      "learning_rate": 4.680670045983903e-05,
+      "loss": 0.8179,
+      "step": 9318
+    },
+    {
+      "epoch": 0.16520648594991513,
+      "grad_norm": 2.375,
+      "learning_rate": 4.6805334553950064e-05,
+      "loss": 0.7996,
+      "step": 9320
+    },
+    {
+      "epoch": 0.1652419379855267,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.6803968375935076e-05,
+      "loss": 0.7934,
+      "step": 9322
+    },
+    {
+      "epoch": 0.16527739002113828,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.68026019258111e-05,
+      "loss": 0.8005,
+      "step": 9324
+    },
+    {
+      "epoch": 0.16531284205674984,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.6801235203595195e-05,
+      "loss": 0.8145,
+      "step": 9326
+    },
+    {
+      "epoch": 0.1653482940923614,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.679986820930441e-05,
+      "loss": 0.8435,
+      "step": 9328
+    },
+    {
+      "epoch": 0.165383746127973,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.679850094295581e-05,
+      "loss": 0.7971,
+      "step": 9330
+    },
+    {
+      "epoch": 0.16541919816358455,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.6797133404566466e-05,
+      "loss": 0.7846,
+      "step": 9332
+    },
+    {
+      "epoch": 0.1654546501991961,
+      "grad_norm": 2.875,
+      "learning_rate": 4.679576559415344e-05,
+      "loss": 0.8079,
+      "step": 9334
+    },
+    {
+      "epoch": 0.1654901022348077,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.679439751173379e-05,
+      "loss": 0.8371,
+      "step": 9336
+    },
+    {
+      "epoch": 0.16552555427041926,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.67930291573246e-05,
+      "loss": 0.8212,
+      "step": 9338
+    },
+    {
+      "epoch": 0.16556100630603082,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.679166053094295e-05,
+      "loss": 0.7708,
+      "step": 9340
+    },
+    {
+      "epoch": 0.1655964583416424,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.679029163260591e-05,
+      "loss": 0.8194,
+      "step": 9342
+    },
+    {
+      "epoch": 0.16563191037725397,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.6788922462330575e-05,
+      "loss": 0.7825,
+      "step": 9344
+    },
+    {
+      "epoch": 0.16566736241286553,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.6787553020134025e-05,
+      "loss": 0.8312,
+      "step": 9346
+    },
+    {
+      "epoch": 0.16570281444847712,
+      "grad_norm": 2.625,
+      "learning_rate": 4.6786183306033346e-05,
+      "loss": 0.849,
+      "step": 9348
+    },
+    {
+      "epoch": 0.16573826648408868,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.678481332004564e-05,
+      "loss": 0.8005,
+      "step": 9350
+    },
+    {
+      "epoch": 0.16577371851970024,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.6783443062188e-05,
+      "loss": 0.8119,
+      "step": 9352
+    },
+    {
+      "epoch": 0.16580917055531183,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.678207253247753e-05,
+      "loss": 0.8433,
+      "step": 9354
+    },
+    {
+      "epoch": 0.1658446225909234,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.6780701730931334e-05,
+      "loss": 0.8299,
+      "step": 9356
+    },
+    {
+      "epoch": 0.16588007462653495,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.6779330657566513e-05,
+      "loss": 0.8287,
+      "step": 9358
+    },
+    {
+      "epoch": 0.16591552666214654,
+      "grad_norm": 2.875,
+      "learning_rate": 4.677795931240018e-05,
+      "loss": 0.8628,
+      "step": 9360
+    },
+    {
+      "epoch": 0.1659509786977581,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.6776587695449455e-05,
+      "loss": 0.8101,
+      "step": 9362
+    },
+    {
+      "epoch": 0.16598643073336966,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.677521580673145e-05,
+      "loss": 0.7773,
+      "step": 9364
+    },
+    {
+      "epoch": 0.16602188276898125,
+      "grad_norm": 2.5,
+      "learning_rate": 4.677384364626328e-05,
+      "loss": 0.7934,
+      "step": 9366
+    },
+    {
+      "epoch": 0.1660573348045928,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.6772471214062086e-05,
+      "loss": 0.8239,
+      "step": 9368
+    },
+    {
+      "epoch": 0.16609278684020437,
+      "grad_norm": 3.125,
+      "learning_rate": 4.6771098510144984e-05,
+      "loss": 0.7893,
+      "step": 9370
+    },
+    {
+      "epoch": 0.16612823887581596,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.676972553452911e-05,
+      "loss": 0.8154,
+      "step": 9372
+    },
+    {
+      "epoch": 0.16616369091142752,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.676835228723159e-05,
+      "loss": 0.8276,
+      "step": 9374
+    },
+    {
+      "epoch": 0.16619914294703908,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.676697876826957e-05,
+      "loss": 0.8499,
+      "step": 9376
+    },
+    {
+      "epoch": 0.16623459498265067,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.676560497766019e-05,
+      "loss": 0.8199,
+      "step": 9378
+    },
+    {
+      "epoch": 0.16627004701826223,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.6764230915420596e-05,
+      "loss": 0.8217,
+      "step": 9380
+    },
+    {
+      "epoch": 0.1663054990538738,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.676285658156793e-05,
+      "loss": 0.8098,
+      "step": 9382
+    },
+    {
+      "epoch": 0.16634095108948538,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.6761481976119336e-05,
+      "loss": 0.8065,
+      "step": 9384
+    },
+    {
+      "epoch": 0.16637640312509694,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.6760107099091985e-05,
+      "loss": 0.8115,
+      "step": 9386
+    },
+    {
+      "epoch": 0.1664118551607085,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.675873195050304e-05,
+      "loss": 0.7882,
+      "step": 9388
+    },
+    {
+      "epoch": 0.1664473071963201,
+      "grad_norm": 2.625,
+      "learning_rate": 4.6757356530369646e-05,
+      "loss": 0.7722,
+      "step": 9390
+    },
+    {
+      "epoch": 0.16648275923193165,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.6755980838708974e-05,
+      "loss": 0.7995,
+      "step": 9392
+    },
+    {
+      "epoch": 0.1665182112675432,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.6754604875538187e-05,
+      "loss": 0.7809,
+      "step": 9394
+    },
+    {
+      "epoch": 0.1665536633031548,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.6753228640874474e-05,
+      "loss": 0.7824,
+      "step": 9396
+    },
+    {
+      "epoch": 0.16658911533876636,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.6751852134734994e-05,
+      "loss": 0.8413,
+      "step": 9398
+    },
+    {
+      "epoch": 0.16662456737437792,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.6750475357136925e-05,
+      "loss": 0.8434,
+      "step": 9400
+    },
+    {
+      "epoch": 0.1666600194099895,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.6749098308097464e-05,
+      "loss": 0.7929,
+      "step": 9402
+    },
+    {
+      "epoch": 0.16669547144560107,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.674772098763378e-05,
+      "loss": 0.8189,
+      "step": 9404
+    },
+    {
+      "epoch": 0.16673092348121263,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.674634339576306e-05,
+      "loss": 0.8326,
+      "step": 9406
+    },
+    {
+      "epoch": 0.16676637551682422,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.6744965532502514e-05,
+      "loss": 0.8473,
+      "step": 9408
+    },
+    {
+      "epoch": 0.16680182755243578,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.674358739786933e-05,
+      "loss": 0.8468,
+      "step": 9410
+    },
+    {
+      "epoch": 0.16683727958804734,
+      "grad_norm": 2.75,
+      "learning_rate": 4.6742208991880694e-05,
+      "loss": 0.8275,
+      "step": 9412
+    },
+    {
+      "epoch": 0.16687273162365893,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.6740830314553823e-05,
+      "loss": 0.7788,
+      "step": 9414
+    },
+    {
+      "epoch": 0.1669081836592705,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.6739451365905915e-05,
+      "loss": 0.797,
+      "step": 9416
+    },
+    {
+      "epoch": 0.16694363569488205,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.673807214595419e-05,
+      "loss": 0.8269,
+      "step": 9418
+    },
+    {
+      "epoch": 0.16697908773049364,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.6736692654715845e-05,
+      "loss": 0.8637,
+      "step": 9420
+    },
+    {
+      "epoch": 0.1670145397661052,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.673531289220811e-05,
+      "loss": 0.79,
+      "step": 9422
+    },
+    {
+      "epoch": 0.16704999180171676,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.6733932858448184e-05,
+      "loss": 0.784,
+      "step": 9424
+    },
+    {
+      "epoch": 0.16708544383732835,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.6732552553453316e-05,
+      "loss": 0.7979,
+      "step": 9426
+    },
+    {
+      "epoch": 0.1671208958729399,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.6731171977240713e-05,
+      "loss": 0.8525,
+      "step": 9428
+    },
+    {
+      "epoch": 0.16715634790855147,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.672979112982761e-05,
+      "loss": 0.7991,
+      "step": 9430
+    },
+    {
+      "epoch": 0.16719179994416306,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.6728410011231235e-05,
+      "loss": 0.8529,
+      "step": 9432
+    },
+    {
+      "epoch": 0.16722725197977462,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.672702862146884e-05,
+      "loss": 0.8547,
+      "step": 9434
+    },
+    {
+      "epoch": 0.16726270401538618,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.672564696055765e-05,
+      "loss": 0.8292,
+      "step": 9436
+    },
+    {
+      "epoch": 0.16729815605099776,
+      "grad_norm": 2.875,
+      "learning_rate": 4.672426502851491e-05,
+      "loss": 0.7923,
+      "step": 9438
+    },
+    {
+      "epoch": 0.16733360808660933,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.672288282535786e-05,
+      "loss": 0.7652,
+      "step": 9440
+    },
+    {
+      "epoch": 0.1673690601222209,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.6721500351103766e-05,
+      "loss": 0.8313,
+      "step": 9442
+    },
+    {
+      "epoch": 0.16740451215783247,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.672011760576987e-05,
+      "loss": 0.8155,
+      "step": 9444
+    },
+    {
+      "epoch": 0.16743996419344404,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.671873458937342e-05,
+      "loss": 0.8007,
+      "step": 9446
+    },
+    {
+      "epoch": 0.1674754162290556,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.6717351301931704e-05,
+      "loss": 0.8482,
+      "step": 9448
+    },
+    {
+      "epoch": 0.16751086826466718,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.671596774346196e-05,
+      "loss": 0.8175,
+      "step": 9450
+    },
+    {
+      "epoch": 0.16754632030027874,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.671458391398146e-05,
+      "loss": 0.8452,
+      "step": 9452
+    },
+    {
+      "epoch": 0.1675817723358903,
+      "grad_norm": 2.875,
+      "learning_rate": 4.6713199813507474e-05,
+      "loss": 0.8062,
+      "step": 9454
+    },
+    {
+      "epoch": 0.1676172243715019,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.6711815442057276e-05,
+      "loss": 0.8506,
+      "step": 9456
+    },
+    {
+      "epoch": 0.16765267640711345,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.671043079964815e-05,
+      "loss": 0.7725,
+      "step": 9458
+    },
+    {
+      "epoch": 0.16768812844272502,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.670904588629736e-05,
+      "loss": 0.8275,
+      "step": 9460
+    },
+    {
+      "epoch": 0.1677235804783366,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.6707660702022205e-05,
+      "loss": 0.8068,
+      "step": 9462
+    },
+    {
+      "epoch": 0.16775903251394816,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.670627524683997e-05,
+      "loss": 0.8035,
+      "step": 9464
+    },
+    {
+      "epoch": 0.16779448454955972,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.6704889520767935e-05,
+      "loss": 0.7704,
+      "step": 9466
+    },
+    {
+      "epoch": 0.1678299365851713,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.67035035238234e-05,
+      "loss": 0.8211,
+      "step": 9468
+    },
+    {
+      "epoch": 0.16786538862078287,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.6702117256023665e-05,
+      "loss": 0.846,
+      "step": 9470
+    },
+    {
+      "epoch": 0.16790084065639443,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.6700730717386024e-05,
+      "loss": 0.8129,
+      "step": 9472
+    },
+    {
+      "epoch": 0.16793629269200602,
+      "grad_norm": 2.875,
+      "learning_rate": 4.6699343907927785e-05,
+      "loss": 0.8938,
+      "step": 9474
+    },
+    {
+      "epoch": 0.16797174472761758,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.669795682766625e-05,
+      "loss": 0.7938,
+      "step": 9476
+    },
+    {
+      "epoch": 0.16800719676322914,
+      "grad_norm": 3.125,
+      "learning_rate": 4.6696569476618736e-05,
+      "loss": 0.8289,
+      "step": 9478
+    },
+    {
+      "epoch": 0.1680426487988407,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.669518185480255e-05,
+      "loss": 0.8243,
+      "step": 9480
+    },
+    {
+      "epoch": 0.1680781008344523,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.669379396223502e-05,
+      "loss": 0.8155,
+      "step": 9482
+    },
+    {
+      "epoch": 0.16811355287006385,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.669240579893346e-05,
+      "loss": 0.7802,
+      "step": 9484
+    },
+    {
+      "epoch": 0.16814900490567541,
+      "grad_norm": 2.75,
+      "learning_rate": 4.669101736491519e-05,
+      "loss": 0.8266,
+      "step": 9486
+    },
+    {
+      "epoch": 0.168184456941287,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.6689628660197534e-05,
+      "loss": 0.7776,
+      "step": 9488
+    },
+    {
+      "epoch": 0.16821990897689856,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.668823968479784e-05,
+      "loss": 0.8245,
+      "step": 9490
+    },
+    {
+      "epoch": 0.16825536101251012,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.668685043873343e-05,
+      "loss": 0.7765,
+      "step": 9492
+    },
+    {
+      "epoch": 0.1682908130481217,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.6685460922021644e-05,
+      "loss": 0.8347,
+      "step": 9494
+    },
+    {
+      "epoch": 0.16832626508373327,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.668407113467983e-05,
+      "loss": 0.8044,
+      "step": 9496
+    },
+    {
+      "epoch": 0.16836171711934483,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.668268107672531e-05,
+      "loss": 0.7938,
+      "step": 9498
+    },
+    {
+      "epoch": 0.16839716915495642,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.6681290748175457e-05,
+      "loss": 0.8243,
+      "step": 9500
+    },
+    {
+      "epoch": 0.16843262119056798,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.6679900149047604e-05,
+      "loss": 0.8175,
+      "step": 9502
+    },
+    {
+      "epoch": 0.16846807322617954,
+      "grad_norm": 2.5,
+      "learning_rate": 4.667850927935912e-05,
+      "loss": 0.8063,
+      "step": 9504
+    },
+    {
+      "epoch": 0.16850352526179113,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.6677118139127354e-05,
+      "loss": 0.7868,
+      "step": 9506
+    },
+    {
+      "epoch": 0.1685389772974027,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.6675726728369664e-05,
+      "loss": 0.8022,
+      "step": 9508
+    },
+    {
+      "epoch": 0.16857442933301425,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.667433504710342e-05,
+      "loss": 0.8211,
+      "step": 9510
+    },
+    {
+      "epoch": 0.16860988136862584,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.667294309534599e-05,
+      "loss": 0.8062,
+      "step": 9512
+    },
+    {
+      "epoch": 0.1686453334042374,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.6671550873114744e-05,
+      "loss": 0.7539,
+      "step": 9514
+    },
+    {
+      "epoch": 0.16868078543984896,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.667015838042705e-05,
+      "loss": 0.832,
+      "step": 9516
+    },
+    {
+      "epoch": 0.16871623747546055,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.6668765617300306e-05,
+      "loss": 0.773,
+      "step": 9518
+    },
+    {
+      "epoch": 0.1687516895110721,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.6667372583751875e-05,
+      "loss": 0.832,
+      "step": 9520
+    },
+    {
+      "epoch": 0.16878714154668367,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.6665979279799146e-05,
+      "loss": 0.8179,
+      "step": 9522
+    },
+    {
+      "epoch": 0.16882259358229526,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.666458570545951e-05,
+      "loss": 0.7374,
+      "step": 9524
+    },
+    {
+      "epoch": 0.16885804561790682,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.6663191860750354e-05,
+      "loss": 0.834,
+      "step": 9526
+    },
+    {
+      "epoch": 0.16889349765351838,
+      "grad_norm": 2.875,
+      "learning_rate": 4.666179774568909e-05,
+      "loss": 0.8472,
+      "step": 9528
+    },
+    {
+      "epoch": 0.16892894968912997,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.666040336029308e-05,
+      "loss": 0.7924,
+      "step": 9530
+    },
+    {
+      "epoch": 0.16896440172474153,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.6659008704579756e-05,
+      "loss": 0.8083,
+      "step": 9532
+    },
+    {
+      "epoch": 0.1689998537603531,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.665761377856652e-05,
+      "loss": 0.8192,
+      "step": 9534
+    },
+    {
+      "epoch": 0.16903530579596468,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.6656218582270775e-05,
+      "loss": 0.8148,
+      "step": 9536
+    },
+    {
+      "epoch": 0.16907075783157624,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.665482311570992e-05,
+      "loss": 0.8115,
+      "step": 9538
+    },
+    {
+      "epoch": 0.1691062098671878,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.6653427378901395e-05,
+      "loss": 0.8163,
+      "step": 9540
+    },
+    {
+      "epoch": 0.1691416619027994,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.665203137186261e-05,
+      "loss": 0.8194,
+      "step": 9542
+    },
+    {
+      "epoch": 0.16917711393841095,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.665063509461097e-05,
+      "loss": 0.7891,
+      "step": 9544
+    },
+    {
+      "epoch": 0.1692125659740225,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.664923854716392e-05,
+      "loss": 0.7962,
+      "step": 9546
+    },
+    {
+      "epoch": 0.1692480180096341,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.664784172953888e-05,
+      "loss": 0.7783,
+      "step": 9548
+    },
+    {
+      "epoch": 0.16928347004524566,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.664644464175328e-05,
+      "loss": 0.8456,
+      "step": 9550
+    },
+    {
+      "epoch": 0.16931892208085722,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.664504728382457e-05,
+      "loss": 0.7876,
+      "step": 9552
+    },
+    {
+      "epoch": 0.1693543741164688,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.664364965577018e-05,
+      "loss": 0.8043,
+      "step": 9554
+    },
+    {
+      "epoch": 0.16938982615208037,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.664225175760754e-05,
+      "loss": 0.7579,
+      "step": 9556
+    },
+    {
+      "epoch": 0.16942527818769193,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.6640853589354114e-05,
+      "loss": 0.8425,
+      "step": 9558
+    },
+    {
+      "epoch": 0.16946073022330352,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.663945515102733e-05,
+      "loss": 0.755,
+      "step": 9560
+    },
+    {
+      "epoch": 0.16949618225891508,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.663805644264467e-05,
+      "loss": 0.815,
+      "step": 9562
+    },
+    {
+      "epoch": 0.16953163429452664,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.663665746422356e-05,
+      "loss": 0.7912,
+      "step": 9564
+    },
+    {
+      "epoch": 0.16956708633013823,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.6635258215781476e-05,
+      "loss": 0.8059,
+      "step": 9566
+    },
+    {
+      "epoch": 0.1696025383657498,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.663385869733587e-05,
+      "loss": 0.8317,
+      "step": 9568
+    },
+    {
+      "epoch": 0.16963799040136135,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.663245890890423e-05,
+      "loss": 0.7885,
+      "step": 9570
+    },
+    {
+      "epoch": 0.16967344243697294,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.663105885050399e-05,
+      "loss": 0.7997,
+      "step": 9572
+    },
+    {
+      "epoch": 0.1697088944725845,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.662965852215265e-05,
+      "loss": 0.8182,
+      "step": 9574
+    },
+    {
+      "epoch": 0.16974434650819606,
+      "grad_norm": 2.625,
+      "learning_rate": 4.6628257923867676e-05,
+      "loss": 0.7913,
+      "step": 9576
+    },
+    {
+      "epoch": 0.16977979854380765,
+      "grad_norm": 2.625,
+      "learning_rate": 4.6626857055666546e-05,
+      "loss": 0.7741,
+      "step": 9578
+    },
+    {
+      "epoch": 0.1698152505794192,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.662545591756675e-05,
+      "loss": 0.7387,
+      "step": 9580
+    },
+    {
+      "epoch": 0.16985070261503077,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.6624054509585755e-05,
+      "loss": 0.8389,
+      "step": 9582
+    },
+    {
+      "epoch": 0.16988615465064236,
+      "grad_norm": 3.0,
+      "learning_rate": 4.6622652831741074e-05,
+      "loss": 0.8325,
+      "step": 9584
+    },
+    {
+      "epoch": 0.16992160668625392,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.6621250884050195e-05,
+      "loss": 0.7699,
+      "step": 9586
+    },
+    {
+      "epoch": 0.16995705872186548,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.6619848666530594e-05,
+      "loss": 0.8166,
+      "step": 9588
+    },
+    {
+      "epoch": 0.16999251075747707,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.6618446179199795e-05,
+      "loss": 0.832,
+      "step": 9590
+    },
+    {
+      "epoch": 0.17002796279308863,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.661704342207529e-05,
+      "loss": 0.7962,
+      "step": 9592
+    },
+    {
+      "epoch": 0.1700634148287002,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.661564039517458e-05,
+      "loss": 0.7887,
+      "step": 9594
+    },
+    {
+      "epoch": 0.17009886686431178,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.661423709851518e-05,
+      "loss": 0.817,
+      "step": 9596
+    },
+    {
+      "epoch": 0.17013431889992334,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.661283353211461e-05,
+      "loss": 0.8294,
+      "step": 9598
+    },
+    {
+      "epoch": 0.1701697709355349,
+      "grad_norm": 2.5,
+      "learning_rate": 4.661142969599037e-05,
+      "loss": 0.7648,
+      "step": 9600
+    },
+    {
+      "epoch": 0.17020522297114649,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.661002559016e-05,
+      "loss": 0.8282,
+      "step": 9602
+    },
+    {
+      "epoch": 0.17024067500675805,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.6608621214641e-05,
+      "loss": 0.8384,
+      "step": 9604
+    },
+    {
+      "epoch": 0.1702761270423696,
+      "grad_norm": 2.875,
+      "learning_rate": 4.660721656945092e-05,
+      "loss": 0.7908,
+      "step": 9606
+    },
+    {
+      "epoch": 0.1703115790779812,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.6605811654607265e-05,
+      "loss": 0.8134,
+      "step": 9608
+    },
+    {
+      "epoch": 0.17034703111359276,
+      "grad_norm": 3.0,
+      "learning_rate": 4.660440647012759e-05,
+      "loss": 0.7794,
+      "step": 9610
+    },
+    {
+      "epoch": 0.17038248314920432,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.6603001016029425e-05,
+      "loss": 0.8215,
+      "step": 9612
+    },
+    {
+      "epoch": 0.1704179351848159,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.66015952923303e-05,
+      "loss": 0.8334,
+      "step": 9614
+    },
+    {
+      "epoch": 0.17045338722042747,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.660018929904776e-05,
+      "loss": 0.7975,
+      "step": 9616
+    },
+    {
+      "epoch": 0.17048883925603903,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.659878303619937e-05,
+      "loss": 0.8501,
+      "step": 9618
+    },
+    {
+      "epoch": 0.17052429129165061,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.659737650380265e-05,
+      "loss": 0.7872,
+      "step": 9620
+    },
+    {
+      "epoch": 0.17055974332726218,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.6595969701875184e-05,
+      "loss": 0.8015,
+      "step": 9622
+    },
+    {
+      "epoch": 0.17059519536287374,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.6594562630434505e-05,
+      "loss": 0.7942,
+      "step": 9624
+    },
+    {
+      "epoch": 0.17063064739848532,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.659315528949819e-05,
+      "loss": 0.7957,
+      "step": 9626
+    },
+    {
+      "epoch": 0.17066609943409689,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.659174767908379e-05,
+      "loss": 0.8032,
+      "step": 9628
+    },
+    {
+      "epoch": 0.17070155146970845,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.659033979920888e-05,
+      "loss": 0.8284,
+      "step": 9630
+    },
+    {
+      "epoch": 0.17073700350532003,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.658893164989102e-05,
+      "loss": 0.8575,
+      "step": 9632
+    },
+    {
+      "epoch": 0.1707724555409316,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.658752323114779e-05,
+      "loss": 0.818,
+      "step": 9634
+    },
+    {
+      "epoch": 0.17080790757654316,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.6586114542996776e-05,
+      "loss": 0.8272,
+      "step": 9636
+    },
+    {
+      "epoch": 0.17084335961215474,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.6584705585455534e-05,
+      "loss": 0.8401,
+      "step": 9638
+    },
+    {
+      "epoch": 0.1708788116477663,
+      "grad_norm": 2.625,
+      "learning_rate": 4.6583296358541675e-05,
+      "loss": 0.8057,
+      "step": 9640
+    },
+    {
+      "epoch": 0.17091426368337787,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.658188686227277e-05,
+      "loss": 0.776,
+      "step": 9642
+    },
+    {
+      "epoch": 0.17094971571898945,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.658047709666641e-05,
+      "loss": 0.8276,
+      "step": 9644
+    },
+    {
+      "epoch": 0.17098516775460101,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.657906706174019e-05,
+      "loss": 0.7901,
+      "step": 9646
+    },
+    {
+      "epoch": 0.17102061979021257,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.65776567575117e-05,
+      "loss": 0.8314,
+      "step": 9648
+    },
+    {
+      "epoch": 0.17105607182582414,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.6576246183998564e-05,
+      "loss": 0.7915,
+      "step": 9650
+    },
+    {
+      "epoch": 0.17109152386143572,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.657483534121836e-05,
+      "loss": 0.7905,
+      "step": 9652
+    },
+    {
+      "epoch": 0.17112697589704728,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.65734242291887e-05,
+      "loss": 0.7404,
+      "step": 9654
