diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,60951 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 8703,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.890166634385125,
+      "learning_rate": 3.816793893129771e-08,
+      "loss": 0.8609,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.018995826185668,
+      "learning_rate": 7.633587786259542e-08,
+      "loss": 0.8911,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.508199926906657,
+      "learning_rate": 1.1450381679389314e-07,
+      "loss": 0.9919,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.0106614528657305,
+      "learning_rate": 1.5267175572519085e-07,
+      "loss": 0.8818,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.636240967549296,
+      "learning_rate": 1.9083969465648858e-07,
+      "loss": 0.7282,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.719152783662321,
+      "learning_rate": 2.2900763358778629e-07,
+      "loss": 0.8999,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.597213351442028,
+      "learning_rate": 2.67175572519084e-07,
+      "loss": 0.957,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.944866045431503,
+      "learning_rate": 3.053435114503817e-07,
+      "loss": 0.9637,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.89239455707263,
+      "learning_rate": 3.4351145038167945e-07,
+      "loss": 0.9206,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.98108745270799,
+      "learning_rate": 3.8167938931297716e-07,
+      "loss": 0.8326,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.828696327021003,
+      "learning_rate": 4.1984732824427486e-07,
+      "loss": 0.9589,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.997403458064708,
+      "learning_rate": 4.5801526717557257e-07,
+      "loss": 0.9752,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.847153101830792,
+      "learning_rate": 4.961832061068702e-07,
+      "loss": 0.7562,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.023329294672322,
+      "learning_rate": 5.34351145038168e-07,
+      "loss": 0.8106,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.315806169936015,
+      "learning_rate": 5.725190839694656e-07,
+      "loss": 0.8366,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.389040742138449,
+      "learning_rate": 6.106870229007634e-07,
+      "loss": 0.7775,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.2905532425509296,
+      "learning_rate": 6.48854961832061e-07,
+      "loss": 0.8314,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.384506492690539,
+      "learning_rate": 6.870229007633589e-07,
+      "loss": 0.8064,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.4455705186879664,
+      "learning_rate": 7.251908396946565e-07,
+      "loss": 0.8606,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.9725698276763084,
+      "learning_rate": 7.633587786259543e-07,
+      "loss": 0.9061,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.3974428327769393,
+      "learning_rate": 8.01526717557252e-07,
+      "loss": 0.8059,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.192106865062429,
+      "learning_rate": 8.396946564885497e-07,
+      "loss": 0.8809,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.982366905763799,
+      "learning_rate": 8.778625954198474e-07,
+      "loss": 0.8447,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7176580466533529,
+      "learning_rate": 9.160305343511451e-07,
+      "loss": 0.9067,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.8850707390908394,
+      "learning_rate": 9.54198473282443e-07,
+      "loss": 0.7836,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.790902338373744,
+      "learning_rate": 9.923664122137404e-07,
+      "loss": 0.8438,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.8786070612893173,
+      "learning_rate": 1.0305343511450382e-06,
+      "loss": 0.7237,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.8003467080314954,
+      "learning_rate": 1.068702290076336e-06,
+      "loss": 0.7366,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.2752989644003745,
+      "learning_rate": 1.1068702290076337e-06,
+      "loss": 0.936,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.5606371594667445,
+      "learning_rate": 1.1450381679389313e-06,
+      "loss": 0.78,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.136902784546001,
+      "learning_rate": 1.1832061068702292e-06,
+      "loss": 0.8018,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.57679604702584,
+      "learning_rate": 1.2213740458015268e-06,
+      "loss": 0.8801,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.245119355469159,
+      "learning_rate": 1.2595419847328243e-06,
+      "loss": 0.7583,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4026489118889587,
+      "learning_rate": 1.297709923664122e-06,
+      "loss": 0.7171,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.428940956414717,
+      "learning_rate": 1.33587786259542e-06,
+      "loss": 0.6858,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.8484387941028255,
+      "learning_rate": 1.3740458015267178e-06,
+      "loss": 0.9102,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.9407270086799553,
+      "learning_rate": 1.4122137404580156e-06,
+      "loss": 0.808,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4972884035093785,
+      "learning_rate": 1.450381679389313e-06,
+      "loss": 0.713,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.5162915456359167,
+      "learning_rate": 1.4885496183206109e-06,
+      "loss": 0.6245,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.312310670357567,
+      "learning_rate": 1.5267175572519086e-06,
+      "loss": 0.7021,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.8754392083742295,
+      "learning_rate": 1.5648854961832064e-06,
+      "loss": 0.6495,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.2239917188403506,
+      "learning_rate": 1.603053435114504e-06,
+      "loss": 0.702,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.676483058379557,
+      "learning_rate": 1.6412213740458017e-06,
+      "loss": 0.7584,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.805202422080153,
+      "learning_rate": 1.6793893129770995e-06,
+      "loss": 0.7396,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.383085335865205,
+      "learning_rate": 1.7175572519083972e-06,
+      "loss": 0.7019,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9248306809339086,
+      "learning_rate": 1.7557251908396948e-06,
+      "loss": 0.7109,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1128613585287783,
+      "learning_rate": 1.7938931297709925e-06,
+      "loss": 0.7431,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1156451383954384,
+      "learning_rate": 1.8320610687022903e-06,
+      "loss": 0.7432,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.664197035688077,
+      "learning_rate": 1.870229007633588e-06,
+      "loss": 0.7306,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.369083502192842,
+      "learning_rate": 1.908396946564886e-06,
+      "loss": 0.615,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1686783449142677,
+      "learning_rate": 1.946564885496183e-06,
+      "loss": 0.6047,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.675193720413499,
+      "learning_rate": 1.984732824427481e-06,
+      "loss": 0.7068,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.459552106291937,
+      "learning_rate": 2.0229007633587786e-06,
+      "loss": 0.6662,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8768042034160564,
+      "learning_rate": 2.0610687022900764e-06,
+      "loss": 0.7876,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1084045313726265,
+      "learning_rate": 2.099236641221374e-06,
+      "loss": 0.6153,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.588159791258937,
+      "learning_rate": 2.137404580152672e-06,
+      "loss": 0.7345,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.966400564645722,
+      "learning_rate": 2.1755725190839697e-06,
+      "loss": 0.7583,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.390132558042109,
+      "learning_rate": 2.2137404580152674e-06,
+      "loss": 0.689,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.945110784389781,
+      "learning_rate": 2.2519083969465648e-06,
+      "loss": 0.7347,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3982587905852006,
+      "learning_rate": 2.2900763358778625e-06,
+      "loss": 0.6469,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0933810726730107,
+      "learning_rate": 2.3282442748091603e-06,
+      "loss": 0.853,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.527176685046032,
+      "learning_rate": 2.3664122137404585e-06,
+      "loss": 0.5491,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1702929851115678,
+      "learning_rate": 2.4045801526717562e-06,
+      "loss": 0.6383,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.066659740815133,
+      "learning_rate": 2.4427480916030536e-06,
+      "loss": 0.6553,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1884778761942196,
+      "learning_rate": 2.4809160305343513e-06,
+      "loss": 0.6634,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.453588640126428,
+      "learning_rate": 2.5190839694656487e-06,
+      "loss": 0.8715,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1286451543432965,
+      "learning_rate": 2.5572519083969464e-06,
+      "loss": 0.623,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.340669242361174,
+      "learning_rate": 2.595419847328244e-06,
+      "loss": 0.7098,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4298367262198837,
+      "learning_rate": 2.633587786259542e-06,
+      "loss": 0.6573,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4471078307447036,
+      "learning_rate": 2.67175572519084e-06,
+      "loss": 0.5916,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2027739797428683,
+      "learning_rate": 2.709923664122138e-06,
+      "loss": 0.5689,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.007638432464223,
+      "learning_rate": 2.7480916030534356e-06,
+      "loss": 0.6749,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0837842414821965,
+      "learning_rate": 2.7862595419847334e-06,
+      "loss": 0.6385,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8096065367593126,
+      "learning_rate": 2.824427480916031e-06,
+      "loss": 0.7018,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3172792776332614,
+      "learning_rate": 2.862595419847328e-06,
+      "loss": 0.5772,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.965324143483437,
+      "learning_rate": 2.900763358778626e-06,
+      "loss": 0.6377,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6936668913072466,
+      "learning_rate": 2.938931297709924e-06,
+      "loss": 0.5813,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.418234080082684,
+      "learning_rate": 2.9770992366412218e-06,
+      "loss": 0.6751,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3319233899484746,
+      "learning_rate": 3.0152671755725195e-06,
+      "loss": 0.6503,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3942978180672085,
+      "learning_rate": 3.0534351145038173e-06,
+      "loss": 0.6531,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.909082818677905,
+      "learning_rate": 3.091603053435115e-06,
+      "loss": 0.7365,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2267671435459437,
+      "learning_rate": 3.129770992366413e-06,
+      "loss": 0.61,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0250670457089046,
+      "learning_rate": 3.1679389312977097e-06,
+      "loss": 0.6631,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5454647514145714,
+      "learning_rate": 3.206106870229008e-06,
+      "loss": 0.5921,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2101376482641772,
+      "learning_rate": 3.2442748091603056e-06,
+      "loss": 0.9002,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1290488223839767,
+      "learning_rate": 3.2824427480916034e-06,
+      "loss": 0.8682,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.895619631338279,
+      "learning_rate": 3.320610687022901e-06,
+      "loss": 0.7295,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9264045009895512,
+      "learning_rate": 3.358778625954199e-06,
+      "loss": 0.6988,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1640218352358804,
+      "learning_rate": 3.3969465648854967e-06,
+      "loss": 0.6314,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1300174045709626,
+      "learning_rate": 3.4351145038167944e-06,
+      "loss": 0.5751,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.497241911438091,
+      "learning_rate": 3.473282442748092e-06,
+      "loss": 0.6143,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7747203732070918,
+      "learning_rate": 3.5114503816793895e-06,
+      "loss": 0.5083,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1608759596655642,
+      "learning_rate": 3.5496183206106873e-06,
+      "loss": 0.7008,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8148636087730368,
+      "learning_rate": 3.587786259541985e-06,
+      "loss": 0.5878,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.873025837356046,
+      "learning_rate": 3.625954198473283e-06,
+      "loss": 0.7022,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9611886021610716,
+      "learning_rate": 3.6641221374045806e-06,
+      "loss": 0.6229,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.732768406214643,
+      "learning_rate": 3.7022900763358783e-06,
+      "loss": 0.559,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.405107944095294,
+      "learning_rate": 3.740458015267176e-06,
+      "loss": 0.5774,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1085123456674095,
+      "learning_rate": 3.778625954198474e-06,
+      "loss": 0.6115,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3321930179653205,
+      "learning_rate": 3.816793893129772e-06,
+      "loss": 0.5939,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1244689736270033,
+      "learning_rate": 3.8549618320610685e-06,
+      "loss": 0.5375,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4722976302596957,
+      "learning_rate": 3.893129770992366e-06,
+      "loss": 0.4882,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3236423717961716,
+      "learning_rate": 3.931297709923664e-06,
+      "loss": 0.5391,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.620323160323361,
+      "learning_rate": 3.969465648854962e-06,
+      "loss": 0.6207,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9699979151630913,
+      "learning_rate": 4.0076335877862595e-06,
+      "loss": 0.7104,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.634046358661291,
+      "learning_rate": 4.045801526717557e-06,
+      "loss": 0.61,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2638414693594338,
+      "learning_rate": 4.083969465648855e-06,
+      "loss": 0.4458,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0383745075393829,
+      "learning_rate": 4.122137404580153e-06,
+      "loss": 0.7978,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.031294801958909,
+      "learning_rate": 4.1603053435114506e-06,
+      "loss": 0.594,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7089036452594197,
+      "learning_rate": 4.198473282442748e-06,
+      "loss": 0.6571,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2025902216433675,
+      "learning_rate": 4.236641221374046e-06,
+      "loss": 0.688,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.067129792466558,
+      "learning_rate": 4.274809160305344e-06,
+      "loss": 0.6357,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.434797515113872,
+      "learning_rate": 4.312977099236642e-06,
+      "loss": 0.6088,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2531317484144475,
+      "learning_rate": 4.351145038167939e-06,
+      "loss": 0.5689,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.755682999468428,
+      "learning_rate": 4.389312977099237e-06,
+      "loss": 0.6887,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6736769324060123,
+      "learning_rate": 4.427480916030535e-06,
+      "loss": 0.4289,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.359416741890606,
+      "learning_rate": 4.465648854961833e-06,
+      "loss": 0.6355,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5295402364438138,
+      "learning_rate": 4.5038167938931296e-06,
+      "loss": 0.6899,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.478525249113735,
+      "learning_rate": 4.541984732824427e-06,
+      "loss": 0.5144,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6421844712036044,
+      "learning_rate": 4.580152671755725e-06,
+      "loss": 0.6299,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.362963612843913,
+      "learning_rate": 4.618320610687023e-06,
+      "loss": 0.6567,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9279529844167191,
+      "learning_rate": 4.656488549618321e-06,
+      "loss": 0.6648,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9394517814905574,
+      "learning_rate": 4.694656488549618e-06,
+      "loss": 0.692,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.714420499303174,
+      "learning_rate": 4.732824427480917e-06,
+      "loss": 0.5938,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8280708450828147,
+      "learning_rate": 4.770992366412215e-06,
+      "loss": 0.5425,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.548074124104295,
+      "learning_rate": 4.8091603053435125e-06,
+      "loss": 0.5346,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.644079723835075,
+      "learning_rate": 4.847328244274809e-06,
+      "loss": 0.6376,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.87570972704621,
+      "learning_rate": 4.885496183206107e-06,
+      "loss": 0.5417,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2660984143895315,
+      "learning_rate": 4.923664122137405e-06,
+      "loss": 0.6728,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9953465116776197,
+      "learning_rate": 4.961832061068703e-06,
+      "loss": 0.6377,
+      "step": 130
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.080578277369364,
+      "learning_rate": 5e-06,
+      "loss": 0.6673,
+      "step": 131
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9444206455609145,
+      "learning_rate": 5.038167938931297e-06,
+      "loss": 0.6614,
+      "step": 132
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0632209046595285,
+      "learning_rate": 5.076335877862596e-06,
+      "loss": 0.6051,
+      "step": 133
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.266717827575863,
+      "learning_rate": 5.114503816793893e-06,
+      "loss": 0.6118,
+      "step": 134
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1370183075180718,
+      "learning_rate": 5.1526717557251914e-06,
+      "loss": 0.5897,
+      "step": 135
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.190800019826852,
+      "learning_rate": 5.190839694656488e-06,
+      "loss": 0.5255,
+      "step": 136
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8083730614590907,
+      "learning_rate": 5.229007633587787e-06,
+      "loss": 0.5118,
+      "step": 137
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1907072189869132,
+      "learning_rate": 5.267175572519084e-06,
+      "loss": 0.6584,
+      "step": 138
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0283430253391495,
+      "learning_rate": 5.3053435114503825e-06,
+      "loss": 0.658,
+      "step": 139
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9369334661258206,
+      "learning_rate": 5.34351145038168e-06,
+      "loss": 0.4788,
+      "step": 140
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1970211066108503,
+      "learning_rate": 5.381679389312977e-06,
+      "loss": 0.6033,
+      "step": 141
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.769573750414448,
+      "learning_rate": 5.419847328244276e-06,
+      "loss": 0.511,
+      "step": 142
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0832132330439515,
+      "learning_rate": 5.458015267175573e-06,
+      "loss": 0.5828,
+      "step": 143
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7745778751743415,
+      "learning_rate": 5.496183206106871e-06,
+      "loss": 0.6611,
+      "step": 144
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.146549747704848,
+      "learning_rate": 5.534351145038168e-06,
+      "loss": 0.66,
+      "step": 145
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.199201288842087,
+      "learning_rate": 5.572519083969467e-06,
+      "loss": 0.5655,
+      "step": 146
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.243549117473414,
+      "learning_rate": 5.610687022900764e-06,
+      "loss": 0.5957,
+      "step": 147
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0292520315634683,
+      "learning_rate": 5.648854961832062e-06,
+      "loss": 0.5581,
+      "step": 148
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.690921987953104,
+      "learning_rate": 5.687022900763359e-06,
+      "loss": 0.5743,
+      "step": 149
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1856130218803993,
+      "learning_rate": 5.725190839694656e-06,
+      "loss": 0.5946,
+      "step": 150
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.472080613760002,
+      "learning_rate": 5.763358778625955e-06,
+      "loss": 0.631,
+      "step": 151
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.96750544521273,
+      "learning_rate": 5.801526717557252e-06,
+      "loss": 0.5649,
+      "step": 152
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.700201324810448,
+      "learning_rate": 5.83969465648855e-06,
+      "loss": 0.5695,
+      "step": 153
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.415854942225705,
+      "learning_rate": 5.877862595419848e-06,
+      "loss": 0.5823,
+      "step": 154
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.90376096224584,
+      "learning_rate": 5.916030534351146e-06,
+      "loss": 0.6253,
+      "step": 155
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.626426782028894,
+      "learning_rate": 5.9541984732824435e-06,
+      "loss": 0.5143,
+      "step": 156
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3791432741123533,
+      "learning_rate": 5.992366412213741e-06,
+      "loss": 0.5803,
+      "step": 157
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.205890973911255,
+      "learning_rate": 6.030534351145039e-06,
+      "loss": 0.6824,
+      "step": 158
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8596542664171676,
+      "learning_rate": 6.068702290076336e-06,
+      "loss": 0.5981,
+      "step": 159
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.2277976360702185,
+      "learning_rate": 6.1068702290076346e-06,
+      "loss": 0.6759,
+      "step": 160
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2559753041168205,
+      "learning_rate": 6.1450381679389315e-06,
+      "loss": 0.6683,
+      "step": 161
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0295168911383654,
+      "learning_rate": 6.18320610687023e-06,
+      "loss": 0.6725,
+      "step": 162
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9910566130593819,
+      "learning_rate": 6.221374045801527e-06,
+      "loss": 0.702,
+      "step": 163
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2146347813921103,
+      "learning_rate": 6.259541984732826e-06,
+      "loss": 0.6066,
+      "step": 164
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4797392947537316,
+      "learning_rate": 6.2977099236641225e-06,
+      "loss": 0.6176,
+      "step": 165
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2455278363546824,
+      "learning_rate": 6.335877862595419e-06,
+      "loss": 0.6191,
+      "step": 166
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.90962756558809,
+      "learning_rate": 6.374045801526718e-06,
+      "loss": 0.4593,
+      "step": 167
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2384529189745477,
+      "learning_rate": 6.412213740458016e-06,
+      "loss": 0.6232,
+      "step": 168
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2377380490237555,
+      "learning_rate": 6.4503816793893135e-06,
+      "loss": 0.5157,
+      "step": 169
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.403674760526213,
+      "learning_rate": 6.488549618320611e-06,
+      "loss": 0.682,
+      "step": 170
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0848187572084997,
+      "learning_rate": 6.526717557251909e-06,
+      "loss": 0.5558,
+      "step": 171
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.955984059636698,
+      "learning_rate": 6.564885496183207e-06,
+      "loss": 0.5878,
+      "step": 172
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.061522538312033,
+      "learning_rate": 6.6030534351145046e-06,
+      "loss": 0.7833,
+      "step": 173
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7321834637588127,
+      "learning_rate": 6.641221374045802e-06,
+      "loss": 0.54,
+      "step": 174
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0042068571296685,
+      "learning_rate": 6.679389312977099e-06,
+      "loss": 0.533,
+      "step": 175
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.771325606052799,
+      "learning_rate": 6.717557251908398e-06,
+      "loss": 0.5431,
+      "step": 176
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1456729366048592,
+      "learning_rate": 6.755725190839695e-06,
+      "loss": 0.6094,
+      "step": 177
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1799578412565372,
+      "learning_rate": 6.793893129770993e-06,
+      "loss": 0.6049,
+      "step": 178
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.614696765039329,
+      "learning_rate": 6.83206106870229e-06,
+      "loss": 0.5772,
+      "step": 179
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1295583526286466,
+      "learning_rate": 6.870229007633589e-06,
+      "loss": 0.558,
+      "step": 180
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2013065626882415,
+      "learning_rate": 6.908396946564886e-06,
+      "loss": 0.5555,
+      "step": 181
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.215697624462296,
+      "learning_rate": 6.946564885496184e-06,
+      "loss": 0.5802,
+      "step": 182
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.89029255805644,
+      "learning_rate": 6.984732824427481e-06,
+      "loss": 0.6081,
+      "step": 183
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1176122493733756,
+      "learning_rate": 7.022900763358779e-06,
+      "loss": 0.5276,
+      "step": 184
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3369678727718837,
+      "learning_rate": 7.061068702290077e-06,
+      "loss": 0.5832,
+      "step": 185
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.758125192525931,
+      "learning_rate": 7.0992366412213746e-06,
+      "loss": 0.5762,
+      "step": 186
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.145207402153636,
+      "learning_rate": 7.137404580152672e-06,
+      "loss": 0.5505,
+      "step": 187
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.163194900291638,
+      "learning_rate": 7.17557251908397e-06,
+      "loss": 0.6057,
+      "step": 188
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.298119938602785,
+      "learning_rate": 7.213740458015268e-06,
+      "loss": 0.6132,
+      "step": 189
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5740904622803455,
+      "learning_rate": 7.251908396946566e-06,
+      "loss": 0.5028,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7504979431362204,
+      "learning_rate": 7.290076335877863e-06,
+      "loss": 0.6823,
+      "step": 191
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7771636257394814,
+      "learning_rate": 7.328244274809161e-06,
+      "loss": 0.6498,
+      "step": 192
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.740196274248293,
+      "learning_rate": 7.366412213740458e-06,
+      "loss": 0.5384,
+      "step": 193
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5915674486945917,
+      "learning_rate": 7.404580152671757e-06,
+      "loss": 0.6461,
+      "step": 194
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.902793071844288,
+      "learning_rate": 7.4427480916030536e-06,
+      "loss": 0.569,
+      "step": 195
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9940485866822664,
+      "learning_rate": 7.480916030534352e-06,
+      "loss": 0.5714,
+      "step": 196
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.350033204872447,
+      "learning_rate": 7.519083969465649e-06,
+      "loss": 0.5913,
+      "step": 197
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.324940727123465,
+      "learning_rate": 7.557251908396948e-06,
+      "loss": 0.6054,
+      "step": 198
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8656943462922542,
+      "learning_rate": 7.595419847328245e-06,
+      "loss": 0.5413,
+      "step": 199
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2045990742282098,
+      "learning_rate": 7.633587786259543e-06,
+      "loss": 0.54,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9996523640336585,
+      "learning_rate": 7.671755725190841e-06,
+      "loss": 0.5698,
+      "step": 201
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6789297122205507,
+      "learning_rate": 7.709923664122137e-06,
+      "loss": 0.5535,
+      "step": 202
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.409617792143921,
+      "learning_rate": 7.748091603053436e-06,
+      "loss": 0.5816,
+      "step": 203
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4573156676174337,
+      "learning_rate": 7.786259541984733e-06,
+      "loss": 0.5612,
+      "step": 204
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5934502129606405,
+      "learning_rate": 7.824427480916032e-06,
+      "loss": 0.5842,
+      "step": 205
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4850518085540276,
+      "learning_rate": 7.862595419847328e-06,
+      "loss": 0.5914,
+      "step": 206
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8889619954933417,
+      "learning_rate": 7.900763358778627e-06,
+      "loss": 0.5697,
+      "step": 207
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.282188924360206,
+      "learning_rate": 7.938931297709924e-06,
+      "loss": 0.5801,
+      "step": 208
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.086007485826392,
+      "learning_rate": 7.977099236641223e-06,
+      "loss": 0.4892,
+      "step": 209
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7916431702271172,
+      "learning_rate": 8.015267175572519e-06,
+      "loss": 0.5171,
+      "step": 210
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.319663191915118,
+      "learning_rate": 8.053435114503817e-06,
+      "loss": 0.4724,
+      "step": 211
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.187240688307719,
+      "learning_rate": 8.091603053435115e-06,
+      "loss": 0.6156,
+      "step": 212
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0433167984071057,
+      "learning_rate": 8.129770992366412e-06,
+      "loss": 0.6329,
+      "step": 213
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.654672821595755,
+      "learning_rate": 8.16793893129771e-06,
+      "loss": 0.5735,
+      "step": 214
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.153876898483677,
+      "learning_rate": 8.206106870229008e-06,
+      "loss": 0.4993,
+      "step": 215
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9590377121549656,
+      "learning_rate": 8.244274809160306e-06,
+      "loss": 0.5775,
+      "step": 216
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0244038668946165,
+      "learning_rate": 8.282442748091603e-06,
+      "loss": 0.6411,
+      "step": 217
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0297199850465115,
+      "learning_rate": 8.320610687022901e-06,
+      "loss": 0.6323,
+      "step": 218
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.419648325354996,
+      "learning_rate": 8.358778625954199e-06,
+      "loss": 0.6267,
+      "step": 219
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.923402702717104,
+      "learning_rate": 8.396946564885497e-06,
+      "loss": 0.546,
+      "step": 220
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2657269922148893,
+      "learning_rate": 8.435114503816794e-06,
+      "loss": 0.5049,
+      "step": 221
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.094469610542637,
+      "learning_rate": 8.473282442748092e-06,
+      "loss": 0.6634,
+      "step": 222
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1265986245033823,
+      "learning_rate": 8.51145038167939e-06,
+      "loss": 0.5247,
+      "step": 223
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.835965225509286,
+      "learning_rate": 8.549618320610688e-06,
+      "loss": 0.3997,
+      "step": 224
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.969936874430152,
+      "learning_rate": 8.587786259541985e-06,
+      "loss": 0.5997,
+      "step": 225
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7795713607826644,
+      "learning_rate": 8.625954198473283e-06,
+      "loss": 0.5052,
+      "step": 226
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.763988706293592,
+      "learning_rate": 8.664122137404581e-06,
+      "loss": 0.6107,
+      "step": 227
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.419915329943222,
+      "learning_rate": 8.702290076335879e-06,
+      "loss": 0.5928,
+      "step": 228
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3912195233065687,
+      "learning_rate": 8.740458015267176e-06,
+      "loss": 0.5121,
+      "step": 229
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.481112229986324,
+      "learning_rate": 8.778625954198474e-06,
+      "loss": 0.5493,
+      "step": 230
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9970811468466948,
+      "learning_rate": 8.816793893129772e-06,
+      "loss": 0.8413,
+      "step": 231
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6643719746413304,
+      "learning_rate": 8.85496183206107e-06,
+      "loss": 0.6671,
+      "step": 232
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4795027111569565,
+      "learning_rate": 8.893129770992368e-06,
+      "loss": 0.6332,
+      "step": 233
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.320890057855446,
+      "learning_rate": 8.931297709923665e-06,
+      "loss": 0.5952,
+      "step": 234
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1581194646705244,
+      "learning_rate": 8.969465648854963e-06,
+      "loss": 0.6559,
+      "step": 235
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6637875252727024,
+      "learning_rate": 9.007633587786259e-06,
+      "loss": 0.6121,
+      "step": 236
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3427402107652973,
+      "learning_rate": 9.045801526717559e-06,
+      "loss": 0.5524,
+      "step": 237
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2473693458726878,
+      "learning_rate": 9.083969465648855e-06,
+      "loss": 0.5285,
+      "step": 238
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.013251975305994,
+      "learning_rate": 9.122137404580154e-06,
+      "loss": 0.5774,
+      "step": 239
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8801664268868556,
+      "learning_rate": 9.16030534351145e-06,
+      "loss": 0.4703,
+      "step": 240
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8842559284333256,
+      "learning_rate": 9.19847328244275e-06,
+      "loss": 0.648,
+      "step": 241
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2471632106994104,
+      "learning_rate": 9.236641221374046e-06,
+      "loss": 0.5878,
+      "step": 242
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3769084743658397,
+      "learning_rate": 9.274809160305345e-06,
+      "loss": 0.6307,
+      "step": 243
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4009481548988285,
+      "learning_rate": 9.312977099236641e-06,
+      "loss": 0.5087,
+      "step": 244
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.9143294407835905,
+      "learning_rate": 9.351145038167939e-06,
+      "loss": 0.4744,
+      "step": 245
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3208513550143612,
+      "learning_rate": 9.389312977099237e-06,
+      "loss": 0.8089,
+      "step": 246
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.04438068976235,
+      "learning_rate": 9.427480916030534e-06,
+      "loss": 0.5826,
+      "step": 247
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.873175343575843,
+      "learning_rate": 9.465648854961834e-06,
+      "loss": 0.579,
+      "step": 248
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.024774012889935,
+      "learning_rate": 9.50381679389313e-06,
+      "loss": 0.5229,
+      "step": 249
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8831178030571767,
+      "learning_rate": 9.54198473282443e-06,
+      "loss": 0.4917,
+      "step": 250
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.629240897707987,
+      "learning_rate": 9.580152671755725e-06,
+      "loss": 0.5385,
+      "step": 251
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.880285496436412,
+      "learning_rate": 9.618320610687025e-06,
+      "loss": 0.5479,
+      "step": 252
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1072919562603794,
+      "learning_rate": 9.656488549618321e-06,
+      "loss": 0.5006,
+      "step": 253
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.153731884165473,
+      "learning_rate": 9.694656488549619e-06,
+      "loss": 0.5696,
+      "step": 254
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.846966274165905,
+      "learning_rate": 9.732824427480917e-06,
+      "loss": 0.6431,
+      "step": 255
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2173124690287693,
+      "learning_rate": 9.770992366412214e-06,
+      "loss": 0.5526,
+      "step": 256
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.091312726274496,
+      "learning_rate": 9.809160305343512e-06,
+      "loss": 0.5499,
+      "step": 257
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.654764801407029,
+      "learning_rate": 9.84732824427481e-06,
+      "loss": 0.5851,
+      "step": 258
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.857701175019791,
+      "learning_rate": 9.885496183206108e-06,
+      "loss": 0.6158,
+      "step": 259
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.319637965431276,
+      "learning_rate": 9.923664122137405e-06,
+      "loss": 0.5985,
+      "step": 260
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3883908579604807,
+      "learning_rate": 9.961832061068703e-06,
+      "loss": 0.5516,
+      "step": 261
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.57325364523273,
+      "learning_rate": 1e-05,
+      "loss": 0.5416,
+      "step": 262
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0248354753736764,
+      "learning_rate": 9.999999653700435e-06,
+      "loss": 0.6027,
+      "step": 263
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9730043990472355,
+      "learning_rate": 9.99999861480178e-06,
+      "loss": 0.6252,
+      "step": 264
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.118427279578174,
+      "learning_rate": 9.999996883304185e-06,
+      "loss": 0.5788,
+      "step": 265
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7881154523851575,
+      "learning_rate": 9.999994459207888e-06,
+      "loss": 0.6092,
+      "step": 266
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.264879859219285,
+      "learning_rate": 9.999991342513225e-06,
+      "loss": 0.5427,
+      "step": 267
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.048261063151237,
+      "learning_rate": 9.999987533220625e-06,
+      "loss": 0.5661,
+      "step": 268
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0040447097897056,
+      "learning_rate": 9.99998303133062e-06,
+      "loss": 0.4841,
+      "step": 269
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.598392609486478,
+      "learning_rate": 9.999977836843832e-06,
+      "loss": 0.5857,
+      "step": 270
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7250674643525077,
+      "learning_rate": 9.999971949760978e-06,
+      "loss": 0.6312,
+      "step": 271
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2529182778988164,
+      "learning_rate": 9.999965370082877e-06,
+      "loss": 0.7782,
+      "step": 272
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.828344703223877,
+      "learning_rate": 9.999958097810438e-06,
+      "loss": 0.581,
+      "step": 273
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4503651911763353,
+      "learning_rate": 9.99995013294467e-06,
+      "loss": 0.7074,
+      "step": 274
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1203970365001243,
+      "learning_rate": 9.999941475486676e-06,
+      "loss": 0.5359,
+      "step": 275
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0760521610611997,
+      "learning_rate": 9.999932125437653e-06,
+      "loss": 0.5741,
+      "step": 276
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7191643640792456,
+      "learning_rate": 9.9999220827989e-06,
+      "loss": 0.6516,
+      "step": 277
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.410014609662457,
+      "learning_rate": 9.999911347571805e-06,
+      "loss": 0.4869,
+      "step": 278
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9030209071839324,
+      "learning_rate": 9.999899919757856e-06,
+      "loss": 0.6146,
+      "step": 279
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.229525366650312,
+      "learning_rate": 9.999887799358638e-06,
+      "loss": 0.5973,
+      "step": 280
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1925446294784217,
+      "learning_rate": 9.999874986375826e-06,
+      "loss": 0.5777,
+      "step": 281
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9595952562877816,
+      "learning_rate": 9.999861480811197e-06,
+      "loss": 0.5594,
+      "step": 282
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6551425234617567,
+      "learning_rate": 9.999847282666623e-06,
+      "loss": 0.6058,
+      "step": 283
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.476059875470107,
+      "learning_rate": 9.999832391944069e-06,
+      "loss": 0.6035,
+      "step": 284
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2523132640825916,
+      "learning_rate": 9.999816808645598e-06,
+      "loss": 0.7623,
+      "step": 285
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1409126701603602,
+      "learning_rate": 9.999800532773367e-06,
+      "loss": 0.4839,
+      "step": 286
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6160244285665035,
+      "learning_rate": 9.999783564329634e-06,
+      "loss": 0.6115,
+      "step": 287
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.078427311079447,
+      "learning_rate": 9.999765903316746e-06,
+      "loss": 0.589,
+      "step": 288
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0321051193973245,
+      "learning_rate": 9.999747549737153e-06,
+      "loss": 0.6369,
+      "step": 289
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6592486244892273,
+      "learning_rate": 9.999728503593395e-06,
+      "loss": 0.6332,
+      "step": 290
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7256452984569186,
+      "learning_rate": 9.999708764888109e-06,
+      "loss": 0.6107,
+      "step": 291
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1748697429490256,
+      "learning_rate": 9.999688333624034e-06,
+      "loss": 0.6385,
+      "step": 292
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1673660380036472,
+      "learning_rate": 9.999667209803994e-06,
+      "loss": 0.4918,
+      "step": 293
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.040215141055379,
+      "learning_rate": 9.999645393430918e-06,
+      "loss": 0.5293,
+      "step": 294
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.110658684469769,
+      "learning_rate": 9.999622884507831e-06,
+      "loss": 0.5277,
+      "step": 295
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0246698956170928,
+      "learning_rate": 9.999599683037847e-06,
+      "loss": 0.6048,
+      "step": 296
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.469820273450985,
+      "learning_rate": 9.999575789024179e-06,
+      "loss": 0.5437,
+      "step": 297
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.461701150196677,
+      "learning_rate": 9.99955120247014e-06,
+      "loss": 0.4948,
+      "step": 298
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.976559385983663,
+      "learning_rate": 9.999525923379133e-06,
+      "loss": 0.5753,
+      "step": 299
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.11564352088973,
+      "learning_rate": 9.999499951754663e-06,
+      "loss": 0.5357,
+      "step": 300
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9559177558880618,
+      "learning_rate": 9.999473287600326e-06,
+      "loss": 0.5466,
+      "step": 301
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.321679877882175,
+      "learning_rate": 9.999445930919813e-06,
+      "loss": 0.5651,
+      "step": 302
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.332491225290583,
+      "learning_rate": 9.999417881716918e-06,
+      "loss": 0.4829,
+      "step": 303
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.608866943416747,
+      "learning_rate": 9.999389139995521e-06,
+      "loss": 0.603,
+      "step": 304
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1026461959102627,
+      "learning_rate": 9.999359705759607e-06,
+      "loss": 0.6674,
+      "step": 305
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.969653313306091,
+      "learning_rate": 9.999329579013254e-06,
+      "loss": 0.5813,
+      "step": 306
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.104900216286705,
+      "learning_rate": 9.999298759760634e-06,
+      "loss": 0.5826,
+      "step": 307
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.197111123007983,
+      "learning_rate": 9.999267248006013e-06,
+      "loss": 0.5492,
+      "step": 308
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.191814484436333,
+      "learning_rate": 9.999235043753761e-06,
+      "loss": 0.55,
+      "step": 309
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0641562916736063,
+      "learning_rate": 9.999202147008336e-06,
+      "loss": 0.668,
+      "step": 310
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.432220568076151,
+      "learning_rate": 9.999168557774294e-06,
+      "loss": 0.4833,
+      "step": 311
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.846028110854124,
+      "learning_rate": 9.999134276056293e-06,
+      "loss": 0.4844,
+      "step": 312
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2196168761593773,
+      "learning_rate": 9.999099301859074e-06,
+      "loss": 0.5817,
+      "step": 313
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0590875049277537,
+      "learning_rate": 9.999063635187487e-06,
+      "loss": 0.6188,
+      "step": 314
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3756564260944026,
+      "learning_rate": 9.999027276046471e-06,
+      "loss": 0.6247,
+      "step": 315
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9224280943537198,
+      "learning_rate": 9.998990224441062e-06,
+      "loss": 0.4924,
+      "step": 316
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2898728158583266,
+      "learning_rate": 9.998952480376397e-06,
+      "loss": 0.5806,
+      "step": 317
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2459749555861923,
+      "learning_rate": 9.998914043857696e-06,
+      "loss": 0.6464,
+      "step": 318
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2878193590407045,
+      "learning_rate": 9.998874914890289e-06,
+      "loss": 0.564,
+      "step": 319
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.022189877704837,
+      "learning_rate": 9.998835093479593e-06,
+      "loss": 0.6183,
+      "step": 320
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7852386263884152,
+      "learning_rate": 9.998794579631127e-06,
+      "loss": 0.7164,
+      "step": 321
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4952600513748306,
+      "learning_rate": 9.998753373350503e-06,
+      "loss": 0.6157,
+      "step": 322
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.607331735140249,
+      "learning_rate": 9.998711474643426e-06,
+      "loss": 0.6304,
+      "step": 323
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.199678650417005,
+      "learning_rate": 9.9986688835157e-06,
+      "loss": 0.5094,
+      "step": 324
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8105075138624742,
+      "learning_rate": 9.998625599973228e-06,
+      "loss": 0.6517,
+      "step": 325
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2766919691830596,
+      "learning_rate": 9.998581624022004e-06,
+      "loss": 0.7044,
+      "step": 326
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.325303305407908,
+      "learning_rate": 9.998536955668117e-06,
+      "loss": 0.5699,
+      "step": 327
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.198260263775083,
+      "learning_rate": 9.99849159491776e-06,
+      "loss": 0.5399,
+      "step": 328
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8429983667419458,
+      "learning_rate": 9.99844554177721e-06,
+      "loss": 0.4809,
+      "step": 329
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.278670123122288,
+      "learning_rate": 9.998398796252851e-06,
+      "loss": 0.5422,
+      "step": 330
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.076292643074287,
+      "learning_rate": 9.998351358351154e-06,
+      "loss": 0.6621,
+      "step": 331
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.049602712120544,
+      "learning_rate": 9.998303228078695e-06,
+      "loss": 0.5686,
+      "step": 332
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9683589965307053,
+      "learning_rate": 9.998254405442139e-06,
+      "loss": 0.599,
+      "step": 333
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3829146251777886,
+      "learning_rate": 9.998204890448247e-06,
+      "loss": 0.6134,
+      "step": 334
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8621611538553502,
+      "learning_rate": 9.99815468310388e-06,
+      "loss": 0.5618,
+      "step": 335
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9501427117240704,
+      "learning_rate": 9.99810378341599e-06,
+      "loss": 0.5592,
+      "step": 336
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2173458141975777,
+      "learning_rate": 9.998052191391633e-06,
+      "loss": 0.6047,
+      "step": 337
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1083197050433076,
+      "learning_rate": 9.99799990703795e-06,
+      "loss": 0.6123,
+      "step": 338
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3886049367478783,
+      "learning_rate": 9.997946930362186e-06,
+      "loss": 0.6042,
+      "step": 339
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.27622322310789,
+      "learning_rate": 9.99789326137168e-06,
+      "loss": 0.6229,
+      "step": 340
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.198740791702434,
+      "learning_rate": 9.997838900073864e-06,
+      "loss": 0.5225,
+      "step": 341
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.648172777903177,
+      "learning_rate": 9.997783846476268e-06,
+      "loss": 0.5987,
+      "step": 342
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1818729169978681,
+      "learning_rate": 9.997728100586522e-06,
+      "loss": 0.8191,
+      "step": 343
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.093180448968955,
+      "learning_rate": 9.997671662412343e-06,
+      "loss": 0.5669,
+      "step": 344
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.943815523422976,
+      "learning_rate": 9.997614531961552e-06,
+      "loss": 0.5554,
+      "step": 345
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8552072142104046,
+      "learning_rate": 9.99755670924206e-06,
+      "loss": 0.5835,
+      "step": 346
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5023891592952405,
+      "learning_rate": 9.99749819426188e-06,
+      "loss": 0.4249,
+      "step": 347
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.270491889889242,
+      "learning_rate": 9.997438987029115e-06,
+      "loss": 0.5382,
+      "step": 348
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.182820993235715,
+      "learning_rate": 9.997379087551968e-06,
+      "loss": 0.6573,
+      "step": 349
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.184657188264395,
+      "learning_rate": 9.997318495838734e-06,
+      "loss": 0.5796,
+      "step": 350
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.868828141168527,
+      "learning_rate": 9.997257211897808e-06,
+      "loss": 0.5867,
+      "step": 351
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.384406691103516,
+      "learning_rate": 9.99719523573768e-06,
+      "loss": 0.5157,
+      "step": 352
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.709636548182967,
+      "learning_rate": 9.997132567366931e-06,
+      "loss": 0.5179,
+      "step": 353
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.331907539317475,
+      "learning_rate": 9.997069206794246e-06,
+      "loss": 0.5786,
+      "step": 354
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.261363318381477,
+      "learning_rate": 9.9970051540284e-06,
+      "loss": 0.5264,
+      "step": 355
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.226390362268574,
+      "learning_rate": 9.996940409078265e-06,
+      "loss": 0.6287,
+      "step": 356
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.062614445723752,
+      "learning_rate": 9.99687497195281e-06,
+      "loss": 0.5419,
+      "step": 357
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5967069904878985,
+      "learning_rate": 9.9968088426611e-06,
+      "loss": 0.539,
+      "step": 358
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6998291583003247,
+      "learning_rate": 9.996742021212294e-06,
+      "loss": 0.6529,
+      "step": 359
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9269182375971687,
+      "learning_rate": 9.996674507615648e-06,
+      "loss": 0.6596,
+      "step": 360
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.285071141667131,
+      "learning_rate": 9.996606301880516e-06,
+      "loss": 0.5829,
+      "step": 361
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.293834911033014,
+      "learning_rate": 9.996537404016345e-06,
+      "loss": 0.4819,
+      "step": 362
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0478396922772317,
+      "learning_rate": 9.996467814032675e-06,
+      "loss": 0.4966,
+      "step": 363
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3024877272670117,
+      "learning_rate": 9.996397531939152e-06,
+      "loss": 0.6098,
+      "step": 364
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.763621642572886,
+      "learning_rate": 9.996326557745508e-06,
+      "loss": 0.5945,
+      "step": 365
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5785769695565834,
+      "learning_rate": 9.996254891461574e-06,
+      "loss": 0.5297,
+      "step": 366
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8393543352293973,
+      "learning_rate": 9.996182533097277e-06,
+      "loss": 0.6217,
+      "step": 367
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0539379083716467,
+      "learning_rate": 9.996109482662642e-06,
+      "loss": 0.6506,
+      "step": 368
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.893797567852461,
+      "learning_rate": 9.996035740167787e-06,
+      "loss": 0.585,
+      "step": 369
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.001101655861041,
+      "learning_rate": 9.995961305622925e-06,
+      "loss": 0.566,
+      "step": 370
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8874219676833905,
+      "learning_rate": 9.995886179038369e-06,
+      "loss": 0.4936,
+      "step": 371
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.027274851009981,
+      "learning_rate": 9.995810360424526e-06,
+      "loss": 0.58,
+      "step": 372
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.566604108630607,
+      "learning_rate": 9.995733849791895e-06,
+      "loss": 0.5217,
+      "step": 373
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0514399791053917,
+      "learning_rate": 9.995656647151077e-06,
+      "loss": 0.5771,
+      "step": 374
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3472983920797503,
+      "learning_rate": 9.995578752512767e-06,
+      "loss": 0.5467,
+      "step": 375
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5229914916968206,
+      "learning_rate": 9.995500165887753e-06,
+      "loss": 0.5958,
+      "step": 376
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3124315722023434,
+      "learning_rate": 9.995420887286922e-06,
+      "loss": 0.5041,
+      "step": 377
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7179919951485985,
+      "learning_rate": 9.995340916721252e-06,
+      "loss": 0.5433,
+      "step": 378
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.059021561522691,
+      "learning_rate": 9.995260254201826e-06,
+      "loss": 0.5805,
+      "step": 379
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7593338327926205,
+      "learning_rate": 9.995178899739813e-06,
+      "loss": 0.5509,
+      "step": 380
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2984305438292174,
+      "learning_rate": 9.995096853346486e-06,
+      "loss": 0.5658,
+      "step": 381
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7423223104243695,
+      "learning_rate": 9.995014115033207e-06,
+      "loss": 0.7361,
+      "step": 382
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0950176812170533,
+      "learning_rate": 9.994930684811439e-06,
+      "loss": 0.5415,
+      "step": 383
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.919678110671075,
+      "learning_rate": 9.994846562692735e-06,
+      "loss": 0.5431,
+      "step": 384
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8681446147061849,
+      "learning_rate": 9.994761748688752e-06,
+      "loss": 0.5691,
+      "step": 385
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.062704405777913,
+      "learning_rate": 9.994676242811236e-06,
+      "loss": 0.6474,
+      "step": 386
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8885353129770497,
+      "learning_rate": 9.994590045072034e-06,
+      "loss": 0.6155,
+      "step": 387
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1689055124084313,
+      "learning_rate": 9.994503155483081e-06,
+      "loss": 0.4243,
+      "step": 388
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.733543389171772,
+      "learning_rate": 9.99441557405642e-06,
+      "loss": 0.5591,
+      "step": 389
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.107080824162522,
+      "learning_rate": 9.994327300804177e-06,
+      "loss": 0.6166,
+      "step": 390
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 15.726449859330081,
+      "learning_rate": 9.99423833573858e-06,
+      "loss": 0.6016,
+      "step": 391
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.728308180095228,
+      "learning_rate": 9.994148678871953e-06,
+      "loss": 0.523,
+      "step": 392
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8387161864811226,
+      "learning_rate": 9.994058330216718e-06,
+      "loss": 0.5845,
+      "step": 393
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.843815542623466,
+      "learning_rate": 9.993967289785388e-06,
+      "loss": 0.6234,
+      "step": 394
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0220458623776523,
+      "learning_rate": 9.993875557590574e-06,
+      "loss": 0.7276,
+      "step": 395
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6849354741323905,
+      "learning_rate": 9.993783133644981e-06,
+      "loss": 0.6013,
+      "step": 396
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3627782083614925,
+      "learning_rate": 9.993690017961415e-06,
+      "loss": 0.6327,
+      "step": 397
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3801988820231963,
+      "learning_rate": 9.993596210552773e-06,
+      "loss": 0.5637,
+      "step": 398
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.838917952864235,
+      "learning_rate": 9.993501711432047e-06,
+      "loss": 0.5547,
+      "step": 399
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.339429495063228,
+      "learning_rate": 9.993406520612331e-06,
+      "loss": 0.536,
+      "step": 400
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.606356132851422,
+      "learning_rate": 9.993310638106808e-06,
+      "loss": 0.5137,
+      "step": 401
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9492895763624338,
+      "learning_rate": 9.99321406392876e-06,
+      "loss": 0.4944,
+      "step": 402
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8675912259371232,
+      "learning_rate": 9.993116798091565e-06,
+      "loss": 0.5491,
+      "step": 403
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5594731195330893,
+      "learning_rate": 9.993018840608695e-06,
+      "loss": 0.5727,
+      "step": 404
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0021074192983543,
+      "learning_rate": 9.99292019149372e-06,
+      "loss": 0.4979,
+      "step": 405
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.01303725913655,
+      "learning_rate": 9.992820850760306e-06,
+      "loss": 0.6247,
+      "step": 406
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.8334361587407724,
+      "learning_rate": 9.992720818422212e-06,
+      "loss": 0.6064,
+      "step": 407
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.188197789732399,
+      "learning_rate": 9.992620094493294e-06,
+      "loss": 0.5418,
+      "step": 408
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0003261952541838,
+      "learning_rate": 9.992518678987506e-06,
+      "loss": 0.6508,
+      "step": 409
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.960764124327958,
+      "learning_rate": 9.992416571918896e-06,
+      "loss": 0.5531,
+      "step": 410
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5373917503926418,
+      "learning_rate": 9.992313773301607e-06,
+      "loss": 0.5359,
+      "step": 411
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7579365451438598,
+      "learning_rate": 9.992210283149878e-06,
+      "loss": 0.5731,
+      "step": 412
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4042748754043357,
+      "learning_rate": 9.992106101478046e-06,
+      "loss": 0.636,
+      "step": 413
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7986792837706107,
+      "learning_rate": 9.992001228300541e-06,
+      "loss": 0.602,
+      "step": 414
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8286483782543528,
+      "learning_rate": 9.991895663631891e-06,
+      "loss": 0.5572,
+      "step": 415
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2546056882238448,
+      "learning_rate": 9.991789407486719e-06,
+      "loss": 0.5046,
+      "step": 416
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0282524147248493,
+      "learning_rate": 9.99168245987974e-06,
+      "loss": 0.5858,
+      "step": 417
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7585294508497795,
+      "learning_rate": 9.991574820825773e-06,
+      "loss": 0.6255,
+      "step": 418
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6141906569557247,
+      "learning_rate": 9.991466490339727e-06,
+      "loss": 0.6995,
+      "step": 419
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1058030398570418,
+      "learning_rate": 9.991357468436607e-06,
+      "loss": 0.8318,
+      "step": 420
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9495268892063249,
+      "learning_rate": 9.991247755131514e-06,
+      "loss": 0.6128,
+      "step": 421
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9432917880837769,
+      "learning_rate": 9.991137350439647e-06,
+      "loss": 0.7418,
+      "step": 422
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7765649438340525,
+      "learning_rate": 9.991026254376302e-06,
+      "loss": 0.6068,
+      "step": 423
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8415893778881127,
+      "learning_rate": 9.990914466956861e-06,
+      "loss": 0.5946,
+      "step": 424
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9023198943976065,
+      "learning_rate": 9.990801988196812e-06,
+      "loss": 0.5789,
+      "step": 425
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.873539080210441,
+      "learning_rate": 9.990688818111739e-06,
+      "loss": 0.581,
+      "step": 426
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9376791768743804,
+      "learning_rate": 9.990574956717313e-06,
+      "loss": 0.5454,
+      "step": 427
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.103342281771859,
+      "learning_rate": 9.99046040402931e-06,
+      "loss": 0.716,
+      "step": 428
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2618298551056268,
+      "learning_rate": 9.990345160063594e-06,
+      "loss": 0.8317,
+      "step": 429
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8074318950085166,
+      "learning_rate": 9.990229224836131e-06,
+      "loss": 0.5353,
+      "step": 430
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7202651792713155,
+      "learning_rate": 9.990112598362982e-06,
+      "loss": 0.6242,
+      "step": 431
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.21084013093517,
+      "learning_rate": 9.989995280660298e-06,
+      "loss": 0.5555,
+      "step": 432
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5055722170558843,
+      "learning_rate": 9.989877271744335e-06,
+      "loss": 0.5689,
+      "step": 433
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.655672632114759,
+      "learning_rate": 9.989758571631434e-06,
+      "loss": 0.5412,
+      "step": 434
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1796896497123655,
+      "learning_rate": 9.989639180338041e-06,
+      "loss": 0.6623,
+      "step": 435
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.035869877449723,
+      "learning_rate": 9.989519097880693e-06,
+      "loss": 0.5494,
+      "step": 436
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.821996513942337,
+      "learning_rate": 9.989398324276022e-06,
+      "loss": 0.5444,
+      "step": 437
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0337328320932437,
+      "learning_rate": 9.989276859540761e-06,
+      "loss": 0.522,
+      "step": 438
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8629849412543438,
+      "learning_rate": 9.989154703691735e-06,
+      "loss": 0.5428,
+      "step": 439
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.950781614154026,
+      "learning_rate": 9.98903185674586e-06,
+      "loss": 0.5241,
+      "step": 440
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9777109355800144,
+      "learning_rate": 9.98890831872016e-06,
+      "loss": 0.7578,
+      "step": 441
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9589448959919118,
+      "learning_rate": 9.988784089631742e-06,
+      "loss": 0.4597,
+      "step": 442
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6582181087938404,
+      "learning_rate": 9.988659169497816e-06,
+      "loss": 0.503,
+      "step": 443
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4628060652114927,
+      "learning_rate": 9.988533558335687e-06,
+      "loss": 0.4823,
+      "step": 444
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8714000503936454,
+      "learning_rate": 9.988407256162751e-06,
+      "loss": 0.5479,
+      "step": 445
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.038680229046823,
+      "learning_rate": 9.988280262996507e-06,
+      "loss": 0.6211,
+      "step": 446
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9990596903410613,
+      "learning_rate": 9.988152578854546e-06,
+      "loss": 0.5205,
+      "step": 447
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2408487160859374,
+      "learning_rate": 9.988024203754554e-06,
+      "loss": 0.5349,
+      "step": 448
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.345167209951377,
+      "learning_rate": 9.987895137714312e-06,
+      "loss": 0.5654,
+      "step": 449
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5886920839224363,
+      "learning_rate": 9.9877653807517e-06,
+      "loss": 0.4432,
+      "step": 450
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5227163023034136,
+      "learning_rate": 9.98763493288469e-06,
+      "loss": 0.5042,
+      "step": 451
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8525528818109034,
+      "learning_rate": 9.987503794131358e-06,
+      "loss": 0.4666,
+      "step": 452
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6459779759608986,
+      "learning_rate": 9.987371964509859e-06,
+      "loss": 0.674,
+      "step": 453
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.927029726884124,
+      "learning_rate": 9.98723944403846e-06,
+      "loss": 0.6311,
+      "step": 454
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5910725050736625,
+      "learning_rate": 9.987106232735519e-06,
+      "loss": 0.582,
+      "step": 455
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2551403430256682,
+      "learning_rate": 9.986972330619485e-06,
+      "loss": 0.5943,
+      "step": 456
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8514670538312779,
+      "learning_rate": 9.986837737708907e-06,
+      "loss": 0.5484,
+      "step": 457
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0741685658042917,
+      "learning_rate": 9.98670245402243e-06,
+      "loss": 0.537,
+      "step": 458
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7589610541469263,
+      "learning_rate": 9.986566479578795e-06,
+      "loss": 0.4586,
+      "step": 459
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.328485838345172,
+      "learning_rate": 9.986429814396831e-06,
+      "loss": 0.5975,
+      "step": 460
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.339460657509247,
+      "learning_rate": 9.986292458495474e-06,
+      "loss": 0.6011,
+      "step": 461
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.383369483186488,
+      "learning_rate": 9.986154411893752e-06,
+      "loss": 0.4756,
+      "step": 462
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.923182102750748,
+      "learning_rate": 9.986015674610782e-06,
+      "loss": 0.6031,
+      "step": 463
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7906864713091486,
+      "learning_rate": 9.985876246665784e-06,
+      "loss": 0.5685,
+      "step": 464
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0371027699558106,
+      "learning_rate": 9.985736128078073e-06,
+      "loss": 0.5506,
+      "step": 465
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0234052447574102,
+      "learning_rate": 9.985595318867057e-06,
+      "loss": 0.5514,
+      "step": 466
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8008704431824296,
+      "learning_rate": 9.985453819052241e-06,
+      "loss": 0.5111,
+      "step": 467
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2749688087157405,
+      "learning_rate": 9.985311628653224e-06,
+      "loss": 0.6589,
+      "step": 468
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.248782408003784,
+      "learning_rate": 9.985168747689706e-06,
+      "loss": 0.4992,
+      "step": 469
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.227040156295523,
+      "learning_rate": 9.985025176181476e-06,
+      "loss": 0.7168,
+      "step": 470
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6846112289588329,
+      "learning_rate": 9.984880914148421e-06,
+      "loss": 0.6324,
+      "step": 471
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7463735413812818,
+      "learning_rate": 9.984735961610525e-06,
+      "loss": 0.6665,
+      "step": 472
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6721300206774572,
+      "learning_rate": 9.984590318587869e-06,
+      "loss": 0.4869,
+      "step": 473
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.707871299162142,
+      "learning_rate": 9.984443985100625e-06,
+      "loss": 0.5766,
+      "step": 474
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4823788121158326,
+      "learning_rate": 9.984296961169062e-06,
+      "loss": 0.5238,
+      "step": 475
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3401369073823535,
+      "learning_rate": 9.984149246813548e-06,
+      "loss": 0.5448,
+      "step": 476
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9538483696775497,
+      "learning_rate": 9.984000842054543e-06,
+      "loss": 0.4805,
+      "step": 477
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.080547703494317,
+      "learning_rate": 9.983851746912605e-06,
+      "loss": 0.5866,
+      "step": 478
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7923393332855886,
+      "learning_rate": 9.983701961408386e-06,
+      "loss": 0.4749,
+      "step": 479
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1611675163842454,
+      "learning_rate": 9.983551485562635e-06,
+      "loss": 0.61,
+      "step": 480
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0889012912703855,
+      "learning_rate": 9.983400319396195e-06,
+      "loss": 0.5138,
+      "step": 481
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7966073065248644,
+      "learning_rate": 9.983248462930007e-06,
+      "loss": 0.4545,
+      "step": 482
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8321383959330082,
+      "learning_rate": 9.983095916185104e-06,
+      "loss": 0.6,
+      "step": 483
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9010991046299766,
+      "learning_rate": 9.982942679182617e-06,
+      "loss": 0.6006,
+      "step": 484
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6934192548515272,
+      "learning_rate": 9.982788751943774e-06,
+      "loss": 0.5647,
+      "step": 485
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1723359435729503,
+      "learning_rate": 9.982634134489897e-06,
+      "loss": 0.5833,
+      "step": 486
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.443317030687206,
+      "learning_rate": 9.982478826842402e-06,
+      "loss": 0.6543,
+      "step": 487
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.237864220129367,
+      "learning_rate": 9.982322829022804e-06,
+      "loss": 0.6084,
+      "step": 488
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.678665386363977,
+      "learning_rate": 9.98216614105271e-06,
+      "loss": 0.5761,
+      "step": 489
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.664362696197754,
+      "learning_rate": 9.982008762953823e-06,
+      "loss": 0.6619,
+      "step": 490
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.64478336216436,
+      "learning_rate": 9.981850694747948e-06,
+      "loss": 0.567,
+      "step": 491
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8067859424901966,
+      "learning_rate": 9.981691936456975e-06,
+      "loss": 0.5719,
+      "step": 492
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3652311385538964,
+      "learning_rate": 9.9815324881029e-06,
+      "loss": 0.534,
+      "step": 493
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.403731211526843,
+      "learning_rate": 9.981372349707806e-06,
+      "loss": 0.5663,
+      "step": 494
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9666329410778112,
+      "learning_rate": 9.981211521293878e-06,
+      "loss": 0.5089,
+      "step": 495
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7800792585259921,
+      "learning_rate": 9.981050002883392e-06,
+      "loss": 0.514,
+      "step": 496
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8486852927173552,
+      "learning_rate": 9.980887794498725e-06,
+      "loss": 0.577,
+      "step": 497
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.056354988417175,
+      "learning_rate": 9.980724896162339e-06,
+      "loss": 0.4936,
+      "step": 498
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.9894753824333082,
+      "learning_rate": 9.980561307896806e-06,
+      "loss": 0.5935,
+      "step": 499
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3374831211052096,
+      "learning_rate": 9.980397029724782e-06,
+      "loss": 0.5741,