+    },
+    {
+      "epoch": 0.17116242793265884,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.657201284792721e-05,
+      "loss": 0.829,
+      "step": 9656
+    },
+    {
+      "epoch": 0.17119787996827043,
+      "grad_norm": 2.375,
+      "learning_rate": 4.657060119745149e-05,
+      "loss": 0.7781,
+      "step": 9658
+    },
+    {
+      "epoch": 0.171233332003882,
+      "grad_norm": 2.5,
+      "learning_rate": 4.6569189277779154e-05,
+      "loss": 0.785,
+      "step": 9660
+    },
+    {
+      "epoch": 0.17126878403949355,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.6567777088927836e-05,
+      "loss": 0.7846,
+      "step": 9662
+    },
+    {
+      "epoch": 0.17130423607510514,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.656636463091515e-05,
+      "loss": 0.8225,
+      "step": 9664
+    },
+    {
+      "epoch": 0.1713396881107167,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.656495190375872e-05,
+      "loss": 0.7614,
+      "step": 9666
+    },
+    {
+      "epoch": 0.17137514014632826,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.656353890747619e-05,
+      "loss": 0.8443,
+      "step": 9668
+    },
+    {
+      "epoch": 0.17141059218193985,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.656212564208518e-05,
+      "loss": 0.8243,
+      "step": 9670
+    },
+    {
+      "epoch": 0.1714460442175514,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.6560712107603334e-05,
+      "loss": 0.7915,
+      "step": 9672
+    },
+    {
+      "epoch": 0.17148149625316297,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.655929830404829e-05,
+      "loss": 0.7426,
+      "step": 9674
+    },
+    {
+      "epoch": 0.17151694828877456,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.65578842314377e-05,
+      "loss": 0.7934,
+      "step": 9676
+    },
+    {
+      "epoch": 0.17155240032438612,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.65564698897892e-05,
+      "loss": 0.8462,
+      "step": 9678
+    },
+    {
+      "epoch": 0.17158785235999768,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.6555055279120444e-05,
+      "loss": 0.8183,
+      "step": 9680
+    },
+    {
+      "epoch": 0.17162330439560927,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.655364039944909e-05,
+      "loss": 0.821,
+      "step": 9682
+    },
+    {
+      "epoch": 0.17165875643122083,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.6552225250792794e-05,
+      "loss": 0.7928,
+      "step": 9684
+    },
+    {
+      "epoch": 0.1716942084668324,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.655080983316922e-05,
+      "loss": 0.845,
+      "step": 9686
+    },
+    {
+      "epoch": 0.17172966050244398,
+      "grad_norm": 2.75,
+      "learning_rate": 4.654939414659602e-05,
+      "loss": 0.8374,
+      "step": 9688
+    },
+    {
+      "epoch": 0.17176511253805554,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.654797819109087e-05,
+      "loss": 0.8192,
+      "step": 9690
+    },
+    {
+      "epoch": 0.1718005645736671,
+      "grad_norm": 2.75,
+      "learning_rate": 4.654656196667145e-05,
+      "loss": 0.7864,
+      "step": 9692
+    },
+    {
+      "epoch": 0.1718360166092787,
+      "grad_norm": 2.625,
+      "learning_rate": 4.654514547335541e-05,
+      "loss": 0.8356,
+      "step": 9694
+    },
+    {
+      "epoch": 0.17187146864489025,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.6543728711160456e-05,
+      "loss": 0.819,
+      "step": 9696
+    },
+    {
+      "epoch": 0.1719069206805018,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.654231168010425e-05,
+      "loss": 0.7734,
+      "step": 9698
+    },
+    {
+      "epoch": 0.1719423727161134,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.654089438020448e-05,
+      "loss": 0.8568,
+      "step": 9700
+    },
+    {
+      "epoch": 0.17197782475172496,
+      "grad_norm": 3.0,
+      "learning_rate": 4.653947681147883e-05,
+      "loss": 0.8154,
+      "step": 9702
+    },
+    {
+      "epoch": 0.17201327678733652,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.6538058973945004e-05,
+      "loss": 0.7839,
+      "step": 9704
+    },
+    {
+      "epoch": 0.1720487288229481,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.6536640867620686e-05,
+      "loss": 0.8061,
+      "step": 9706
+    },
+    {
+      "epoch": 0.17208418085855967,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.653522249252357e-05,
+      "loss": 0.8139,
+      "step": 9708
+    },
+    {
+      "epoch": 0.17211963289417123,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.6533803848671366e-05,
+      "loss": 0.8192,
+      "step": 9710
+    },
+    {
+      "epoch": 0.17215508492978282,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.6532384936081777e-05,
+      "loss": 0.8006,
+      "step": 9712
+    },
+    {
+      "epoch": 0.17219053696539438,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.65309657547725e-05,
+      "loss": 0.8358,
+      "step": 9714
+    },
+    {
+      "epoch": 0.17222598900100594,
+      "grad_norm": 2.75,
+      "learning_rate": 4.652954630476127e-05,
+      "loss": 0.8254,
+      "step": 9716
+    },
+    {
+      "epoch": 0.17226144103661753,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.652812658606578e-05,
+      "loss": 0.832,
+      "step": 9718
+    },
+    {
+      "epoch": 0.1722968930722291,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.652670659870375e-05,
+      "loss": 0.7853,
+      "step": 9720
+    },
+    {
+      "epoch": 0.17233234510784065,
+      "grad_norm": 3.125,
+      "learning_rate": 4.652528634269291e-05,
+      "loss": 0.8334,
+      "step": 9722
+    },
+    {
+      "epoch": 0.17236779714345224,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.6523865818050984e-05,
+      "loss": 0.8368,
+      "step": 9724
+    },
+    {
+      "epoch": 0.1724032491790638,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.6522445024795694e-05,
+      "loss": 0.8178,
+      "step": 9726
+    },
+    {
+      "epoch": 0.17243870121467536,
+      "grad_norm": 2.875,
+      "learning_rate": 4.6521023962944765e-05,
+      "loss": 0.819,
+      "step": 9728
+    },
+    {
+      "epoch": 0.17247415325028695,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.651960263251594e-05,
+      "loss": 0.8075,
+      "step": 9730
+    },
+    {
+      "epoch": 0.1725096052858985,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.6518181033526966e-05,
+      "loss": 0.8005,
+      "step": 9732
+    },
+    {
+      "epoch": 0.17254505732151007,
+      "grad_norm": 2.875,
+      "learning_rate": 4.6516759165995563e-05,
+      "loss": 0.8483,
+      "step": 9734
+    },
+    {
+      "epoch": 0.17258050935712166,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.651533702993949e-05,
+      "loss": 0.8286,
+      "step": 9736
+    },
+    {
+      "epoch": 0.17261596139273322,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.65139146253765e-05,
+      "loss": 0.7821,
+      "step": 9738
+    },
+    {
+      "epoch": 0.17265141342834478,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.6512491952324334e-05,
+      "loss": 0.851,
+      "step": 9740
+    },
+    {
+      "epoch": 0.17268686546395637,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.6511069010800745e-05,
+      "loss": 0.7941,
+      "step": 9742
+    },
+    {
+      "epoch": 0.17272231749956793,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.6509645800823494e-05,
+      "loss": 0.7729,
+      "step": 9744
+    },
+    {
+      "epoch": 0.1727577695351795,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.650822232241034e-05,
+      "loss": 0.8009,
+      "step": 9746
+    },
+    {
+      "epoch": 0.17279322157079108,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.650679857557906e-05,
+      "loss": 0.7401,
+      "step": 9748
+    },
+    {
+      "epoch": 0.17282867360640264,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.6505374560347415e-05,
+      "loss": 0.7513,
+      "step": 9750
+    },
+    {
+      "epoch": 0.1728641256420142,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.650395027673317e-05,
+      "loss": 0.8464,
+      "step": 9752
+    },
+    {
+      "epoch": 0.1728995776776258,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.650252572475411e-05,
+      "loss": 0.8083,
+      "step": 9754
+    },
+    {
+      "epoch": 0.17293502971323735,
+      "grad_norm": 3.0,
+      "learning_rate": 4.6501100904427996e-05,
+      "loss": 0.7998,
+      "step": 9756
+    },
+    {
+      "epoch": 0.1729704817488489,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.649967581577263e-05,
+      "loss": 0.7697,
+      "step": 9758
+    },
+    {
+      "epoch": 0.1730059337844605,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.649825045880579e-05,
+      "loss": 0.7944,
+      "step": 9760
+    },
+    {
+      "epoch": 0.17304138582007206,
+      "grad_norm": 2.875,
+      "learning_rate": 4.649682483354525e-05,
+      "loss": 0.8214,
+      "step": 9762
+    },
+    {
+      "epoch": 0.17307683785568362,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.649539894000883e-05,
+      "loss": 0.7868,
+      "step": 9764
+    },
+    {
+      "epoch": 0.1731122898912952,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.6493972778214294e-05,
+      "loss": 0.7774,
+      "step": 9766
+    },
+    {
+      "epoch": 0.17314774192690677,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.649254634817946e-05,
+      "loss": 0.8818,
+      "step": 9768
+    },
+    {
+      "epoch": 0.17318319396251833,
+      "grad_norm": 2.5,
+      "learning_rate": 4.6491119649922124e-05,
+      "loss": 0.7984,
+      "step": 9770
+    },
+    {
+      "epoch": 0.17321864599812992,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.64896926834601e-05,
+      "loss": 0.7955,
+      "step": 9772
+    },
+    {
+      "epoch": 0.17325409803374148,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.648826544881117e-05,
+      "loss": 0.8068,
+      "step": 9774
+    },
+    {
+      "epoch": 0.17328955006935304,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.648683794599318e-05,
+      "loss": 0.8155,
+      "step": 9776
+    },
+    {
+      "epoch": 0.17332500210496463,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.648541017502392e-05,
+      "loss": 0.8327,
+      "step": 9778
+    },
+    {
+      "epoch": 0.1733604541405762,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.648398213592121e-05,
+      "loss": 0.8585,
+      "step": 9780
+    },
+    {
+      "epoch": 0.17339590617618775,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.648255382870288e-05,
+      "loss": 0.8518,
+      "step": 9782
+    },
+    {
+      "epoch": 0.17343135821179934,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.648112525338676e-05,
+      "loss": 0.7884,
+      "step": 9784
+    },
+    {
+      "epoch": 0.1734668102474109,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.647969640999066e-05,
+      "loss": 0.8202,
+      "step": 9786
+    },
+    {
+      "epoch": 0.17350226228302246,
+      "grad_norm": 2.875,
+      "learning_rate": 4.6478267298532434e-05,
+      "loss": 0.8754,
+      "step": 9788
+    },
+    {
+      "epoch": 0.17353771431863405,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.6476837919029904e-05,
+      "loss": 0.8141,
+      "step": 9790
+    },
+    {
+      "epoch": 0.1735731663542456,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.647540827150091e-05,
+      "loss": 0.826,
+      "step": 9792
+    },
+    {
+      "epoch": 0.17360861838985717,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.647397835596329e-05,
+      "loss": 0.804,
+      "step": 9794
+    },
+    {
+      "epoch": 0.17364407042546876,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.647254817243489e-05,
+      "loss": 0.8226,
+      "step": 9796
+    },
+    {
+      "epoch": 0.17367952246108032,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.647111772093356e-05,
+      "loss": 0.8509,
+      "step": 9798
+    },
+    {
+      "epoch": 0.17371497449669188,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.646968700147717e-05,
+      "loss": 0.7859,
+      "step": 9800
+    },
+    {
+      "epoch": 0.17375042653230346,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.6468256014083546e-05,
+      "loss": 0.7955,
+      "step": 9802
+    },
+    {
+      "epoch": 0.17378587856791503,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.6466824758770555e-05,
+      "loss": 0.8422,
+      "step": 9804
+    },
+    {
+      "epoch": 0.17382133060352659,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.6465393235556066e-05,
+      "loss": 0.8353,
+      "step": 9806
+    },
+    {
+      "epoch": 0.17385678263913817,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.6463961444457934e-05,
+      "loss": 0.8482,
+      "step": 9808
+    },
+    {
+      "epoch": 0.17389223467474973,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.646252938549405e-05,
+      "loss": 0.798,
+      "step": 9810
+    },
+    {
+      "epoch": 0.1739276867103613,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.646109705868226e-05,
+      "loss": 0.8061,
+      "step": 9812
+    },
+    {
+      "epoch": 0.17396313874597288,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.645966446404044e-05,
+      "loss": 0.756,
+      "step": 9814
+    },
+    {
+      "epoch": 0.17399859078158444,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.645823160158649e-05,
+      "loss": 0.8085,
+      "step": 9816
+    },
+    {
+      "epoch": 0.174034042817196,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.645679847133827e-05,
+      "loss": 0.83,
+      "step": 9818
+    },
+    {
+      "epoch": 0.17406949485280757,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.645536507331368e-05,
+      "loss": 0.7976,
+      "step": 9820
+    },
+    {
+      "epoch": 0.17410494688841915,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.64539314075306e-05,
+      "loss": 0.8144,
+      "step": 9822
+    },
+    {
+      "epoch": 0.17414039892403071,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.645249747400693e-05,
+      "loss": 0.7811,
+      "step": 9824
+    },
+    {
+      "epoch": 0.17417585095964228,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.645106327276056e-05,
+      "loss": 0.7529,
+      "step": 9826
+    },
+    {
+      "epoch": 0.17421130299525386,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.6449628803809384e-05,
+      "loss": 0.7915,
+      "step": 9828
+    },
+    {
+      "epoch": 0.17424675503086542,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.644819406717131e-05,
+      "loss": 0.7945,
+      "step": 9830
+    },
+    {
+      "epoch": 0.17428220706647699,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.6446759062864236e-05,
+      "loss": 0.8408,
+      "step": 9832
+    },
+    {
+      "epoch": 0.17431765910208857,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.644532379090608e-05,
+      "loss": 0.7478,
+      "step": 9834
+    },
+    {
+      "epoch": 0.17435311113770013,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.644388825131475e-05,
+      "loss": 0.7955,
+      "step": 9836
+    },
+    {
+      "epoch": 0.1743885631733117,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.6442452444108166e-05,
+      "loss": 0.8422,
+      "step": 9838
+    },
+    {
+      "epoch": 0.17442401520892328,
+      "grad_norm": 2.625,
+      "learning_rate": 4.644101636930423e-05,
+      "loss": 0.7869,
+      "step": 9840
+    },
+    {
+      "epoch": 0.17445946724453484,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.643958002692088e-05,
+      "loss": 0.8166,
+      "step": 9842
+    },
+    {
+      "epoch": 0.1744949192801464,
+      "grad_norm": 2.625,
+      "learning_rate": 4.643814341697604e-05,
+      "loss": 0.7817,
+      "step": 9844
+    },
+    {
+      "epoch": 0.174530371315758,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.6436706539487636e-05,
+      "loss": 0.8151,
+      "step": 9846
+    },
+    {
+      "epoch": 0.17456582335136955,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.64352693944736e-05,
+      "loss": 0.7965,
+      "step": 9848
+    },
+    {
+      "epoch": 0.17460127538698111,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.643383198195186e-05,
+      "loss": 0.7691,
+      "step": 9850
+    },
+    {
+      "epoch": 0.1746367274225927,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.643239430194036e-05,
+      "loss": 0.779,
+      "step": 9852
+    },
+    {
+      "epoch": 0.17467217945820426,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.6430956354457054e-05,
+      "loss": 0.7911,
+      "step": 9854
+    },
+    {
+      "epoch": 0.17470763149381582,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.642951813951987e-05,
+      "loss": 0.7949,
+      "step": 9856
+    },
+    {
+      "epoch": 0.1747430835294274,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.642807965714676e-05,
+      "loss": 0.8238,
+      "step": 9858
+    },
+    {
+      "epoch": 0.17477853556503897,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.642664090735569e-05,
+      "loss": 0.7866,
+      "step": 9860
+    },
+    {
+      "epoch": 0.17481398760065053,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.64252018901646e-05,
+      "loss": 0.813,
+      "step": 9862
+    },
+    {
+      "epoch": 0.17484943963626212,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.642376260559145e-05,
+      "loss": 0.7915,
+      "step": 9864
+    },
+    {
+      "epoch": 0.17488489167187368,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.6422323053654205e-05,
+      "loss": 0.8049,
+      "step": 9866
+    },
+    {
+      "epoch": 0.17492034370748524,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.6420883234370826e-05,
+      "loss": 0.8036,
+      "step": 9868
+    },
+    {
+      "epoch": 0.17495579574309683,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.641944314775929e-05,
+      "loss": 0.824,
+      "step": 9870
+    },
+    {
+      "epoch": 0.1749912477787084,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.6418002793837564e-05,
+      "loss": 0.8099,
+      "step": 9872
+    },
+    {
+      "epoch": 0.17502669981431995,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.641656217262362e-05,
+      "loss": 0.8045,
+      "step": 9874
+    },
+    {
+      "epoch": 0.17506215184993154,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.6415121284135453e-05,
+      "loss": 0.8309,
+      "step": 9876
+    },
+    {
+      "epoch": 0.1750976038855431,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.641368012839102e-05,
+      "loss": 0.8056,
+      "step": 9878
+    },
+    {
+      "epoch": 0.17513305592115466,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.641223870540833e-05,
+      "loss": 0.7938,
+      "step": 9880
+    },
+    {
+      "epoch": 0.17516850795676625,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.641079701520535e-05,
+      "loss": 0.7869,
+      "step": 9882
+    },
+    {
+      "epoch": 0.1752039599923778,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.64093550578001e-05,
+      "loss": 0.7944,
+      "step": 9884
+    },
+    {
+      "epoch": 0.17523941202798937,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.640791283321054e-05,
+      "loss": 0.7952,
+      "step": 9886
+    },
+    {
+      "epoch": 0.17527486406360096,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.64064703414547e-05,
+      "loss": 0.8147,
+      "step": 9888
+    },
+    {
+      "epoch": 0.17531031609921252,
+      "grad_norm": 3.0,
+      "learning_rate": 4.6405027582550556e-05,
+      "loss": 0.8132,
+      "step": 9890
+    },
+    {
+      "epoch": 0.17534576813482408,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.640358455651613e-05,
+      "loss": 0.8292,
+      "step": 9892
+    },
+    {
+      "epoch": 0.17538122017043567,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.640214126336943e-05,
+      "loss": 0.7597,
+      "step": 9894
+    },
+    {
+      "epoch": 0.17541667220604723,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.640069770312846e-05,
+      "loss": 0.7683,
+      "step": 9896
+    },
+    {
+      "epoch": 0.1754521242416588,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.639925387581125e-05,
+      "loss": 0.7725,
+      "step": 9898
+    },
+    {
+      "epoch": 0.17548757627727038,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.6397809781435805e-05,
+      "loss": 0.7968,
+      "step": 9900
+    },
+    {
+      "epoch": 0.17552302831288194,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.639636542002015e-05,
+      "loss": 0.7619,
+      "step": 9902
+    },
+    {
+      "epoch": 0.1755584803484935,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.639492079158231e-05,
+      "loss": 0.8393,
+      "step": 9904
+    },
+    {
+      "epoch": 0.1755939323841051,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.6393475896140306e-05,
+      "loss": 0.8553,
+      "step": 9906
+    },
+    {
+      "epoch": 0.17562938441971665,
+      "grad_norm": 2.75,
+      "learning_rate": 4.639203073371219e-05,
+      "loss": 0.8141,
+      "step": 9908
+    },
+    {
+      "epoch": 0.1756648364553282,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.639058530431598e-05,
+      "loss": 0.799,
+      "step": 9910
+    },
+    {
+      "epoch": 0.1757002884909398,
+      "grad_norm": 2.875,
+      "learning_rate": 4.638913960796973e-05,
+      "loss": 0.8397,
+      "step": 9912
+    },
+    {
+      "epoch": 0.17573574052655136,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.6387693644691464e-05,
+      "loss": 0.7984,
+      "step": 9914
+    },
+    {
+      "epoch": 0.17577119256216292,
+      "grad_norm": 2.375,
+      "learning_rate": 4.638624741449924e-05,
+      "loss": 0.7958,
+      "step": 9916
+    },
+    {
+      "epoch": 0.1758066445977745,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.63848009174111e-05,
+      "loss": 0.8291,
+      "step": 9918
+    },
+    {
+      "epoch": 0.17584209663338607,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.63833541534451e-05,
+      "loss": 0.8011,
+      "step": 9920
+    },
+    {
+      "epoch": 0.17587754866899763,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.63819071226193e-05,
+      "loss": 0.7897,
+      "step": 9922
+    },
+    {
+      "epoch": 0.17591300070460922,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.638045982495174e-05,
+      "loss": 0.8142,
+      "step": 9924
+    },
+    {
+      "epoch": 0.17594845274022078,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.637901226046051e-05,
+      "loss": 0.785,
+      "step": 9926
+    },
+    {
+      "epoch": 0.17598390477583234,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.6377564429163645e-05,
+      "loss": 0.7962,
+      "step": 9928
+    },
+    {
+      "epoch": 0.17601935681144393,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.6376116331079235e-05,
+      "loss": 0.7905,
+      "step": 9930
+    },
+    {
+      "epoch": 0.1760548088470555,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.637466796622535e-05,
+      "loss": 0.7657,
+      "step": 9932
+    },
+    {
+      "epoch": 0.17609026088266705,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.637321933462006e-05,
+      "loss": 0.795,
+      "step": 9934
+    },
+    {
+      "epoch": 0.17612571291827864,
+      "grad_norm": 2.625,
+      "learning_rate": 4.6371770436281436e-05,
+      "loss": 0.8107,
+      "step": 9936
+    },
+    {
+      "epoch": 0.1761611649538902,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.637032127122757e-05,
+      "loss": 0.777,
+      "step": 9938
+    },
+    {
+      "epoch": 0.17619661698950176,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.636887183947655e-05,
+      "loss": 0.8037,
+      "step": 9940
+    },
+    {
+      "epoch": 0.17623206902511335,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.6367422141046455e-05,
+      "loss": 0.8038,
+      "step": 9942
+    },
+    {
+      "epoch": 0.1762675210607249,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.6365972175955394e-05,
+      "loss": 0.8317,
+      "step": 9944
+    },
+    {
+      "epoch": 0.17630297309633647,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.636452194422144e-05,
+      "loss": 0.7815,
+      "step": 9946
+    },
+    {
+      "epoch": 0.17633842513194806,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.6363071445862704e-05,
+      "loss": 0.8249,
+      "step": 9948
+    },
+    {
+      "epoch": 0.17637387716755962,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.636162068089729e-05,
+      "loss": 0.7989,
+      "step": 9950
+    },
+    {
+      "epoch": 0.17640932920317118,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.636016964934329e-05,
+      "loss": 0.8027,
+      "step": 9952
+    },
+    {
+      "epoch": 0.17644478123878277,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.635871835121883e-05,
+      "loss": 0.8011,
+      "step": 9954
+    },
+    {
+      "epoch": 0.17648023327439433,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.6357266786542006e-05,
+      "loss": 0.7947,
+      "step": 9956
+    },
+    {
+      "epoch": 0.1765156853100059,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.6355814955330954e-05,
+      "loss": 0.8287,
+      "step": 9958
+    },
+    {
+      "epoch": 0.17655113734561748,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.635436285760377e-05,
+      "loss": 0.7452,
+      "step": 9960
+    },
+    {
+      "epoch": 0.17658658938122904,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.635291049337859e-05,
+      "loss": 0.784,
+      "step": 9962
+    },
+    {
+      "epoch": 0.1766220414168406,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.635145786267353e-05,
+      "loss": 0.7922,
+      "step": 9964
+    },
+    {
+      "epoch": 0.17665749345245219,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.635000496550672e-05,
+      "loss": 0.8115,
+      "step": 9966
+    },
+    {
+      "epoch": 0.17669294548806375,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.63485518018963e-05,
+      "loss": 0.7737,
+      "step": 9968
+    },
+    {
+      "epoch": 0.1767283975236753,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.6347098371860396e-05,
+      "loss": 0.8111,
+      "step": 9970
+    },
+    {
+      "epoch": 0.1767638495592869,
+      "grad_norm": 2.625,
+      "learning_rate": 4.634564467541715e-05,
+      "loss": 0.7941,
+      "step": 9972
+    },
+    {
+      "epoch": 0.17679930159489846,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.634419071258472e-05,
+      "loss": 0.8295,
+      "step": 9974
+    },
+    {
+      "epoch": 0.17683475363051002,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.634273648338122e-05,
+      "loss": 0.8232,
+      "step": 9976
+    },
+    {
+      "epoch": 0.1768702056661216,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.6341281987824817e-05,
+      "loss": 0.7908,
+      "step": 9978
+    },
+    {
+      "epoch": 0.17690565770173317,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.6339827225933665e-05,
+      "loss": 0.8186,
+      "step": 9980
+    },
+    {
+      "epoch": 0.17694110973734473,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.633837219772591e-05,
+      "loss": 0.8195,
+      "step": 9982
+    },
+    {
+      "epoch": 0.17697656177295631,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.633691690321971e-05,
+      "loss": 0.8031,
+      "step": 9984
+    },
+    {
+      "epoch": 0.17701201380856788,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.633546134243324e-05,
+      "loss": 0.8055,
+      "step": 9986
+    },
+    {
+      "epoch": 0.17704746584417944,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.633400551538465e-05,
+      "loss": 0.798,
+      "step": 9988
+    },
+    {
+      "epoch": 0.177082917879791,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.633254942209212e-05,
+      "loss": 0.8628,
+      "step": 9990
+    },
+    {
+      "epoch": 0.17711836991540258,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.633109306257381e-05,