+      "step": 500
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0030874862557044,
+      "learning_rate": 9.980232061669025e-06,
+      "loss": 0.5441,
+      "step": 501
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9073596474713068,
+      "learning_rate": 9.980066403752386e-06,
+      "loss": 0.5972,
+      "step": 502
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8774714806592239,
+      "learning_rate": 9.97990005599781e-06,
+      "loss": 0.5625,
+      "step": 503
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3870696207094775,
+      "learning_rate": 9.97973301842834e-06,
+      "loss": 0.5669,
+      "step": 504
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.045454811429022,
+      "learning_rate": 9.979565291067117e-06,
+      "loss": 0.4997,
+      "step": 505
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.846742309221474,
+      "learning_rate": 9.979396873937372e-06,
+      "loss": 0.5514,
+      "step": 506
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.328248967421797,
+      "learning_rate": 9.979227767062434e-06,
+      "loss": 0.5725,
+      "step": 507
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.492304612898177,
+      "learning_rate": 9.979057970465727e-06,
+      "loss": 0.4954,
+      "step": 508
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8449710513830246,
+      "learning_rate": 9.978887484170775e-06,
+      "loss": 0.5348,
+      "step": 509
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.249942402600172,
+      "learning_rate": 9.978716308201188e-06,
+      "loss": 0.5527,
+      "step": 510
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.435225984382256,
+      "learning_rate": 9.978544442580683e-06,
+      "loss": 0.5554,
+      "step": 511
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9967300292137642,
+      "learning_rate": 9.978371887333062e-06,
+      "loss": 0.6985,
+      "step": 512
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.098075739938484,
+      "learning_rate": 9.97819864248223e-06,
+      "loss": 0.5478,
+      "step": 513
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0108043273276905,
+      "learning_rate": 9.978024708052185e-06,
+      "loss": 0.5786,
+      "step": 514
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8886420429682862,
+      "learning_rate": 9.97785008406702e-06,
+      "loss": 0.5368,
+      "step": 515
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4769256597522853,
+      "learning_rate": 9.977674770550922e-06,
+      "loss": 0.5564,
+      "step": 516
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0953448724306982,
+      "learning_rate": 9.977498767528177e-06,
+      "loss": 0.5797,
+      "step": 517
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9274272714522065,
+      "learning_rate": 9.977322075023165e-06,
+      "loss": 0.5284,
+      "step": 518
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.290609229407683,
+      "learning_rate": 9.977144693060364e-06,
+      "loss": 0.6271,
+      "step": 519
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2994469922299414,
+      "learning_rate": 9.97696662166434e-06,
+      "loss": 0.657,
+      "step": 520
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4479291045284794,
+      "learning_rate": 9.97678786085976e-06,
+      "loss": 0.5339,
+      "step": 521
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7979868268627963,
+      "learning_rate": 9.97660841067139e-06,
+      "loss": 0.6454,
+      "step": 522
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0243030318189406,
+      "learning_rate": 9.976428271124084e-06,
+      "loss": 0.5819,
+      "step": 523
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9829068690224358,
+      "learning_rate": 9.976247442242796e-06,
+      "loss": 0.5287,
+      "step": 524
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.131388712989478,
+      "learning_rate": 9.976065924052574e-06,
+      "loss": 0.5158,
+      "step": 525
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.448431338062217,
+      "learning_rate": 9.975883716578563e-06,
+      "loss": 0.5648,
+      "step": 526
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.528208843117592,
+      "learning_rate": 9.975700819846e-06,
+      "loss": 0.7358,
+      "step": 527
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0094011920873736,
+      "learning_rate": 9.975517233880223e-06,
+      "loss": 0.5864,
+      "step": 528
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7850968093471051,
+      "learning_rate": 9.975332958706659e-06,
+      "loss": 0.5882,
+      "step": 529
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9995053929934854,
+      "learning_rate": 9.975147994350836e-06,
+      "loss": 0.4404,
+      "step": 530
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.263966489162853,
+      "learning_rate": 9.974962340838375e-06,
+      "loss": 0.4928,
+      "step": 531
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2152130393922254,
+      "learning_rate": 9.97477599819499e-06,
+      "loss": 0.4812,
+      "step": 532
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7523741435856102,
+      "learning_rate": 9.974588966446498e-06,
+      "loss": 0.6108,
+      "step": 533
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9739011635537678,
+      "learning_rate": 9.974401245618804e-06,
+      "loss": 0.6127,
+      "step": 534
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1225189214538274,
+      "learning_rate": 9.974212835737908e-06,
+      "loss": 0.5691,
+      "step": 535
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.315986828717084,
+      "learning_rate": 9.974023736829915e-06,
+      "loss": 0.5326,
+      "step": 536
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1004171697715934,
+      "learning_rate": 9.973833948921014e-06,
+      "loss": 0.6036,
+      "step": 537
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0670416615712317,
+      "learning_rate": 9.973643472037495e-06,
+      "loss": 0.538,
+      "step": 538
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.841554168594649,
+      "learning_rate": 9.973452306205745e-06,
+      "loss": 0.4935,
+      "step": 539
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0343671015390488,
+      "learning_rate": 9.973260451452242e-06,
+      "loss": 0.5615,
+      "step": 540
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2651768782888326,
+      "learning_rate": 9.973067907803564e-06,
+      "loss": 0.5148,
+      "step": 541
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.016895763006731,
+      "learning_rate": 9.97287467528638e-06,
+      "loss": 0.5124,
+      "step": 542
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.000680555027487,
+      "learning_rate": 9.972680753927457e-06,
+      "loss": 0.5512,
+      "step": 543
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0337247130499536,
+      "learning_rate": 9.972486143753658e-06,
+      "loss": 0.4864,
+      "step": 544
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.901068470180426,
+      "learning_rate": 9.972290844791939e-06,
+      "loss": 0.4951,
+      "step": 545
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.788125393228748,
+      "learning_rate": 9.972094857069355e-06,
+      "loss": 0.5923,
+      "step": 546
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7392139323251832,
+      "learning_rate": 9.97189818061305e-06,
+      "loss": 0.5834,
+      "step": 547
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.995990366492805,
+      "learning_rate": 9.971700815450272e-06,
+      "loss": 0.5108,
+      "step": 548
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.124328921818091,
+      "learning_rate": 9.971502761608356e-06,
+      "loss": 0.5594,
+      "step": 549
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9822774996278716,
+      "learning_rate": 9.97130401911474e-06,
+      "loss": 0.6016,
+      "step": 550
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.168248153135091,
+      "learning_rate": 9.971104587996954e-06,
+      "loss": 0.5476,
+      "step": 551
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.920272165963318,
+      "learning_rate": 9.97090446828262e-06,
+      "loss": 0.517,
+      "step": 552
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7401555794657373,
+      "learning_rate": 9.970703659999459e-06,
+      "loss": 0.5068,
+      "step": 553
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2512367458212705,
+      "learning_rate": 9.97050216317529e-06,
+      "loss": 0.5231,
+      "step": 554
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9294410448670494,
+      "learning_rate": 9.97029997783802e-06,
+      "loss": 0.5074,
+      "step": 555
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6861335932259194,
+      "learning_rate": 9.970097104015661e-06,
+      "loss": 0.5134,
+      "step": 556
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7293325596022315,
+      "learning_rate": 9.969893541736308e-06,
+      "loss": 0.5173,
+      "step": 557
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2897855167340326,
+      "learning_rate": 9.969689291028166e-06,
+      "loss": 0.4951,
+      "step": 558
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6993206170666062,
+      "learning_rate": 9.969484351919523e-06,
+      "loss": 0.5454,
+      "step": 559
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9847058035952536,
+      "learning_rate": 9.969278724438768e-06,
+      "loss": 0.7885,
+      "step": 560
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.077648530226415,
+      "learning_rate": 9.969072408614385e-06,
+      "loss": 0.6087,
+      "step": 561
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.119218554323957,
+      "learning_rate": 9.968865404474952e-06,
+      "loss": 0.5378,
+      "step": 562
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9691991776761808,
+      "learning_rate": 9.968657712049144e-06,
+      "loss": 0.5568,
+      "step": 563
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.445572947457805,
+      "learning_rate": 9.968449331365732e-06,
+      "loss": 0.5358,
+      "step": 564
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0075204873900834,
+      "learning_rate": 9.968240262453577e-06,
+      "loss": 0.6798,
+      "step": 565
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5988639406998886,
+      "learning_rate": 9.968030505341642e-06,
+      "loss": 0.5685,
+      "step": 566
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7955216967813392,
+      "learning_rate": 9.967820060058982e-06,
+      "loss": 0.5385,
+      "step": 567
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7126955413396432,
+      "learning_rate": 9.967608926634748e-06,
+      "loss": 0.5401,
+      "step": 568
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7935880187862583,
+      "learning_rate": 9.967397105098187e-06,
+      "loss": 0.4806,
+      "step": 569
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.901029781106834,
+      "learning_rate": 9.967184595478637e-06,
+      "loss": 0.53,
+      "step": 570
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4289977395109075,
+      "learning_rate": 9.966971397805538e-06,
+      "loss": 0.5637,
+      "step": 571
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1260725695208773,
+      "learning_rate": 9.966757512108422e-06,
+      "loss": 0.6531,
+      "step": 572
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.095846757003945,
+      "learning_rate": 9.966542938416916e-06,
+      "loss": 0.5741,
+      "step": 573
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7959300946548455,
+      "learning_rate": 9.966327676760741e-06,
+      "loss": 0.5972,
+      "step": 574
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.949517420718697,
+      "learning_rate": 9.966111727169717e-06,
+      "loss": 0.6073,
+      "step": 575
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7890006154838127,
+      "learning_rate": 9.965895089673757e-06,
+      "loss": 0.616,
+      "step": 576
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.703044495347558,
+      "learning_rate": 9.965677764302869e-06,
+      "loss": 0.4951,
+      "step": 577
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0557594779185466,
+      "learning_rate": 9.965459751087156e-06,
+      "loss": 0.7773,
+      "step": 578
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.511284441150251,
+      "learning_rate": 9.965241050056821e-06,
+      "loss": 0.5461,
+      "step": 579
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.987912906953852,
+      "learning_rate": 9.965021661242153e-06,
+      "loss": 0.6269,
+      "step": 580
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.163326114097087,
+      "learning_rate": 9.964801584673548e-06,
+      "loss": 0.6195,
+      "step": 581
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.643189298274756,
+      "learning_rate": 9.964580820381484e-06,
+      "loss": 0.5452,
+      "step": 582
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8040576097433332,
+      "learning_rate": 9.964359368396545e-06,
+      "loss": 0.4765,
+      "step": 583
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6349169764913407,
+      "learning_rate": 9.964137228749409e-06,
+      "loss": 0.5971,
+      "step": 584
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.789740294636757,
+      "learning_rate": 9.963914401470842e-06,
+      "loss": 0.548,
+      "step": 585
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.668166353592765,
+      "learning_rate": 9.96369088659171e-06,
+      "loss": 0.4713,
+      "step": 586
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2800896199502083,
+      "learning_rate": 9.96346668414298e-06,
+      "loss": 0.5463,
+      "step": 587
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9164841038556226,
+      "learning_rate": 9.963241794155701e-06,
+      "loss": 0.5258,
+      "step": 588
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2434150448698325,
+      "learning_rate": 9.96301621666103e-06,
+      "loss": 0.5389,
+      "step": 589
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.89420857530226,
+      "learning_rate": 9.962789951690213e-06,
+      "loss": 0.4505,
+      "step": 590
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.200242949607029,
+      "learning_rate": 9.96256299927459e-06,
+      "loss": 0.8496,
+      "step": 591
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.813874014307887,
+      "learning_rate": 9.9623353594456e-06,
+      "loss": 0.4867,
+      "step": 592
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8131360669421115,
+      "learning_rate": 9.962107032234775e-06,
+      "loss": 0.4694,
+      "step": 593
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.851464975827319,
+      "learning_rate": 9.961878017673746e-06,
+      "loss": 0.6189,
+      "step": 594
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.01308798272705,
+      "learning_rate": 9.961648315794231e-06,
+      "loss": 0.5765,
+      "step": 595
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0047703251731694,
+      "learning_rate": 9.961417926628051e-06,
+      "loss": 0.5202,
+      "step": 596
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8702049380112458,
+      "learning_rate": 9.96118685020712e-06,
+      "loss": 0.5994,
+      "step": 597
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0908730176025663,
+      "learning_rate": 9.960955086563447e-06,
+      "loss": 0.4698,
+      "step": 598
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9250156354503047,
+      "learning_rate": 9.960722635729131e-06,
+      "loss": 0.4996,
+      "step": 599
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9327650319495364,
+      "learning_rate": 9.96048949773638e-06,
+      "loss": 0.5255,
+      "step": 600
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.413015836197884,
+      "learning_rate": 9.960255672617478e-06,
+      "loss": 0.5819,
+      "step": 601
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.958274460474823,
+      "learning_rate": 9.96002116040482e-06,
+      "loss": 0.7668,
+      "step": 602
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5927236709717243,
+      "learning_rate": 9.959785961130892e-06,
+      "loss": 0.4978,
+      "step": 603
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9099822747105775,
+      "learning_rate": 9.95955007482827e-06,
+      "loss": 0.7573,
+      "step": 604
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.01824485284765,
+      "learning_rate": 9.959313501529633e-06,
+      "loss": 0.5239,
+      "step": 605
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1559745511651243,
+      "learning_rate": 9.959076241267747e-06,
+      "loss": 0.5919,
+      "step": 606
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2206890952663016,
+      "learning_rate": 9.95883829407548e-06,
+      "loss": 0.5849,
+      "step": 607
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.07008264681777,
+      "learning_rate": 9.95859965998579e-06,
+      "loss": 0.514,
+      "step": 608
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1568321442281873,
+      "learning_rate": 9.958360339031734e-06,
+      "loss": 0.5549,
+      "step": 609
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7154895186161723,
+      "learning_rate": 9.958120331246464e-06,
+      "loss": 0.6352,
+      "step": 610
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0154440267987495,
+      "learning_rate": 9.957879636663224e-06,
+      "loss": 0.498,
+      "step": 611
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3563732777713553,
+      "learning_rate": 9.957638255315354e-06,
+      "loss": 0.5744,
+      "step": 612
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.464455407316683,
+      "learning_rate": 9.957396187236292e-06,
+      "loss": 0.532,
+      "step": 613
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7318349809915812,
+      "learning_rate": 9.95715343245957e-06,
+      "loss": 0.6437,
+      "step": 614
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.444444998095691,
+      "learning_rate": 9.956909991018813e-06,
+      "loss": 0.4807,
+      "step": 615
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7174291491536446,
+      "learning_rate": 9.956665862947743e-06,
+      "loss": 0.5887,
+      "step": 616
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3789116517547813,
+      "learning_rate": 9.956421048280174e-06,
+      "loss": 0.45,
+      "step": 617
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.484620723750597,
+      "learning_rate": 9.956175547050022e-06,
+      "loss": 0.5533,
+      "step": 618
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.122355559497418,
+      "learning_rate": 9.955929359291291e-06,
+      "loss": 0.6512,
+      "step": 619
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8120982758466313,
+      "learning_rate": 9.955682485038084e-06,
+      "loss": 0.5641,
+      "step": 620
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6795786049320425,
+      "learning_rate": 9.955434924324596e-06,
+      "loss": 0.5031,
+      "step": 621
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0326293280023964,
+      "learning_rate": 9.955186677185122e-06,
+      "loss": 0.672,
+      "step": 622
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3474402745204173,
+      "learning_rate": 9.954937743654048e-06,
+      "loss": 0.5264,
+      "step": 623
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3474965006576283,
+      "learning_rate": 9.954688123765856e-06,
+      "loss": 0.526,
+      "step": 624
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2449115807143245,
+      "learning_rate": 9.954437817555122e-06,
+      "loss": 0.8155,
+      "step": 625
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.605339464482112,
+      "learning_rate": 9.95418682505652e-06,
+      "loss": 0.4435,
+      "step": 626
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.029216592351817,
+      "learning_rate": 9.953935146304817e-06,
+      "loss": 0.5129,
+      "step": 627
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.989176654518757,
+      "learning_rate": 9.953682781334876e-06,
+      "loss": 0.8205,
+      "step": 628
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7390455433770013,
+      "learning_rate": 9.953429730181653e-06,
+      "loss": 0.4952,
+      "step": 629
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0630352549450754,
+      "learning_rate": 9.953175992880204e-06,
+      "loss": 0.5644,
+      "step": 630
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8098763223905001,
+      "learning_rate": 9.952921569465675e-06,
+      "loss": 0.5738,
+      "step": 631
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.134643882066886,
+      "learning_rate": 9.952666459973304e-06,
+      "loss": 0.4949,
+      "step": 632
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2757808428129422,
+      "learning_rate": 9.952410664438436e-06,
+      "loss": 0.6466,
+      "step": 633
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8313078306214612,
+      "learning_rate": 9.952154182896499e-06,
+      "loss": 0.5182,
+      "step": 634
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1764358137619535,
+      "learning_rate": 9.951897015383023e-06,
+      "loss": 0.5172,
+      "step": 635
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.325567238496445,
+      "learning_rate": 9.951639161933631e-06,
+      "loss": 0.4904,
+      "step": 636
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7906520515697446,
+      "learning_rate": 9.951380622584039e-06,
+      "loss": 0.5627,
+      "step": 637
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9679335448695763,
+      "learning_rate": 9.95112139737006e-06,
+      "loss": 0.5486,
+      "step": 638
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2632138637419326,
+      "learning_rate": 9.950861486327604e-06,
+      "loss": 0.5282,
+      "step": 639
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.454021479588278,
+      "learning_rate": 9.950600889492672e-06,
+      "loss": 0.5918,
+      "step": 640
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5184412986496905,
+      "learning_rate": 9.950339606901362e-06,
+      "loss": 0.5734,
+      "step": 641
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7889424889705194,
+      "learning_rate": 9.950077638589867e-06,
+      "loss": 0.5436,
+      "step": 642
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8935818395296444,
+      "learning_rate": 9.949814984594475e-06,
+      "loss": 0.5078,
+      "step": 643
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4126857663649313,
+      "learning_rate": 9.949551644951569e-06,
+      "loss": 0.5145,
+      "step": 644
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2060769275218712,
+      "learning_rate": 9.949287619697625e-06,
+      "loss": 0.5834,
+      "step": 645
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0914569325619548,
+      "learning_rate": 9.94902290886922e-06,
+      "loss": 0.5225,
+      "step": 646
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.871294008503719,
+      "learning_rate": 9.948757512503015e-06,
+      "loss": 0.6368,
+      "step": 647
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.019332062064822,
+      "learning_rate": 9.948491430635779e-06,
+      "loss": 0.7444,
+      "step": 648
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9714214625326023,
+      "learning_rate": 9.948224663304367e-06,
+      "loss": 0.5102,
+      "step": 649
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5546874952338325,
+      "learning_rate": 9.94795721054573e-06,
+      "loss": 0.5388,
+      "step": 650
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.171739076400427,
+      "learning_rate": 9.94768907239692e-06,
+      "loss": 0.5669,
+      "step": 651
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.6071706335259175,
+      "learning_rate": 9.947420248895077e-06,
+      "loss": 0.6496,
+      "step": 652
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.045292602663791,
+      "learning_rate": 9.947150740077436e-06,
+      "loss": 0.5385,
+      "step": 653
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7594494563789622,
+      "learning_rate": 9.94688054598133e-06,
+      "loss": 0.542,
+      "step": 654
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0979323485136585,
+      "learning_rate": 9.94660966664419e-06,
+      "loss": 0.5031,
+      "step": 655
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.236019032635491,
+      "learning_rate": 9.946338102103536e-06,
+      "loss": 0.6364,
+      "step": 656
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.971355208204479,
+      "learning_rate": 9.946065852396984e-06,
+      "loss": 0.5337,
+      "step": 657
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.578291872063506,
+      "learning_rate": 9.945792917562245e-06,
+      "loss": 0.5368,
+      "step": 658
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0568071478387573,
+      "learning_rate": 9.94551929763713e-06,
+      "loss": 0.5566,
+      "step": 659
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9567326825421567,
+      "learning_rate": 9.945244992659539e-06,
+      "loss": 0.5507,
+      "step": 660
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0904525357199466,
+      "learning_rate": 9.944970002667466e-06,
+      "loss": 0.7381,
+      "step": 661
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.733509053147223,
+      "learning_rate": 9.944694327699007e-06,
+      "loss": 0.5748,
+      "step": 662
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.556355348898383,
+      "learning_rate": 9.944417967792343e-06,
+      "loss": 0.523,
+      "step": 663
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1214029699130985,
+      "learning_rate": 9.944140922985761e-06,
+      "loss": 0.4548,
+      "step": 664
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.55978427339231,
+      "learning_rate": 9.943863193317635e-06,
+      "loss": 0.5298,
+      "step": 665
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2936620226210076,
+      "learning_rate": 9.943584778826434e-06,
+      "loss": 0.5022,
+      "step": 666
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.20556628878954,
+      "learning_rate": 9.943305679550727e-06,
+      "loss": 0.5552,
+      "step": 667
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9831012442096033,
+      "learning_rate": 9.943025895529174e-06,
+      "loss": 0.7514,
+      "step": 668
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.180583779045343,
+      "learning_rate": 9.942745426800529e-06,
+      "loss": 0.5527,
+      "step": 669
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2616320787242095,
+      "learning_rate": 9.942464273403643e-06,
+      "loss": 0.5243,
+      "step": 670
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8321299529506032,
+      "learning_rate": 9.942182435377463e-06,
+      "loss": 0.5327,
+      "step": 671
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0128739587228384,
+      "learning_rate": 9.941899912761028e-06,
+      "loss": 0.4551,
+      "step": 672
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9881347246035441,
+      "learning_rate": 9.941616705593473e-06,
+      "loss": 0.4875,
+      "step": 673
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1229868156106777,
+      "learning_rate": 9.941332813914027e-06,
+      "loss": 0.4959,
+      "step": 674
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4462541819488464,
+      "learning_rate": 9.941048237762016e-06,
+      "loss": 0.604,
+      "step": 675
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0822584649130524,
+      "learning_rate": 9.94076297717686e-06,
+      "loss": 0.4954,
+      "step": 676
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9270875399485574,
+      "learning_rate": 9.94047703219807e-06,
+      "loss": 0.4783,
+      "step": 677
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9277392122471173,
+      "learning_rate": 9.94019040286526e-06,
+      "loss": 0.5306,
+      "step": 678
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.098715895988353,
+      "learning_rate": 9.939903089218129e-06,
+      "loss": 0.5976,
+      "step": 679
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0620666409701256,
+      "learning_rate": 9.939615091296479e-06,
+      "loss": 0.8239,
+      "step": 680
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9029685346928389,
+      "learning_rate": 9.939326409140201e-06,
+      "loss": 0.4987,
+      "step": 681
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7605793327274526,
+      "learning_rate": 9.939037042789284e-06,
+      "loss": 0.5361,
+      "step": 682
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8372909946594722,
+      "learning_rate": 9.938746992283812e-06,
+      "loss": 0.465,
+      "step": 683
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6081343759820523,
+      "learning_rate": 9.938456257663963e-06,
+      "loss": 0.4518,
+      "step": 684
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.126485581946189,
+      "learning_rate": 9.938164838970007e-06,
+      "loss": 0.4543,
+      "step": 685
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9428120973747975,
+      "learning_rate": 9.937872736242314e-06,
+      "loss": 0.5348,
+      "step": 686
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0163273214225184,
+      "learning_rate": 9.937579949521342e-06,
+      "loss": 0.5541,
+      "step": 687
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.057377891791151,
+      "learning_rate": 9.937286478847655e-06,
+      "loss": 0.5675,
+      "step": 688
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2465249306335044,
+      "learning_rate": 9.936992324261898e-06,
+      "loss": 0.5981,
+      "step": 689
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9262038385341438,
+      "learning_rate": 9.936697485804818e-06,
+      "loss": 0.5529,
+      "step": 690
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0024051993436713,
+      "learning_rate": 9.93640196351726e-06,
+      "loss": 0.6581,
+      "step": 691
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.115884017243316,
+      "learning_rate": 9.936105757440155e-06,
+      "loss": 0.4934,
+      "step": 692
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.398871482458367,
+      "learning_rate": 9.935808867614536e-06,
+      "loss": 0.5358,
+      "step": 693
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.007300873595809,
+      "learning_rate": 9.935511294081528e-06,
+      "loss": 0.6184,
+      "step": 694
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8477158987942803,
+      "learning_rate": 9.93521303688235e-06,
+      "loss": 0.6032,
+      "step": 695
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5343449039322081,
+      "learning_rate": 9.934914096058317e-06,
+      "loss": 0.5647,
+      "step": 696
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7804778680838254,
+      "learning_rate": 9.934614471650838e-06,
+      "loss": 0.5273,
+      "step": 697
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.172940278642374,
+      "learning_rate": 9.934314163701417e-06,
+      "loss": 0.6595,
+      "step": 698
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.342133488016177,
+      "learning_rate": 9.934013172251654e-06,
+      "loss": 0.5122,
+      "step": 699
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8481879425363652,
+      "learning_rate": 9.93371149734324e-06,
+      "loss": 0.546,
+      "step": 700
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8531759546579984,
+      "learning_rate": 9.933409139017963e-06,
+      "loss": 0.5769,
+      "step": 701
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9451313212524495,
+      "learning_rate": 9.933106097317707e-06,
+      "loss": 0.612,
+      "step": 702
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.55230137621701,
+      "learning_rate": 9.93280237228445e-06,
+      "loss": 0.5932,
+      "step": 703
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.044791314072983,
+      "learning_rate": 9.93249796396026e-06,
+      "loss": 0.5762,
+      "step": 704
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8104729706868379,
+      "learning_rate": 9.932192872387309e-06,
+      "loss": 0.5768,
+      "step": 705
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.995672554932377,
+      "learning_rate": 9.931887097607857e-06,
+      "loss": 0.6117,
+      "step": 706
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1278005357159704,
+      "learning_rate": 9.931580639664256e-06,
+      "loss": 0.4826,
+      "step": 707
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1791385156669003,
+      "learning_rate": 9.931273498598958e-06,
+      "loss": 0.4879,
+      "step": 708
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3206698093850275,
+      "learning_rate": 9.930965674454512e-06,
+      "loss": 0.5934,
+      "step": 709
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2777966100747844,
+      "learning_rate": 9.930657167273552e-06,
+      "loss": 0.4988,
+      "step": 710
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8826281627701882,
+      "learning_rate": 9.930347977098818e-06,
+      "loss": 0.5451,
+      "step": 711
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.226633155844569,
+      "learning_rate": 9.930038103973134e-06,
+      "loss": 0.6042,
+      "step": 712
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1601930532455857,
+      "learning_rate": 9.929727547939427e-06,
+      "loss": 0.4846,
+      "step": 713
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8442303257123716,
+      "learning_rate": 9.929416309040713e-06,
+      "loss": 0.4684,
+      "step": 714
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.063520633778046,
+      "learning_rate": 9.929104387320107e-06,
+      "loss": 0.5203,
+      "step": 715
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7980427447754983,
+      "learning_rate": 9.928791782820814e-06,
+      "loss": 0.4908,
+      "step": 716
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.512675798750525,
+      "learning_rate": 9.928478495586136e-06,
+      "loss": 0.5003,
+      "step": 717
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.173303590706775,
+      "learning_rate": 9.928164525659471e-06,
+      "loss": 0.6356,
+      "step": 718
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1788977791717667,
+      "learning_rate": 9.92784987308431e-06,
+      "loss": 0.6497,
+      "step": 719
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.826391375036422,
+      "learning_rate": 9.92753453790424e-06,
+      "loss": 0.4982,
+      "step": 720
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.179762023739226,
+      "learning_rate": 9.927218520162936e-06,
+      "loss": 0.5629,
+      "step": 721
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1048496410133914,
+      "learning_rate": 9.926901819904179e-06,
+      "loss": 0.594,
+      "step": 722
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.660875620099305,
+      "learning_rate": 9.926584437171833e-06,
+      "loss": 0.597,
+      "step": 723
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.168769712974328,
+      "learning_rate": 9.926266372009864e-06,
+      "loss": 0.5471,
+      "step": 724
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3994913270534344,
+      "learning_rate": 9.925947624462331e-06,
+      "loss": 0.6612,
+      "step": 725
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8698950319088523,
+      "learning_rate": 9.925628194573387e-06,
+      "loss": 0.43,
+      "step": 726
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8823893842736936,
+      "learning_rate": 9.925308082387278e-06,
+      "loss": 0.6223,
+      "step": 727
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.025879052142635,
+      "learning_rate": 9.924987287948347e-06,
+      "loss": 0.6342,
+      "step": 728
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2683153638883216,
+      "learning_rate": 9.92466581130103e-06,
+      "loss": 0.5773,
+      "step": 729
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8839728083943041,
+      "learning_rate": 9.924343652489856e-06,
+      "loss": 0.574,
+      "step": 730
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8674703703850188,
+      "learning_rate": 9.924020811559455e-06,
+      "loss": 0.5426,
+      "step": 731
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8251327391842784,
+      "learning_rate": 9.923697288554541e-06,
+      "loss": 0.491,
+      "step": 732
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3171599450221123,
+      "learning_rate": 9.923373083519932e-06,
+      "loss": 0.5069,
+      "step": 733
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8393428399702203,
+      "learning_rate": 9.923048196500537e-06,
+      "loss": 0.5713,
+      "step": 734
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4959149153027789,
+      "learning_rate": 9.92272262754136e-06,
+      "loss": 0.4881,
+      "step": 735
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.384836534371826,
+      "learning_rate": 9.922396376687496e-06,
+      "loss": 0.5902,
+      "step": 736
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.078912666958259,
+      "learning_rate": 9.922069443984137e-06,
+      "loss": 0.5072,
+      "step": 737
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.230468392380398,
+      "learning_rate": 9.921741829476574e-06,
+      "loss": 0.6295,
+      "step": 738
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.111905315128315,
+      "learning_rate": 9.921413533210183e-06,
+      "loss": 0.5493,
+      "step": 739
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9886021343673919,
+      "learning_rate": 9.921084555230443e-06,
+      "loss": 0.5414,
+      "step": 740
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.35063839581549,
+      "learning_rate": 9.920754895582923e-06,
+      "loss": 0.5427,
+      "step": 741
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6884971089657492,
+      "learning_rate": 9.920424554313287e-06,
+      "loss": 0.5673,
+      "step": 742
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2040526045487945,
+      "learning_rate": 9.920093531467292e-06,
+      "loss": 0.6176,
+      "step": 743
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.830093499337659,
+      "learning_rate": 9.919761827090794e-06,
+      "loss": 0.5724,
+      "step": 744
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.73956602881056,
+      "learning_rate": 9.919429441229741e-06,
+      "loss": 0.6113,
+      "step": 745
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6224037099847817,
+      "learning_rate": 9.919096373930173e-06,
+      "loss": 0.5316,
+      "step": 746
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7836489274995084,
+      "learning_rate": 9.918762625238227e-06,
+      "loss": 0.6392,
+      "step": 747
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8953826140673886,
+      "learning_rate": 9.918428195200137e-06,
+      "loss": 0.4769,
+      "step": 748
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0202531597217592,
+      "learning_rate": 9.918093083862221e-06,
+      "loss": 0.5416,
+      "step": 749
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6618625542023637,
+      "learning_rate": 9.917757291270906e-06,
+      "loss": 0.5019,
+      "step": 750
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.402843337035946,
+      "learning_rate": 9.917420817472701e-06,
+      "loss": 0.6129,
+      "step": 751
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1313367286528635,
+      "learning_rate": 9.917083662514218e-06,
+      "loss": 0.599,
+      "step": 752
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8259048180525739,
+      "learning_rate": 9.916745826442155e-06,
+      "loss": 0.5103,
+      "step": 753
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8131221490423983,
+      "learning_rate": 9.916407309303315e-06,
+      "loss": 0.5025,
+      "step": 754
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0687164677680507,
+      "learning_rate": 9.916068111144584e-06,
+      "loss": 0.5791,
+      "step": 755
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0056526659623533,
+      "learning_rate": 9.915728232012948e-06,
+      "loss": 0.5507,
+      "step": 756
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.204055682525048,
+      "learning_rate": 9.915387671955492e-06,
+      "loss": 0.4871,
+      "step": 757
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.163946230808035,
+      "learning_rate": 9.915046431019386e-06,
+      "loss": 0.607,
+      "step": 758
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8721184287224315,
+      "learning_rate": 9.9147045092519e-06,
+      "loss": 0.5558,
+      "step": 759
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0394258219841412,
+      "learning_rate": 9.914361906700395e-06,
+      "loss": 0.5533,
+      "step": 760
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2110918203526078,
+      "learning_rate": 9.914018623412332e-06,
+      "loss": 0.5598,
+      "step": 761
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.975737341190045,
+      "learning_rate": 9.91367465943526e-06,
+      "loss": 0.5323,
+      "step": 762
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.914946697474532,
+      "learning_rate": 9.913330014816825e-06,
+      "loss": 0.48,
+      "step": 763
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3215453586939656,
+      "learning_rate": 9.912984689604767e-06,
+      "loss": 0.591,
+      "step": 764
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.71369353912503,
+      "learning_rate": 9.91263868384692e-06,
+      "loss": 0.6416,
+      "step": 765
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8021380576463637,
+      "learning_rate": 9.912291997591214e-06,
+      "loss": 0.5271,
+      "step": 766
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.501968820056032,
+      "learning_rate": 9.911944630885673e-06,
+      "loss": 0.5592,
+      "step": 767
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9508127903138144,
+      "learning_rate": 9.91159658377841e-06,
+      "loss": 0.5236,
+      "step": 768
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.9344108037566103,
+      "learning_rate": 9.91124785631764e-06,
+      "loss": 0.5318,
+      "step": 769
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8071322055126993,
+      "learning_rate": 9.910898448551667e-06,
+      "loss": 0.5318,
+      "step": 770
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9819159417999153,
+      "learning_rate": 9.910548360528894e-06,
+      "loss": 0.6113,
+      "step": 771
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8651408487871985,
+      "learning_rate": 9.91019759229781e-06,
+      "loss": 0.6614,
+      "step": 772
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8931814155451214,
+      "learning_rate": 9.909846143907007e-06,
+      "loss": 0.453,
+      "step": 773
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.250629188912134,
+      "learning_rate": 9.909494015405165e-06,
+      "loss": 0.5246,
+      "step": 774
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.356545651971741,
+      "learning_rate": 9.909141206841063e-06,
+      "loss": 0.5608,
+      "step": 775
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6067207075793863,
+      "learning_rate": 9.908787718263573e-06,
+      "loss": 0.5624,
+      "step": 776
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5715543892755393,
+      "learning_rate": 9.908433549721657e-06,
+      "loss": 0.4727,
+      "step": 777
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.814019611720708,
+      "learning_rate": 9.908078701264377e-06,
+      "loss": 0.5259,
+      "step": 778
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9615486322382332,
+      "learning_rate": 9.907723172940885e-06,
+      "loss": 0.4919,
+      "step": 779
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3423456682474426,
+      "learning_rate": 9.907366964800429e-06,
+      "loss": 0.6536,
+      "step": 780
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2794434458657546,
+      "learning_rate": 9.90701007689235e-06,
+      "loss": 0.5372,
+      "step": 781
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4870751556327084,
+      "learning_rate": 9.906652509266086e-06,
+      "loss": 0.5785,
+      "step": 782
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.44398977849809,
+      "learning_rate": 9.906294261971167e-06,
+      "loss": 0.55,
+      "step": 783
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.688777216456219,
+      "learning_rate": 9.905935335057215e-06,
+      "loss": 0.5801,
+      "step": 784
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1364305601121654,
+      "learning_rate": 9.905575728573952e-06,
+      "loss": 0.5739,
+      "step": 785
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5522044750988018,
+      "learning_rate": 9.905215442571189e-06,
+      "loss": 0.5388,
+      "step": 786
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.399659223136487,
+      "learning_rate": 9.904854477098829e-06,
+      "loss": 0.5485,
+      "step": 787
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.221108025977278,
+      "learning_rate": 9.904492832206879e-06,
+      "loss": 0.4584,
+      "step": 788
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1366722251738843,
+      "learning_rate": 9.90413050794543e-06,
+      "loss": 0.5765,
+      "step": 789
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.178803801307301,
+      "learning_rate": 9.903767504364674e-06,
+      "loss": 0.4955,
+      "step": 790
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4051784151081432,
+      "learning_rate": 9.903403821514893e-06,
+      "loss": 0.5564,
+      "step": 791
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9680915771905254,
+      "learning_rate": 9.903039459446463e-06,
+      "loss": 0.5101,
+      "step": 792
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.04411908521223,
+      "learning_rate": 9.902674418209856e-06,
+      "loss": 0.5609,
+      "step": 793
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.255772234723594,
+      "learning_rate": 9.902308697855638e-06,
+      "loss": 0.6796,
+      "step": 794
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3993617801407505,
+      "learning_rate": 9.901942298434469e-06,
+      "loss": 0.5999,
+      "step": 795
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3125065096164716,
+      "learning_rate": 9.901575219997101e-06,
+      "loss": 0.5687,
+      "step": 796
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6085419162717534,
+      "learning_rate": 9.901207462594383e-06,
+      "loss": 0.5943,
+      "step": 797
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0920400823301573,
+      "learning_rate": 9.900839026277256e-06,
+      "loss": 0.6622,
+      "step": 798
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1861997520184646,
+      "learning_rate": 9.900469911096756e-06,
+      "loss": 0.4737,
+      "step": 799
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.375693675651978,
+      "learning_rate": 9.900100117104011e-06,
+      "loss": 0.602,
+      "step": 800
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2502230168747253,
+      "learning_rate": 9.899729644350249e-06,
+      "loss": 0.5393,
+      "step": 801
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.114481174134644,
+      "learning_rate": 9.899358492886784e-06,
+      "loss": 0.4948,
+      "step": 802
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9471575108186827,
+      "learning_rate": 9.898986662765029e-06,
+      "loss": 0.5589,
+      "step": 803
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.162272939817687,
+      "learning_rate": 9.898614154036491e-06,
+      "loss": 0.5211,
+      "step": 804
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.04777132394763,
+      "learning_rate": 9.898240966752768e-06,
+      "loss": 0.4473,
+      "step": 805
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4256387684862872,
+      "learning_rate": 9.897867100965555e-06,
+      "loss": 0.6285,
+      "step": 806
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9545077404605793,
+      "learning_rate": 9.89749255672664e-06,
+      "loss": 0.5509,
+      "step": 807
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.930443187041612,
+      "learning_rate": 9.897117334087904e-06,
+      "loss": 0.5743,
+      "step": 808
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8622711635673348,
+      "learning_rate": 9.896741433101322e-06,
+      "loss": 0.5263,
+      "step": 809
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.104415358110688,
+      "learning_rate": 9.896364853818967e-06,
+      "loss": 0.5606,
+      "step": 810
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.9804163392740417,
+      "learning_rate": 9.895987596293e-06,
+      "loss": 0.5242,
+      "step": 811
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1821664235484057,
+      "learning_rate": 9.895609660575678e-06,
+      "loss": 0.5937,
+      "step": 812
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.047322065001753,
+      "learning_rate": 9.895231046719354e-06,
+      "loss": 0.6513,
+      "step": 813
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5285621751661074,
+      "learning_rate": 9.894851754776473e-06,
+      "loss": 0.5082,
+      "step": 814
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6050560960277096,
+      "learning_rate": 9.894471784799575e-06,
+      "loss": 0.5519,
+      "step": 815
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4468718588917024,
+      "learning_rate": 9.894091136841294e-06,
+      "loss": 0.5595,
+      "step": 816
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.263929432390896,
+      "learning_rate": 9.893709810954354e-06,
+      "loss": 0.6088,
+      "step": 817
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5060644876972598,
+      "learning_rate": 9.893327807191581e-06,
+      "loss": 0.5988,
+      "step": 818
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.212660303552677,
+      "learning_rate": 9.892945125605888e-06,
+      "loss": 0.5156,
+      "step": 819
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.629945351422265,
+      "learning_rate": 9.892561766250284e-06,
+      "loss": 0.5808,
+      "step": 820
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1968672159859475,
+      "learning_rate": 9.89217772917787e-06,
+      "loss": 0.6845,
+      "step": 821
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2512796937148254,
+      "learning_rate": 9.891793014441844e-06,
+      "loss": 0.5687,
+      "step": 822
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6830955802885534,
+      "learning_rate": 9.891407622095498e-06,
+      "loss": 0.6377,
+      "step": 823
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.894746516916812,
+      "learning_rate": 9.891021552192215e-06,
+      "loss": 0.6678,
+      "step": 824
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.971413953696215,
+      "learning_rate": 9.890634804785473e-06,
+      "loss": 0.5345,
+      "step": 825
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9816815986119727,
+      "learning_rate": 9.890247379928845e-06,
+      "loss": 0.604,
+      "step": 826
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8958919628301971,
+      "learning_rate": 9.889859277675999e-06,
+      "loss": 0.5321,
+      "step": 827
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.368763408783505,
+      "learning_rate": 9.889470498080691e-06,
+      "loss": 0.4718,
+      "step": 828
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9264470940603833,
+      "learning_rate": 9.889081041196777e-06,
+      "loss": 0.4672,
+      "step": 829
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.337750119077244,
+      "learning_rate": 9.888690907078205e-06,
+      "loss": 0.5995,
+      "step": 830
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.134855133374358,
+      "learning_rate": 9.888300095779013e-06,
+      "loss": 0.6687,
+      "step": 831
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5573954400696732,
+      "learning_rate": 9.887908607353341e-06,
+      "loss": 0.4879,
+      "step": 832
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9280029362037718,
+      "learning_rate": 9.887516441855413e-06,
+      "loss": 0.4589,
+      "step": 833
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.14199820530583,
+      "learning_rate": 9.887123599339555e-06,
+      "loss": 0.5969,
+      "step": 834
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.845683916256456,
+      "learning_rate": 9.886730079860182e-06,
+      "loss": 0.5413,
+      "step": 835
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6804140424519625,
+      "learning_rate": 9.886335883471804e-06,
+      "loss": 0.5706,
+      "step": 836
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7436964357422922,
+      "learning_rate": 9.885941010229028e-06,
+      "loss": 0.5755,
+      "step": 837
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.030870656340766,
+      "learning_rate": 9.885545460186548e-06,
+      "loss": 0.4875,
+      "step": 838
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.275484580724343,
+      "learning_rate": 9.885149233399158e-06,
+      "loss": 0.4827,
+      "step": 839
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.083240178100821,
+      "learning_rate": 9.884752329921743e-06,
+      "loss": 0.5062,
+      "step": 840
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1288713566535558,
+      "learning_rate": 9.88435474980928e-06,
+      "loss": 0.5906,
+      "step": 841
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0204345005821027,
+      "learning_rate": 9.883956493116842e-06,
+      "loss": 0.5751,
+      "step": 842
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6466305426953567,
+      "learning_rate": 9.883557559899599e-06,
+      "loss": 0.5665,
+      "step": 843
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8533477132542155,
+      "learning_rate": 9.883157950212807e-06,
+      "loss": 0.4665,
+      "step": 844
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4708687595532766,
+      "learning_rate": 9.882757664111822e-06,
+      "loss": 0.5441,
+      "step": 845
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2994338199228217,
+      "learning_rate": 9.882356701652092e-06,
+      "loss": 0.4874,
+      "step": 846
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9579351519231976,
+      "learning_rate": 9.881955062889155e-06,
+      "loss": 0.5224,
+      "step": 847
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9931887200717259,
+      "learning_rate": 9.88155274787865e-06,
+      "loss": 0.735,
+      "step": 848
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3310893594085265,
+      "learning_rate": 9.881149756676302e-06,
+      "loss": 0.5277,
+      "step": 849
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9117576941358045,
+      "learning_rate": 9.880746089337938e-06,
+      "loss": 0.6142,
+      "step": 850
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.62749544087934,
+      "learning_rate": 9.88034174591947e-06,
+      "loss": 0.6049,
+      "step": 851
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6992686547562001,
+      "learning_rate": 9.879936726476908e-06,
+      "loss": 0.5118,
+      "step": 852
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8881631794332134,
+      "learning_rate": 9.879531031066355e-06,
+      "loss": 0.5697,
+      "step": 853
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5117542962535326,
+      "learning_rate": 9.87912465974401e-06,
+      "loss": 0.7017,
+      "step": 854
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.772221473871344,
+      "learning_rate": 9.878717612566163e-06,
+      "loss": 0.4996,
+      "step": 855
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7693683763283217,
+      "learning_rate": 9.878309889589197e-06,
+      "loss": 0.4482,
+      "step": 856
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.083813715152233,
+      "learning_rate": 9.87790149086959e-06,
+      "loss": 0.5495,
+      "step": 857
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0428863186336743,
+      "learning_rate": 9.877492416463913e-06,
+      "loss": 0.5361,
+      "step": 858
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.167621007070851,
+      "learning_rate": 9.87708266642883e-06,
+      "loss": 0.446,
+      "step": 859
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0051783378371573,
+      "learning_rate": 9.876672240821103e-06,
+      "loss": 0.6286,
+      "step": 860
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7316501723126323,
+      "learning_rate": 9.87626113969758e-06,
+      "loss": 0.5663,
+      "step": 861
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9826503785060554,
+      "learning_rate": 9.87584936311521e-06,
+      "loss": 0.4846,
+      "step": 862
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.88056999337644,
+      "learning_rate": 9.87543691113103e-06,
+      "loss": 0.5047,
+      "step": 863
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.904497096726396,
+      "learning_rate": 9.875023783802174e-06,
+      "loss": 0.6063,
+      "step": 864
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7947849651683292,
+      "learning_rate": 9.874609981185868e-06,
+      "loss": 0.5665,
+      "step": 865
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.274184819131868,
+      "learning_rate": 9.87419550333943e-06,
+      "loss": 0.519,
+      "step": 866
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.176736515231948,
+      "learning_rate": 9.873780350320276e-06,
+      "loss": 0.5503,
+      "step": 867
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.082572846073231,
+      "learning_rate": 9.873364522185913e-06,
+      "loss": 0.5768,
+      "step": 868
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0723792133072667,
+      "learning_rate": 9.87294801899394e-06,
+      "loss": 0.5119,
+      "step": 869
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9594449988199611,
+      "learning_rate": 9.872530840802052e-06,
+      "loss": 0.6024,
+      "step": 870
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6720242689909397,
+      "learning_rate": 9.872112987668034e-06,
+      "loss": 0.5155,
+      "step": 871
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5745320908660818,
+      "learning_rate": 9.87169445964977e-06,
+      "loss": 0.4948,
+      "step": 872
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3699734020127146,
+      "learning_rate": 9.871275256805234e-06,
+      "loss": 0.5871,
+      "step": 873
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.447242454615751,
+      "learning_rate": 9.870855379192492e-06,
+      "loss": 0.5256,
+      "step": 874
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9619054723644462,
+      "learning_rate": 9.870434826869707e-06,
+      "loss": 0.4347,
+      "step": 875
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.413207594444614,
+      "learning_rate": 9.870013599895135e-06,
+      "loss": 0.5814,
+      "step": 876
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.88488074195696,
+      "learning_rate": 9.86959169832712e-06,
+      "loss": 0.6093,
+      "step": 877
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1020798272220043,
+      "learning_rate": 9.869169122224107e-06,
+      "loss": 0.7854,
+      "step": 878
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0332124961262945,
+      "learning_rate": 9.86874587164463e-06,
+      "loss": 0.6893,
+      "step": 879
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0329005627045382,
+      "learning_rate": 9.86832194664732e-06,
+      "loss": 0.5611,
+      "step": 880
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6668858806494669,
+      "learning_rate": 9.867897347290895e-06,
+      "loss": 0.4961,
+      "step": 881
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.96260381361014,
+      "learning_rate": 9.867472073634175e-06,
+      "loss": 0.5655,
+      "step": 882
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0521870674033225,
+      "learning_rate": 9.867046125736066e-06,
+      "loss": 0.6272,
+      "step": 883
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5514603181433215,
+      "learning_rate": 9.866619503655569e-06,
+      "loss": 0.4832,
+      "step": 884
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0541862830110125,
+      "learning_rate": 9.866192207451781e-06,
+      "loss": 0.4537,
+      "step": 885
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1940097597358115,
+      "learning_rate": 9.865764237183894e-06,
+      "loss": 0.5032,
+      "step": 886
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.081322935477042,
+      "learning_rate": 9.865335592911185e-06,
+      "loss": 0.5184,
+      "step": 887
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.623336425535113,
+      "learning_rate": 9.864906274693033e-06,
+      "loss": 0.6915,
+      "step": 888
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.859948260015348,
+      "learning_rate": 9.864476282588908e-06,
+      "loss": 0.5413,
+      "step": 889
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.876717891088719,
+      "learning_rate": 9.86404561665837e-06,
+      "loss": 0.4841,
+      "step": 890
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.416190909696552,
+      "learning_rate": 9.863614276961076e-06,
+      "loss": 0.5213,
+      "step": 891
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8779983877324762,
+      "learning_rate": 9.863182263556775e-06,
+      "loss": 0.5661,
+      "step": 892
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.93331573223808,
+      "learning_rate": 9.862749576505307e-06,
+      "loss": 0.6173,
+      "step": 893
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.040800357035209,
+      "learning_rate": 9.862316215866612e-06,
+      "loss": 0.6335,
+      "step": 894
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.175378755291958,
+      "learning_rate": 9.861882181700716e-06,
+      "loss": 0.599,
+      "step": 895
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.484611585682195,
+      "learning_rate": 9.861447474067743e-06,
+      "loss": 0.5247,
+      "step": 896
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7951302581855397,
+      "learning_rate": 9.861012093027906e-06,
+      "loss": 0.5696,
+      "step": 897
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5526691120467024,
+      "learning_rate": 9.860576038641519e-06,
+      "loss": 0.4391,
+      "step": 898
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6546786440469714,
+      "learning_rate": 9.860139310968977e-06,
+      "loss": 0.5454,
+      "step": 899
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.202884123716279,
+      "learning_rate": 9.859701910070782e-06,
+      "loss": 0.5243,
+      "step": 900
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.139351235017453,
+      "learning_rate": 9.85926383600752e-06,
+      "loss": 0.5235,
+      "step": 901
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3337599317768776,
+      "learning_rate": 9.858825088839875e-06,
+      "loss": 0.6117,
+      "step": 902
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.983665523016773,
+      "learning_rate": 9.858385668628617e-06,
+      "loss": 0.61,
+      "step": 903
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9826605737745353,
+      "learning_rate": 9.85794557543462e-06,
+      "loss": 0.5841,
+      "step": 904
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0793920356780644,
+      "learning_rate": 9.85750480931884e-06,
+      "loss": 0.6438,
+      "step": 905
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2568681660668313,
+      "learning_rate": 9.857063370342338e-06,
+      "loss": 0.4774,
+      "step": 906
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8305427229825173,
+      "learning_rate": 9.856621258566259e-06,
+      "loss": 0.5997,
+      "step": 907
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.08893580885194,
+      "learning_rate": 9.856178474051845e-06,
+      "loss": 0.5366,
+      "step": 908
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.567135023441297,
+      "learning_rate": 9.855735016860428e-06,
+      "loss": 0.624,
+      "step": 909
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8916638741404452,
+      "learning_rate": 9.85529088705344e-06,
+      "loss": 0.4536,
+      "step": 910
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1411075703471507,
+      "learning_rate": 9.854846084692397e-06,
+      "loss": 0.5809,
+      "step": 911
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.357596381735398,
+      "learning_rate": 9.854400609838916e-06,
+      "loss": 0.5888,
+      "step": 912
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4936677321580134,
+      "learning_rate": 9.853954462554703e-06,
+      "loss": 0.49,
+      "step": 913
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.011769222956356,
+      "learning_rate": 9.853507642901558e-06,
+      "loss": 0.5093,
+      "step": 914
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9819385128415954,
+      "learning_rate": 9.853060150941377e-06,
+      "loss": 0.4335,
+      "step": 915
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.675833708963052,
+      "learning_rate": 9.852611986736144e-06,
+      "loss": 0.5649,
+      "step": 916
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.10717545681445,
+      "learning_rate": 9.852163150347937e-06,
+      "loss": 0.5694,
+      "step": 917
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0181411420473,
+      "learning_rate": 9.851713641838934e-06,
+      "loss": 0.57,
+      "step": 918
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7055665116202148,
+      "learning_rate": 9.851263461271394e-06,
+      "loss": 0.5162,
+      "step": 919
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7282290100619306,
+      "learning_rate": 9.850812608707683e-06,
+      "loss": 0.541,
+      "step": 920
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9128856440649002,
+      "learning_rate": 9.850361084210247e-06,
+      "loss": 0.5061,
+      "step": 921
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.061521210692046,
+      "learning_rate": 9.849908887841635e-06,
+      "loss": 0.4863,
+      "step": 922
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.259488530051225,
+      "learning_rate": 9.849456019664486e-06,
+      "loss": 0.6277,
+      "step": 923
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.794885586084605,
+      "learning_rate": 9.849002479741525e-06,
+      "loss": 0.4633,
+      "step": 924
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0804153275399906,
+      "learning_rate": 9.848548268135583e-06,
+      "loss": 0.5343,
+      "step": 925
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7718272690563932,
+      "learning_rate": 9.848093384909573e-06,
+      "loss": 0.5315,
+      "step": 926
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8418474426835068,
+      "learning_rate": 9.847637830126508e-06,
+      "loss": 0.6539,
+      "step": 927
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9505070614896556,
+      "learning_rate": 9.84718160384949e-06,
+      "loss": 0.4598,
+      "step": 928
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.506743969882724,
+      "learning_rate": 9.846724706141718e-06,
+      "loss": 0.5976,
+      "step": 929
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0031381787617617,
+      "learning_rate": 9.846267137066476e-06,
+      "loss": 0.7782,
+      "step": 930
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5846599054207,
+      "learning_rate": 9.845808896687152e-06,
+      "loss": 0.4449,
+      "step": 931
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6276921120644192,
+      "learning_rate": 9.845349985067218e-06,
+      "loss": 0.4,
+      "step": 932
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9670455745506532,
+      "learning_rate": 9.844890402270243e-06,
+      "loss": 0.5269,
+      "step": 933
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2858606166463047,
+      "learning_rate": 9.84443014835989e-06,
+      "loss": 0.4882,
+      "step": 934
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5439573276502268,
+      "learning_rate": 9.84396922339991e-06,
+      "loss": 0.4892,
+      "step": 935
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2016239119196155,
+      "learning_rate": 9.843507627454152e-06,
+      "loss": 0.6017,
+      "step": 936
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.431450848772123,
+      "learning_rate": 9.843045360586559e-06,
+      "loss": 0.5045,
+      "step": 937
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8870915149366394,
+      "learning_rate": 9.842582422861158e-06,
+      "loss": 0.4405,
+      "step": 938
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1532647667628426,
+      "learning_rate": 9.842118814342081e-06,
+      "loss": 0.6124,
+      "step": 939
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.92346935831187,
+      "learning_rate": 9.841654535093544e-06,
+      "loss": 0.5297,
+      "step": 940
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1723861706838217,
+      "learning_rate": 9.841189585179859e-06,
+      "loss": 0.4803,
+      "step": 941
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6258217938948032,
+      "learning_rate": 9.840723964665432e-06,
+      "loss": 0.5511,
+      "step": 942
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7738155453002094,
+      "learning_rate": 9.84025767361476e-06,
+      "loss": 0.532,
+      "step": 943
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0738336552303536,
+      "learning_rate": 9.839790712092431e-06,
+      "loss": 0.6309,
+      "step": 944
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.510841463333185,
+      "learning_rate": 9.839323080163134e-06,
+      "loss": 0.6706,
+      "step": 945
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4639620806369558,
+      "learning_rate": 9.838854777891639e-06,
+      "loss": 0.5444,
+      "step": 946
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1003446221040263,
+      "learning_rate": 9.83838580534282e-06,
+      "loss": 0.5484,
+      "step": 947
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.219206753720893,
+      "learning_rate": 9.837916162581638e-06,
+      "loss": 0.44,
+      "step": 948
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8707519973476032,
+      "learning_rate": 9.837445849673145e-06,
+      "loss": 0.5154,
+      "step": 949
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.948187874266642,
+      "learning_rate": 9.836974866682494e-06,
+      "loss": 0.5119,
+      "step": 950
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.914781098342308,
+      "learning_rate": 9.83650321367492e-06,
+      "loss": 0.441,
+      "step": 951
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6996017858315513,
+      "learning_rate": 9.83603089071576e-06,
+      "loss": 0.4772,
+      "step": 952
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3032078793750927,
+      "learning_rate": 9.835557897870435e-06,
+      "loss": 0.5559,
+      "step": 953
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9491132367918018,
+      "learning_rate": 9.835084235204471e-06,
+      "loss": 0.4787,
+      "step": 954
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.912878126767747,
+      "learning_rate": 9.834609902783477e-06,
+      "loss": 0.5808,
+      "step": 955
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7471564771229053,
+      "learning_rate": 9.834134900673153e-06,
+      "loss": 0.5492,
+      "step": 956
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 14.441560558354487,
+      "learning_rate": 9.833659228939302e-06,
+      "loss": 0.5247,
+      "step": 957
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5143158235820677,
+      "learning_rate": 9.833182887647811e-06,
+      "loss": 0.5236,
+      "step": 958
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.691288573466352,
+      "learning_rate": 9.832705876864665e-06,
+      "loss": 0.4817,
+      "step": 959
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7456129629728314,
+      "learning_rate": 9.832228196655938e-06,
+      "loss": 0.5525,
+      "step": 960
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8487467436169844,
+      "learning_rate": 9.831749847087798e-06,
+      "loss": 0.6213,
+      "step": 961
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.738038570015744,
+      "learning_rate": 9.831270828226505e-06,
+      "loss": 0.5483,
+      "step": 962
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0962130298462873,
+      "learning_rate": 9.830791140138414e-06,
+      "loss": 0.5434,
+      "step": 963
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9203518131355894,
+      "learning_rate": 9.830310782889972e-06,
+      "loss": 0.6674,
+      "step": 964
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7137178676939158,
+      "learning_rate": 9.829829756547715e-06,
+      "loss": 0.5118,
+      "step": 965
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6125222980536797,
+      "learning_rate": 9.829348061178278e-06,
+      "loss": 0.5765,
+      "step": 966
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9994414209720692,
+      "learning_rate": 9.828865696848384e-06,
+      "loss": 0.5677,
+      "step": 967
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.025442572970201,
+      "learning_rate": 9.828382663624849e-06,
+      "loss": 0.5682,
+      "step": 968
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.1175025761190795,
+      "learning_rate": 9.827898961574584e-06,
+      "loss": 0.5028,
+      "step": 969
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7298350849749924,
+      "learning_rate": 9.827414590764593e-06,
+      "loss": 0.5815,
+      "step": 970
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7746048319465775,
+      "learning_rate": 9.826929551261968e-06,
+      "loss": 0.5549,
+      "step": 971
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1380357139233985,
+      "learning_rate": 9.826443843133898e-06,
+      "loss": 0.6138,
+      "step": 972
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.164530910366711,
+      "learning_rate": 9.82595746644766e-06,
+      "loss": 0.5356,
+      "step": 973
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.273978627254315,
+      "learning_rate": 9.825470421270632e-06,
+      "loss": 0.5469,
+      "step": 974
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6984966506367596,
+      "learning_rate": 9.824982707670277e-06,
+      "loss": 0.5147,
+      "step": 975
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.299334232263952,
+      "learning_rate": 9.824494325714154e-06,
+      "loss": 0.5117,
+      "step": 976
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5743987992660804,
+      "learning_rate": 9.82400527546991e-06,
+      "loss": 0.591,
+      "step": 977
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8580755112203982,
+      "learning_rate": 9.823515557005293e-06,
+      "loss": 0.512,
+      "step": 978
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8638682917143672,
+      "learning_rate": 9.823025170388135e-06,
+      "loss": 0.547,
+      "step": 979
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1109071694765453,
+      "learning_rate": 9.822534115686367e-06,
+      "loss": 0.4884,
+      "step": 980
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.044077980664823,
+      "learning_rate": 9.822042392968007e-06,
+      "loss": 0.7641,
+      "step": 981
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8155165003466838,
+      "learning_rate": 9.82155000230117e-06,
+      "loss": 0.5415,
+      "step": 982
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9345341258804145,
+      "learning_rate": 9.821056943754064e-06,
+      "loss": 0.7727,
+      "step": 983
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9525081260293182,
+      "learning_rate": 9.820563217394985e-06,
+      "loss": 0.4924,
+      "step": 984
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.083293606210164,
+      "learning_rate": 9.820068823292323e-06,