+      "loss": 0.8108,
+      "step": 9992
+    },
+    {
+      "epoch": 0.17715382195101415,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.632963643684791e-05,
+      "loss": 0.791,
+      "step": 9994
+    },
+    {
+      "epoch": 0.1771892739866257,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.632817954493258e-05,
+      "loss": 0.8108,
+      "step": 9996
+    },
+    {
+      "epoch": 0.1772247260222373,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.632672238684602e-05,
+      "loss": 0.8331,
+      "step": 9998
+    },
+    {
+      "epoch": 0.17726017805784886,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.6325264962606395e-05,
+      "loss": 0.7822,
+      "step": 10000
+    },
+    {
+      "epoch": 0.17729563009346042,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.6323807272231915e-05,
+      "loss": 0.7935,
+      "step": 10002
+    },
+    {
+      "epoch": 0.177331082129072,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.6322349315740756e-05,
+      "loss": 0.7826,
+      "step": 10004
+    },
+    {
+      "epoch": 0.17736653416468356,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.632089109315113e-05,
+      "loss": 0.8238,
+      "step": 10006
+    },
+    {
+      "epoch": 0.17740198620029513,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.631943260448122e-05,
+      "loss": 0.8166,
+      "step": 10008
+    },
+    {
+      "epoch": 0.1774374382359067,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.631797384974922e-05,
+      "loss": 0.7851,
+      "step": 10010
+    },
+    {
+      "epoch": 0.17747289027151827,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.631651482897336e-05,
+      "loss": 0.7606,
+      "step": 10012
+    },
+    {
+      "epoch": 0.17750834230712983,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.631505554217183e-05,
+      "loss": 0.8473,
+      "step": 10014
+    },
+    {
+      "epoch": 0.17754379434274142,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.6313595989362844e-05,
+      "loss": 0.8125,
+      "step": 10016
+    },
+    {
+      "epoch": 0.17757924637835298,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.631213617056462e-05,
+      "loss": 0.7986,
+      "step": 10018
+    },
+    {
+      "epoch": 0.17761469841396454,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.6310676085795376e-05,
+      "loss": 0.8402,
+      "step": 10020
+    },
+    {
+      "epoch": 0.17765015044957613,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.630921573507333e-05,
+      "loss": 0.8056,
+      "step": 10022
+    },
+    {
+      "epoch": 0.1776856024851877,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.630775511841672e-05,
+      "loss": 0.7992,
+      "step": 10024
+    },
+    {
+      "epoch": 0.17772105452079925,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.630629423584376e-05,
+      "loss": 0.8099,
+      "step": 10026
+    },
+    {
+      "epoch": 0.17775650655641084,
+      "grad_norm": 2.75,
+      "learning_rate": 4.6304833087372676e-05,
+      "loss": 0.8018,
+      "step": 10028
+    },
+    {
+      "epoch": 0.1777919585920224,
+      "grad_norm": 2.75,
+      "learning_rate": 4.6303371673021726e-05,
+      "loss": 0.8144,
+      "step": 10030
+    },
+    {
+      "epoch": 0.17782741062763396,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.630190999280912e-05,
+      "loss": 0.8091,
+      "step": 10032
+    },
+    {
+      "epoch": 0.17786286266324555,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.630044804675313e-05,
+      "loss": 0.8219,
+      "step": 10034
+    },
+    {
+      "epoch": 0.1778983146988571,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.629898583487198e-05,
+      "loss": 0.8331,
+      "step": 10036
+    },
+    {
+      "epoch": 0.17793376673446867,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.629752335718391e-05,
+      "loss": 0.8037,
+      "step": 10038
+    },
+    {
+      "epoch": 0.17796921877008026,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.62960606137072e-05,
+      "loss": 0.8163,
+      "step": 10040
+    },
+    {
+      "epoch": 0.17800467080569182,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.6294597604460086e-05,
+      "loss": 0.7786,
+      "step": 10042
+    },
+    {
+      "epoch": 0.17804012284130338,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.629313432946083e-05,
+      "loss": 0.8375,
+      "step": 10044
+    },
+    {
+      "epoch": 0.17807557487691497,
+      "grad_norm": 2.75,
+      "learning_rate": 4.629167078872769e-05,
+      "loss": 0.8391,
+      "step": 10046
+    },
+    {
+      "epoch": 0.17811102691252653,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.629020698227893e-05,
+      "loss": 0.7991,
+      "step": 10048
+    },
+    {
+      "epoch": 0.1781464789481381,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.6288742910132834e-05,
+      "loss": 0.7886,
+      "step": 10050
+    },
+    {
+      "epoch": 0.17818193098374968,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.628727857230765e-05,
+      "loss": 0.8491,
+      "step": 10052
+    },
+    {
+      "epoch": 0.17821738301936124,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.628581396882166e-05,
+      "loss": 0.7983,
+      "step": 10054
+    },
+    {
+      "epoch": 0.1782528350549728,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.628434909969315e-05,
+      "loss": 0.8217,
+      "step": 10056
+    },
+    {
+      "epoch": 0.1782882870905844,
+      "grad_norm": 2.625,
+      "learning_rate": 4.62828839649404e-05,
+      "loss": 0.8234,
+      "step": 10058
+    },
+    {
+      "epoch": 0.17832373912619595,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.628141856458168e-05,
+      "loss": 0.8098,
+      "step": 10060
+    },
+    {
+      "epoch": 0.1783591911618075,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.6279952898635305e-05,
+      "loss": 0.7831,
+      "step": 10062
+    },
+    {
+      "epoch": 0.1783946431974191,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.627848696711954e-05,
+      "loss": 0.8053,
+      "step": 10064
+    },
+    {
+      "epoch": 0.17843009523303066,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.6277020770052695e-05,
+      "loss": 0.8616,
+      "step": 10066
+    },
+    {
+      "epoch": 0.17846554726864222,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.627555430745305e-05,
+      "loss": 0.8083,
+      "step": 10068
+    },
+    {
+      "epoch": 0.1785009993042538,
+      "grad_norm": 2.75,
+      "learning_rate": 4.6274087579338934e-05,
+      "loss": 0.7734,
+      "step": 10070
+    },
+    {
+      "epoch": 0.17853645133986537,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.6272620585728626e-05,
+      "loss": 0.7799,
+      "step": 10072
+    },
+    {
+      "epoch": 0.17857190337547693,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.627115332664045e-05,
+      "loss": 0.7787,
+      "step": 10074
+    },
+    {
+      "epoch": 0.17860735541108852,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.626968580209271e-05,
+      "loss": 0.7949,
+      "step": 10076
+    },
+    {
+      "epoch": 0.17864280744670008,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.6268218012103716e-05,
+      "loss": 0.8278,
+      "step": 10078
+    },
+    {
+      "epoch": 0.17867825948231164,
+      "grad_norm": 3.0,
+      "learning_rate": 4.6266749956691794e-05,
+      "loss": 0.8582,
+      "step": 10080
+    },
+    {
+      "epoch": 0.17871371151792323,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.626528163587527e-05,
+      "loss": 0.8595,
+      "step": 10082
+    },
+    {
+      "epoch": 0.1787491635535348,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.626381304967244e-05,
+      "loss": 0.8001,
+      "step": 10084
+    },
+    {
+      "epoch": 0.17878461558914635,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.626234419810167e-05,
+      "loss": 0.7742,
+      "step": 10086
+    },
+    {
+      "epoch": 0.17882006762475794,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.626087508118127e-05,
+      "loss": 0.7925,
+      "step": 10088
+    },
+    {
+      "epoch": 0.1788555196603695,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.625940569892958e-05,
+      "loss": 0.8243,
+      "step": 10090
+    },
+    {
+      "epoch": 0.17889097169598106,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.6257936051364927e-05,
+      "loss": 0.8089,
+      "step": 10092
+    },
+    {
+      "epoch": 0.17892642373159265,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.625646613850566e-05,
+      "loss": 0.815,
+      "step": 10094
+    },
+    {
+      "epoch": 0.1789618757672042,
+      "grad_norm": 3.0,
+      "learning_rate": 4.6254995960370126e-05,
+      "loss": 0.8038,
+      "step": 10096
+    },
+    {
+      "epoch": 0.17899732780281577,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.625352551697667e-05,
+      "loss": 0.7894,
+      "step": 10098
+    },
+    {
+      "epoch": 0.17903277983842736,
+      "grad_norm": 2.875,
+      "learning_rate": 4.6252054808343645e-05,
+      "loss": 0.7992,
+      "step": 10100
+    },
+    {
+      "epoch": 0.17906823187403892,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.62505838344894e-05,
+      "loss": 0.7953,
+      "step": 10102
+    },
+    {
+      "epoch": 0.17910368390965048,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.62491125954323e-05,
+      "loss": 0.8013,
+      "step": 10104
+    },
+    {
+      "epoch": 0.17913913594526207,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.624764109119069e-05,
+      "loss": 0.8029,
+      "step": 10106
+    },
+    {
+      "epoch": 0.17917458798087363,
+      "grad_norm": 2.875,
+      "learning_rate": 4.624616932178295e-05,
+      "loss": 0.8491,
+      "step": 10108
+    },
+    {
+      "epoch": 0.1792100400164852,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.624469728722744e-05,
+      "loss": 0.7669,
+      "step": 10110
+    },
+    {
+      "epoch": 0.17924549205209678,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.624322498754253e-05,
+      "loss": 0.7937,
+      "step": 10112
+    },
+    {
+      "epoch": 0.17928094408770834,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.624175242274661e-05,
+      "loss": 0.8431,
+      "step": 10114
+    },
+    {
+      "epoch": 0.1793163961233199,
+      "grad_norm": 2.625,
+      "learning_rate": 4.624027959285804e-05,
+      "loss": 0.8262,
+      "step": 10116
+    },
+    {
+      "epoch": 0.1793518481589315,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.6238806497895194e-05,
+      "loss": 0.8436,
+      "step": 10118
+    },
+    {
+      "epoch": 0.17938730019454305,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.623733313787647e-05,
+      "loss": 0.8014,
+      "step": 10120
+    },
+    {
+      "epoch": 0.1794227522301546,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.623585951282026e-05,
+      "loss": 0.7842,
+      "step": 10122
+    },
+    {
+      "epoch": 0.1794582042657662,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.623438562274494e-05,
+      "loss": 0.7823,
+      "step": 10124
+    },
+    {
+      "epoch": 0.17949365630137776,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.623291146766892e-05,
+      "loss": 0.8083,
+      "step": 10126
+    },
+    {
+      "epoch": 0.17952910833698932,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.623143704761057e-05,
+      "loss": 0.7822,
+      "step": 10128
+    },
+    {
+      "epoch": 0.1795645603726009,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.622996236258832e-05,
+      "loss": 0.8341,
+      "step": 10130
+    },
+    {
+      "epoch": 0.17960001240821247,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.622848741262056e-05,
+      "loss": 0.8124,
+      "step": 10132
+    },
+    {
+      "epoch": 0.17963546444382403,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.6227012197725695e-05,
+      "loss": 0.8148,
+      "step": 10134
+    },
+    {
+      "epoch": 0.17967091647943562,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.622553671792213e-05,
+      "loss": 0.7565,
+      "step": 10136
+    },
+    {
+      "epoch": 0.17970636851504718,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.6224060973228314e-05,
+      "loss": 0.7274,
+      "step": 10138
+    },
+    {
+      "epoch": 0.17974182055065874,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.622258496366262e-05,
+      "loss": 0.8538,
+      "step": 10140
+    },
+    {
+      "epoch": 0.17977727258627033,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.622110868924349e-05,
+      "loss": 0.8225,
+      "step": 10142
+    },
+    {
+      "epoch": 0.1798127246218819,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.6219632149989336e-05,
+      "loss": 0.7981,
+      "step": 10144
+    },
+    {
+      "epoch": 0.17984817665749345,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.62181553459186e-05,
+      "loss": 0.8034,
+      "step": 10146
+    },
+    {
+      "epoch": 0.17988362869310504,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.6216678277049705e-05,
+      "loss": 0.8419,
+      "step": 10148
+    },
+    {
+      "epoch": 0.1799190807287166,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.621520094340108e-05,
+      "loss": 0.7972,
+      "step": 10150
+    },
+    {
+      "epoch": 0.17995453276432816,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.6213723344991163e-05,
+      "loss": 0.7692,
+      "step": 10152
+    },
+    {
+      "epoch": 0.17998998479993975,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.621224548183841e-05,
+      "loss": 0.7733,
+      "step": 10154
+    },
+    {
+      "epoch": 0.1800254368355513,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.621076735396124e-05,
+      "loss": 0.7879,
+      "step": 10156
+    },
+    {
+      "epoch": 0.18006088887116287,
+      "grad_norm": 2.875,
+      "learning_rate": 4.620928896137812e-05,
+      "loss": 0.7648,
+      "step": 10158
+    },
+    {
+      "epoch": 0.18009634090677445,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.620781030410749e-05,
+      "loss": 0.8248,
+      "step": 10160
+    },
+    {
+      "epoch": 0.18013179294238602,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.62063313821678e-05,
+      "loss": 0.7791,
+      "step": 10162
+    },
+    {
+      "epoch": 0.18016724497799758,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.6204852195577506e-05,
+      "loss": 0.8162,
+      "step": 10164
+    },
+    {
+      "epoch": 0.18020269701360914,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.620337274435508e-05,
+      "loss": 0.8146,
+      "step": 10166
+    },
+    {
+      "epoch": 0.18023814904922072,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.6201893028518986e-05,
+      "loss": 0.825,
+      "step": 10168
+    },
+    {
+      "epoch": 0.18027360108483229,
+      "grad_norm": 2.75,
+      "learning_rate": 4.620041304808767e-05,
+      "loss": 0.7898,
+      "step": 10170
+    },
+    {
+      "epoch": 0.18030905312044385,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.619893280307962e-05,
+      "loss": 0.793,
+      "step": 10172
+    },
+    {
+      "epoch": 0.18034450515605543,
+      "grad_norm": 2.875,
+      "learning_rate": 4.619745229351331e-05,
+      "loss": 0.7651,
+      "step": 10174
+    },
+    {
+      "epoch": 0.180379957191667,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.61959715194072e-05,
+      "loss": 0.7769,
+      "step": 10176
+    },
+    {
+      "epoch": 0.18041540922727856,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.619449048077979e-05,
+      "loss": 0.8221,
+      "step": 10178
+    },
+    {
+      "epoch": 0.18045086126289014,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.619300917764955e-05,
+      "loss": 0.7848,
+      "step": 10180
+    },
+    {
+      "epoch": 0.1804863132985017,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.6191527610034965e-05,
+      "loss": 0.7816,
+      "step": 10182
+    },
+    {
+      "epoch": 0.18052176533411327,
+      "grad_norm": 2.75,
+      "learning_rate": 4.619004577795453e-05,
+      "loss": 0.7701,
+      "step": 10184
+    },
+    {
+      "epoch": 0.18055721736972485,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.618856368142674e-05,
+      "loss": 0.8016,
+      "step": 10186
+    },
+    {
+      "epoch": 0.18059266940533641,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.6187081320470096e-05,
+      "loss": 0.7553,
+      "step": 10188
+    },
+    {
+      "epoch": 0.18062812144094798,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.6185598695103075e-05,
+      "loss": 0.7912,
+      "step": 10190
+    },
+    {
+      "epoch": 0.18066357347655956,
+      "grad_norm": 2.875,
+      "learning_rate": 4.6184115805344206e-05,
+      "loss": 0.8097,
+      "step": 10192
+    },
+    {
+      "epoch": 0.18069902551217112,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.6182632651211976e-05,
+      "loss": 0.7898,
+      "step": 10194
+    },
+    {
+      "epoch": 0.18073447754778268,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.618114923272491e-05,
+      "loss": 0.8266,
+      "step": 10196
+    },
+    {
+      "epoch": 0.18076992958339427,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.6179665549901506e-05,
+      "loss": 0.7944,
+      "step": 10198
+    },
+    {
+      "epoch": 0.18080538161900583,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.617818160276029e-05,
+      "loss": 0.7928,
+      "step": 10200
+    },
+    {
+      "epoch": 0.1808408336546174,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.617669739131979e-05,
+      "loss": 0.8188,
+      "step": 10202
+    },
+    {
+      "epoch": 0.18087628569022898,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.617521291559851e-05,
+      "loss": 0.7336,
+      "step": 10204
+    },
+    {
+      "epoch": 0.18091173772584054,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.617372817561497e-05,
+      "loss": 0.7776,
+      "step": 10206
+    },
+    {
+      "epoch": 0.1809471897614521,
+      "grad_norm": 2.75,
+      "learning_rate": 4.617224317138773e-05,
+      "loss": 0.7534,
+      "step": 10208
+    },
+    {
+      "epoch": 0.1809826417970637,
+      "grad_norm": 3.0,
+      "learning_rate": 4.6170757902935296e-05,
+      "loss": 0.7864,
+      "step": 10210
+    },
+    {
+      "epoch": 0.18101809383267525,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.616927237027622e-05,
+      "loss": 0.8448,
+      "step": 10212
+    },
+    {
+      "epoch": 0.1810535458682868,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.616778657342903e-05,
+      "loss": 0.7974,
+      "step": 10214
+    },
+    {
+      "epoch": 0.1810889979038984,
+      "grad_norm": 2.625,
+      "learning_rate": 4.616630051241227e-05,
+      "loss": 0.8253,
+      "step": 10216
+    },
+    {
+      "epoch": 0.18112444993950996,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.616481418724449e-05,
+      "loss": 0.8596,
+      "step": 10218
+    },
+    {
+      "epoch": 0.18115990197512152,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.616332759794424e-05,
+      "loss": 0.8015,
+      "step": 10220
+    },
+    {
+      "epoch": 0.1811953540107331,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.616184074453006e-05,
+      "loss": 0.857,
+      "step": 10222
+    },
+    {
+      "epoch": 0.18123080604634467,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.616035362702053e-05,
+      "loss": 0.781,
+      "step": 10224
+    },
+    {
+      "epoch": 0.18126625808195623,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.615886624543418e-05,
+      "loss": 0.8163,
+      "step": 10226
+    },
+    {
+      "epoch": 0.18130171011756782,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.615737859978959e-05,
+      "loss": 0.8093,
+      "step": 10228
+    },
+    {
+      "epoch": 0.18133716215317938,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.615589069010533e-05,
+      "loss": 0.8303,
+      "step": 10230
+    },
+    {
+      "epoch": 0.18137261418879094,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.615440251639995e-05,
+      "loss": 0.8302,
+      "step": 10232
+    },
+    {
+      "epoch": 0.18140806622440253,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.6152914078692046e-05,
+      "loss": 0.8092,
+      "step": 10234
+    },
+    {
+      "epoch": 0.1814435182600141,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.615142537700017e-05,
+      "loss": 0.8085,
+      "step": 10236
+    },
+    {
+      "epoch": 0.18147897029562565,
+      "grad_norm": 2.75,
+      "learning_rate": 4.614993641134291e-05,
+      "loss": 0.7855,
+      "step": 10238
+    },
+    {
+      "epoch": 0.18151442233123724,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.614844718173885e-05,
+      "loss": 0.8175,
+      "step": 10240
+    },
+    {
+      "epoch": 0.1815498743668488,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.6146957688206585e-05,
+      "loss": 0.8237,
+      "step": 10242
+    },
+    {
+      "epoch": 0.18158532640246036,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.614546793076467e-05,
+      "loss": 0.8189,
+      "step": 10244
+    },
+    {
+      "epoch": 0.18162077843807195,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.614397790943174e-05,
+      "loss": 0.7813,
+      "step": 10246
+    },
+    {
+      "epoch": 0.1816562304736835,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.6142487624226364e-05,
+      "loss": 0.8565,
+      "step": 10248
+    },
+    {
+      "epoch": 0.18169168250929507,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.614099707516715e-05,
+      "loss": 0.7952,
+      "step": 10250
+    },
+    {
+      "epoch": 0.18172713454490666,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.6139506262272684e-05,
+      "loss": 0.772,
+      "step": 10252
+    },
+    {
+      "epoch": 0.18176258658051822,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.613801518556159e-05,
+      "loss": 0.8336,
+      "step": 10254
+    },
+    {
+      "epoch": 0.18179803861612978,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.613652384505247e-05,
+      "loss": 0.8098,
+      "step": 10256
+    },
+    {
+      "epoch": 0.18183349065174137,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.613503224076393e-05,
+      "loss": 0.7743,
+      "step": 10258
+    },
+    {
+      "epoch": 0.18186894268735293,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.613354037271459e-05,
+      "loss": 0.8096,
+      "step": 10260
+    },
+    {
+      "epoch": 0.1819043947229645,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.6132048240923075e-05,
+      "loss": 0.8208,
+      "step": 10262
+    },
+    {
+      "epoch": 0.18193984675857608,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.6130555845408e-05,
+      "loss": 0.8287,
+      "step": 10264
+    },
+    {
+      "epoch": 0.18197529879418764,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.612906318618798e-05,
+      "loss": 0.8224,
+      "step": 10266
+    },
+    {
+      "epoch": 0.1820107508297992,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.612757026328166e-05,
+      "loss": 0.8103,
+      "step": 10268
+    },
+    {
+      "epoch": 0.1820462028654108,
+      "grad_norm": 2.875,
+      "learning_rate": 4.6126077076707665e-05,
+      "loss": 0.8066,
+      "step": 10270
+    },
+    {
+      "epoch": 0.18208165490102235,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.612458362648462e-05,
+      "loss": 0.7646,
+      "step": 10272
+    },
+    {
+      "epoch": 0.1821171069366339,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.612308991263118e-05,
+      "loss": 0.8349,
+      "step": 10274
+    },
+    {
+      "epoch": 0.1821525589722455,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.612159593516597e-05,
+      "loss": 0.8153,
+      "step": 10276
+    },
+    {
+      "epoch": 0.18218801100785706,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.612010169410764e-05,
+      "loss": 0.7964,
+      "step": 10278
+    },
+    {
+      "epoch": 0.18222346304346862,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.611860718947485e-05,
+      "loss": 0.7736,
+      "step": 10280
+    },
+    {
+      "epoch": 0.1822589150790802,
+      "grad_norm": 2.75,
+      "learning_rate": 4.6117112421286235e-05,
+      "loss": 0.819,
+      "step": 10282
+    },
+    {
+      "epoch": 0.18229436711469177,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.611561738956046e-05,
+      "loss": 0.8259,
+      "step": 10284
+    },
+    {
+      "epoch": 0.18232981915030333,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.611412209431617e-05,
+      "loss": 0.8199,
+      "step": 10286
+    },
+    {
+      "epoch": 0.18236527118591492,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.6112626535572035e-05,
+      "loss": 0.7539,
+      "step": 10288
+    },
+    {
+      "epoch": 0.18240072322152648,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.611113071334673e-05,
+      "loss": 0.7697,
+      "step": 10290
+    },
+    {
+      "epoch": 0.18243617525713804,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.610963462765889e-05,
+      "loss": 0.8111,
+      "step": 10292
+    },
+    {
+      "epoch": 0.18247162729274963,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.6108138278527226e-05,
+      "loss": 0.8284,
+      "step": 10294
+    },
+    {
+      "epoch": 0.1825070793283612,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.610664166597039e-05,
+      "loss": 0.8089,
+      "step": 10296
+    },
+    {
+      "epoch": 0.18254253136397275,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.610514479000706e-05,
+      "loss": 0.8126,
+      "step": 10298
+    },
+    {
+      "epoch": 0.18257798339958434,
+      "grad_norm": 2.875,
+      "learning_rate": 4.610364765065591e-05,
+      "loss": 0.8213,
+      "step": 10300
+    },
+    {
+      "epoch": 0.1826134354351959,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.610215024793564e-05,
+      "loss": 0.7928,
+      "step": 10302
+    },
+    {
+      "epoch": 0.18264888747080746,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.6100652581864925e-05,
+      "loss": 0.7968,
+      "step": 10304
+    },
+    {
+      "epoch": 0.18268433950641905,
+      "grad_norm": 2.625,
+      "learning_rate": 4.6099154652462474e-05,
+      "loss": 0.8257,
+      "step": 10306
+    },
+    {
+      "epoch": 0.1827197915420306,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.609765645974695e-05,
+      "loss": 0.8239,
+      "step": 10308
+    },
+    {
+      "epoch": 0.18275524357764217,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.609615800373708e-05,
+      "loss": 0.8294,
+      "step": 10310
+    },
+    {
+      "epoch": 0.18279069561325376,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.609465928445155e-05,
+      "loss": 0.7955,
+      "step": 10312
+    },
+    {
+      "epoch": 0.18282614764886532,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.609316030190906e-05,
+      "loss": 0.7991,
+      "step": 10314
+    },
+    {
+      "epoch": 0.18286159968447688,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.609166105612833e-05,
+      "loss": 0.7639,
+      "step": 10316
+    },
+    {
+      "epoch": 0.18289705172008847,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.6090161547128065e-05,
+      "loss": 0.8347,
+      "step": 10318
+    },
+    {
+      "epoch": 0.18293250375570003,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.6088661774926975e-05,
+      "loss": 0.7859,
+      "step": 10320
+    },
+    {
+      "epoch": 0.1829679557913116,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.608716173954377e-05,
+      "loss": 0.8036,
+      "step": 10322
+    },
+    {
+      "epoch": 0.18300340782692318,
+      "grad_norm": 2.5,
+      "learning_rate": 4.6085661440997185e-05,
+      "loss": 0.7947,
+      "step": 10324
+    },
+    {
+      "epoch": 0.18303885986253474,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.608416087930594e-05,
+      "loss": 0.7943,
+      "step": 10326
+    },
+    {
+      "epoch": 0.1830743118981463,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.608266005448876e-05,
+      "loss": 0.8049,
+      "step": 10328
+    },
+    {
+      "epoch": 0.18310976393375789,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.608115896656437e-05,
+      "loss": 0.7881,
+      "step": 10330
+    },
+    {
+      "epoch": 0.18314521596936945,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.6079657615551495e-05,
+      "loss": 0.8153,
+      "step": 10332
+    },
+    {
+      "epoch": 0.183180668004981,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.60781560014689e-05,
+      "loss": 0.8135,
+      "step": 10334
+    },
+    {
+      "epoch": 0.18321612004059257,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.607665412433531e-05,
+      "loss": 0.7861,
+      "step": 10336
+    },
+    {
+      "epoch": 0.18325157207620416,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.607515198416945e-05,
+      "loss": 0.8023,
+      "step": 10338
+    },
+    {
+      "epoch": 0.18328702411181572,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.6073649580990096e-05,
+      "loss": 0.7736,
+      "step": 10340
+    },
+    {
+      "epoch": 0.18332247614742728,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.607214691481598e-05,
+      "loss": 0.825,
+      "step": 10342
+    },
+    {
+      "epoch": 0.18335792818303887,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.6070643985665864e-05,
+      "loss": 0.8063,
+      "step": 10344
+    },
+    {
+      "epoch": 0.18339338021865043,
+      "grad_norm": 2.875,
+      "learning_rate": 4.6069140793558495e-05,
+      "loss": 0.806,
+      "step": 10346
+    },
+    {
+      "epoch": 0.183428832254262,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.606763733851264e-05,
+      "loss": 0.8334,
+      "step": 10348
+    },
+    {
+      "epoch": 0.18346428428987357,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.606613362054706e-05,
+      "loss": 0.8018,
+      "step": 10350
+    },
+    {
+      "epoch": 0.18349973632548514,
+      "grad_norm": 3.0,
+      "learning_rate": 4.6064629639680514e-05,
+      "loss": 0.8372,
+      "step": 10352
+    },
+    {
+      "epoch": 0.1835351883610967,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.606312539593178e-05,
+      "loss": 0.8061,
+      "step": 10354
+    },
+    {
+      "epoch": 0.18357064039670828,
+      "grad_norm": 2.75,
+      "learning_rate": 4.606162088931963e-05,
+      "loss": 0.814,
+      "step": 10356
+    },
+    {
+      "epoch": 0.18360609243231985,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.606011611986283e-05,
+      "loss": 0.8303,
+      "step": 10358
+    },
+    {
+      "epoch": 0.1836415444679314,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.605861108758018e-05,
+      "loss": 0.813,
+      "step": 10360
+    },
+    {
+      "epoch": 0.183676996503543,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.6057105792490446e-05,
+      "loss": 0.8015,
+      "step": 10362
+    },
+    {
+      "epoch": 0.18371244853915455,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.605560023461242e-05,
+      "loss": 0.7731,
+      "step": 10364
+    },
+    {
+      "epoch": 0.18374790057476612,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.6054094413964876e-05,
+      "loss": 0.768,
+      "step": 10366
+    },
+    {
+      "epoch": 0.1837833526103777,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.605258833056663e-05,
+      "loss": 0.7992,
+      "step": 10368
+    },
+    {
+      "epoch": 0.18381880464598926,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.605108198443647e-05,
+      "loss": 0.81,
+      "step": 10370
+    },
+    {
+      "epoch": 0.18385425668160083,
+      "grad_norm": 2.625,
+      "learning_rate": 4.604957537559318e-05,
+      "loss": 0.8038,
+      "step": 10372
+    },
+    {
+      "epoch": 0.1838897087172124,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.604806850405559e-05,
+      "loss": 0.8335,
+      "step": 10374
+    },
+    {
+      "epoch": 0.18392516075282397,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.604656136984247e-05,
+      "loss": 0.8094,
+      "step": 10376
+    },
+    {
+      "epoch": 0.18396061278843553,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.6045053972972654e-05,
+      "loss": 0.8018,
+      "step": 10378
+    },
+    {
+      "epoch": 0.18399606482404712,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.604354631346495e-05,
+      "loss": 0.7497,
+      "step": 10380
+    },
+    {
+      "epoch": 0.18403151685965868,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.6042038391338174e-05,
+      "loss": 0.7949,
+      "step": 10382
+    },
+    {
+      "epoch": 0.18406696889527024,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.6040530206611146e-05,
+      "loss": 0.7979,
+      "step": 10384
+    },
+    {
+      "epoch": 0.18410242093088183,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.603902175930267e-05,
+      "loss": 0.8234,
+      "step": 10386
+    },
+    {
+      "epoch": 0.1841378729664934,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.603751304943159e-05,
+      "loss": 0.8141,
+      "step": 10388
+    },
+    {
+      "epoch": 0.18417332500210495,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.603600407701673e-05,
+      "loss": 0.779,
+      "step": 10390
+    },
+    {
+      "epoch": 0.18420877703771654,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.603449484207692e-05,
+      "loss": 0.7836,
+      "step": 10392
+    },
+    {
+      "epoch": 0.1842442290733281,
+      "grad_norm": 2.75,
+      "learning_rate": 4.6032985344631e-05,
+      "loss": 0.8044,
+      "step": 10394
+    },
+    {
+      "epoch": 0.18427968110893966,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.60314755846978e-05,
+      "loss": 0.8166,
+      "step": 10396
+    },
+    {
+      "epoch": 0.18431513314455125,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.602996556229616e-05,
+      "loss": 0.8144,
+      "step": 10398
+    },
+    {
+      "epoch": 0.1843505851801628,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.6028455277444936e-05,
+      "loss": 0.7835,
+      "step": 10400
+    },
+    {
+      "epoch": 0.18438603721577437,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.602694473016297e-05,
+      "loss": 0.8057,
+      "step": 10402
+    },
+    {
+      "epoch": 0.18442148925138596,
+      "grad_norm": 2.875,
+      "learning_rate": 4.6025433920469117e-05,
+      "loss": 0.7827,
+      "step": 10404
+    },
+    {
+      "epoch": 0.18445694128699752,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.6023922848382215e-05,
+      "loss": 0.7865,
+      "step": 10406
+    },
+    {
+      "epoch": 0.18449239332260908,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.6022411513921146e-05,
+      "loss": 0.8172,
+      "step": 10408
+    },
+    {
+      "epoch": 0.18452784535822067,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.602089991710475e-05,
+      "loss": 0.8207,
+      "step": 10410
+    },
+    {
+      "epoch": 0.18456329739383223,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.6019388057951916e-05,
+      "loss": 0.8287,
+      "step": 10412
+    },
+    {
+      "epoch": 0.1845987494294438,
+      "grad_norm": 2.875,
+      "learning_rate": 4.6017875936481494e-05,
+      "loss": 0.801,
+      "step": 10414
+    },
+    {
+      "epoch": 0.18463420146505538,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.601636355271235e-05,
+      "loss": 0.8247,
+      "step": 10416
+    },
+    {
+      "epoch": 0.18466965350066694,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.601485090666337e-05,
+      "loss": 0.8076,
+      "step": 10418
+    },
+    {
+      "epoch": 0.1847051055362785,
+      "grad_norm": 2.5,
+      "learning_rate": 4.601333799835343e-05,
+      "loss": 0.8205,
+      "step": 10420
+    },
+    {
+      "epoch": 0.1847405575718901,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.60118248278014e-05,
+      "loss": 0.7802,
+      "step": 10422
+    },
+    {
+      "epoch": 0.18477600960750165,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.601031139502619e-05,
+      "loss": 0.791,
+      "step": 10424
+    },
+    {
+      "epoch": 0.1848114616431132,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.6008797700046647e-05,
+      "loss": 0.8082,
+      "step": 10426
+    },
+    {
+      "epoch": 0.1848469136787248,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.6007283742881704e-05,
+      "loss": 0.7884,
+      "step": 10428
+    },
+    {
+      "epoch": 0.18488236571433636,
+      "grad_norm": 2.75,
+      "learning_rate": 4.6005769523550226e-05,
+      "loss": 0.7965,
+      "step": 10430
+    },
+    {
+      "epoch": 0.18491781774994792,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.600425504207112e-05,
+      "loss": 0.7819,
+      "step": 10432
+    },
+    {
+      "epoch": 0.1849532697855595,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.600274029846329e-05,
+      "loss": 0.8161,
+      "step": 10434
+    },
+    {
+      "epoch": 0.18498872182117107,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.600122529274563e-05,
+      "loss": 0.7533,
+      "step": 10436
+    },
+    {
+      "epoch": 0.18502417385678263,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.599971002493706e-05,
+      "loss": 0.7857,
+      "step": 10438
+    },
+    {
+      "epoch": 0.18505962589239422,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.599819449505647e-05,
+      "loss": 0.741,
+      "step": 10440
+    },
+    {
+      "epoch": 0.18509507792800578,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.5996678703122794e-05,
+      "loss": 0.7814,
+      "step": 10442
+    },
+    {
+      "epoch": 0.18513052996361734,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.5995162649154944e-05,
+      "loss": 0.8245,
+      "step": 10444
+    },
+    {
+      "epoch": 0.18516598199922893,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.5993646333171837e-05,
+      "loss": 0.8011,
+      "step": 10446
+    },
+    {
+      "epoch": 0.1852014340348405,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.599212975519239e-05,
+      "loss": 0.7667,
+      "step": 10448
+    },
+    {
+      "epoch": 0.18523688607045205,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.5990612915235545e-05,
+      "loss": 0.8283,
+      "step": 10450
+    },
+    {
+      "epoch": 0.18527233810606364,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.598909581332021e-05,
+      "loss": 0.8087,
+      "step": 10452
+    },
+    {
+      "epoch": 0.1853077901416752,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.598757844946534e-05,
+      "loss": 0.8038,
+      "step": 10454
+    },
+    {
+      "epoch": 0.18534324217728676,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.598606082368986e-05,
+      "loss": 0.8039,
+      "step": 10456
+    },
+    {
+      "epoch": 0.18537869421289835,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.5984542936012716e-05,
+      "loss": 0.7779,
+      "step": 10458
+    },
+    {
+      "epoch": 0.1854141462485099,
+      "grad_norm": 3.0,
+      "learning_rate": 4.598302478645284e-05,
+      "loss": 0.8191,
+      "step": 10460
+    },
+    {
+      "epoch": 0.18544959828412147,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.5981506375029194e-05,
+      "loss": 0.7917,
+      "step": 10462
+    },
+    {
+      "epoch": 0.18548505031973306,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.597998770176071e-05,
+      "loss": 0.8076,
+      "step": 10464
+    },
+    {
+      "epoch": 0.18552050235534462,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.597846876666635e-05,
+      "loss": 0.7947,
+      "step": 10466
+    },
+    {
+      "epoch": 0.18555595439095618,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.597694956976508e-05,
+      "loss": 0.8099,
+      "step": 10468
+    },
+    {
+      "epoch": 0.18559140642656777,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.597543011107584e-05,
+      "loss": 0.8352,
+      "step": 10470
+    },
+    {
+      "epoch": 0.18562685846217933,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.59739103906176e-05,
+      "loss": 0.7795,
+      "step": 10472
+    },
+    {
+      "epoch": 0.1856623104977909,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.597239040840933e-05,
+      "loss": 0.7894,
+      "step": 10474
+    },
+    {
+      "epoch": 0.18569776253340248,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.5970870164469995e-05,
+      "loss": 0.8411,
+      "step": 10476
+    },
+    {
+      "epoch": 0.18573321456901404,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.5969349658818575e-05,
+      "loss": 0.7486,
+      "step": 10478
+    },
+    {
+      "epoch": 0.1857686666046256,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.596782889147403e-05,
+      "loss": 0.8125,
+      "step": 10480
+    },
+    {
+      "epoch": 0.1858041186402372,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.5966307862455344e-05,
+      "loss": 0.8195,
+      "step": 10482
+    },
+    {
+      "epoch": 0.18583957067584875,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.596478657178151e-05,
+      "loss": 0.8261,
+      "step": 10484
+    },
+    {
+      "epoch": 0.1858750227114603,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.5963265019471504e-05,
+      "loss": 0.809,
+      "step": 10486
+    },
+    {
+      "epoch": 0.1859104747470719,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.596174320554432e-05,
+      "loss": 0.8152,
+      "step": 10488
+    },
+    {
+      "epoch": 0.18594592678268346,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.5960221130018946e-05,
+      "loss": 0.7875,
+      "step": 10490
+    },
+    {
+      "epoch": 0.18598137881829502,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.5958698792914364e-05,
+      "loss": 0.8312,
+      "step": 10492
+    },
+    {
+      "epoch": 0.1860168308539066,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.595717619424961e-05,
+      "loss": 0.746,
+      "step": 10494
+    },
+    {
+      "epoch": 0.18605228288951817,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.595565333404365e-05,
+      "loss": 0.8165,
+      "step": 10496
+    },
+    {
+      "epoch": 0.18608773492512973,
+      "grad_norm": 2.625,
+      "learning_rate": 4.59541302123155e-05,
+      "loss": 0.7756,
+      "step": 10498
+    },
+    {
+      "epoch": 0.18612318696074132,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.595260682908417e-05,
+      "loss": 0.7439,
+      "step": 10500
+    },
+    {
+      "epoch": 0.18615863899635288,
+      "grad_norm": 2.875,
+      "learning_rate": 4.595108318436867e-05,
+      "loss": 0.8294,
+      "step": 10502
+    },
+    {
+      "epoch": 0.18619409103196444,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.594955927818802e-05,
+      "loss": 0.7616,
+      "step": 10504
+    },
+    {
+      "epoch": 0.186229543067576,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.5948035110561236e-05,
+      "loss": 0.7847,
+      "step": 10506
+    },
+    {
+      "epoch": 0.1862649951031876,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.5946510681507326e-05,
+      "loss": 0.8121,
+      "step": 10508
+    },
+    {
+      "epoch": 0.18630044713879915,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.5944985991045333e-05,
+      "loss": 0.8213,
+      "step": 10510
+    },
+    {
+      "epoch": 0.1863358991744107,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.594346103919428e-05,
+      "loss": 0.8258,
+      "step": 10512
+    },
+    {
+      "epoch": 0.1863713512100223,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.594193582597319e-05,
+      "loss": 0.8081,
+      "step": 10514
+    },
+    {
+      "epoch": 0.18640680324563386,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.594041035140111e-05,
+      "loss": 0.7971,
+      "step": 10516
+    },
+    {
+      "epoch": 0.18644225528124542,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.593888461549706e-05,
+      "loss": 0.7671,
+      "step": 10518
+    },
+    {
+      "epoch": 0.186477707316857,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.59373586182801e-05,
+      "loss": 0.8175,
+      "step": 10520
+    },
+    {
+      "epoch": 0.18651315935246857,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.593583235976926e-05,
+      "loss": 0.7981,
+      "step": 10522
+    },
+    {
+      "epoch": 0.18654861138808013,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.593430583998359e-05,
+      "loss": 0.8102,
+      "step": 10524
+    },
+    {
+      "epoch": 0.18658406342369172,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.5932779058942154e-05,
+      "loss": 0.76,
+      "step": 10526
+    },
+    {
+      "epoch": 0.18661951545930328,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.5931252016663985e-05,
+      "loss": 0.8077,
+      "step": 10528
+    },
+    {
+      "epoch": 0.18665496749491484,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.592972471316815e-05,
+      "loss": 0.8312,
+      "step": 10530
+    },
+    {
+      "epoch": 0.18669041953052642,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.5928197148473726e-05,
+      "loss": 0.7986,
+      "step": 10532
+    },
+    {
+      "epoch": 0.18672587156613799,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.5926669322599746e-05,
+      "loss": 0.8203,
+      "step": 10534
+    },
+    {
+      "epoch": 0.18676132360174955,
+      "grad_norm": 2.875,
+      "learning_rate": 4.5925141235565294e-05,
+      "loss": 0.776,
+      "step": 10536
+    },
+    {
+      "epoch": 0.18679677563736113,
+      "grad_norm": 2.875,
+      "learning_rate": 4.592361288738945e-05,
+      "loss": 0.8089,
+      "step": 10538
+    },
+    {
+      "epoch": 0.1868322276729727,
+      "grad_norm": 2.875,
+      "learning_rate": 4.592208427809125e-05,
+      "loss": 0.7796,
+      "step": 10540
+    },
+    {
+      "epoch": 0.18686767970858426,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.592055540768981e-05,
+      "loss": 0.7896,
+      "step": 10542
+    },
+    {
+      "epoch": 0.18690313174419584,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.59190262762042e-05,
+      "loss": 0.8076,
+      "step": 10544
+    },
+    {
+      "epoch": 0.1869385837798074,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.591749688365349e-05,
+      "loss": 0.8327,
+      "step": 10546
+    },
+    {
+      "epoch": 0.18697403581541897,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.5915967230056786e-05,
+      "loss": 0.766,
+      "step": 10548
+    },
+    {
+      "epoch": 0.18700948785103055,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.591443731543316e-05,
+      "loss": 0.8005,
+      "step": 10550
+    },
+    {
+      "epoch": 0.18704493988664211,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.591290713980172e-05,
+      "loss": 0.8138,
+      "step": 10552
+    },
+    {
+      "epoch": 0.18708039192225367,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.591137670318155e-05,
+      "loss": 0.8359,
+      "step": 10554
+    },
+    {
+      "epoch": 0.18711584395786526,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.590984600559175e-05,
+      "loss": 0.8147,
+      "step": 10556
+    },
+    {
+      "epoch": 0.18715129599347682,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.590831504705143e-05,
+      "loss": 0.8025,
+      "step": 10558
+    },
+    {
+      "epoch": 0.18718674802908838,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.590678382757969e-05,
+      "loss": 0.8017,
+      "step": 10560
+    },
+    {
+      "epoch": 0.18722220006469997,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.590525234719565e-05,
+      "loss": 0.7923,
+      "step": 10562
+    },
+    {
+      "epoch": 0.18725765210031153,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.590372060591841e-05,
+      "loss": 0.7839,
+      "step": 10564
+    },
+    {
+      "epoch": 0.1872931041359231,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.5902188603767094e-05,
+      "loss": 0.814,
+      "step": 10566
+    },
+    {
+      "epoch": 0.18732855617153468,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.590065634076082e-05,
+      "loss": 0.8513,
+      "step": 10568
+    },
+    {
+      "epoch": 0.18736400820714624,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.58991238169187e-05,
+      "loss": 0.7772,
+      "step": 10570
+    },
+    {
+      "epoch": 0.1873994602427578,
+      "grad_norm": 3.0,
+      "learning_rate": 4.589759103225987e-05,
+      "loss": 0.8171,
+      "step": 10572
+    },
+    {
+      "epoch": 0.1874349122783694,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.589605798680346e-05,
+      "loss": 0.8193,
+      "step": 10574
+    },
+    {
+      "epoch": 0.18747036431398095,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.5894524680568596e-05,
+      "loss": 0.8185,
+      "step": 10576
+    },
+    {
+      "epoch": 0.1875058163495925,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.589299111357441e-05,
+      "loss": 0.788,
+      "step": 10578
+    },
+    {
+      "epoch": 0.1875412683852041,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.589145728584006e-05,
+      "loss": 0.7515,
+      "step": 10580
+    },
+    {
+      "epoch": 0.18757672042081566,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.588992319738466e-05,
+      "loss": 0.7798,
+      "step": 10582
+    },
+    {
+      "epoch": 0.18761217245642722,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.588838884822738e-05,
+      "loss": 0.7761,
+      "step": 10584
+    },
+    {
+      "epoch": 0.1876476244920388,
+      "grad_norm": 3.0,
+      "learning_rate": 4.5886854238387364e-05,
+      "loss": 0.8103,
+      "step": 10586
+    },
+    {
+      "epoch": 0.18768307652765037,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.588531936788375e-05,
+      "loss": 0.8216,
+      "step": 10588
+    },
+    {
+      "epoch": 0.18771852856326193,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.588378423673569e-05,
+      "loss": 0.8116,
+      "step": 10590
+    },
+    {
+      "epoch": 0.18775398059887352,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.588224884496237e-05,
+      "loss": 0.8287,
+      "step": 10592
+    },
+    {
+      "epoch": 0.18778943263448508,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.588071319258293e-05,
+      "loss": 0.7659,
+      "step": 10594
+    },
+    {
+      "epoch": 0.18782488467009664,
+      "grad_norm": 2.5,
+      "learning_rate": 4.587917727961652e-05,
+      "loss": 0.8064,
+      "step": 10596
+    },
+    {
+      "epoch": 0.18786033670570823,
+      "grad_norm": 2.75,
+      "learning_rate": 4.587764110608235e-05,
+      "loss": 0.7971,
+      "step": 10598
+    },
+    {
+      "epoch": 0.1878957887413198,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.5876104671999556e-05,
+      "loss": 0.8327,
+      "step": 10600
+    },
+    {
+      "epoch": 0.18793124077693135,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.5874567977387326e-05,
+      "loss": 0.7434,
+      "step": 10602
+    },
+    {
+      "epoch": 0.18796669281254294,
+      "grad_norm": 2.625,
+      "learning_rate": 4.5873031022264834e-05,
+      "loss": 0.8094,
+      "step": 10604
+    },
+    {
+      "epoch": 0.1880021448481545,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.5871493806651265e-05,
+      "loss": 0.7896,
+      "step": 10606
+    },
+    {
+      "epoch": 0.18803759688376606,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.58699563305658e-05,
+      "loss": 0.792,
+      "step": 10608
+    },
+    {
+      "epoch": 0.18807304891937765,
+      "grad_norm": 2.75,
+      "learning_rate": 4.586841859402763e-05,
+      "loss": 0.8111,
+      "step": 10610
+    },
+    {
+      "epoch": 0.1881085009549892,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.586688059705593e-05,
+      "loss": 0.8222,
+      "step": 10612
+    },
+    {
+      "epoch": 0.18814395299060077,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.586534233966992e-05,
+      "loss": 0.8149,
+      "step": 10614
+    },
+    {
+      "epoch": 0.18817940502621236,
+      "grad_norm": 2.625,
+      "learning_rate": 4.5863803821888775e-05,
+      "loss": 0.7729,
+      "step": 10616
+    },
+    {
+      "epoch": 0.18821485706182392,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.586226504373171e-05,
+      "loss": 0.8154,
+      "step": 10618
+    },
+    {
+      "epoch": 0.18825030909743548,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.5860726005217924e-05,
+      "loss": 0.8175,
+      "step": 10620
+    },
+    {
+      "epoch": 0.18828576113304707,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.585918670636662e-05,
+      "loss": 0.8531,
+      "step": 10622
+    },
+    {
+      "epoch": 0.18832121316865863,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.5857647147197e-05,
+      "loss": 0.7576,
+      "step": 10624
+    },
+    {
+      "epoch": 0.1883566652042702,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.5856107327728305e-05,
+      "loss": 0.7785,
+      "step": 10626
+    },
+    {
+      "epoch": 0.18839211723988178,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.5854567247979727e-05,
+      "loss": 0.7932,
+      "step": 10628
+    },
+    {
+      "epoch": 0.18842756927549334,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.5853026907970484e-05,
+      "loss": 0.7907,
+      "step": 10630
+    },
+    {
+      "epoch": 0.1884630213111049,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.585148630771983e-05,
+      "loss": 0.8394,
+      "step": 10632
+    },
+    {
+      "epoch": 0.1884984733467165,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.584994544724695e-05,
+      "loss": 0.8028,
+      "step": 10634
+    },
+    {
+      "epoch": 0.18853392538232805,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.5848404326571104e-05,
+      "loss": 0.8247,
+      "step": 10636
+    },
+    {
+      "epoch": 0.1885693774179396,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.584686294571151e-05,
+      "loss": 0.7953,
+      "step": 10638
+    },
+    {
+      "epoch": 0.1886048294535512,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.584532130468741e-05,
+      "loss": 0.8111,
+      "step": 10640
+    },
+    {
+      "epoch": 0.18864028148916276,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.584377940351804e-05,
+      "loss": 0.7983,
+      "step": 10642
+    },
+    {
+      "epoch": 0.18867573352477432,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.584223724222265e-05,
+      "loss": 0.8223,
+      "step": 10644
+    },
+    {
+      "epoch": 0.1887111855603859,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.5840694820820476e-05,
+      "loss": 0.8173,
+      "step": 10646
+    },
+    {
+      "epoch": 0.18874663759599747,
+      "grad_norm": 2.625,
+      "learning_rate": 4.583915213933077e-05,
+      "loss": 0.7821,
+      "step": 10648
+    },
+    {
+      "epoch": 0.18878208963160903,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.583760919777279e-05,
+      "loss": 0.8053,
+      "step": 10650
+    },
+    {
+      "epoch": 0.18881754166722062,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.583606599616579e-05,
+      "loss": 0.7501,
+      "step": 10652
+    },
+    {
+      "epoch": 0.18885299370283218,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.5834522534529015e-05,
+      "loss": 0.771,
+      "step": 10654
+    },
+    {
+      "epoch": 0.18888844573844374,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.5832978812881744e-05,
+      "loss": 0.8112,
+      "step": 10656
+    },
+    {
+      "epoch": 0.18892389777405533,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.583143483124324e-05,
+      "loss": 0.8547,
+      "step": 10658
+    },
+    {
+      "epoch": 0.1889593498096669,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.582989058963276e-05,
+      "loss": 0.806,
+      "step": 10660
+    },
+    {
+      "epoch": 0.18899480184527845,
+      "grad_norm": 2.875,
+      "learning_rate": 4.5828346088069596e-05,
+      "loss": 0.7513,