+      "loss": 0.5863,
+      "step": 985
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7043308629573626,
+      "learning_rate": 9.819573761514562e-06,
+      "loss": 0.5528,
+      "step": 986
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.947890519378382,
+      "learning_rate": 9.819078032130278e-06,
+      "loss": 0.3827,
+      "step": 987
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0580015993585363,
+      "learning_rate": 9.818581635208141e-06,
+      "loss": 0.7479,
+      "step": 988
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9249410911869198,
+      "learning_rate": 9.81808457081691e-06,
+      "loss": 0.5544,
+      "step": 989
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5582906404529258,
+      "learning_rate": 9.817586839025439e-06,
+      "loss": 0.5231,
+      "step": 990
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0019333354053246,
+      "learning_rate": 9.817088439902673e-06,
+      "loss": 0.4995,
+      "step": 991
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9761655942962517,
+      "learning_rate": 9.81658937351765e-06,
+      "loss": 0.5132,
+      "step": 992
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3176701843037417,
+      "learning_rate": 9.816089639939503e-06,
+      "loss": 0.604,
+      "step": 993
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8930942080860391,
+      "learning_rate": 9.815589239237452e-06,
+      "loss": 0.5321,
+      "step": 994
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7072557581459284,
+      "learning_rate": 9.815088171480815e-06,
+      "loss": 0.5383,
+      "step": 995
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.67107596558961,
+      "learning_rate": 9.814586436738998e-06,
+      "loss": 0.641,
+      "step": 996
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1096878604045863,
+      "learning_rate": 9.8140840350815e-06,
+      "loss": 0.5115,
+      "step": 997
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6414556169990353,
+      "learning_rate": 9.813580966577916e-06,
+      "loss": 0.5595,
+      "step": 998
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9958978624521275,
+      "learning_rate": 9.813077231297931e-06,
+      "loss": 0.6052,
+      "step": 999
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.76854022219441,
+      "learning_rate": 9.812572829311322e-06,
+      "loss": 0.5108,
+      "step": 1000
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.941674310946483,
+      "learning_rate": 9.812067760687957e-06,
+      "loss": 0.689,
+      "step": 1001
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3063096635957803,
+      "learning_rate": 9.811562025497801e-06,
+      "loss": 0.5033,
+      "step": 1002
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2500743412745656,
+      "learning_rate": 9.811055623810906e-06,
+      "loss": 0.585,
+      "step": 1003
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5046944295941724,
+      "learning_rate": 9.810548555697419e-06,
+      "loss": 0.5725,
+      "step": 1004
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.353133760303956,
+      "learning_rate": 9.810040821227577e-06,
+      "loss": 0.4871,
+      "step": 1005
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4087881231531962,
+      "learning_rate": 9.809532420471716e-06,
+      "loss": 0.4698,
+      "step": 1006
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9266873696174018,
+      "learning_rate": 9.809023353500258e-06,
+      "loss": 0.4479,
+      "step": 1007
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5945729709139935,
+      "learning_rate": 9.808513620383715e-06,
+      "loss": 0.6815,
+      "step": 1008
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.22999088069897,
+      "learning_rate": 9.808003221192701e-06,
+      "loss": 0.4737,
+      "step": 1009
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.856188351218398,
+      "learning_rate": 9.807492155997913e-06,
+      "loss": 0.5685,
+      "step": 1010
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3487601762725494,
+      "learning_rate": 9.806980424870142e-06,
+      "loss": 0.541,
+      "step": 1011
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3261918867485076,
+      "learning_rate": 9.806468027880278e-06,
+      "loss": 0.3967,
+      "step": 1012
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.949313300590524,
+      "learning_rate": 9.805954965099294e-06,
+      "loss": 0.5587,
+      "step": 1013
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5211627589293955,
+      "learning_rate": 9.80544123659826e-06,
+      "loss": 0.498,
+      "step": 1014
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0690696886413193,
+      "learning_rate": 9.80492684244834e-06,
+      "loss": 0.5682,
+      "step": 1015
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8183912807707936,
+      "learning_rate": 9.804411782720786e-06,
+      "loss": 0.571,
+      "step": 1016
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8692643637832056,
+      "learning_rate": 9.803896057486942e-06,
+      "loss": 0.5453,
+      "step": 1017
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5892962792207372,
+      "learning_rate": 9.803379666818249e-06,
+      "loss": 0.5541,
+      "step": 1018
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6927202272783397,
+      "learning_rate": 9.802862610786237e-06,
+      "loss": 0.4732,
+      "step": 1019
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7437796497847078,
+      "learning_rate": 9.802344889462528e-06,
+      "loss": 0.4851,
+      "step": 1020
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8537865279881567,
+      "learning_rate": 9.801826502918836e-06,
+      "loss": 0.489,
+      "step": 1021
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0277338882267102,
+      "learning_rate": 9.80130745122697e-06,
+      "loss": 0.7973,
+      "step": 1022
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.031507196818382,
+      "learning_rate": 9.800787734458827e-06,
+      "loss": 0.5246,
+      "step": 1023
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7059599738915348,
+      "learning_rate": 9.800267352686398e-06,
+      "loss": 0.6705,
+      "step": 1024
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5229903092206543,
+      "learning_rate": 9.799746305981766e-06,
+      "loss": 0.4647,
+      "step": 1025
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.759909907868852,
+      "learning_rate": 9.799224594417109e-06,
+      "loss": 0.5765,
+      "step": 1026
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9096652099559779,
+      "learning_rate": 9.79870221806469e-06,
+      "loss": 0.731,
+      "step": 1027
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.903135673912431,
+      "learning_rate": 9.798179176996873e-06,
+      "loss": 0.6371,
+      "step": 1028
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8329565257170692,
+      "learning_rate": 9.797655471286106e-06,
+      "loss": 0.477,
+      "step": 1029
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6803635725336725,
+      "learning_rate": 9.797131101004935e-06,
+      "loss": 0.5654,
+      "step": 1030
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.717672175188184,
+      "learning_rate": 9.796606066225996e-06,
+      "loss": 0.5536,
+      "step": 1031
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2102161061776915,
+      "learning_rate": 9.796080367022013e-06,
+      "loss": 0.5547,
+      "step": 1032
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1385793732095264,
+      "learning_rate": 9.795554003465809e-06,
+      "loss": 0.5941,
+      "step": 1033
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0963068014442805,
+      "learning_rate": 9.795026975630295e-06,
+      "loss": 0.6031,
+      "step": 1034
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.369299885865623,
+      "learning_rate": 9.794499283588475e-06,
+      "loss": 0.5379,
+      "step": 1035
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.104055127260396,
+      "learning_rate": 9.793970927413446e-06,
+      "loss": 0.6338,
+      "step": 1036
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0858628883242756,
+      "learning_rate": 9.793441907178393e-06,
+      "loss": 0.6015,
+      "step": 1037
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.088834293040691,
+      "learning_rate": 9.792912222956597e-06,
+      "loss": 0.6317,
+      "step": 1038
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0129724876822115,
+      "learning_rate": 9.792381874821431e-06,
+      "loss": 0.4724,
+      "step": 1039
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7630642652722168,
+      "learning_rate": 9.791850862846358e-06,
+      "loss": 0.5239,
+      "step": 1040
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9889206355862572,
+      "learning_rate": 9.791319187104932e-06,
+      "loss": 0.4562,
+      "step": 1041
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5286500563588563,
+      "learning_rate": 9.790786847670803e-06,
+      "loss": 0.5211,
+      "step": 1042
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.74850734360135,
+      "learning_rate": 9.79025384461771e-06,
+      "loss": 0.5711,
+      "step": 1043
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.653281161769676,
+      "learning_rate": 9.789720178019483e-06,
+      "loss": 0.4858,
+      "step": 1044
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7634185634687276,
+      "learning_rate": 9.789185847950048e-06,
+      "loss": 0.5161,
+      "step": 1045
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8256191538644848,
+      "learning_rate": 9.788650854483418e-06,
+      "loss": 0.5129,
+      "step": 1046
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7558127218892805,
+      "learning_rate": 9.788115197693702e-06,
+      "loss": 0.6176,
+      "step": 1047
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.975730899011124,
+      "learning_rate": 9.787578877655097e-06,
+      "loss": 0.5984,
+      "step": 1048
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9143506115996631,
+      "learning_rate": 9.787041894441895e-06,
+      "loss": 0.5623,
+      "step": 1049
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8367601667559863,
+      "learning_rate": 9.78650424812848e-06,
+      "loss": 0.4863,
+      "step": 1050
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7923229183103575,
+      "learning_rate": 9.785965938789324e-06,
+      "loss": 0.5821,
+      "step": 1051
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7746745276691676,
+      "learning_rate": 9.785426966498997e-06,
+      "loss": 0.4848,
+      "step": 1052
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6056208176167677,
+      "learning_rate": 9.784887331332153e-06,
+      "loss": 0.5214,
+      "step": 1053
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8536291617414775,
+      "learning_rate": 9.784347033363548e-06,
+      "loss": 0.5252,
+      "step": 1054
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7311654246122938,
+      "learning_rate": 9.783806072668018e-06,
+      "loss": 0.5526,
+      "step": 1055
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.624988059364019,
+      "learning_rate": 9.7832644493205e-06,
+      "loss": 0.4739,
+      "step": 1056
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.044997115353559,
+      "learning_rate": 9.782722163396019e-06,
+      "loss": 0.5295,
+      "step": 1057
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9606906286989311,
+      "learning_rate": 9.782179214969693e-06,
+      "loss": 0.7692,
+      "step": 1058
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6145329688566057,
+      "learning_rate": 9.781635604116731e-06,
+      "loss": 0.5612,
+      "step": 1059
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.723799689000995,
+      "learning_rate": 9.78109133091243e-06,
+      "loss": 0.543,
+      "step": 1060
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5940788671836397,
+      "learning_rate": 9.780546395432188e-06,
+      "loss": 0.511,
+      "step": 1061
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6206052200842596,
+      "learning_rate": 9.780000797751489e-06,
+      "loss": 0.6019,
+      "step": 1062
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5176849367769005,
+      "learning_rate": 9.779454537945906e-06,
+      "loss": 0.653,
+      "step": 1063
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.158508397858273,
+      "learning_rate": 9.778907616091108e-06,
+      "loss": 0.5375,
+      "step": 1064
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8051967241073346,
+      "learning_rate": 9.778360032262855e-06,
+      "loss": 0.5596,
+      "step": 1065
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.279184037575354,
+      "learning_rate": 9.777811786536997e-06,
+      "loss": 0.6517,
+      "step": 1066
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6064123951606217,
+      "learning_rate": 9.777262878989479e-06,
+      "loss": 0.5744,
+      "step": 1067
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5453461124778967,
+      "learning_rate": 9.776713309696335e-06,
+      "loss": 0.5056,
+      "step": 1068
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7925892276212625,
+      "learning_rate": 9.77616307873369e-06,
+      "loss": 0.5721,
+      "step": 1069
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2840604791139123,
+      "learning_rate": 9.775612186177762e-06,
+      "loss": 0.5716,
+      "step": 1070
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.820901935750855,
+      "learning_rate": 9.775060632104862e-06,
+      "loss": 0.5253,
+      "step": 1071
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7963163768707844,
+      "learning_rate": 9.77450841659139e-06,
+      "loss": 0.4414,
+      "step": 1072
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9375016601534734,
+      "learning_rate": 9.77395553971384e-06,
+      "loss": 0.6388,
+      "step": 1073
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1537371407988473,
+      "learning_rate": 9.773402001548794e-06,
+      "loss": 0.4858,
+      "step": 1074
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.82832547448926,
+      "learning_rate": 9.77284780217293e-06,
+      "loss": 0.5245,
+      "step": 1075
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9660033833108304,
+      "learning_rate": 9.772292941663015e-06,
+      "loss": 0.5771,
+      "step": 1076
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9363558696402283,
+      "learning_rate": 9.771737420095908e-06,
+      "loss": 0.5608,
+      "step": 1077
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6636530012885635,
+      "learning_rate": 9.77118123754856e-06,
+      "loss": 0.5623,
+      "step": 1078
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7082104492170138,
+      "learning_rate": 9.770624394098015e-06,
+      "loss": 0.5576,
+      "step": 1079
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.953556156305212,
+      "learning_rate": 9.770066889821403e-06,
+      "loss": 0.5186,
+      "step": 1080
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.652021280426804,
+      "learning_rate": 9.769508724795953e-06,
+      "loss": 0.5925,
+      "step": 1081
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.866883944372643,
+      "learning_rate": 9.768949899098981e-06,
+      "loss": 0.5017,
+      "step": 1082
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8288837934201587,
+      "learning_rate": 9.768390412807894e-06,
+      "loss": 0.4962,
+      "step": 1083
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.747587610245608,
+      "learning_rate": 9.767830266000194e-06,
+      "loss": 0.5369,
+      "step": 1084
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8446737517151766,
+      "learning_rate": 9.76726945875347e-06,
+      "loss": 0.6093,
+      "step": 1085
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0528630017289824,
+      "learning_rate": 9.766707991145407e-06,
+      "loss": 0.7344,
+      "step": 1086
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.498483165262252,
+      "learning_rate": 9.766145863253778e-06,
+      "loss": 0.5354,
+      "step": 1087
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.525873089272019,
+      "learning_rate": 9.765583075156451e-06,
+      "loss": 0.5475,
+      "step": 1088
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8915856670547946,
+      "learning_rate": 9.76501962693138e-06,
+      "loss": 0.5054,
+      "step": 1089
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.202894083963947,
+      "learning_rate": 9.764455518656617e-06,
+      "loss": 0.6191,
+      "step": 1090
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.048323948953868,
+      "learning_rate": 9.7638907504103e-06,
+      "loss": 0.6221,
+      "step": 1091
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.859390596744628,
+      "learning_rate": 9.763325322270663e-06,
+      "loss": 0.5485,
+      "step": 1092
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6296253935496907,
+      "learning_rate": 9.762759234316026e-06,
+      "loss": 0.5687,
+      "step": 1093
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.106137614144312,
+      "learning_rate": 9.762192486624805e-06,
+      "loss": 0.5949,
+      "step": 1094
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 18.77649770770319,
+      "learning_rate": 9.761625079275506e-06,
+      "loss": 0.561,
+      "step": 1095
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.135307080314248,
+      "learning_rate": 9.761057012346724e-06,
+      "loss": 0.5587,
+      "step": 1096
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0694752423475116,
+      "learning_rate": 9.760488285917152e-06,
+      "loss": 0.8036,
+      "step": 1097
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6986331298285788,
+      "learning_rate": 9.759918900065564e-06,
+      "loss": 0.4926,
+      "step": 1098
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0883980159974667,
+      "learning_rate": 9.759348854870836e-06,
+      "loss": 0.5514,
+      "step": 1099
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2420830049049103,
+      "learning_rate": 9.75877815041193e-06,
+      "loss": 0.4438,
+      "step": 1100
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7496028999049695,
+      "learning_rate": 9.758206786767897e-06,
+      "loss": 0.5964,
+      "step": 1101
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8413880447299287,
+      "learning_rate": 9.757634764017885e-06,
+      "loss": 0.541,
+      "step": 1102
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8311128352460027,
+      "learning_rate": 9.75706208224113e-06,
+      "loss": 0.5505,
+      "step": 1103
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.925705189991499,
+      "learning_rate": 9.756488741516958e-06,
+      "loss": 0.5832,
+      "step": 1104
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5435021417437853,
+      "learning_rate": 9.75591474192479e-06,
+      "loss": 0.5065,
+      "step": 1105
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.798812617974512,
+      "learning_rate": 9.755340083544138e-06,
+      "loss": 0.4975,
+      "step": 1106
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6728234741591104,
+      "learning_rate": 9.754764766454598e-06,
+      "loss": 0.5916,
+      "step": 1107
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6835967492709578,
+      "learning_rate": 9.754188790735867e-06,
+      "loss": 0.5969,
+      "step": 1108
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1458369809479296,
+      "learning_rate": 9.75361215646773e-06,
+      "loss": 0.5379,
+      "step": 1109
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9384734720719554,
+      "learning_rate": 9.753034863730058e-06,
+      "loss": 0.5354,
+      "step": 1110
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9286885421673534,
+      "learning_rate": 9.752456912602821e-06,
+      "loss": 0.549,
+      "step": 1111
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0214374862873097,
+      "learning_rate": 9.751878303166076e-06,
+      "loss": 0.5268,
+      "step": 1112
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1005791139988825,
+      "learning_rate": 9.75129903549997e-06,
+      "loss": 0.5557,
+      "step": 1113
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0012338151437288,
+      "learning_rate": 9.750719109684746e-06,
+      "loss": 0.7965,
+      "step": 1114
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1964835138276273,
+      "learning_rate": 9.750138525800732e-06,
+      "loss": 0.5391,
+      "step": 1115
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8551273669675743,
+      "learning_rate": 9.749557283928354e-06,
+      "loss": 0.5415,
+      "step": 1116
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1977387206178816,
+      "learning_rate": 9.748975384148123e-06,
+      "loss": 0.6123,
+      "step": 1117
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8800907563680433,
+      "learning_rate": 9.748392826540645e-06,
+      "loss": 0.7572,
+      "step": 1118
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0690865746295914,
+      "learning_rate": 9.747809611186614e-06,
+      "loss": 0.565,
+      "step": 1119
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5268434924081653,
+      "learning_rate": 9.747225738166818e-06,
+      "loss": 0.5321,
+      "step": 1120
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.25232847379261,
+      "learning_rate": 9.746641207562137e-06,
+      "loss": 0.5756,
+      "step": 1121
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0827120147142124,
+      "learning_rate": 9.746056019453536e-06,
+      "loss": 0.5586,
+      "step": 1122
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1805738741584695,
+      "learning_rate": 9.745470173922078e-06,
+      "loss": 0.5393,
+      "step": 1123
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.598898635078897,
+      "learning_rate": 9.744883671048912e-06,
+      "loss": 0.5129,
+      "step": 1124
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2741894126822686,
+      "learning_rate": 9.744296510915285e-06,
+      "loss": 0.4429,
+      "step": 1125
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7305633629946673,
+      "learning_rate": 9.743708693602526e-06,
+      "loss": 0.5911,
+      "step": 1126
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0696916797211156,
+      "learning_rate": 9.743120219192057e-06,
+      "loss": 0.551,
+      "step": 1127
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7268648989762236,
+      "learning_rate": 9.7425310877654e-06,
+      "loss": 0.6358,
+      "step": 1128
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7161246235019867,
+      "learning_rate": 9.741941299404157e-06,
+      "loss": 0.4771,
+      "step": 1129
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4477345262586394,
+      "learning_rate": 9.741350854190028e-06,
+      "loss": 0.5041,
+      "step": 1130
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8153622433324048,
+      "learning_rate": 9.740759752204798e-06,
+      "loss": 0.5874,
+      "step": 1131
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.153402684992974,
+      "learning_rate": 9.74016799353035e-06,
+      "loss": 0.5805,
+      "step": 1132
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.032897770318943,
+      "learning_rate": 9.739575578248652e-06,
+      "loss": 0.8022,
+      "step": 1133
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.489877394969998,
+      "learning_rate": 9.738982506441765e-06,
+      "loss": 0.5259,
+      "step": 1134
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7767384672829438,
+      "learning_rate": 9.738388778191842e-06,
+      "loss": 0.4944,
+      "step": 1135
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0808299114652224,
+      "learning_rate": 9.737794393581125e-06,
+      "loss": 0.5524,
+      "step": 1136
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0024919052189283,
+      "learning_rate": 9.737199352691952e-06,
+      "loss": 0.5834,
+      "step": 1137
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9283759813613832,
+      "learning_rate": 9.736603655606744e-06,
+      "loss": 0.5039,
+      "step": 1138
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.610968904185747,
+      "learning_rate": 9.73600730240802e-06,
+      "loss": 0.5993,
+      "step": 1139
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4684272282164907,
+      "learning_rate": 9.735410293178382e-06,
+      "loss": 0.5081,
+      "step": 1140
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3822223808973453,
+      "learning_rate": 9.73481262800053e-06,
+      "loss": 0.5757,
+      "step": 1141
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5326342351823343,
+      "learning_rate": 9.734214306957255e-06,
+      "loss": 0.5317,
+      "step": 1142
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7957597692658716,
+      "learning_rate": 9.733615330131432e-06,
+      "loss": 0.5374,
+      "step": 1143
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8075284868496324,
+      "learning_rate": 9.733015697606036e-06,
+      "loss": 0.5659,
+      "step": 1144
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4806346999379807,
+      "learning_rate": 9.732415409464124e-06,
+      "loss": 0.5322,
+      "step": 1145
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.019984092695893,
+      "learning_rate": 9.73181446578885e-06,
+      "loss": 0.5741,
+      "step": 1146
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.192169125700534,
+      "learning_rate": 9.731212866663453e-06,
+      "loss": 0.5418,
+      "step": 1147
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7722865380127528,
+      "learning_rate": 9.730610612171272e-06,
+      "loss": 0.583,
+      "step": 1148
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8945520109358525,
+      "learning_rate": 9.730007702395728e-06,
+      "loss": 0.5234,
+      "step": 1149
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6772111050392062,
+      "learning_rate": 9.729404137420335e-06,
+      "loss": 0.5266,
+      "step": 1150
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.32089407070629,
+      "learning_rate": 9.7287999173287e-06,
+      "loss": 0.5213,
+      "step": 1151
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1159783669073793,
+      "learning_rate": 9.728195042204522e-06,
+      "loss": 0.5247,
+      "step": 1152
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7223782909228391,
+      "learning_rate": 9.727589512131583e-06,
+      "loss": 0.5422,
+      "step": 1153
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2783065543852263,
+      "learning_rate": 9.726983327193764e-06,
+      "loss": 0.577,
+      "step": 1154
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9389354648862989,
+      "learning_rate": 9.726376487475035e-06,
+      "loss": 0.6023,
+      "step": 1155
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1984364573021744,
+      "learning_rate": 9.725768993059452e-06,
+      "loss": 0.5586,
+      "step": 1156
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.048867249349575,
+      "learning_rate": 9.725160844031168e-06,
+      "loss": 0.6111,
+      "step": 1157
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7932938579703839,
+      "learning_rate": 9.724552040474421e-06,
+      "loss": 0.6174,
+      "step": 1158
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.354668452481259,
+      "learning_rate": 9.723942582473545e-06,
+      "loss": 0.5754,
+      "step": 1159
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8310381019549715,
+      "learning_rate": 9.723332470112959e-06,
+      "loss": 0.6152,
+      "step": 1160
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9904506036507443,
+      "learning_rate": 9.722721703477178e-06,
+      "loss": 0.5711,
+      "step": 1161
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.645971158890774,
+      "learning_rate": 9.722110282650805e-06,
+      "loss": 0.5389,
+      "step": 1162
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.435873000889471,
+      "learning_rate": 9.721498207718533e-06,
+      "loss": 0.4873,
+      "step": 1163
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.105386946230954,
+      "learning_rate": 9.720885478765147e-06,
+      "loss": 0.565,
+      "step": 1164
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7215944844426054,
+      "learning_rate": 9.720272095875523e-06,
+      "loss": 0.5849,
+      "step": 1165
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7952740292959033,
+      "learning_rate": 9.719658059134624e-06,
+      "loss": 0.4681,
+      "step": 1166
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8564598935283374,
+      "learning_rate": 9.719043368627511e-06,
+      "loss": 0.4824,
+      "step": 1167
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.164212558837668,
+      "learning_rate": 9.718428024439326e-06,
+      "loss": 0.5809,
+      "step": 1168
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.310640636623551,
+      "learning_rate": 9.717812026655308e-06,
+      "loss": 0.4249,
+      "step": 1169
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7256183847388988,
+      "learning_rate": 9.717195375360786e-06,
+      "loss": 0.5632,
+      "step": 1170
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8898905408970166,
+      "learning_rate": 9.716578070641178e-06,
+      "loss": 0.5783,
+      "step": 1171
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5991991034036723,
+      "learning_rate": 9.715960112581992e-06,
+      "loss": 0.5394,
+      "step": 1172
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7620975711020934,
+      "learning_rate": 9.715341501268828e-06,
+      "loss": 0.5958,
+      "step": 1173
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6573539687568224,
+      "learning_rate": 9.714722236787377e-06,
+      "loss": 0.5954,
+      "step": 1174
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0955382122552138,
+      "learning_rate": 9.714102319223417e-06,
+      "loss": 0.5094,
+      "step": 1175
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9211957238906248,
+      "learning_rate": 9.71348174866282e-06,
+      "loss": 0.5573,
+      "step": 1176
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.104246197541215,
+      "learning_rate": 9.71286052519155e-06,
+      "loss": 0.5491,
+      "step": 1177
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3347678672012937,
+      "learning_rate": 9.712238648895655e-06,
+      "loss": 0.5579,
+      "step": 1178
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0362854768757415,
+      "learning_rate": 9.711616119861278e-06,
+      "loss": 0.5169,
+      "step": 1179
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.37708880810182,
+      "learning_rate": 9.710992938174653e-06,
+      "loss": 0.5069,
+      "step": 1180
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8067092512213696,
+      "learning_rate": 9.710369103922101e-06,
+      "loss": 0.5704,
+      "step": 1181
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0319874695569398,
+      "learning_rate": 9.709744617190039e-06,
+      "loss": 0.8075,
+      "step": 1182
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1246490879670197,
+      "learning_rate": 9.709119478064965e-06,
+      "loss": 0.5285,
+      "step": 1183
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9896056288122467,
+      "learning_rate": 9.708493686633479e-06,
+      "loss": 0.818,
+      "step": 1184
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0602780374139904,
+      "learning_rate": 9.70786724298226e-06,
+      "loss": 0.5404,
+      "step": 1185
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.123247347910237,
+      "learning_rate": 9.707240147198089e-06,
+      "loss": 0.6294,
+      "step": 1186
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7564545726358114,
+      "learning_rate": 9.706612399367828e-06,
+      "loss": 0.5064,
+      "step": 1187
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0146727228599492,
+      "learning_rate": 9.705983999578433e-06,
+      "loss": 0.5689,
+      "step": 1188
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7106021183468805,
+      "learning_rate": 9.705354947916947e-06,
+      "loss": 0.5307,
+      "step": 1189
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5624719435084782,
+      "learning_rate": 9.704725244470509e-06,
+      "loss": 0.5062,
+      "step": 1190
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7638372444772765,
+      "learning_rate": 9.704094889326347e-06,
+      "loss": 0.5707,
+      "step": 1191
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7416476443128306,
+      "learning_rate": 9.703463882571775e-06,
+      "loss": 0.4612,
+      "step": 1192
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8788622027847808,
+      "learning_rate": 9.7028322242942e-06,
+      "loss": 0.5972,
+      "step": 1193
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5947278035285033,
+      "learning_rate": 9.70219991458112e-06,
+      "loss": 0.4544,
+      "step": 1194
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6834339839276224,
+      "learning_rate": 9.701566953520123e-06,
+      "loss": 0.4592,
+      "step": 1195
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6980620087124993,
+      "learning_rate": 9.700933341198885e-06,
+      "loss": 0.5742,
+      "step": 1196
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9665219047094553,
+      "learning_rate": 9.700299077705176e-06,
+      "loss": 0.5163,
+      "step": 1197
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0496349298879504,
+      "learning_rate": 9.699664163126851e-06,
+      "loss": 0.4949,
+      "step": 1198
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6394137764220442,
+      "learning_rate": 9.699028597551862e-06,
+      "loss": 0.568,
+      "step": 1199
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9774859862632355,
+      "learning_rate": 9.698392381068244e-06,
+      "loss": 0.5271,
+      "step": 1200
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0477768094349402,
+      "learning_rate": 9.697755513764128e-06,
+      "loss": 0.5543,
+      "step": 1201
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0234679012026042,
+      "learning_rate": 9.697117995727732e-06,
+      "loss": 0.5556,
+      "step": 1202
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7828763959907534,
+      "learning_rate": 9.696479827047364e-06,
+      "loss": 0.5791,
+      "step": 1203
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5487664238465073,
+      "learning_rate": 9.695841007811424e-06,
+      "loss": 0.4605,
+      "step": 1204
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1994498325911764,
+      "learning_rate": 9.695201538108403e-06,
+      "loss": 0.5738,
+      "step": 1205
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3968025704719187,
+      "learning_rate": 9.694561418026875e-06,
+      "loss": 0.5175,
+      "step": 1206
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.635182898671791,
+      "learning_rate": 9.693920647655515e-06,
+      "loss": 0.4882,
+      "step": 1207
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.377929644036782,
+      "learning_rate": 9.693279227083079e-06,
+      "loss": 0.5363,
+      "step": 1208
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8005594942717784,
+      "learning_rate": 9.692637156398417e-06,
+      "loss": 0.5271,
+      "step": 1209
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6219754110794222,
+      "learning_rate": 9.69199443569047e-06,
+      "loss": 0.5057,
+      "step": 1210
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1370649640349915,
+      "learning_rate": 9.691351065048266e-06,
+      "loss": 0.5397,
+      "step": 1211
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9834854095762893,
+      "learning_rate": 9.690707044560924e-06,
+      "loss": 0.5886,
+      "step": 1212
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0244444395084003,
+      "learning_rate": 9.690062374317656e-06,
+      "loss": 0.5308,
+      "step": 1213
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9977791311500488,
+      "learning_rate": 9.68941705440776e-06,
+      "loss": 0.5539,
+      "step": 1214
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8303079553934596,
+      "learning_rate": 9.688771084920625e-06,
+      "loss": 0.5131,
+      "step": 1215
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7581418194598397,
+      "learning_rate": 9.688124465945732e-06,
+      "loss": 0.5365,
+      "step": 1216
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.347117307573303,
+      "learning_rate": 9.68747719757265e-06,
+      "loss": 0.4837,
+      "step": 1217
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3103987493003872,
+      "learning_rate": 9.686829279891037e-06,
+      "loss": 0.5223,
+      "step": 1218
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2173581871658352,
+      "learning_rate": 9.686180712990647e-06,
+      "loss": 0.7106,
+      "step": 1219
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5793123924693708,
+      "learning_rate": 9.685531496961314e-06,
+      "loss": 0.5359,
+      "step": 1220
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7801466418323264,
+      "learning_rate": 9.684881631892971e-06,
+      "loss": 0.55,
+      "step": 1221
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4745760379179944,
+      "learning_rate": 9.684231117875634e-06,
+      "loss": 0.4955,
+      "step": 1222
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0682686278338953,
+      "learning_rate": 9.683579954999415e-06,
+      "loss": 0.5503,
+      "step": 1223
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6723597508200834,
+      "learning_rate": 9.68292814335451e-06,
+      "loss": 0.4286,
+      "step": 1224
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7455934620170128,
+      "learning_rate": 9.682275683031213e-06,
+      "loss": 0.4976,
+      "step": 1225
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.386345487079645,
+      "learning_rate": 9.681622574119898e-06,
+      "loss": 0.5515,
+      "step": 1226
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2863482924147216,
+      "learning_rate": 9.680968816711033e-06,
+      "loss": 0.5484,
+      "step": 1227
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8404935293643752,
+      "learning_rate": 9.680314410895182e-06,
+      "loss": 0.5135,
+      "step": 1228
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.681102520837694,
+      "learning_rate": 9.679659356762987e-06,
+      "loss": 0.4648,
+      "step": 1229
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8005599638084562,
+      "learning_rate": 9.679003654405188e-06,
+      "loss": 0.569,
+      "step": 1230
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0065301837428198,
+      "learning_rate": 9.678347303912615e-06,
+      "loss": 0.7712,
+      "step": 1231
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4657913817255996,
+      "learning_rate": 9.677690305376182e-06,
+      "loss": 0.5298,
+      "step": 1232
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4236628566333467,
+      "learning_rate": 9.6770326588869e-06,
+      "loss": 0.4208,
+      "step": 1233
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7502833414627141,
+      "learning_rate": 9.676374364535864e-06,
+      "loss": 0.5277,
+      "step": 1234
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8853252144724015,
+      "learning_rate": 9.67571542241426e-06,
+      "loss": 0.5274,
+      "step": 1235
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0394596918590175,
+      "learning_rate": 9.675055832613365e-06,
+      "loss": 0.5738,
+      "step": 1236
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8305880186273233,
+      "learning_rate": 9.674395595224546e-06,
+      "loss": 0.651,
+      "step": 1237
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.690649734674144,
+      "learning_rate": 9.67373471033926e-06,
+      "loss": 0.5621,
+      "step": 1238
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6784242707116488,
+      "learning_rate": 9.673073178049051e-06,
+      "loss": 0.4988,
+      "step": 1239
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.741767587848836,
+      "learning_rate": 9.672410998445553e-06,
+      "loss": 0.6918,
+      "step": 1240
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5737718714374644,
+      "learning_rate": 9.671748171620497e-06,
+      "loss": 0.5579,
+      "step": 1241
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9985290132272966,
+      "learning_rate": 9.67108469766569e-06,
+      "loss": 0.5544,
+      "step": 1242
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6771506632645983,
+      "learning_rate": 9.67042057667304e-06,
+      "loss": 0.5922,
+      "step": 1243
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8333996955696386,
+      "learning_rate": 9.669755808734541e-06,
+      "loss": 0.5705,
+      "step": 1244
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7005159543780555,
+      "learning_rate": 9.669090393942277e-06,
+      "loss": 0.5212,
+      "step": 1245
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6265421547783332,
+      "learning_rate": 9.66842433238842e-06,
+      "loss": 0.5335,
+      "step": 1246
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8499712273374371,
+      "learning_rate": 9.667757624165231e-06,
+      "loss": 0.734,
+      "step": 1247
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.55555220403889,
+      "learning_rate": 9.667090269365066e-06,
+      "loss": 0.5439,
+      "step": 1248
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8486241773687218,
+      "learning_rate": 9.666422268080366e-06,
+      "loss": 0.5254,
+      "step": 1249
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9496732048260292,
+      "learning_rate": 9.665753620403661e-06,
+      "loss": 0.6488,
+      "step": 1250
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8081945718074874,
+      "learning_rate": 9.665084326427575e-06,
+      "loss": 0.4925,
+      "step": 1251
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6960173732175188,
+      "learning_rate": 9.664414386244812e-06,
+      "loss": 0.5717,
+      "step": 1252
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1104227427882822,
+      "learning_rate": 9.663743799948178e-06,
+      "loss": 0.621,
+      "step": 1253
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7765506700447908,
+      "learning_rate": 9.66307256763056e-06,
+      "loss": 0.5686,
+      "step": 1254
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.687176184215572,
+      "learning_rate": 9.66240068938494e-06,
+      "loss": 0.5834,
+      "step": 1255
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8520980958277116,
+      "learning_rate": 9.661728165304381e-06,
+      "loss": 0.502,
+      "step": 1256
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.616056889833274,
+      "learning_rate": 9.661054995482045e-06,
+      "loss": 0.558,
+      "step": 1257
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5986388959168787,
+      "learning_rate": 9.660381180011177e-06,
+      "loss": 0.4505,
+      "step": 1258
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1478693169988756,
+      "learning_rate": 9.659706718985118e-06,
+      "loss": 0.5014,
+      "step": 1259
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8684866220535443,
+      "learning_rate": 9.65903161249729e-06,
+      "loss": 0.6017,
+      "step": 1260
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9030299868505797,
+      "learning_rate": 9.658355860641212e-06,
+      "loss": 0.722,
+      "step": 1261
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9959181349520354,
+      "learning_rate": 9.657679463510483e-06,
+      "loss": 0.5953,
+      "step": 1262
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9569197100920068,
+      "learning_rate": 9.657002421198805e-06,
+      "loss": 0.5841,
+      "step": 1263
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.565820515759259,
+      "learning_rate": 9.656324733799955e-06,
+      "loss": 0.565,
+      "step": 1264
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8092120582056619,
+      "learning_rate": 9.655646401407813e-06,
+      "loss": 0.6031,
+      "step": 1265
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.000394155729214,
+      "learning_rate": 9.654967424116335e-06,
+      "loss": 0.5453,
+      "step": 1266
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7257342102439934,
+      "learning_rate": 9.654287802019578e-06,
+      "loss": 0.597,
+      "step": 1267
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7129074037705536,
+      "learning_rate": 9.653607535211677e-06,
+      "loss": 0.5042,
+      "step": 1268
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.703515189652405,
+      "learning_rate": 9.65292662378687e-06,
+      "loss": 0.6198,
+      "step": 1269
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.141899223080931,
+      "learning_rate": 9.652245067839472e-06,
+      "loss": 0.6045,
+      "step": 1270
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3538866802682046,
+      "learning_rate": 9.651562867463892e-06,
+      "loss": 0.5692,
+      "step": 1271
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9302148343853345,
+      "learning_rate": 9.65088002275463e-06,
+      "loss": 0.6228,
+      "step": 1272
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9483725325277228,
+      "learning_rate": 9.650196533806272e-06,
+      "loss": 0.5384,
+      "step": 1273
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7064954059021051,
+      "learning_rate": 9.649512400713497e-06,
+      "loss": 0.6075,
+      "step": 1274
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8465982724463552,
+      "learning_rate": 9.64882762357107e-06,
+      "loss": 0.5342,
+      "step": 1275
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.447802164711969,
+      "learning_rate": 9.648142202473844e-06,
+      "loss": 0.4944,
+      "step": 1276
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9284366628673724,
+      "learning_rate": 9.647456137516766e-06,
+      "loss": 0.7562,
+      "step": 1277
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8975927056846607,
+      "learning_rate": 9.646769428794869e-06,
+      "loss": 0.5168,
+      "step": 1278
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.821651679304054,
+      "learning_rate": 9.646082076403276e-06,
+      "loss": 0.504,
+      "step": 1279
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.352561983560717,
+      "learning_rate": 9.645394080437197e-06,
+      "loss": 0.4853,
+      "step": 1280
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.683215841892426,
+      "learning_rate": 9.644705440991935e-06,
+      "loss": 0.5744,
+      "step": 1281
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9920554386622165,
+      "learning_rate": 9.644016158162881e-06,
+      "loss": 0.5115,
+      "step": 1282
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4687303058955226,
+      "learning_rate": 9.643326232045512e-06,
+      "loss": 0.5171,
+      "step": 1283
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2449786537853833,
+      "learning_rate": 9.642635662735397e-06,
+      "loss": 0.6039,
+      "step": 1284
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6897826890743508,
+      "learning_rate": 9.641944450328196e-06,
+      "loss": 0.5233,
+      "step": 1285
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7833681210350623,
+      "learning_rate": 9.641252594919653e-06,
+      "loss": 0.5536,
+      "step": 1286
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6342141858776325,
+      "learning_rate": 9.640560096605605e-06,
+      "loss": 0.5006,
+      "step": 1287
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9214249787056155,
+      "learning_rate": 9.639866955481975e-06,
+      "loss": 0.5234,
+      "step": 1288
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.917029874904909,
+      "learning_rate": 9.639173171644778e-06,
+      "loss": 0.795,
+      "step": 1289
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.805088008871878,
+      "learning_rate": 9.638478745190118e-06,
+      "loss": 0.4273,
+      "step": 1290
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5912050812401852,
+      "learning_rate": 9.637783676214186e-06,
+      "loss": 0.5884,
+      "step": 1291
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0056041682416565,
+      "learning_rate": 9.63708796481326e-06,
+      "loss": 0.5395,
+      "step": 1292
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.408238102553004,
+      "learning_rate": 9.636391611083712e-06,
+      "loss": 0.5822,
+      "step": 1293
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7940473401151356,
+      "learning_rate": 9.635694615122004e-06,
+      "loss": 0.6245,
+      "step": 1294
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.576843151269019,
+      "learning_rate": 9.63499697702468e-06,
+      "loss": 0.5923,
+      "step": 1295
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.077790163759468,
+      "learning_rate": 9.634298696888376e-06,
+      "loss": 0.5623,
+      "step": 1296
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5173566733612762,
+      "learning_rate": 9.633599774809822e-06,
+      "loss": 0.497,
+      "step": 1297
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.6186170364128305,
+      "learning_rate": 9.632900210885827e-06,
+      "loss": 0.5691,
+      "step": 1298
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6991767005197203,
+      "learning_rate": 9.632200005213299e-06,
+      "loss": 0.5421,
+      "step": 1299
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7307332167395466,
+      "learning_rate": 9.631499157889226e-06,
+      "loss": 0.5393,
+      "step": 1300
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9141365955991985,
+      "learning_rate": 9.630797669010694e-06,
+      "loss": 0.4972,
+      "step": 1301
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8219624525909301,
+      "learning_rate": 9.630095538674871e-06,
+      "loss": 0.5443,
+      "step": 1302
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8239486039604076,
+      "learning_rate": 9.629392766979016e-06,
+      "loss": 0.5789,
+      "step": 1303
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.686258332096098,
+      "learning_rate": 9.628689354020474e-06,
+      "loss": 0.4382,
+      "step": 1304
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5847306289008554,
+      "learning_rate": 9.627985299896688e-06,
+      "loss": 0.4443,
+      "step": 1305
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1747028821828125,
+      "learning_rate": 9.62728060470518e-06,
+      "loss": 0.5046,
+      "step": 1306
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.091612566972885,
+      "learning_rate": 9.626575268543561e-06,
+      "loss": 0.4844,
+      "step": 1307
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0874893210845076,
+      "learning_rate": 9.62586929150954e-06,
+      "loss": 0.4793,
+      "step": 1308
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.359802442764472,
+      "learning_rate": 9.625162673700906e-06,
+      "loss": 0.4877,
+      "step": 1309
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.331252042764603,
+      "learning_rate": 9.624455415215537e-06,
+      "loss": 0.5427,
+      "step": 1310
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8011672255880258,
+      "learning_rate": 9.623747516151406e-06,
+      "loss": 0.4814,
+      "step": 1311
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.508310298385075,
+      "learning_rate": 9.62303897660657e-06,
+      "loss": 0.4219,
+      "step": 1312
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7003640367450878,
+      "learning_rate": 9.622329796679175e-06,
+      "loss": 0.5733,
+      "step": 1313
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1838289810681117,
+      "learning_rate": 9.621619976467459e-06,
+      "loss": 0.6036,
+      "step": 1314
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8459241852753459,
+      "learning_rate": 9.620909516069744e-06,
+      "loss": 0.4744,
+      "step": 1315
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0316668217648224,
+      "learning_rate": 9.620198415584441e-06,
+      "loss": 0.4295,
+      "step": 1316
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1893943314331343,
+      "learning_rate": 9.619486675110055e-06,
+      "loss": 0.5323,
+      "step": 1317
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5583331387922668,
+      "learning_rate": 9.618774294745178e-06,
+      "loss": 0.4657,
+      "step": 1318
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.748098941054415,
+      "learning_rate": 9.618061274588481e-06,
+      "loss": 0.5773,
+      "step": 1319
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8781004289935406,
+      "learning_rate": 9.617347614738738e-06,
+      "loss": 0.5809,
+      "step": 1320
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5016398856823465,
+      "learning_rate": 9.616633315294803e-06,
+      "loss": 0.5738,
+      "step": 1321
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8070642676942736,
+      "learning_rate": 9.61591837635562e-06,
+      "loss": 0.593,
+      "step": 1322
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.8759527414630104,
+      "learning_rate": 9.615202798020224e-06,
+      "loss": 0.6197,
+      "step": 1323
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5907853665896872,
+      "learning_rate": 9.614486580387737e-06,
+      "loss": 0.5527,
+      "step": 1324
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4058247734643388,
+      "learning_rate": 9.613769723557366e-06,
+      "loss": 0.5374,
+      "step": 1325
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9395281241149795,
+      "learning_rate": 9.613052227628414e-06,
+      "loss": 0.7563,
+      "step": 1326
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.550992372654103,
+      "learning_rate": 9.612334092700264e-06,
+      "loss": 0.581,
+      "step": 1327
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8151426063325695,
+      "learning_rate": 9.611615318872396e-06,
+      "loss": 0.5768,
+      "step": 1328
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8696696096733496,
+      "learning_rate": 9.610895906244373e-06,
+      "loss": 0.5449,
+      "step": 1329
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.160433503560469,
+      "learning_rate": 9.610175854915846e-06,
+      "loss": 0.5867,
+      "step": 1330
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.661538476041728,
+      "learning_rate": 9.60945516498656e-06,
+      "loss": 0.4589,
+      "step": 1331
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9172872587964246,
+      "learning_rate": 9.608733836556343e-06,
+      "loss": 0.736,
+      "step": 1332
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0354648502978274,
+      "learning_rate": 9.608011869725111e-06,
+      "loss": 0.6312,
+      "step": 1333
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2342857897992436,
+      "learning_rate": 9.607289264592874e-06,
+      "loss": 0.5859,
+      "step": 1334
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.07927537838627,
+      "learning_rate": 9.606566021259726e-06,
+      "loss": 0.4496,
+      "step": 1335
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9153277794732088,
+      "learning_rate": 9.605842139825851e-06,
+      "loss": 0.7503,
+      "step": 1336
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.036853709615021,
+      "learning_rate": 9.605117620391522e-06,
+      "loss": 0.6517,
+      "step": 1337
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1171520480442165,
+      "learning_rate": 9.604392463057097e-06,
+      "loss": 0.5511,
+      "step": 1338
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9747557856238598,
+      "learning_rate": 9.603666667923024e-06,
+      "loss": 0.6502,
+      "step": 1339
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6538402051067016,
+      "learning_rate": 9.602940235089841e-06,
+      "loss": 0.5077,
+      "step": 1340
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.135139721275134,
+      "learning_rate": 9.602213164658177e-06,
+      "loss": 0.4967,
+      "step": 1341
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2259483627688725,
+      "learning_rate": 9.60148545672874e-06,
+      "loss": 0.57,
+      "step": 1342
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8727203899347886,
+      "learning_rate": 9.600757111402336e-06,
+      "loss": 0.5683,
+      "step": 1343
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8819895259237334,
+      "learning_rate": 9.600028128779853e-06,
+      "loss": 0.6488,
+      "step": 1344
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.541759175953644,
+      "learning_rate": 9.599298508962272e-06,
+      "loss": 0.6319,
+      "step": 1345
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.168890978221971,
+      "learning_rate": 9.598568252050655e-06,
+      "loss": 0.5149,
+      "step": 1346
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6705386031984615,
+      "learning_rate": 9.597837358146163e-06,
+      "loss": 0.5339,
+      "step": 1347
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6261352861228484,
+      "learning_rate": 9.597105827350035e-06,
+      "loss": 0.5364,
+      "step": 1348
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6660173173576758,
+      "learning_rate": 9.596373659763605e-06,
+      "loss": 0.5311,
+      "step": 1349
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7296864317719287,
+      "learning_rate": 9.59564085548829e-06,
+      "loss": 0.5809,
+      "step": 1350
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9743902359678382,
+      "learning_rate": 9.594907414625602e-06,
+      "loss": 0.5152,
+      "step": 1351
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8559917506672614,
+      "learning_rate": 9.594173337277134e-06,
+      "loss": 0.6136,
+      "step": 1352
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5856374992614448,
+      "learning_rate": 9.59343862354457e-06,
+      "loss": 0.5505,
+      "step": 1353
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.12921172670809,
+      "learning_rate": 9.592703273529684e-06,
+      "loss": 0.4847,
+      "step": 1354
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9823831914429788,
+      "learning_rate": 9.591967287334337e-06,
+      "loss": 0.4316,
+      "step": 1355
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0361992065456316,
+      "learning_rate": 9.591230665060476e-06,
+      "loss": 0.5168,
+      "step": 1356
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8389999731052249,
+      "learning_rate": 9.590493406810138e-06,
+      "loss": 0.7443,
+      "step": 1357
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.061037444653029,
+      "learning_rate": 9.589755512685451e-06,
+      "loss": 0.5526,
+      "step": 1358
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.303185799440571,
+      "learning_rate": 9.589016982788622e-06,
+      "loss": 0.5756,
+      "step": 1359
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.22577788698618,
+      "learning_rate": 9.588277817221956e-06,
+      "loss": 0.504,
+      "step": 1360
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.495231079842881,
+      "learning_rate": 9.587538016087842e-06,
+      "loss": 0.6001,
+      "step": 1361
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8828021045574168,
+      "learning_rate": 9.586797579488758e-06,
+      "loss": 0.5679,
+      "step": 1362
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6529132646256395,
+      "learning_rate": 9.586056507527266e-06,
+      "loss": 0.5296,
+      "step": 1363
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9965361476626804,
+      "learning_rate": 9.585314800306022e-06,
+      "loss": 0.5778,
+      "step": 1364
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7630374307097239,
+      "learning_rate": 9.584572457927766e-06,
+      "loss": 0.5877,
+      "step": 1365
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9561769411059107,
+      "learning_rate": 9.583829480495325e-06,
+      "loss": 0.5537,
+      "step": 1366
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8296568376509585,
+      "learning_rate": 9.583085868111622e-06,
+      "loss": 0.5761,
+      "step": 1367
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.410707494646512,
+      "learning_rate": 9.582341620879655e-06,
+      "loss": 0.4984,
+      "step": 1368
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7337146166604787,
+      "learning_rate": 9.581596738902521e-06,
+      "loss": 0.4464,
+      "step": 1369
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7860843511421733,
+      "learning_rate": 9.580851222283401e-06,
+      "loss": 0.5717,
+      "step": 1370
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.249520606836769,
+      "learning_rate": 9.580105071125564e-06,
+      "loss": 0.6088,
+      "step": 1371
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.000195050626322,
+      "learning_rate": 9.579358285532364e-06,
+      "loss": 0.4956,
+      "step": 1372
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9052364492719027,
+      "learning_rate": 9.578610865607249e-06,
+      "loss": 0.5011,
+      "step": 1373
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.896794260454515,
+      "learning_rate": 9.577862811453748e-06,
+      "loss": 0.4662,
+      "step": 1374
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.660013676739649,
+      "learning_rate": 9.577114123175486e-06,
+      "loss": 0.5505,
+      "step": 1375
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.150120646673957,
+      "learning_rate": 9.576364800876167e-06,
+      "loss": 0.501,
+      "step": 1376
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9557497549709233,
+      "learning_rate": 9.575614844659588e-06,
+      "loss": 0.6225,
+      "step": 1377
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8280569348577205,
+      "learning_rate": 9.574864254629634e-06,
+      "loss": 0.4841,
+      "step": 1378
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9463265637025862,
+      "learning_rate": 9.574113030890274e-06,
+      "loss": 0.4751,
+      "step": 1379
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9331697762572653,
+      "learning_rate": 9.573361173545572e-06,
+      "loss": 0.4618,
+      "step": 1380
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.850508198484521,
+      "learning_rate": 9.57260868269967e-06,
+      "loss": 0.5106,
+      "step": 1381
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.10499177313721,
+      "learning_rate": 9.571855558456807e-06,
+      "loss": 0.6001,
+      "step": 1382
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2096432909627723,
+      "learning_rate": 9.571101800921304e-06,
+      "loss": 0.601,
+      "step": 1383
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9332979120207041,
+      "learning_rate": 9.57034741019757e-06,
+      "loss": 0.5168,
+      "step": 1384
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0459003983315287,
+      "learning_rate": 9.569592386390105e-06,
+      "loss": 0.7835,
+      "step": 1385
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9714242504480233,
+      "learning_rate": 9.568836729603495e-06,
+      "loss": 0.5747,
+      "step": 1386
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.600528143316716,
+      "learning_rate": 9.56808043994241e-06,
+      "loss": 0.577,
+      "step": 1387
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.006816380734164,
+      "learning_rate": 9.567323517511617e-06,
+      "loss": 0.5495,
+      "step": 1388
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1614670648200107,
+      "learning_rate": 9.566565962415958e-06,
+      "loss": 0.6416,
+      "step": 1389
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.372382590415793,
+      "learning_rate": 9.565807774760376e-06,
+      "loss": 0.4682,
+      "step": 1390
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0571712473570565,
+      "learning_rate": 9.56504895464989e-06,
+      "loss": 0.5685,
+      "step": 1391
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1470730720800186,
+      "learning_rate": 9.564289502189615e-06,
+      "loss": 0.5494,
+      "step": 1392
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.879077283106543,
+      "learning_rate": 9.563529417484747e-06,
+      "loss": 0.6456,
+      "step": 1393
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.85725869169753,
+      "learning_rate": 9.562768700640575e-06,
+      "loss": 0.5873,
+      "step": 1394
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1230601623032794,
+      "learning_rate": 9.562007351762473e-06,
+      "loss": 0.5205,
+      "step": 1395
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.263549718980947,
+      "learning_rate": 9.561245370955903e-06,
+      "loss": 0.531,
+      "step": 1396
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6700140466273794,
+      "learning_rate": 9.560482758326414e-06,
+      "loss": 0.516,
+      "step": 1397
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.890852254087225,
+      "learning_rate": 9.559719513979645e-06,
+      "loss": 0.4899,
+      "step": 1398
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9632468246607881,
+      "learning_rate": 9.558955638021314e-06,
+      "loss": 0.5203,
+      "step": 1399
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8608194044045343,
+      "learning_rate": 9.558191130557242e-06,
+      "loss": 0.4981,
+      "step": 1400
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.941016713952777,
+      "learning_rate": 9.557425991693323e-06,
+      "loss": 0.7146,
+      "step": 1401
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.479110677104224,
+      "learning_rate": 9.556660221535545e-06,
+      "loss": 0.5136,
+      "step": 1402
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.854634248025494,
+      "learning_rate": 9.55589382018998e-06,
+      "loss": 0.4831,
+      "step": 1403
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.821923908562663,
+      "learning_rate": 9.555126787762796e-06,
+      "loss": 0.5212,
+      "step": 1404
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2483254989009502,
+      "learning_rate": 9.554359124360236e-06,
+      "loss": 0.5665,
+      "step": 1405
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0033334961158134,
+      "learning_rate": 9.55359083008864e-06,
+      "loss": 0.5693,
+      "step": 1406
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9040616941878393,
+      "learning_rate": 9.55282190505443e-06,
+      "loss": 0.7642,
+      "step": 1407
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4685571795671053,
+      "learning_rate": 9.552052349364118e-06,
+      "loss": 0.5584,
+      "step": 1408
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1040518367050782,
+      "learning_rate": 9.551282163124304e-06,
+      "loss": 0.5804,
+      "step": 1409
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6710792422808451,
+      "learning_rate": 9.550511346441674e-06,
+      "loss": 0.5671,
+      "step": 1410
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.191627122312166,
+      "learning_rate": 9.549739899422998e-06,
+      "loss": 0.4611,
+      "step": 1411
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.705327642401606,
+      "learning_rate": 9.548967822175142e-06,
+      "loss": 0.5022,
+      "step": 1412
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8347727946182772,
+      "learning_rate": 9.548195114805047e-06,
+      "loss": 0.5471,
+      "step": 1413
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9764802925520053,
+      "learning_rate": 9.547421777419756e-06,
+      "loss": 0.4352,
+      "step": 1414
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6946602197003582,
+      "learning_rate": 9.546647810126388e-06,
+      "loss": 0.4884,
+      "step": 1415
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9233943474299515,
+      "learning_rate": 9.545873213032151e-06,
+      "loss": 0.6501,
+      "step": 1416
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1000008561961976,
+      "learning_rate": 9.545097986244345e-06,
+      "loss": 0.6233,
+      "step": 1417
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.194788739905355,
+      "learning_rate": 9.544322129870354e-06,
+      "loss": 0.524,
+      "step": 1418
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1592055036121645,
+      "learning_rate": 9.543545644017646e-06,
+      "loss": 0.5565,
+      "step": 1419
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0296609059043385,
+      "learning_rate": 9.542768528793784e-06,
+      "loss": 0.5466,
+      "step": 1420
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.729969446737493,
+      "learning_rate": 9.541990784306414e-06,
+      "loss": 0.5495,
+      "step": 1421
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7698462955069232,
+      "learning_rate": 9.541212410663266e-06,
+      "loss": 0.6134,
+      "step": 1422
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.555090294658996,
+      "learning_rate": 9.54043340797216e-06,
+      "loss": 0.4947,
+      "step": 1423
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9366223874743607,
+      "learning_rate": 9.539653776341007e-06,
+      "loss": 0.5837,
+      "step": 1424
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5921490088478234,
+      "learning_rate": 9.538873515877797e-06,
+      "loss": 0.4956,
+      "step": 1425
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9509843780336904,
+      "learning_rate": 9.538092626690613e-06,
+      "loss": 0.7544,
+      "step": 1426
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1505978970517656,
+      "learning_rate": 9.537311108887626e-06,
+      "loss": 0.5252,
+      "step": 1427
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0597680570537076,
+      "learning_rate": 9.536528962577092e-06,
+      "loss": 0.5424,
+      "step": 1428
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8016401978593481,
+      "learning_rate": 9.535746187867349e-06,
+      "loss": 0.5543,
+      "step": 1429
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.704742676852404,
+      "learning_rate": 9.53496278486683e-06,
+      "loss": 0.5873,
+      "step": 1430
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0903093995692044,
+      "learning_rate": 9.534178753684054e-06,
+      "loss": 0.53,
+      "step": 1431
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.060378143366305,
+      "learning_rate": 9.533394094427619e-06,
+      "loss": 0.5089,
+      "step": 1432
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9597813868398153,
+      "learning_rate": 9.53260880720622e-06,
+      "loss": 0.5024,
+      "step": 1433
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0713033160538337,
+      "learning_rate": 9.531822892128637e-06,
+      "loss": 0.577,
+      "step": 1434
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3398236723030874,
+      "learning_rate": 9.531036349303729e-06,
+      "loss": 0.471,
+      "step": 1435
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5449609511427382,
+      "learning_rate": 9.530249178840452e-06,
+      "loss": 0.4751,
+      "step": 1436
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2202636298457588,
+      "learning_rate": 9.529461380847842e-06,
+      "loss": 0.5656,
+      "step": 1437
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0479970582076232,
+      "learning_rate": 9.528672955435027e-06,
+      "loss": 0.5446,
+      "step": 1438
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9239044730878956,
+      "learning_rate": 9.527883902711219e-06,
+      "loss": 0.5564,
+      "step": 1439
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6887756653610342,
+      "learning_rate": 9.527094222785717e-06,
+      "loss": 0.4266,
+      "step": 1440
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.487628035495983,
+      "learning_rate": 9.526303915767906e-06,
+      "loss": 0.5408,
+      "step": 1441
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.7064563976933,
+      "learning_rate": 9.525512981767263e-06,
+      "loss": 0.5075,
+      "step": 1442
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9369214160079387,
+      "learning_rate": 9.524721420893344e-06,
+      "loss": 0.543,
+      "step": 1443
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.077238591866259,
+      "learning_rate": 9.523929233255797e-06,
+      "loss": 0.4671,
+      "step": 1444
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7772843833931513,
+      "learning_rate": 9.523136418964356e-06,
+      "loss": 0.5712,
+      "step": 1445
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2758023744504237,
+      "learning_rate": 9.522342978128843e-06,
+      "loss": 0.4608,
+      "step": 1446
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9311861858877077,
+      "learning_rate": 9.521548910859163e-06,
+      "loss": 0.5748,
+      "step": 1447
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.909715683777971,
+      "learning_rate": 9.520754217265311e-06,
+      "loss": 0.556,
+      "step": 1448
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1421173885534506,
+      "learning_rate": 9.519958897457368e-06,
+      "loss": 0.5377,
+      "step": 1449
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.109948896939215,
+      "learning_rate": 9.519162951545501e-06,
+      "loss": 0.4631,
+      "step": 1450
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2241729431217605,
+      "learning_rate": 9.518366379639968e-06,
+      "loss": 0.6092,
+      "step": 1451
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6252645290041907,
+      "learning_rate": 9.517569181851103e-06,
+      "loss": 0.4747,
+      "step": 1452
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2498382162419936,
+      "learning_rate": 9.516771358289339e-06,
+      "loss": 0.5081,
+      "step": 1453
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0888134535102534,
+      "learning_rate": 9.515972909065187e-06,
+      "loss": 0.7124,
+      "step": 1454
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9467670810506359,
+      "learning_rate": 9.515173834289253e-06,
+      "loss": 0.7239,
+      "step": 1455
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1855249702453428,
+      "learning_rate": 9.51437413407222e-06,
+      "loss": 0.6236,
+      "step": 1456
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8127158875825493,
+      "learning_rate": 9.513573808524864e-06,
+      "loss": 0.519,
+      "step": 1457
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.972085765145068,
+      "learning_rate": 9.512772857758044e-06,
+      "loss": 0.5843,
+      "step": 1458
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.112764304907531,
+      "learning_rate": 9.511971281882711e-06,
+      "loss": 0.5616,
+      "step": 1459
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.030054421483051,
+      "learning_rate": 9.511169081009897e-06,
+      "loss": 0.5482,
+      "step": 1460
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8287901603843935,
+      "learning_rate": 9.510366255250722e-06,
+      "loss": 0.5266,
+      "step": 1461
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.000496495801951,
+      "learning_rate": 9.509562804716396e-06,
+      "loss": 0.5519,
+      "step": 1462
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.775526543316991,
+      "learning_rate": 9.508758729518213e-06,
+      "loss": 0.6783,
+      "step": 1463
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7284959623174554,
+      "learning_rate": 9.50795402976755e-06,
+      "loss": 0.516,
+      "step": 1464
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.00937464269589,
+      "learning_rate": 9.507148705575876e-06,
+      "loss": 0.3977,
+      "step": 1465
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8227419757331833,
+      "learning_rate": 9.506342757054744e-06,
+      "loss": 0.5012,
+      "step": 1466
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.11901413287132,