+      "step": 10662
+    },
+    {
+      "epoch": 0.18903025388089004,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.5826801326573006e-05,
+      "loss": 0.8386,
+      "step": 10664
+    },
+    {
+      "epoch": 0.1890657059165016,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.582525630516227e-05,
+      "loss": 0.7845,
+      "step": 10666
+    },
+    {
+      "epoch": 0.18910115795211316,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.582371102385667e-05,
+      "loss": 0.8075,
+      "step": 10668
+    },
+    {
+      "epoch": 0.18913660998772475,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.5822165482675505e-05,
+      "loss": 0.8587,
+      "step": 10670
+    },
+    {
+      "epoch": 0.1891720620233363,
+      "grad_norm": 2.75,
+      "learning_rate": 4.5820619681638046e-05,
+      "loss": 0.7742,
+      "step": 10672
+    },
+    {
+      "epoch": 0.18920751405894787,
+      "grad_norm": 2.625,
+      "learning_rate": 4.581907362076359e-05,
+      "loss": 0.839,
+      "step": 10674
+    },
+    {
+      "epoch": 0.18924296609455943,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.581752730007143e-05,
+      "loss": 0.8301,
+      "step": 10676
+    },
+    {
+      "epoch": 0.18927841813017102,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.5815980719580864e-05,
+      "loss": 0.7948,
+      "step": 10678
+    },
+    {
+      "epoch": 0.18931387016578258,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.5814433879311194e-05,
+      "loss": 0.8258,
+      "step": 10680
+    },
+    {
+      "epoch": 0.18934932220139414,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.581288677928173e-05,
+      "loss": 0.8405,
+      "step": 10682
+    },
+    {
+      "epoch": 0.18938477423700573,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.581133941951177e-05,
+      "loss": 0.8275,
+      "step": 10684
+    },
+    {
+      "epoch": 0.1894202262726173,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.580979180002063e-05,
+      "loss": 0.7894,
+      "step": 10686
+    },
+    {
+      "epoch": 0.18945567830822885,
+      "grad_norm": 2.75,
+      "learning_rate": 4.580824392082762e-05,
+      "loss": 0.7929,
+      "step": 10688
+    },
+    {
+      "epoch": 0.18949113034384044,
+      "grad_norm": 2.625,
+      "learning_rate": 4.580669578195206e-05,
+      "loss": 0.8554,
+      "step": 10690
+    },
+    {
+      "epoch": 0.189526582379452,
+      "grad_norm": 2.75,
+      "learning_rate": 4.580514738341328e-05,
+      "loss": 0.7792,
+      "step": 10692
+    },
+    {
+      "epoch": 0.18956203441506356,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.580359872523058e-05,
+      "loss": 0.8322,
+      "step": 10694
+    },
+    {
+      "epoch": 0.18959748645067515,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.580204980742331e-05,
+      "loss": 0.8135,
+      "step": 10696
+    },
+    {
+      "epoch": 0.1896329384862867,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.580050063001079e-05,
+      "loss": 0.8146,
+      "step": 10698
+    },
+    {
+      "epoch": 0.18966839052189827,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.579895119301235e-05,
+      "loss": 0.8429,
+      "step": 10700
+    },
+    {
+      "epoch": 0.18970384255750986,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.579740149644734e-05,
+      "loss": 0.7867,
+      "step": 10702
+    },
+    {
+      "epoch": 0.18973929459312142,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.579585154033509e-05,
+      "loss": 0.8102,
+      "step": 10704
+    },
+    {
+      "epoch": 0.18977474662873298,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.5794301324694934e-05,
+      "loss": 0.8017,
+      "step": 10706
+    },
+    {
+      "epoch": 0.18981019866434456,
+      "grad_norm": 2.625,
+      "learning_rate": 4.579275084954623e-05,
+      "loss": 0.7597,
+      "step": 10708
+    },
+    {
+      "epoch": 0.18984565069995613,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.579120011490834e-05,
+      "loss": 0.8081,
+      "step": 10710
+    },
+    {
+      "epoch": 0.1898811027355677,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.5789649120800587e-05,
+      "loss": 0.7912,
+      "step": 10712
+    },
+    {
+      "epoch": 0.18991655477117927,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.5788097867242355e-05,
+      "loss": 0.7823,
+      "step": 10714
+    },
+    {
+      "epoch": 0.18995200680679084,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.578654635425298e-05,
+      "loss": 0.8361,
+      "step": 10716
+    },
+    {
+      "epoch": 0.1899874588424024,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.578499458185185e-05,
+      "loss": 0.7681,
+      "step": 10718
+    },
+    {
+      "epoch": 0.19002291087801398,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.578344255005831e-05,
+      "loss": 0.8167,
+      "step": 10720
+    },
+    {
+      "epoch": 0.19005836291362554,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.578189025889173e-05,
+      "loss": 0.8504,
+      "step": 10722
+    },
+    {
+      "epoch": 0.1900938149492371,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.578033770837149e-05,
+      "loss": 0.7749,
+      "step": 10724
+    },
+    {
+      "epoch": 0.1901292669848487,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.577878489851697e-05,
+      "loss": 0.8276,
+      "step": 10726
+    },
+    {
+      "epoch": 0.19016471902046025,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.577723182934754e-05,
+      "loss": 0.8341,
+      "step": 10728
+    },
+    {
+      "epoch": 0.19020017105607182,
+      "grad_norm": 2.75,
+      "learning_rate": 4.577567850088258e-05,
+      "loss": 0.7664,
+      "step": 10730
+    },
+    {
+      "epoch": 0.1902356230916834,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.577412491314149e-05,
+      "loss": 0.8106,
+      "step": 10732
+    },
+    {
+      "epoch": 0.19027107512729496,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.577257106614364e-05,
+      "loss": 0.8244,
+      "step": 10734
+    },
+    {
+      "epoch": 0.19030652716290652,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.577101695990843e-05,
+      "loss": 0.8214,
+      "step": 10736
+    },
+    {
+      "epoch": 0.1903419791985181,
+      "grad_norm": 2.875,
+      "learning_rate": 4.5769462594455256e-05,
+      "loss": 0.8786,
+      "step": 10738
+    },
+    {
+      "epoch": 0.19037743123412967,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.5767907969803514e-05,
+      "loss": 0.8052,
+      "step": 10740
+    },
+    {
+      "epoch": 0.19041288326974123,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.5766353085972605e-05,
+      "loss": 0.8261,
+      "step": 10742
+    },
+    {
+      "epoch": 0.19044833530535282,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.5764797942981944e-05,
+      "loss": 0.7922,
+      "step": 10744
+    },
+    {
+      "epoch": 0.19048378734096438,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.576324254085092e-05,
+      "loss": 0.7848,
+      "step": 10746
+    },
+    {
+      "epoch": 0.19051923937657594,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.576168687959895e-05,
+      "loss": 0.8277,
+      "step": 10748
+    },
+    {
+      "epoch": 0.19055469141218753,
+      "grad_norm": 2.625,
+      "learning_rate": 4.5760130959245464e-05,
+      "loss": 0.8246,
+      "step": 10750
+    },
+    {
+      "epoch": 0.1905901434477991,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.575857477980986e-05,
+      "loss": 0.7686,
+      "step": 10752
+    },
+    {
+      "epoch": 0.19062559548341065,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.5757018341311565e-05,
+      "loss": 0.7949,
+      "step": 10754
+    },
+    {
+      "epoch": 0.19066104751902224,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.575546164377e-05,
+      "loss": 0.8289,
+      "step": 10756
+    },
+    {
+      "epoch": 0.1906964995546338,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.575390468720461e-05,
+      "loss": 0.8116,
+      "step": 10758
+    },
+    {
+      "epoch": 0.19073195159024536,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.5752347471634804e-05,
+      "loss": 0.8185,
+      "step": 10760
+    },
+    {
+      "epoch": 0.19076740362585695,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.5750789997080035e-05,
+      "loss": 0.7798,
+      "step": 10762
+    },
+    {
+      "epoch": 0.1908028556614685,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.5749232263559716e-05,
+      "loss": 0.8127,
+      "step": 10764
+    },
+    {
+      "epoch": 0.19083830769708007,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.5747674271093306e-05,
+      "loss": 0.8089,
+      "step": 10766
+    },
+    {
+      "epoch": 0.19087375973269166,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.5746116019700234e-05,
+      "loss": 0.7595,
+      "step": 10768
+    },
+    {
+      "epoch": 0.19090921176830322,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.574455750939997e-05,
+      "loss": 0.7865,
+      "step": 10770
+    },
+    {
+      "epoch": 0.19094466380391478,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.574299874021194e-05,
+      "loss": 0.8002,
+      "step": 10772
+    },
+    {
+      "epoch": 0.19098011583952637,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.574143971215561e-05,
+      "loss": 0.7734,
+      "step": 10774
+    },
+    {
+      "epoch": 0.19101556787513793,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.573988042525042e-05,
+      "loss": 0.794,
+      "step": 10776
+    },
+    {
+      "epoch": 0.1910510199107495,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.573832087951586e-05,
+      "loss": 0.7835,
+      "step": 10778
+    },
+    {
+      "epoch": 0.19108647194636108,
+      "grad_norm": 2.75,
+      "learning_rate": 4.5736761074971366e-05,
+      "loss": 0.814,
+      "step": 10780
+    },
+    {
+      "epoch": 0.19112192398197264,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.573520101163641e-05,
+      "loss": 0.8026,
+      "step": 10782
+    },
+    {
+      "epoch": 0.1911573760175842,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.5733640689530465e-05,
+      "loss": 0.7686,
+      "step": 10784
+    },
+    {
+      "epoch": 0.1911928280531958,
+      "grad_norm": 2.75,
+      "learning_rate": 4.5732080108673007e-05,
+      "loss": 0.8028,
+      "step": 10786
+    },
+    {
+      "epoch": 0.19122828008880735,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.573051926908351e-05,
+      "loss": 0.8185,
+      "step": 10788
+    },
+    {
+      "epoch": 0.1912637321244189,
+      "grad_norm": 3.25,
+      "learning_rate": 4.5728958170781446e-05,
+      "loss": 0.8286,
+      "step": 10790
+    },
+    {
+      "epoch": 0.1912991841600305,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.57273968137863e-05,
+      "loss": 0.8065,
+      "step": 10792
+    },
+    {
+      "epoch": 0.19133463619564206,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.572583519811756e-05,
+      "loss": 0.8441,
+      "step": 10794
+    },
+    {
+      "epoch": 0.19137008823125362,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.572427332379472e-05,
+      "loss": 0.7712,
+      "step": 10796
+    },
+    {
+      "epoch": 0.1914055402668652,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.572271119083726e-05,
+      "loss": 0.7992,
+      "step": 10798
+    },
+    {
+      "epoch": 0.19144099230247677,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.5721148799264676e-05,
+      "loss": 0.7992,
+      "step": 10800
+    },
+    {
+      "epoch": 0.19147644433808833,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.571958614909648e-05,
+      "loss": 0.7961,
+      "step": 10802
+    },
+    {
+      "epoch": 0.19151189637369992,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.571802324035216e-05,
+      "loss": 0.7667,
+      "step": 10804
+    },
+    {
+      "epoch": 0.19154734840931148,
+      "grad_norm": 2.625,
+      "learning_rate": 4.5716460073051224e-05,
+      "loss": 0.788,
+      "step": 10806
+    },
+    {
+      "epoch": 0.19158280044492304,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.571489664721318e-05,
+      "loss": 0.7764,
+      "step": 10808
+    },
+    {
+      "epoch": 0.19161825248053463,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.571333296285755e-05,
+      "loss": 0.7794,
+      "step": 10810
+    },
+    {
+      "epoch": 0.1916537045161462,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.571176902000383e-05,
+      "loss": 0.7814,
+      "step": 10812
+    },
+    {
+      "epoch": 0.19168915655175775,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.5710204818671546e-05,
+      "loss": 0.825,
+      "step": 10814
+    },
+    {
+      "epoch": 0.19172460858736934,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.570864035888022e-05,
+      "loss": 0.8684,
+      "step": 10816
+    },
+    {
+      "epoch": 0.1917600606229809,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.570707564064938e-05,
+      "loss": 0.7767,
+      "step": 10818
+    },
+    {
+      "epoch": 0.19179551265859246,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.570551066399854e-05,
+      "loss": 0.8271,
+      "step": 10820
+    },
+    {
+      "epoch": 0.19183096469420405,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.570394542894725e-05,
+      "loss": 0.7728,
+      "step": 10822
+    },
+    {
+      "epoch": 0.1918664167298156,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.5702379935515026e-05,
+      "loss": 0.7949,
+      "step": 10824
+    },
+    {
+      "epoch": 0.19190186876542717,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.570081418372142e-05,
+      "loss": 0.8263,
+      "step": 10826
+    },
+    {
+      "epoch": 0.19193732080103876,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.569924817358596e-05,
+      "loss": 0.7999,
+      "step": 10828
+    },
+    {
+      "epoch": 0.19197277283665032,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.5697681905128195e-05,
+      "loss": 0.794,
+      "step": 10830
+    },
+    {
+      "epoch": 0.19200822487226188,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.569611537836767e-05,
+      "loss": 0.7843,
+      "step": 10832
+    },
+    {
+      "epoch": 0.19204367690787347,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.569454859332394e-05,
+      "loss": 0.806,
+      "step": 10834
+    },
+    {
+      "epoch": 0.19207912894348503,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.569298155001655e-05,
+      "loss": 0.8376,
+      "step": 10836
+    },
+    {
+      "epoch": 0.1921145809790966,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.569141424846506e-05,
+      "loss": 0.8096,
+      "step": 10838
+    },
+    {
+      "epoch": 0.19215003301470818,
+      "grad_norm": 2.625,
+      "learning_rate": 4.568984668868903e-05,
+      "loss": 0.8156,
+      "step": 10840
+    },
+    {
+      "epoch": 0.19218548505031974,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.568827887070802e-05,
+      "loss": 0.7778,
+      "step": 10842
+    },
+    {
+      "epoch": 0.1922209370859313,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.5686710794541595e-05,
+      "loss": 0.808,
+      "step": 10844
+    },
+    {
+      "epoch": 0.19225638912154286,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.568514246020934e-05,
+      "loss": 0.7928,
+      "step": 10846
+    },
+    {
+      "epoch": 0.19229184115715445,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.568357386773081e-05,
+      "loss": 0.7913,
+      "step": 10848
+    },
+    {
+      "epoch": 0.192327293192766,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.5682005017125584e-05,
+      "loss": 0.7929,
+      "step": 10850
+    },
+    {
+      "epoch": 0.19236274522837757,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.568043590841325e-05,
+      "loss": 0.809,
+      "step": 10852
+    },
+    {
+      "epoch": 0.19239819726398916,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.567886654161338e-05,
+      "loss": 0.7859,
+      "step": 10854
+    },
+    {
+      "epoch": 0.19243364929960072,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.567729691674556e-05,
+      "loss": 0.7684,
+      "step": 10856
+    },
+    {
+      "epoch": 0.19246910133521228,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.5675727033829386e-05,
+      "loss": 0.8059,
+      "step": 10858
+    },
+    {
+      "epoch": 0.19250455337082387,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.567415689288444e-05,
+      "loss": 0.8213,
+      "step": 10860
+    },
+    {
+      "epoch": 0.19254000540643543,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.5672586493930325e-05,
+      "loss": 0.7972,
+      "step": 10862
+    },
+    {
+      "epoch": 0.192575457442047,
+      "grad_norm": 2.625,
+      "learning_rate": 4.567101583698663e-05,
+      "loss": 0.8258,
+      "step": 10864
+    },
+    {
+      "epoch": 0.19261090947765858,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.5669444922072965e-05,
+      "loss": 0.8301,
+      "step": 10866
+    },
+    {
+      "epoch": 0.19264636151327014,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.5667873749208946e-05,
+      "loss": 0.8251,
+      "step": 10868
+    },
+    {
+      "epoch": 0.1926818135488817,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.566630231841416e-05,
+      "loss": 0.8066,
+      "step": 10870
+    },
+    {
+      "epoch": 0.19271726558449329,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.566473062970821e-05,
+      "loss": 0.7789,
+      "step": 10872
+    },
+    {
+      "epoch": 0.19275271762010485,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.566315868311074e-05,
+      "loss": 0.8069,
+      "step": 10874
+    },
+    {
+      "epoch": 0.1927881696557164,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.5661586478641356e-05,
+      "loss": 0.7928,
+      "step": 10876
+    },
+    {
+      "epoch": 0.192823621691328,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.5660014016319674e-05,
+      "loss": 0.7653,
+      "step": 10878
+    },
+    {
+      "epoch": 0.19285907372693956,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.5658441296165316e-05,
+      "loss": 0.8492,
+      "step": 10880
+    },
+    {
+      "epoch": 0.19289452576255112,
+      "grad_norm": 2.75,
+      "learning_rate": 4.5656868318197914e-05,
+      "loss": 0.7798,
+      "step": 10882
+    },
+    {
+      "epoch": 0.1929299777981627,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.56552950824371e-05,
+      "loss": 0.7742,
+      "step": 10884
+    },
+    {
+      "epoch": 0.19296542983377427,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.5653721588902506e-05,
+      "loss": 0.779,
+      "step": 10886
+    },
+    {
+      "epoch": 0.19300088186938583,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.565214783761377e-05,
+      "loss": 0.8233,
+      "step": 10888
+    },
+    {
+      "epoch": 0.19303633390499741,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.5650573828590525e-05,
+      "loss": 0.8177,
+      "step": 10890
+    },
+    {
+      "epoch": 0.19307178594060898,
+      "grad_norm": 2.625,
+      "learning_rate": 4.5648999561852424e-05,
+      "loss": 0.8479,
+      "step": 10892
+    },
+    {
+      "epoch": 0.19310723797622054,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.564742503741911e-05,
+      "loss": 0.8367,
+      "step": 10894
+    },
+    {
+      "epoch": 0.19314269001183212,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.564585025531023e-05,
+      "loss": 0.7768,
+      "step": 10896
+    },
+    {
+      "epoch": 0.19317814204744369,
+      "grad_norm": 2.75,
+      "learning_rate": 4.564427521554544e-05,
+      "loss": 0.8172,
+      "step": 10898
+    },
+    {
+      "epoch": 0.19321359408305525,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.564269991814439e-05,
+      "loss": 0.8101,
+      "step": 10900
+    },
+    {
+      "epoch": 0.19324904611866683,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.564112436312675e-05,
+      "loss": 0.8061,
+      "step": 10902
+    },
+    {
+      "epoch": 0.1932844981542784,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.563954855051218e-05,
+      "loss": 0.8271,
+      "step": 10904
+    },
+    {
+      "epoch": 0.19331995018988996,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.563797248032034e-05,
+      "loss": 0.7935,
+      "step": 10906
+    },
+    {
+      "epoch": 0.19335540222550154,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.5636396152570906e-05,
+      "loss": 0.8145,
+      "step": 10908
+    },
+    {
+      "epoch": 0.1933908542611131,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.5634819567283536e-05,
+      "loss": 0.8593,
+      "step": 10910
+    },
+    {
+      "epoch": 0.19342630629672466,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.563324272447792e-05,
+      "loss": 0.7838,
+      "step": 10912
+    },
+    {
+      "epoch": 0.19346175833233625,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.563166562417374e-05,
+      "loss": 0.8105,
+      "step": 10914
+    },
+    {
+      "epoch": 0.19349721036794781,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.563008826639066e-05,
+      "loss": 0.8196,
+      "step": 10916
+    },
+    {
+      "epoch": 0.19353266240355937,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.5628510651148385e-05,
+      "loss": 0.8281,
+      "step": 10918
+    },
+    {
+      "epoch": 0.19356811443917096,
+      "grad_norm": 2.625,
+      "learning_rate": 4.562693277846658e-05,
+      "loss": 0.7671,
+      "step": 10920
+    },
+    {
+      "epoch": 0.19360356647478252,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.562535464836496e-05,
+      "loss": 0.8377,
+      "step": 10922
+    },
+    {
+      "epoch": 0.19363901851039408,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.562377626086321e-05,
+      "loss": 0.7881,
+      "step": 10924
+    },
+    {
+      "epoch": 0.19367447054600567,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.5622197615981025e-05,
+      "loss": 0.815,
+      "step": 10926
+    },
+    {
+      "epoch": 0.19370992258161723,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.5620618713738114e-05,
+      "loss": 0.7683,
+      "step": 10928
+    },
+    {
+      "epoch": 0.1937453746172288,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.561903955415417e-05,
+      "loss": 0.7652,
+      "step": 10930
+    },
+    {
+      "epoch": 0.19378082665284038,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.5617460137248915e-05,
+      "loss": 0.8001,
+      "step": 10932
+    },
+    {
+      "epoch": 0.19381627868845194,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.5615880463042036e-05,
+      "loss": 0.8007,
+      "step": 10934
+    },
+    {
+      "epoch": 0.1938517307240635,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.561430053155328e-05,
+      "loss": 0.7932,
+      "step": 10936
+    },
+    {
+      "epoch": 0.1938871827596751,
+      "grad_norm": 2.625,
+      "learning_rate": 4.561272034280234e-05,
+      "loss": 0.7869,
+      "step": 10938
+    },
+    {
+      "epoch": 0.19392263479528665,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.561113989680894e-05,
+      "loss": 0.8114,
+      "step": 10940
+    },
+    {
+      "epoch": 0.1939580868308982,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.560955919359281e-05,
+      "loss": 0.7966,
+      "step": 10942
+    },
+    {
+      "epoch": 0.1939935388665098,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.560797823317368e-05,
+      "loss": 0.8255,
+      "step": 10944
+    },
+    {
+      "epoch": 0.19402899090212136,
+      "grad_norm": 2.875,
+      "learning_rate": 4.560639701557127e-05,
+      "loss": 0.8436,
+      "step": 10946
+    },
+    {
+      "epoch": 0.19406444293773292,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.560481554080531e-05,
+      "loss": 0.7665,
+      "step": 10948
+    },
+    {
+      "epoch": 0.1940998949733445,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.5603233808895554e-05,
+      "loss": 0.7843,
+      "step": 10950
+    },
+    {
+      "epoch": 0.19413534700895607,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.560165181986172e-05,
+      "loss": 0.8479,
+      "step": 10952
+    },
+    {
+      "epoch": 0.19417079904456763,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.5600069573723577e-05,
+      "loss": 0.7911,
+      "step": 10954
+    },
+    {
+      "epoch": 0.19420625108017922,
+      "grad_norm": 2.375,
+      "learning_rate": 4.559848707050085e-05,
+      "loss": 0.7645,
+      "step": 10956
+    },
+    {
+      "epoch": 0.19424170311579078,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.559690431021329e-05,
+      "loss": 0.8018,
+      "step": 10958
+    },
+    {
+      "epoch": 0.19427715515140234,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.559532129288066e-05,
+      "loss": 0.8233,
+      "step": 10960
+    },
+    {
+      "epoch": 0.19431260718701393,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.559373801852271e-05,
+      "loss": 0.809,
+      "step": 10962
+    },
+    {
+      "epoch": 0.1943480592226255,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.5592154487159197e-05,
+      "loss": 0.7707,
+      "step": 10964
+    },
+    {
+      "epoch": 0.19438351125823705,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.559057069880988e-05,
+      "loss": 0.7766,
+      "step": 10966
+    },
+    {
+      "epoch": 0.19441896329384864,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.558898665349453e-05,
+      "loss": 0.8001,
+      "step": 10968
+    },
+    {
+      "epoch": 0.1944544153294602,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.558740235123292e-05,
+      "loss": 0.7537,
+      "step": 10970
+    },
+    {
+      "epoch": 0.19448986736507176,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.5585817792044815e-05,
+      "loss": 0.8424,
+      "step": 10972
+    },
+    {
+      "epoch": 0.19452531940068335,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.558423297595e-05,
+      "loss": 0.8104,
+      "step": 10974
+    },
+    {
+      "epoch": 0.1945607714362949,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.558264790296823e-05,
+      "loss": 0.8017,
+      "step": 10976
+    },
+    {
+      "epoch": 0.19459622347190647,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.558106257311932e-05,
+      "loss": 0.8555,
+      "step": 10978
+    },
+    {
+      "epoch": 0.19463167550751806,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.557947698642302e-05,
+      "loss": 0.8092,
+      "step": 10980
+    },
+    {
+      "epoch": 0.19466712754312962,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.557789114289913e-05,
+      "loss": 0.8264,
+      "step": 10982
+    },
+    {
+      "epoch": 0.19470257957874118,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.557630504256746e-05,
+      "loss": 0.7817,
+      "step": 10984
+    },
+    {
+      "epoch": 0.19473803161435277,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.5574718685447784e-05,
+      "loss": 0.7646,
+      "step": 10986
+    },
+    {
+      "epoch": 0.19477348364996433,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.55731320715599e-05,
+      "loss": 0.7924,
+      "step": 10988
+    },
+    {
+      "epoch": 0.1948089356855759,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.557154520092361e-05,
+      "loss": 0.7924,
+      "step": 10990
+    },
+    {
+      "epoch": 0.19484438772118748,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.5569958073558724e-05,
+      "loss": 0.8446,
+      "step": 10992
+    },
+    {
+      "epoch": 0.19487983975679904,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.556837068948505e-05,
+      "loss": 0.83,
+      "step": 10994
+    },
+    {
+      "epoch": 0.1949152917924106,
+      "grad_norm": 3.0,
+      "learning_rate": 4.556678304872239e-05,
+      "loss": 0.7711,
+      "step": 10996
+    },
+    {
+      "epoch": 0.1949507438280222,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.556519515129056e-05,
+      "loss": 0.8246,
+      "step": 10998
+    },
+    {
+      "epoch": 0.19498619586363375,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.556360699720938e-05,
+      "loss": 0.8017,
+      "step": 11000
+    },
+    {
+      "epoch": 0.1950216478992453,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.556201858649867e-05,
+      "loss": 0.838,
+      "step": 11002
+    },
+    {
+      "epoch": 0.1950570999348569,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.556042991917825e-05,
+      "loss": 0.7748,
+      "step": 11004
+    },
+    {
+      "epoch": 0.19509255197046846,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.555884099526794e-05,
+      "loss": 0.8404,
+      "step": 11006
+    },
+    {
+      "epoch": 0.19512800400608002,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.555725181478758e-05,
+      "loss": 0.8018,
+      "step": 11008
+    },
+    {
+      "epoch": 0.1951634560416916,
+      "grad_norm": 2.75,
+      "learning_rate": 4.5555662377757e-05,
+      "loss": 0.8392,
+      "step": 11010
+    },
+    {
+      "epoch": 0.19519890807730317,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.5554072684196035e-05,
+      "loss": 0.8423,
+      "step": 11012
+    },
+    {
+      "epoch": 0.19523436011291473,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.555248273412453e-05,
+      "loss": 0.8131,
+      "step": 11014
+    },
+    {
+      "epoch": 0.1952698121485263,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.555089252756232e-05,
+      "loss": 0.7641,
+      "step": 11016
+    },
+    {
+      "epoch": 0.19530526418413788,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.554930206452924e-05,
+      "loss": 0.7994,
+      "step": 11018
+    },
+    {
+      "epoch": 0.19534071621974944,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.554771134504516e-05,
+      "loss": 0.7984,
+      "step": 11020
+    },
+    {
+      "epoch": 0.195376168255361,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.554612036912992e-05,
+      "loss": 0.7979,
+      "step": 11022
+    },
+    {
+      "epoch": 0.1954116202909726,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.554452913680338e-05,
+      "loss": 0.7939,
+      "step": 11024
+    },
+    {
+      "epoch": 0.19544707232658415,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.5542937648085394e-05,
+      "loss": 0.8338,
+      "step": 11026
+    },
+    {
+      "epoch": 0.1954825243621957,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.5541345902995825e-05,
+      "loss": 0.8299,
+      "step": 11028
+    },
+    {
+      "epoch": 0.1955179763978073,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.553975390155454e-05,
+      "loss": 0.8436,
+      "step": 11030
+    },
+    {
+      "epoch": 0.19555342843341886,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.55381616437814e-05,
+      "loss": 0.7852,
+      "step": 11032
+    },
+    {
+      "epoch": 0.19558888046903042,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.553656912969628e-05,
+      "loss": 0.8327,
+      "step": 11034
+    },
+    {
+      "epoch": 0.195624332504642,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.553497635931905e-05,
+      "loss": 0.7849,
+      "step": 11036
+    },
+    {
+      "epoch": 0.19565978454025357,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.55333833326696e-05,
+      "loss": 0.801,
+      "step": 11038
+    },
+    {
+      "epoch": 0.19569523657586513,
+      "grad_norm": 3.0,
+      "learning_rate": 4.55317900497678e-05,
+      "loss": 0.8602,
+      "step": 11040
+    },
+    {
+      "epoch": 0.19573068861147672,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.553019651063354e-05,
+      "loss": 0.7818,
+      "step": 11042
+    },
+    {
+      "epoch": 0.19576614064708828,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.55286027152867e-05,
+      "loss": 0.7704,
+      "step": 11044
+    },
+    {
+      "epoch": 0.19580159268269984,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.5527008663747176e-05,
+      "loss": 0.7926,
+      "step": 11046
+    },
+    {
+      "epoch": 0.19583704471831143,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.552541435603486e-05,
+      "loss": 0.8085,
+      "step": 11048
+    },
+    {
+      "epoch": 0.195872496753923,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.5523819792169646e-05,
+      "loss": 0.7907,
+      "step": 11050
+    },
+    {
+      "epoch": 0.19590794878953455,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.5522224972171435e-05,
+      "loss": 0.8319,
+      "step": 11052
+    },
+    {
+      "epoch": 0.19594340082514614,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.5520629896060134e-05,
+      "loss": 0.8224,
+      "step": 11054
+    },
+    {
+      "epoch": 0.1959788528607577,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.5519034563855646e-05,
+      "loss": 0.7893,
+      "step": 11056
+    },
+    {
+      "epoch": 0.19601430489636926,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.551743897557788e-05,
+      "loss": 0.8568,
+      "step": 11058
+    },
+    {
+      "epoch": 0.19604975693198085,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.551584313124675e-05,
+      "loss": 0.7999,
+      "step": 11060
+    },
+    {
+      "epoch": 0.1960852089675924,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.5514247030882165e-05,
+      "loss": 0.818,
+      "step": 11062
+    },
+    {
+      "epoch": 0.19612066100320397,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.551265067450405e-05,
+      "loss": 0.7775,
+      "step": 11064
+    },
+    {
+      "epoch": 0.19615611303881555,
+      "grad_norm": 2.875,
+      "learning_rate": 4.551105406213233e-05,
+      "loss": 0.8353,
+      "step": 11066
+    },
+    {
+      "epoch": 0.19619156507442712,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.550945719378693e-05,
+      "loss": 0.8182,
+      "step": 11068
+    },
+    {
+      "epoch": 0.19622701711003868,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.550786006948777e-05,
+      "loss": 0.7792,
+      "step": 11070
+    },
+    {
+      "epoch": 0.19626246914565026,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.5506262689254796e-05,
+      "loss": 0.819,
+      "step": 11072
+    },
+    {
+      "epoch": 0.19629792118126183,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.550466505310793e-05,
+      "loss": 0.8382,
+      "step": 11074
+    },
+    {
+      "epoch": 0.19633337321687339,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.550306716106712e-05,
+      "loss": 0.8173,
+      "step": 11076
+    },
+    {
+      "epoch": 0.19636882525248497,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.55014690131523e-05,
+      "loss": 0.8505,
+      "step": 11078
+    },
+    {
+      "epoch": 0.19640427728809653,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.549987060938341e-05,
+      "loss": 0.8275,
+      "step": 11080
+    },
+    {
+      "epoch": 0.1964397293237081,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.5498271949780414e-05,
+      "loss": 0.765,
+      "step": 11082
+    },
+    {
+      "epoch": 0.19647518135931968,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.5496673034363246e-05,
+      "loss": 0.7818,
+      "step": 11084
+    },
+    {
+      "epoch": 0.19651063339493124,
+      "grad_norm": 2.625,
+      "learning_rate": 4.549507386315187e-05,
+      "loss": 0.7932,
+      "step": 11086
+    },
+    {
+      "epoch": 0.1965460854305428,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.5493474436166236e-05,
+      "loss": 0.817,
+      "step": 11088
+    },
+    {
+      "epoch": 0.1965815374661544,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.549187475342632e-05,
+      "loss": 0.805,
+      "step": 11090
+    },
+    {
+      "epoch": 0.19661698950176595,
+      "grad_norm": 2.5,
+      "learning_rate": 4.549027481495207e-05,
+      "loss": 0.7974,
+      "step": 11092
+    },
+    {
+      "epoch": 0.19665244153737751,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.548867462076346e-05,
+      "loss": 0.8194,
+      "step": 11094
+    },
+    {
+      "epoch": 0.1966878935729891,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.548707417088046e-05,
+      "loss": 0.8188,
+      "step": 11096
+    },
+    {
+      "epoch": 0.19672334560860066,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.5485473465323035e-05,
+      "loss": 0.8054,
+      "step": 11098
+    },
+    {
+      "epoch": 0.19675879764421222,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.548387250411117e-05,
+      "loss": 0.8056,
+      "step": 11100
+    },
+    {
+      "epoch": 0.1967942496798238,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.5482271287264845e-05,
+      "loss": 0.7905,
+      "step": 11102
+    },
+    {
+      "epoch": 0.19682970171543537,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.5480669814804036e-05,
+      "loss": 0.7922,
+      "step": 11104
+    },
+    {
+      "epoch": 0.19686515375104693,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.5479068086748746e-05,
+      "loss": 0.8214,
+      "step": 11106
+    },
+    {
+      "epoch": 0.19690060578665852,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.5477466103118936e-05,
+      "loss": 0.7875,
+      "step": 11108
+    },
+    {
+      "epoch": 0.19693605782227008,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.547586386393462e-05,
+      "loss": 0.816,
+      "step": 11110
+    },
+    {
+      "epoch": 0.19697150985788164,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.547426136921579e-05,
+      "loss": 0.8257,
+      "step": 11112
+    },
+    {
+      "epoch": 0.19700696189349323,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.5472658618982446e-05,
+      "loss": 0.7914,
+      "step": 11114
+    },
+    {
+      "epoch": 0.1970424139291048,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.5471055613254574e-05,
+      "loss": 0.8,
+      "step": 11116
+    },
+    {
+      "epoch": 0.19707786596471635,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.54694523520522e-05,
+      "loss": 0.7953,
+      "step": 11118
+    },
+    {
+      "epoch": 0.19711331800032794,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.546784883539533e-05,
+      "loss": 0.7683,
+      "step": 11120
+    },
+    {
+      "epoch": 0.1971487700359395,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.546624506330396e-05,
+      "loss": 0.7772,
+      "step": 11122
+    },
+    {
+      "epoch": 0.19718422207155106,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.546464103579812e-05,
+      "loss": 0.7998,
+      "step": 11124
+    },
+    {
+      "epoch": 0.19721967410716265,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.546303675289782e-05,
+      "loss": 0.8329,
+      "step": 11126
+    },
+    {
+      "epoch": 0.1972551261427742,
+      "grad_norm": 2.875,
+      "learning_rate": 4.5461432214623084e-05,
+      "loss": 0.7998,
+      "step": 11128
+    },
+    {
+      "epoch": 0.19729057817838577,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.545982742099394e-05,
+      "loss": 0.8297,
+      "step": 11130
+    },
+    {
+      "epoch": 0.19732603021399736,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.54582223720304e-05,
+      "loss": 0.8171,
+      "step": 11132
+    },
+    {
+      "epoch": 0.19736148224960892,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.545661706775251e-05,
+      "loss": 0.7716,
+      "step": 11134
+    },
+    {
+      "epoch": 0.19739693428522048,
+      "grad_norm": 2.625,
+      "learning_rate": 4.54550115081803e-05,
+      "loss": 0.7913,
+      "step": 11136
+    },
+    {
+      "epoch": 0.19743238632083207,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.545340569333382e-05,
+      "loss": 0.7837,
+      "step": 11138
+    },
+    {
+      "epoch": 0.19746783835644363,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.545179962323308e-05,
+      "loss": 0.8132,
+      "step": 11140
+    },
+    {
+      "epoch": 0.1975032903920552,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.545019329789815e-05,
+      "loss": 0.8121,
+      "step": 11142
+    },
+    {
+      "epoch": 0.19753874242766678,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.5448586717349065e-05,
+      "loss": 0.8451,
+      "step": 11144
+    },
+    {
+      "epoch": 0.19757419446327834,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.5446979881605874e-05,
+      "loss": 0.7968,
+      "step": 11146
+    },
+    {
+      "epoch": 0.1976096464988899,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.5445372790688634e-05,
+      "loss": 0.7621,
+      "step": 11148
+    },
+    {
+      "epoch": 0.1976450985345015,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.5443765444617404e-05,
+      "loss": 0.805,
+      "step": 11150
+    },
+    {
+      "epoch": 0.19768055057011305,
+      "grad_norm": 2.75,
+      "learning_rate": 4.544215784341224e-05,
+      "loss": 0.8009,
+      "step": 11152
+    },
+    {
+      "epoch": 0.1977160026057246,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.544054998709319e-05,
+      "loss": 0.8013,
+      "step": 11154
+    },
+    {
+      "epoch": 0.1977514546413362,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.543894187568035e-05,
+      "loss": 0.804,
+      "step": 11156
+    },
+    {
+      "epoch": 0.19778690667694776,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.5437333509193765e-05,
+      "loss": 0.8002,
+      "step": 11158
+    },
+    {
+      "epoch": 0.19782235871255932,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.543572488765351e-05,
+      "loss": 0.8441,
+      "step": 11160
+    },
+    {
+      "epoch": 0.1978578107481709,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.5434116011079675e-05,
+      "loss": 0.8051,
+      "step": 11162
+    },
+    {
+      "epoch": 0.19789326278378247,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.543250687949232e-05,
+      "loss": 0.7755,
+      "step": 11164
+    },
+    {
+      "epoch": 0.19792871481939403,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.543089749291154e-05,
+      "loss": 0.7464,
+      "step": 11166
+    },
+    {
+      "epoch": 0.19796416685500562,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.5429287851357416e-05,
+      "loss": 0.8078,
+      "step": 11168
+    },
+    {
+      "epoch": 0.19799961889061718,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.542767795485003e-05,
+      "loss": 0.7877,
+      "step": 11170
+    },
+    {
+      "epoch": 0.19803507092622874,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.542606780340948e-05,
+      "loss": 0.8076,
+      "step": 11172
+    },
+    {
+      "epoch": 0.19807052296184033,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.5424457397055856e-05,
+      "loss": 0.7988,
+      "step": 11174
+    },
+    {
+      "epoch": 0.1981059749974519,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.542284673580927e-05,
+      "loss": 0.8249,
+      "step": 11176
+    },
+    {
+      "epoch": 0.19814142703306345,
+      "grad_norm": 2.5,
+      "learning_rate": 4.5421235819689796e-05,
+      "loss": 0.7981,
+      "step": 11178
+    },
+    {
+      "epoch": 0.19817687906867504,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.541962464871756e-05,
+      "loss": 0.7837,
+      "step": 11180
+    },
+    {
+      "epoch": 0.1982123311042866,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.541801322291266e-05,
+      "loss": 0.756,
+      "step": 11182
+    },
+    {
+      "epoch": 0.19824778313989816,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.54164015422952e-05,
+      "loss": 0.8059,
+      "step": 11184
+    },
+    {
+      "epoch": 0.19828323517550972,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.541478960688531e-05,
+      "loss": 0.8035,
+      "step": 11186
+    },
+    {
+      "epoch": 0.1983186872111213,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.5413177416703094e-05,
+      "loss": 0.8364,
+      "step": 11188
+    },
+    {
+      "epoch": 0.19835413924673287,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.541156497176868e-05,
+      "loss": 0.7633,
+      "step": 11190
+    },
+    {
+      "epoch": 0.19838959128234443,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.540995227210218e-05,
+      "loss": 0.8082,
+      "step": 11192
+    },
+    {
+      "epoch": 0.19842504331795602,
+      "grad_norm": 2.875,
+      "learning_rate": 4.540833931772373e-05,
+      "loss": 0.8092,
+      "step": 11194
+    },
+    {
+      "epoch": 0.19846049535356758,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.540672610865346e-05,
+      "loss": 0.8039,
+      "step": 11196
+    },
+    {
+      "epoch": 0.19849594738917914,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.540511264491149e-05,
+      "loss": 0.8232,
+      "step": 11198
+    },
+    {
+      "epoch": 0.19853139942479073,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.540349892651797e-05,
+      "loss": 0.8326,
+      "step": 11200
+    },
+    {
+      "epoch": 0.1985668514604023,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.5401884953493035e-05,
+      "loss": 0.8045,
+      "step": 11202
+    },
+    {
+      "epoch": 0.19860230349601385,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.540027072585682e-05,
+      "loss": 0.8091,
+      "step": 11204
+    },
+    {
+      "epoch": 0.19863775553162544,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.539865624362948e-05,
+      "loss": 0.8383,
+      "step": 11206
+    },
+    {
+      "epoch": 0.198673207567237,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.5397041506831154e-05,
+      "loss": 0.7917,
+      "step": 11208
+    },
+    {
+      "epoch": 0.19870865960284856,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.5395426515482005e-05,
+      "loss": 0.7516,
+      "step": 11210
+    },
+    {
+      "epoch": 0.19874411163846015,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.5393811269602173e-05,
+      "loss": 0.7987,
+      "step": 11212
+    },
+    {
+      "epoch": 0.1987795636740717,
+      "grad_norm": 2.625,
+      "learning_rate": 4.539219576921183e-05,
+      "loss": 0.7911,
+      "step": 11214
+    },
+    {
+      "epoch": 0.19881501570968327,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.539058001433113e-05,
+      "loss": 0.8121,
+      "step": 11216
+    },
+    {
+      "epoch": 0.19885046774529486,
+      "grad_norm": 2.625,
+      "learning_rate": 4.538896400498024e-05,
+      "loss": 0.7945,
+      "step": 11218
+    },
+    {
+      "epoch": 0.19888591978090642,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.538734774117932e-05,
+      "loss": 0.8229,
+      "step": 11220
+    },
+    {
+      "epoch": 0.19892137181651798,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.538573122294856e-05,
+      "loss": 0.8015,
+      "step": 11222
+    },
+    {
+      "epoch": 0.19895682385212957,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.53841144503081e-05,
+      "loss": 0.7746,
+      "step": 11224
+    },
+    {
+      "epoch": 0.19899227588774113,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.538249742327815e-05,
+      "loss": 0.8317,
+      "step": 11226
+    },
+    {
+      "epoch": 0.1990277279233527,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.5380880141878876e-05,
+      "loss": 0.7737,
+      "step": 11228
+    },
+    {
+      "epoch": 0.19906317995896428,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.5379262606130465e-05,
+      "loss": 0.7977,
+      "step": 11230
+    },
+    {
+      "epoch": 0.19909863199457584,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.53776448160531e-05,
+      "loss": 0.7838,
+      "step": 11232
+    },
+    {
+      "epoch": 0.1991340840301874,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.537602677166697e-05,
+      "loss": 0.7875,
+      "step": 11234
+    },
+    {
+      "epoch": 0.19916953606579899,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.537440847299227e-05,
+      "loss": 0.798,
+      "step": 11236
+    },
+    {
+      "epoch": 0.19920498810141055,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.53727899200492e-05,
+      "loss": 0.8197,
+      "step": 11238
+    },
+    {
+      "epoch": 0.1992404401370221,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.537117111285795e-05,
+      "loss": 0.7943,
+      "step": 11240
+    },
+    {
+      "epoch": 0.1992758921726337,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.536955205143873e-05,
+      "loss": 0.858,
+      "step": 11242
+    },
+    {
+      "epoch": 0.19931134420824526,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.536793273581174e-05,
+      "loss": 0.781,
+      "step": 11244
+    },
+    {
+      "epoch": 0.19934679624385682,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.5366313165997196e-05,
+      "loss": 0.7907,
+      "step": 11246
+    },
+    {
+      "epoch": 0.1993822482794684,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.5364693342015306e-05,
+      "loss": 0.8009,
+      "step": 11248
+    },
+    {
+      "epoch": 0.19941770031507997,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.536307326388628e-05,
+      "loss": 0.8137,
+      "step": 11250
+    },
+    {
+      "epoch": 0.19945315235069153,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.536145293163034e-05,
+      "loss": 0.7723,
+      "step": 11252
+    },
+    {
+      "epoch": 0.19948860438630311,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.535983234526772e-05,
+      "loss": 0.7916,
+      "step": 11254
+    },
+    {
+      "epoch": 0.19952405642191468,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.5358211504818625e-05,
+      "loss": 0.7999,
+      "step": 11256
+    },
+    {
+      "epoch": 0.19955950845752624,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.535659041030329e-05,
+      "loss": 0.7989,
+      "step": 11258
+    },
+    {
+      "epoch": 0.19959496049313782,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.535496906174195e-05,
+      "loss": 0.8241,
+      "step": 11260
+    },
+    {
+      "epoch": 0.19963041252874938,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.535334745915483e-05,
+      "loss": 0.8053,
+      "step": 11262
+    },
+    {
+      "epoch": 0.19966586456436095,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.535172560256218e-05,
+      "loss": 0.8119,
+      "step": 11264
+    },
+    {
+      "epoch": 0.19970131659997253,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.535010349198423e-05,
+      "loss": 0.7971,
+      "step": 11266
+    },
+    {
+      "epoch": 0.1997367686355841,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.5348481127441226e-05,
+      "loss": 0.8,
+      "step": 11268
+    },
+    {
+      "epoch": 0.19977222067119565,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.534685850895342e-05,
+      "loss": 0.7838,
+      "step": 11270
+    },
+    {
+      "epoch": 0.19980767270680724,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.534523563654105e-05,
+      "loss": 0.7988,
+      "step": 11272
+    },
+    {
+      "epoch": 0.1998431247424188,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.5343612510224374e-05,
+      "loss": 0.8068,
+      "step": 11274
+    },
+    {
+      "epoch": 0.19987857677803036,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.534198913002367e-05,
+      "loss": 0.8058,
+      "step": 11276
+    },
+    {
+      "epoch": 0.19991402881364195,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.534036549595916e-05,
+      "loss": 0.8102,
+      "step": 11278
+    },
+    {
+      "epoch": 0.1999494808492535,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.533874160805113e-05,
+      "loss": 0.7908,
+      "step": 11280
+    },
+    {
+      "epoch": 0.19998493288486507,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.5337117466319843e-05,
+      "loss": 0.7979,
+      "step": 11282
+    },
+    {
+      "epoch": 0.20002038492047666,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.533549307078557e-05,
+      "loss": 0.8293,
+      "step": 11284
+    },
+    {
+      "epoch": 0.20005583695608822,
+      "grad_norm": 2.625,
+      "learning_rate": 4.5333868421468574e-05,
+      "loss": 0.8047,
+      "step": 11286
+    },
+    {
+      "epoch": 0.20009128899169978,
+      "grad_norm": 2.5,
+      "learning_rate": 4.533224351838914e-05,
+      "loss": 0.7941,
+      "step": 11288
+    },
+    {
+      "epoch": 0.20012674102731137,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.533061836156753e-05,
+      "loss": 0.7903,
+      "step": 11290
+    },
+    {
+      "epoch": 0.20016219306292293,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.5328992951024054e-05,
+      "loss": 0.8487,
+      "step": 11292
+    },
+    {
+      "epoch": 0.2001976450985345,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.532736728677897e-05,
+      "loss": 0.8333,
+      "step": 11294
+    },
+    {
+      "epoch": 0.20023309713414608,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.5325741368852576e-05,
+      "loss": 0.8218,
+      "step": 11296
+    },
+    {
+      "epoch": 0.20026854916975764,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.532411519726517e-05,
+      "loss": 0.7909,
+      "step": 11298
+    },
+    {
+      "epoch": 0.2003040012053692,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.532248877203703e-05,
+      "loss": 0.84,
+      "step": 11300
+    },
+    {
+      "epoch": 0.2003394532409808,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.532086209318846e-05,
+      "loss": 0.8219,
+      "step": 11302
+    },
+    {
+      "epoch": 0.20037490527659235,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.531923516073978e-05,
+      "loss": 0.7958,
+      "step": 11304
+    },
+    {
+      "epoch": 0.2004103573122039,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.5317607974711265e-05,
+      "loss": 0.8303,
+      "step": 11306
+    },
+    {
+      "epoch": 0.2004458093478155,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.5315980535123246e-05,
+      "loss": 0.817,
+      "step": 11308
+    },
+    {
+      "epoch": 0.20048126138342706,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.531435284199601e-05,
+      "loss": 0.8366,
+      "step": 11310
+    },
+    {
+      "epoch": 0.20051671341903862,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.5312724895349885e-05,
+      "loss": 0.7865,
+      "step": 11312
+    },
+    {
+      "epoch": 0.2005521654546502,
+      "grad_norm": 2.625,
+      "learning_rate": 4.531109669520519e-05,
+      "loss": 0.8117,
+      "step": 11314
+    },
+    {
+      "epoch": 0.20058761749026177,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.530946824158223e-05,
+      "loss": 0.7994,
+      "step": 11316
+    },
+    {
+      "epoch": 0.20062306952587333,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.530783953450134e-05,
+      "loss": 0.7917,
+      "step": 11318
+    },
+    {
+      "epoch": 0.20065852156148492,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.530621057398284e-05,
+      "loss": 0.8028,
+      "step": 11320
+    },
+    {
+      "epoch": 0.20069397359709648,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.530458136004706e-05,
+      "loss": 0.8238,
+      "step": 11322
+    },
+    {
+      "epoch": 0.20072942563270804,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.530295189271434e-05,
+      "loss": 0.8099,
+      "step": 11324
+    },
+    {
+      "epoch": 0.20076487766831963,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.530132217200501e-05,
+      "loss": 0.8288,
+      "step": 11326
+    },
+    {
+      "epoch": 0.2008003297039312,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.52996921979394e-05,
+      "loss": 0.8228,
+      "step": 11328
+    },
+    {
+      "epoch": 0.20083578173954275,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.5298061970537865e-05,
+      "loss": 0.8187,
+      "step": 11330
+    },
+    {
+      "epoch": 0.20087123377515434,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.529643148982074e-05,
+      "loss": 0.7506,
+      "step": 11332
+    },