+      "learning_rate": 9.505536184315793e-06,
+      "loss": 0.5069,
+      "step": 1467
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.53597531641787,
+      "learning_rate": 9.50472898747075e-06,
+      "loss": 0.5201,
+      "step": 1468
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8385979069973468,
+      "learning_rate": 9.50392116663143e-06,
+      "loss": 0.5532,
+      "step": 1469
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9340394447527698,
+      "learning_rate": 9.503112721909728e-06,
+      "loss": 0.5417,
+      "step": 1470
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8211635236483439,
+      "learning_rate": 9.502303653417631e-06,
+      "loss": 0.5537,
+      "step": 1471
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8567616213882414,
+      "learning_rate": 9.501493961267213e-06,
+      "loss": 0.6387,
+      "step": 1472
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.561795593439265,
+      "learning_rate": 9.500683645570632e-06,
+      "loss": 0.648,
+      "step": 1473
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1408629257592136,
+      "learning_rate": 9.499872706440132e-06,
+      "loss": 0.5465,
+      "step": 1474
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.781320494085114,
+      "learning_rate": 9.499061143988042e-06,
+      "loss": 0.5453,
+      "step": 1475
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9065847596129448,
+      "learning_rate": 9.498248958326783e-06,
+      "loss": 0.6311,
+      "step": 1476
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8454794441091658,
+      "learning_rate": 9.497436149568858e-06,
+      "loss": 0.6189,
+      "step": 1477
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.054435406232731,
+      "learning_rate": 9.496622717826855e-06,
+      "loss": 0.527,
+      "step": 1478
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.847704212112147,
+      "learning_rate": 9.495808663213454e-06,
+      "loss": 0.4713,
+      "step": 1479
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6161879655352474,
+      "learning_rate": 9.494993985841414e-06,
+      "loss": 0.5122,
+      "step": 1480
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6288866221277847,
+      "learning_rate": 9.494178685823586e-06,
+      "loss": 0.5416,
+      "step": 1481
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1751590584173246,
+      "learning_rate": 9.493362763272906e-06,
+      "loss": 0.531,
+      "step": 1482
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8942983798603406,
+      "learning_rate": 9.492546218302392e-06,
+      "loss": 0.536,
+      "step": 1483
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7793768760876623,
+      "learning_rate": 9.491729051025157e-06,
+      "loss": 0.6665,
+      "step": 1484
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9606206687667782,
+      "learning_rate": 9.49091126155439e-06,
+      "loss": 0.5227,
+      "step": 1485
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0015944592733046,
+      "learning_rate": 9.490092850003372e-06,
+      "loss": 0.5494,
+      "step": 1486
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.90739247745898,
+      "learning_rate": 9.489273816485472e-06,
+      "loss": 0.4668,
+      "step": 1487
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.630449565138997,
+      "learning_rate": 9.488454161114138e-06,
+      "loss": 0.523,
+      "step": 1488
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.257967856576836,
+      "learning_rate": 9.487633884002914e-06,
+      "loss": 0.5064,
+      "step": 1489
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.06221848584041,
+      "learning_rate": 9.48681298526542e-06,
+      "loss": 0.6519,
+      "step": 1490
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5631694786725043,
+      "learning_rate": 9.48599146501537e-06,
+      "loss": 0.5348,
+      "step": 1491
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5116495517499198,
+      "learning_rate": 9.485169323366556e-06,
+      "loss": 0.6165,
+      "step": 1492
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.737551992032248,
+      "learning_rate": 9.484346560432867e-06,
+      "loss": 0.5424,
+      "step": 1493
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.734580512119994,
+      "learning_rate": 9.48352317632827e-06,
+      "loss": 0.5018,
+      "step": 1494
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.415610930314321,
+      "learning_rate": 9.482699171166816e-06,
+      "loss": 0.5908,
+      "step": 1495
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8142746571788346,
+      "learning_rate": 9.481874545062651e-06,
+      "loss": 0.4933,
+      "step": 1496
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.771807041567568,
+      "learning_rate": 9.48104929813e-06,
+      "loss": 0.5534,
+      "step": 1497
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9252584612298708,
+      "learning_rate": 9.480223430483176e-06,
+      "loss": 0.756,
+      "step": 1498
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1829521908306595,
+      "learning_rate": 9.47939694223658e-06,
+      "loss": 0.5595,
+      "step": 1499
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.321874578093091,
+      "learning_rate": 9.478569833504694e-06,
+      "loss": 0.4621,
+      "step": 1500
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2840611952651813,
+      "learning_rate": 9.47774210440209e-06,
+      "loss": 0.5139,
+      "step": 1501
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6475694356517705,
+      "learning_rate": 9.476913755043427e-06,
+      "loss": 0.5039,
+      "step": 1502
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7602111456002916,
+      "learning_rate": 9.476084785543444e-06,
+      "loss": 0.5565,
+      "step": 1503
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0025178336996734,
+      "learning_rate": 9.475255196016972e-06,
+      "loss": 0.5418,
+      "step": 1504
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7505514269438753,
+      "learning_rate": 9.474424986578928e-06,
+      "loss": 0.4238,
+      "step": 1505
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.006269175920709,
+      "learning_rate": 9.473594157344307e-06,
+      "loss": 0.6045,
+      "step": 1506
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0209803761513654,
+      "learning_rate": 9.4727627084282e-06,
+      "loss": 0.5503,
+      "step": 1507
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4692240938051726,
+      "learning_rate": 9.471930639945777e-06,
+      "loss": 0.5782,
+      "step": 1508
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.605852202185983,
+      "learning_rate": 9.471097952012296e-06,
+      "loss": 0.5746,
+      "step": 1509
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.082110555252948,
+      "learning_rate": 9.4702646447431e-06,
+      "loss": 0.6455,
+      "step": 1510
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.500299430628737,
+      "learning_rate": 9.46943071825362e-06,
+      "loss": 0.4985,
+      "step": 1511
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6393743041118873,
+      "learning_rate": 9.468596172659372e-06,
+      "loss": 0.6134,
+      "step": 1512
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.809760471963379,
+      "learning_rate": 9.467761008075957e-06,
+      "loss": 0.4374,
+      "step": 1513
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.451962105545382,
+      "learning_rate": 9.466925224619059e-06,
+      "loss": 0.5095,
+      "step": 1514
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.532438955618862,
+      "learning_rate": 9.466088822404454e-06,
+      "loss": 0.552,
+      "step": 1515
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7574621909644783,
+      "learning_rate": 9.465251801547998e-06,
+      "loss": 0.4963,
+      "step": 1516
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.929622554690285,
+      "learning_rate": 9.464414162165635e-06,
+      "loss": 0.486,
+      "step": 1517
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.445082309452605,
+      "learning_rate": 9.463575904373397e-06,
+      "loss": 0.5817,
+      "step": 1518
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3133468491973166,
+      "learning_rate": 9.462737028287398e-06,
+      "loss": 0.5551,
+      "step": 1519
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6198102076375613,
+      "learning_rate": 9.461897534023838e-06,
+      "loss": 0.5179,
+      "step": 1520
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.116514234181379,
+      "learning_rate": 9.461057421699004e-06,
+      "loss": 0.6356,
+      "step": 1521
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.349094653833558,
+      "learning_rate": 9.460216691429271e-06,
+      "loss": 0.5676,
+      "step": 1522
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9766733759532904,
+      "learning_rate": 9.459375343331091e-06,
+      "loss": 0.7801,
+      "step": 1523
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6030340735159212,
+      "learning_rate": 9.458533377521014e-06,
+      "loss": 0.517,
+      "step": 1524
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6988472470307165,
+      "learning_rate": 9.457690794115664e-06,
+      "loss": 0.5458,
+      "step": 1525
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2751385754829427,
+      "learning_rate": 9.456847593231758e-06,
+      "loss": 0.5583,
+      "step": 1526
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.385446649891306,
+      "learning_rate": 9.456003774986096e-06,
+      "loss": 0.6169,
+      "step": 1527
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.169669422604865,
+      "learning_rate": 9.45515933949556e-06,
+      "loss": 0.5129,
+      "step": 1528
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7805941627597726,
+      "learning_rate": 9.454314286877127e-06,
+      "loss": 0.569,
+      "step": 1529
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6358505985341094,
+      "learning_rate": 9.45346861724785e-06,
+      "loss": 0.4823,
+      "step": 1530
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7894858945504952,
+      "learning_rate": 9.45262233072487e-06,
+      "loss": 0.5521,
+      "step": 1531
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.652156941345048,
+      "learning_rate": 9.451775427425417e-06,
+      "loss": 0.5548,
+      "step": 1532
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8043718676510645,
+      "learning_rate": 9.450927907466803e-06,
+      "loss": 0.4128,
+      "step": 1533
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9885338788116866,
+      "learning_rate": 9.450079770966424e-06,
+      "loss": 0.4633,
+      "step": 1534
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.986915889820981,
+      "learning_rate": 9.449231018041769e-06,
+      "loss": 0.4737,
+      "step": 1535
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0226185998553654,
+      "learning_rate": 9.448381648810403e-06,
+      "loss": 0.4827,
+      "step": 1536
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5378208451849988,
+      "learning_rate": 9.447531663389982e-06,
+      "loss": 0.5685,
+      "step": 1537
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1989547700602237,
+      "learning_rate": 9.446681061898244e-06,
+      "loss": 0.5481,
+      "step": 1538
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7461483412079766,
+      "learning_rate": 9.445829844453017e-06,
+      "loss": 0.5131,
+      "step": 1539
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0038722623007386,
+      "learning_rate": 9.444978011172207e-06,
+      "loss": 0.7803,
+      "step": 1540
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7657144239580935,
+      "learning_rate": 9.444125562173816e-06,
+      "loss": 0.5738,
+      "step": 1541
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.977315713817142,
+      "learning_rate": 9.443272497575922e-06,
+      "loss": 0.5779,
+      "step": 1542
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7349565433400973,
+      "learning_rate": 9.442418817496689e-06,
+      "loss": 0.5682,
+      "step": 1543
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.072917492730993,
+      "learning_rate": 9.441564522054372e-06,
+      "loss": 0.4662,
+      "step": 1544
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9603366244114617,
+      "learning_rate": 9.440709611367308e-06,
+      "loss": 0.5404,
+      "step": 1545
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3058408812745883,
+      "learning_rate": 9.439854085553914e-06,
+      "loss": 0.517,
+      "step": 1546
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.942141637564404,
+      "learning_rate": 9.438997944732705e-06,
+      "loss": 0.5759,
+      "step": 1547
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.888176375381119,
+      "learning_rate": 9.438141189022267e-06,
+      "loss": 0.5845,
+      "step": 1548
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7958421091183836,
+      "learning_rate": 9.43728381854128e-06,
+      "loss": 0.5848,
+      "step": 1549
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8418586436738982,
+      "learning_rate": 9.436425833408509e-06,
+      "loss": 0.5457,
+      "step": 1550
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.312040209207688,
+      "learning_rate": 9.435567233742799e-06,
+      "loss": 0.4754,
+      "step": 1551
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8213631686188005,
+      "learning_rate": 9.434708019663085e-06,
+      "loss": 0.595,
+      "step": 1552
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6958279726854661,
+      "learning_rate": 9.433848191288384e-06,
+      "loss": 0.5166,
+      "step": 1553
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6039608029137833,
+      "learning_rate": 9.432987748737798e-06,
+      "loss": 0.595,
+      "step": 1554
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.444212090530702,
+      "learning_rate": 9.432126692130518e-06,
+      "loss": 0.6208,
+      "step": 1555
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.914056758420974,
+      "learning_rate": 9.431265021585816e-06,
+      "loss": 0.5174,
+      "step": 1556
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5720779055197163,
+      "learning_rate": 9.430402737223051e-06,
+      "loss": 0.5405,
+      "step": 1557
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8353091615135224,
+      "learning_rate": 9.429539839161665e-06,
+      "loss": 0.6386,
+      "step": 1558
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6682560882299255,
+      "learning_rate": 9.428676327521189e-06,
+      "loss": 0.5587,
+      "step": 1559
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2528850932549336,
+      "learning_rate": 9.427812202421236e-06,
+      "loss": 0.5145,
+      "step": 1560
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4866252642879645,
+      "learning_rate": 9.426947463981502e-06,
+      "loss": 0.6121,
+      "step": 1561
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.748776826380648,
+      "learning_rate": 9.426082112321773e-06,
+      "loss": 0.6093,
+      "step": 1562
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.209771080896768,
+      "learning_rate": 9.425216147561916e-06,
+      "loss": 0.5364,
+      "step": 1563
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.870438844556082,
+      "learning_rate": 9.424349569821884e-06,
+      "loss": 0.7473,
+      "step": 1564
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.060885811366647,
+      "learning_rate": 9.423482379221717e-06,
+      "loss": 0.4351,
+      "step": 1565
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7458260772542664,
+      "learning_rate": 9.422614575881536e-06,
+      "loss": 0.528,
+      "step": 1566
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8680515003621059,
+      "learning_rate": 9.421746159921553e-06,
+      "loss": 0.561,
+      "step": 1567
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.810134387043382,
+      "learning_rate": 9.420877131462053e-06,
+      "loss": 0.5177,
+      "step": 1568
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.281103211342526,
+      "learning_rate": 9.420007490623422e-06,
+      "loss": 0.5588,
+      "step": 1569
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2006922877456607,
+      "learning_rate": 9.419137237526116e-06,
+      "loss": 0.5113,
+      "step": 1570
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.137370959466019,
+      "learning_rate": 9.418266372290689e-06,
+      "loss": 0.5734,
+      "step": 1571
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.390238091119377,
+      "learning_rate": 9.417394895037768e-06,
+      "loss": 0.5692,
+      "step": 1572
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7867108325303955,
+      "learning_rate": 9.416522805888072e-06,
+      "loss": 0.5484,
+      "step": 1573
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7287156106410118,
+      "learning_rate": 9.415650104962399e-06,
+      "loss": 0.5086,
+      "step": 1574
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.614209509115682,
+      "learning_rate": 9.414776792381639e-06,
+      "loss": 0.4683,
+      "step": 1575
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8526686678311381,
+      "learning_rate": 9.413902868266764e-06,
+      "loss": 0.6403,
+      "step": 1576
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1436525981823706,
+      "learning_rate": 9.413028332738827e-06,
+      "loss": 0.6158,
+      "step": 1577
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.997408998716668,
+      "learning_rate": 9.41215318591897e-06,
+      "loss": 0.557,
+      "step": 1578
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.682274264355876,
+      "learning_rate": 9.411277427928419e-06,
+      "loss": 0.5273,
+      "step": 1579
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9915799640097176,
+      "learning_rate": 9.410401058888482e-06,
+      "loss": 0.5122,
+      "step": 1580
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.863239011162514,
+      "learning_rate": 9.409524078920553e-06,
+      "loss": 0.5355,
+      "step": 1581
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2768149552548866,
+      "learning_rate": 9.408646488146113e-06,
+      "loss": 0.5698,
+      "step": 1582
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0316930813079797,
+      "learning_rate": 9.407768286686726e-06,
+      "loss": 0.4587,
+      "step": 1583
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.185120014444133,
+      "learning_rate": 9.40688947466404e-06,
+      "loss": 0.522,
+      "step": 1584
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8355065585628945,
+      "learning_rate": 9.406010052199786e-06,
+      "loss": 0.5596,
+      "step": 1585
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7964244886665925,
+      "learning_rate": 9.405130019415782e-06,
+      "loss": 0.575,
+      "step": 1586
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6286436422823765,
+      "learning_rate": 9.404249376433932e-06,
+      "loss": 0.5221,
+      "step": 1587
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9629959819228031,
+      "learning_rate": 9.403368123376222e-06,
+      "loss": 0.5028,
+      "step": 1588
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8467952421348366,
+      "learning_rate": 9.402486260364721e-06,
+      "loss": 0.5833,
+      "step": 1589
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2950549190414185,
+      "learning_rate": 9.401603787521584e-06,
+      "loss": 0.5262,
+      "step": 1590
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9048919972223175,
+      "learning_rate": 9.400720704969055e-06,
+      "loss": 0.4974,
+      "step": 1591
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.752228683873074,
+      "learning_rate": 9.399837012829456e-06,
+      "loss": 0.5506,
+      "step": 1592
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.9953779798529143,
+      "learning_rate": 9.398952711225195e-06,
+      "loss": 0.4445,
+      "step": 1593
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2018795280289054,
+      "learning_rate": 9.398067800278767e-06,
+      "loss": 0.4596,
+      "step": 1594
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6681465740057362,
+      "learning_rate": 9.397182280112748e-06,
+      "loss": 0.463,
+      "step": 1595
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5427933492272992,
+      "learning_rate": 9.396296150849804e-06,
+      "loss": 0.4764,
+      "step": 1596
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0167461007259617,
+      "learning_rate": 9.395409412612677e-06,
+      "loss": 0.5193,
+      "step": 1597
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7445542085222687,
+      "learning_rate": 9.394522065524199e-06,
+      "loss": 0.5209,
+      "step": 1598
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.570325645842292,
+      "learning_rate": 9.393634109707286e-06,
+      "loss": 0.5666,
+      "step": 1599
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.62731632193813,
+      "learning_rate": 9.392745545284938e-06,
+      "loss": 0.4471,
+      "step": 1600
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9883075038456708,
+      "learning_rate": 9.391856372380238e-06,
+      "loss": 0.5026,
+      "step": 1601
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2141677244465354,
+      "learning_rate": 9.390966591116351e-06,
+      "loss": 0.4177,
+      "step": 1602
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9429072382062087,
+      "learning_rate": 9.390076201616536e-06,
+      "loss": 0.5277,
+      "step": 1603
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9524843693486186,
+      "learning_rate": 9.389185204004123e-06,
+      "loss": 0.776,
+      "step": 1604
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4436621564967007,
+      "learning_rate": 9.388293598402538e-06,
+      "loss": 0.5519,
+      "step": 1605
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.580810861388631,
+      "learning_rate": 9.387401384935282e-06,
+      "loss": 0.4495,
+      "step": 1606
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.660961280407153,
+      "learning_rate": 9.386508563725947e-06,
+      "loss": 0.486,
+      "step": 1607
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5498932179204852,
+      "learning_rate": 9.385615134898206e-06,
+      "loss": 0.4832,
+      "step": 1608
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.656627323274859,
+      "learning_rate": 9.384721098575815e-06,
+      "loss": 0.6313,
+      "step": 1609
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3286936415258985,
+      "learning_rate": 9.383826454882618e-06,
+      "loss": 0.4837,
+      "step": 1610
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.242165112596168,
+      "learning_rate": 9.38293120394254e-06,
+      "loss": 0.6287,
+      "step": 1611
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9618078407314004,
+      "learning_rate": 9.38203534587959e-06,
+      "loss": 0.5228,
+      "step": 1612
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.592663318979876,
+      "learning_rate": 9.381138880817862e-06,
+      "loss": 0.5712,
+      "step": 1613
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3376267464565847,
+      "learning_rate": 9.380241808881536e-06,
+      "loss": 0.6743,
+      "step": 1614
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8649294539568677,
+      "learning_rate": 9.379344130194873e-06,
+      "loss": 0.6241,
+      "step": 1615
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1033972423351037,
+      "learning_rate": 9.378445844882222e-06,
+      "loss": 0.5421,
+      "step": 1616
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0276002128068393,
+      "learning_rate": 9.377546953068008e-06,
+      "loss": 0.5465,
+      "step": 1617
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.949817547082076,
+      "learning_rate": 9.37664745487675e-06,
+      "loss": 0.4586,
+      "step": 1618
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.78858636720677,
+      "learning_rate": 9.375747350433044e-06,
+      "loss": 0.6349,
+      "step": 1619
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7454502884578886,
+      "learning_rate": 9.374846639861573e-06,
+      "loss": 0.492,
+      "step": 1620
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6503649648430325,
+      "learning_rate": 9.373945323287102e-06,
+      "loss": 0.5457,
+      "step": 1621
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8499920739614937,
+      "learning_rate": 9.373043400834482e-06,
+      "loss": 0.6042,
+      "step": 1622
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0167520935463745,
+      "learning_rate": 9.37214087262865e-06,
+      "loss": 0.5347,
+      "step": 1623
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5100709617349641,
+      "learning_rate": 9.37123773879462e-06,
+      "loss": 0.5089,
+      "step": 1624
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8008478191103365,
+      "learning_rate": 9.370333999457498e-06,
+      "loss": 0.5735,
+      "step": 1625
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3598057277155364,
+      "learning_rate": 9.369429654742463e-06,
+      "loss": 0.6149,
+      "step": 1626
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0311060968027648,
+      "learning_rate": 9.368524704774793e-06,
+      "loss": 0.5339,
+      "step": 1627
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6588586436955497,
+      "learning_rate": 9.367619149679836e-06,
+      "loss": 0.4918,
+      "step": 1628
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5312167814682616,
+      "learning_rate": 9.366712989583031e-06,
+      "loss": 0.4911,
+      "step": 1629
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6772441422717488,
+      "learning_rate": 9.3658062246099e-06,
+      "loss": 0.5281,
+      "step": 1630
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2196791581857616,
+      "learning_rate": 9.364898854886044e-06,
+      "loss": 0.5219,
+      "step": 1631
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.690329251618206,
+      "learning_rate": 9.363990880537157e-06,
+      "loss": 0.4907,
+      "step": 1632
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1773979013974643,
+      "learning_rate": 9.363082301689008e-06,
+      "loss": 0.702,
+      "step": 1633
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9129287728075911,
+      "learning_rate": 9.362173118467455e-06,
+      "loss": 0.7323,
+      "step": 1634
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.832091081125249,
+      "learning_rate": 9.361263330998436e-06,
+      "loss": 0.5309,
+      "step": 1635
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.845024142967831,
+      "learning_rate": 9.360352939407977e-06,
+      "loss": 0.5611,
+      "step": 1636
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7155968780584063,
+      "learning_rate": 9.359441943822185e-06,
+      "loss": 0.5773,
+      "step": 1637
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.940951756223291,
+      "learning_rate": 9.358530344367247e-06,
+      "loss": 0.4815,
+      "step": 1638
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4746351623637635,
+      "learning_rate": 9.357618141169444e-06,
+      "loss": 0.5174,
+      "step": 1639
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.010168292301214,
+      "learning_rate": 9.35670533435513e-06,
+      "loss": 0.5211,
+      "step": 1640
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7887664725592907,
+      "learning_rate": 9.355791924050746e-06,
+      "loss": 0.503,
+      "step": 1641
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9662081867662127,
+      "learning_rate": 9.35487791038282e-06,
+      "loss": 0.5247,
+      "step": 1642
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9667746818100165,
+      "learning_rate": 9.35396329347796e-06,
+      "loss": 0.4986,
+      "step": 1643
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.755291694452646,
+      "learning_rate": 9.35304807346286e-06,
+      "loss": 0.6297,
+      "step": 1644
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5268666469571373,
+      "learning_rate": 9.352132250464294e-06,
+      "loss": 0.5353,
+      "step": 1645
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.182152519595133,
+      "learning_rate": 9.351215824609123e-06,
+      "loss": 0.5012,
+      "step": 1646
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.071807968427525,
+      "learning_rate": 9.350298796024288e-06,
+      "loss": 0.5629,
+      "step": 1647
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7609081932742119,
+      "learning_rate": 9.349381164836818e-06,
+      "loss": 0.527,
+      "step": 1648
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.03081093691396,
+      "learning_rate": 9.348462931173824e-06,
+      "loss": 0.5172,
+      "step": 1649
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0200849732321906,
+      "learning_rate": 9.347544095162495e-06,
+      "loss": 0.6014,
+      "step": 1650
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2375401706846145,
+      "learning_rate": 9.346624656930113e-06,
+      "loss": 0.5171,
+      "step": 1651
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.598803077642052,
+      "learning_rate": 9.345704616604036e-06,
+      "loss": 0.5759,
+      "step": 1652
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1723427389105816,
+      "learning_rate": 9.344783974311709e-06,
+      "loss": 0.469,
+      "step": 1653
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7788756360752505,
+      "learning_rate": 9.343862730180657e-06,
+      "loss": 0.5035,
+      "step": 1654
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.489989827105883,
+      "learning_rate": 9.342940884338492e-06,
+      "loss": 0.6048,
+      "step": 1655
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5205891378341054,
+      "learning_rate": 9.342018436912908e-06,
+      "loss": 0.5814,
+      "step": 1656
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1676788234413187,
+      "learning_rate": 9.341095388031684e-06,
+      "loss": 0.5329,
+      "step": 1657
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9402058529923814,
+      "learning_rate": 9.340171737822677e-06,
+      "loss": 0.5218,
+      "step": 1658
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.881072234578272,
+      "learning_rate": 9.339247486413832e-06,
+      "loss": 0.5825,
+      "step": 1659
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9175369724697766,
+      "learning_rate": 9.338322633933178e-06,
+      "loss": 0.4724,
+      "step": 1660
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 20.418053166467853,
+      "learning_rate": 9.337397180508825e-06,
+      "loss": 0.4652,
+      "step": 1661
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9262230133338871,
+      "learning_rate": 9.336471126268965e-06,
+      "loss": 0.5184,
+      "step": 1662
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0559451459029967,
+      "learning_rate": 9.335544471341876e-06,
+      "loss": 0.4706,
+      "step": 1663
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.948126151951804,
+      "learning_rate": 9.334617215855916e-06,
+      "loss": 0.5583,
+      "step": 1664
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0959032273625624,
+      "learning_rate": 9.33368935993953e-06,
+      "loss": 0.5169,
+      "step": 1665
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7551353465570954,
+      "learning_rate": 9.332760903721248e-06,
+      "loss": 0.5775,
+      "step": 1666
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.015298559665234,
+      "learning_rate": 9.331831847329674e-06,
+      "loss": 0.5793,
+      "step": 1667
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7841466972445137,
+      "learning_rate": 9.3309021908935e-06,
+      "loss": 0.5258,
+      "step": 1668
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5071785752632825,
+      "learning_rate": 9.329971934541508e-06,
+      "loss": 0.5644,
+      "step": 1669
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4689997155356616,
+      "learning_rate": 9.329041078402553e-06,
+      "loss": 0.555,
+      "step": 1670
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8628550108931088,
+      "learning_rate": 9.328109622605579e-06,
+      "loss": 0.5878,
+      "step": 1671
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.83086627279445,
+      "learning_rate": 9.327177567279608e-06,
+      "loss": 0.5181,
+      "step": 1672
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.890704231586985,
+      "learning_rate": 9.326244912553749e-06,
+      "loss": 0.7284,
+      "step": 1673
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.85034561764086,
+      "learning_rate": 9.325311658557195e-06,
+      "loss": 0.5443,
+      "step": 1674
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7523869508674763,
+      "learning_rate": 9.32437780541922e-06,
+      "loss": 0.5438,
+      "step": 1675
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.120203311956958,
+      "learning_rate": 9.323443353269179e-06,
+      "loss": 0.5628,
+      "step": 1676
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8949944133321155,
+      "learning_rate": 9.322508302236515e-06,
+      "loss": 0.4906,
+      "step": 1677
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6104717066554324,
+      "learning_rate": 9.321572652450749e-06,
+      "loss": 0.6067,
+      "step": 1678
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.907027282582565,
+      "learning_rate": 9.320636404041487e-06,
+      "loss": 0.5042,
+      "step": 1679
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.556490597953133,
+      "learning_rate": 9.31969955713842e-06,
+      "loss": 0.5551,
+      "step": 1680
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7604952199030588,
+      "learning_rate": 9.318762111871318e-06,
+      "loss": 0.567,
+      "step": 1681
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9717231670496491,
+      "learning_rate": 9.317824068370036e-06,
+      "loss": 0.5675,
+      "step": 1682
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7838151625845646,
+      "learning_rate": 9.316885426764512e-06,
+      "loss": 0.4793,
+      "step": 1683
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1260110021628784,
+      "learning_rate": 9.315946187184765e-06,
+      "loss": 0.4534,
+      "step": 1684
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9524502686507368,
+      "learning_rate": 9.315006349760903e-06,
+      "loss": 0.7155,
+      "step": 1685
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6977042332185168,
+      "learning_rate": 9.314065914623106e-06,
+      "loss": 0.5344,
+      "step": 1686
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7123875041222583,
+      "learning_rate": 9.313124881901648e-06,
+      "loss": 0.4912,
+      "step": 1687
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8031208988745432,
+      "learning_rate": 9.312183251726876e-06,
+      "loss": 0.6784,
+      "step": 1688
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6460891424841892,
+      "learning_rate": 9.311241024229227e-06,
+      "loss": 0.5118,
+      "step": 1689
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.857826997164157,
+      "learning_rate": 9.31029819953922e-06,
+      "loss": 0.5509,
+      "step": 1690
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5030341662851896,
+      "learning_rate": 9.309354777787452e-06,
+      "loss": 0.4867,
+      "step": 1691
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4581579820689228,
+      "learning_rate": 9.308410759104606e-06,
+      "loss": 0.4773,
+      "step": 1692
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1294693420619235,
+      "learning_rate": 9.307466143621449e-06,
+      "loss": 0.4801,
+      "step": 1693
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8206610963410648,
+      "learning_rate": 9.306520931468828e-06,
+      "loss": 0.5204,
+      "step": 1694
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7567220059452366,
+      "learning_rate": 9.305575122777672e-06,
+      "loss": 0.605,
+      "step": 1695
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7471833702748962,
+      "learning_rate": 9.304628717678997e-06,
+      "loss": 0.5759,
+      "step": 1696
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1298582970523965,
+      "learning_rate": 9.303681716303896e-06,
+      "loss": 0.5313,
+      "step": 1697
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.249811427482779,
+      "learning_rate": 9.302734118783551e-06,
+      "loss": 0.569,
+      "step": 1698
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0545328509448202,
+      "learning_rate": 9.30178592524922e-06,
+      "loss": 0.8219,
+      "step": 1699
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7480512481127208,
+      "learning_rate": 9.300837135832249e-06,
+      "loss": 0.6424,
+      "step": 1700
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.671875010015276,
+      "learning_rate": 9.299887750664062e-06,
+      "loss": 0.5655,
+      "step": 1701
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.140857094371549,
+      "learning_rate": 9.298937769876168e-06,
+      "loss": 0.5079,
+      "step": 1702
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9804452311861018,
+      "learning_rate": 9.29798719360016e-06,
+      "loss": 0.7256,
+      "step": 1703
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1748119547094826,
+      "learning_rate": 9.297036021967709e-06,
+      "loss": 0.4758,
+      "step": 1704
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.173287989562393,
+      "learning_rate": 9.296084255110574e-06,
+      "loss": 0.6222,
+      "step": 1705
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0189381478411104,
+      "learning_rate": 9.295131893160591e-06,
+      "loss": 0.6305,
+      "step": 1706
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6480246533419487,
+      "learning_rate": 9.294178936249682e-06,
+      "loss": 0.5096,
+      "step": 1707
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.033585701299999,
+      "learning_rate": 9.29322538450985e-06,
+      "loss": 0.5171,
+      "step": 1708
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6493562347377557,
+      "learning_rate": 9.292271238073182e-06,
+      "loss": 0.5271,
+      "step": 1709
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5186542710373736,
+      "learning_rate": 9.291316497071847e-06,
+      "loss": 0.4936,
+      "step": 1710
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.8299734116166455,
+      "learning_rate": 9.290361161638093e-06,
+      "loss": 0.4407,
+      "step": 1711
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0904959960450777,
+      "learning_rate": 9.289405231904255e-06,
+      "loss": 0.4839,
+      "step": 1712
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1669664535299336,
+      "learning_rate": 9.288448708002743e-06,
+      "loss": 0.5695,
+      "step": 1713
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8587318527713061,
+      "learning_rate": 9.287491590066064e-06,
+      "loss": 0.5323,
+      "step": 1714
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7633797237150333,
+      "learning_rate": 9.286533878226789e-06,
+      "loss": 0.6068,
+      "step": 1715
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2555066996358035,
+      "learning_rate": 9.285575572617586e-06,
+      "loss": 0.5615,
+      "step": 1716
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.949895636328444,
+      "learning_rate": 9.284616673371196e-06,
+      "loss": 0.6067,
+      "step": 1717
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6641944382141005,
+      "learning_rate": 9.283657180620446e-06,
+      "loss": 0.5577,
+      "step": 1718
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7511235280427524,
+      "learning_rate": 9.282697094498245e-06,
+      "loss": 0.6046,
+      "step": 1719
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.414553788212881,
+      "learning_rate": 9.281736415137586e-06,
+      "loss": 0.462,
+      "step": 1720
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6693156003841387,
+      "learning_rate": 9.280775142671539e-06,
+      "loss": 0.4884,
+      "step": 1721
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.916775836886097,
+      "learning_rate": 9.279813277233261e-06,
+      "loss": 0.6012,
+      "step": 1722
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7620966205766,
+      "learning_rate": 9.278850818955989e-06,
+      "loss": 0.5516,
+      "step": 1723
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1150048556025753,
+      "learning_rate": 9.277887767973044e-06,
+      "loss": 0.4484,
+      "step": 1724
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.38236577808735,
+      "learning_rate": 9.276924124417825e-06,
+      "loss": 0.5757,
+      "step": 1725
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3706686831043804,
+      "learning_rate": 9.275959888423817e-06,
+      "loss": 0.5205,
+      "step": 1726
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9224324308109333,
+      "learning_rate": 9.274995060124587e-06,
+      "loss": 0.4409,
+      "step": 1727
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.01397500517764,
+      "learning_rate": 9.27402963965378e-06,
+      "loss": 0.6055,
+      "step": 1728
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7855739472627676,
+      "learning_rate": 9.273063627145129e-06,
+      "loss": 0.6416,
+      "step": 1729
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1938142516751835,
+      "learning_rate": 9.272097022732444e-06,
+      "loss": 0.493,
+      "step": 1730
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2671632617175543,
+      "learning_rate": 9.271129826549618e-06,
+      "loss": 0.5706,
+      "step": 1731
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.745878934945339,
+      "learning_rate": 9.27016203873063e-06,
+      "loss": 0.547,
+      "step": 1732
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2405446983690793,
+      "learning_rate": 9.269193659409537e-06,
+      "loss": 0.4102,
+      "step": 1733
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6376638649941833,
+      "learning_rate": 9.268224688720475e-06,
+      "loss": 0.4033,
+      "step": 1734
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.898115728206501,
+      "learning_rate": 9.26725512679767e-06,
+      "loss": 0.517,
+      "step": 1735
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.100911007122096,
+      "learning_rate": 9.266284973775423e-06,
+      "loss": 0.5908,
+      "step": 1736
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6888217399057366,
+      "learning_rate": 9.265314229788122e-06,
+      "loss": 0.4371,
+      "step": 1737
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.261139539657666,
+      "learning_rate": 9.264342894970232e-06,
+      "loss": 0.5368,
+      "step": 1738
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0070994367713797,
+      "learning_rate": 9.263370969456303e-06,
+      "loss": 0.5619,
+      "step": 1739
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.233244602205971,
+      "learning_rate": 9.262398453380964e-06,
+      "loss": 0.6043,
+      "step": 1740
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1344209362683615,
+      "learning_rate": 9.261425346878932e-06,
+      "loss": 0.5722,
+      "step": 1741
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.312045015174331,
+      "learning_rate": 9.260451650084997e-06,
+      "loss": 0.4329,
+      "step": 1742
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.73223118750901,
+      "learning_rate": 9.259477363134038e-06,
+      "loss": 0.5158,
+      "step": 1743
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.314947432089919,
+      "learning_rate": 9.258502486161011e-06,
+      "loss": 0.5371,
+      "step": 1744
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.488166647225511,
+      "learning_rate": 9.25752701930096e-06,
+      "loss": 0.4757,
+      "step": 1745
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9617956614253966,
+      "learning_rate": 9.256550962689003e-06,
+      "loss": 0.735,
+      "step": 1746
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0685417327972937,
+      "learning_rate": 9.255574316460342e-06,
+      "loss": 0.5729,
+      "step": 1747
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9397338365648071,
+      "learning_rate": 9.254597080750268e-06,
+      "loss": 0.5669,
+      "step": 1748
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.140901289905813,
+      "learning_rate": 9.25361925569414e-06,
+      "loss": 0.5026,
+      "step": 1749
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1999199991383493,
+      "learning_rate": 9.25264084142741e-06,
+      "loss": 0.5072,
+      "step": 1750
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7285888415076514,
+      "learning_rate": 9.251661838085606e-06,
+      "loss": 0.604,
+      "step": 1751
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.494550271524926,
+      "learning_rate": 9.250682245804342e-06,
+      "loss": 0.5594,
+      "step": 1752
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.153887537450256,
+      "learning_rate": 9.249702064719308e-06,
+      "loss": 0.5721,
+      "step": 1753
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8075900480034057,
+      "learning_rate": 9.248721294966284e-06,
+      "loss": 0.6166,
+      "step": 1754
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0685179132930416,
+      "learning_rate": 9.247739936681118e-06,
+      "loss": 0.4231,
+      "step": 1755
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.522963444088429,
+      "learning_rate": 9.246757989999754e-06,
+      "loss": 0.4632,
+      "step": 1756
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7720080104646432,
+      "learning_rate": 9.245775455058207e-06,
+      "loss": 0.5268,
+      "step": 1757
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.777016192418897,
+      "learning_rate": 9.24479233199258e-06,
+      "loss": 0.5514,
+      "step": 1758
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6858353612969699,
+      "learning_rate": 9.243808620939057e-06,
+      "loss": 0.6307,
+      "step": 1759
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8225147284944496,
+      "learning_rate": 9.242824322033895e-06,
+      "loss": 0.5193,
+      "step": 1760
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.238989755907847,
+      "learning_rate": 9.241839435413445e-06,
+      "loss": 0.6767,
+      "step": 1761
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.657645169142675,
+      "learning_rate": 9.24085396121413e-06,
+      "loss": 0.5105,
+      "step": 1762
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.472699153319154,
+      "learning_rate": 9.239867899572459e-06,
+      "loss": 0.6072,
+      "step": 1763
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0460582834217007,
+      "learning_rate": 9.238881250625023e-06,
+      "loss": 0.5136,
+      "step": 1764
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7452890341890717,
+      "learning_rate": 9.237894014508487e-06,
+      "loss": 0.5396,
+      "step": 1765
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.541047991053986,
+      "learning_rate": 9.236906191359608e-06,
+      "loss": 0.4281,
+      "step": 1766
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9218260778816876,
+      "learning_rate": 9.235917781315217e-06,
+      "loss": 0.5829,
+      "step": 1767
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6567259338940978,
+      "learning_rate": 9.23492878451223e-06,
+      "loss": 0.4657,
+      "step": 1768
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.12920076329061,
+      "learning_rate": 9.233939201087639e-06,
+      "loss": 0.4404,
+      "step": 1769
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7231129848252806,
+      "learning_rate": 9.232949031178524e-06,
+      "loss": 0.4778,
+      "step": 1770
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6687219688921977,
+      "learning_rate": 9.231958274922042e-06,
+      "loss": 0.5585,
+      "step": 1771
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8497823571547065,
+      "learning_rate": 9.230966932455434e-06,
+      "loss": 0.6104,
+      "step": 1772
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.061172769883759,
+      "learning_rate": 9.22997500391602e-06,
+      "loss": 0.5214,
+      "step": 1773
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.020129858974131,
+      "learning_rate": 9.228982489441199e-06,
+      "loss": 0.5454,
+      "step": 1774
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3850469036444344,
+      "learning_rate": 9.227989389168454e-06,
+      "loss": 0.5998,
+      "step": 1775
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.690129713596542,
+      "learning_rate": 9.226995703235355e-06,
+      "loss": 0.6152,
+      "step": 1776
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6039194247374198,
+      "learning_rate": 9.226001431779543e-06,
+      "loss": 0.5586,
+      "step": 1777
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.025359666134105,
+      "learning_rate": 9.225006574938745e-06,
+      "loss": 0.4411,
+      "step": 1778
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.137278956157683,
+      "learning_rate": 9.224011132850765e-06,
+      "loss": 0.4883,
+      "step": 1779
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.050831612541425,
+      "learning_rate": 9.223015105653497e-06,
+      "loss": 0.5034,
+      "step": 1780
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.332124191851134,
+      "learning_rate": 9.222018493484907e-06,
+      "loss": 0.5034,
+      "step": 1781
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.151898173779744,
+      "learning_rate": 9.221021296483047e-06,
+      "loss": 0.5179,
+      "step": 1782
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0201980247756803,
+      "learning_rate": 9.220023514786047e-06,
+      "loss": 0.399,
+      "step": 1783
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1814300163939624,
+      "learning_rate": 9.219025148532124e-06,
+      "loss": 0.6111,
+      "step": 1784
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.197113075633367,
+      "learning_rate": 9.218026197859565e-06,
+      "loss": 0.4652,
+      "step": 1785
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7903012509168792,
+      "learning_rate": 9.217026662906747e-06,
+      "loss": 0.5569,
+      "step": 1786
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9326665734843288,
+      "learning_rate": 9.216026543812129e-06,
+      "loss": 0.7471,
+      "step": 1787
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.126023062470786,
+      "learning_rate": 9.215025840714243e-06,
+      "loss": 0.5279,
+      "step": 1788
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0579227510180287,
+      "learning_rate": 9.214024553751709e-06,
+      "loss": 0.6602,
+      "step": 1789
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9626414416159172,
+      "learning_rate": 9.21302268306322e-06,
+      "loss": 0.5305,
+      "step": 1790
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2437894388769,
+      "learning_rate": 9.212020228787562e-06,
+      "loss": 0.4894,
+      "step": 1791
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.256827513396298,
+      "learning_rate": 9.21101719106359e-06,
+      "loss": 0.4984,
+      "step": 1792
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9117355346291962,
+      "learning_rate": 9.210013570030246e-06,
+      "loss": 0.5877,
+      "step": 1793
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9346739038510565,
+      "learning_rate": 9.209009365826553e-06,
+      "loss": 0.4944,
+      "step": 1794
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2923896260919716,
+      "learning_rate": 9.20800457859161e-06,
+      "loss": 0.5518,
+      "step": 1795
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.935080120734737,
+      "learning_rate": 9.206999208464602e-06,
+      "loss": 0.5728,
+      "step": 1796
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2387378544723875,
+      "learning_rate": 9.205993255584793e-06,
+      "loss": 0.4694,
+      "step": 1797
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.6141508127479476,
+      "learning_rate": 9.204986720091527e-06,
+      "loss": 0.5316,
+      "step": 1798
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.923212355555867,
+      "learning_rate": 9.203979602124227e-06,
+      "loss": 0.7379,
+      "step": 1799
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7174323897023884,
+      "learning_rate": 9.202971901822401e-06,
+      "loss": 0.5369,
+      "step": 1800
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.149912136910846,
+      "learning_rate": 9.201963619325637e-06,
+      "loss": 0.5556,
+      "step": 1801
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7152865798518133,
+      "learning_rate": 9.200954754773598e-06,
+      "loss": 0.5244,
+      "step": 1802
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9933439146724623,
+      "learning_rate": 9.199945308306037e-06,
+      "loss": 0.4657,
+      "step": 1803
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.385353556019281,
+      "learning_rate": 9.198935280062777e-06,
+      "loss": 0.4397,
+      "step": 1804
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8009881603970972,
+      "learning_rate": 9.19792467018373e-06,
+      "loss": 0.6378,
+      "step": 1805
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1227904905286,
+      "learning_rate": 9.196913478808884e-06,
+      "loss": 0.6093,
+      "step": 1806
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.65541491251585,
+      "learning_rate": 9.19590170607831e-06,
+      "loss": 0.55,
+      "step": 1807
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8004198144346177,
+      "learning_rate": 9.19488935213216e-06,
+      "loss": 0.5754,
+      "step": 1808
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6392718422746173,
+      "learning_rate": 9.193876417110663e-06,
+      "loss": 0.5246,
+      "step": 1809
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6041962325593957,
+      "learning_rate": 9.19286290115413e-06,
+      "loss": 0.4805,
+      "step": 1810
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.152945962884701,
+      "learning_rate": 9.191848804402953e-06,
+      "loss": 0.5872,
+      "step": 1811
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0157569167531815,
+      "learning_rate": 9.190834126997608e-06,
+      "loss": 0.5304,
+      "step": 1812
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8933135369308454,
+      "learning_rate": 9.189818869078646e-06,
+      "loss": 0.4374,
+      "step": 1813
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0081378690035843,
+      "learning_rate": 9.188803030786699e-06,
+      "loss": 0.5237,
+      "step": 1814
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.816258066899212,
+      "learning_rate": 9.18778661226248e-06,
+      "loss": 0.5353,
+      "step": 1815
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9725331292589412,
+      "learning_rate": 9.186769613646788e-06,
+      "loss": 0.5072,
+      "step": 1816
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9242474553624498,
+      "learning_rate": 9.185752035080493e-06,
+      "loss": 0.5533,
+      "step": 1817
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8027038075045165,
+      "learning_rate": 9.184733876704551e-06,
+      "loss": 0.483,
+      "step": 1818
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.488182535391086,
+      "learning_rate": 9.183715138659996e-06,
+      "loss": 0.6066,
+      "step": 1819
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0794839724835548,
+      "learning_rate": 9.182695821087946e-06,
+      "loss": 0.5022,
+      "step": 1820
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5755164555833303,
+      "learning_rate": 9.181675924129595e-06,
+      "loss": 0.5197,
+      "step": 1821
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9283660167109985,
+      "learning_rate": 9.180655447926219e-06,
+      "loss": 0.7784,
+      "step": 1822
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2116940209469713,
+      "learning_rate": 9.179634392619174e-06,
+      "loss": 0.6143,
+      "step": 1823
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.139523940555377,
+      "learning_rate": 9.178612758349899e-06,
+      "loss": 0.5217,
+      "step": 1824
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9511526129103196,
+      "learning_rate": 9.177590545259907e-06,
+      "loss": 0.4964,
+      "step": 1825
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.823612264443561,
+      "learning_rate": 9.176567753490795e-06,
+      "loss": 0.6966,
+      "step": 1826
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4795243834549616,
+      "learning_rate": 9.175544383184243e-06,
+      "loss": 0.4929,
+      "step": 1827
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2350200431591474,
+      "learning_rate": 9.174520434482006e-06,
+      "loss": 0.5514,
+      "step": 1828
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7938537056568982,
+      "learning_rate": 9.173495907525922e-06,
+      "loss": 0.5881,
+      "step": 1829
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.041122078062016,
+      "learning_rate": 9.172470802457906e-06,
+      "loss": 0.5162,
+      "step": 1830
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7715885562248541,
+      "learning_rate": 9.17144511941996e-06,
+      "loss": 0.4667,
+      "step": 1831
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.208043086641995,
+      "learning_rate": 9.170418858554156e-06,
+      "loss": 0.52,
+      "step": 1832
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7486752137742332,
+      "learning_rate": 9.169392020002655e-06,
+      "loss": 0.5684,
+      "step": 1833
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2117531554836023,
+      "learning_rate": 9.168364603907693e-06,
+      "loss": 0.5671,
+      "step": 1834
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1347463663878847,
+      "learning_rate": 9.167336610411588e-06,
+      "loss": 0.4207,
+      "step": 1835
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9142562532132055,
+      "learning_rate": 9.166308039656737e-06,
+      "loss": 0.5062,
+      "step": 1836
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.007774560296191,
+      "learning_rate": 9.16527889178562e-06,
+      "loss": 0.5626,
+      "step": 1837
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.492070546092896,
+      "learning_rate": 9.16424916694079e-06,
+      "loss": 0.5432,
+      "step": 1838
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.987932903431278,
+      "learning_rate": 9.163218865264889e-06,
+      "loss": 0.4863,
+      "step": 1839
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9981386912086843,
+      "learning_rate": 9.162187986900631e-06,
+      "loss": 0.4791,
+      "step": 1840
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.592624829293792,
+      "learning_rate": 9.161156531990814e-06,
+      "loss": 0.5567,
+      "step": 1841
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9262053257022613,
+      "learning_rate": 9.160124500678313e-06,
+      "loss": 0.7334,
+      "step": 1842
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.745789896780652,
+      "learning_rate": 9.159091893106089e-06,
+      "loss": 0.597,
+      "step": 1843
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.559812399649817,
+      "learning_rate": 9.158058709417176e-06,
+      "loss": 0.4866,
+      "step": 1844
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4744705737025416,
+      "learning_rate": 9.15702494975469e-06,
+      "loss": 0.5537,
+      "step": 1845
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7076292301830707,
+      "learning_rate": 9.15599061426183e-06,
+      "loss": 0.4611,
+      "step": 1846
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.216686976791891,
+      "learning_rate": 9.154955703081868e-06,
+      "loss": 0.532,
+      "step": 1847
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.194850077687543,
+      "learning_rate": 9.153920216358161e-06,
+      "loss": 0.5214,
+      "step": 1848
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.849176543655242,
+      "learning_rate": 9.152884154234147e-06,
+      "loss": 0.5672,
+      "step": 1849
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.571982433504018,
+      "learning_rate": 9.151847516853338e-06,
+      "loss": 0.5682,
+      "step": 1850
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4476724548427258,
+      "learning_rate": 9.15081030435933e-06,
+      "loss": 0.4616,
+      "step": 1851
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.170319376968797,
+      "learning_rate": 9.149772516895798e-06,
+      "loss": 0.5425,
+      "step": 1852
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.698999346373502,
+      "learning_rate": 9.148734154606497e-06,
+      "loss": 0.5585,
+      "step": 1853
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6812894428682459,
+      "learning_rate": 9.147695217635258e-06,
+      "loss": 0.5393,
+      "step": 1854
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6961757583843824,
+      "learning_rate": 9.146655706125995e-06,
+      "loss": 0.5085,
+      "step": 1855
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.186448213325977,
+      "learning_rate": 9.145615620222705e-06,
+      "loss": 0.492,
+      "step": 1856
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9774798959394002,
+      "learning_rate": 9.144574960069454e-06,
+      "loss": 0.5308,
+      "step": 1857
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8391895024784863,
+      "learning_rate": 9.143533725810398e-06,
+      "loss": 0.5597,
+      "step": 1858
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4588263380412037,
+      "learning_rate": 9.142491917589768e-06,
+      "loss": 0.6383,
+      "step": 1859
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.907131153713141,
+      "learning_rate": 9.141449535551878e-06,
+      "loss": 0.7582,
+      "step": 1860
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.021368692690415,
+      "learning_rate": 9.140406579841113e-06,
+      "loss": 0.5481,
+      "step": 1861
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5719269848717543,
+      "learning_rate": 9.139363050601946e-06,
+      "loss": 0.467,
+      "step": 1862
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8715011163912674,
+      "learning_rate": 9.138318947978927e-06,
+      "loss": 0.5686,
+      "step": 1863
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 38.77794395126524,
+      "learning_rate": 9.137274272116683e-06,
+      "loss": 0.5895,
+      "step": 1864
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.867779733294635,
+      "learning_rate": 9.136229023159924e-06,
+      "loss": 0.4682,
+      "step": 1865
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8127864666263446,
+      "learning_rate": 9.135183201253436e-06,
+      "loss": 0.7177,
+      "step": 1866
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8425028393910223,
+      "learning_rate": 9.134136806542089e-06,
+      "loss": 0.5679,
+      "step": 1867
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7783589068517192,
+      "learning_rate": 9.133089839170827e-06,
+      "loss": 0.5049,
+      "step": 1868
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.356543996534519,
+      "learning_rate": 9.132042299284675e-06,
+      "loss": 0.4768,
+      "step": 1869
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8374804126372851,
+      "learning_rate": 9.13099418702874e-06,
+      "loss": 0.7606,
+      "step": 1870
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8325439443698737,
+      "learning_rate": 9.129945502548207e-06,
+      "loss": 0.5195,
+      "step": 1871
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.85856708583394,
+      "learning_rate": 9.128896245988338e-06,
+      "loss": 0.5826,
+      "step": 1872
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.6343025245998666,
+      "learning_rate": 9.127846417494476e-06,
+      "loss": 0.5934,
+      "step": 1873
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1865048544002312,
+      "learning_rate": 9.126796017212043e-06,
+      "loss": 0.4989,
+      "step": 1874
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9021958568108963,
+      "learning_rate": 9.12574504528654e-06,
+      "loss": 0.6248,
+      "step": 1875
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7645176368857158,
+      "learning_rate": 9.124693501863548e-06,
+      "loss": 0.4595,
+      "step": 1876
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1204261240904514,
+      "learning_rate": 9.123641387088728e-06,
+      "loss": 0.489,
+      "step": 1877
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.229406726615884,
+      "learning_rate": 9.122588701107816e-06,
+      "loss": 0.4985,
+      "step": 1878
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.958105451295423,
+      "learning_rate": 9.121535444066631e-06,
+      "loss": 0.5233,
+      "step": 1879
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8092095078807724,
+      "learning_rate": 9.12048161611107e-06,
+      "loss": 0.6936,
+      "step": 1880
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1205924474160938,
+      "learning_rate": 9.11942721738711e-06,
+      "loss": 0.5167,
+      "step": 1881
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.3999100818341494,
+      "learning_rate": 9.118372248040806e-06,
+      "loss": 0.5918,
+      "step": 1882
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7527760489266877,
+      "learning_rate": 9.11731670821829e-06,
+      "loss": 0.5206,
+      "step": 1883
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7445246125433562,
+      "learning_rate": 9.116260598065776e-06,
+      "loss": 0.5597,
+      "step": 1884
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2084289500002594,
+      "learning_rate": 9.11520391772956e-06,
+      "loss": 0.4984,
+      "step": 1885
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.900056047123036,
+      "learning_rate": 9.114146667356008e-06,
+      "loss": 0.5778,
+      "step": 1886
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3733798801148764,
+      "learning_rate": 9.113088847091572e-06,
+      "loss": 0.5175,
+      "step": 1887
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2387827326535477,
+      "learning_rate": 9.112030457082782e-06,
+      "loss": 0.4892,
+      "step": 1888
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8164512957048708,
+      "learning_rate": 9.110971497476245e-06,
+      "loss": 0.5144,
+      "step": 1889
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.066164608150523,
+      "learning_rate": 9.10991196841865e-06,
+      "loss": 0.5728,
+      "step": 1890
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8826150265260178,
+      "learning_rate": 9.108851870056759e-06,
+      "loss": 0.7528,
+      "step": 1891
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2096333128539767,
+      "learning_rate": 9.107791202537419e-06,
+      "loss": 0.5318,
+      "step": 1892
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9240021967684576,
+      "learning_rate": 9.106729966007552e-06,
+      "loss": 0.7282,
+      "step": 1893
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.015164754561475,
+      "learning_rate": 9.105668160614163e-06,
+      "loss": 0.4649,
+      "step": 1894
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1511659744168843,
+      "learning_rate": 9.104605786504332e-06,
+      "loss": 0.5673,
+      "step": 1895
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.358178994873392,
+      "learning_rate": 9.103542843825217e-06,
+      "loss": 0.465,
+      "step": 1896
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5521116290661428,
+      "learning_rate": 9.102479332724058e-06,
+      "loss": 0.5447,
+      "step": 1897
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.10997096003024,
+      "learning_rate": 9.101415253348173e-06,
+      "loss": 0.5374,
+      "step": 1898
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9962027174419892,
+      "learning_rate": 9.100350605844957e-06,
+      "loss": 0.5406,
+      "step": 1899
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6243297471546538,
+      "learning_rate": 9.099285390361886e-06,
+      "loss": 0.5796,
+      "step": 1900
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2121660419747107,
+      "learning_rate": 9.098219607046511e-06,
+      "loss": 0.6073,
+      "step": 1901
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9955861573741578,
+      "learning_rate": 9.097153256046469e-06,
+      "loss": 0.547,
+      "step": 1902
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1364326659776434,
+      "learning_rate": 9.096086337509466e-06,
+      "loss": 0.4824,
+      "step": 1903
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.192690978932753,
+      "learning_rate": 9.095018851583292e-06,
+      "loss": 0.5849,
+      "step": 1904
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7830087415320337,
+      "learning_rate": 9.093950798415819e-06,
+      "loss": 0.5749,
+      "step": 1905
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.770271052780923,
+      "learning_rate": 9.092882178154988e-06,
+      "loss": 0.4663,
+      "step": 1906
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9445653866631203,
+      "learning_rate": 9.091812990948827e-06,
+      "loss": 0.4966,
+      "step": 1907
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.021649522757026,
+      "learning_rate": 9.09074323694544e-06,
+      "loss": 0.7717,
+      "step": 1908
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8794644831130864,
+      "learning_rate": 9.089672916293006e-06,
+      "loss": 0.5512,
+      "step": 1909
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7493547962086655,
+      "learning_rate": 9.088602029139789e-06,
+      "loss": 0.5204,
+      "step": 1910
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9728629963329603,
+      "learning_rate": 9.087530575634127e-06,
+      "loss": 0.6523,
+      "step": 1911
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9484474067945332,
+      "learning_rate": 9.086458555924439e-06,
+      "loss": 0.6191,
+      "step": 1912
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8422586991681948,
+      "learning_rate": 9.085385970159218e-06,
+      "loss": 0.5835,
+      "step": 1913
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.461546559963499,
+      "learning_rate": 9.084312818487042e-06,
+      "loss": 0.4302,
+      "step": 1914
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8370421559220482,
+      "learning_rate": 9.08323910105656e-06,
+      "loss": 0.5327,
+      "step": 1915
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3277267708761578,
+      "learning_rate": 9.082164818016506e-06,
+      "loss": 0.5237,
+      "step": 1916
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.698622440526287,
+      "learning_rate": 9.081089969515689e-06,
+      "loss": 0.4815,
+      "step": 1917
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6669862760948075,
+      "learning_rate": 9.080014555702993e-06,
+      "loss": 0.5327,
+      "step": 1918
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8013039744692143,
+      "learning_rate": 9.078938576727393e-06,
+      "loss": 0.4548,
+      "step": 1919
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1190401318482635,
+      "learning_rate": 9.077862032737923e-06,
+      "loss": 0.4389,
+      "step": 1920
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8306882820321422,
+      "learning_rate": 9.076784923883712e-06,
+      "loss": 0.6679,
+      "step": 1921
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9206827597398815,
+      "learning_rate": 9.07570725031396e-06,
+      "loss": 0.581,
+      "step": 1922
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1044052355430543,
+      "learning_rate": 9.074629012177946e-06,
+      "loss": 0.5197,
+      "step": 1923
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8211858826799479,
+      "learning_rate": 9.073550209625026e-06,
+      "loss": 0.6589,
+      "step": 1924
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.64151622473254,
+      "learning_rate": 9.072470842804636e-06,
+      "loss": 0.4113,
+      "step": 1925
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.916598599985274,
+      "learning_rate": 9.071390911866291e-06,
+      "loss": 0.534,
+      "step": 1926
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.231100355315972,
+      "learning_rate": 9.070310416959582e-06,
+      "loss": 0.4303,
+      "step": 1927
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5289379666551688,
+      "learning_rate": 9.06922935823418e-06,
+      "loss": 0.4689,
+      "step": 1928
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7143397383272994,
+      "learning_rate": 9.068147735839831e-06,
+      "loss": 0.5486,
+      "step": 1929
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.33881139914712,
+      "learning_rate": 9.067065549926362e-06,
+      "loss": 0.5312,
+      "step": 1930
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8960024957907573,
+      "learning_rate": 9.065982800643679e-06,
+      "loss": 0.6016,
+      "step": 1931
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.785514434261121,
+      "learning_rate": 9.064899488141761e-06,
+      "loss": 0.4924,
+      "step": 1932
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8671983616213554,
+      "learning_rate": 9.06381561257067e-06,
+      "loss": 0.5363,
+      "step": 1933
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.880933228690065,
+      "learning_rate": 9.062731174080546e-06,
+      "loss": 0.551,
+      "step": 1934
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9133027538157963,
+      "learning_rate": 9.061646172821602e-06,
+      "loss": 0.5565,
+      "step": 1935
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.849106292022935,
+      "learning_rate": 9.060560608944134e-06,
+      "loss": 0.536,
+      "step": 1936
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.213159310538447,
+      "learning_rate": 9.059474482598513e-06,
+      "loss": 0.515,
+      "step": 1937
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9411566263430953,
+      "learning_rate": 9.05838779393519e-06,
+      "loss": 0.5548,
+      "step": 1938
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9459753390672059,
+      "learning_rate": 9.057300543104694e-06,
+      "loss": 0.5633,
+      "step": 1939
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.088177135655805,
+      "learning_rate": 9.05621273025763e-06,
+      "loss": 0.5007,
+      "step": 1940
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.721449262555121,
+      "learning_rate": 9.05512435554468e-06,
+      "loss": 0.5571,
+      "step": 1941
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.065905032378851,
+      "learning_rate": 9.054035419116606e-06,
+      "loss": 0.4857,
+      "step": 1942
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9504759539256835,
+      "learning_rate": 9.052945921124248e-06,
+      "loss": 0.4633,
+      "step": 1943
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8560338515869066,
+      "learning_rate": 9.051855861718524e-06,
+      "loss": 0.561,
+      "step": 1944
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4671426828258154,
+      "learning_rate": 9.050765241050428e-06,
+      "loss": 0.568,
+      "step": 1945
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6397981485264328,
+      "learning_rate": 9.04967405927103e-06,
+      "loss": 0.4312,
+      "step": 1946
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9972076886572796,
+      "learning_rate": 9.048582316531485e-06,
+      "loss": 0.5349,
+      "step": 1947
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.051096277529259,
+      "learning_rate": 9.047490012983018e-06,
+      "loss": 0.5323,
+      "step": 1948
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.965470589495238,
+      "learning_rate": 9.046397148776936e-06,
+      "loss": 0.4902,
+      "step": 1949
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6663527598837038,
+      "learning_rate": 9.045303724064622e-06,
+      "loss": 0.6704,
+      "step": 1950
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2056114953419663,
+      "learning_rate": 9.044209738997536e-06,
+      "loss": 0.5089,
+      "step": 1951
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.721142554293289,
+      "learning_rate": 9.043115193727217e-06,
+      "loss": 0.5348,
+      "step": 1952
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0201122793669208,
+      "learning_rate": 9.042020088405283e-06,
+      "loss": 0.7489,
+      "step": 1953
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.006774542673198,
+      "learning_rate": 9.040924423183426e-06,
+      "loss": 0.5338,
+      "step": 1954
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.792945940839733,
+      "learning_rate": 9.039828198213417e-06,
+      "loss": 0.5401,
+      "step": 1955
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1968251391207456,
+      "learning_rate": 9.038731413647107e-06,
+      "loss": 0.4626,
+      "step": 1956
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.685138503893445,
+      "learning_rate": 9.037634069636421e-06,
+      "loss": 0.5999,
+      "step": 1957
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.238135664801939,
+      "learning_rate": 9.036536166333362e-06,
+      "loss": 0.5119,
+      "step": 1958
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.111904037343309,
+      "learning_rate": 9.035437703890013e-06,
+      "loss": 0.5815,
+      "step": 1959
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8229903814935794,
+      "learning_rate": 9.034338682458532e-06,
+      "loss": 0.6909,
+      "step": 1960
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6850429482854024,
+      "learning_rate": 9.033239102191156e-06,
+      "loss": 0.5839,
+      "step": 1961
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.534874169611228,
+      "learning_rate": 9.032138963240196e-06,
+      "loss": 0.5452,
+      "step": 1962
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7692708409492117,
+      "learning_rate": 9.031038265758047e-06,
+      "loss": 0.5316,
+      "step": 1963
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1581888028680707,
+      "learning_rate": 9.029937009897176e-06,
+      "loss": 0.5427,
+      "step": 1964
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.912847642581488,
+      "learning_rate": 9.028835195810129e-06,
+      "loss": 0.5466,
+      "step": 1965
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6850714608750732,
+      "learning_rate": 9.027732823649526e-06,
+      "loss": 0.6231,
+      "step": 1966
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.49959562907448,
+      "learning_rate": 9.026629893568072e-06,
+      "loss": 0.6016,
+      "step": 1967
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.586796918276492,
+      "learning_rate": 9.02552640571854e-06,
+      "loss": 0.5038,
+      "step": 1968
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9385622516231051,
+      "learning_rate": 9.02442236025379e-06,
+      "loss": 0.6131,
+      "step": 1969
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.068435608703494,
+      "learning_rate": 9.023317757326753e-06,
+      "loss": 0.5491,
+      "step": 1970
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.2004764087517055,
+      "learning_rate": 9.022212597090434e-06,
+      "loss": 0.5518,
+      "step": 1971
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5917492754666265,
+      "learning_rate": 9.021106879697925e-06,
+      "loss": 0.6244,
+      "step": 1972
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0962273704440277,
+      "learning_rate": 9.020000605302385e-06,
+      "loss": 0.5446,
+      "step": 1973
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6645339005009094,
+      "learning_rate": 9.018893774057061e-06,
+      "loss": 0.5601,
+      "step": 1974
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8231461292672424,
+      "learning_rate": 9.017786386115263e-06,
+      "loss": 0.6358,
+      "step": 1975
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7987163489856062,
+      "learning_rate": 9.016678441630393e-06,
+      "loss": 0.5367,
+      "step": 1976
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2369345857824317,
+      "learning_rate": 9.015569940755922e-06,
+      "loss": 0.5329,
+      "step": 1977
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.728066836602479,
+      "learning_rate": 9.014460883645398e-06,
+      "loss": 0.6002,
+      "step": 1978
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6059944357341411,
+      "learning_rate": 9.013351270452446e-06,
+      "loss": 0.4742,
+      "step": 1979
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7823855944385212,
+      "learning_rate": 9.012241101330772e-06,
+      "loss": 0.5767,
+      "step": 1980
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0201033799993966,
+      "learning_rate": 9.011130376434157e-06,
+      "loss": 0.4586,
+      "step": 1981
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0534029345220777,
+      "learning_rate": 9.010019095916456e-06,
+      "loss": 0.5922,
+      "step": 1982
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1261607367013187,
+      "learning_rate": 9.008907259931603e-06,
+      "loss": 0.5441,
+      "step": 1983
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.065925327568663,
+      "learning_rate": 9.007794868633613e-06,
+      "loss": 0.5081,
+      "step": 1984
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.041424168139034,
+      "learning_rate": 9.00668192217657e-06,
+      "loss": 0.5511,
+      "step": 1985
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2997858035561634,
+      "learning_rate": 9.005568420714643e-06,
+      "loss": 0.6062,
+      "step": 1986
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8412917708484329,
+      "learning_rate": 9.00445436440207e-06,
+      "loss": 0.4952,
+      "step": 1987
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.089107395469012,
+      "learning_rate": 9.003339753393174e-06,
+      "loss": 0.4571,
+      "step": 1988
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4793857270263975,
+      "learning_rate": 9.002224587842348e-06,
+      "loss": 0.5813,
+      "step": 1989
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8213706345617011,
+      "learning_rate": 9.001108867904066e-06,
+      "loss": 0.5654,
+      "step": 1990
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.213176110546956,
+      "learning_rate": 8.999992593732876e-06,
+      "loss": 0.5236,
+      "step": 1991
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9147480873118935,
+      "learning_rate": 8.998875765483403e-06,
+      "loss": 0.7171,
+      "step": 1992
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.553084948631206,
+      "learning_rate": 8.997758383310353e-06,
+      "loss": 0.559,
+      "step": 1993
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.299159937606802,
+      "learning_rate": 8.996640447368505e-06,
+      "loss": 0.6135,
+      "step": 1994
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2903173319664947,
+      "learning_rate": 8.995521957812713e-06,
+      "loss": 0.522,
+      "step": 1995
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7712206512869257,
+      "learning_rate": 8.994402914797913e-06,
+      "loss": 0.6154,
+      "step": 1996
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.647062710587506,
+      "learning_rate": 8.993283318479114e-06,
+      "loss": 0.5562,
+      "step": 1997
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4820557946984594,
+      "learning_rate": 8.992163169011398e-06,
+      "loss": 0.5354,
+      "step": 1998
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9501538495177364,
+      "learning_rate": 8.991042466549934e-06,
+      "loss": 0.4955,
+      "step": 1999
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5354235506764096,
+      "learning_rate": 8.989921211249959e-06,
+      "loss": 0.5564,
+      "step": 2000
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8812252462931816,
+      "learning_rate": 8.988799403266787e-06,
+      "loss": 0.5429,
+      "step": 2001
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0598170169677092,
+      "learning_rate": 8.987677042755813e-06,
+      "loss": 0.4832,
+      "step": 2002
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7735614800416573,
+      "learning_rate": 8.986554129872506e-06,
+      "loss": 0.6027,
+      "step": 2003
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7839874046135726,
+      "learning_rate": 8.985430664772412e-06,
+      "loss": 0.3973,
+      "step": 2004
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.920618782500927,
+      "learning_rate": 8.984306647611152e-06,
+      "loss": 0.4952,
+      "step": 2005
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4874926107984585,
+      "learning_rate": 8.983182078544426e-06,
+      "loss": 0.5013,
+      "step": 2006
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.870436872918295,
+      "learning_rate": 8.982056957728007e-06,
+      "loss": 0.44,
+      "step": 2007
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7663225469186314,
+      "learning_rate": 8.980931285317748e-06,
+      "loss": 0.5122,
+      "step": 2008
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.787792430643811,
+      "learning_rate": 8.979805061469578e-06,
+      "loss": 0.5355,
+      "step": 2009
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7720685138763685,
+      "learning_rate": 8.978678286339499e-06,
+      "loss": 0.5106,
+      "step": 2010
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.049347466256778,
+      "learning_rate": 8.977550960083594e-06,
+      "loss": 0.4953,
+      "step": 2011
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3512648083501,
+      "learning_rate": 8.976423082858019e-06,
+      "loss": 0.4878,
+      "step": 2012
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9080251530236207,
+      "learning_rate": 8.975294654819007e-06,
+      "loss": 0.5177,
+      "step": 2013
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8155878514484811,
+      "learning_rate": 8.974165676122868e-06,
+      "loss": 0.4294,
+      "step": 2014
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0606182954851757,
+      "learning_rate": 8.973036146925988e-06,
+      "loss": 0.5454,
+      "step": 2015
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.162875142123971,
+      "learning_rate": 8.971906067384828e-06,
+      "loss": 0.6403,
+      "step": 2016
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.361838190447896,
+      "learning_rate": 8.970775437655929e-06,
+      "loss": 0.5584,
+      "step": 2017
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.650402602528094,
+      "learning_rate": 8.969644257895903e-06,
+      "loss": 0.4942,
+      "step": 2018
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.790318007947418,
+      "learning_rate": 8.968512528261442e-06,
+      "loss": 0.5611,
+      "step": 2019
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4218513169375613,
+      "learning_rate": 8.967380248909314e-06,
+      "loss": 0.5619,
+      "step": 2020
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.675111813391207,
+      "learning_rate": 8.966247419996361e-06,
+      "loss": 0.5237,
+      "step": 2021
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.959342748663956,
+      "learning_rate": 8.965114041679501e-06,
+      "loss": 0.5651,
+      "step": 2022
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.571613547302985,
+      "learning_rate": 8.96398011411573e-06,
+      "loss": 0.5118,
+      "step": 2023
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4611593014253406,
+      "learning_rate": 8.962845637462124e-06,
+      "loss": 0.5403,
+      "step": 2024
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9672891699008197,
+      "learning_rate": 8.961710611875825e-06,
+      "loss": 0.5571,
+      "step": 2025
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7357065807967837,
+      "learning_rate": 8.960575037514056e-06,
+      "loss": 0.5661,
+      "step": 2026
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.746420480725577,
+      "learning_rate": 8.95943891453412e-06,
+      "loss": 0.4555,
+      "step": 2027
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3178158614336666,
+      "learning_rate": 8.958302243093393e-06,
+      "loss": 0.459,
+      "step": 2028
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.886116006509517,
+      "learning_rate": 8.957165023349324e-06,
+      "loss": 0.5282,
+      "step": 2029
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9081119130058164,
+      "learning_rate": 8.95602725545944e-06,
+      "loss": 0.5952,
+      "step": 2030
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8423987633434151,
+      "learning_rate": 8.954888939581348e-06,
+      "loss": 0.4258,
+      "step": 2031
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2143503900539963,
+      "learning_rate": 8.953750075872724e-06,
+      "loss": 0.5114,
+      "step": 2032
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.450856412023587,
+      "learning_rate": 8.952610664491323e-06,
+      "loss": 0.4967,
+      "step": 2033
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9805190908088535,
+      "learning_rate": 8.95147070559498e-06,
+      "loss": 0.5003,
+      "step": 2034
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.677948063167133,
+      "learning_rate": 8.950330199341596e-06,
+      "loss": 0.4622,
+      "step": 2035
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9831307116449997,
+      "learning_rate": 8.94918914588916e-06,
+      "loss": 0.4746,
+      "step": 2036
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9312163915182219,
+      "learning_rate": 8.948047545395726e-06,
+      "loss": 0.7038,
+      "step": 2037
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.886471612416682,
+      "learning_rate": 8.946905398019431e-06,
+      "loss": 0.5847,
+      "step": 2038
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.391104786889129,
+      "learning_rate": 8.945762703918483e-06,
+      "loss": 0.4701,
+      "step": 2039
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.243904195104066,
+      "learning_rate": 8.944619463251168e-06,
+      "loss": 0.6354,
+      "step": 2040
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.340947548213193,
+      "learning_rate": 8.94347567617585e-06,
+      "loss": 0.6067,
+      "step": 2041
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8960986517902896,
+      "learning_rate": 8.942331342850963e-06,
+      "loss": 0.5122,
+      "step": 2042
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8489366828549778,
+      "learning_rate": 8.941186463435022e-06,
+      "loss": 0.5327,
+      "step": 2043
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9016716320714124,
+      "learning_rate": 8.940041038086614e-06,
+      "loss": 0.5367,
+      "step": 2044
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.981146067552342,
+      "learning_rate": 8.938895066964404e-06,
+      "loss": 0.5019,
+      "step": 2045
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.173745512259389,
+      "learning_rate": 8.937748550227133e-06,
+      "loss": 0.5757,
+      "step": 2046
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6241116538216447,
+      "learning_rate": 8.936601488033612e-06,
+      "loss": 0.5454,
+      "step": 2047
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6610816194666764,
+      "learning_rate": 8.935453880542737e-06,
+      "loss": 0.6177,
+      "step": 2048
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6670705798856436,
+      "learning_rate": 8.934305727913471e-06,
+      "loss": 0.5054,
+      "step": 2049
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9224513510586698,
+      "learning_rate": 8.933157030304857e-06,
+      "loss": 0.5067,
+      "step": 2050
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.018367274285547,
+      "learning_rate": 8.932007787876013e-06,
+      "loss": 0.5187,
+      "step": 2051
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.748596918233636,
+      "learning_rate": 8.930858000786131e-06,
+      "loss": 0.4713,
+      "step": 2052
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.831321661548211,
+      "learning_rate": 8.929707669194481e-06,
+      "loss": 0.5209,
+      "step": 2053
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.001707857952489,
+      "learning_rate": 8.928556793260403e-06,
+      "loss": 0.5032,
+      "step": 2054
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.430408329174184,
+      "learning_rate": 8.92740537314332e-06,
+      "loss": 0.5754,
+      "step": 2055
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8925528705235564,
+      "learning_rate": 8.926253409002724e-06,
+      "loss": 0.5093,
+      "step": 2056
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.225437857819384,
+      "learning_rate": 8.925100900998186e-06,
+      "loss": 0.4646,
+      "step": 2057
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1354516504698555,
+      "learning_rate": 8.923947849289351e-06,
+      "loss": 0.5638,
+      "step": 2058
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9132921665934566,
+      "learning_rate": 8.92279425403594e-06,
+      "loss": 0.4661,
+      "step": 2059
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6125730492900616,
+      "learning_rate": 8.921640115397748e-06,
+      "loss": 0.4956,
+      "step": 2060
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3167689133474267,
+      "learning_rate": 8.920485433534647e-06,
+      "loss": 0.5575,
+      "step": 2061
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.826781907095146,
+      "learning_rate": 8.919330208606583e-06,
+      "loss": 0.6106,
+      "step": 2062
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.321674037173777,
+      "learning_rate": 8.918174440773577e-06,
+      "loss": 0.4041,
+      "step": 2063
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.638864169267249,
+      "learning_rate": 8.917018130195725e-06,
+      "loss": 0.4187,
+      "step": 2064
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.056884140708432,
+      "learning_rate": 8.915861277033202e-06,
+      "loss": 0.5797,
+      "step": 2065
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.665890828806845,
+      "learning_rate": 8.914703881446252e-06,
+      "loss": 0.5421,
+      "step": 2066
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1259883238062707,
+      "learning_rate": 8.913545943595198e-06,
+      "loss": 0.4276,
+      "step": 2067
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6910264743155885,
+      "learning_rate": 8.912387463640439e-06,
+      "loss": 0.5054,
+      "step": 2068
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.941895797687128,
+      "learning_rate": 8.911228441742444e-06,
+      "loss": 0.5703,
+      "step": 2069
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5329820578195132,
+      "learning_rate": 8.910068878061764e-06,
+      "loss": 0.5277,
+      "step": 2070
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.848208964911629,
+      "learning_rate": 8.908908772759022e-06,
+      "loss": 0.5345,
+      "step": 2071
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2046412090931597,
+      "learning_rate": 8.90774812599491e-06,
+      "loss": 0.413,
+      "step": 2072
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6467209154823024,
+      "learning_rate": 8.906586937930208e-06,
+      "loss": 0.4922,
+      "step": 2073
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8436067594171583,
+      "learning_rate": 8.905425208725758e-06,
+      "loss": 0.5338,
+      "step": 2074
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1419270448086674,
+      "learning_rate": 8.904262938542485e-06,
+      "loss": 0.5431,
+      "step": 2075
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.856177815319082,
+      "learning_rate": 8.903100127541386e-06,
+      "loss": 0.5217,
+      "step": 2076
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9119980684578037,
+      "learning_rate": 8.901936775883535e-06,
+      "loss": 0.5393,
+      "step": 2077
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7133432676260245,
+      "learning_rate": 8.900772883730075e-06,
+      "loss": 0.5472,
+      "step": 2078
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3545169619710604,
+      "learning_rate": 8.899608451242233e-06,
+      "loss": 0.5834,
+      "step": 2079
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7687590085648386,
+      "learning_rate": 8.898443478581302e-06,
+      "loss": 0.5649,
+      "step": 2080
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.268035795443232,
+      "learning_rate": 8.897277965908657e-06,
+      "loss": 0.5661,
+      "step": 2081
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7808180118052503,
+      "learning_rate": 8.896111913385742e-06,
+      "loss": 0.574,
+      "step": 2082
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6078770321909537,
+      "learning_rate": 8.89494532117408e-06,
+      "loss": 0.5393,
+      "step": 2083
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0148917361849326,
+      "learning_rate": 8.893778189435267e-06,
+      "loss": 0.4931,
+      "step": 2084
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.753782231373047,
+      "learning_rate": 8.892610518330973e-06,
+      "loss": 0.5214,
+      "step": 2085
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6471398211881594,
+      "learning_rate": 8.891442308022946e-06,
+      "loss": 0.4875,
+      "step": 2086
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7043554882043275,
+      "learning_rate": 8.890273558673003e-06,
+      "loss": 0.3684,
+      "step": 2087
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.835616868568219,
+      "learning_rate": 8.889104270443041e-06,
+      "loss": 0.4754,
+      "step": 2088
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.056681168139256,
+      "learning_rate": 8.887934443495028e-06,
+      "loss": 0.47,
+      "step": 2089
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3950943575263306,
+      "learning_rate": 8.88676407799101e-06,
+      "loss": 0.5694,
+      "step": 2090
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1414395650603617,
+      "learning_rate": 8.885593174093105e-06,
+      "loss": 0.4689,
+      "step": 2091
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.8822869988179045,
+      "learning_rate": 8.884421731963506e-06,
+      "loss": 0.4789,
+      "step": 2092
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5007876769100963,
+      "learning_rate": 8.883249751764482e-06,
+      "loss": 0.561,
+      "step": 2093
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0611037260584157,
+      "learning_rate": 8.882077233658377e-06,
+      "loss": 0.4767,
+      "step": 2094
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7386468820656826,
+      "learning_rate": 8.880904177807604e-06,
+      "loss": 0.5117,
+      "step": 2095
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9437316431529302,
+      "learning_rate": 8.879730584374655e-06,
+      "loss": 0.7624,
+      "step": 2096
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9177220870121228,
+      "learning_rate": 8.8785564535221e-06,
+      "loss": 0.5588,
+      "step": 2097
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9231574683404207,
+      "learning_rate": 8.877381785412575e-06,
+      "loss": 0.5373,
+      "step": 2098
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3243126270036267,
+      "learning_rate": 8.876206580208798e-06,
+      "loss": 0.4653,
+      "step": 2099
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.650622911641895,
+      "learning_rate": 8.875030838073557e-06,
+      "loss": 0.4928,
+      "step": 2100
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6628312074767986,
+      "learning_rate": 8.873854559169714e-06,
+      "loss": 0.5065,
+      "step": 2101
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2558874968478078,
+      "learning_rate": 8.872677743660209e-06,
+      "loss": 0.434,
+      "step": 2102
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.195442737439743,
+      "learning_rate": 8.871500391708055e-06,
+      "loss": 0.4853,
+      "step": 2103
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.029974243697887,
+      "learning_rate": 8.870322503476337e-06,
+      "loss": 0.5586,
+      "step": 2104
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.773540159716805,
+      "learning_rate": 8.869144079128215e-06,
+      "loss": 0.6171,
+      "step": 2105
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.102428276750225,
+      "learning_rate": 8.867965118826926e-06,
+      "loss": 0.5663,
+      "step": 2106
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.724713434218149,
+      "learning_rate": 8.866785622735779e-06,
+      "loss": 0.5822,
+      "step": 2107
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0739950932009736,
+      "learning_rate": 8.865605591018156e-06,
+      "loss": 0.6491,
+      "step": 2108
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.351888506544126,
+      "learning_rate": 8.864425023837517e-06,
+      "loss": 0.5319,
+      "step": 2109
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2844933481250864,
+      "learning_rate": 8.863243921357394e-06,
+      "loss": 0.6397,
+      "step": 2110
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6797180404802847,
+      "learning_rate": 8.862062283741391e-06,
+      "loss": 0.514,
+      "step": 2111
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8059568255763552,
+      "learning_rate": 8.86088011115319e-06,
+      "loss": 0.5067,
+      "step": 2112
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9218594911112099,
+      "learning_rate": 8.859697403756544e-06,
+      "loss": 0.7539,
+      "step": 2113
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0891119818489234,
+      "learning_rate": 8.858514161715281e-06,
+      "loss": 0.5758,
+      "step": 2114
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.5208706710422684,
+      "learning_rate": 8.857330385193308e-06,
+      "loss": 0.5146,
+      "step": 2115
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7286589153350835,
+      "learning_rate": 8.856146074354594e-06,
+      "loss": 0.3946,
+      "step": 2116
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2660669768271604,
+      "learning_rate": 8.854961229363197e-06,
+      "loss": 0.5251,
+      "step": 2117
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8813426087604475,
+      "learning_rate": 8.853775850383237e-06,
+      "loss": 0.5208,
+      "step": 2118
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.657859256573508,
+      "learning_rate": 8.852589937578913e-06,
+      "loss": 0.4701,
+      "step": 2119
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8968142310605052,
+      "learning_rate": 8.8514034911145e-06,
+      "loss": 0.5376,
+      "step": 2120
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.067689794443828,
+      "learning_rate": 8.850216511154342e-06,
+      "loss": 0.5216,
+      "step": 2121
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2943415002118406,
+      "learning_rate": 8.849028997862858e-06,
+      "loss": 0.5498,
+      "step": 2122
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8104947919952035,
+      "learning_rate": 8.847840951404545e-06,
+      "loss": 0.5608,
+      "step": 2123
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8435694332373755,
+      "learning_rate": 8.84665237194397e-06,
+      "loss": 0.495,
+      "step": 2124
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0456575263110457,
+      "learning_rate": 8.845463259645774e-06,
+      "loss": 0.5385,
+      "step": 2125
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.028474822832002,
+      "learning_rate": 8.844273614674675e-06,
+      "loss": 0.5732,
+      "step": 2126
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.090798974212873,
+      "learning_rate": 8.843083437195458e-06,
+      "loss": 0.6024,
+      "step": 2127
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.668936380744818,
+      "learning_rate": 8.841892727372991e-06,
+      "loss": 0.4695,
+      "step": 2128
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.751565514804749,
+      "learning_rate": 8.84070148537221e-06,
+      "loss": 0.5978,
+      "step": 2129
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.381452027385268,
+      "learning_rate": 8.839509711358122e-06,
+      "loss": 0.4935,
+      "step": 2130
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.616691960192636,
+      "learning_rate": 8.838317405495815e-06,
+      "loss": 0.4206,
+      "step": 2131
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9391272839328788,
+      "learning_rate": 8.837124567950446e-06,
+      "loss": 0.4313,
+      "step": 2132
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.940467547341228,
+      "learning_rate": 8.835931198887247e-06,
+      "loss": 0.4885,
+      "step": 2133
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.735120370072147,
+      "learning_rate": 8.83473729847152e-06,
+      "loss": 0.5789,
+      "step": 2134
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.308185033562398,
+      "learning_rate": 8.833542866868649e-06,
+      "loss": 0.5634,
+      "step": 2135
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.476049147621378,
+      "learning_rate": 8.832347904244082e-06,
+      "loss": 0.4692,
+      "step": 2136
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8363709659987673,
+      "learning_rate": 8.83115241076335e-06,
+      "loss": 0.6006,
+      "step": 2137
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7435882880990088,
+      "learning_rate": 8.829956386592047e-06,
+      "loss": 0.4676,
+      "step": 2138
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8961689323625917,
+      "learning_rate": 8.82875983189585e-06,
+      "loss": 0.6213,
+      "step": 2139
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.467323656668091,
+      "learning_rate": 8.827562746840506e-06,
+      "loss": 0.5385,
+      "step": 2140
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9765200786224564,
+      "learning_rate": 8.82636513159183e-06,
+      "loss": 0.5186,
+      "step": 2141
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.043292190539482,
+      "learning_rate": 8.825166986315721e-06,
+      "loss": 0.4912,
+      "step": 2142
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6994357588549216,
+      "learning_rate": 8.82396831117814e-06,
+      "loss": 0.4896,
+      "step": 2143
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5122144161829592,
+      "learning_rate": 8.822769106345135e-06,
+      "loss": 0.4832,
+      "step": 2144
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8802874125682632,
+      "learning_rate": 8.821569371982815e-06,
+      "loss": 0.4965,
+      "step": 2145
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8758906265636286,
+      "learning_rate": 8.820369108257366e-06,
+      "loss": 0.6344,
+      "step": 2146
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6135766318094578,
+      "learning_rate": 8.819168315335051e-06,
+      "loss": 0.4487,
+      "step": 2147
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.465716548330362,
+      "learning_rate": 8.817966993382202e-06,
+      "loss": 0.5274,
+      "step": 2148
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2927988617937913,
+      "learning_rate": 8.816765142565226e-06,
+      "loss": 0.4902,
+      "step": 2149
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1612341561826502,
+      "learning_rate": 8.815562763050603e-06,
+      "loss": 0.5428,
+      "step": 2150
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5270600645006445,
+      "learning_rate": 8.814359855004889e-06,
+      "loss": 0.4204,
+      "step": 2151
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.944573290441661,
+      "learning_rate": 8.813156418594706e-06,
+      "loss": 0.5515,
+      "step": 2152
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4098364177809137,
+      "learning_rate": 8.811952453986758e-06,
+      "loss": 0.4848,
+      "step": 2153
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0745868643885954,
+      "learning_rate": 8.810747961347816e-06,
+      "loss": 0.4777,
+      "step": 2154
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5357421598195296,
+      "learning_rate": 8.809542940844727e-06,
+      "loss": 0.5258,
+      "step": 2155
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9120088992301487,
+      "learning_rate": 8.808337392644408e-06,
+      "loss": 0.55,
+      "step": 2156
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5265032109948617,
+      "learning_rate": 8.807131316913856e-06,
+      "loss": 0.5494,
+      "step": 2157
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2353621047324186,
+      "learning_rate": 8.80592471382013e-06,
+      "loss": 0.5527,
+      "step": 2158
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5875737619756314,
+      "learning_rate": 8.804717583530373e-06,
+      "loss": 0.6356,
+      "step": 2159
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9408757215609294,
+      "learning_rate": 8.803509926211796e-06,
+      "loss": 0.4903,
+      "step": 2160
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1317363658218245,
+      "learning_rate": 8.802301742031682e-06,
+      "loss": 0.5679,
+      "step": 2161
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8986004927487693,
+      "learning_rate": 8.80109303115739e-06,
+      "loss": 0.6436,
+      "step": 2162
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7164833239885695,
+      "learning_rate": 8.799883793756349e-06,
+      "loss": 0.4657,
+      "step": 2163
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.55405859465503,
+      "learning_rate": 8.798674029996064e-06,
+      "loss": 0.5128,
+      "step": 2164
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.843873193538699,
+      "learning_rate": 8.79746374004411e-06,
+      "loss": 0.5871,
+      "step": 2165
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.743985771775315,
+      "learning_rate": 8.796252924068135e-06,
+      "loss": 0.4725,
+      "step": 2166
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.475662863261369,
+      "learning_rate": 8.795041582235864e-06,
+      "loss": 0.4312,
+      "step": 2167
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7019028946632286,
+      "learning_rate": 8.79382971471509e-06,
+      "loss": 0.5354,
+      "step": 2168
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9559217907331918,
+      "learning_rate": 8.792617321673682e-06,
+      "loss": 0.761,
+      "step": 2169
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7933554600853572,
+      "learning_rate": 8.791404403279577e-06,
+      "loss": 0.5847,
+      "step": 2170
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.77837729947464,
+      "learning_rate": 8.790190959700793e-06,
+      "loss": 0.4593,
+      "step": 2171
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5767361983000359,
+      "learning_rate": 8.788976991105414e-06,
+      "loss": 0.523,
+      "step": 2172
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.097111106381782,
+      "learning_rate": 8.787762497661598e-06,
+      "loss": 0.5108,
+      "step": 2173
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0104568643185368,
+      "learning_rate": 8.786547479537574e-06,
+      "loss": 0.5184,
+      "step": 2174
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7355274394341116,
+      "learning_rate": 8.785331936901652e-06,
+      "loss": 0.4882,
+      "step": 2175
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8474643448813841,
+      "learning_rate": 8.784115869922206e-06,
+      "loss": 0.7046,
+      "step": 2176
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6421077280285923,
+      "learning_rate": 8.782899278767685e-06,
+      "loss": 0.5535,
+      "step": 2177
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.171385895464419,
+      "learning_rate": 8.78168216360661e-06,
+      "loss": 0.4936,
+      "step": 2178
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0497486600950787,
+      "learning_rate": 8.780464524607577e-06,
+      "loss": 0.5039,
+      "step": 2179
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8961815971980327,
+      "learning_rate": 8.779246361939253e-06,
+      "loss": 0.4925,
+      "step": 2180
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.885680286055336,
+      "learning_rate": 8.778027675770378e-06,
+      "loss": 0.5283,
+      "step": 2181
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.7321524773604655,
+      "learning_rate": 8.776808466269761e-06,
+      "loss": 0.6226,
+      "step": 2182
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2737846385809206,
+      "learning_rate": 8.775588733606293e-06,
+      "loss": 0.5431,
+      "step": 2183
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.192855319747856,
+      "learning_rate": 8.774368477948926e-06,
+      "loss": 0.5401,
+      "step": 2184
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0143240585962805,
+      "learning_rate": 8.773147699466692e-06,
+      "loss": 0.504,
+      "step": 2185
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2946469364563207,
+      "learning_rate": 8.771926398328691e-06,
+      "loss": 0.6081,
+      "step": 2186
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6942268083033827,
+      "learning_rate": 8.770704574704099e-06,
+      "loss": 0.4941,
+      "step": 2187
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3061919630684082,
+      "learning_rate": 8.769482228762163e-06,
+      "loss": 0.4969,
+      "step": 2188
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6007160236864182,
+      "learning_rate": 8.7682593606722e-06,
+      "loss": 0.5078,
+      "step": 2189
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8529048852575227,
+      "learning_rate": 8.767035970603606e-06,
+      "loss": 0.5192,
+      "step": 2190
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5756828269871974,
+      "learning_rate": 8.765812058725839e-06,
+      "loss": 0.5312,
+      "step": 2191
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.117732348710527,
+      "learning_rate": 8.764587625208439e-06,
+      "loss": 0.5425,
+      "step": 2192
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.902167141575285,
+      "learning_rate": 8.763362670221014e-06,
+      "loss": 0.5293,
+      "step": 2193
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.688685656513306,
+      "learning_rate": 8.762137193933241e-06,
+      "loss": 0.4827,
+      "step": 2194
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8012190614592067,
+      "learning_rate": 8.760911196514879e-06,
+      "loss": 0.5124,
+      "step": 2195
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0205830095607547,
+      "learning_rate": 8.759684678135746e-06,
+      "loss": 0.4967,
+      "step": 2196
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7050477917729272,
+      "learning_rate": 8.758457638965745e-06,
+      "loss": 0.4034,
+      "step": 2197
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3596790857427643,
+      "learning_rate": 8.757230079174843e-06,
+      "loss": 0.6195,
+      "step": 2198
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8843918112964191,
+      "learning_rate": 8.75600199893308e-06,
+      "loss": 0.7156,
+      "step": 2199
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7178972747712413,
+      "learning_rate": 8.754773398410572e-06,
+      "loss": 0.5336,
+      "step": 2200
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.08424504958815,
+      "learning_rate": 8.753544277777501e-06,
+      "loss": 0.5337,
+      "step": 2201
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2678211962938457,
+      "learning_rate": 8.752314637204129e-06,
+      "loss": 0.4385,
+      "step": 2202
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0941350796752887,
+      "learning_rate": 8.751084476860782e-06,
+      "loss": 0.5387,
+      "step": 2203
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.471688181518448,
+      "learning_rate": 8.749853796917864e-06,
+      "loss": 0.5868,
+      "step": 2204
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3698609119640057,
+      "learning_rate": 8.748622597545847e-06,
+      "loss": 0.4551,
+      "step": 2205
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.645698414460901,
+      "learning_rate": 8.747390878915277e-06,
+      "loss": 0.5019,
+      "step": 2206
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6839910020961422,
+      "learning_rate": 8.746158641196771e-06,
+      "loss": 0.5178,
+      "step": 2207
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.480306180187886,
+      "learning_rate": 8.74492588456102e-06,
+      "loss": 0.5211,
+      "step": 2208
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2970142696231868,
+      "learning_rate": 8.743692609178785e-06,
+      "loss": 0.5558,
+      "step": 2209
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.99137543856927,
+      "learning_rate": 8.742458815220895e-06,
+      "loss": 0.5267,
+      "step": 2210
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.592975607747482,
+      "learning_rate": 8.74122450285826e-06,
+      "loss": 0.4719,
+      "step": 2211
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8770709924374213,
+      "learning_rate": 8.739989672261855e-06,
+      "loss": 0.7568,
+      "step": 2212
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8140789143636114,
+      "learning_rate": 8.738754323602728e-06,
+      "loss": 0.4539,
+      "step": 2213
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.231591365591911,
+      "learning_rate": 8.737518457052e-06,
+      "loss": 0.4743,
+      "step": 2214
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4534637528693772,
+      "learning_rate": 8.736282072780863e-06,
+      "loss": 0.5025,
+      "step": 2215
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.155101418896837,
+      "learning_rate": 8.73504517096058e-06,
+      "loss": 0.6165,
+      "step": 2216
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.059042497669161,
+      "learning_rate": 8.733807751762486e-06,
+      "loss": 0.6516,
+      "step": 2217
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3300278755963486,
+      "learning_rate": 8.73256981535799e-06,
+      "loss": 0.5414,
+      "step": 2218
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5562905470061332,
+      "learning_rate": 8.73133136191857e-06,
+      "loss": 0.4712,
+      "step": 2219
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1734657590601176,
+      "learning_rate": 8.730092391615776e-06,
+      "loss": 0.5508,
+      "step": 2220
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7901974916144585,
+      "learning_rate": 8.728852904621227e-06,
+      "loss": 0.4144,
+      "step": 2221
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.363687218327753,
+      "learning_rate": 8.727612901106623e-06,
+      "loss": 0.541,
+      "step": 2222
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.144040794938621,
+      "learning_rate": 8.726372381243726e-06,
+      "loss": 0.6474,
+      "step": 2223
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.056204823364584,
+      "learning_rate": 8.72513134520437e-06,
+      "loss": 0.4832,
+      "step": 2224
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8902168082286424,
+      "learning_rate": 8.723889793160465e-06,
+      "loss": 0.4652,
+      "step": 2225
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7531965264178277,
+      "learning_rate": 8.722647725283993e-06,
+      "loss": 0.5001,
+      "step": 2226
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.032572650704824,
+      "learning_rate": 8.721405141747001e-06,
+      "loss": 0.6845,
+      "step": 2227
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7632299519330001,
+      "learning_rate": 8.720162042721614e-06,
+      "loss": 0.4882,
+      "step": 2228
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.148595113215755,
+      "learning_rate": 8.718918428380025e-06,
+      "loss": 0.5327,
+      "step": 2229
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6120850373403077,
+      "learning_rate": 8.7176742988945e-06,
+      "loss": 0.4537,
+      "step": 2230
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9390430684289897,
+      "learning_rate": 8.716429654437375e-06,
+      "loss": 0.5465,
+      "step": 2231
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8441414791594257,
+      "learning_rate": 8.715184495181057e-06,
+      "loss": 0.4298,
+      "step": 2232
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2890720875046706,
+      "learning_rate": 8.713938821298027e-06,
+      "loss": 0.588,
+      "step": 2233
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7129195311069367,
+      "learning_rate": 8.712692632960835e-06,
+      "loss": 0.5146,
+      "step": 2234
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.000566059950348,
+      "learning_rate": 8.711445930342101e-06,
+      "loss": 0.5246,
+      "step": 2235
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.796835364993895,
+      "learning_rate": 8.710198713614522e-06,
+      "loss": 0.4904,
+      "step": 2236
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4175145125566138,
+      "learning_rate": 8.708950982950858e-06,
+      "loss": 0.5606,
+      "step": 2237
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9912996063405612,
+      "learning_rate": 8.707702738523948e-06,
+      "loss": 0.4584,
+      "step": 2238
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7217716365216131,
+      "learning_rate": 8.706453980506695e-06,
+      "loss": 0.6064,
+      "step": 2239
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.46553766365681,
+      "learning_rate": 8.70520470907208e-06,
+      "loss": 0.5136,
+      "step": 2240
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1466332129871484,
+      "learning_rate": 8.70395492439315e-06,
+      "loss": 0.5926,
+      "step": 2241
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.940713149852987,
+      "learning_rate": 8.702704626643024e-06,
+      "loss": 0.5317,
+      "step": 2242
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.428791463967903,
+      "learning_rate": 8.701453815994896e-06,
+      "loss": 0.4816,
+      "step": 2243
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2641387992645954,
+      "learning_rate": 8.700202492622025e-06,
+      "loss": 0.5314,
+      "step": 2244
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6115483631304746,
+      "learning_rate": 8.698950656697748e-06,
+      "loss": 0.6277,
+      "step": 2245
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.18190515704846,
+      "learning_rate": 8.697698308395466e-06,
+      "loss": 0.6492,
+      "step": 2246
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.287975324377037,
+      "learning_rate": 8.696445447888652e-06,
+      "loss": 0.5203,
+      "step": 2247
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5992541174246715,
+      "learning_rate": 8.695192075350857e-06,
+      "loss": 0.5286,
+      "step": 2248
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7529838509972064,
+      "learning_rate": 8.693938190955698e-06,
+      "loss": 0.5399,
+      "step": 2249
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8510962430389155,
+      "learning_rate": 8.692683794876857e-06,
+      "loss": 0.6027,
+      "step": 2250
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.698955514609625,
+      "learning_rate": 8.691428887288098e-06,
+      "loss": 0.5755,
+      "step": 2251
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3616661531634584,
+      "learning_rate": 8.69017346836325e-06,
+      "loss": 0.4903,
+      "step": 2252
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7542187771166797,
+      "learning_rate": 8.68891753827621e-06,
+      "loss": 0.5065,
+      "step": 2253
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.125090840983156,
+      "learning_rate": 8.687661097200952e-06,
+      "loss": 0.5286,
+      "step": 2254
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8286988748350745,
+      "learning_rate": 8.686404145311517e-06,
+      "loss": 0.527,
+      "step": 2255
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7091268135298812,
+      "learning_rate": 8.68514668278202e-06,
+      "loss": 0.5475,
+      "step": 2256
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7618261876615084,
+      "learning_rate": 8.683888709786642e-06,
+      "loss": 0.4964,
+      "step": 2257
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.024407347856764,
+      "learning_rate": 8.682630226499638e-06,
+      "loss": 0.4604,
+      "step": 2258
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8878587632013193,
+      "learning_rate": 8.681371233095334e-06,
+      "loss": 0.4661,
+      "step": 2259
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6492027137339098,
+      "learning_rate": 8.680111729748122e-06,
+      "loss": 0.4795,
+      "step": 2260
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.867430839523205,
+      "learning_rate": 8.678851716632473e-06,
+      "loss": 0.4868,
+      "step": 2261
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0513182290720677,
+      "learning_rate": 8.677591193922921e-06,
+      "loss": 0.4907,
+      "step": 2262
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8850911019899388,
+      "learning_rate": 8.676330161794073e-06,
+      "loss": 0.509,
+      "step": 2263
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.329901247599794,
+      "learning_rate": 8.675068620420609e-06,
+      "loss": 0.5417,
+      "step": 2264
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9145404068485052,
+      "learning_rate": 8.673806569977274e-06,
+      "loss": 0.5234,
+      "step": 2265
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8430370504766014,
+      "learning_rate": 8.67254401063889e-06,
+      "loss": 0.6163,
+      "step": 2266
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.754607210358045,
+      "learning_rate": 8.671280942580347e-06,
+      "loss": 0.5156,
+      "step": 2267
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.733711129031052,
+      "learning_rate": 8.670017365976602e-06,
+      "loss": 0.6012,
+      "step": 2268
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6500512240249596,
+      "learning_rate": 8.66875328100269e-06,
+      "loss": 0.5834,
+      "step": 2269
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1601639218713133,
+      "learning_rate": 8.667488687833705e-06,
+      "loss": 0.5338,
+      "step": 2270
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4410368829293607,
+      "learning_rate": 8.666223586644824e-06,
+      "loss": 0.5383,
+      "step": 2271
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.181020542323042,
+      "learning_rate": 8.664957977611289e-06,
+      "loss": 0.5309,
+      "step": 2272
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.764369803452699,
+      "learning_rate": 8.663691860908406e-06,
+      "loss": 0.5299,
+      "step": 2273
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7744264623422263,
+      "learning_rate": 8.662425236711562e-06,
+      "loss": 0.5963,
+      "step": 2274
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8626387677397316,
+      "learning_rate": 8.66115810519621e-06,
+      "loss": 0.5903,
+      "step": 2275
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8575449438674587,
+      "learning_rate": 8.65989046653787e-06,
+      "loss": 0.5327,
+      "step": 2276
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.096722196049806,
+      "learning_rate": 8.658622320912138e-06,
+      "loss": 0.5095,
+      "step": 2277
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.005229864248042,
+      "learning_rate": 8.657353668494674e-06,
+      "loss": 0.4656,
+      "step": 2278
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0949252956119375,
+      "learning_rate": 8.656084509461215e-06,
+      "loss": 0.5408,
+      "step": 2279
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5034306042431147,
+      "learning_rate": 8.654814843987563e-06,
+      "loss": 0.5324,
+      "step": 2280
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0897484214647433,
+      "learning_rate": 8.653544672249589e-06,
+      "loss": 0.5591,
+      "step": 2281
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.080335148701125,
+      "learning_rate": 8.652273994423244e-06,
+      "loss": 0.569,
+      "step": 2282
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.008595984347143,
+      "learning_rate": 8.651002810684535e-06,
+      "loss": 0.471,
+      "step": 2283
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.031010690628849,
+      "learning_rate": 8.64973112120955e-06,
+      "loss": 0.5708,
+      "step": 2284
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9948332206862486,
+      "learning_rate": 8.648458926174441e-06,
+      "loss": 0.7799,
+      "step": 2285
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0396332652716818,
+      "learning_rate": 8.647186225755435e-06,
+      "loss": 0.4751,
+      "step": 2286
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7548931373175751,
+      "learning_rate": 8.645913020128825e-06,
+      "loss": 0.5011,
+      "step": 2287
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0038049262412314,
+      "learning_rate": 8.644639309470975e-06,
+      "loss": 0.6107,
+      "step": 2288
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8680987325093685,
+      "learning_rate": 8.643365093958317e-06,
+      "loss": 0.5834,
+      "step": 2289
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.910833530384027,
+      "learning_rate": 8.64209037376736e-06,
+      "loss": 0.4698,
+      "step": 2290
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0785730280830625,
+      "learning_rate": 8.640815149074673e-06,
+      "loss": 0.5217,
+      "step": 2291
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0848790976934275,
+      "learning_rate": 8.639539420056902e-06,
+      "loss": 0.5671,
+      "step": 2292
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8483965979659747,
+      "learning_rate": 8.638263186890763e-06,
+      "loss": 0.5116,
+      "step": 2293
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3884745834370387,
+      "learning_rate": 8.636986449753035e-06,
+      "loss": 0.5732,
+      "step": 2294
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9756541069138531,
+      "learning_rate": 8.635709208820576e-06,
+      "loss": 0.6132,
+      "step": 2295
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.93832137629286,
+      "learning_rate": 8.634431464270308e-06,
+      "loss": 0.5225,
+      "step": 2296
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8992525614542254,
+      "learning_rate": 8.63315321627922e-06,
+      "loss": 0.7111,
+      "step": 2297
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.105046768195289,
+      "learning_rate": 8.63187446502438e-06,
+      "loss": 0.4815,
+      "step": 2298
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.070912678065797,
+      "learning_rate": 8.630595210682918e-06,
+      "loss": 0.5886,
+      "step": 2299
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7278541391421343,
+      "learning_rate": 8.629315453432034e-06,
+      "loss": 0.5294,
+      "step": 2300
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.56117071641269,
+      "learning_rate": 8.628035193449005e-06,
+      "loss": 0.4999,
+      "step": 2301
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8518025863754612,
+      "learning_rate": 8.626754430911169e-06,
+      "loss": 0.6341,
+      "step": 2302
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7061760875116814,
+      "learning_rate": 8.625473165995935e-06,
+      "loss": 0.5476,
+      "step": 2303
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4226513940645313,
+      "learning_rate": 8.624191398880788e-06,
+      "loss": 0.5364,
+      "step": 2304
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8058284646508111,
+      "learning_rate": 8.622909129743275e-06,
+      "loss": 0.4586,
+      "step": 2305
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.328935292027457,
+      "learning_rate": 8.621626358761018e-06,
+      "loss": 0.5346,
+      "step": 2306
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4249907731636644,
+      "learning_rate": 8.620343086111704e-06,
+      "loss": 0.5571,
+      "step": 2307
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.843600469132354,
+      "learning_rate": 8.619059311973095e-06,
+      "loss": 0.5594,
+      "step": 2308
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7841753620414624,
+      "learning_rate": 8.617775036523014e-06,
+      "loss": 0.5243,
+      "step": 2309
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9806771663835443,
+      "learning_rate": 8.616490259939364e-06,
+      "loss": 0.6093,
+      "step": 2310
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.10706723258942,
+      "learning_rate": 8.615204982400108e-06,
+      "loss": 0.5568,
+      "step": 2311
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.107515498865774,
+      "learning_rate": 8.613919204083286e-06,
+      "loss": 0.4663,
+      "step": 2312
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8084497120770857,
+      "learning_rate": 8.612632925166999e-06,
+      "loss": 0.5211,
+      "step": 2313
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9652682538730981,
+      "learning_rate": 8.611346145829427e-06,
+      "loss": 0.5909,
+      "step": 2314
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7769822561053341,
+      "learning_rate": 8.61005886624881e-06,
+      "loss": 0.529,
+      "step": 2315
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7625781251490886,
+      "learning_rate": 8.608771086603466e-06,
+      "loss": 0.5589,
+      "step": 2316
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7232410926821211,
+      "learning_rate": 8.607482807071777e-06,
+      "loss": 0.4574,
+      "step": 2317
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0769521026562483,
+      "learning_rate": 8.606194027832192e-06,
+      "loss": 0.5277,
+      "step": 2318
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1481645897457073,
+      "learning_rate": 8.604904749063237e-06,
+      "loss": 0.553,
+      "step": 2319
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6829855750905547,
+      "learning_rate": 8.6036149709435e-06,
+      "loss": 0.499,
+      "step": 2320
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9474046302458858,
+      "learning_rate": 8.60232469365164e-06,
+      "loss": 0.527,
+      "step": 2321
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.279835887908195,
+      "learning_rate": 8.601033917366389e-06,
+      "loss": 0.5544,
+      "step": 2322
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9463099680217628,
+      "learning_rate": 8.599742642266544e-06,
+      "loss": 0.58,
+      "step": 2323
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9580128016449191,
+      "learning_rate": 8.59845086853097e-06,
+      "loss": 0.7037,
+      "step": 2324
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.24531561197382,
+      "learning_rate": 8.597158596338605e-06,
+      "loss": 0.5977,
+      "step": 2325
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8514204764596625,
+      "learning_rate": 8.595865825868455e-06,
+      "loss": 0.5622,
+      "step": 2326
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2475751940197632,
+      "learning_rate": 8.594572557299594e-06,
+      "loss": 0.5065,
+      "step": 2327
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9158471428342122,
+      "learning_rate": 8.593278790811164e-06,
+      "loss": 0.7299,
+      "step": 2328
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.813242227312915,
+      "learning_rate": 8.591984526582378e-06,
+      "loss": 0.4803,
+      "step": 2329
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.645786698014883,
+      "learning_rate": 8.59068976479252e-06,
+      "loss": 0.4947,
+      "step": 2330
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.93319902528064,
+      "learning_rate": 8.589394505620935e-06,
+      "loss": 0.5693,
+      "step": 2331
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9880754684872175,
+      "learning_rate": 8.588098749247045e-06,
+      "loss": 0.5421,
+      "step": 2332
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.533670935320121,
+      "learning_rate": 8.586802495850339e-06,
+      "loss": 0.5184,
+      "step": 2333
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9809616055970773,
+      "learning_rate": 8.585505745610372e-06,
+      "loss": 0.575,
+      "step": 2334
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8981138984920811,
+      "learning_rate": 8.58420849870677e-06,
+      "loss": 0.4851,
+      "step": 2335
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.086166785264737,
+      "learning_rate": 8.582910755319228e-06,
+      "loss": 0.6033,
+      "step": 2336
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8260800458352073,
+      "learning_rate": 8.581612515627509e-06,
+      "loss": 0.4789,
+      "step": 2337
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.659097195306069,
+      "learning_rate": 8.580313779811444e-06,
+      "loss": 0.5152,
+      "step": 2338
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.18358648363183,
+      "learning_rate": 8.579014548050934e-06,
+      "loss": 0.5565,
+      "step": 2339
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5496818491857096,
+      "learning_rate": 8.57771482052595e-06,
+      "loss": 0.5464,
+      "step": 2340
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8022724038625273,
+      "learning_rate": 8.576414597416527e-06,
+      "loss": 0.5251,
+      "step": 2341
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.799153134386961,
+      "learning_rate": 8.575113878902776e-06,
+      "loss": 0.5009,
+      "step": 2342
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4829811034135174,
+      "learning_rate": 8.573812665164867e-06,
+      "loss": 0.5081,
+      "step": 2343
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4509325387654677,
+      "learning_rate": 8.572510956383048e-06,
+      "loss": 0.4958,
+      "step": 2344
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.097664324527126,
+      "learning_rate": 8.57120875273763e-06,
+      "loss": 0.617,
+      "step": 2345
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.348344261473166,
+      "learning_rate": 8.569906054408994e-06,
+      "loss": 0.6082,
+      "step": 2346
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.672057391635827,
+      "learning_rate": 8.568602861577589e-06,
+      "loss": 0.5956,
+      "step": 2347
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0069915450531437,
+      "learning_rate": 8.567299174423936e-06,
+      "loss": 0.5148,
+      "step": 2348
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9114641404024724,
+      "learning_rate": 8.565994993128617e-06,
+      "loss": 0.4413,
+      "step": 2349
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0249486878362815,
+      "learning_rate": 8.564690317872289e-06,
+      "loss": 0.4527,
+      "step": 2350
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.330168692292167,
+      "learning_rate": 8.563385148835677e-06,
+      "loss": 0.4136,
+      "step": 2351
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4738136289782906,
+      "learning_rate": 8.562079486199571e-06,
+      "loss": 0.6065,
+      "step": 2352
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3352898443116574,
+      "learning_rate": 8.56077333014483e-06,
+      "loss": 0.5319,
+      "step": 2353
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.003509521835223,
+      "learning_rate": 8.559466680852386e-06,
+      "loss": 0.4713,
+      "step": 2354
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6677954078827002,
+      "learning_rate": 8.558159538503234e-06,
+      "loss": 0.4585,
+      "step": 2355
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8270435954514523,
+      "learning_rate": 8.556851903278437e-06,
+      "loss": 0.5158,
+      "step": 2356
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5977877718483098,
+      "learning_rate": 8.555543775359132e-06,
+      "loss": 0.4162,
+      "step": 2357
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9839522944195285,
+      "learning_rate": 8.55423515492652e-06,
+      "loss": 0.5185,
+      "step": 2358
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9658390005673964,
+      "learning_rate": 8.552926042161868e-06,
+      "loss": 0.7499,
+      "step": 2359
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.87423004591279,
+      "learning_rate": 8.551616437246515e-06,
+      "loss": 0.6412,
+      "step": 2360
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8031405105241216,
+      "learning_rate": 8.550306340361872e-06,
+      "loss": 0.5286,
+      "step": 2361
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.651957743700889,
+      "learning_rate": 8.548995751689406e-06,
+      "loss": 0.4529,
+      "step": 2362
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5661871773717315,
+      "learning_rate": 8.547684671410665e-06,
+      "loss": 0.4987,
+      "step": 2363
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9232183513384353,
+      "learning_rate": 8.54637309970726e-06,
+      "loss": 0.5896,
+      "step": 2364
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9020394888930264,
+      "learning_rate": 8.545061036760863e-06,
+      "loss": 0.7447,
+      "step": 2365
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9091640547450839,
+      "learning_rate": 8.543748482753229e-06,
+      "loss": 0.5526,
+      "step": 2366
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8455647781615331,
+      "learning_rate": 8.542435437866166e-06,
+      "loss": 0.72,
+      "step": 2367
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6507528582420827,
+      "learning_rate": 8.541121902281562e-06,
+      "loss": 0.4673,
+      "step": 2368
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.679356007242605,
+      "learning_rate": 8.539807876181363e-06,
+      "loss": 0.5614,
+      "step": 2369
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9176614700352168,
+      "learning_rate": 8.538493359747592e-06,
+      "loss": 0.4389,
+      "step": 2370
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2720074443151357,
+      "learning_rate": 8.537178353162334e-06,
+      "loss": 0.4873,
+      "step": 2371
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.888301903142672,
+      "learning_rate": 8.535862856607742e-06,
+      "loss": 0.5817,
+      "step": 2372
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7757660656049783,
+      "learning_rate": 8.534546870266041e-06,
+      "loss": 0.5896,
+      "step": 2373
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8228889122298935,
+      "learning_rate": 8.533230394319518e-06,
+      "loss": 0.5099,
+      "step": 2374
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.90402632513311,
+      "learning_rate": 8.531913428950533e-06,
+      "loss": 0.7348,
+      "step": 2375
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9428790711454256,
+      "learning_rate": 8.530595974341512e-06,
+      "loss": 0.4974,
+      "step": 2376
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8614049628728114,
+      "learning_rate": 8.529278030674947e-06,
+      "loss": 0.7531,
+      "step": 2377
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5968356956824492,
+      "learning_rate": 8.527959598133403e-06,
+      "loss": 0.4661,
+      "step": 2378
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.125853941818109,
+      "learning_rate": 8.526640676899505e-06,
+      "loss": 0.5615,
+      "step": 2379
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0342213784685574,
+      "learning_rate": 8.525321267155952e-06,
+      "loss": 0.5489,
+      "step": 2380
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.795539591551683,
+      "learning_rate": 8.524001369085506e-06,
+      "loss": 0.6115,
+      "step": 2381
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8519074764310129,
+      "learning_rate": 8.522680982871002e-06,
+      "loss": 0.686,
+      "step": 2382
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.188530242911944,
+      "learning_rate": 8.521360108695339e-06,
+      "loss": 0.5787,
+      "step": 2383
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.872561488825852,
+      "learning_rate": 8.520038746741482e-06,
+      "loss": 0.4544,
+      "step": 2384
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.013866660018719,
+      "learning_rate": 8.518716897192469e-06,
+      "loss": 0.536,
+      "step": 2385
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8494793195374575,
+      "learning_rate": 8.5173945602314e-06,
+      "loss": 0.5939,
+      "step": 2386
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9832176665558487,
+      "learning_rate": 8.516071736041447e-06,
+      "loss": 0.4244,
+      "step": 2387
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.066641057995938,
+      "learning_rate": 8.514748424805844e-06,
+      "loss": 0.5044,
+      "step": 2388
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5084331331194893,
+      "learning_rate": 8.5134246267079e-06,
+      "loss": 0.5055,
+      "step": 2389
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0971906905476803,
+      "learning_rate": 8.512100341930985e-06,
+      "loss": 0.5392,
+      "step": 2390
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2142955934709714,
+      "learning_rate": 8.510775570658538e-06,
+      "loss": 0.5431,
+      "step": 2391
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6255438205354569,
+      "learning_rate": 8.509450313074065e-06,
+      "loss": 0.4706,
+      "step": 2392
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1440566210055514,
+      "learning_rate": 8.508124569361147e-06,
+      "loss": 0.5368,
+      "step": 2393
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.580442036293708,
+      "learning_rate": 8.50679833970342e-06,
+      "loss": 0.5688,
+      "step": 2394
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7681383725357012,
+      "learning_rate": 8.505471624284593e-06,
+      "loss": 0.4565,
+      "step": 2395
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8951924092729338,
+      "learning_rate": 8.504144423288443e-06,
+      "loss": 0.5529,
+      "step": 2396
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2937428508779907,
+      "learning_rate": 8.502816736898816e-06,
+      "loss": 0.5319,
+      "step": 2397
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7388437012805387,
+      "learning_rate": 8.50148856529962e-06,
+      "loss": 0.5622,
+      "step": 2398
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.419220188531338,
+      "learning_rate": 8.500159908674836e-06,
+      "loss": 0.6096,
+      "step": 2399
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8797098042071638,
+      "learning_rate": 8.498830767208507e-06,
+      "loss": 0.4853,
+      "step": 2400
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.035074734741205,
+      "learning_rate": 8.497501141084746e-06,
+      "loss": 0.5668,
+      "step": 2401
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6767940478062733,
+      "learning_rate": 8.496171030487734e-06,
+      "loss": 0.5177,
+      "step": 2402
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7984699733963958,
+      "learning_rate": 8.494840435601714e-06,
+      "loss": 0.5678,
+      "step": 2403
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.344009074710361,
+      "learning_rate": 8.493509356611005e-06,
+      "loss": 0.6263,
+      "step": 2404
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8735917421469612,
+      "learning_rate": 8.492177793699982e-06,
+      "loss": 0.5414,
+      "step": 2405
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8241710857775186,
+      "learning_rate": 8.490845747053098e-06,
+      "loss": 0.5194,
+      "step": 2406
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2042176118048005,
+      "learning_rate": 8.489513216854866e-06,
+      "loss": 0.4905,
+      "step": 2407
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1300988496740922,
+      "learning_rate": 8.488180203289867e-06,
+      "loss": 0.5169,
+      "step": 2408
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.660090732493645,
+      "learning_rate": 8.48684670654275e-06,
+      "loss": 0.5449,
+      "step": 2409
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.085723350946313,
+      "learning_rate": 8.485512726798231e-06,
+      "loss": 0.4833,
+      "step": 2410
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.765503452891893,
+      "learning_rate": 8.484178264241093e-06,
+      "loss": 0.4435,
+      "step": 2411
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.636353243404151,
+      "learning_rate": 8.482843319056187e-06,
+      "loss": 0.4904,
+      "step": 2412
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0386061911027396,
+      "learning_rate": 8.481507891428425e-06,
+      "loss": 0.5441,
+      "step": 2413
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1725796535929502,
+      "learning_rate": 8.480171981542794e-06,
+      "loss": 0.5793,
+      "step": 2414
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.836450667300905,
+      "learning_rate": 8.478835589584344e-06,
+      "loss": 0.4223,
+      "step": 2415
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1505424564248234,
+      "learning_rate": 8.477498715738188e-06,
+      "loss": 0.4926,
+      "step": 2416
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.840454782914558,
+      "learning_rate": 8.476161360189514e-06,
+      "loss": 0.4809,
+      "step": 2417
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8818612913039194,
+      "learning_rate": 8.47482352312357e-06,
+      "loss": 0.5504,
+      "step": 2418
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.4565971716407367,
+      "learning_rate": 8.473485204725675e-06,
+      "loss": 0.4423,
+      "step": 2419
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0969523865163753,
+      "learning_rate": 8.47214640518121e-06,
+      "loss": 0.5154,
+      "step": 2420
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0695240527446206,
+      "learning_rate": 8.470807124675626e-06,
+      "loss": 0.6035,
+      "step": 2421
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0045141237032547,
+      "learning_rate": 8.46946736339444e-06,
+      "loss": 0.4661,
+      "step": 2422
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.099087722014555,
+      "learning_rate": 8.468127121523236e-06,
+      "loss": 0.5171,
+      "step": 2423
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7878960983531456,
+      "learning_rate": 8.466786399247663e-06,
+      "loss": 0.4581,
+      "step": 2424
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.971256078618696,
+      "learning_rate": 8.465445196753441e-06,
+      "loss": 0.5106,
+      "step": 2425
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4516316684893837,
+      "learning_rate": 8.464103514226349e-06,
+      "loss": 0.5064,
+      "step": 2426
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.319775154737638,
+      "learning_rate": 8.462761351852238e-06,
+      "loss": 0.5651,
+      "step": 2427
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8302544208145304,
+      "learning_rate": 8.461418709817026e-06,
+      "loss": 0.4807,
+      "step": 2428
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.944570195235223,
+      "learning_rate": 8.460075588306692e-06,
+      "loss": 0.7355,
+      "step": 2429
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.557827092040908,