+    {
+      "epoch": 0.2009066858107659,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.5294800755808385e-05,
+      "loss": 0.7772,
+      "step": 11334
+    },
+    {
+      "epoch": 0.20094213784637746,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.5293169768521135e-05,
+      "loss": 0.804,
+      "step": 11336
+    },
+    {
+      "epoch": 0.20097758988198905,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.529153852797936e-05,
+      "loss": 0.8106,
+      "step": 11338
+    },
+    {
+      "epoch": 0.2010130419176006,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.528990703420341e-05,
+      "loss": 0.8061,
+      "step": 11340
+    },
+    {
+      "epoch": 0.20104849395321217,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.528827528721364e-05,
+      "loss": 0.7881,
+      "step": 11342
+    },
+    {
+      "epoch": 0.20108394598882376,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.528664328703043e-05,
+      "loss": 0.7986,
+      "step": 11344
+    },
+    {
+      "epoch": 0.20111939802443532,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.528501103367413e-05,
+      "loss": 0.7846,
+      "step": 11346
+    },
+    {
+      "epoch": 0.20115485006004688,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.5283378527165125e-05,
+      "loss": 0.8148,
+      "step": 11348
+    },
+    {
+      "epoch": 0.20119030209565847,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.528174576752377e-05,
+      "loss": 0.7728,
+      "step": 11350
+    },
+    {
+      "epoch": 0.20122575413127003,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.528011275477045e-05,
+      "loss": 0.7757,
+      "step": 11352
+    },
+    {
+      "epoch": 0.2012612061668816,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.5278479488925563e-05,
+      "loss": 0.8256,
+      "step": 11354
+    },
+    {
+      "epoch": 0.20129665820249315,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.527684597000946e-05,
+      "loss": 0.7902,
+      "step": 11356
+    },
+    {
+      "epoch": 0.20133211023810474,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.527521219804255e-05,
+      "loss": 0.8071,
+      "step": 11358
+    },
+    {
+      "epoch": 0.2013675622737163,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.527357817304522e-05,
+      "loss": 0.8052,
+      "step": 11360
+    },
+    {
+      "epoch": 0.20140301430932786,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.527194389503784e-05,
+      "loss": 0.8477,
+      "step": 11362
+    },
+    {
+      "epoch": 0.20143846634493945,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.527030936404084e-05,
+      "loss": 0.7812,
+      "step": 11364
+    },
+    {
+      "epoch": 0.201473918380551,
+      "grad_norm": 3.25,
+      "learning_rate": 4.5268674580074594e-05,
+      "loss": 0.7922,
+      "step": 11366
+    },
+    {
+      "epoch": 0.20150937041616257,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.5267039543159504e-05,
+      "loss": 0.8091,
+      "step": 11368
+    },
+    {
+      "epoch": 0.20154482245177416,
+      "grad_norm": 2.75,
+      "learning_rate": 4.5265404253316e-05,
+      "loss": 0.8252,
+      "step": 11370
+    },
+    {
+      "epoch": 0.20158027448738572,
+      "grad_norm": 3.125,
+      "learning_rate": 4.526376871056446e-05,
+      "loss": 0.7924,
+      "step": 11372
+    },
+    {
+      "epoch": 0.20161572652299728,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.5262132914925303e-05,
+      "loss": 0.8151,
+      "step": 11374
+    },
+    {
+      "epoch": 0.20165117855860887,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.526049686641896e-05,
+      "loss": 0.8193,
+      "step": 11376
+    },
+    {
+      "epoch": 0.20168663059422043,
+      "grad_norm": 2.75,
+      "learning_rate": 4.525886056506582e-05,
+      "loss": 0.8402,
+      "step": 11378
+    },
+    {
+      "epoch": 0.201722082629832,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.5257224010886335e-05,
+      "loss": 0.8217,
+      "step": 11380
+    },
+    {
+      "epoch": 0.20175753466544358,
+      "grad_norm": 2.625,
+      "learning_rate": 4.525558720390091e-05,
+      "loss": 0.8155,
+      "step": 11382
+    },
+    {
+      "epoch": 0.20179298670105514,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.525395014412997e-05,
+      "loss": 0.7995,
+      "step": 11384
+    },
+    {
+      "epoch": 0.2018284387366667,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.525231283159395e-05,
+      "loss": 0.7724,
+      "step": 11386
+    },
+    {
+      "epoch": 0.2018638907722783,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.525067526631329e-05,
+      "loss": 0.7979,
+      "step": 11388
+    },
+    {
+      "epoch": 0.20189934280788985,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.524903744830842e-05,
+      "loss": 0.8032,
+      "step": 11390
+    },
+    {
+      "epoch": 0.2019347948435014,
+      "grad_norm": 2.75,
+      "learning_rate": 4.5247399377599773e-05,
+      "loss": 0.8137,
+      "step": 11392
+    },
+    {
+      "epoch": 0.201970246879113,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.524576105420781e-05,
+      "loss": 0.7718,
+      "step": 11394
+    },
+    {
+      "epoch": 0.20200569891472456,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.524412247815296e-05,
+      "loss": 0.8087,
+      "step": 11396
+    },
+    {
+      "epoch": 0.20204115095033612,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.524248364945568e-05,
+      "loss": 0.8201,
+      "step": 11398
+    },
+    {
+      "epoch": 0.2020766029859477,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.5240844568136415e-05,
+      "loss": 0.7793,
+      "step": 11400
+    },
+    {
+      "epoch": 0.20211205502155927,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.5239205234215634e-05,
+      "loss": 0.8182,
+      "step": 11402
+    },
+    {
+      "epoch": 0.20214750705717083,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.523756564771378e-05,
+      "loss": 0.7857,
+      "step": 11404
+    },
+    {
+      "epoch": 0.20218295909278242,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.523592580865132e-05,
+      "loss": 0.7496,
+      "step": 11406
+    },
+    {
+      "epoch": 0.20221841112839398,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.523428571704873e-05,
+      "loss": 0.7862,
+      "step": 11408
+    },
+    {
+      "epoch": 0.20225386316400554,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.523264537292646e-05,
+      "loss": 0.7743,
+      "step": 11410
+    },
+    {
+      "epoch": 0.20228931519961713,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.5231004776305e-05,
+      "loss": 0.7779,
+      "step": 11412
+    },
+    {
+      "epoch": 0.2023247672352287,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.52293639272048e-05,
+      "loss": 0.7965,
+      "step": 11414
+    },
+    {
+      "epoch": 0.20236021927084025,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.522772282564637e-05,
+      "loss": 0.8085,
+      "step": 11416
+    },
+    {
+      "epoch": 0.20239567130645184,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.522608147165016e-05,
+      "loss": 0.8085,
+      "step": 11418
+    },
+    {
+      "epoch": 0.2024311233420634,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.522443986523667e-05,
+      "loss": 0.8009,
+      "step": 11420
+    },
+    {
+      "epoch": 0.20246657537767496,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.522279800642638e-05,
+      "loss": 0.8008,
+      "step": 11422
+    },
+    {
+      "epoch": 0.20250202741328654,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.522115589523978e-05,
+      "loss": 0.826,
+      "step": 11424
+    },
+    {
+      "epoch": 0.2025374794488981,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.521951353169737e-05,
+      "loss": 0.7952,
+      "step": 11426
+    },
+    {
+      "epoch": 0.20257293148450967,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.521787091581964e-05,
+      "loss": 0.7572,
+      "step": 11428
+    },
+    {
+      "epoch": 0.20260838352012125,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.5216228047627096e-05,
+      "loss": 0.8417,
+      "step": 11430
+    },
+    {
+      "epoch": 0.20264383555573282,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.5214584927140236e-05,
+      "loss": 0.8195,
+      "step": 11432
+    },
+    {
+      "epoch": 0.20267928759134438,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.521294155437957e-05,
+      "loss": 0.7914,
+      "step": 11434
+    },
+    {
+      "epoch": 0.20271473962695596,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.52112979293656e-05,
+      "loss": 0.763,
+      "step": 11436
+    },
+    {
+      "epoch": 0.20275019166256752,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.520965405211884e-05,
+      "loss": 0.8315,
+      "step": 11438
+    },
+    {
+      "epoch": 0.20278564369817909,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.520800992265981e-05,
+      "loss": 0.7956,
+      "step": 11440
+    },
+    {
+      "epoch": 0.20282109573379067,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.520636554100902e-05,
+      "loss": 0.7905,
+      "step": 11442
+    },
+    {
+      "epoch": 0.20285654776940223,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.5204720907187004e-05,
+      "loss": 0.7493,
+      "step": 11444
+    },
+    {
+      "epoch": 0.2028919998050138,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.5203076021214274e-05,
+      "loss": 0.8327,
+      "step": 11446
+    },
+    {
+      "epoch": 0.20292745184062538,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.520143088311136e-05,
+      "loss": 0.8388,
+      "step": 11448
+    },
+    {
+      "epoch": 0.20296290387623694,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.5199785492898805e-05,
+      "loss": 0.794,
+      "step": 11450
+    },
+    {
+      "epoch": 0.2029983559118485,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.519813985059712e-05,
+      "loss": 0.8091,
+      "step": 11452
+    },
+    {
+      "epoch": 0.2030338079474601,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.519649395622687e-05,
+      "loss": 0.8035,
+      "step": 11454
+    },
+    {
+      "epoch": 0.20306925998307165,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.5194847809808585e-05,
+      "loss": 0.8301,
+      "step": 11456
+    },
+    {
+      "epoch": 0.20310471201868321,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.519320141136279e-05,
+      "loss": 0.7984,
+      "step": 11458
+    },
+    {
+      "epoch": 0.2031401640542948,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.519155476091006e-05,
+      "loss": 0.8362,
+      "step": 11460
+    },
+    {
+      "epoch": 0.20317561608990636,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.518990785847093e-05,
+      "loss": 0.8017,
+      "step": 11462
+    },
+    {
+      "epoch": 0.20321106812551792,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.5188260704065955e-05,
+      "loss": 0.8518,
+      "step": 11464
+    },
+    {
+      "epoch": 0.2032465201611295,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.518661329771569e-05,
+      "loss": 0.8023,
+      "step": 11466
+    },
+    {
+      "epoch": 0.20328197219674107,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.51849656394407e-05,
+      "loss": 0.7892,
+      "step": 11468
+    },
+    {
+      "epoch": 0.20331742423235263,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.518331772926154e-05,
+      "loss": 0.7387,
+      "step": 11470
+    },
+    {
+      "epoch": 0.20335287626796422,
+      "grad_norm": 2.875,
+      "learning_rate": 4.518166956719877e-05,
+      "loss": 0.7539,
+      "step": 11472
+    },
+    {
+      "epoch": 0.20338832830357578,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.518002115327298e-05,
+      "loss": 0.8094,
+      "step": 11474
+    },
+    {
+      "epoch": 0.20342378033918734,
+      "grad_norm": 2.875,
+      "learning_rate": 4.517837248750473e-05,
+      "loss": 0.787,
+      "step": 11476
+    },
+    {
+      "epoch": 0.20345923237479893,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.517672356991458e-05,
+      "loss": 0.7691,
+      "step": 11478
+    },
+    {
+      "epoch": 0.2034946844104105,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.517507440052313e-05,
+      "loss": 0.8491,
+      "step": 11480
+    },
+    {
+      "epoch": 0.20353013644602205,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.517342497935096e-05,
+      "loss": 0.8421,
+      "step": 11482
+    },
+    {
+      "epoch": 0.20356558848163364,
+      "grad_norm": 3.0,
+      "learning_rate": 4.517177530641864e-05,
+      "loss": 0.8048,
+      "step": 11484
+    },
+    {
+      "epoch": 0.2036010405172452,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.517012538174676e-05,
+      "loss": 0.8004,
+      "step": 11486
+    },
+    {
+      "epoch": 0.20363649255285676,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.516847520535593e-05,
+      "loss": 0.806,
+      "step": 11488
+    },
+    {
+      "epoch": 0.20367194458846835,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.516682477726673e-05,
+      "loss": 0.7854,
+      "step": 11490
+    },
+    {
+      "epoch": 0.2037073966240799,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.5165174097499755e-05,
+      "loss": 0.7726,
+      "step": 11492
+    },
+    {
+      "epoch": 0.20374284865969147,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.5163523166075594e-05,
+      "loss": 0.8349,
+      "step": 11494
+    },
+    {
+      "epoch": 0.20377830069530306,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.516187198301488e-05,
+      "loss": 0.7926,
+      "step": 11496
+    },
+    {
+      "epoch": 0.20381375273091462,
+      "grad_norm": 2.75,
+      "learning_rate": 4.516022054833819e-05,
+      "loss": 0.8755,
+      "step": 11498
+    },
+    {
+      "epoch": 0.20384920476652618,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.515856886206616e-05,
+      "loss": 0.8133,
+      "step": 11500
+    },
+    {
+      "epoch": 0.20388465680213777,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.5156916924219385e-05,
+      "loss": 0.8109,
+      "step": 11502
+    },
+    {
+      "epoch": 0.20392010883774933,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.515526473481848e-05,
+      "loss": 0.8204,
+      "step": 11504
+    },
+    {
+      "epoch": 0.2039555608733609,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.515361229388407e-05,
+      "loss": 0.78,
+      "step": 11506
+    },
+    {
+      "epoch": 0.20399101290897248,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.515195960143678e-05,
+      "loss": 0.8417,
+      "step": 11508
+    },
+    {
+      "epoch": 0.20402646494458404,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.515030665749723e-05,
+      "loss": 0.8101,
+      "step": 11510
+    },
+    {
+      "epoch": 0.2040619169801956,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.514865346208605e-05,
+      "loss": 0.8105,
+      "step": 11512
+    },
+    {
+      "epoch": 0.2040973690158072,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.514700001522387e-05,
+      "loss": 0.8042,
+      "step": 11514
+    },
+    {
+      "epoch": 0.20413282105141875,
+      "grad_norm": 2.625,
+      "learning_rate": 4.514534631693133e-05,
+      "loss": 0.7869,
+      "step": 11516
+    },
+    {
+      "epoch": 0.2041682730870303,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.5143692367229065e-05,
+      "loss": 0.8014,
+      "step": 11518
+    },
+    {
+      "epoch": 0.2042037251226419,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.5142038166137706e-05,
+      "loss": 0.8336,
+      "step": 11520
+    },
+    {
+      "epoch": 0.20423917715825346,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.5140383713677916e-05,
+      "loss": 0.7753,
+      "step": 11522
+    },
+    {
+      "epoch": 0.20427462919386502,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.513872900987032e-05,
+      "loss": 0.8049,
+      "step": 11524
+    },
+    {
+      "epoch": 0.2043100812294766,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.513707405473559e-05,
+      "loss": 0.793,
+      "step": 11526
+    },
+    {
+      "epoch": 0.20434553326508817,
+      "grad_norm": 2.875,
+      "learning_rate": 4.5135418848294366e-05,
+      "loss": 0.7965,
+      "step": 11528
+    },
+    {
+      "epoch": 0.20438098530069973,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.5133763390567316e-05,
+      "loss": 0.7977,
+      "step": 11530
+    },
+    {
+      "epoch": 0.2044164373363113,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.513210768157508e-05,
+      "loss": 0.8373,
+      "step": 11532
+    },
+    {
+      "epoch": 0.20445188937192288,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.5130451721338344e-05,
+      "loss": 0.8149,
+      "step": 11534
+    },
+    {
+      "epoch": 0.20448734140753444,
+      "grad_norm": 2.75,
+      "learning_rate": 4.5128795509877764e-05,
+      "loss": 0.7718,
+      "step": 11536
+    },
+    {
+      "epoch": 0.204522793443146,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.5127139047214006e-05,
+      "loss": 0.7862,
+      "step": 11538
+    },
+    {
+      "epoch": 0.2045582454787576,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.5125482333367744e-05,
+      "loss": 0.7925,
+      "step": 11540
+    },
+    {
+      "epoch": 0.20459369751436915,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.512382536835965e-05,
+      "loss": 0.8175,
+      "step": 11542
+    },
+    {
+      "epoch": 0.2046291495499807,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.512216815221041e-05,
+      "loss": 0.8373,
+      "step": 11544
+    },
+    {
+      "epoch": 0.2046646015855923,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.51205106849407e-05,
+      "loss": 0.8322,
+      "step": 11546
+    },
+    {
+      "epoch": 0.20470005362120386,
+      "grad_norm": 2.75,
+      "learning_rate": 4.511885296657121e-05,
+      "loss": 0.7942,
+      "step": 11548
+    },
+    {
+      "epoch": 0.20473550565681542,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.511719499712264e-05,
+      "loss": 0.7904,
+      "step": 11550
+    },
+    {
+      "epoch": 0.204770957692427,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.511553677661564e-05,
+      "loss": 0.8381,
+      "step": 11552
+    },
+    {
+      "epoch": 0.20480640972803857,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.5113878305070945e-05,
+      "loss": 0.7896,
+      "step": 11554
+    },
+    {
+      "epoch": 0.20484186176365013,
+      "grad_norm": 2.875,
+      "learning_rate": 4.5112219582509244e-05,
+      "loss": 0.8513,
+      "step": 11556
+    },
+    {
+      "epoch": 0.20487731379926172,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.511056060895122e-05,
+      "loss": 0.7517,
+      "step": 11558
+    },
+    {
+      "epoch": 0.20491276583487328,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.510890138441759e-05,
+      "loss": 0.7744,
+      "step": 11560
+    },
+    {
+      "epoch": 0.20494821787048484,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.5107241908929066e-05,
+      "loss": 0.8118,
+      "step": 11562
+    },
+    {
+      "epoch": 0.20498366990609643,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.510558218250635e-05,
+      "loss": 0.8291,
+      "step": 11564
+    },
+    {
+      "epoch": 0.205019121941708,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.5103922205170144e-05,
+      "loss": 0.794,
+      "step": 11566
+    },
+    {
+      "epoch": 0.20505457397731955,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.510226197694119e-05,
+      "loss": 0.8009,
+      "step": 11568
+    },
+    {
+      "epoch": 0.20509002601293114,
+      "grad_norm": 2.625,
+      "learning_rate": 4.510060149784019e-05,
+      "loss": 0.7495,
+      "step": 11570
+    },
+    {
+      "epoch": 0.2051254780485427,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.509894076788787e-05,
+      "loss": 0.8229,
+      "step": 11572
+    },
+    {
+      "epoch": 0.20516093008415426,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.509727978710495e-05,
+      "loss": 0.8195,
+      "step": 11574
+    },
+    {
+      "epoch": 0.20519638211976585,
+      "grad_norm": 2.625,
+      "learning_rate": 4.509561855551217e-05,
+      "loss": 0.8167,
+      "step": 11576
+    },
+    {
+      "epoch": 0.2052318341553774,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.509395707313026e-05,
+      "loss": 0.7554,
+      "step": 11578
+    },
+    {
+      "epoch": 0.20526728619098897,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.509229533997994e-05,
+      "loss": 0.8383,
+      "step": 11580
+    },
+    {
+      "epoch": 0.20530273822660056,
+      "grad_norm": 2.875,
+      "learning_rate": 4.509063335608196e-05,
+      "loss": 0.7817,
+      "step": 11582
+    },
+    {
+      "epoch": 0.20533819026221212,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.5088971121457066e-05,
+      "loss": 0.7454,
+      "step": 11584
+    },
+    {
+      "epoch": 0.20537364229782368,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.508730863612599e-05,
+      "loss": 0.8176,
+      "step": 11586
+    },
+    {
+      "epoch": 0.20540909433343527,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.50856459001095e-05,
+      "loss": 0.821,
+      "step": 11588
+    },
+    {
+      "epoch": 0.20544454636904683,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.5083982913428324e-05,
+      "loss": 0.843,
+      "step": 11590
+    },
+    {
+      "epoch": 0.2054799984046584,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.508231967610322e-05,
+      "loss": 0.8111,
+      "step": 11592
+    },
+    {
+      "epoch": 0.20551545044026998,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.5080656188154955e-05,
+      "loss": 0.7918,
+      "step": 11594
+    },
+    {
+      "epoch": 0.20555090247588154,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.507899244960429e-05,
+      "loss": 0.7829,
+      "step": 11596
+    },
+    {
+      "epoch": 0.2055863545114931,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.5077328460471965e-05,
+      "loss": 0.8103,
+      "step": 11598
+    },
+    {
+      "epoch": 0.20562180654710469,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.5075664220778766e-05,
+      "loss": 0.8046,
+      "step": 11600
+    },
+    {
+      "epoch": 0.20565725858271625,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.5073999730545466e-05,
+      "loss": 0.7595,
+      "step": 11602
+    },
+    {
+      "epoch": 0.2056927106183278,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.507233498979283e-05,
+      "loss": 0.8069,
+      "step": 11604
+    },
+    {
+      "epoch": 0.2057281626539394,
+      "grad_norm": 2.5,
+      "learning_rate": 4.507066999854164e-05,
+      "loss": 0.8049,
+      "step": 11606
+    },
+    {
+      "epoch": 0.20576361468955096,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.506900475681266e-05,
+      "loss": 0.7755,
+      "step": 11608
+    },
+    {
+      "epoch": 0.20579906672516252,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.506733926462668e-05,
+      "loss": 0.794,
+      "step": 11610
+    },
+    {
+      "epoch": 0.2058345187607741,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.5065673522004495e-05,
+      "loss": 0.7593,
+      "step": 11612
+    },
+    {
+      "epoch": 0.20586997079638567,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.5064007528966865e-05,
+      "loss": 0.7871,
+      "step": 11614
+    },
+    {
+      "epoch": 0.20590542283199723,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.506234128553461e-05,
+      "loss": 0.8246,
+      "step": 11616
+    },
+    {
+      "epoch": 0.20594087486760881,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.506067479172852e-05,
+      "loss": 0.8485,
+      "step": 11618
+    },
+    {
+      "epoch": 0.20597632690322037,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.505900804756938e-05,
+      "loss": 0.8238,
+      "step": 11620
+    },
+    {
+      "epoch": 0.20601177893883194,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.5057341053078004e-05,
+      "loss": 0.7727,
+      "step": 11622
+    },
+    {
+      "epoch": 0.20604723097444352,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.505567380827519e-05,
+      "loss": 0.7404,
+      "step": 11624
+    },
+    {
+      "epoch": 0.20608268301005508,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.505400631318174e-05,
+      "loss": 0.7966,
+      "step": 11626
+    },
+    {
+      "epoch": 0.20611813504566665,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.505233856781846e-05,
+      "loss": 0.8268,
+      "step": 11628
+    },
+    {
+      "epoch": 0.20615358708127823,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.5050670572206186e-05,
+      "loss": 0.803,
+      "step": 11630
+    },
+    {
+      "epoch": 0.2061890391168898,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.504900232636571e-05,
+      "loss": 0.7863,
+      "step": 11632
+    },
+    {
+      "epoch": 0.20622449115250135,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.5047333830317865e-05,
+      "loss": 0.8045,
+      "step": 11634
+    },
+    {
+      "epoch": 0.20625994318811294,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.504566508408347e-05,
+      "loss": 0.8201,
+      "step": 11636
+    },
+    {
+      "epoch": 0.2062953952237245,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.5043996087683346e-05,
+      "loss": 0.7946,
+      "step": 11638
+    },
+    {
+      "epoch": 0.20633084725933606,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.504232684113833e-05,
+      "loss": 0.8265,
+      "step": 11640
+    },
+    {
+      "epoch": 0.20636629929494765,
+      "grad_norm": 2.375,
+      "learning_rate": 4.504065734446925e-05,
+      "loss": 0.794,
+      "step": 11642
+    },
+    {
+      "epoch": 0.2064017513305592,
+      "grad_norm": 2.625,
+      "learning_rate": 4.503898759769694e-05,
+      "loss": 0.7796,
+      "step": 11644
+    },
+    {
+      "epoch": 0.20643720336617077,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.5037317600842235e-05,
+      "loss": 0.7687,
+      "step": 11646
+    },
+    {
+      "epoch": 0.20647265540178236,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.503564735392598e-05,
+      "loss": 0.78,
+      "step": 11648
+    },
+    {
+      "epoch": 0.20650810743739392,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.503397685696902e-05,
+      "loss": 0.7891,
+      "step": 11650
+    },
+    {
+      "epoch": 0.20654355947300548,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.5032306109992204e-05,
+      "loss": 0.7959,
+      "step": 11652
+    },
+    {
+      "epoch": 0.20657901150861707,
+      "grad_norm": 3.0,
+      "learning_rate": 4.503063511301638e-05,
+      "loss": 0.7963,
+      "step": 11654
+    },
+    {
+      "epoch": 0.20661446354422863,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.50289638660624e-05,
+      "loss": 0.7754,
+      "step": 11656
+    },
+    {
+      "epoch": 0.2066499155798402,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.502729236915112e-05,
+      "loss": 0.7979,
+      "step": 11658
+    },
+    {
+      "epoch": 0.20668536761545178,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.50256206223034e-05,
+      "loss": 0.8032,
+      "step": 11660
+    },
+    {
+      "epoch": 0.20672081965106334,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.502394862554011e-05,
+      "loss": 0.8072,
+      "step": 11662
+    },
+    {
+      "epoch": 0.2067562716866749,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.5022276378882125e-05,
+      "loss": 0.8484,
+      "step": 11664
+    },
+    {
+      "epoch": 0.2067917237222865,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.5020603882350286e-05,
+      "loss": 0.7754,
+      "step": 11666
+    },
+    {
+      "epoch": 0.20682717575789805,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.501893113596548e-05,