+      "learning_rate": 8.458731987507287e-06,
+      "loss": 0.4978,
+      "step": 2430
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9066442989457917,
+      "learning_rate": 8.457387907604926e-06,
+      "loss": 0.5129,
+      "step": 2431
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.824117342532886,
+      "learning_rate": 8.456043348785792e-06,
+      "loss": 0.4858,
+      "step": 2432
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1823806355264614,
+      "learning_rate": 8.45469831123613e-06,
+      "loss": 0.6024,
+      "step": 2433
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.718317119718294,
+      "learning_rate": 8.453352795142259e-06,
+      "loss": 0.458,
+      "step": 2434
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2935947015750284,
+      "learning_rate": 8.452006800690554e-06,
+      "loss": 0.4594,
+      "step": 2435
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.981192322218991,
+      "learning_rate": 8.450660328067467e-06,
+      "loss": 0.5005,
+      "step": 2436
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9653612682724262,
+      "learning_rate": 8.449313377459509e-06,
+      "loss": 0.5971,
+      "step": 2437
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7324664975658497,
+      "learning_rate": 8.447965949053258e-06,
+      "loss": 0.534,
+      "step": 2438
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1587540128561247,
+      "learning_rate": 8.446618043035361e-06,
+      "loss": 0.552,
+      "step": 2439
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.447199903022612,
+      "learning_rate": 8.44526965959253e-06,
+      "loss": 0.4602,
+      "step": 2440
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.132934600271061,
+      "learning_rate": 8.443920798911544e-06,
+      "loss": 0.5594,
+      "step": 2441
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.850505870285045,
+      "learning_rate": 8.442571461179243e-06,
+      "loss": 0.5521,
+      "step": 2442
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8880157930258608,
+      "learning_rate": 8.441221646582542e-06,
+      "loss": 0.6758,
+      "step": 2443
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6200432361001056,
+      "learning_rate": 8.439871355308413e-06,
+      "loss": 0.4329,
+      "step": 2444
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3736382223822274,
+      "learning_rate": 8.438520587543901e-06,
+      "loss": 0.5428,
+      "step": 2445
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.231367064261845,
+      "learning_rate": 8.43716934347611e-06,
+      "loss": 0.5207,
+      "step": 2446
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.614487878671032,
+      "learning_rate": 8.43581762329222e-06,
+      "loss": 0.4072,
+      "step": 2447
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.011835224106064,
+      "learning_rate": 8.434465427179465e-06,
+      "loss": 0.5087,
+      "step": 2448
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.56122732967349,
+      "learning_rate": 8.433112755325156e-06,
+      "loss": 0.4995,
+      "step": 2449
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.149617913932202,
+      "learning_rate": 8.431759607916663e-06,
+      "loss": 0.5051,
+      "step": 2450
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.684465741301117,
+      "learning_rate": 8.430405985141422e-06,
+      "loss": 0.5627,
+      "step": 2451
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2400415390707393,
+      "learning_rate": 8.429051887186938e-06,
+      "loss": 0.5304,
+      "step": 2452
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7080036651392563,
+      "learning_rate": 8.427697314240783e-06,
+      "loss": 0.5206,
+      "step": 2453
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.019754708991868,
+      "learning_rate": 8.426342266490588e-06,
+      "loss": 0.4974,
+      "step": 2454
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7825288062337001,
+      "learning_rate": 8.424986744124055e-06,
+      "loss": 0.4479,
+      "step": 2455
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.553590391333902,
+      "learning_rate": 8.423630747328952e-06,
+      "loss": 0.5018,
+      "step": 2456
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.694477646166225,
+      "learning_rate": 8.422274276293112e-06,
+      "loss": 0.5876,
+      "step": 2457
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9353920663937663,
+      "learning_rate": 8.42091733120443e-06,
+      "loss": 0.5347,
+      "step": 2458
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9125974724980463,
+      "learning_rate": 8.419559912250873e-06,
+      "loss": 0.6372,
+      "step": 2459
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6363066527834271,
+      "learning_rate": 8.41820201962047e-06,
+      "loss": 0.517,
+      "step": 2460
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.770651132381932,
+      "learning_rate": 8.416843653501314e-06,
+      "loss": 0.3946,
+      "step": 2461
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7447269718246932,
+      "learning_rate": 8.415484814081567e-06,
+      "loss": 0.5645,
+      "step": 2462
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2579212353928275,
+      "learning_rate": 8.414125501549456e-06,
+      "loss": 0.5223,
+      "step": 2463
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.101878526298239,
+      "learning_rate": 8.412765716093273e-06,
+      "loss": 0.5117,
+      "step": 2464
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7959466524879582,
+      "learning_rate": 8.41140545790137e-06,
+      "loss": 0.479,
+      "step": 2465
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8632316215309777,
+      "learning_rate": 8.410044727162177e-06,
+      "loss": 0.5766,
+      "step": 2466
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.30572320937174,
+      "learning_rate": 8.408683524064178e-06,
+      "loss": 0.4351,
+      "step": 2467
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9783787516107614,
+      "learning_rate": 8.407321848795928e-06,
+      "loss": 0.3994,
+      "step": 2468
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9813384395591465,
+      "learning_rate": 8.405959701546046e-06,
+      "loss": 0.5501,
+      "step": 2469
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0587082069470823,
+      "learning_rate": 8.404597082503216e-06,
+      "loss": 0.4959,
+      "step": 2470
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0284767220773756,
+      "learning_rate": 8.403233991856187e-06,
+      "loss": 0.4635,
+      "step": 2471
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3076544070795073,
+      "learning_rate": 8.401870429793775e-06,
+      "loss": 0.5459,
+      "step": 2472
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.136617051911346,
+      "learning_rate": 8.400506396504862e-06,
+      "loss": 0.4896,
+      "step": 2473
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.793823749821937,
+      "learning_rate": 8.39914189217839e-06,
+      "loss": 0.4948,
+      "step": 2474
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9877011529799353,
+      "learning_rate": 8.397776917003373e-06,
+      "loss": 0.4339,
+      "step": 2475
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1414104109797414,
+      "learning_rate": 8.396411471168885e-06,
+      "loss": 0.4235,
+      "step": 2476
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0776739748904367,
+      "learning_rate": 8.39504555486407e-06,
+      "loss": 0.4944,
+      "step": 2477
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.602588870494956,
+      "learning_rate": 8.39367916827813e-06,
+      "loss": 0.4304,
+      "step": 2478
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9596982500983087,
+      "learning_rate": 8.392312311600342e-06,
+      "loss": 0.5162,
+      "step": 2479
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9459732395447518,
+      "learning_rate": 8.39094498502004e-06,
+      "loss": 0.5254,
+      "step": 2480
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.113600715524274,
+      "learning_rate": 8.389577188726624e-06,
+      "loss": 0.5244,
+      "step": 2481
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8300524743861903,
+      "learning_rate": 8.388208922909565e-06,
+      "loss": 0.6096,
+      "step": 2482
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6704663676474898,
+      "learning_rate": 8.386840187758392e-06,
+      "loss": 0.4765,
+      "step": 2483
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.283560580900437,
+      "learning_rate": 8.385470983462702e-06,
+      "loss": 0.4914,
+      "step": 2484
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.000735906709495,
+      "learning_rate": 8.384101310212159e-06,
+      "loss": 0.6591,
+      "step": 2485
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.186069554083827,
+      "learning_rate": 8.382731168196488e-06,
+      "loss": 0.5405,
+      "step": 2486
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7359400705871633,
+      "learning_rate": 8.381360557605482e-06,
+      "loss": 0.4225,
+      "step": 2487
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7413768391461313,
+      "learning_rate": 8.379989478628995e-06,
+      "loss": 0.4555,
+      "step": 2488
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4667667335876529,
+      "learning_rate": 8.378617931456954e-06,
+      "loss": 0.7928,
+      "step": 2489
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.735496535214493,
+      "learning_rate": 8.37724591627934e-06,
+      "loss": 0.5059,
+      "step": 2490
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.839994480530658,
+      "learning_rate": 8.375873433286208e-06,
+      "loss": 0.52,
+      "step": 2491
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.678638473734695,
+      "learning_rate": 8.374500482667672e-06,
+      "loss": 0.5166,
+      "step": 2492
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2162092490751757,
+      "learning_rate": 8.373127064613915e-06,
+      "loss": 0.5076,
+      "step": 2493
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3200971074787304,
+      "learning_rate": 8.371753179315179e-06,
+      "loss": 0.5531,
+      "step": 2494
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7216342085849654,
+      "learning_rate": 8.370378826961778e-06,
+      "loss": 0.4487,
+      "step": 2495
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1423237589216275,
+      "learning_rate": 8.369004007744087e-06,
+      "loss": 0.5298,
+      "step": 2496
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.645257198526256,
+      "learning_rate": 8.367628721852543e-06,
+      "loss": 0.5699,
+      "step": 2497
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8134696604772362,
+      "learning_rate": 8.36625296947765e-06,
+      "loss": 0.477,
+      "step": 2498
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5414641186239586,
+      "learning_rate": 8.36487675080998e-06,
+      "loss": 0.4389,
+      "step": 2499
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5252367505118072,
+      "learning_rate": 8.363500066040166e-06,
+      "loss": 0.4013,
+      "step": 2500
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.495368216145309,
+      "learning_rate": 8.362122915358905e-06,
+      "loss": 0.528,
+      "step": 2501
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3571232321605056,
+      "learning_rate": 8.360745298956961e-06,
+      "loss": 0.5404,
+      "step": 2502
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1125992136789797,
+      "learning_rate": 8.35936721702516e-06,
+      "loss": 0.6349,
+      "step": 2503
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8142371765059333,
+      "learning_rate": 8.357988669754394e-06,
+      "loss": 0.5174,
+      "step": 2504
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8118932932164502,
+      "learning_rate": 8.356609657335618e-06,
+      "loss": 0.5023,
+      "step": 2505
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3076764111032952,
+      "learning_rate": 8.355230179959854e-06,
+      "loss": 0.605,
+      "step": 2506
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9235744931715504,
+      "learning_rate": 8.353850237818186e-06,
+      "loss": 0.5272,
+      "step": 2507
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.078398450030257,
+      "learning_rate": 8.352469831101766e-06,
+      "loss": 0.4528,
+      "step": 2508
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9531604578798716,
+      "learning_rate": 8.351088960001803e-06,
+      "loss": 0.5168,
+      "step": 2509
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1028441277671313,
+      "learning_rate": 8.34970762470958e-06,
+      "loss": 0.5571,
+      "step": 2510
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.35250411204121,
+      "learning_rate": 8.348325825416437e-06,
+      "loss": 0.4797,
+      "step": 2511
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.195447050556698,
+      "learning_rate": 8.346943562313778e-06,
+      "loss": 0.6018,
+      "step": 2512
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4852129332469812,
+      "learning_rate": 8.34556083559308e-06,
+      "loss": 0.4733,
+      "step": 2513
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.060677580368984,
+      "learning_rate": 8.344177645445873e-06,
+      "loss": 0.6135,
+      "step": 2514
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7837756150638142,
+      "learning_rate": 8.342793992063756e-06,
+      "loss": 0.3539,
+      "step": 2515
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.995582646914153,
+      "learning_rate": 8.341409875638396e-06,
+      "loss": 0.5847,
+      "step": 2516
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9757836101816952,
+      "learning_rate": 8.340025296361519e-06,
+      "loss": 0.4839,
+      "step": 2517
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.081990463370359,
+      "learning_rate": 8.338640254424914e-06,
+      "loss": 0.5003,
+      "step": 2518
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.996171181958468,
+      "learning_rate": 8.337254750020442e-06,
+      "loss": 0.5643,
+      "step": 2519
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8140496662868268,
+      "learning_rate": 8.33586878334002e-06,
+      "loss": 0.5516,
+      "step": 2520
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3990925075918694,
+      "learning_rate": 8.334482354575632e-06,
+      "loss": 0.7757,
+      "step": 2521
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9449977506334513,
+      "learning_rate": 8.333095463919325e-06,
+      "loss": 0.5291,
+      "step": 2522
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9491452355773233,
+      "learning_rate": 8.33170811156321e-06,
+      "loss": 0.5566,
+      "step": 2523
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5212534621021248,
+      "learning_rate": 8.330320297699467e-06,
+      "loss": 0.5415,
+      "step": 2524
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.554886217100242,
+      "learning_rate": 8.328932022520333e-06,
+      "loss": 0.5234,
+      "step": 2525
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3468339085643706,
+      "learning_rate": 8.32754328621811e-06,
+      "loss": 0.5143,
+      "step": 2526
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4750427602865726,
+      "learning_rate": 8.326154088985167e-06,
+      "loss": 0.5359,
+      "step": 2527
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.930860651733599,
+      "learning_rate": 8.324764431013939e-06,
+      "loss": 0.4767,
+      "step": 2528
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1631323063997323,
+      "learning_rate": 8.323374312496915e-06,
+      "loss": 0.5287,
+      "step": 2529
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7063873078612262,
+      "learning_rate": 8.321983733626658e-06,
+      "loss": 0.5289,
+      "step": 2530
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5864732121922116,
+      "learning_rate": 8.32059269459579e-06,
+      "loss": 0.579,
+      "step": 2531
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.9602400789474994,
+      "learning_rate": 8.319201195596997e-06,
+      "loss": 0.6436,
+      "step": 2532
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8289823560157406,
+      "learning_rate": 8.317809236823029e-06,
+      "loss": 0.5489,
+      "step": 2533
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6231852476729802,
+      "learning_rate": 8.3164168184667e-06,
+      "loss": 0.5161,
+      "step": 2534
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5366764576126193,
+      "learning_rate": 8.315023940720887e-06,
+      "loss": 0.5531,
+      "step": 2535
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9488627060720387,
+      "learning_rate": 8.313630603778534e-06,
+      "loss": 0.4573,
+      "step": 2536
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.924078240217396,
+      "learning_rate": 8.31223680783264e-06,
+      "loss": 0.4541,
+      "step": 2537
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.225838060266021,
+      "learning_rate": 8.310842553076282e-06,
+      "loss": 0.4971,
+      "step": 2538
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.4014459222026394,
+      "learning_rate": 8.309447839702583e-06,
+      "loss": 0.5184,
+      "step": 2539
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.3570305539406524,
+      "learning_rate": 8.308052667904743e-06,
+      "loss": 0.5054,
+      "step": 2540
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.057641848920971,
+      "learning_rate": 8.306657037876022e-06,
+      "loss": 0.5212,
+      "step": 2541
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4410708190005943,
+      "learning_rate": 8.30526094980974e-06,
+      "loss": 0.5786,
+      "step": 2542
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6974692521333536,
+      "learning_rate": 8.303864403899284e-06,
+      "loss": 0.588,
+      "step": 2543
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.702940776148582,
+      "learning_rate": 8.302467400338103e-06,
+      "loss": 0.4504,
+      "step": 2544
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9055720062028387,
+      "learning_rate": 8.301069939319709e-06,
+      "loss": 0.5871,
+      "step": 2545
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4313222718474585,
+      "learning_rate": 8.29967202103768e-06,
+      "loss": 0.4795,
+      "step": 2546
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1847211600284964,
+      "learning_rate": 8.298273645685654e-06,
+      "loss": 0.5752,
+      "step": 2547
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5620276305431986,
+      "learning_rate": 8.296874813457333e-06,
+      "loss": 0.5981,
+      "step": 2548
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0790635751117557,
+      "learning_rate": 8.295475524546483e-06,
+      "loss": 0.492,
+      "step": 2549
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6700540372456516,
+      "learning_rate": 8.294075779146937e-06,
+      "loss": 0.4114,
+      "step": 2550
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1810573750147086,
+      "learning_rate": 8.292675577452582e-06,
+      "loss": 0.5181,
+      "step": 2551
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.834411970565417,
+      "learning_rate": 8.291274919657378e-06,
+      "loss": 0.5127,
+      "step": 2552
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.4206051480736432,
+      "learning_rate": 8.289873805955342e-06,
+      "loss": 0.4816,
+      "step": 2553
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8007577976221842,
+      "learning_rate": 8.288472236540556e-06,
+      "loss": 0.5099,
+      "step": 2554
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.17998306811133,
+      "learning_rate": 8.287070211607164e-06,
+      "loss": 0.5325,
+      "step": 2555
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8541425438226002,
+      "learning_rate": 8.285667731349377e-06,
+      "loss": 0.5163,
+      "step": 2556
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9837987984166925,
+      "learning_rate": 8.284264795961464e-06,
+      "loss": 0.4859,
+      "step": 2557
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1085327767185698,
+      "learning_rate": 8.282861405637763e-06,
+      "loss": 0.5985,
+      "step": 2558
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.197077982294476,
+      "learning_rate": 8.281457560572665e-06,
+      "loss": 0.4856,
+      "step": 2559
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0160435763015268,
+      "learning_rate": 8.280053260960636e-06,
+      "loss": 0.5432,
+      "step": 2560
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.424517128906401,
+      "learning_rate": 8.278648506996197e-06,
+      "loss": 0.7695,
+      "step": 2561
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.48023049048259,
+      "learning_rate": 8.277243298873936e-06,
+      "loss": 0.4928,
+      "step": 2562
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2092091953518698,
+      "learning_rate": 8.2758376367885e-06,
+      "loss": 0.4321,
+      "step": 2563
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.112813309972629,
+      "learning_rate": 8.274431520934602e-06,
+      "loss": 0.4979,
+      "step": 2564
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9282668287096787,
+      "learning_rate": 8.273024951507017e-06,
+      "loss": 0.5302,
+      "step": 2565
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8446823223621954,
+      "learning_rate": 8.271617928700581e-06,
+      "loss": 0.7312,
+      "step": 2566
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.617690319350906,
+      "learning_rate": 8.270210452710198e-06,
+      "loss": 0.5737,
+      "step": 2567
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0552496777118328,
+      "learning_rate": 8.268802523730827e-06,
+      "loss": 0.5568,
+      "step": 2568
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0232775617679954,
+      "learning_rate": 8.2673941419575e-06,
+      "loss": 0.5011,
+      "step": 2569
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.024146167493914,
+      "learning_rate": 8.265985307585301e-06,
+      "loss": 0.5274,
+      "step": 2570
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3118394486841143,
+      "learning_rate": 8.264576020809383e-06,
+      "loss": 0.5356,
+      "step": 2571
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.610072451524174,
+      "learning_rate": 8.26316628182496e-06,
+      "loss": 0.5092,
+      "step": 2572
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.256709912825738,
+      "learning_rate": 8.261756090827308e-06,
+      "loss": 0.5468,
+      "step": 2573
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.861063601434612,
+      "learning_rate": 8.260345448011768e-06,
+      "loss": 0.5414,
+      "step": 2574
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7829656141030537,
+      "learning_rate": 8.258934353573742e-06,
+      "loss": 0.5507,
+      "step": 2575
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.220009496041894,
+      "learning_rate": 8.257522807708693e-06,
+      "loss": 0.5474,
+      "step": 2576
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.653471236538373,
+      "learning_rate": 8.256110810612148e-06,
+      "loss": 0.5424,
+      "step": 2577
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9681653631174034,
+      "learning_rate": 8.254698362479698e-06,
+      "loss": 0.4672,
+      "step": 2578
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0595366649340368,
+      "learning_rate": 8.253285463506995e-06,
+      "loss": 0.5706,
+      "step": 2579
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.793100338503336,
+      "learning_rate": 8.251872113889754e-06,
+      "loss": 0.4998,
+      "step": 2580
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.87952322526159,
+      "learning_rate": 8.250458313823749e-06,
+      "loss": 0.4832,
+      "step": 2581
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1921515783496046,
+      "learning_rate": 8.249044063504824e-06,
+      "loss": 0.4682,
+      "step": 2582
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4465863214620356,
+      "learning_rate": 8.247629363128876e-06,
+      "loss": 0.4665,
+      "step": 2583
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.074286932433618,
+      "learning_rate": 8.24621421289187e-06,
+      "loss": 0.4985,
+      "step": 2584
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9520032705047459,
+      "learning_rate": 8.244798612989837e-06,
+      "loss": 0.5981,
+      "step": 2585
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.357868501764867,
+      "learning_rate": 8.24338256361886e-06,
+      "loss": 0.4091,
+      "step": 2586
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0763479993867717,
+      "learning_rate": 8.241966064975091e-06,
+      "loss": 0.5052,
+      "step": 2587
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.118292695663577,
+      "learning_rate": 8.240549117254746e-06,
+      "loss": 0.4825,
+      "step": 2588
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1767757588181538,
+      "learning_rate": 8.239131720654099e-06,
+      "loss": 0.4345,
+      "step": 2589
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.023881024929553,
+      "learning_rate": 8.237713875369485e-06,
+      "loss": 0.4986,
+      "step": 2590
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4511741843294326,
+      "learning_rate": 8.236295581597307e-06,
+      "loss": 0.5393,
+      "step": 2591
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8330080456413795,
+      "learning_rate": 8.234876839534025e-06,
+      "loss": 0.4641,
+      "step": 2592
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.141542332447975,
+      "learning_rate": 8.233457649376165e-06,
+      "loss": 0.5672,
+      "step": 2593
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9362241133091074,
+      "learning_rate": 8.23203801132031e-06,
+      "loss": 0.5554,
+      "step": 2594
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2754758547507206,
+      "learning_rate": 8.230617925563108e-06,
+      "loss": 0.4555,
+      "step": 2595
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.900594829201475,
+      "learning_rate": 8.229197392301274e-06,
+      "loss": 0.5462,
+      "step": 2596
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2577907696076145,
+      "learning_rate": 8.227776411731574e-06,
+      "loss": 0.479,
+      "step": 2597
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4724473080863394,
+      "learning_rate": 8.226354984050846e-06,
+      "loss": 0.5822,
+      "step": 2598
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.572981476225515,
+      "learning_rate": 8.224933109455984e-06,
+      "loss": 0.5546,
+      "step": 2599
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9099495729709979,
+      "learning_rate": 8.223510788143946e-06,
+      "loss": 0.5689,
+      "step": 2600
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4522404105463105,
+      "learning_rate": 8.222088020311753e-06,
+      "loss": 0.5573,
+      "step": 2601
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9608137780272412,
+      "learning_rate": 8.220664806156485e-06,
+      "loss": 0.4721,
+      "step": 2602
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.748922548536042,
+      "learning_rate": 8.219241145875284e-06,
+      "loss": 0.4902,
+      "step": 2603
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.515391983218839,
+      "learning_rate": 8.21781703966536e-06,
+      "loss": 0.5165,
+      "step": 2604
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8026762625518005,
+      "learning_rate": 8.216392487723974e-06,
+      "loss": 0.5162,
+      "step": 2605
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.525923698956431,
+      "learning_rate": 8.21496749024846e-06,
+      "loss": 0.4184,
+      "step": 2606
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.690732644247078,
+      "learning_rate": 8.213542047436207e-06,
+      "loss": 0.5216,
+      "step": 2607
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.932992151033144,
+      "learning_rate": 8.212116159484663e-06,
+      "loss": 0.4179,
+      "step": 2608
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8949851989688322,
+      "learning_rate": 8.210689826591348e-06,
+      "loss": 0.5248,
+      "step": 2609
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9084182291196878,
+      "learning_rate": 8.20926304895383e-06,
+      "loss": 0.4672,
+      "step": 2610
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0229182635627345,
+      "learning_rate": 8.207835826769754e-06,
+      "loss": 0.5381,
+      "step": 2611
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.368575564456848,
+      "learning_rate": 8.206408160236814e-06,
+      "loss": 0.4981,
+      "step": 2612
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.80304552071042,
+      "learning_rate": 8.204980049552771e-06,
+      "loss": 0.4464,
+      "step": 2613
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8289280162883566,
+      "learning_rate": 8.203551494915447e-06,
+      "loss": 0.4563,
+      "step": 2614
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3997514625336303,
+      "learning_rate": 8.202122496522724e-06,
+      "loss": 0.5362,
+      "step": 2615
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.915580868660194,
+      "learning_rate": 8.200693054572549e-06,
+      "loss": 0.454,
+      "step": 2616
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0165465694625697,
+      "learning_rate": 8.199263169262926e-06,
+      "loss": 0.7581,
+      "step": 2617
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5175960025874917,
+      "learning_rate": 8.197832840791921e-06,
+      "loss": 0.4169,
+      "step": 2618
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.309293361666766,
+      "learning_rate": 8.196402069357667e-06,
+      "loss": 0.5654,
+      "step": 2619
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.913954291665579,
+      "learning_rate": 8.194970855158351e-06,
+      "loss": 0.5939,
+      "step": 2620
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5134998548093526,
+      "learning_rate": 8.193539198392223e-06,
+      "loss": 0.6259,
+      "step": 2621
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.643991665914572,
+      "learning_rate": 8.192107099257604e-06,
+      "loss": 0.4816,
+      "step": 2622
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.201914171500576,
+      "learning_rate": 8.190674557952859e-06,
+      "loss": 0.5591,
+      "step": 2623
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9421403735108577,
+      "learning_rate": 8.189241574676428e-06,
+      "loss": 0.7072,
+      "step": 2624
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.026448528775754,
+      "learning_rate": 8.187808149626805e-06,
+      "loss": 0.5318,
+      "step": 2625
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1255734077205064,
+      "learning_rate": 8.18637428300255e-06,
+      "loss": 0.4483,
+      "step": 2626
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3026666589377163,
+      "learning_rate": 8.184939975002282e-06,
+      "loss": 0.5126,
+      "step": 2627
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.924866681636218,
+      "learning_rate": 8.183505225824678e-06,
+      "loss": 0.4764,
+      "step": 2628
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5455062793008416,
+      "learning_rate": 8.182070035668483e-06,
+      "loss": 0.4667,
+      "step": 2629
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1014178301836886,
+      "learning_rate": 8.180634404732499e-06,
+      "loss": 0.51,
+      "step": 2630
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6326201780747789,
+      "learning_rate": 8.179198333215588e-06,
+      "loss": 0.4979,
+      "step": 2631
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1127585387318475,
+      "learning_rate": 8.177761821316673e-06,
+      "loss": 0.4186,
+      "step": 2632
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.5667817685790095,
+      "learning_rate": 8.17632486923474e-06,
+      "loss": 0.4789,
+      "step": 2633
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2282953532535705,
+      "learning_rate": 8.174887477168838e-06,
+      "loss": 0.5056,
+      "step": 2634
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2731624794648244,
+      "learning_rate": 8.173449645318073e-06,
+      "loss": 0.4992,
+      "step": 2635
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7622816661750307,
+      "learning_rate": 8.172011373881613e-06,
+      "loss": 0.4283,
+      "step": 2636
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8124244627370563,
+      "learning_rate": 8.170572663058685e-06,
+      "loss": 0.4379,
+      "step": 2637
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.812719529052574,
+      "learning_rate": 8.169133513048581e-06,
+      "loss": 0.5041,
+      "step": 2638
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.046043517141781,
+      "learning_rate": 8.167693924050654e-06,
+      "loss": 0.5004,
+      "step": 2639
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.141722516858288,
+      "learning_rate": 8.166253896264313e-06,
+      "loss": 0.5229,
+      "step": 2640
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.31208919722211,
+      "learning_rate": 8.164813429889028e-06,
+      "loss": 0.6349,
+      "step": 2641
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.191106200937493,
+      "learning_rate": 8.163372525124337e-06,
+      "loss": 0.5277,
+      "step": 2642
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.578088089010064,
+      "learning_rate": 8.161931182169831e-06,
+      "loss": 0.5299,
+      "step": 2643
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4168045269213048,
+      "learning_rate": 8.160489401225164e-06,
+      "loss": 0.5204,
+      "step": 2644
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.136566182545421,
+      "learning_rate": 8.159047182490055e-06,
+      "loss": 0.4372,
+      "step": 2645
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3239674995998194,
+      "learning_rate": 8.157604526164277e-06,
+      "loss": 0.5406,
+      "step": 2646
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9675343438418094,
+      "learning_rate": 8.156161432447667e-06,
+      "loss": 0.5072,
+      "step": 2647
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9496801825564734,
+      "learning_rate": 8.154717901540122e-06,
+      "loss": 0.4905,
+      "step": 2648
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.828644232880727,
+      "learning_rate": 8.153273933641598e-06,
+      "loss": 0.4398,
+      "step": 2649
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1060654131164345,
+      "learning_rate": 8.151829528952116e-06,
+      "loss": 0.5236,
+      "step": 2650
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8552252104557339,
+      "learning_rate": 8.150384687671754e-06,
+      "loss": 0.7433,
+      "step": 2651
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3678985640533976,
+      "learning_rate": 8.148939410000651e-06,
+      "loss": 0.5357,
+      "step": 2652
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.509625024280866,
+      "learning_rate": 8.147493696139005e-06,
+      "loss": 0.5589,
+      "step": 2653
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.101670409898381,
+      "learning_rate": 8.146047546287077e-06,
+      "loss": 0.587,
+      "step": 2654
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9633985830700957,
+      "learning_rate": 8.144600960645188e-06,
+      "loss": 0.6321,
+      "step": 2655
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.479628384598657,
+      "learning_rate": 8.143153939413722e-06,
+      "loss": 0.4181,
+      "step": 2656
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9862100390016741,
+      "learning_rate": 8.141706482793113e-06,
+      "loss": 0.5357,
+      "step": 2657
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1669535494021783,
+      "learning_rate": 8.140258590983867e-06,
+      "loss": 0.5263,
+      "step": 2658
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2025931739623035,
+      "learning_rate": 8.138810264186547e-06,
+      "loss": 0.447,
+      "step": 2659
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6826077503564942,
+      "learning_rate": 8.137361502601771e-06,
+      "loss": 0.5164,
+      "step": 2660
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.223374492700929,
+      "learning_rate": 8.135912306430222e-06,
+      "loss": 0.4507,
+      "step": 2661
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2391538014463412,
+      "learning_rate": 8.134462675872645e-06,
+      "loss": 0.4993,
+      "step": 2662
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.166778801813421,
+      "learning_rate": 8.13301261112984e-06,
+      "loss": 0.5364,
+      "step": 2663
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8973864942428746,
+      "learning_rate": 8.131562112402673e-06,
+      "loss": 0.5187,
+      "step": 2664
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5920960687867973,
+      "learning_rate": 8.130111179892062e-06,
+      "loss": 0.4947,
+      "step": 2665
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1031256648416106,
+      "learning_rate": 8.128659813798993e-06,
+      "loss": 0.5763,
+      "step": 2666
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8341656228175178,
+      "learning_rate": 8.12720801432451e-06,
+      "loss": 0.4709,
+      "step": 2667
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3210476864035043,
+      "learning_rate": 8.125755781669713e-06,
+      "loss": 0.5819,
+      "step": 2668
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.461526355609641,
+      "learning_rate": 8.124303116035768e-06,
+      "loss": 0.5239,
+      "step": 2669
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.924909982007976,
+      "learning_rate": 8.122850017623896e-06,
+      "loss": 0.532,
+      "step": 2670
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4378221968775082,
+      "learning_rate": 8.121396486635379e-06,
+      "loss": 0.5152,
+      "step": 2671
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7209751460374247,
+      "learning_rate": 8.119942523271562e-06,
+      "loss": 0.5047,
+      "step": 2672
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1142838178598167,
+      "learning_rate": 8.118488127733848e-06,
+      "loss": 0.5039,
+      "step": 2673
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0849872068419386,
+      "learning_rate": 8.117033300223698e-06,
+      "loss": 0.4756,
+      "step": 2674
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.114156491338893,
+      "learning_rate": 8.115578040942636e-06,
+      "loss": 0.4856,
+      "step": 2675
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8830685358429977,
+      "learning_rate": 8.114122350092242e-06,
+      "loss": 0.7568,
+      "step": 2676
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0350935414465514,
+      "learning_rate": 8.11266622787416e-06,
+      "loss": 0.4157,
+      "step": 2677
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.6954817884646785,
+      "learning_rate": 8.11120967449009e-06,
+      "loss": 0.5191,
+      "step": 2678
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1680373895461678,
+      "learning_rate": 8.109752690141797e-06,
+      "loss": 0.42,
+      "step": 2679
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7632888267007614,
+      "learning_rate": 8.1082952750311e-06,
+      "loss": 0.584,
+      "step": 2680
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.592766198098536,
+      "learning_rate": 8.106837429359879e-06,
+      "loss": 0.6261,
+      "step": 2681
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.285498419495638,
+      "learning_rate": 8.105379153330075e-06,
+      "loss": 0.4616,
+      "step": 2682
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9979798897872922,
+      "learning_rate": 8.103920447143689e-06,
+      "loss": 0.488,
+      "step": 2683
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8184448715286545,
+      "learning_rate": 8.10246131100278e-06,
+      "loss": 0.5482,
+      "step": 2684
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4075334417898797,
+      "learning_rate": 8.101001745109466e-06,
+      "loss": 0.5369,
+      "step": 2685
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.998913152808339,
+      "learning_rate": 8.09954174966593e-06,
+      "loss": 0.4817,
+      "step": 2686
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7075597433061724,
+      "learning_rate": 8.098081324874407e-06,
+      "loss": 0.5357,
+      "step": 2687
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.7572738965317525,
+      "learning_rate": 8.096620470937196e-06,
+      "loss": 0.5349,
+      "step": 2688
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.975533596088515,
+      "learning_rate": 8.095159188056654e-06,
+      "loss": 0.4405,
+      "step": 2689
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.456773770118364,
+      "learning_rate": 8.093697476435196e-06,
+      "loss": 0.4993,
+      "step": 2690
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.483942110684858,
+      "learning_rate": 8.0922353362753e-06,
+      "loss": 0.5165,
+      "step": 2691
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.401080190148116,
+      "learning_rate": 8.0907727677795e-06,
+      "loss": 0.481,
+      "step": 2692
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.295217433827886,
+      "learning_rate": 8.089309771150391e-06,
+      "loss": 0.4102,
+      "step": 2693
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0742333357729574,
+      "learning_rate": 8.08784634659063e-06,
+      "loss": 0.5975,
+      "step": 2694
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9327466394488196,
+      "learning_rate": 8.086382494302927e-06,
+      "loss": 0.7196,
+      "step": 2695
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3095289021661425,
+      "learning_rate": 8.084918214490054e-06,
+      "loss": 0.4926,
+      "step": 2696
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.850572146198129,
+      "learning_rate": 8.083453507354846e-06,
+      "loss": 0.4379,
+      "step": 2697
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.993891193207234,
+      "learning_rate": 8.081988373100192e-06,
+      "loss": 0.476,
+      "step": 2698
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.812939091119548,
+      "learning_rate": 8.08052281192904e-06,
+      "loss": 0.4542,
+      "step": 2699
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3602484852291896,
+      "learning_rate": 8.079056824044405e-06,
+      "loss": 0.4792,
+      "step": 2700
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.80114180217629,
+      "learning_rate": 8.077590409649351e-06,
+      "loss": 0.4957,
+      "step": 2701
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9001716608028922,
+      "learning_rate": 8.076123568947006e-06,
+      "loss": 0.5238,
+      "step": 2702
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4884668539980117,
+      "learning_rate": 8.074656302140558e-06,
+      "loss": 0.5645,
+      "step": 2703
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1848586805360437,
+      "learning_rate": 8.07318860943325e-06,
+      "loss": 0.6117,
+      "step": 2704
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8737626556460176,
+      "learning_rate": 8.071720491028388e-06,
+      "loss": 0.4409,
+      "step": 2705
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.627295299813453,
+      "learning_rate": 8.070251947129337e-06,
+      "loss": 0.521,
+      "step": 2706
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0278307900889434,
+      "learning_rate": 8.068782977939518e-06,
+      "loss": 0.551,
+      "step": 2707
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7160545691083138,
+      "learning_rate": 8.067313583662413e-06,
+      "loss": 0.5531,
+      "step": 2708
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8424686694430443,
+      "learning_rate": 8.06584376450156e-06,
+      "loss": 0.5644,
+      "step": 2709
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.873073946817738,
+      "learning_rate": 8.06437352066056e-06,
+      "loss": 0.7099,
+      "step": 2710
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.435571342920133,
+      "learning_rate": 8.06290285234307e-06,
+      "loss": 0.5966,
+      "step": 2711
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9885806864176903,
+      "learning_rate": 8.061431759752809e-06,
+      "loss": 0.4403,
+      "step": 2712
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4188466040353287,
+      "learning_rate": 8.059960243093551e-06,
+      "loss": 0.4744,
+      "step": 2713
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0671385230118164,
+      "learning_rate": 8.05848830256913e-06,
+      "loss": 0.5413,
+      "step": 2714
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9323694346639906,
+      "learning_rate": 8.057015938383438e-06,
+      "loss": 0.4844,
+      "step": 2715
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6759499190251672,
+      "learning_rate": 8.05554315074043e-06,
+      "loss": 0.4624,
+      "step": 2716
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3345206746226324,
+      "learning_rate": 8.05406993984411e-06,
+      "loss": 0.516,
+      "step": 2717
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.546702723028085,
+      "learning_rate": 8.052596305898555e-06,
+      "loss": 0.3679,
+      "step": 2718
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0698493310403165,
+      "learning_rate": 8.051122249107885e-06,
+      "loss": 0.4803,
+      "step": 2719
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.298032294374249,
+      "learning_rate": 8.049647769676291e-06,
+      "loss": 0.4668,
+      "step": 2720
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8229886794485424,
+      "learning_rate": 8.048172867808018e-06,
+      "loss": 0.5083,
+      "step": 2721
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0294921317316788,
+      "learning_rate": 8.046697543707364e-06,
+      "loss": 0.5161,
+      "step": 2722
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0472378728028997,
+      "learning_rate": 8.045221797578698e-06,
+      "loss": 0.6513,
+      "step": 2723
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.13750463813746,
+      "learning_rate": 8.043745629626433e-06,
+      "loss": 0.5348,
+      "step": 2724
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8630121861921047,
+      "learning_rate": 8.04226904005505e-06,
+      "loss": 0.4848,
+      "step": 2725
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.133918818543244,
+      "learning_rate": 8.040792029069089e-06,
+      "loss": 0.4513,
+      "step": 2726
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.273553407620105,
+      "learning_rate": 8.039314596873141e-06,
+      "loss": 0.6088,
+      "step": 2727
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.5674563715576912,
+      "learning_rate": 8.037836743671863e-06,
+      "loss": 0.5608,
+      "step": 2728
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.661128706263857,
+      "learning_rate": 8.036358469669962e-06,
+      "loss": 0.4331,
+      "step": 2729
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7991540685105603,
+      "learning_rate": 8.034879775072215e-06,
+      "loss": 0.5399,
+      "step": 2730
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.800810400125649,
+      "learning_rate": 8.033400660083448e-06,
+      "loss": 0.4012,
+      "step": 2731
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0229030226586047,
+      "learning_rate": 8.031921124908545e-06,
+      "loss": 0.4888,
+      "step": 2732
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.467249055650737,
+      "learning_rate": 8.030441169752452e-06,
+      "loss": 0.575,
+      "step": 2733
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.248913832923308,
+      "learning_rate": 8.028960794820176e-06,
+      "loss": 0.4882,
+      "step": 2734
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.404628334176481,
+      "learning_rate": 8.027480000316773e-06,
+      "loss": 0.5549,
+      "step": 2735
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8413735537611442,
+      "learning_rate": 8.025998786447364e-06,
+      "loss": 0.5055,
+      "step": 2736
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.609735715145815,
+      "learning_rate": 8.024517153417129e-06,
+      "loss": 0.6208,
+      "step": 2737
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3130414938541075,
+      "learning_rate": 8.023035101431303e-06,
+      "loss": 0.595,
+      "step": 2738
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.400789501084477,
+      "learning_rate": 8.021552630695176e-06,
+      "loss": 0.6384,
+      "step": 2739
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1878091553452323,
+      "learning_rate": 8.020069741414103e-06,
+      "loss": 0.414,
+      "step": 2740
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.91606653270814,
+      "learning_rate": 8.018586433793492e-06,
+      "loss": 0.5733,
+      "step": 2741
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1628194132270306,
+      "learning_rate": 8.01710270803881e-06,
+      "loss": 0.5219,
+      "step": 2742
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.930665557151176,
+      "learning_rate": 8.015618564355585e-06,
+      "loss": 0.56,
+      "step": 2743
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8983264009292906,
+      "learning_rate": 8.014134002949399e-06,
+      "loss": 0.4663,
+      "step": 2744
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0660160621890356,
+      "learning_rate": 8.012649024025892e-06,
+      "loss": 0.5093,
+      "step": 2745
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.482865082761326,
+      "learning_rate": 8.011163627790765e-06,
+      "loss": 0.5785,
+      "step": 2746
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9170283553749623,
+      "learning_rate": 8.009677814449773e-06,
+      "loss": 0.535,
+      "step": 2747
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.512741740230055,
+      "learning_rate": 8.008191584208732e-06,
+      "loss": 0.6837,
+      "step": 2748
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7527201018949867,
+      "learning_rate": 8.006704937273513e-06,
+      "loss": 0.5023,
+      "step": 2749
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.665934520558981,
+      "learning_rate": 8.005217873850048e-06,
+      "loss": 0.4276,
+      "step": 2750
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.00337124871565,
+      "learning_rate": 8.003730394144322e-06,
+      "loss": 0.4756,
+      "step": 2751
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1055865720560973,
+      "learning_rate": 8.002242498362384e-06,
+      "loss": 0.631,
+      "step": 2752
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3361706375862212,
+      "learning_rate": 8.000754186710333e-06,
+      "loss": 0.508,
+      "step": 2753
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2415534899797693,
+      "learning_rate": 7.999265459394334e-06,
+      "loss": 0.5675,
+      "step": 2754
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.9876418285797244,
+      "learning_rate": 7.997776316620603e-06,
+      "loss": 0.5199,
+      "step": 2755
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1292822971195653,
+      "learning_rate": 7.996286758595413e-06,
+      "loss": 0.5067,
+      "step": 2756
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9961904909637553,
+      "learning_rate": 7.994796785525103e-06,
+      "loss": 0.5629,
+      "step": 2757
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5422441621330054,
+      "learning_rate": 7.993306397616061e-06,
+      "loss": 0.5239,
+      "step": 2758
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.339139495398308,
+      "learning_rate": 7.991815595074733e-06,
+      "loss": 0.5247,
+      "step": 2759
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9217820361181952,
+      "learning_rate": 7.990324378107628e-06,
+      "loss": 0.5198,
+      "step": 2760
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.34673339581103,
+      "learning_rate": 7.98883274692131e-06,
+      "loss": 0.5867,
+      "step": 2761
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.25636343561003,
+      "learning_rate": 7.987340701722395e-06,
+      "loss": 0.6272,
+      "step": 2762
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6447224408726435,
+      "learning_rate": 7.985848242717564e-06,
+      "loss": 0.5116,
+      "step": 2763
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.779417522671597,
+      "learning_rate": 7.984355370113553e-06,
+      "loss": 0.5527,
+      "step": 2764
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9295455417084404,
+      "learning_rate": 7.982862084117152e-06,
+      "loss": 0.7216,
+      "step": 2765
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.928475627983069,
+      "learning_rate": 7.98136838493521e-06,
+      "loss": 0.5175,
+      "step": 2766
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.054350552554032,
+      "learning_rate": 7.97987427277464e-06,
+      "loss": 0.4977,
+      "step": 2767
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3071197434238906,
+      "learning_rate": 7.978379747842398e-06,
+      "loss": 0.552,
+      "step": 2768
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8658521399209966,
+      "learning_rate": 7.97688481034551e-06,
+      "loss": 0.7213,
+      "step": 2769
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5858970200435887,
+      "learning_rate": 7.975389460491054e-06,
+      "loss": 0.4605,
+      "step": 2770
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8410170875727503,
+      "learning_rate": 7.973893698486166e-06,
+      "loss": 0.5975,
+      "step": 2771
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.281415110380626,
+      "learning_rate": 7.972397524538036e-06,
+      "loss": 0.5134,
+      "step": 2772
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1662379473963287,
+      "learning_rate": 7.970900938853918e-06,
+      "loss": 0.5741,
+      "step": 2773
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.489914144369624,
+      "learning_rate": 7.969403941641117e-06,
+      "loss": 0.5369,
+      "step": 2774
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9439346769260484,
+      "learning_rate": 7.967906533106994e-06,
+      "loss": 0.5503,
+      "step": 2775
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.228612661596518,
+      "learning_rate": 7.966408713458973e-06,
+      "loss": 0.5183,
+      "step": 2776
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9583391601775046,
+      "learning_rate": 7.964910482904532e-06,
+      "loss": 0.5577,
+      "step": 2777
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.172710734632264,
+      "learning_rate": 7.963411841651202e-06,
+      "loss": 0.5274,
+      "step": 2778
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8345351998090547,
+      "learning_rate": 7.961912789906579e-06,
+      "loss": 0.4859,
+      "step": 2779
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6811891351617008,
+      "learning_rate": 7.960413327878309e-06,
+      "loss": 0.5568,
+      "step": 2780
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0586414514304536,
+      "learning_rate": 7.958913455774097e-06,
+      "loss": 0.5107,
+      "step": 2781
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.476799113416005,
+      "learning_rate": 7.957413173801706e-06,
+      "loss": 0.484,
+      "step": 2782
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.029924826038305,
+      "learning_rate": 7.955912482168956e-06,
+      "loss": 0.5333,
+      "step": 2783
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2184977206129366,
+      "learning_rate": 7.954411381083717e-06,
+      "loss": 0.5,
+      "step": 2784
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.075023537607446,
+      "learning_rate": 7.952909870753928e-06,
+      "loss": 0.4747,
+      "step": 2785
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9026421317046905,
+      "learning_rate": 7.951407951387575e-06,
+      "loss": 0.6914,
+      "step": 2786
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.754194531422513,
+      "learning_rate": 7.949905623192702e-06,
+      "loss": 0.5321,
+      "step": 2787
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.870129097248555,
+      "learning_rate": 7.948402886377415e-06,
+      "loss": 0.5972,
+      "step": 2788
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4313102873153656,
+      "learning_rate": 7.94689974114987e-06,
+      "loss": 0.496,
+      "step": 2789
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2216135299142925,
+      "learning_rate": 7.945396187718284e-06,
+      "loss": 0.6397,
+      "step": 2790
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7887964692358387,
+      "learning_rate": 7.943892226290929e-06,
+      "loss": 0.5073,
+      "step": 2791
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4775370164035797,
+      "learning_rate": 7.94238785707613e-06,
+      "loss": 0.5477,
+      "step": 2792
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3841762780066196,
+      "learning_rate": 7.940883080282276e-06,
+      "loss": 0.5021,
+      "step": 2793
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.840327680179315,
+      "learning_rate": 7.939377896117808e-06,
+      "loss": 0.6436,
+      "step": 2794
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.86383657667886,
+      "learning_rate": 7.937872304791222e-06,
+      "loss": 0.5248,
+      "step": 2795
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5563136517955662,
+      "learning_rate": 7.936366306511074e-06,
+      "loss": 0.4609,
+      "step": 2796
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.223963798314906,
+      "learning_rate": 7.934859901485973e-06,
+      "loss": 0.5053,
+      "step": 2797
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9513097465351306,
+      "learning_rate": 7.933353089924586e-06,
+      "loss": 0.4631,
+      "step": 2798
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7381478779910948,
+      "learning_rate": 7.93184587203564e-06,
+      "loss": 0.4679,
+      "step": 2799
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.306291311604797,
+      "learning_rate": 7.93033824802791e-06,
+      "loss": 0.4704,
+      "step": 2800
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0992274939116955,
+      "learning_rate": 7.928830218110233e-06,
+      "loss": 0.5107,
+      "step": 2801
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5565434477154236,
+      "learning_rate": 7.9273217824915e-06,
+      "loss": 0.4521,
+      "step": 2802
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5943522873238118,
+      "learning_rate": 7.925812941380663e-06,
+      "loss": 0.4312,
+      "step": 2803
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6038082595038707,
+      "learning_rate": 7.924303694986723e-06,
+      "loss": 0.5406,
+      "step": 2804
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.067468817240631,
+      "learning_rate": 7.922794043518742e-06,
+      "loss": 0.578,
+      "step": 2805
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.893179960484438,
+      "learning_rate": 7.921283987185836e-06,
+      "loss": 0.5362,
+      "step": 2806
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8686549080765698,
+      "learning_rate": 7.919773526197178e-06,
+      "loss": 0.4546,
+      "step": 2807
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2669060659135227,
+      "learning_rate": 7.918262660761999e-06,
+      "loss": 0.5553,
+      "step": 2808
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7914401011706638,
+      "learning_rate": 7.916751391089579e-06,
+      "loss": 0.4995,
+      "step": 2809
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3464973615101257,
+      "learning_rate": 7.915239717389264e-06,
+      "loss": 0.495,
+      "step": 2810
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3686119086956654,
+      "learning_rate": 7.913727639870446e-06,
+      "loss": 0.5977,
+      "step": 2811
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2303095388299963,
+      "learning_rate": 7.912215158742581e-06,
+      "loss": 0.5549,
+      "step": 2812
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.197381227343946,
+      "learning_rate": 7.910702274215176e-06,
+      "loss": 0.5626,
+      "step": 2813
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.567462243597013,
+      "learning_rate": 7.909188986497797e-06,
+      "loss": 0.526,
+      "step": 2814
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9303974175298675,
+      "learning_rate": 7.907675295800062e-06,
+      "loss": 0.584,
+      "step": 2815
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.745199598342414,
+      "learning_rate": 7.906161202331652e-06,
+      "loss": 0.4787,
+      "step": 2816
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8187286103340488,
+      "learning_rate": 7.904646706302292e-06,
+      "loss": 0.5737,
+      "step": 2817
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6503837972845514,
+      "learning_rate": 7.903131807921776e-06,
+      "loss": 0.4773,
+      "step": 2818
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.014884374213472,
+      "learning_rate": 7.901616507399943e-06,
+      "loss": 0.4838,
+      "step": 2819
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3014735304666822,
+      "learning_rate": 7.900100804946695e-06,
+      "loss": 0.5783,
+      "step": 2820
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9541029595367323,
+      "learning_rate": 7.898584700771984e-06,
+      "loss": 0.5126,
+      "step": 2821
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.7922591750437857,
+      "learning_rate": 7.897068195085825e-06,
+      "loss": 0.508,
+      "step": 2822
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.333529609122084,
+      "learning_rate": 7.895551288098278e-06,
+      "loss": 0.5425,
+      "step": 2823
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6769325083547368,
+      "learning_rate": 7.894033980019471e-06,
+      "loss": 0.4845,
+      "step": 2824
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.83972346676095,
+      "learning_rate": 7.892516271059577e-06,
+      "loss": 0.5696,
+      "step": 2825
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7205715956722716,
+      "learning_rate": 7.89099816142883e-06,
+      "loss": 0.4592,
+      "step": 2826
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7138444200888199,
+      "learning_rate": 7.88947965133752e-06,
+      "loss": 0.4277,
+      "step": 2827
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3180624398108725,
+      "learning_rate": 7.887960740995988e-06,
+      "loss": 0.6045,
+      "step": 2828
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3163712164472154,
+      "learning_rate": 7.886441430614635e-06,
+      "loss": 0.55,
+      "step": 2829
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.650152235286882,
+      "learning_rate": 7.884921720403914e-06,
+      "loss": 0.4923,
+      "step": 2830
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3412441826971344,
+      "learning_rate": 7.883401610574338e-06,
+      "loss": 0.5816,
+      "step": 2831
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0717124813637398,
+      "learning_rate": 7.881881101336467e-06,
+      "loss": 0.5606,
+      "step": 2832
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8809679457647487,
+      "learning_rate": 7.880360192900928e-06,
+      "loss": 0.735,
+      "step": 2833
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.795443244861914,
+      "learning_rate": 7.878838885478393e-06,
+      "loss": 0.4835,
+      "step": 2834
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0785830319635688,
+      "learning_rate": 7.877317179279593e-06,
+      "loss": 0.5664,
+      "step": 2835
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6909259881174155,
+      "learning_rate": 7.875795074515316e-06,
+      "loss": 0.3613,
+      "step": 2836
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.040848180384166,
+      "learning_rate": 7.874272571396404e-06,
+      "loss": 0.446,
+      "step": 2837
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.147286690884782,
+      "learning_rate": 7.872749670133754e-06,
+      "loss": 0.5337,
+      "step": 2838
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.9512494779990552,
+      "learning_rate": 7.871226370938316e-06,
+      "loss": 0.516,
+      "step": 2839
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7564960507728684,
+      "learning_rate": 7.869702674021098e-06,
+      "loss": 0.4278,
+      "step": 2840
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0586388926302437,
+      "learning_rate": 7.868178579593165e-06,
+      "loss": 0.4846,
+      "step": 2841
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0518253062349205,
+      "learning_rate": 7.86665408786563e-06,
+      "loss": 0.6266,
+      "step": 2842
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9195303083858667,
+      "learning_rate": 7.865129199049667e-06,
+      "loss": 0.7107,
+      "step": 2843
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2937679083383267,
+      "learning_rate": 7.863603913356505e-06,
+      "loss": 0.6104,
+      "step": 2844
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6653358538520346,
+      "learning_rate": 7.862078230997425e-06,
+      "loss": 0.4646,
+      "step": 2845
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9752400262255085,
+      "learning_rate": 7.860552152183763e-06,
+      "loss": 0.5434,
+      "step": 2846
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.120938523981115,
+      "learning_rate": 7.859025677126914e-06,
+      "loss": 0.4834,
+      "step": 2847
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8307417329711975,
+      "learning_rate": 7.857498806038321e-06,
+      "loss": 0.5691,
+      "step": 2848
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8207558744223356,
+      "learning_rate": 7.85597153912949e-06,
+      "loss": 0.5926,
+      "step": 2849
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.496267171099766,
+      "learning_rate": 7.854443876611976e-06,
+      "loss": 0.5429,
+      "step": 2850
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.606526518298688,
+      "learning_rate": 7.852915818697391e-06,
+      "loss": 0.585,
+      "step": 2851
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.222585073046454,
+      "learning_rate": 7.851387365597401e-06,
+      "loss": 0.7776,
+      "step": 2852
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.490542455946593,
+      "learning_rate": 7.849858517523725e-06,
+      "loss": 0.6018,
+      "step": 2853
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.963991175454294,
+      "learning_rate": 7.848329274688143e-06,
+      "loss": 0.5209,
+      "step": 2854
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.04660360536957,
+      "learning_rate": 7.84679963730248e-06,
+      "loss": 0.4392,
+      "step": 2855
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9347306604899075,
+      "learning_rate": 7.845269605578628e-06,
+      "loss": 0.446,
+      "step": 2856
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8521769781341793,
+      "learning_rate": 7.84373917972852e-06,
+      "loss": 0.4683,
+      "step": 2857
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.105793466668921,
+      "learning_rate": 7.842208359964157e-06,
+      "loss": 0.5612,
+      "step": 2858
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8853773851640279,
+      "learning_rate": 7.840677146497582e-06,
+      "loss": 0.4859,
+      "step": 2859
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.983068486139454,
+      "learning_rate": 7.8391455395409e-06,
+      "loss": 0.6218,
+      "step": 2860
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6540258285639426,
+      "learning_rate": 7.83761353930627e-06,
+      "loss": 0.4793,
+      "step": 2861
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0413710881394036,
+      "learning_rate": 7.836081146005906e-06,
+      "loss": 0.5717,
+      "step": 2862
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2535283194454703,
+      "learning_rate": 7.83454835985207e-06,
+      "loss": 0.4258,
+      "step": 2863
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.51213697458294,
+      "learning_rate": 7.833015181057088e-06,
+      "loss": 0.5227,
+      "step": 2864
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9837458860826667,
+      "learning_rate": 7.831481609833333e-06,
+      "loss": 0.7266,
+      "step": 2865
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5202573275118496,
+      "learning_rate": 7.829947646393237e-06,
+      "loss": 0.6385,
+      "step": 2866
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4495264437524926,
+      "learning_rate": 7.828413290949282e-06,
+      "loss": 0.4387,
+      "step": 2867
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.936619109920517,
+      "learning_rate": 7.826878543714007e-06,
+      "loss": 0.6082,
+      "step": 2868
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9382337242367331,
+      "learning_rate": 7.825343404900008e-06,
+      "loss": 0.5534,
+      "step": 2869
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4971381755755377,
+      "learning_rate": 7.823807874719929e-06,
+      "loss": 0.4565,
+      "step": 2870
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1234957063362,
+      "learning_rate": 7.82227195338647e-06,
+      "loss": 0.5726,
+      "step": 2871
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.67202790166371,
+      "learning_rate": 7.820735641112394e-06,
+      "loss": 0.4659,
+      "step": 2872
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.066033714524526,
+      "learning_rate": 7.819198938110501e-06,
+      "loss": 0.5954,
+      "step": 2873
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.489886034718662,
+      "learning_rate": 7.817661844593661e-06,
+      "loss": 0.5677,
+      "step": 2874
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6799771980626135,
+      "learning_rate": 7.816124360774792e-06,
+      "loss": 0.5234,
+      "step": 2875
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.007174919027362,
+      "learning_rate": 7.814586486866862e-06,
+      "loss": 0.4791,
+      "step": 2876
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0221138993847907,
+      "learning_rate": 7.8130482230829e-06,
+      "loss": 0.5348,
+      "step": 2877
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0484933471639537,
+      "learning_rate": 7.811509569635984e-06,
+      "loss": 0.5348,
+      "step": 2878
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.3966521239363003,
+      "learning_rate": 7.80997052673925e-06,
+      "loss": 0.4443,
+      "step": 2879
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.301959768027492,
+      "learning_rate": 7.808431094605887e-06,
+      "loss": 0.5895,
+      "step": 2880
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8765655337998974,
+      "learning_rate": 7.806891273449134e-06,
+      "loss": 0.4786,
+      "step": 2881
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1533827667974417,
+      "learning_rate": 7.80535106348229e-06,
+      "loss": 0.5444,
+      "step": 2882
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4780372306733187,
+      "learning_rate": 7.803810464918699e-06,
+      "loss": 0.5539,
+      "step": 2883
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.053174106982524,
+      "learning_rate": 7.802269477971771e-06,
+      "loss": 0.4593,
+      "step": 2884
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.7452782843645367,
+      "learning_rate": 7.80072810285496e-06,
+      "loss": 0.5226,
+      "step": 2885
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1701497276994215,
+      "learning_rate": 7.799186339781774e-06,
+      "loss": 0.5726,
+      "step": 2886
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9765197897320606,
+      "learning_rate": 7.797644188965785e-06,
+      "loss": 0.5338,
+      "step": 2887
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.165322483903886,
+      "learning_rate": 7.796101650620605e-06,
+      "loss": 0.4888,
+      "step": 2888
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7802218277226602,
+      "learning_rate": 7.79455872495991e-06,
+      "loss": 0.4659,
+      "step": 2889
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.078627091424822,
+      "learning_rate": 7.793015412197424e-06,
+      "loss": 0.5878,
+      "step": 2890
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1987500174222228,
+      "learning_rate": 7.791471712546928e-06,
+      "loss": 0.3887,
+      "step": 2891
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2349132037906094,
+      "learning_rate": 7.789927626222253e-06,
+      "loss": 0.5394,
+      "step": 2892
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9491747347423336,
+      "learning_rate": 7.788383153437286e-06,
+      "loss": 0.5352,
+      "step": 2893
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1520661463093,
+      "learning_rate": 7.786838294405968e-06,
+      "loss": 0.6683,
+      "step": 2894
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0177019206945705,
+      "learning_rate": 7.78529304934229e-06,
+      "loss": 0.5486,
+      "step": 2895
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8733395828896169,
+      "learning_rate": 7.783747418460305e-06,
+      "loss": 0.5539,
+      "step": 2896
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.818741066646174,
+      "learning_rate": 7.782201401974107e-06,
+      "loss": 0.4972,
+      "step": 2897
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.169323240902594,
+      "learning_rate": 7.780655000097854e-06,
+      "loss": 0.561,
+      "step": 2898
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.481132685320535,
+      "learning_rate": 7.779108213045752e-06,
+      "loss": 0.5459,
+      "step": 2899
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.483788380927395,
+      "learning_rate": 7.777561041032061e-06,
+      "loss": 0.5005,
+      "step": 2900
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4194458517288875,
+      "learning_rate": 7.776013484271096e-06,
+      "loss": 0.4974,
+      "step": 2901
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8954373011375965,
+      "learning_rate": 7.774465542977224e-06,
+      "loss": 0.5491,
+      "step": 2902
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1365244094847817,
+      "learning_rate": 7.772917217364866e-06,
+      "loss": 0.4709,
+      "step": 2903
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7232532986010074,
+      "learning_rate": 7.771368507648494e-06,
+      "loss": 0.5166,
+      "step": 2904
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.049455832821686,
+      "learning_rate": 7.769819414042639e-06,
+      "loss": 0.5499,
+      "step": 2905
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.012171878517844,
+      "learning_rate": 7.768269936761875e-06,
+      "loss": 0.5468,
+      "step": 2906
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.474627658858279,
+      "learning_rate": 7.76672007602084e-06,
+      "loss": 0.5115,
+      "step": 2907
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2474340456451163,
+      "learning_rate": 7.76516983203422e-06,
+      "loss": 0.4763,
+      "step": 2908
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.98043557433873,
+      "learning_rate": 7.763619205016754e-06,
+      "loss": 0.5194,
+      "step": 2909
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.228230179586194,
+      "learning_rate": 7.762068195183234e-06,
+      "loss": 0.6098,
+      "step": 2910
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.860624149234416,
+      "learning_rate": 7.760516802748506e-06,
+      "loss": 0.6057,
+      "step": 2911
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.132851809397818,
+      "learning_rate": 7.75896502792747e-06,
+      "loss": 0.5063,
+      "step": 2912
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.044307510309009,
+      "learning_rate": 7.757412870935074e-06,
+      "loss": 0.5451,
+      "step": 2913
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.639350364709035,
+      "learning_rate": 7.755860331986326e-06,
+      "loss": 0.4569,
+      "step": 2914
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2857697468157556,
+      "learning_rate": 7.75430741129628e-06,
+      "loss": 0.5052,
+      "step": 2915
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9947975274243404,
+      "learning_rate": 7.752754109080051e-06,
+      "loss": 0.5661,
+      "step": 2916
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7206767443045559,
+      "learning_rate": 7.751200425552801e-06,
+      "loss": 0.5384,
+      "step": 2917
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8260874410773151,
+      "learning_rate": 7.749646360929741e-06,
+      "loss": 0.5626,
+      "step": 2918
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.435843437185004,
+      "learning_rate": 7.748091915426145e-06,
+      "loss": 0.5373,
+      "step": 2919
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.061451427930398,
+      "learning_rate": 7.746537089257332e-06,
+      "loss": 0.5114,
+      "step": 2920
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5931050831720555,
+      "learning_rate": 7.744981882638678e-06,
+      "loss": 0.4758,
+      "step": 2921
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3228744853219463,
+      "learning_rate": 7.743426295785608e-06,
+      "loss": 0.4034,
+      "step": 2922
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1897378820783415,
+      "learning_rate": 7.741870328913602e-06,
+      "loss": 0.5802,
+      "step": 2923
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7166339781316216,
+      "learning_rate": 7.740313982238196e-06,
+      "loss": 0.5451,
+      "step": 2924
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4997109601899017,
+      "learning_rate": 7.73875725597497e-06,
+      "loss": 0.49,
+      "step": 2925
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1203236791245705,
+      "learning_rate": 7.737200150339564e-06,
+      "loss": 0.7592,
+      "step": 2926
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.742797129883485,
+      "learning_rate": 7.735642665547667e-06,
+      "loss": 0.3658,
+      "step": 2927
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3012555458803137,
+      "learning_rate": 7.734084801815022e-06,
+      "loss": 0.6232,
+      "step": 2928
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4691819321460167,
+      "learning_rate": 7.732526559357423e-06,
+      "loss": 0.561,
+      "step": 2929
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.611099599992232,
+      "learning_rate": 7.730967938390718e-06,
+      "loss": 0.4972,
+      "step": 2930
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.72783732716072,
+      "learning_rate": 7.729408939130809e-06,
+      "loss": 0.4314,
+      "step": 2931
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8368836825789712,
+      "learning_rate": 7.727849561793643e-06,
+      "loss": 0.4891,
+      "step": 2932
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8445229386917201,
+      "learning_rate": 7.726289806595231e-06,
+      "loss": 0.5145,
+      "step": 2933
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.128903757136383,
+      "learning_rate": 7.724729673751628e-06,
+      "loss": 0.4367,
+      "step": 2934
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5988610619431016,
+      "learning_rate": 7.72316916347894e-06,
+      "loss": 0.5272,
+      "step": 2935
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3602582964595364,
+      "learning_rate": 7.721608275993334e-06,
+      "loss": 0.5232,
+      "step": 2936
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.167494589365408,
+      "learning_rate": 7.720047011511018e-06,
+      "loss": 0.5174,
+      "step": 2937
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.497134122359307,
+      "learning_rate": 7.718485370248264e-06,
+      "loss": 0.5033,
+      "step": 2938
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9261206907560326,
+      "learning_rate": 7.716923352421385e-06,
+      "loss": 0.4349,
+      "step": 2939
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.383231649251385,
+      "learning_rate": 7.715360958246753e-06,
+      "loss": 0.4846,
+      "step": 2940
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.890924212846894,
+      "learning_rate": 7.713798187940794e-06,
+      "loss": 0.4957,
+      "step": 2941
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8517308708455993,
+      "learning_rate": 7.712235041719979e-06,
+      "loss": 0.5856,
+      "step": 2942
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3859422380751445,
+      "learning_rate": 7.710671519800836e-06,
+      "loss": 0.545,
+      "step": 2943
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6501928972180138,
+      "learning_rate": 7.709107622399945e-06,
+      "loss": 0.5267,
+      "step": 2944
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.9882644263317366,
+      "learning_rate": 7.707543349733932e-06,
+      "loss": 0.4831,
+      "step": 2945
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8356616527512446,
+      "learning_rate": 7.705978702019486e-06,
+      "loss": 0.5728,
+      "step": 2946
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.791547352696187,
+      "learning_rate": 7.704413679473338e-06,
+      "loss": 0.4833,
+      "step": 2947
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0565533923255033,
+      "learning_rate": 7.702848282312275e-06,
+      "loss": 0.4865,
+      "step": 2948
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.926220838080352,
+      "learning_rate": 7.701282510753137e-06,
+      "loss": 0.4934,
+      "step": 2949
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.158863602545339,
+      "learning_rate": 7.699716365012813e-06,
+      "loss": 0.5122,
+      "step": 2950
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7349318596712946,
+      "learning_rate": 7.698149845308245e-06,
+      "loss": 0.4975,
+      "step": 2951
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0703992921889913,
+      "learning_rate": 7.696582951856428e-06,
+      "loss": 0.4926,
+      "step": 2952
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9216645456245265,
+      "learning_rate": 7.69501568487441e-06,
+      "loss": 0.536,
+      "step": 2953
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3466368473922428,
+      "learning_rate": 7.693448044579284e-06,
+      "loss": 0.5794,
+      "step": 2954
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.630253512621041,
+      "learning_rate": 7.6918800311882e-06,
+      "loss": 0.5676,
+      "step": 2955
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8924825897882203,
+      "learning_rate": 7.690311644918362e-06,
+      "loss": 0.4483,
+      "step": 2956
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.322509102806364,
+      "learning_rate": 7.68874288598702e-06,
+      "loss": 0.4541,
+      "step": 2957
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4574223382848848,
+      "learning_rate": 7.687173754611481e-06,
+      "loss": 0.4947,
+      "step": 2958
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.5607622510857704,
+      "learning_rate": 7.685604251009097e-06,
+      "loss": 0.5787,
+      "step": 2959
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2096089635268163,
+      "learning_rate": 7.684034375397278e-06,
+      "loss": 0.4584,
+      "step": 2960
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2804338401435427,
+      "learning_rate": 7.682464127993483e-06,
+      "loss": 0.6141,
+      "step": 2961
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1756159314044514,
+      "learning_rate": 7.68089350901522e-06,
+      "loss": 0.47,
+      "step": 2962
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.287938095293658,
+      "learning_rate": 7.679322518680054e-06,
+      "loss": 0.5433,
+      "step": 2963
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.929173897204981,
+      "learning_rate": 7.677751157205597e-06,
+      "loss": 0.5351,
+      "step": 2964
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9663847241081105,
+      "learning_rate": 7.676179424809512e-06,
+      "loss": 0.3873,
+      "step": 2965
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9227217638591527,
+      "learning_rate": 7.674607321709517e-06,
+      "loss": 0.5108,
+      "step": 2966
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.243532509885131,
+      "learning_rate": 7.673034848123379e-06,
+      "loss": 0.5614,
+      "step": 2967
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7371399234676472,
+      "learning_rate": 7.67146200426892e-06,
+      "loss": 0.5275,
+      "step": 2968
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9269322442188325,
+      "learning_rate": 7.669888790364002e-06,
+      "loss": 0.7532,
+      "step": 2969
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8511701629478972,
+      "learning_rate": 7.668315206626554e-06,
+      "loss": 0.5932,
+      "step": 2970
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.064551487376635,
+      "learning_rate": 7.666741253274545e-06,
+      "loss": 0.5058,
+      "step": 2971
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.874255400463469,
+      "learning_rate": 7.665166930525999e-06,
+      "loss": 0.7496,
+      "step": 2972
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.200789282396879,
+      "learning_rate": 7.663592238598992e-06,
+      "loss": 0.6205,
+      "step": 2973
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0999519128644257,
+      "learning_rate": 7.66201717771165e-06,
+      "loss": 0.6294,
+      "step": 2974
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.541270548818184,
+      "learning_rate": 7.660441748082148e-06,
+      "loss": 0.6071,
+      "step": 2975
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.790211970550135,
+      "learning_rate": 7.658865949928717e-06,
+      "loss": 0.5314,
+      "step": 2976
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.123254001834856,
+      "learning_rate": 7.657289783469637e-06,
+      "loss": 0.4634,
+      "step": 2977
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5908427041928235,
+      "learning_rate": 7.655713248923233e-06,
+      "loss": 0.4845,
+      "step": 2978
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9049266356984043,
+      "learning_rate": 7.654136346507892e-06,
+      "loss": 0.7384,
+      "step": 2979
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6002503760260605,
+      "learning_rate": 7.652559076442043e-06,
+      "loss": 0.5255,
+      "step": 2980
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.090864433519095,
+      "learning_rate": 7.650981438944169e-06,
+      "loss": 0.4484,
+      "step": 2981
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2949779244741935,
+      "learning_rate": 7.649403434232807e-06,
+      "loss": 0.5032,
+      "step": 2982
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0295852578748144,
+      "learning_rate": 7.647825062526539e-06,
+      "loss": 0.5644,
+      "step": 2983
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8570147353674065,
+      "learning_rate": 7.646246324044005e-06,
+      "loss": 0.48,
+      "step": 2984
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2901509684985295,
+      "learning_rate": 7.644667219003885e-06,
+      "loss": 0.4782,
+      "step": 2985
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.7965285771962836,
+      "learning_rate": 7.643087747624923e-06,
+      "loss": 0.5782,
+      "step": 2986
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9451023823398457,
+      "learning_rate": 7.641507910125901e-06,
+      "loss": 0.4883,
+      "step": 2987
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0068245925838193,
+      "learning_rate": 7.639927706725661e-06,
+      "loss": 0.4565,
+      "step": 2988
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8587125057933247,
+      "learning_rate": 7.638347137643094e-06,
+      "loss": 0.5599,
+      "step": 2989
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0719682769291112,
+      "learning_rate": 7.636766203097137e-06,
+      "loss": 0.5059,
+      "step": 2990
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8915907712243234,
+      "learning_rate": 7.635184903306783e-06,
+      "loss": 0.6083,
+      "step": 2991
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.461162270295933,
+      "learning_rate": 7.633603238491072e-06,
+      "loss": 0.4568,
+      "step": 2992
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9377791600861256,
+      "learning_rate": 7.632021208869099e-06,
+      "loss": 0.5317,
+      "step": 2993
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1755013673172137,
+      "learning_rate": 7.630438814660002e-06,
+      "loss": 0.5211,
+      "step": 2994
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.041833325931362,
+      "learning_rate": 7.628856056082976e-06,
+      "loss": 0.5576,
+      "step": 2995
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8244062603031144,
+      "learning_rate": 7.6272729333572656e-06,
+      "loss": 0.4423,
+      "step": 2996
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9113523393591565,
+      "learning_rate": 7.625689446702162e-06,
+      "loss": 0.5588,
+      "step": 2997
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7932671127752189,
+      "learning_rate": 7.6241055963370115e-06,
+      "loss": 0.4953,
+      "step": 2998
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.084520604488172,
+      "learning_rate": 7.622521382481208e-06,
+      "loss": 0.5054,
+      "step": 2999
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.585848774893426,
+      "learning_rate": 7.620936805354198e-06,
+      "loss": 0.5723,
+      "step": 3000
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0960427361570337,
+      "learning_rate": 7.619351865175475e-06,
+      "loss": 0.6097,
+      "step": 3001
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8099090711212205,
+      "learning_rate": 7.617766562164586e-06,
+      "loss": 0.504,
+      "step": 3002
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3004536172358447,
+      "learning_rate": 7.6161808965411255e-06,
+      "loss": 0.5545,
+      "step": 3003
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9187119060330333,
+      "learning_rate": 7.61459486852474e-06,
+      "loss": 0.7088,
+      "step": 3004
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1926570913718986,
+      "learning_rate": 7.6130084783351255e-06,
+      "loss": 0.5005,
+      "step": 3005
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 26.3348170685861,
+      "learning_rate": 7.61142172619203e-06,
+      "loss": 0.5001,
+      "step": 3006
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8228877120025713,
+      "learning_rate": 7.60983461231525e-06,
+      "loss": 0.5537,
+      "step": 3007
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.934026830395377,
+      "learning_rate": 7.608247136924631e-06,
+      "loss": 0.5071,
+      "step": 3008
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.428847184622086,
+      "learning_rate": 7.606659300240069e-06,
+      "loss": 0.6156,
+      "step": 3009
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9415161648949937,
+      "learning_rate": 7.605071102481515e-06,
+      "loss": 0.5086,
+      "step": 3010
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7949844429470543,
+      "learning_rate": 7.603482543868961e-06,
+      "loss": 0.4972,
+      "step": 3011
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4293649844133642,
+      "learning_rate": 7.601893624622457e-06,
+      "loss": 0.58,
+      "step": 3012
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5984728768570198,
+      "learning_rate": 7.6003043449620985e-06,
+      "loss": 0.5248,
+      "step": 3013
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4567970827934973,
+      "learning_rate": 7.598714705108032e-06,
+      "loss": 0.4393,
+      "step": 3014
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4661073780936915,
+      "learning_rate": 7.597124705280457e-06,
+      "loss": 0.4549,
+      "step": 3015
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1431633035870603,
+      "learning_rate": 7.595534345699614e-06,
+      "loss": 0.5268,
+      "step": 3016
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9115151673070327,
+      "learning_rate": 7.5939436265858045e-06,
+      "loss": 0.4838,
+      "step": 3017
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 17.118261576765867,
+      "learning_rate": 7.592352548159374e-06,
+      "loss": 0.4676,
+      "step": 3018
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8395927292192455,
+      "learning_rate": 7.590761110640718e-06,
+      "loss": 0.5961,
+      "step": 3019
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7868876702102325,
+      "learning_rate": 7.58916931425028e-06,
+      "loss": 0.5947,
+      "step": 3020
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9374642558959907,
+      "learning_rate": 7.587577159208558e-06,
+      "loss": 0.4821,
+      "step": 3021
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9117629781844931,
+      "learning_rate": 7.585984645736095e-06,
+      "loss": 0.4922,
+      "step": 3022
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8877449361570138,
+      "learning_rate": 7.584391774053488e-06,
+      "loss": 0.554,
+      "step": 3023
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8349479844942105,
+      "learning_rate": 7.5827985443813786e-06,
+      "loss": 0.4578,
+      "step": 3024
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.816384865181815,
+      "learning_rate": 7.581204956940463e-06,
+      "loss": 0.4929,
+      "step": 3025
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.726748230334919,
+      "learning_rate": 7.5796110119514844e-06,
+      "loss": 0.4977,
+      "step": 3026
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.517570093908069,
+      "learning_rate": 7.578016709635236e-06,
+      "loss": 0.465,
+      "step": 3027
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.013953462256144,
+      "learning_rate": 7.576422050212557e-06,
+      "loss": 0.4812,
+      "step": 3028
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2966066388658475,
+      "learning_rate": 7.574827033904344e-06,
+      "loss": 0.5277,
+      "step": 3029
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8432401384048425,
+      "learning_rate": 7.5732316609315345e-06,
+      "loss": 0.5696,
+      "step": 3030
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4051250639458654,
+      "learning_rate": 7.571635931515122e-06,
+      "loss": 0.5518,
+      "step": 3031
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6452145479989446,
+      "learning_rate": 7.570039845876146e-06,
+      "loss": 0.4789,
+      "step": 3032
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5612735395296857,
+      "learning_rate": 7.568443404235694e-06,
+      "loss": 0.5883,
+      "step": 3033
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.828912842747167,
+      "learning_rate": 7.566846606814909e-06,
+      "loss": 0.5538,
+      "step": 3034
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7412551767514126,
+      "learning_rate": 7.565249453834976e-06,
+      "loss": 0.5182,
+      "step": 3035
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.080776631449496,
+      "learning_rate": 7.563651945517132e-06,
+      "loss": 0.4987,
+      "step": 3036
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6185454076096106,
+      "learning_rate": 7.562054082082666e-06,
+      "loss": 0.5159,
+      "step": 3037
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2219963911960017,
+      "learning_rate": 7.560455863752911e-06,
+      "loss": 0.5868,
+      "step": 3038
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.013537456890812,
+      "learning_rate": 7.5588572907492545e-06,
+      "loss": 0.7644,
+      "step": 3039
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4314280730839473,
+      "learning_rate": 7.557258363293127e-06,
+      "loss": 0.6849,
+      "step": 3040
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8380739790951384,
+      "learning_rate": 7.555659081606016e-06,
+      "loss": 0.4928,
+      "step": 3041
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8403860002601804,
+      "learning_rate": 7.55405944590945e-06,
+      "loss": 0.4819,
+      "step": 3042
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0560123603629687,
+      "learning_rate": 7.5524594564250144e-06,
+      "loss": 0.5549,
+      "step": 3043
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8737898680550322,
+      "learning_rate": 7.550859113374336e-06,
+      "loss": 0.5082,
+      "step": 3044
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6362494072049818,
+      "learning_rate": 7.549258416979094e-06,
+      "loss": 0.4231,
+      "step": 3045
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.067393085383789,
+      "learning_rate": 7.547657367461019e-06,
+      "loss": 0.512,
+      "step": 3046
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.0198623639358795,
+      "learning_rate": 7.546055965041885e-06,
+      "loss": 0.6001,
+      "step": 3047
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0966414988714015,
+      "learning_rate": 7.54445420994352e-06,
+      "loss": 0.6089,
+      "step": 3048
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2089263817668106,
+      "learning_rate": 7.542852102387799e-06,
+      "loss": 0.5655,
+      "step": 3049
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9415941691021346,
+      "learning_rate": 7.541249642596645e-06,
+      "loss": 0.5378,
+      "step": 3050
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5719762847019463,
+      "learning_rate": 7.539646830792031e-06,
+      "loss": 0.4866,
+      "step": 3051
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.320773468786632,
+      "learning_rate": 7.538043667195979e-06,
+      "loss": 0.4347,
+      "step": 3052
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6828875222191377,
+      "learning_rate": 7.536440152030556e-06,
+      "loss": 0.4795,
+      "step": 3053
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.862764114394205,
+      "learning_rate": 7.534836285517883e-06,
+      "loss": 0.4713,
+      "step": 3054
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8646135813375817,
+      "learning_rate": 7.533232067880127e-06,
+      "loss": 0.4715,
+      "step": 3055
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7378958280272816,
+      "learning_rate": 7.531627499339504e-06,
+      "loss": 0.501,
+      "step": 3056
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7869871849195075,
+      "learning_rate": 7.530022580118278e-06,
+      "loss": 0.6667,
+      "step": 3057
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.900233915189838,
+      "learning_rate": 7.528417310438762e-06,
+      "loss": 0.5488,
+      "step": 3058
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1167739385272806,
+      "learning_rate": 7.526811690523319e-06,
+      "loss": 0.4699,
+      "step": 3059
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0897236073300878,
+      "learning_rate": 7.525205720594357e-06,
+      "loss": 0.5616,
+      "step": 3060
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7832049851391216,
+      "learning_rate": 7.5235994008743385e-06,
+      "loss": 0.5226,
+      "step": 3061
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.31305898735799,
+      "learning_rate": 7.521992731585766e-06,
+      "loss": 0.5587,
+      "step": 3062
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6816633306633766,
+      "learning_rate": 7.520385712951197e-06,
+      "loss": 0.5729,
+      "step": 3063
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2866652276576844,
+      "learning_rate": 7.518778345193236e-06,
+      "loss": 0.5684,
+      "step": 3064
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.344125482088026,
+      "learning_rate": 7.517170628534536e-06,
+      "loss": 0.5798,
+      "step": 3065
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9487499593237727,
+      "learning_rate": 7.515562563197794e-06,
+      "loss": 0.5593,
+      "step": 3066
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.325991750956368,
+      "learning_rate": 7.5139541494057635e-06,
+      "loss": 0.4926,
+      "step": 3067
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8550113577945218,
+      "learning_rate": 7.512345387381239e-06,
+      "loss": 0.4587,
+      "step": 3068
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.055398430362565,
+      "learning_rate": 7.510736277347067e-06,
+      "loss": 0.6278,
+      "step": 3069
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3835524235896774,
+      "learning_rate": 7.50912681952614e-06,
+      "loss": 0.4817,
+      "step": 3070
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0788569424648284,
+      "learning_rate": 7.507517014141401e-06,
+      "loss": 0.5252,
+      "step": 3071
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.323438556864119,
+      "learning_rate": 7.50590686141584e-06,
+      "loss": 0.5603,
+      "step": 3072
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9882217353541358,
+      "learning_rate": 7.504296361572494e-06,
+      "loss": 0.4634,
+      "step": 3073
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1398379051274556,
+      "learning_rate": 7.502685514834449e-06,
+      "loss": 0.4196,
+      "step": 3074
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6060615484709055,
+      "learning_rate": 7.501074321424842e-06,
+      "loss": 0.5019,
+      "step": 3075
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6815272515601314,
+      "learning_rate": 7.499462781566851e-06,
+      "loss": 0.5502,
+      "step": 3076
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.544292161334717,
+      "learning_rate": 7.49785089548371e-06,
+      "loss": 0.4574,
+      "step": 3077
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7462186380631068,
+      "learning_rate": 7.496238663398695e-06,
+      "loss": 0.4485,
+      "step": 3078
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.737583939454655,
+      "learning_rate": 7.494626085535132e-06,
+      "loss": 0.5146,
+      "step": 3079
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.790394097681309,
+      "learning_rate": 7.493013162116398e-06,
+      "loss": 0.4545,
+      "step": 3080
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4580651655804155,
+      "learning_rate": 7.49139989336591e-06,
+      "loss": 0.6078,
+      "step": 3081
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7449025654937735,
+      "learning_rate": 7.489786279507142e-06,
+      "loss": 0.5181,
+      "step": 3082
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.442941055781812,
+      "learning_rate": 7.48817232076361e-06,
+      "loss": 0.4203,
+      "step": 3083
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1978016641583755,
+      "learning_rate": 7.486558017358877e-06,
+      "loss": 0.5042,
+      "step": 3084
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.817690979932559,
+      "learning_rate": 7.484943369516558e-06,
+      "loss": 0.5954,
+      "step": 3085
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.442067351059011,
+      "learning_rate": 7.4833283774603174e-06,
+      "loss": 0.4733,
+      "step": 3086
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6067478074928188,
+      "learning_rate": 7.481713041413857e-06,
+      "loss": 0.6012,
+      "step": 3087
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8989158828469794,
+      "learning_rate": 7.480097361600937e-06,
+      "loss": 0.5297,
+      "step": 3088
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9253868412195685,
+      "learning_rate": 7.4784813382453595e-06,
+      "loss": 0.4931,
+      "step": 3089
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8934673982622288,
+      "learning_rate": 7.476864971570976e-06,
+      "loss": 0.5084,
+      "step": 3090
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9215448411472478,
+      "learning_rate": 7.475248261801687e-06,
+      "loss": 0.5307,
+      "step": 3091
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8266729622017928,
+      "learning_rate": 7.473631209161436e-06,
+      "loss": 0.4431,
+      "step": 3092
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.748961198995004,
+      "learning_rate": 7.472013813874219e-06,
+      "loss": 0.5396,
+      "step": 3093
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.12753897676637,
+      "learning_rate": 7.470396076164078e-06,
+      "loss": 0.544,
+      "step": 3094
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9251274020674929,
+      "learning_rate": 7.468777996255099e-06,
+      "loss": 0.4158,
+      "step": 3095
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3470826798567734,
+      "learning_rate": 7.46715957437142e-06,
+      "loss": 0.4367,
+      "step": 3096
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.4768803671420305,
+      "learning_rate": 7.465540810737224e-06,
+      "loss": 0.4344,
+      "step": 3097
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6092499994121563,
+      "learning_rate": 7.463921705576741e-06,
+      "loss": 0.4636,
+      "step": 3098
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5467406565561836,
+      "learning_rate": 7.4623022591142516e-06,
+      "loss": 0.5106,
+      "step": 3099
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.874925429396424,
+      "learning_rate": 7.460682471574079e-06,
+      "loss": 0.5718,
+      "step": 3100
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4800778518046664,
+      "learning_rate": 7.4590623431805965e-06,
+      "loss": 0.4719,
+      "step": 3101
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1852046980973134,
+      "learning_rate": 7.457441874158224e-06,
+      "loss": 0.489,
+      "step": 3102
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1271942521140907,
+      "learning_rate": 7.45582106473143e-06,
+      "loss": 0.5666,
+      "step": 3103
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3588116038229594,
+      "learning_rate": 7.454199915124727e-06,
+      "loss": 0.5412,
+      "step": 3104
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7927395601518523,
+      "learning_rate": 7.452578425562677e-06,
+      "loss": 0.5562,
+      "step": 3105
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.006609263574854,
+      "learning_rate": 7.450956596269887e-06,
+      "loss": 0.4832,
+      "step": 3106
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9512060861775211,
+      "learning_rate": 7.449334427471014e-06,
+      "loss": 0.5648,
+      "step": 3107
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2965786497543523,
+      "learning_rate": 7.447711919390761e-06,
+      "loss": 0.4988,
+      "step": 3108
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.111194626432888,
+      "learning_rate": 7.446089072253877e-06,
+      "loss": 0.5307,
+      "step": 3109
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9068368380845009,
+      "learning_rate": 7.444465886285157e-06,
+      "loss": 0.5114,
+      "step": 3110
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.582403869676365,
+      "learning_rate": 7.4428423617094485e-06,
+      "loss": 0.5016,
+      "step": 3111
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.389292042633605,
+      "learning_rate": 7.441218498751637e-06,
+      "loss": 0.5436,
+      "step": 3112
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1219242483238445,
+      "learning_rate": 7.439594297636663e-06,
+      "loss": 0.6301,
+      "step": 3113
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.267207976764517,
+      "learning_rate": 7.437969758589508e-06,
+      "loss": 0.4732,
+      "step": 3114
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.687644657326908,
+      "learning_rate": 7.436344881835205e-06,
+      "loss": 0.5308,
+      "step": 3115
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.809851846195055,
+      "learning_rate": 7.434719667598831e-06,
+      "loss": 0.5155,
+      "step": 3116
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.951326683921181,
+      "learning_rate": 7.43309411610551e-06,
+      "loss": 0.5165,
+      "step": 3117
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9739660998919277,
+      "learning_rate": 7.431468227580415e-06,
+      "loss": 0.6536,
+      "step": 3118
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6708568918918405,
+      "learning_rate": 7.42984200224876e-06,
+      "loss": 0.5641,
+      "step": 3119
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.926609162126331,
+      "learning_rate": 7.428215440335814e-06,
+      "loss": 0.5283,
+      "step": 3120
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5048361681299824,
+      "learning_rate": 7.426588542066885e-06,
+      "loss": 0.4949,
+      "step": 3121
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.816417055949335,
+      "learning_rate": 7.424961307667331e-06,
+      "loss": 0.6014,
+      "step": 3122
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.431129522000982,
+      "learning_rate": 7.423333737362558e-06,
+      "loss": 0.5607,
+      "step": 3123
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9607745394510776,
+      "learning_rate": 7.421705831378014e-06,
+      "loss": 0.4455,
+      "step": 3124
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.971731739656572,
+      "learning_rate": 7.420077589939199e-06,
+      "loss": 0.4937,
+      "step": 3125
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0899099082163612,
+      "learning_rate": 7.4184490132716534e-06,
+      "loss": 0.4584,
+      "step": 3126
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9735869363107756,
+      "learning_rate": 7.41682010160097e-06,
+      "loss": 0.5652,
+      "step": 3127
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6375971702833931,
+      "learning_rate": 7.415190855152786e-06,
+      "loss": 0.5475,
+      "step": 3128
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.45023314669999,
+      "learning_rate": 7.413561274152783e-06,
+      "loss": 0.4809,
+      "step": 3129
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.056422857625645,
+      "learning_rate": 7.411931358826689e-06,
+      "loss": 0.5171,
+      "step": 3130
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5054868958806142,
+      "learning_rate": 7.410301109400281e-06,
+      "loss": 0.4808,
+      "step": 3131
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.3828056300272458,
+      "learning_rate": 7.4086705260993814e-06,
+      "loss": 0.519,
+      "step": 3132
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7108145895349829,
+      "learning_rate": 7.4070396091498575e-06,
+      "loss": 0.5167,
+      "step": 3133
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1811405072517664,
+      "learning_rate": 7.405408358777624e-06,
+      "loss": 0.4639,
+      "step": 3134
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9485573921471266,
+      "learning_rate": 7.403776775208641e-06,
+      "loss": 0.5804,
+      "step": 3135
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.118709744754571,
+      "learning_rate": 7.402144858668915e-06,
+      "loss": 0.5453,
+      "step": 3136
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9540590906381115,
+      "learning_rate": 7.4005126093845005e-06,
+      "loss": 0.7287,
+      "step": 3137
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8759706432653271,
+      "learning_rate": 7.398880027581494e-06,
+      "loss": 0.606,
+      "step": 3138
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9061848246277939,
+      "learning_rate": 7.397247113486044e-06,
+      "loss": 0.738,
+      "step": 3139
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8814007133360884,
+      "learning_rate": 7.395613867324336e-06,
+      "loss": 0.4775,
+      "step": 3140
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.914763021175095,
+      "learning_rate": 7.393980289322611e-06,
+      "loss": 0.5371,
+      "step": 3141
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8351047473956608,
+      "learning_rate": 7.3923463797071515e-06,
+      "loss": 0.4919,
+      "step": 3142
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.610456681635869,
+      "learning_rate": 7.390712138704286e-06,
+      "loss": 0.6067,
+      "step": 3143
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8276151344518674,
+      "learning_rate": 7.389077566540388e-06,
+      "loss": 0.4659,
+      "step": 3144
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1450913710434842,
+      "learning_rate": 7.38744266344188e-06,
+      "loss": 0.5543,
+      "step": 3145
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3989405299325575,
+      "learning_rate": 7.385807429635229e-06,
+      "loss": 0.4398,
+      "step": 3146
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.823231640939177,
+      "learning_rate": 7.384171865346944e-06,
+      "loss": 0.5893,
+      "step": 3147
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4459555289103547,
+      "learning_rate": 7.382535970803586e-06,
+      "loss": 0.5044,
+      "step": 3148
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.099212623753445,
+      "learning_rate": 7.380899746231758e-06,
+      "loss": 0.525,
+      "step": 3149
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6789875267504135,
+      "learning_rate": 7.379263191858109e-06,
+      "loss": 0.5282,
+      "step": 3150
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7343144179524803,
+      "learning_rate": 7.377626307909336e-06,
+      "loss": 0.5301,
+      "step": 3151
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8952217035453864,
+      "learning_rate": 7.375989094612177e-06,
+      "loss": 0.4757,
+      "step": 3152
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.161547708729708,
+      "learning_rate": 7.374351552193421e-06,
+      "loss": 0.4779,
+      "step": 3153
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.243781677545647,
+      "learning_rate": 7.372713680879901e-06,
+      "loss": 0.5456,
+      "step": 3154
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9182894616810096,
+      "learning_rate": 7.371075480898491e-06,
+      "loss": 0.5009,
+      "step": 3155
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7326257731793997,
+      "learning_rate": 7.369436952476116e-06,
+      "loss": 0.6112,
+      "step": 3156
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.031224117428107,
+      "learning_rate": 7.367798095839745e-06,
+      "loss": 0.4417,
+      "step": 3157
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.726987330858252,
+      "learning_rate": 7.366158911216391e-06,
+      "loss": 0.5807,
+      "step": 3158
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9804120454591008,
+      "learning_rate": 7.364519398833115e-06,
+      "loss": 0.4744,
+      "step": 3159
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4187826328488,
+      "learning_rate": 7.3628795589170224e-06,
+      "loss": 0.4684,
+      "step": 3160
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6058890337316674,
+      "learning_rate": 7.361239391695261e-06,
+      "loss": 0.5961,
+      "step": 3161
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9740778067132343,
+      "learning_rate": 7.35959889739503e-06,
+      "loss": 0.586,
+      "step": 3162
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9628608964201661,
+      "learning_rate": 7.357958076243567e-06,
+      "loss": 0.4806,
+      "step": 3163
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0106988005801811,
+      "learning_rate": 7.35631692846816e-06,
+      "loss": 0.7417,
+      "step": 3164
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.126958672674491,
+      "learning_rate": 7.354675454296141e-06,
+      "loss": 0.5103,
+      "step": 3165
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7176010213065342,
+      "learning_rate": 7.353033653954885e-06,
+      "loss": 0.5621,
+      "step": 3166
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.695755804760129,
+      "learning_rate": 7.351391527671815e-06,
+      "loss": 0.493,
+      "step": 3167
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8645049880854505,
+      "learning_rate": 7.349749075674399e-06,
+      "loss": 0.5014,
+      "step": 3168
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.718375430885016,
+      "learning_rate": 7.348106298190145e-06,
+      "loss": 0.4857,
+      "step": 3169
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.038685116778116,
+      "learning_rate": 7.346463195446617e-06,
+      "loss": 0.5542,
+      "step": 3170
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7030806662947264,
+      "learning_rate": 7.3448197676714115e-06,
+      "loss": 0.5403,
+      "step": 3171
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6220301492539222,
+      "learning_rate": 7.343176015092177e-06,
+      "loss": 0.5198,
+      "step": 3172
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9492196877989625,
+      "learning_rate": 7.341531937936608e-06,
+      "loss": 0.4857,
+      "step": 3173
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0090274678653395,
+      "learning_rate": 7.33988753643244e-06,
+      "loss": 0.6522,
+      "step": 3174
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9027644548490903,
+      "learning_rate": 7.3382428108074566e-06,
+      "loss": 0.4974,
+      "step": 3175
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.5710390636701534,
+      "learning_rate": 7.336597761289484e-06,
+      "loss": 0.5544,
+      "step": 3176
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5275365158756697,
+      "learning_rate": 7.334952388106393e-06,
+      "loss": 0.5084,
+      "step": 3177
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0494201519833197,
+      "learning_rate": 7.3333066914861024e-06,
+      "loss": 0.4988,
+      "step": 3178
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3934631584707122,
+      "learning_rate": 7.331660671656574e-06,
+      "loss": 0.6034,
+      "step": 3179
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.723782594664949,
+      "learning_rate": 7.330014328845813e-06,
+      "loss": 0.524,
+      "step": 3180
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.8114999798888345,
+      "learning_rate": 7.328367663281869e-06,
+      "loss": 0.591,
+      "step": 3181
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8561883525509666,
+      "learning_rate": 7.326720675192841e-06,
+      "loss": 0.4606,
+      "step": 3182
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.94805703412477,
+      "learning_rate": 7.325073364806867e-06,
+      "loss": 0.6989,
+      "step": 3183
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.973679307183509,
+      "learning_rate": 7.323425732352134e-06,
+      "loss": 0.5307,
+      "step": 3184
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.775740756994591,
+      "learning_rate": 7.321777778056871e-06,
+      "loss": 0.531,
+      "step": 3185
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0502821528170987,
+      "learning_rate": 7.320129502149353e-06,
+      "loss": 0.5389,
+      "step": 3186
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2253858743489636,
+      "learning_rate": 7.318480904857897e-06,
+      "loss": 0.5612,
+      "step": 3187
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.621153424068711,
+      "learning_rate": 7.316831986410868e-06,
+      "loss": 0.5588,
+      "step": 3188
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9833168580716858,
+      "learning_rate": 7.315182747036674e-06,
+      "loss": 0.4838,
+      "step": 3189
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8524564170062607,
+      "learning_rate": 7.313533186963767e-06,
+      "loss": 0.7002,
+      "step": 3190
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.03107634064579,
+      "learning_rate": 7.311883306420643e-06,
+      "loss": 0.5449,
+      "step": 3191
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8705110042887458,
+      "learning_rate": 7.310233105635843e-06,
+      "loss": 0.4446,
+      "step": 3192
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.123840300323815,
+      "learning_rate": 7.308582584837955e-06,
+      "loss": 0.4796,
+      "step": 3193
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9752455487408522,
+      "learning_rate": 7.306931744255607e-06,
+      "loss": 0.5302,
+      "step": 3194
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7101197308004368,
+      "learning_rate": 7.305280584117471e-06,
+      "loss": 0.5844,
+      "step": 3195
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8749406624922914,
+      "learning_rate": 7.303629104652271e-06,
+      "loss": 0.7354,
+      "step": 3196
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0811736418184013,
+      "learning_rate": 7.301977306088766e-06,
+      "loss": 0.5283,
+      "step": 3197
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6971218815444455,
+      "learning_rate": 7.300325188655762e-06,
+      "loss": 0.4601,
+      "step": 3198
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.852497922082125,
+      "learning_rate": 7.298672752582111e-06,
+      "loss": 0.5307,
+      "step": 3199
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.550515058265413,
+      "learning_rate": 7.29701999809671e-06,
+      "loss": 0.4976,
+      "step": 3200
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0126859120858676,
+      "learning_rate": 7.295366925428494e-06,
+      "loss": 0.5404,
+      "step": 3201
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.841462450703764,
+      "learning_rate": 7.293713534806451e-06,
+      "loss": 0.5051,
+      "step": 3202
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8862402059201049,
+      "learning_rate": 7.292059826459607e-06,
+      "loss": 0.5322,
+      "step": 3203
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6034025192903605,
+      "learning_rate": 7.290405800617033e-06,
+      "loss": 0.4708,
+      "step": 3204
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.238645064785399,
+      "learning_rate": 7.288751457507844e-06,
+      "loss": 0.4763,
+      "step": 3205
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.049591709278908,
+      "learning_rate": 7.287096797361197e-06,
+      "loss": 0.5366,
+      "step": 3206
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.532309091186629,
+      "learning_rate": 7.285441820406301e-06,
+      "loss": 0.5132,
+      "step": 3207
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.863228770837838,
+      "learning_rate": 7.283786526872398e-06,
+      "loss": 0.5385,
+      "step": 3208
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.868284190461831,
+      "learning_rate": 7.282130916988781e-06,
+      "loss": 0.5288,
+      "step": 3209
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6412857947209902,
+      "learning_rate": 7.280474990984785e-06,
+      "loss": 0.4763,
+      "step": 3210
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9197115324563732,
+      "learning_rate": 7.278818749089789e-06,
+      "loss": 0.7684,
+      "step": 3211
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3460030317762084,
+      "learning_rate": 7.277162191533213e-06,
+      "loss": 0.5638,
+      "step": 3212
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6850650615946843,
+      "learning_rate": 7.275505318544527e-06,
+      "loss": 0.6287,
+      "step": 3213
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1930545690492353,
+      "learning_rate": 7.273848130353237e-06,
+      "loss": 0.4971,
+      "step": 3214
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.3505132156274433,
+      "learning_rate": 7.272190627188897e-06,
+      "loss": 0.4765,
+      "step": 3215
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2824895546465958,
+      "learning_rate": 7.270532809281106e-06,
+      "loss": 0.5819,
+      "step": 3216
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.0073510658478195,
+      "learning_rate": 7.268874676859503e-06,
+      "loss": 0.4771,
+      "step": 3217
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9652088793024016,
+      "learning_rate": 7.267216230153773e-06,
+      "loss": 0.4579,
+      "step": 3218
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0067956395934017,
+      "learning_rate": 7.265557469393643e-06,
+      "loss": 0.5462,
+      "step": 3219
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.234519215570638,
+      "learning_rate": 7.263898394808885e-06,
+      "loss": 0.479,
+      "step": 3220
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8937308113026818,
+      "learning_rate": 7.262239006629315e-06,
+      "loss": 0.4489,
+      "step": 3221
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7872955854870838,
+      "learning_rate": 7.260579305084787e-06,
+      "loss": 0.4731,
+      "step": 3222
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4114170190253597,
+      "learning_rate": 7.2589192904052065e-06,
+      "loss": 0.5488,
+      "step": 3223
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8632435357532933,
+      "learning_rate": 7.257258962820517e-06,
+      "loss": 0.4936,
+      "step": 3224
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5889494245152085,
+      "learning_rate": 7.255598322560707e-06,
+      "loss": 0.4966,
+      "step": 3225
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4704757724789483,
+      "learning_rate": 7.253937369855808e-06,
+      "loss": 0.4387,
+      "step": 3226
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7549965397048326,
+      "learning_rate": 7.252276104935896e-06,
+      "loss": 0.5216,
+      "step": 3227
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.809916264275648,
+      "learning_rate": 7.250614528031087e-06,
+      "loss": 0.5805,
+      "step": 3228
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9489089900126306,
+      "learning_rate": 7.248952639371543e-06,
+      "loss": 0.5042,
+      "step": 3229
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8802118945933053,
+      "learning_rate": 7.247290439187471e-06,
+      "loss": 0.5728,
+      "step": 3230
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.074895522856019,
+      "learning_rate": 7.245627927709114e-06,
+      "loss": 0.6152,
+      "step": 3231
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.592295403445032,
+      "learning_rate": 7.243965105166766e-06,
+      "loss": 0.5559,
+      "step": 3232
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7190186318556324,
+      "learning_rate": 7.242301971790762e-06,
+      "loss": 0.5651,
+      "step": 3233
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.085765370388405,
+      "learning_rate": 7.240638527811474e-06,
+      "loss": 0.5603,
+      "step": 3234
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7110710804156666,
+      "learning_rate": 7.238974773459328e-06,
+      "loss": 0.4211,
+      "step": 3235
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8137208200112913,
+      "learning_rate": 7.237310708964783e-06,
+      "loss": 0.4364,
+      "step": 3236
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.023132542880251,
+      "learning_rate": 7.235646334558346e-06,
+      "loss": 0.4862,
+      "step": 3237
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6848600855234326,
+      "learning_rate": 7.233981650470567e-06,
+      "loss": 0.4482,
+      "step": 3238
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8868382311741896,
+      "learning_rate": 7.2323166569320366e-06,
+      "loss": 0.5264,
+      "step": 3239
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0148909892719913,
+      "learning_rate": 7.230651354173389e-06,
+      "loss": 0.5207,
+      "step": 3240
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.885194804203804,
+      "learning_rate": 7.228985742425302e-06,
+      "loss": 0.5363,
+      "step": 3241
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8144530508181507,
+      "learning_rate": 7.227319821918496e-06,
+      "loss": 0.568,
+      "step": 3242
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.149548295102153,
+      "learning_rate": 7.225653592883734e-06,
+      "loss": 0.5095,
+      "step": 3243
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1772612870531507,
+      "learning_rate": 7.2239870555518224e-06,
+      "loss": 0.5496,
+      "step": 3244
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8398368862289578,
+      "learning_rate": 7.222320210153608e-06,
+      "loss": 0.475,
+      "step": 3245
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1894193382149774,
+      "learning_rate": 7.2206530569199855e-06,
+      "loss": 0.4299,
+      "step": 3246
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9107026587726634,
+      "learning_rate": 7.218985596081884e-06,
+      "loss": 0.5145,
+      "step": 3247
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.038233063128196,
+      "learning_rate": 7.217317827870283e-06,
+      "loss": 0.5319,
+      "step": 3248
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2651182052750625,
+      "learning_rate": 7.2156497525162e-06,
+      "loss": 0.4441,
+      "step": 3249
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1163274374191965,
+      "learning_rate": 7.213981370250698e-06,
+      "loss": 0.5158,
+      "step": 3250
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.5688874956877426,
+      "learning_rate": 7.212312681304879e-06,
+      "loss": 0.4803,
+      "step": 3251
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8526500809353896,
+      "learning_rate": 7.2106436859098904e-06,
+      "loss": 0.522,
+      "step": 3252
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8788397472015697,
+      "learning_rate": 7.208974384296921e-06,
+      "loss": 0.5756,
+      "step": 3253
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9713596369811346,
+      "learning_rate": 7.207304776697204e-06,
+      "loss": 0.4271,
+      "step": 3254
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.789984616457143,
+      "learning_rate": 7.205634863342011e-06,
+      "loss": 0.4678,
+      "step": 3255
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8325535741031778,
+      "learning_rate": 7.203964644462658e-06,
+      "loss": 0.7334,
+      "step": 3256
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.062446456036115,
+      "learning_rate": 7.202294120290505e-06,
+      "loss": 0.4655,
+      "step": 3257
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8397795903457739,
+      "learning_rate": 7.200623291056952e-06,
+      "loss": 0.4547,
+      "step": 3258
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.003938829803272,
+      "learning_rate": 7.198952156993441e-06,
+      "loss": 0.4871,
+      "step": 3259
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.89124909279938,
+      "learning_rate": 7.197280718331459e-06,
+      "loss": 0.7692,
+      "step": 3260
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9020077513978273,
+      "learning_rate": 7.195608975302532e-06,
+      "loss": 0.4372,
+      "step": 3261
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0370801612671094,
+      "learning_rate": 7.19393692813823e-06,
+      "loss": 0.5475,
+      "step": 3262
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.801995486360701,
+      "learning_rate": 7.192264577070165e-06,
+      "loss": 0.4844,
+      "step": 3263
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7835752072198074,
+      "learning_rate": 7.19059192232999e-06,
+      "loss": 0.3805,
+      "step": 3264
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 15.601082772477739,
+      "learning_rate": 7.188918964149402e-06,
+      "loss": 0.4931,
+      "step": 3265
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.012785201442268,
+      "learning_rate": 7.187245702760137e-06,
+      "loss": 0.582,
+      "step": 3266
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7760690634172722,
+      "learning_rate": 7.185572138393977e-06,
+      "loss": 0.5204,
+      "step": 3267
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9022646832204473,
+      "learning_rate": 7.183898271282743e-06,
+      "loss": 0.4624,
+      "step": 3268
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.634757795140507,
+      "learning_rate": 7.182224101658299e-06,
+      "loss": 0.5325,
+      "step": 3269
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8767580000759363,
+      "learning_rate": 7.180549629752551e-06,
+      "loss": 0.5552,
+      "step": 3270
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.828377519318436,
+      "learning_rate": 7.178874855797445e-06,
+      "loss": 0.4504,
+      "step": 3271
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.247467794038215,
+      "learning_rate": 7.1771997800249715e-06,
+      "loss": 0.5546,
+      "step": 3272
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.19397919747565,
+      "learning_rate": 7.1755244026671625e-06,
+      "loss": 0.4685,
+      "step": 3273
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9535158630812461,
+      "learning_rate": 7.173848723956088e-06,
+      "loss": 0.4488,
+      "step": 3274
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7406526882238667,
+      "learning_rate": 7.172172744123867e-06,
+      "loss": 0.4142,
+      "step": 3275
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8592206525877195,
+      "learning_rate": 7.170496463402652e-06,
+      "loss": 0.532,
+      "step": 3276
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8887567127952223,
+      "learning_rate": 7.1688198820246425e-06,
+      "loss": 0.4276,
+      "step": 3277
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0180699588180255,
+      "learning_rate": 7.16714300022208e-06,
+      "loss": 0.4776,
+      "step": 3278
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8021955720026592,
+      "learning_rate": 7.165465818227244e-06,
+      "loss": 0.5022,
+      "step": 3279
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0094149440716094,
+      "learning_rate": 7.163788336272458e-06,
+      "loss": 0.47,
+      "step": 3280
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9017665413054776,
+      "learning_rate": 7.162110554590087e-06,
+      "loss": 0.5316,
+      "step": 3281
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2026234243488587,
+      "learning_rate": 7.160432473412535e-06,
+      "loss": 0.4397,
+      "step": 3282
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.356663742354321,
+      "learning_rate": 7.158754092972252e-06,
+      "loss": 0.4609,
+      "step": 3283
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0332566541098696,
+      "learning_rate": 7.157075413501725e-06,
+      "loss": 0.6345,
+      "step": 3284
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5678032929905938,
+      "learning_rate": 7.155396435233486e-06,
+      "loss": 0.4311,
+      "step": 3285
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8825154841327794,
+      "learning_rate": 7.153717158400107e-06,
+      "loss": 0.4679,
+      "step": 3286
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8969582820091169,
+      "learning_rate": 7.152037583234198e-06,
+      "loss": 0.5157,
+      "step": 3287
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8512636929650883,
+      "learning_rate": 7.150357709968417e-06,
+      "loss": 0.418,
+      "step": 3288
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8099448106161273,
+      "learning_rate": 7.14867753883546e-06,
+      "loss": 0.5412,
+      "step": 3289
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4534207721145647,
+      "learning_rate": 7.146997070068062e-06,
+      "loss": 0.5074,
+      "step": 3290
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7880624734714443,
+      "learning_rate": 7.145316303898999e-06,
+      "loss": 0.5573,
+      "step": 3291
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6807351687304894,
+      "learning_rate": 7.143635240561095e-06,
+      "loss": 0.4644,
+      "step": 3292
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.081068964161852,
+      "learning_rate": 7.14195388028721e-06,
+      "loss": 0.5173,
+      "step": 3293
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.293983820807733,
+      "learning_rate": 7.1402722233102425e-06,
+      "loss": 0.5717,
+      "step": 3294
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0848089250062345,
+      "learning_rate": 7.138590269863139e-06,
+      "loss": 0.4972,
+      "step": 3295
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7213707390818764,
+      "learning_rate": 7.136908020178881e-06,
+      "loss": 0.5583,
+      "step": 3296
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6598338037195575,
+      "learning_rate": 7.1352254744904945e-06,
+      "loss": 0.514,
+      "step": 3297
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.734873940679305,
+      "learning_rate": 7.133542633031044e-06,
+      "loss": 0.4629,
+      "step": 3298
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9460876505155544,
+      "learning_rate": 7.131859496033638e-06,
+      "loss": 0.4747,
+      "step": 3299
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.521511835497702,
+      "learning_rate": 7.130176063731424e-06,
+      "loss": 0.5332,
+      "step": 3300
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.245208563722081,
+      "learning_rate": 7.128492336357591e-06,
+      "loss": 0.5573,
+      "step": 3301
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.101163022208161,
+      "learning_rate": 7.126808314145367e-06,
+      "loss": 0.4659,
+      "step": 3302
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8883660155409427,
+      "learning_rate": 7.125123997328025e-06,
+      "loss": 0.488,
+      "step": 3303
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6917231504833827,
+      "learning_rate": 7.123439386138874e-06,
+      "loss": 0.4455,
+      "step": 3304
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2753292729551196,
+      "learning_rate": 7.121754480811268e-06,
+      "loss": 0.5959,
+      "step": 3305
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.351000731916369,
+      "learning_rate": 7.1200692815785985e-06,
+      "loss": 0.5244,
+      "step": 3306
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9874802463165762,
+      "learning_rate": 7.118383788674299e-06,
+      "loss": 0.5239,
+      "step": 3307
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.145282082179849,
+      "learning_rate": 7.116698002331843e-06,
+      "loss": 0.5579,
+      "step": 3308
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.808496389465618,
+      "learning_rate": 7.115011922784748e-06,
+      "loss": 0.4003,
+      "step": 3309
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8347951753979703,
+      "learning_rate": 7.113325550266568e-06,
+      "loss": 0.5092,
+      "step": 3310
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6199587495911798,
+      "learning_rate": 7.111638885010897e-06,
+      "loss": 0.5317,
+      "step": 3311
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.771617775677887,
+      "learning_rate": 7.109951927251375e-06,
+      "loss": 0.4248,
+      "step": 3312
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.814254324992072,
+      "learning_rate": 7.108264677221678e-06,
+      "loss": 0.4322,
+      "step": 3313
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9586739025325208,
+      "learning_rate": 7.106577135155522e-06,
+      "loss": 0.5439,
+      "step": 3314
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.459431482883816,
+      "learning_rate": 7.104889301286666e-06,
+      "loss": 0.516,
+      "step": 3315
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9247091224471418,
+      "learning_rate": 7.10320117584891e-06,
+      "loss": 0.4875,
+      "step": 3316
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.477395775077742,
+      "learning_rate": 7.101512759076089e-06,
+      "loss": 0.53,
+      "step": 3317
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8133035019620085,
+      "learning_rate": 7.099824051202085e-06,
+      "loss": 0.4995,
+      "step": 3318
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8168220514325955,
+      "learning_rate": 7.0981350524608185e-06,
+      "loss": 0.4669,
+      "step": 3319
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7199759712727876,
+      "learning_rate": 7.096445763086247e-06,
+      "loss": 0.4781,
+      "step": 3320
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4478372631076475,
+      "learning_rate": 7.094756183312372e-06,
+      "loss": 0.5902,
+      "step": 3321
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.445252417717323,
+      "learning_rate": 7.093066313373233e-06,
+      "loss": 0.4546,
+      "step": 3322
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.134598194439933,
+      "learning_rate": 7.091376153502911e-06,
+      "loss": 0.494,
+      "step": 3323
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7549817590983094,
+      "learning_rate": 7.089685703935527e-06,
+      "loss": 0.4912,
+      "step": 3324
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.873593100951613,
+      "learning_rate": 7.087994964905241e-06,
+      "loss": 0.4553,
+      "step": 3325
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.726683483613141,
+      "learning_rate": 7.086303936646252e-06,
+      "loss": 0.57,
+      "step": 3326
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8760484714681507,
+      "learning_rate": 7.084612619392806e-06,
+      "loss": 0.4988,
+      "step": 3327
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.071186592474359,
+      "learning_rate": 7.08292101337918e-06,
+      "loss": 0.4624,
+      "step": 3328
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.6034368668242145,
+      "learning_rate": 7.081229118839694e-06,
+      "loss": 0.4847,
+      "step": 3329
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.020338677373167,
+      "learning_rate": 7.079536936008713e-06,
+      "loss": 0.5451,
+      "step": 3330
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.026625897949116,
+      "learning_rate": 7.077844465120637e-06,
+      "loss": 0.4378,
+      "step": 3331
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9491601118622488,
+      "learning_rate": 7.076151706409904e-06,
+      "loss": 0.5308,
+      "step": 3332
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7985596183329262,
+      "learning_rate": 7.074458660110996e-06,
+      "loss": 0.5172,
+      "step": 3333
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6302903652287353,
+      "learning_rate": 7.072765326458434e-06,
+      "loss": 0.5382,
+      "step": 3334
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7193246501495403,
+      "learning_rate": 7.0710717056867795e-06,
+      "loss": 0.5372,
+      "step": 3335
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.378660331233637,
+      "learning_rate": 7.0693777980306285e-06,
+      "loss": 0.486,
+      "step": 3336
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.231866403492623,
+      "learning_rate": 7.067683603724624e-06,
+      "loss": 0.5287,
+      "step": 3337
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.829571233886774,
+      "learning_rate": 7.065989123003446e-06,
+      "loss": 0.5689,
+      "step": 3338
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6495981190605775,
+      "learning_rate": 7.064294356101813e-06,
+      "loss": 0.5163,
+      "step": 3339
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9253678636633127,
+      "learning_rate": 7.062599303254481e-06,
+      "loss": 0.4565,
+      "step": 3340
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5668418569876255,
+      "learning_rate": 7.060903964696253e-06,
+      "loss": 0.5402,
+      "step": 3341
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.215486829168166,
+      "learning_rate": 7.0592083406619625e-06,
+      "loss": 0.4735,
+      "step": 3342
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8619760268674858,
+      "learning_rate": 7.057512431386491e-06,
+      "loss": 0.7092,
+      "step": 3343
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8582916463553318,
+      "learning_rate": 7.055816237104753e-06,
+      "loss": 0.5029,
+      "step": 3344
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7808791953412382,
+      "learning_rate": 7.054119758051706e-06,
+      "loss": 0.4882,
+      "step": 3345
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6801590873260863,
+      "learning_rate": 7.052422994462347e-06,
+      "loss": 0.5823,
+      "step": 3346
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7823987879297132,
+      "learning_rate": 7.050725946571709e-06,
+      "loss": 0.4933,
+      "step": 3347
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.184400530057603,
+      "learning_rate": 7.04902861461487e-06,
+      "loss": 0.4894,
+      "step": 3348
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.774840272074094,
+      "learning_rate": 7.04733099882694e-06,
+      "loss": 0.4476,
+      "step": 3349
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9510124060593994,
+      "learning_rate": 7.045633099443075e-06,
+      "loss": 0.4145,
+      "step": 3350
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.431760183714208,
+      "learning_rate": 7.043934916698468e-06,
+      "loss": 0.6732,
+      "step": 3351
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0635805673296357,
+      "learning_rate": 7.04223645082835e-06,
+      "loss": 0.5096,
+      "step": 3352
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8403659969769715,
+      "learning_rate": 7.040537702067993e-06,
+      "loss": 0.4947,
+      "step": 3353
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7130317839654687,
+      "learning_rate": 7.038838670652707e-06,
+      "loss": 0.5062,
+      "step": 3354
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.810994776009769,
+      "learning_rate": 7.037139356817842e-06,
+      "loss": 0.5007,
+      "step": 3355
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3411930614027816,
+      "learning_rate": 7.035439760798785e-06,
+      "loss": 0.5781,
+      "step": 3356
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0230287686438215,
+      "learning_rate": 7.0337398828309644e-06,
+      "loss": 0.533,
+      "step": 3357
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4932216934354003,
+      "learning_rate": 7.032039723149849e-06,
+      "loss": 0.4499,
+      "step": 3358
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7854868030378267,
+      "learning_rate": 7.030339281990943e-06,
+      "loss": 0.6705,
+      "step": 3359
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9843314652864446,
+      "learning_rate": 7.02863855958979e-06,
+      "loss": 0.5876,
+      "step": 3360
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3181806647864467,
+      "learning_rate": 7.0269375561819764e-06,
+      "loss": 0.5392,
+      "step": 3361
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.785424370397955,
+      "learning_rate": 7.025236272003122e-06,
+      "loss": 0.5339,
+      "step": 3362
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2542539743269416,
+      "learning_rate": 7.023534707288891e-06,
+      "loss": 0.4684,
+      "step": 3363
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.359697249439902,
+      "learning_rate": 7.0218328622749836e-06,
+      "loss": 0.5553,
+      "step": 3364
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0954780144779774,
+      "learning_rate": 7.0201307371971375e-06,
+      "loss": 0.6043,
+      "step": 3365
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0241381270734045,
+      "learning_rate": 7.0184283322911314e-06,
+      "loss": 0.6349,
+      "step": 3366
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.14007675523266,
+      "learning_rate": 7.016725647792783e-06,
+      "loss": 0.4981,
+      "step": 3367
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8713838290459028,
+      "learning_rate": 7.015022683937947e-06,
+      "loss": 0.4702,
+      "step": 3368
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2090028356146045,
+      "learning_rate": 7.013319440962516e-06,
+      "loss": 0.5631,
+      "step": 3369
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.707387595798634,
+      "learning_rate": 7.011615919102427e-06,
+      "loss": 0.5185,
+      "step": 3370
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.973856408742496,
+      "learning_rate": 7.009912118593648e-06,
+      "loss": 0.5069,
+      "step": 3371
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9378565374671552,
+      "learning_rate": 7.008208039672191e-06,
+      "loss": 0.5381,
+      "step": 3372
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.054627580064713,
+      "learning_rate": 7.006503682574105e-06,
+      "loss": 0.5188,
+      "step": 3373
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.766154877738583,
+      "learning_rate": 7.0047990475354755e-06,
+      "loss": 0.5217,
+      "step": 3374
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2416538809460973,
+      "learning_rate": 7.00309413479243e-06,
+      "loss": 0.494,
+      "step": 3375
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8974841364632913,
+      "learning_rate": 7.001388944581131e-06,
+      "loss": 0.4418,
+      "step": 3376
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.087674603559062,
+      "learning_rate": 6.999683477137783e-06,
+      "loss": 0.5272,
+      "step": 3377
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8459375335471115,
+      "learning_rate": 6.997977732698625e-06,
+      "loss": 0.5425,
+      "step": 3378
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.335732608265989,
+      "learning_rate": 6.996271711499938e-06,
+      "loss": 0.5632,
+      "step": 3379
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.872561544660747,
+      "learning_rate": 6.9945654137780384e-06,
+      "loss": 0.4418,
+      "step": 3380
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9122515933073285,
+      "learning_rate": 6.992858839769285e-06,
+      "loss": 0.4891,
+      "step": 3381
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.808134879536226,
+      "learning_rate": 6.99115198971007e-06,
+      "loss": 0.497,
+      "step": 3382
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9315048228461427,
+      "learning_rate": 6.989444863836825e-06,
+      "loss": 0.5139,
+      "step": 3383
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0596979396945376,
+      "learning_rate": 6.9877374623860215e-06,
+      "loss": 0.5995,
+      "step": 3384
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.064542258563414,
+      "learning_rate": 6.98602978559417e-06,
+      "loss": 0.5419,
+      "step": 3385
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.369031792877626,
+      "learning_rate": 6.984321833697817e-06,
+      "loss": 0.555,
+      "step": 3386
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8853456989645592,
+      "learning_rate": 6.982613606933547e-06,
+      "loss": 0.5066,
+      "step": 3387
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.278776163921591,
+      "learning_rate": 6.9809051055379825e-06,
+      "loss": 0.6361,
+      "step": 3388
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7858077315788579,
+      "learning_rate": 6.9791963297477875e-06,
+      "loss": 0.5466,
+      "step": 3389
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9413796084415047,
+      "learning_rate": 6.97748727979966e-06,
+      "loss": 0.4507,
+      "step": 3390
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0787315889283353,
+      "learning_rate": 6.975777955930336e-06,
+      "loss": 0.599,
+      "step": 3391
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.309439684043517,
+      "learning_rate": 6.974068358376591e-06,
+      "loss": 0.5042,
+      "step": 3392