+      "loss": 0.8538,
+      "step": 11668
+    },
+    {
+      "epoch": 0.2068626277935096,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.501725813974858e-05,
+      "loss": 0.7993,
+      "step": 11670
+    },
+    {
+      "epoch": 0.2068980798291212,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.501558489372049e-05,
+      "loss": 0.8001,
+      "step": 11672
+    },
+    {
+      "epoch": 0.20693353186473276,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.5013911397902044e-05,
+      "loss": 0.8201,
+      "step": 11674
+    },
+    {
+      "epoch": 0.20696898390034432,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.5012237652314164e-05,
+      "loss": 0.7967,
+      "step": 11676
+    },
+    {
+      "epoch": 0.2070044359359559,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.5010563656977725e-05,
+      "loss": 0.8026,
+      "step": 11678
+    },
+    {
+      "epoch": 0.20703988797156747,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.5008889411913625e-05,
+      "loss": 0.7553,
+      "step": 11680
+    },
+    {
+      "epoch": 0.20707534000717903,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.500721491714274e-05,
+      "loss": 0.7963,
+      "step": 11682
+    },
+    {
+      "epoch": 0.20711079204279062,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.500554017268599e-05,
+      "loss": 0.8529,
+      "step": 11684
+    },
+    {
+      "epoch": 0.20714624407840218,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.5003865178564265e-05,
+      "loss": 0.7984,
+      "step": 11686
+    },
+    {
+      "epoch": 0.20718169611401374,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.500218993479847e-05,
+      "loss": 0.8345,
+      "step": 11688
+    },
+    {
+      "epoch": 0.20721714814962533,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.5000514441409505e-05,
+      "loss": 0.8268,
+      "step": 11690
+    },
+    {
+      "epoch": 0.2072526001852369,
+      "grad_norm": 2.75,
+      "learning_rate": 4.499883869841828e-05,
+      "loss": 0.8061,
+      "step": 11692
+    },
+    {
+      "epoch": 0.20728805222084845,
+      "grad_norm": 2.75,
+      "learning_rate": 4.499716270584573e-05,
+      "loss": 0.8296,
+      "step": 11694
+    },
+    {
+      "epoch": 0.20732350425646004,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.4995486463712735e-05,
+      "loss": 0.7848,
+      "step": 11696
+    },
+    {
+      "epoch": 0.2073589562920716,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.4993809972040246e-05,
+      "loss": 0.8256,
+      "step": 11698
+    },
+    {
+      "epoch": 0.20739440832768316,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.4992133230849176e-05,
+      "loss": 0.835,
+      "step": 11700
+    },
+    {
+      "epoch": 0.20742986036329472,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.499045624016044e-05,
+      "loss": 0.7905,
+      "step": 11702
+    },
+    {
+      "epoch": 0.2074653123989063,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.498877899999498e-05,
+      "loss": 0.7935,
+      "step": 11704
+    },
+    {
+      "epoch": 0.20750076443451787,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.4987101510373717e-05,
+      "loss": 0.7899,
+      "step": 11706
+    },
+    {
+      "epoch": 0.20753621647012943,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.4985423771317595e-05,
+      "loss": 0.7984,
+      "step": 11708
+    },
+    {
+      "epoch": 0.20757166850574102,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.498374578284754e-05,
+      "loss": 0.8174,
+      "step": 11710
+    },
+    {
+      "epoch": 0.20760712054135258,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.49820675449845e-05,
+      "loss": 0.8063,
+      "step": 11712
+    },
+    {
+      "epoch": 0.20764257257696414,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.498038905774942e-05,
+      "loss": 0.8267,
+      "step": 11714
+    },
+    {
+      "epoch": 0.20767802461257573,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.497871032116325e-05,
+      "loss": 0.8247,
+      "step": 11716
+    },
+    {
+      "epoch": 0.2077134766481873,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.497703133524693e-05,
+      "loss": 0.762,
+      "step": 11718
+    },
+    {
+      "epoch": 0.20774892868379885,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.497535210002143e-05,
+      "loss": 0.7811,
+      "step": 11720
+    },
+    {
+      "epoch": 0.20778438071941044,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.497367261550769e-05,
+      "loss": 0.8083,
+      "step": 11722
+    },
+    {
+      "epoch": 0.207819832755022,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.497199288172668e-05,
+      "loss": 0.7593,
+      "step": 11724
+    },
+    {
+      "epoch": 0.20785528479063356,
+      "grad_norm": 2.625,
+      "learning_rate": 4.497031289869936e-05,
+      "loss": 0.7981,
+      "step": 11726
+    },
+    {
+      "epoch": 0.20789073682624515,
+      "grad_norm": 3.0,
+      "learning_rate": 4.4968632666446684e-05,
+      "loss": 0.7701,
+      "step": 11728
+    },
+    {
+      "epoch": 0.2079261888618567,
+      "grad_norm": 2.75,
+      "learning_rate": 4.4966952184989645e-05,
+      "loss": 0.8126,
+      "step": 11730
+    },
+    {
+      "epoch": 0.20796164089746827,
+      "grad_norm": 2.25,
+      "learning_rate": 4.4965271454349186e-05,
+      "loss": 0.8052,
+      "step": 11732
+    },
+    {
+      "epoch": 0.20799709293307986,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.496359047454631e-05,
+      "loss": 0.783,
+      "step": 11734
+    },
+    {
+      "epoch": 0.20803254496869142,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.496190924560197e-05,
+      "loss": 0.8078,
+      "step": 11736
+    },
+    {
+      "epoch": 0.20806799700430298,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.4960227767537175e-05,
+      "loss": 0.8013,
+      "step": 11738
+    },
+    {
+      "epoch": 0.20810344903991457,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.4958546040372896e-05,
+      "loss": 0.7807,
+      "step": 11740
+    },
+    {
+      "epoch": 0.20813890107552613,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.495686406413011e-05,
+      "loss": 0.82,
+      "step": 11742
+    },
+    {
+      "epoch": 0.2081743531111377,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.495518183882982e-05,
+      "loss": 0.7697,
+      "step": 11744
+    },
+    {
+      "epoch": 0.20820980514674928,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.4953499364493015e-05,
+      "loss": 0.8185,
+      "step": 11746
+    },
+    {
+      "epoch": 0.20824525718236084,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.49518166411407e-05,
+      "loss": 0.8237,
+      "step": 11748
+    },
+    {
+      "epoch": 0.2082807092179724,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.4950133668793856e-05,
+      "loss": 0.8437,
+      "step": 11750
+    },
+    {
+      "epoch": 0.208316161253584,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.4948450447473515e-05,
+      "loss": 0.8473,
+      "step": 11752
+    },
+    {
+      "epoch": 0.20835161328919555,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.494676697720066e-05,
+      "loss": 0.8198,
+      "step": 11754
+    },
+    {
+      "epoch": 0.2083870653248071,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.4945083257996306e-05,
+      "loss": 0.7779,
+      "step": 11756
+    },
+    {
+      "epoch": 0.2084225173604187,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.494339928988147e-05,
+      "loss": 0.7301,
+      "step": 11758
+    },
+    {
+      "epoch": 0.20845796939603026,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.494171507287717e-05,
+      "loss": 0.8416,
+      "step": 11760
+    },
+    {
+      "epoch": 0.20849342143164182,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.4940030607004405e-05,
+      "loss": 0.8098,
+      "step": 11762
+    },
+    {
+      "epoch": 0.2085288734672534,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.4938345892284225e-05,
+      "loss": 0.7958,
+      "step": 11764
+    },
+    {
+      "epoch": 0.20856432550286497,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.493666092873763e-05,
+      "loss": 0.781,
+      "step": 11766
+    },
+    {
+      "epoch": 0.20859977753847653,
+      "grad_norm": 2.875,
+      "learning_rate": 4.493497571638567e-05,
+      "loss": 0.8235,
+      "step": 11768
+    },
+    {
+      "epoch": 0.20863522957408812,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.493329025524936e-05,
+      "loss": 0.7772,
+      "step": 11770
+    },
+    {
+      "epoch": 0.20867068160969968,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.4931604545349735e-05,
+      "loss": 0.7812,
+      "step": 11772
+    },
+    {
+      "epoch": 0.20870613364531124,
+      "grad_norm": 2.75,
+      "learning_rate": 4.492991858670784e-05,
+      "loss": 0.8014,
+      "step": 11774
+    },
+    {
+      "epoch": 0.20874158568092283,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.492823237934472e-05,
+      "loss": 0.8308,
+      "step": 11776
+    },
+    {
+      "epoch": 0.20877703771653439,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.4926545923281404e-05,
+      "loss": 0.7867,
+      "step": 11778
+    },
+    {
+      "epoch": 0.20881248975214595,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.492485921853894e-05,
+      "loss": 0.793,
+      "step": 11780
+    },
+    {
+      "epoch": 0.20884794178775754,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.492317226513839e-05,
+      "loss": 0.7952,
+      "step": 11782
+    },
+    {
+      "epoch": 0.2088833938233691,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.4921485063100796e-05,
+      "loss": 0.7814,
+      "step": 11784
+    },
+    {
+      "epoch": 0.20891884585898066,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.491979761244722e-05,
+      "loss": 0.8192,
+      "step": 11786
+    },
+    {
+      "epoch": 0.20895429789459224,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.491810991319873e-05,
+      "loss": 0.8639,
+      "step": 11788
+    },
+    {
+      "epoch": 0.2089897499302038,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.491642196537635e-05,
+      "loss": 0.7892,
+      "step": 11790
+    },
+    {
+      "epoch": 0.20902520196581537,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.491473376900119e-05,
+      "loss": 0.795,
+      "step": 11792
+    },
+    {
+      "epoch": 0.20906065400142695,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.4913045324094306e-05,
+      "loss": 0.8182,
+      "step": 11794
+    },
+    {
+      "epoch": 0.20909610603703851,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.4911356630676756e-05,
+      "loss": 0.8547,
+      "step": 11796
+    },
+    {
+      "epoch": 0.20913155807265008,
+      "grad_norm": 2.75,
+      "learning_rate": 4.4909667688769616e-05,
+      "loss": 0.7721,
+      "step": 11798
+    },
+    {
+      "epoch": 0.20916701010826166,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.490797849839398e-05,
+      "loss": 0.8018,
+      "step": 11800
+    },
+    {
+      "epoch": 0.20920246214387322,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.4906289059570916e-05,
+      "loss": 0.7506,
+      "step": 11802
+    },
+    {
+      "epoch": 0.20923791417948479,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.490459937232151e-05,
+      "loss": 0.8056,
+      "step": 11804
+    },
+    {
+      "epoch": 0.20927336621509637,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.4902909436666855e-05,
+      "loss": 0.8051,
+      "step": 11806
+    },
+    {
+      "epoch": 0.20930881825070793,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.490121925262803e-05,
+      "loss": 0.7755,
+      "step": 11808
+    },
+    {
+      "epoch": 0.2093442702863195,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.489952882022613e-05,
+      "loss": 0.8145,
+      "step": 11810
+    },
+    {
+      "epoch": 0.20937972232193108,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.4897838139482263e-05,
+      "loss": 0.7995,
+      "step": 11812
+    },
+    {
+      "epoch": 0.20941517435754264,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.489614721041751e-05,
+      "loss": 0.7982,
+      "step": 11814
+    },
+    {
+      "epoch": 0.2094506263931542,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.4894456033053e-05,
+      "loss": 0.8647,
+      "step": 11816
+    },
+    {
+      "epoch": 0.2094860784287658,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.4892764607409806e-05,
+      "loss": 0.8212,
+      "step": 11818
+    },
+    {
+      "epoch": 0.20952153046437735,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.489107293350907e-05,
+      "loss": 0.7901,
+      "step": 11820
+    },
+    {
+      "epoch": 0.20955698249998891,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.488938101137188e-05,
+      "loss": 0.8199,
+      "step": 11822
+    },
+    {
+      "epoch": 0.2095924345356005,
+      "grad_norm": 2.875,
+      "learning_rate": 4.4887688841019346e-05,
+      "loss": 0.8028,
+      "step": 11824
+    },
+    {
+      "epoch": 0.20962788657121206,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.488599642247261e-05,
+      "loss": 0.796,
+      "step": 11826
+    },
+    {
+      "epoch": 0.20966333860682362,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.4884303755752775e-05,
+      "loss": 0.8427,
+      "step": 11828
+    },
+    {
+      "epoch": 0.2096987906424352,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.488261084088098e-05,
+      "loss": 0.7922,
+      "step": 11830
+    },
+    {
+      "epoch": 0.20973424267804677,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.488091767787833e-05,
+      "loss": 0.798,
+      "step": 11832
+    },
+    {
+      "epoch": 0.20976969471365833,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.4879224266765974e-05,
+      "loss": 0.8005,
+      "step": 11834
+    },
+    {
+      "epoch": 0.20980514674926992,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.4877530607565045e-05,
+      "loss": 0.7842,
+      "step": 11836
+    },
+    {
+      "epoch": 0.20984059878488148,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.487583670029667e-05,
+      "loss": 0.8524,
+      "step": 11838
+    },
+    {
+      "epoch": 0.20987605082049304,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.487414254498199e-05,
+      "loss": 0.8432,
+      "step": 11840
+    },
+    {
+      "epoch": 0.20991150285610463,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.4872448141642156e-05,
+      "loss": 0.7839,
+      "step": 11842
+    },
+    {
+      "epoch": 0.2099469548917162,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.48707534902983e-05,
+      "loss": 0.7926,
+      "step": 11844
+    },
+    {
+      "epoch": 0.20998240692732775,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.486905859097158e-05,
+      "loss": 0.8395,
+      "step": 11846
+    },
+    {
+      "epoch": 0.21001785896293934,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.486736344368315e-05,
+      "loss": 0.7788,
+      "step": 11848
+    },
+    {
+      "epoch": 0.2100533109985509,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.4865668048454166e-05,
+      "loss": 0.7706,
+      "step": 11850
+    },
+    {
+      "epoch": 0.21008876303416246,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.486397240530578e-05,
+      "loss": 0.8222,
+      "step": 11852
+    },
+    {
+      "epoch": 0.21012421506977405,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.486227651425916e-05,
+      "loss": 0.7723,
+      "step": 11854
+    },
+    {
+      "epoch": 0.2101596671053856,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.486058037533546e-05,
+      "loss": 0.8181,
+      "step": 11856
+    },
+    {
+      "epoch": 0.21019511914099717,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.4858883988555854e-05,
+      "loss": 0.7714,
+      "step": 11858
+    },
+    {
+      "epoch": 0.21023057117660876,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.485718735394151e-05,
+      "loss": 0.8509,
+      "step": 11860
+    },
+    {
+      "epoch": 0.21026602321222032,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.485549047151361e-05,
+      "loss": 0.8047,
+      "step": 11862
+    },
+    {
+      "epoch": 0.21030147524783188,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.485379334129333e-05,
+      "loss": 0.809,
+      "step": 11864
+    },
+    {
+      "epoch": 0.21033692728344347,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.485209596330183e-05,
+      "loss": 0.7935,
+      "step": 11866
+    },
+    {
+      "epoch": 0.21037237931905503,
+      "grad_norm": 2.75,
+      "learning_rate": 4.485039833756032e-05,
+      "loss": 0.7642,
+      "step": 11868
+    },
+    {
+      "epoch": 0.2104078313546666,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.484870046408996e-05,
+      "loss": 0.8066,
+      "step": 11870
+    },
+    {
+      "epoch": 0.21044328339027815,
+      "grad_norm": 2.625,
+      "learning_rate": 4.4847002342911956e-05,
+      "loss": 0.8243,
+      "step": 11872
+    },
+    {
+      "epoch": 0.21047873542588974,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.48453039740475e-05,
+      "loss": 0.8057,
+      "step": 11874
+    },
+    {
+      "epoch": 0.2105141874615013,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.4843605357517786e-05,
+      "loss": 0.7966,
+      "step": 11876
+    },
+    {
+      "epoch": 0.21054963949711286,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.484190649334401e-05,
+      "loss": 0.8227,
+      "step": 11878
+    },
+    {
+      "epoch": 0.21058509153272445,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.484020738154737e-05,
+      "loss": 0.857,
+      "step": 11880
+    },
+    {
+      "epoch": 0.210620543568336,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.4838508022149074e-05,
+      "loss": 0.7818,
+      "step": 11882
+    },
+    {
+      "epoch": 0.21065599560394757,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.483680841517033e-05,
+      "loss": 0.7862,
+      "step": 11884
+    },
+    {
+      "epoch": 0.21069144763955916,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.4835108560632353e-05,
+      "loss": 0.7826,
+      "step": 11886
+    },
+    {
+      "epoch": 0.21072689967517072,
+      "grad_norm": 2.75,
+      "learning_rate": 4.483340845855635e-05,
+      "loss": 0.7946,
+      "step": 11888
+    },
+    {
+      "epoch": 0.21076235171078228,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.4831708108963546e-05,
+      "loss": 0.8464,
+      "step": 11890
+    },
+    {
+      "epoch": 0.21079780374639387,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.483000751187515e-05,
+      "loss": 0.8415,
+      "step": 11892
+    },
+    {
+      "epoch": 0.21083325578200543,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.4828306667312385e-05,
+      "loss": 0.8356,
+      "step": 11894
+    },
+    {
+      "epoch": 0.210868707817617,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.4826605575296486e-05,
+      "loss": 0.8171,
+      "step": 11896
+    },
+    {
+      "epoch": 0.21090415985322858,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.482490423584868e-05,
+      "loss": 0.7678,
+      "step": 11898
+    },
+    {
+      "epoch": 0.21093961188884014,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.48232026489902e-05,
+      "loss": 0.8157,
+      "step": 11900
+    },
+    {
+      "epoch": 0.2109750639244517,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.482150081474229e-05,
+      "loss": 0.8076,
+      "step": 11902
+    },
+    {
+      "epoch": 0.2110105159600633,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.481979873312616e-05,
+      "loss": 0.8188,
+      "step": 11904
+    },
+    {
+      "epoch": 0.21104596799567485,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.481809640416308e-05,
+      "loss": 0.7849,
+      "step": 11906
+    },
+    {
+      "epoch": 0.2110814200312864,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.481639382787428e-05,
+      "loss": 0.8167,
+      "step": 11908
+    },
+    {
+      "epoch": 0.211116872066898,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.481469100428102e-05,
+      "loss": 0.7851,
+      "step": 11910
+    },
+    {
+      "epoch": 0.21115232410250956,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.4812987933404535e-05,
+      "loss": 0.7196,
+      "step": 11912
+    },
+    {
+      "epoch": 0.21118777613812112,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.481128461526609e-05,
+      "loss": 0.7946,
+      "step": 11914
+    },
+    {
+      "epoch": 0.2112232281737327,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.480958104988694e-05,
+      "loss": 0.8444,
+      "step": 11916
+    },
+    {
+      "epoch": 0.21125868020934427,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.4807877237288344e-05,
+      "loss": 0.8068,
+      "step": 11918
+    },
+    {
+      "epoch": 0.21129413224495583,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.4806173177491564e-05,
+      "loss": 0.7822,
+      "step": 11920
+    },
+    {
+      "epoch": 0.21132958428056742,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.480446887051787e-05,
+      "loss": 0.8279,
+      "step": 11922
+    },
+    {
+      "epoch": 0.21136503631617898,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.4802764316388536e-05,
+      "loss": 0.7888,
+      "step": 11924
+    },
+    {
+      "epoch": 0.21140048835179054,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.480105951512482e-05,
+      "loss": 0.7546,
+      "step": 11926
+    },
+    {
+      "epoch": 0.21143594038740213,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.4799354466748e-05,
+      "loss": 0.8287,
+      "step": 11928
+    },
+    {
+      "epoch": 0.2114713924230137,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.479764917127938e-05,
+      "loss": 0.807,
+      "step": 11930
+    },
+    {
+      "epoch": 0.21150684445862525,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.4795943628740204e-05,
+      "loss": 0.7858,
+      "step": 11932
+    },
+    {
+      "epoch": 0.21154229649423684,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.479423783915177e-05,
+      "loss": 0.815,
+      "step": 11934
+    },
+    {
+      "epoch": 0.2115777485298484,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.4792531802535386e-05,
+      "loss": 0.8248,
+      "step": 11936
+    },
+    {
+      "epoch": 0.21161320056545996,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.4790825518912326e-05,
+      "loss": 0.8082,
+      "step": 11938
+    },
+    {
+      "epoch": 0.21164865260107155,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.478911898830388e-05,
+      "loss": 0.7982,
+      "step": 11940
+    },
+    {
+      "epoch": 0.2116841046366831,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.478741221073136e-05,
+      "loss": 0.7951,
+      "step": 11942
+    },
+    {
+      "epoch": 0.21171955667229467,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.478570518621604e-05,
+      "loss": 0.8596,
+      "step": 11944
+    },
+    {
+      "epoch": 0.21175500870790626,
+      "grad_norm": 2.625,
+      "learning_rate": 4.4783997914779254e-05,
+      "loss": 0.7852,
+      "step": 11946
+    },
+    {
+      "epoch": 0.21179046074351782,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.47822903964423e-05,
+      "loss": 0.7785,
+      "step": 11948
+    },
+    {
+      "epoch": 0.21182591277912938,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.478058263122646e-05,
+      "loss": 0.7551,
+      "step": 11950
+    },
+    {
+      "epoch": 0.21186136481474097,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.4778874619153086e-05,
+      "loss": 0.8051,
+      "step": 11952
+    },
+    {
+      "epoch": 0.21189681685035253,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.4777166360243474e-05,
+      "loss": 0.7988,
+      "step": 11954
+    },
+    {
+      "epoch": 0.2119322688859641,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.4775457854518944e-05,
+      "loss": 0.8211,
+      "step": 11956
+    },
+    {
+      "epoch": 0.21196772092157568,
+      "grad_norm": 2.75,
+      "learning_rate": 4.4773749102000814e-05,
+      "loss": 0.7679,
+      "step": 11958
+    },
+    {
+      "epoch": 0.21200317295718724,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.477204010271042e-05,
+      "loss": 0.8257,
+      "step": 11960
+    },
+    {
+      "epoch": 0.2120386249927988,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.477033085666909e-05,
+      "loss": 0.8224,
+      "step": 11962
+    },
+    {
+      "epoch": 0.21207407702841038,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.4768621363898135e-05,
+      "loss": 0.7642,
+      "step": 11964
+    },
+    {
+      "epoch": 0.21210952906402195,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.47669116244189e-05,
+      "loss": 0.7883,
+      "step": 11966
+    },
+    {
+      "epoch": 0.2121449810996335,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.476520163825274e-05,
+      "loss": 0.8087,
+      "step": 11968
+    },
+    {
+      "epoch": 0.2121804331352451,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.476349140542098e-05,
+      "loss": 0.8171,
+      "step": 11970
+    },
+    {
+      "epoch": 0.21221588517085666,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.476178092594495e-05,
+      "loss": 0.7746,
+      "step": 11972
+    },
+    {
+      "epoch": 0.21225133720646822,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.4760070199846016e-05,
+      "loss": 0.8099,
+      "step": 11974
+    },
+    {
+      "epoch": 0.2122867892420798,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.475835922714552e-05,
+      "loss": 0.8398,
+      "step": 11976
+    },
+    {
+      "epoch": 0.21232224127769136,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.475664800786482e-05,
+      "loss": 0.7722,
+      "step": 11978
+    },
+    {
+      "epoch": 0.21235769331330293,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.475493654202527e-05,
+      "loss": 0.7342,
+      "step": 11980
+    },
+    {
+      "epoch": 0.2123931453489145,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.475322482964823e-05,
+      "loss": 0.7683,
+      "step": 11982
+    },
+    {
+      "epoch": 0.21242859738452607,
+      "grad_norm": 2.75,
+      "learning_rate": 4.475151287075505e-05,
+      "loss": 0.8195,
+      "step": 11984
+    },
+    {
+      "epoch": 0.21246404942013764,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.4749800665367104e-05,
+      "loss": 0.8022,
+      "step": 11986
+    },
+    {
+      "epoch": 0.21249950145574922,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.474808821350576e-05,
+      "loss": 0.8183,
+      "step": 11988
+    },
+    {
+      "epoch": 0.21253495349136078,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.474637551519239e-05,
+      "loss": 0.8333,
+      "step": 11990
+    },
+    {
+      "epoch": 0.21257040552697234,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.474466257044837e-05,
+      "loss": 0.7638,
+      "step": 11992
+    },
+    {
+      "epoch": 0.21260585756258393,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.474294937929506e-05,
+      "loss": 0.7644,
+      "step": 11994
+    },
+    {
+      "epoch": 0.2126413095981955,
+      "grad_norm": 2.875,
+      "learning_rate": 4.474123594175387e-05,
+      "loss": 0.8115,
+      "step": 11996
+    },
+    {
+      "epoch": 0.21267676163380705,
+      "grad_norm": 3.25,
+      "learning_rate": 4.4739522257846154e-05,
+      "loss": 0.8353,
+      "step": 11998
+    },
+    {
+      "epoch": 0.21271221366941864,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.4737808327593325e-05,
+      "loss": 0.7944,
+      "step": 12000
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 56414,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "total_flos": 1.9936181361265607e+19,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}