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3893742041052435,
+      "learning_rate": 6.97235848737524e-06,
+      "loss": 0.5213,
+      "step": 3393
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8943864999761024,
+      "learning_rate": 6.970648343163133e-06,
+      "loss": 0.4996,
+      "step": 3394
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.769335739512878,
+      "learning_rate": 6.968937925977158e-06,
+      "loss": 0.5202,
+      "step": 3395
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.2326885589371352,
+      "learning_rate": 6.967227236054244e-06,
+      "loss": 0.4951,
+      "step": 3396
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.754544319304602,
+      "learning_rate": 6.9655162736313535e-06,
+      "loss": 0.4726,
+      "step": 3397
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9703502108468753,
+      "learning_rate": 6.963805038945488e-06,
+      "loss": 0.5987,
+      "step": 3398
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.964420151277497,
+      "learning_rate": 6.962093532233689e-06,
+      "loss": 0.5084,
+      "step": 3399
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7322089983619717,
+      "learning_rate": 6.9603817537330355e-06,
+      "loss": 0.4336,
+      "step": 3400
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.253488241781959,
+      "learning_rate": 6.958669703680639e-06,
+      "loss": 0.4229,
+      "step": 3401
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4461789481107687,
+      "learning_rate": 6.956957382313656e-06,
+      "loss": 0.4456,
+      "step": 3402
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.180205444191373,
+      "learning_rate": 6.955244789869274e-06,
+      "loss": 0.5414,
+      "step": 3403
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.003427563603858,
+      "learning_rate": 6.9535319265847225e-06,
+      "loss": 0.4573,
+      "step": 3404
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1653592258812395,
+      "learning_rate": 6.951818792697267e-06,
+      "loss": 0.5074,
+      "step": 3405
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6388643320473126,
+      "learning_rate": 6.9501053884442106e-06,
+      "loss": 0.5501,
+      "step": 3406
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8880422136748156,
+      "learning_rate": 6.948391714062894e-06,
+      "loss": 0.5518,
+      "step": 3407
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.648228717571414,
+      "learning_rate": 6.946677769790695e-06,
+      "loss": 0.5468,
+      "step": 3408
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9643306629666895,
+      "learning_rate": 6.944963555865028e-06,
+      "loss": 0.439,
+      "step": 3409
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2701453517850134,
+      "learning_rate": 6.943249072523344e-06,
+      "loss": 0.4602,
+      "step": 3410
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7655151445090895,
+      "learning_rate": 6.941534320003139e-06,
+      "loss": 0.4444,
+      "step": 3411
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1631394784188576,
+      "learning_rate": 6.939819298541932e-06,
+      "loss": 0.5443,
+      "step": 3412
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7957351029559017,
+      "learning_rate": 6.9381040083772946e-06,
+      "loss": 0.5593,
+      "step": 3413
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.144413007829309,
+      "learning_rate": 6.936388449746825e-06,
+      "loss": 0.4586,
+      "step": 3414
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.118253158555524,
+      "learning_rate": 6.934672622888163e-06,
+      "loss": 0.4744,
+      "step": 3415
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.098679053046805,
+      "learning_rate": 6.932956528038984e-06,
+      "loss": 0.5025,
+      "step": 3416
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.026012934816335,
+      "learning_rate": 6.931240165437002e-06,
+      "loss": 0.4754,
+      "step": 3417
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1279189545145365,
+      "learning_rate": 6.9295235353199665e-06,
+      "loss": 0.5834,
+      "step": 3418
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9915211969981903,
+      "learning_rate": 6.927806637925665e-06,
+      "loss": 0.6684,
+      "step": 3419
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.808153379635157,
+      "learning_rate": 6.926089473491923e-06,
+      "loss": 0.6168,
+      "step": 3420
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0402742214787764,
+      "learning_rate": 6.924372042256599e-06,
+      "loss": 0.6283,
+      "step": 3421
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2188721858116653,
+      "learning_rate": 6.922654344457594e-06,
+      "loss": 0.417,
+      "step": 3422
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.502934246177953,
+      "learning_rate": 6.920936380332841e-06,
+      "loss": 0.5252,
+      "step": 3423
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.003982033170949,
+      "learning_rate": 6.919218150120315e-06,
+      "loss": 0.4793,
+      "step": 3424
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5994077354030196,
+      "learning_rate": 6.917499654058023e-06,
+      "loss": 0.5034,
+      "step": 3425
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4941298696722516,
+      "learning_rate": 6.91578089238401e-06,
+      "loss": 0.4829,
+      "step": 3426
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.154963348667239,
+      "learning_rate": 6.91406186533636e-06,
+      "loss": 0.5715,
+      "step": 3427
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5455535674718277,
+      "learning_rate": 6.912342573153193e-06,
+      "loss": 0.5363,
+      "step": 3428
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9893435432691113,
+      "learning_rate": 6.910623016072662e-06,
+      "loss": 0.5044,
+      "step": 3429
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.916253278679359,
+      "learning_rate": 6.908903194332963e-06,
+      "loss": 0.4277,
+      "step": 3430
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9955634076029332,
+      "learning_rate": 6.907183108172324e-06,
+      "loss": 0.5436,
+      "step": 3431
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2158499874449897,
+      "learning_rate": 6.905462757829011e-06,
+      "loss": 0.475,
+      "step": 3432
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.178307891950237,
+      "learning_rate": 6.9037421435413275e-06,
+      "loss": 0.4939,
+      "step": 3433
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.331011629950725,
+      "learning_rate": 6.9020212655476116e-06,
+      "loss": 0.5127,
+      "step": 3434
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.553487253834424,
+      "learning_rate": 6.900300124086239e-06,
+      "loss": 0.5205,
+      "step": 3435
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8349367637919298,
+      "learning_rate": 6.898578719395622e-06,
+      "loss": 0.4996,
+      "step": 3436
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7960297236031453,
+      "learning_rate": 6.896857051714211e-06,
+      "loss": 0.5007,
+      "step": 3437
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7197818905948623,
+      "learning_rate": 6.895135121280488e-06,
+      "loss": 0.4412,
+      "step": 3438
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.297819995590865,
+      "learning_rate": 6.893412928332979e-06,
+      "loss": 0.4725,
+      "step": 3439
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0985980814828284,
+      "learning_rate": 6.891690473110237e-06,
+      "loss": 0.5328,
+      "step": 3440
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.483579898201814,
+      "learning_rate": 6.889967755850858e-06,
+      "loss": 0.4426,
+      "step": 3441
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.643349768301987,
+      "learning_rate": 6.888244776793474e-06,
+      "loss": 0.5251,
+      "step": 3442
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8560864552527307,
+      "learning_rate": 6.8865215361767515e-06,
+      "loss": 0.4695,
+      "step": 3443
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.136099014732434,
+      "learning_rate": 6.88479803423939e-06,
+      "loss": 0.4778,
+      "step": 3444
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1344014437579646,
+      "learning_rate": 6.883074271220133e-06,
+      "loss": 0.5328,
+      "step": 3445
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4875853374656764,
+      "learning_rate": 6.881350247357753e-06,
+      "loss": 0.5009,
+      "step": 3446
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8230873513323729,
+      "learning_rate": 6.8796259628910635e-06,
+      "loss": 0.3822,
+      "step": 3447
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.144060516070587,
+      "learning_rate": 6.87790141805891e-06,
+      "loss": 0.5541,
+      "step": 3448
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.004598245836152,
+      "learning_rate": 6.8761766131001795e-06,
+      "loss": 0.4142,
+      "step": 3449
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.095755636348988,
+      "learning_rate": 6.874451548253788e-06,
+      "loss": 0.5364,
+      "step": 3450
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0105263217559926,
+      "learning_rate": 6.872726223758692e-06,
+      "loss": 0.4558,
+      "step": 3451
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9894019796799127,
+      "learning_rate": 6.871000639853886e-06,
+      "loss": 0.4566,
+      "step": 3452
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9206441371536123,
+      "learning_rate": 6.869274796778394e-06,
+      "loss": 0.5627,
+      "step": 3453
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9996733480670144,
+      "learning_rate": 6.86754869477128e-06,
+      "loss": 0.5254,
+      "step": 3454
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4945799916669955,
+      "learning_rate": 6.865822334071646e-06,
+      "loss": 0.4843,
+      "step": 3455
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.603382215490561,
+      "learning_rate": 6.864095714918624e-06,
+      "loss": 0.574,
+      "step": 3456
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.913675786018516,
+      "learning_rate": 6.862368837551387e-06,
+      "loss": 0.5318,
+      "step": 3457
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0742804757686466,
+      "learning_rate": 6.860641702209142e-06,
+      "loss": 0.4984,
+      "step": 3458
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0353713704258203,
+      "learning_rate": 6.858914309131131e-06,
+      "loss": 0.5378,
+      "step": 3459
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.526278264429891,
+      "learning_rate": 6.85718665855663e-06,
+      "loss": 0.5241,
+      "step": 3460
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.805414797561438,
+      "learning_rate": 6.8554587507249555e-06,
+      "loss": 0.5362,
+      "step": 3461
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.262866599877029,
+      "learning_rate": 6.853730585875458e-06,
+      "loss": 0.4453,
+      "step": 3462
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.297052038857309,
+      "learning_rate": 6.852002164247519e-06,
+      "loss": 0.4629,
+      "step": 3463
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.429261348960382,
+      "learning_rate": 6.8502734860805605e-06,
+      "loss": 0.5512,
+      "step": 3464
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0961285978649347,
+      "learning_rate": 6.8485445516140405e-06,
+      "loss": 0.5205,
+      "step": 3465
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4432404905297975,
+      "learning_rate": 6.846815361087449e-06,
+      "loss": 0.5336,
+      "step": 3466
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.489894439252578,
+      "learning_rate": 6.845085914740314e-06,
+      "loss": 0.5266,
+      "step": 3467
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.858592888373152,
+      "learning_rate": 6.8433562128121966e-06,
+      "loss": 0.4932,
+      "step": 3468
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.341560874073902,
+      "learning_rate": 6.841626255542696e-06,
+      "loss": 0.4736,
+      "step": 3469
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6685942199324213,
+      "learning_rate": 6.839896043171446e-06,
+      "loss": 0.452,
+      "step": 3470
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8714172160577403,
+      "learning_rate": 6.838165575938114e-06,
+      "loss": 0.523,
+      "step": 3471
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.223435143902185,
+      "learning_rate": 6.836434854082405e-06,
+      "loss": 0.5385,
+      "step": 3472
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9897231824328618,
+      "learning_rate": 6.8347038778440585e-06,
+      "loss": 0.6362,
+      "step": 3473
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.035006947352832,
+      "learning_rate": 6.83297264746285e-06,
+      "loss": 0.5629,
+      "step": 3474
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.307760585328449,
+      "learning_rate": 6.831241163178586e-06,
+      "loss": 0.4724,
+      "step": 3475
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1444844048934115,
+      "learning_rate": 6.829509425231113e-06,
+      "loss": 0.5549,
+      "step": 3476
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.069914275074382,
+      "learning_rate": 6.827777433860312e-06,
+      "loss": 0.4925,
+      "step": 3477
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1854231243681617,
+      "learning_rate": 6.826045189306099e-06,
+      "loss": 0.5461,
+      "step": 3478
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1559813400174828,
+      "learning_rate": 6.8243126918084205e-06,
+      "loss": 0.5245,
+      "step": 3479
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.632282099536249,
+      "learning_rate": 6.822579941607264e-06,
+      "loss": 0.495,
+      "step": 3480
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3423916719386395,
+      "learning_rate": 6.82084693894265e-06,
+      "loss": 0.4856,
+      "step": 3481
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.912904484087553,
+      "learning_rate": 6.819113684054634e-06,
+      "loss": 0.5006,
+      "step": 3482
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4091797638086523,
+      "learning_rate": 6.817380177183306e-06,
+      "loss": 0.5074,
+      "step": 3483
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9785965166079251,
+      "learning_rate": 6.815646418568789e-06,
+      "loss": 0.4868,
+      "step": 3484
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.048744995881688,
+      "learning_rate": 6.813912408451247e-06,
+      "loss": 0.5939,
+      "step": 3485
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0911700340490222,
+      "learning_rate": 6.812178147070869e-06,
+      "loss": 0.5218,
+      "step": 3486
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2747555308835885,
+      "learning_rate": 6.81044363466789e-06,
+      "loss": 0.4465,
+      "step": 3487
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.094915517223537,
+      "learning_rate": 6.808708871482572e-06,
+      "loss": 0.5002,
+      "step": 3488
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8906848722910898,
+      "learning_rate": 6.806973857755214e-06,
+      "loss": 0.4358,
+      "step": 3489
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3251912395624648,
+      "learning_rate": 6.805238593726151e-06,
+      "loss": 0.4842,
+      "step": 3490
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.073795446509601,
+      "learning_rate": 6.803503079635752e-06,
+      "loss": 0.5359,
+      "step": 3491
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1740849809324923,
+      "learning_rate": 6.8017673157244156e-06,
+      "loss": 0.5325,
+      "step": 3492
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6743320270660744,
+      "learning_rate": 6.800031302232584e-06,
+      "loss": 0.619,
+      "step": 3493
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.052210682658103,
+      "learning_rate": 6.798295039400729e-06,
+      "loss": 0.5002,
+      "step": 3494
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1427699427585734,
+      "learning_rate": 6.796558527469355e-06,
+      "loss": 0.4844,
+      "step": 3495
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7137416161296128,
+      "learning_rate": 6.794821766679006e-06,
+      "loss": 0.4557,
+      "step": 3496
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6209310768289507,
+      "learning_rate": 6.793084757270256e-06,
+      "loss": 0.4675,
+      "step": 3497
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9501731459906053,
+      "learning_rate": 6.791347499483717e-06,
+      "loss": 0.4679,
+      "step": 3498
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6582593174997475,
+      "learning_rate": 6.789609993560032e-06,
+      "loss": 0.5096,
+      "step": 3499
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.193867591080037,
+      "learning_rate": 6.787872239739882e-06,
+      "loss": 0.4189,
+      "step": 3500
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2822910212035423,
+      "learning_rate": 6.786134238263977e-06,
+      "loss": 0.5389,
+      "step": 3501
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6095517419297294,
+      "learning_rate": 6.784395989373068e-06,
+      "loss": 0.4166,
+      "step": 3502
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.347207977992443,
+      "learning_rate": 6.782657493307936e-06,
+      "loss": 0.5289,
+      "step": 3503
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8859078852986118,
+      "learning_rate": 6.780918750309395e-06,
+      "loss": 0.5674,
+      "step": 3504
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9071824286244587,
+      "learning_rate": 6.7791797606183e-06,
+      "loss": 0.5492,
+      "step": 3505
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.134241113572359,
+      "learning_rate": 6.77744052447553e-06,
+      "loss": 0.4575,
+      "step": 3506
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7686059475851854,
+      "learning_rate": 6.775701042122007e-06,
+      "loss": 0.5404,
+      "step": 3507
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5706955363113437,
+      "learning_rate": 6.773961313798685e-06,
+      "loss": 0.3996,
+      "step": 3508
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.626636875653107,
+      "learning_rate": 6.7722213397465475e-06,
+      "loss": 0.5042,
+      "step": 3509
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.583941264270841,
+      "learning_rate": 6.770481120206617e-06,
+      "loss": 0.445,
+      "step": 3510
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8173847825731113,
+      "learning_rate": 6.768740655419949e-06,
+      "loss": 0.5191,
+      "step": 3511
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0672022005145423,
+      "learning_rate": 6.76699994562763e-06,
+      "loss": 0.4363,
+      "step": 3512
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1977340981728313,
+      "learning_rate": 6.765258991070787e-06,
+      "loss": 0.5165,
+      "step": 3513
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6477099652811495,
+      "learning_rate": 6.763517791990572e-06,
+      "loss": 0.472,
+      "step": 3514
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0050808588529097,
+      "learning_rate": 6.7617763486281795e-06,
+      "loss": 0.5548,
+      "step": 3515
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9639114967009688,
+      "learning_rate": 6.760034661224831e-06,
+      "loss": 0.5329,
+      "step": 3516
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8258750637181461,
+      "learning_rate": 6.758292730021788e-06,
+      "loss": 0.6946,
+      "step": 3517
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.889758770682578,
+      "learning_rate": 6.756550555260339e-06,
+      "loss": 0.5132,
+      "step": 3518
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.9612032181346875,
+      "learning_rate": 6.754808137181812e-06,
+      "loss": 0.5496,
+      "step": 3519
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.29602043113293,
+      "learning_rate": 6.753065476027566e-06,
+      "loss": 0.4564,
+      "step": 3520
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.009534383224343,
+      "learning_rate": 6.751322572038993e-06,
+      "loss": 0.5015,
+      "step": 3521
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8233005227356933,
+      "learning_rate": 6.749579425457522e-06,
+      "loss": 0.5085,
+      "step": 3522
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.102848044348859,
+      "learning_rate": 6.7478360365246106e-06,
+      "loss": 0.4057,
+      "step": 3523
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4927043334003782,
+      "learning_rate": 6.746092405481756e-06,
+      "loss": 0.4327,
+      "step": 3524
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5182206924926644,
+      "learning_rate": 6.744348532570482e-06,
+      "loss": 0.5902,
+      "step": 3525
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.630675180211426,
+      "learning_rate": 6.742604418032353e-06,
+      "loss": 0.552,
+      "step": 3526
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1684188275145346,
+      "learning_rate": 6.74086006210896e-06,
+      "loss": 0.562,
+      "step": 3527
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.0464131872586,
+      "learning_rate": 6.739115465041934e-06,
+      "loss": 0.4236,
+      "step": 3528
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8377014985372924,
+      "learning_rate": 6.737370627072934e-06,
+      "loss": 0.411,
+      "step": 3529
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9744618016684727,
+      "learning_rate": 6.735625548443656e-06,
+      "loss": 0.4961,
+      "step": 3530
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.033214852066662,
+      "learning_rate": 6.733880229395828e-06,
+      "loss": 0.5244,
+      "step": 3531
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6204270685576323,
+      "learning_rate": 6.732134670171211e-06,
+      "loss": 0.4187,
+      "step": 3532
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.914051754985948,
+      "learning_rate": 6.730388871011601e-06,
+      "loss": 0.5338,
+      "step": 3533
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6475926085817167,
+      "learning_rate": 6.728642832158823e-06,
+      "loss": 0.578,
+      "step": 3534
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.762647803962594,
+      "learning_rate": 6.726896553854738e-06,
+      "loss": 0.5096,
+      "step": 3535
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8772609053969875,
+      "learning_rate": 6.7251500363412425e-06,
+      "loss": 0.7098,
+      "step": 3536
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0205210120177712,
+      "learning_rate": 6.723403279860262e-06,
+      "loss": 0.5067,
+      "step": 3537
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2546565503696363,
+      "learning_rate": 6.7216562846537584e-06,
+      "loss": 0.501,
+      "step": 3538
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.10002796192174,
+      "learning_rate": 6.719909050963725e-06,
+      "loss": 0.4987,
+      "step": 3539
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.874818537052079,
+      "learning_rate": 6.718161579032186e-06,
+      "loss": 0.5319,
+      "step": 3540
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.916160096548118,
+      "learning_rate": 6.7164138691012035e-06,
+      "loss": 0.5126,
+      "step": 3541
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9798171070336796,
+      "learning_rate": 6.714665921412871e-06,
+      "loss": 0.4996,
+      "step": 3542
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9780709813397543,
+      "learning_rate": 6.71291773620931e-06,
+      "loss": 0.4709,
+      "step": 3543
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0919576107326283,
+      "learning_rate": 6.711169313732682e-06,
+      "loss": 0.6401,
+      "step": 3544
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2922852903103075,
+      "learning_rate": 6.709420654225176e-06,
+      "loss": 0.5517,
+      "step": 3545
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8539758846455174,
+      "learning_rate": 6.707671757929017e-06,
+      "loss": 0.5938,
+      "step": 3546
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4354653405420237,
+      "learning_rate": 6.705922625086464e-06,
+      "loss": 0.5311,
+      "step": 3547
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0892107892444653,
+      "learning_rate": 6.704173255939802e-06,
+      "loss": 0.5119,
+      "step": 3548
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.097176300931407,
+      "learning_rate": 6.702423650731357e-06,
+      "loss": 0.5103,
+      "step": 3549
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.017261819744391,
+      "learning_rate": 6.700673809703483e-06,
+      "loss": 0.4445,
+      "step": 3550
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0455477324112703,
+      "learning_rate": 6.698923733098567e-06,
+      "loss": 0.5001,
+      "step": 3551
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.143977955371404,
+      "learning_rate": 6.697173421159029e-06,
+      "loss": 0.5956,
+      "step": 3552
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.907655264703219,
+      "learning_rate": 6.695422874127323e-06,
+      "loss": 0.4919,
+      "step": 3553
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9116452802190234,
+      "learning_rate": 6.693672092245934e-06,
+      "loss": 0.4434,
+      "step": 3554
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8767790016093753,
+      "learning_rate": 6.69192107575738e-06,
+      "loss": 0.5277,
+      "step": 3555
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.2760262127698647,
+      "learning_rate": 6.6901698249042125e-06,
+      "loss": 0.4357,
+      "step": 3556
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.806160048395752,
+      "learning_rate": 6.688418339929013e-06,
+      "loss": 0.5638,
+      "step": 3557
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9132916256769812,
+      "learning_rate": 6.686666621074398e-06,
+      "loss": 0.5239,
+      "step": 3558
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.294330487449972,
+      "learning_rate": 6.684914668583016e-06,
+      "loss": 0.4701,
+      "step": 3559
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.104014515987271,
+      "learning_rate": 6.683162482697544e-06,
+      "loss": 0.5176,
+      "step": 3560
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.124781326346334,
+      "learning_rate": 6.681410063660696e-06,
+      "loss": 0.4375,
+      "step": 3561
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8841532214368404,
+      "learning_rate": 6.679657411715218e-06,
+      "loss": 0.4961,
+      "step": 3562
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3435669306674405,
+      "learning_rate": 6.677904527103887e-06,
+      "loss": 0.4795,
+      "step": 3563
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9887401121875197,
+      "learning_rate": 6.67615141006951e-06,
+      "loss": 0.4865,
+      "step": 3564
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.051510051685662,
+      "learning_rate": 6.674398060854931e-06,
+      "loss": 0.5532,
+      "step": 3565
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0650627174107883,
+      "learning_rate": 6.6726444797030225e-06,
+      "loss": 0.5987,
+      "step": 3566
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.009636637490451,
+      "learning_rate": 6.6708906668566906e-06,
+      "loss": 0.4522,
+      "step": 3567
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0018100801911562,
+      "learning_rate": 6.669136622558873e-06,
+      "loss": 0.4932,
+      "step": 3568
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.010070154411886,
+      "learning_rate": 6.667382347052539e-06,
+      "loss": 0.5174,
+      "step": 3569
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5741815974182063,
+      "learning_rate": 6.665627840580693e-06,
+      "loss": 0.4748,
+      "step": 3570
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1475107559208446,
+      "learning_rate": 6.663873103386365e-06,
+      "loss": 0.534,
+      "step": 3571
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.040702540317953,
+      "learning_rate": 6.662118135712623e-06,
+      "loss": 0.6129,
+      "step": 3572
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8581917288236094,
+      "learning_rate": 6.660362937802565e-06,
+      "loss": 0.544,
+      "step": 3573
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.095329157755988,
+      "learning_rate": 6.6586075098993196e-06,
+      "loss": 0.4411,
+      "step": 3574
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.960323019417156,
+      "learning_rate": 6.65685185224605e-06,
+      "loss": 0.5614,
+      "step": 3575
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.397698994196487,
+      "learning_rate": 6.655095965085949e-06,
+      "loss": 0.4431,
+      "step": 3576
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8816801739529165,
+      "learning_rate": 6.6533398486622426e-06,
+      "loss": 0.5202,
+      "step": 3577
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.693662545886062,
+      "learning_rate": 6.6515835032181855e-06,
+      "loss": 0.4835,
+      "step": 3578
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0498649909178264,
+      "learning_rate": 6.649826928997068e-06,
+      "loss": 0.5628,
+      "step": 3579
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.005088661384491,
+      "learning_rate": 6.648070126242208e-06,
+      "loss": 0.5108,
+      "step": 3580
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4096309111837146,
+      "learning_rate": 6.646313095196961e-06,
+      "loss": 0.4768,
+      "step": 3581
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.86545807768766,
+      "learning_rate": 6.64455583610471e-06,
+      "loss": 0.4831,
+      "step": 3582
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4341947955179664,
+      "learning_rate": 6.642798349208869e-06,
+      "loss": 0.5218,
+      "step": 3583
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0052273143841948,
+      "learning_rate": 6.6410406347528855e-06,
+      "loss": 0.5088,
+      "step": 3584
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.280938669416697,
+      "learning_rate": 6.639282692980238e-06,
+      "loss": 0.5108,
+      "step": 3585
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.115208642837691,
+      "learning_rate": 6.637524524134434e-06,
+      "loss": 0.5156,
+      "step": 3586
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7269747037326932,
+      "learning_rate": 6.635766128459018e-06,
+      "loss": 0.5422,
+      "step": 3587
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9764975785024523,
+      "learning_rate": 6.63400750619756e-06,
+      "loss": 0.4564,
+      "step": 3588
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2619094645667053,
+      "learning_rate": 6.632248657593667e-06,
+      "loss": 0.516,
+      "step": 3589
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2670843596428365,
+      "learning_rate": 6.630489582890971e-06,
+      "loss": 0.5359,
+      "step": 3590
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6387092591940882,
+      "learning_rate": 6.6287302823331416e-06,
+      "loss": 0.4776,
+      "step": 3591
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.914563819949806,
+      "learning_rate": 6.626970756163875e-06,
+      "loss": 0.5411,
+      "step": 3592
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.099125446353,
+      "learning_rate": 6.625211004626901e-06,
+      "loss": 0.5732,
+      "step": 3593
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0266478571382875,
+      "learning_rate": 6.623451027965981e-06,
+      "loss": 0.5713,
+      "step": 3594
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9074625196503217,
+      "learning_rate": 6.621690826424905e-06,
+      "loss": 0.7419,
+      "step": 3595
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6930001157900096,
+      "learning_rate": 6.619930400247496e-06,
+      "loss": 0.5184,
+      "step": 3596
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.403149628756053,
+      "learning_rate": 6.6181697496776084e-06,
+      "loss": 0.5476,
+      "step": 3597
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6753464077936635,
+      "learning_rate": 6.616408874959128e-06,
+      "loss": 0.4927,
+      "step": 3598
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.011851638293724,
+      "learning_rate": 6.61464777633597e-06,
+      "loss": 0.4882,
+      "step": 3599
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8113012556320098,
+      "learning_rate": 6.612886454052082e-06,
+      "loss": 0.482,
+      "step": 3600
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0067079901524343,
+      "learning_rate": 6.611124908351443e-06,
+      "loss": 0.5304,
+      "step": 3601
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.949846810206903,
+      "learning_rate": 6.609363139478059e-06,
+      "loss": 0.5669,
+      "step": 3602
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9063583331264744,
+      "learning_rate": 6.607601147675973e-06,
+      "loss": 0.5319,
+      "step": 3603
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.373132647372629,
+      "learning_rate": 6.605838933189253e-06,
+      "loss": 0.4831,
+      "step": 3604
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.94091936638512,
+      "learning_rate": 6.604076496262002e-06,
+      "loss": 0.5384,
+      "step": 3605
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1061532958443396,
+      "learning_rate": 6.602313837138353e-06,
+      "loss": 0.3879,
+      "step": 3606
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8652323968864297,
+      "learning_rate": 6.600550956062469e-06,
+      "loss": 0.4835,
+      "step": 3607
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8379089410070378,
+      "learning_rate": 6.598787853278544e-06,
+      "loss": 0.726,
+      "step": 3608
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.346441402606797,
+      "learning_rate": 6.597024529030803e-06,
+      "loss": 0.4963,
+      "step": 3609
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5690653496094755,
+      "learning_rate": 6.5952609835635e-06,
+      "loss": 0.5655,
+      "step": 3610
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.150094313612813,
+      "learning_rate": 6.5934972171209224e-06,
+      "loss": 0.4786,
+      "step": 3611
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.751278416364143,
+      "learning_rate": 6.591733229947387e-06,
+      "loss": 0.5887,
+      "step": 3612
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9687495721572215,
+      "learning_rate": 6.589969022287239e-06,
+      "loss": 0.5172,
+      "step": 3613
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7858800377852178,
+      "learning_rate": 6.588204594384857e-06,
+      "loss": 0.5255,
+      "step": 3614
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.61345197549771,
+      "learning_rate": 6.586439946484651e-06,
+      "loss": 0.4937,
+      "step": 3615
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4301578290487353,
+      "learning_rate": 6.584675078831057e-06,
+      "loss": 0.4192,
+      "step": 3616
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0287578369206996,
+      "learning_rate": 6.582909991668547e-06,
+      "loss": 0.5568,
+      "step": 3617
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5613951474624528,
+      "learning_rate": 6.581144685241619e-06,
+      "loss": 0.4675,
+      "step": 3618
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9929696525055807,
+      "learning_rate": 6.579379159794802e-06,
+      "loss": 0.5019,
+      "step": 3619
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.795420442194661,
+      "learning_rate": 6.577613415572658e-06,
+      "loss": 0.4498,
+      "step": 3620
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2212609121392903,
+      "learning_rate": 6.575847452819777e-06,
+      "loss": 0.5215,
+      "step": 3621
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0003312678826775,
+      "learning_rate": 6.574081271780779e-06,
+      "loss": 0.5995,
+      "step": 3622
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9922345024127028,
+      "learning_rate": 6.572314872700316e-06,
+      "loss": 0.5012,
+      "step": 3623
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.446742448296675,
+      "learning_rate": 6.570548255823071e-06,
+      "loss": 0.5125,
+      "step": 3624
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0341673993421847,
+      "learning_rate": 6.568781421393751e-06,
+      "loss": 0.604,
+      "step": 3625
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.063986166195195,
+      "learning_rate": 6.567014369657102e-06,
+      "loss": 0.4717,
+      "step": 3626
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2571993278317826,
+      "learning_rate": 6.565247100857893e-06,
+      "loss": 0.5973,
+      "step": 3627
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.698173396942107,
+      "learning_rate": 6.563479615240928e-06,
+      "loss": 0.5376,
+      "step": 3628
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4394806105958624,
+      "learning_rate": 6.561711913051037e-06,
+      "loss": 0.5523,
+      "step": 3629
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.8721079208835625,
+      "learning_rate": 6.55994399453308e-06,
+      "loss": 0.4127,
+      "step": 3630
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9283098952033346,
+      "learning_rate": 6.558175859931953e-06,
+      "loss": 0.4671,
+      "step": 3631
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.174395118184736,
+      "learning_rate": 6.556407509492577e-06,
+      "loss": 0.5461,
+      "step": 3632
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.096482074619806,
+      "learning_rate": 6.554638943459901e-06,
+      "loss": 0.5073,
+      "step": 3633
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9572624035740518,
+      "learning_rate": 6.552870162078908e-06,
+      "loss": 0.7498,
+      "step": 3634
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9455719961594937,
+      "learning_rate": 6.55110116559461e-06,
+      "loss": 0.5788,
+      "step": 3635
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5758310580556256,
+      "learning_rate": 6.5493319542520455e-06,
+      "loss": 0.4695,
+      "step": 3636
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8744509713540964,
+      "learning_rate": 6.547562528296287e-06,
+      "loss": 0.598,
+      "step": 3637
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5164898336705903,
+      "learning_rate": 6.545792887972436e-06,
+      "loss": 0.565,
+      "step": 3638
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8707105387309158,
+      "learning_rate": 6.544023033525622e-06,
+      "loss": 0.5342,
+      "step": 3639
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.147616476187572,
+      "learning_rate": 6.542252965201005e-06,
+      "loss": 0.5411,
+      "step": 3640
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.379962658006673,
+      "learning_rate": 6.540482683243774e-06,
+      "loss": 0.5577,
+      "step": 3641
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.672515128821014,
+      "learning_rate": 6.5387121878991475e-06,
+      "loss": 0.4223,
+      "step": 3642
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2887278315237816,
+      "learning_rate": 6.536941479412377e-06,
+      "loss": 0.4323,
+      "step": 3643
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.82695057323536,
+      "learning_rate": 6.535170558028738e-06,
+      "loss": 0.4689,
+      "step": 3644
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.648996029973481,
+      "learning_rate": 6.53339942399354e-06,
+      "loss": 0.4637,
+      "step": 3645
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.821127624252996,
+      "learning_rate": 6.531628077552119e-06,
+      "loss": 0.5525,
+      "step": 3646
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.847745786755322,
+      "learning_rate": 6.5298565189498415e-06,
+      "loss": 0.4647,
+      "step": 3647
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.038774898396566,
+      "learning_rate": 6.528084748432104e-06,
+      "loss": 0.44,
+      "step": 3648
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8695899302635066,
+      "learning_rate": 6.526312766244331e-06,
+      "loss": 0.4647,
+      "step": 3649
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8633588340536484,
+      "learning_rate": 6.52454057263198e-06,
+      "loss": 0.5182,
+      "step": 3650
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.364247034736832,
+      "learning_rate": 6.522768167840532e-06,
+      "loss": 0.5467,
+      "step": 3651
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.9809139324133564,
+      "learning_rate": 6.520995552115502e-06,
+      "loss": 0.6104,
+      "step": 3652
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2045223862495713,
+      "learning_rate": 6.519222725702431e-06,
+      "loss": 0.5866,
+      "step": 3653
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9938558985883759,
+      "learning_rate": 6.517449688846891e-06,
+      "loss": 0.4443,
+      "step": 3654
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8348027498826385,
+      "learning_rate": 6.515676441794483e-06,
+      "loss": 0.434,
+      "step": 3655
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.245963541690037,
+      "learning_rate": 6.513902984790837e-06,
+      "loss": 0.5046,
+      "step": 3656
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.849058738766952,
+      "learning_rate": 6.5121293180816105e-06,
+      "loss": 0.4835,
+      "step": 3657
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0078270602152215,
+      "learning_rate": 6.510355441912493e-06,
+      "loss": 0.5005,
+      "step": 3658
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3606096710817908,
+      "learning_rate": 6.508581356529202e-06,
+      "loss": 0.4694,
+      "step": 3659
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1194421989665924,
+      "learning_rate": 6.5068070621774844e-06,
+      "loss": 0.4459,
+      "step": 3660
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0064534227836988,
+      "learning_rate": 6.5050325591031115e-06,
+      "loss": 0.4768,
+      "step": 3661
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3046584800140186,
+      "learning_rate": 6.5032578475518895e-06,
+      "loss": 0.4392,
+      "step": 3662
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0347630112450745,
+      "learning_rate": 6.501482927769651e-06,
+      "loss": 0.6024,
+      "step": 3663
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9713630177246495,
+      "learning_rate": 6.4997078000022575e-06,
+      "loss": 0.4754,
+      "step": 3664
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9089597407517178,
+      "learning_rate": 6.497932464495599e-06,
+      "loss": 0.5567,
+      "step": 3665
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.908761823603585,
+      "learning_rate": 6.496156921495594e-06,
+      "loss": 0.6325,
+      "step": 3666
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9686657473336888,
+      "learning_rate": 6.494381171248193e-06,
+      "loss": 0.5544,
+      "step": 3667
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.827232493903466,
+      "learning_rate": 6.4926052139993715e-06,
+      "loss": 0.7387,
+      "step": 3668
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9612875277610882,
+      "learning_rate": 6.490829049995133e-06,
+      "loss": 0.5389,
+      "step": 3669
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5884267008321,
+      "learning_rate": 6.489052679481513e-06,
+      "loss": 0.419,
+      "step": 3670
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0533336961851147,
+      "learning_rate": 6.4872761027045735e-06,
+      "loss": 0.539,
+      "step": 3671
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9597899138807566,
+      "learning_rate": 6.485499319910405e-06,
+      "loss": 0.5594,
+      "step": 3672
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0261226240708834,
+      "learning_rate": 6.4837223313451304e-06,
+      "loss": 0.5679,
+      "step": 3673
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9244182240950856,
+      "learning_rate": 6.4819451372548945e-06,
+      "loss": 0.6026,
+      "step": 3674
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9419572010855228,
+      "learning_rate": 6.480167737885874e-06,
+      "loss": 0.759,
+      "step": 3675
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0101465849198905,
+      "learning_rate": 6.478390133484276e-06,
+      "loss": 0.5583,
+      "step": 3676
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0882298301855386,
+      "learning_rate": 6.476612324296332e-06,
+      "loss": 0.4932,
+      "step": 3677
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8358492213126896,
+      "learning_rate": 6.474834310568305e-06,
+      "loss": 0.4862,
+      "step": 3678
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7675586813133983,
+      "learning_rate": 6.473056092546485e-06,
+      "loss": 0.5083,
+      "step": 3679
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0811118086280653,
+      "learning_rate": 6.471277670477189e-06,
+      "loss": 0.5485,
+      "step": 3680
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.938056337446877,
+      "learning_rate": 6.469499044606765e-06,
+      "loss": 0.5168,
+      "step": 3681
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7427343905518542,
+      "learning_rate": 6.467720215181589e-06,
+      "loss": 0.5109,
+      "step": 3682
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.250726862221631,
+      "learning_rate": 6.4659411824480625e-06,
+      "loss": 0.4989,
+      "step": 3683
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.078709845925251,
+      "learning_rate": 6.4641619466526166e-06,
+      "loss": 0.4971,
+      "step": 3684
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.974968026225244,
+      "learning_rate": 6.462382508041714e-06,
+      "loss": 0.5046,
+      "step": 3685
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6572870003968732,
+      "learning_rate": 6.460602866861836e-06,
+      "loss": 0.5854,
+      "step": 3686
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9993132308170822,
+      "learning_rate": 6.458823023359504e-06,
+      "loss": 0.4972,
+      "step": 3687
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2077807862892564,
+      "learning_rate": 6.45704297778126e-06,
+      "loss": 0.5298,
+      "step": 3688
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4557952690465075,
+      "learning_rate": 6.455262730373673e-06,
+      "loss": 0.4567,
+      "step": 3689
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7828039530973063,
+      "learning_rate": 6.453482281383346e-06,
+      "loss": 0.5683,
+      "step": 3690
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0117046421260816,
+      "learning_rate": 6.451701631056905e-06,
+      "loss": 0.5407,
+      "step": 3691
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6056689280607068,
+      "learning_rate": 6.449920779641005e-06,
+      "loss": 0.5058,
+      "step": 3692
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.111117675155408,
+      "learning_rate": 6.4481397273823294e-06,
+      "loss": 0.4929,
+      "step": 3693
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7061020317167797,
+      "learning_rate": 6.446358474527592e-06,
+      "loss": 0.51,
+      "step": 3694
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1274694349672507,
+      "learning_rate": 6.444577021323528e-06,
+      "loss": 0.5032,
+      "step": 3695
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.231329185570803,
+      "learning_rate": 6.442795368016904e-06,
+      "loss": 0.4045,
+      "step": 3696
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.309000293890529,
+      "learning_rate": 6.441013514854517e-06,
+      "loss": 0.5624,
+      "step": 3697
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6826589101233216,
+      "learning_rate": 6.439231462083187e-06,
+      "loss": 0.4769,
+      "step": 3698
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8214224108287267,
+      "learning_rate": 6.437449209949764e-06,
+      "loss": 0.4817,
+      "step": 3699
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8737291048765755,
+      "learning_rate": 6.4356667587011256e-06,
+      "loss": 0.7071,
+      "step": 3700
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.013278207361291,
+      "learning_rate": 6.4338841085841765e-06,
+      "loss": 0.56,
+      "step": 3701
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.830787167445565,
+      "learning_rate": 6.432101259845849e-06,
+      "loss": 0.5148,
+      "step": 3702
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.214672267499787,
+      "learning_rate": 6.430318212733103e-06,
+      "loss": 0.5129,
+      "step": 3703
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.115212816102047,
+      "learning_rate": 6.428534967492926e-06,
+      "loss": 0.4162,
+      "step": 3704
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.903363189742424,
+      "learning_rate": 6.426751524372332e-06,
+      "loss": 0.4495,
+      "step": 3705
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.167483502275897,
+      "learning_rate": 6.4249678836183645e-06,
+      "loss": 0.5339,
+      "step": 3706
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.085569391317056,
+      "learning_rate": 6.423184045478093e-06,
+      "loss": 0.4929,
+      "step": 3707
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8896756208843106,
+      "learning_rate": 6.421400010198613e-06,
+      "loss": 0.4907,
+      "step": 3708
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8934048982976273,
+      "learning_rate": 6.419615778027051e-06,
+      "loss": 0.4848,
+      "step": 3709
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6138143611447546,
+      "learning_rate": 6.417831349210556e-06,
+      "loss": 0.5926,
+      "step": 3710
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.774365432083248,
+      "learning_rate": 6.41604672399631e-06,
+      "loss": 0.4755,
+      "step": 3711
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.379687013158279,
+      "learning_rate": 6.414261902631515e-06,
+      "loss": 0.5975,
+      "step": 3712
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8288929487264844,
+      "learning_rate": 6.412476885363407e-06,
+      "loss": 0.4246,
+      "step": 3713
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9723436058577197,
+      "learning_rate": 6.410691672439246e-06,
+      "loss": 0.5106,
+      "step": 3714
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8311296326695852,
+      "learning_rate": 6.4089062641063175e-06,
+      "loss": 0.4926,
+      "step": 3715
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.737177848721453,
+      "learning_rate": 6.407120660611938e-06,
+      "loss": 0.4994,
+      "step": 3716
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7017674137055212,
+      "learning_rate": 6.40533486220345e-06,
+      "loss": 0.4946,
+      "step": 3717
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.781135263217238,
+      "learning_rate": 6.403548869128218e-06,
+      "loss": 0.4597,
+      "step": 3718
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.017291109682776,
+      "learning_rate": 6.401762681633641e-06,
+      "loss": 0.3983,
+      "step": 3719
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7392309935548593,
+      "learning_rate": 6.39997629996714e-06,
+      "loss": 0.4225,
+      "step": 3720
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.259780252532224,
+      "learning_rate": 6.398189724376165e-06,
+      "loss": 0.6144,
+      "step": 3721
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3282443312515295,
+      "learning_rate": 6.39640295510819e-06,
+      "loss": 0.5528,
+      "step": 3722
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.803462222240754,
+      "learning_rate": 6.39461599241072e-06,
+      "loss": 0.5295,
+      "step": 3723
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9366763207437103,
+      "learning_rate": 6.392828836531284e-06,
+      "loss": 0.4919,
+      "step": 3724
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.185447776765421,
+      "learning_rate": 6.39104148771744e-06,
+      "loss": 0.4678,
+      "step": 3725
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.258509865484344,
+      "learning_rate": 6.389253946216769e-06,
+      "loss": 0.5785,
+      "step": 3726
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.20629079580278,
+      "learning_rate": 6.387466212276882e-06,
+      "loss": 0.5292,
+      "step": 3727
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2954860250216815,
+      "learning_rate": 6.385678286145417e-06,
+      "loss": 0.6209,
+      "step": 3728
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5767189945645126,
+      "learning_rate": 6.383890168070035e-06,
+      "loss": 0.546,
+      "step": 3729
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.149308894538845,
+      "learning_rate": 6.382101858298425e-06,
+      "loss": 0.5018,
+      "step": 3730
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6840931851694483,
+      "learning_rate": 6.380313357078307e-06,
+      "loss": 0.607,
+      "step": 3731
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8772501221990776,
+      "learning_rate": 6.378524664657421e-06,
+      "loss": 0.4405,
+      "step": 3732
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1725162050613873,
+      "learning_rate": 6.376735781283537e-06,
+      "loss": 0.5541,
+      "step": 3733
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8324553652452498,
+      "learning_rate": 6.374946707204452e-06,
+      "loss": 0.4716,
+      "step": 3734
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8863912575458164,
+      "learning_rate": 6.373157442667985e-06,
+      "loss": 0.4478,
+      "step": 3735
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6241930444624073,
+      "learning_rate": 6.37136798792199e-06,
+      "loss": 0.5152,
+      "step": 3736
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8640154340332881,
+      "learning_rate": 6.369578343214337e-06,
+      "loss": 0.6878,
+      "step": 3737
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.864004569774421,
+      "learning_rate": 6.36778850879293e-06,
+      "loss": 0.4185,
+      "step": 3738
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.371126264770795,
+      "learning_rate": 6.3659984849056965e-06,
+      "loss": 0.6347,
+      "step": 3739
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8712028043288276,
+      "learning_rate": 6.3642082718005885e-06,
+      "loss": 0.5452,
+      "step": 3740
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3567129056905043,
+      "learning_rate": 6.362417869725586e-06,
+      "loss": 0.5239,
+      "step": 3741
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.64344631950651,
+      "learning_rate": 6.360627278928697e-06,
+      "loss": 0.6168,
+      "step": 3742
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8133767567419796,
+      "learning_rate": 6.358836499657952e-06,
+      "loss": 0.4792,
+      "step": 3743
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1163142816217655,
+      "learning_rate": 6.357045532161412e-06,
+      "loss": 0.5042,
+      "step": 3744
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0854392069726337,
+      "learning_rate": 6.3552543766871585e-06,
+      "loss": 0.5545,
+      "step": 3745
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8611455570150977,
+      "learning_rate": 6.353463033483305e-06,
+      "loss": 0.4894,
+      "step": 3746
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5649393682942025,
+      "learning_rate": 6.351671502797986e-06,
+      "loss": 0.4545,
+      "step": 3747
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.350993165228112,
+      "learning_rate": 6.349879784879364e-06,
+      "loss": 0.5537,
+      "step": 3748
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.528251394467677,
+      "learning_rate": 6.348087879975627e-06,
+      "loss": 0.5819,
+      "step": 3749
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8639327564428918,
+      "learning_rate": 6.3462957883349915e-06,
+      "loss": 0.5555,
+      "step": 3750
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.015975945110439,
+      "learning_rate": 6.344503510205697e-06,
+      "loss": 0.5727,
+      "step": 3751
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9868670400549004,
+      "learning_rate": 6.342711045836008e-06,
+      "loss": 0.5694,
+      "step": 3752
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7647861416784019,
+      "learning_rate": 6.34091839547422e-06,
+      "loss": 0.5391,
+      "step": 3753
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5520820992026816,
+      "learning_rate": 6.339125559368647e-06,
+      "loss": 0.5516,
+      "step": 3754
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8324978713053337,
+      "learning_rate": 6.337332537767632e-06,
+      "loss": 0.5738,
+      "step": 3755
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8687888923814248,
+      "learning_rate": 6.3355393309195465e-06,
+      "loss": 0.7262,
+      "step": 3756
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0095672102802804,
+      "learning_rate": 6.333745939072784e-06,
+      "loss": 0.4637,
+      "step": 3757
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.173507736904005,
+      "learning_rate": 6.331952362475765e-06,
+      "loss": 0.7386,
+      "step": 3758
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.91568365322517,
+      "learning_rate": 6.3301586013769365e-06,
+      "loss": 0.4145,
+      "step": 3759
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8477133527947616,
+      "learning_rate": 6.328364656024768e-06,
+      "loss": 0.4958,
+      "step": 3760
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7953006514160414,
+      "learning_rate": 6.3265705266677565e-06,
+      "loss": 0.4897,
+      "step": 3761
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9565416312614845,
+      "learning_rate": 6.324776213554428e-06,
+      "loss": 0.4435,
+      "step": 3762
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4445390497054458,
+      "learning_rate": 6.3229817169333266e-06,
+      "loss": 0.6382,
+      "step": 3763
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7488835906128346,
+      "learning_rate": 6.321187037053026e-06,
+      "loss": 0.4781,
+      "step": 3764
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8741922903853565,
+      "learning_rate": 6.319392174162125e-06,
+      "loss": 0.5864,
+      "step": 3765
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2287394702839265,
+      "learning_rate": 6.317597128509251e-06,
+      "loss": 0.5154,
+      "step": 3766
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.391914166356338,
+      "learning_rate": 6.3158019003430495e-06,
+      "loss": 0.5943,
+      "step": 3767
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6914606283809621,
+      "learning_rate": 6.314006489912197e-06,
+      "loss": 0.4725,
+      "step": 3768
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8589368004384728,
+      "learning_rate": 6.3122108974653924e-06,
+      "loss": 0.5432,
+      "step": 3769
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5149163323222976,
+      "learning_rate": 6.310415123251364e-06,
+      "loss": 0.6049,
+      "step": 3770
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7516423580270115,
+      "learning_rate": 6.308619167518858e-06,
+      "loss": 0.5448,
+      "step": 3771
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7385148009610685,
+      "learning_rate": 6.306823030516651e-06,
+      "loss": 0.4671,
+      "step": 3772
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3928864953584217,
+      "learning_rate": 6.305026712493545e-06,
+      "loss": 0.4389,
+      "step": 3773
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8338848402993722,
+      "learning_rate": 6.3032302136983646e-06,
+      "loss": 0.5316,
+      "step": 3774
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.000738083338805,
+      "learning_rate": 6.301433534379961e-06,
+      "loss": 0.5144,
+      "step": 3775
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8774769853039026,
+      "learning_rate": 6.299636674787208e-06,
+      "loss": 0.5199,
+      "step": 3776
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7879066582685983,
+      "learning_rate": 6.297839635169009e-06,
+      "loss": 0.4689,
+      "step": 3777
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.74002426376618,
+      "learning_rate": 6.29604241577429e-06,
+      "loss": 0.4719,
+      "step": 3778
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6138881538542127,
+      "learning_rate": 6.2942450168519996e-06,
+      "loss": 0.5428,
+      "step": 3779
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3746783202444277,
+      "learning_rate": 6.292447438651112e-06,
+      "loss": 0.5243,
+      "step": 3780
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9507871483620898,
+      "learning_rate": 6.290649681420631e-06,
+      "loss": 0.4889,
+      "step": 3781
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8901834225395278,
+      "learning_rate": 6.288851745409578e-06,
+      "loss": 0.4584,
+      "step": 3782
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.583051358633804,
+      "learning_rate": 6.287053630867006e-06,
+      "loss": 0.5043,
+      "step": 3783
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8014333270356484,
+      "learning_rate": 6.285255338041987e-06,
+      "loss": 0.4925,
+      "step": 3784
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8209157504101245,
+      "learning_rate": 6.283456867183622e-06,
+      "loss": 0.5989,
+      "step": 3785
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4362382331503003,
+      "learning_rate": 6.281658218541032e-06,
+      "loss": 0.4379,
+      "step": 3786
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0355770411998937,
+      "learning_rate": 6.27985939236337e-06,
+      "loss": 0.4979,
+      "step": 3787
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.139830492876621,
+      "learning_rate": 6.278060388899805e-06,
+      "loss": 0.3805,
+      "step": 3788
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4295413368290024,
+      "learning_rate": 6.276261208399536e-06,
+      "loss": 0.5213,
+      "step": 3789
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6167326508652153,
+      "learning_rate": 6.274461851111787e-06,
+      "loss": 0.5089,
+      "step": 3790
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7260417984423884,
+      "learning_rate": 6.272662317285802e-06,
+      "loss": 0.3962,
+      "step": 3791
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7017241814747766,
+      "learning_rate": 6.270862607170854e-06,
+      "loss": 0.4707,
+      "step": 3792
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.041661017396624,
+      "learning_rate": 6.269062721016237e-06,
+      "loss": 0.4871,
+      "step": 3793
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.873489150120475,
+      "learning_rate": 6.267262659071273e-06,
+      "loss": 0.4834,
+      "step": 3794
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3659072297258232,
+      "learning_rate": 6.265462421585304e-06,
+      "loss": 0.4675,
+      "step": 3795
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9575804561535535,
+      "learning_rate": 6.2636620088077e-06,
+      "loss": 0.5542,
+      "step": 3796
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5259495034663435,
+      "learning_rate": 6.261861420987853e-06,
+      "loss": 0.5677,
+      "step": 3797
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.3645814052454455,
+      "learning_rate": 6.26006065837518e-06,
+      "loss": 0.5263,
+      "step": 3798
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.275796972363486,
+      "learning_rate": 6.258259721219125e-06,
+      "loss": 0.4225,
+      "step": 3799
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.806674260234892,
+      "learning_rate": 6.2564586097691485e-06,
+      "loss": 0.5176,
+      "step": 3800
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.991372843305997,
+      "learning_rate": 6.2546573242747455e-06,
+      "loss": 0.511,
+      "step": 3801
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.22078520290855,
+      "learning_rate": 6.252855864985425e-06,
+      "loss": 0.5603,
+      "step": 3802
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.55186548266212,
+      "learning_rate": 6.251054232150728e-06,
+      "loss": 0.4304,
+      "step": 3803
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4471899547208684,
+      "learning_rate": 6.249252426020217e-06,
+      "loss": 0.5718,
+      "step": 3804
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6293331107990483,
+      "learning_rate": 6.2474504468434745e-06,
+      "loss": 0.5235,
+      "step": 3805
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9283970961487356,
+      "learning_rate": 6.245648294870112e-06,
+      "loss": 0.5254,
+      "step": 3806
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5080090258814565,
+      "learning_rate": 6.243845970349764e-06,
+      "loss": 0.5455,
+      "step": 3807
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4439307809299704,
+      "learning_rate": 6.242043473532088e-06,
+      "loss": 0.4383,
+      "step": 3808
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9035877540212305,
+      "learning_rate": 6.240240804666765e-06,
+      "loss": 0.4485,
+      "step": 3809
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0610210174697943,
+      "learning_rate": 6.2384379640034994e-06,
+      "loss": 0.4874,
+      "step": 3810
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6562402578770676,
+      "learning_rate": 6.236634951792023e-06,
+      "loss": 0.5006,
+      "step": 3811
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7588003962779382,
+      "learning_rate": 6.234831768282088e-06,
+      "loss": 0.5282,
+      "step": 3812
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2248111168867855,
+      "learning_rate": 6.2330284137234685e-06,
+      "loss": 0.5074,
+      "step": 3813
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7488561137757463,
+      "learning_rate": 6.231224888365968e-06,
+      "loss": 0.4907,
+      "step": 3814
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6829937035558102,
+      "learning_rate": 6.229421192459408e-06,
+      "loss": 0.5216,
+      "step": 3815
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9353687068888177,
+      "learning_rate": 6.227617326253638e-06,
+      "loss": 0.5838,
+      "step": 3816
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6451312426968963,
+      "learning_rate": 6.225813289998528e-06,
+      "loss": 0.5444,
+      "step": 3817
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7162030540676056,
+      "learning_rate": 6.224009083943973e-06,
+      "loss": 0.4303,
+      "step": 3818
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9861443789982134,
+      "learning_rate": 6.222204708339893e-06,
+      "loss": 0.4693,
+      "step": 3819
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9965815139772727,
+      "learning_rate": 6.220400163436228e-06,
+      "loss": 0.4434,
+      "step": 3820
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.305594901880178,
+      "learning_rate": 6.218595449482945e-06,
+      "loss": 0.5039,
+      "step": 3821
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7473432684601102,
+      "learning_rate": 6.21679056673003e-06,
+      "loss": 0.5352,
+      "step": 3822
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.184967971802951,
+      "learning_rate": 6.2149855154274965e-06,
+      "loss": 0.4816,
+      "step": 3823
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8565031989209765,
+      "learning_rate": 6.2131802958253805e-06,
+      "loss": 0.7019,
+      "step": 3824
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9323659347596283,
+      "learning_rate": 6.2113749081737396e-06,
+      "loss": 0.5781,
+      "step": 3825
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2086475064706663,
+      "learning_rate": 6.209569352722657e-06,
+      "loss": 0.5641,
+      "step": 3826
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9465680591827748,
+      "learning_rate": 6.2077636297222355e-06,
+      "loss": 0.5193,
+      "step": 3827
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4454648734132904,
+      "learning_rate": 6.2059577394226056e-06,
+      "loss": 0.4375,
+      "step": 3828
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8245885515136018,
+      "learning_rate": 6.20415168207392e-06,
+      "loss": 0.5405,
+      "step": 3829
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7168392268114387,
+      "learning_rate": 6.202345457926351e-06,
+      "loss": 0.4489,
+      "step": 3830
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6512890497199604,
+      "learning_rate": 6.200539067230097e-06,
+      "loss": 0.6089,
+      "step": 3831
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.928512631271115,
+      "learning_rate": 6.198732510235379e-06,
+      "loss": 0.5198,
+      "step": 3832
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7641150785719257,
+      "learning_rate": 6.196925787192443e-06,
+      "loss": 0.5174,
+      "step": 3833
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5121647751783573,
+      "learning_rate": 6.195118898351553e-06,
+      "loss": 0.5276,
+      "step": 3834
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.273570010598364,
+      "learning_rate": 6.193311843963001e-06,
+      "loss": 0.4879,
+      "step": 3835
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8569612037831926,
+      "learning_rate": 6.191504624277097e-06,
+      "loss": 0.5346,
+      "step": 3836
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.4690857750330486,
+      "learning_rate": 6.1896972395441814e-06,
+      "loss": 0.5943,
+      "step": 3837
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.418822478651311,
+      "learning_rate": 6.187889690014609e-06,
+      "loss": 0.4427,
+      "step": 3838
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9274803187518802,
+      "learning_rate": 6.186081975938763e-06,
+      "loss": 0.7193,
+      "step": 3839
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0402237168644195,
+      "learning_rate": 6.184274097567047e-06,
+      "loss": 0.5503,
+      "step": 3840
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5978603816512145,
+      "learning_rate": 6.1824660551498875e-06,
+      "loss": 0.5296,
+      "step": 3841
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.467422380097626,
+      "learning_rate": 6.1806578489377345e-06,
+      "loss": 0.5457,
+      "step": 3842
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.093518207155721,
+      "learning_rate": 6.178849479181061e-06,
+      "loss": 0.569,
+      "step": 3843
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.84911064856627,
+      "learning_rate": 6.177040946130364e-06,
+      "loss": 0.5102,
+      "step": 3844
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0113320404640076,
+      "learning_rate": 6.175232250036157e-06,
+      "loss": 0.5654,
+      "step": 3845
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2507538000064056,
+      "learning_rate": 6.173423391148983e-06,
+      "loss": 0.5763,
+      "step": 3846
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8660644953352767,
+      "learning_rate": 6.171614369719406e-06,
+      "loss": 0.5064,
+      "step": 3847
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2868547737774976,
+      "learning_rate": 6.169805185998006e-06,
+      "loss": 0.5642,
+      "step": 3848
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.27957553249128,
+      "learning_rate": 6.167995840235396e-06,
+      "loss": 0.5377,
+      "step": 3849
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.475969198433679,
+      "learning_rate": 6.166186332682203e-06,
+      "loss": 0.6373,
+      "step": 3850
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.171813102639267,
+      "learning_rate": 6.164376663589082e-06,
+      "loss": 0.5223,
+      "step": 3851
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4307753912261485,
+      "learning_rate": 6.162566833206707e-06,
+      "loss": 0.5823,
+      "step": 3852
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2576737970165253,
+      "learning_rate": 6.160756841785776e-06,
+      "loss": 0.4477,
+      "step": 3853
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0771923569351647,
+      "learning_rate": 6.15894668957701e-06,
+      "loss": 0.5393,
+      "step": 3854
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7071485325132538,
+      "learning_rate": 6.157136376831147e-06,
+      "loss": 0.4766,
+      "step": 3855
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0059189851166965,
+      "learning_rate": 6.155325903798954e-06,
+      "loss": 0.5366,
+      "step": 3856
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.631687727586758,
+      "learning_rate": 6.1535152707312174e-06,
+      "loss": 0.5519,
+      "step": 3857
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9040434323865552,
+      "learning_rate": 6.151704477878745e-06,
+      "loss": 0.5294,
+      "step": 3858
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.864504990458134,
+      "learning_rate": 6.149893525492368e-06,
+      "loss": 0.5515,
+      "step": 3859
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0313320135666597,
+      "learning_rate": 6.1480824138229375e-06,
+      "loss": 0.5834,
+      "step": 3860
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1947539900172055,
+      "learning_rate": 6.1462711431213315e-06,
+      "loss": 0.4075,
+      "step": 3861
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.394372712067144,
+      "learning_rate": 6.144459713638444e-06,
+      "loss": 0.5111,
+      "step": 3862
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.015999114357917,
+      "learning_rate": 6.142648125625198e-06,
+      "loss": 0.4619,
+      "step": 3863
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.93096483459641,
+      "learning_rate": 6.140836379332529e-06,
+      "loss": 0.4672,
+      "step": 3864
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8856694074513514,
+      "learning_rate": 6.1390244750114036e-06,
+      "loss": 0.4368,
+      "step": 3865
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7103978418204258,
+      "learning_rate": 6.137212412912804e-06,
+      "loss": 0.4154,
+      "step": 3866
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6423033873328343,
+      "learning_rate": 6.1354001932877395e-06,
+      "loss": 0.4603,
+      "step": 3867
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6733956841884265,
+      "learning_rate": 6.133587816387236e-06,
+      "loss": 0.4852,
+      "step": 3868
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.2377017830208623,
+      "learning_rate": 6.131775282462345e-06,
+      "loss": 0.5035,
+      "step": 3869
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.946295426927518,
+      "learning_rate": 6.129962591764137e-06,
+      "loss": 0.5665,
+      "step": 3870
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8598381292596793,
+      "learning_rate": 6.128149744543708e-06,
+      "loss": 0.5054,
+      "step": 3871
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.255009701730728,
+      "learning_rate": 6.12633674105217e-06,
+      "loss": 0.5304,
+      "step": 3872
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.003969247050497,
+      "learning_rate": 6.124523581540662e-06,
+      "loss": 0.4773,
+      "step": 3873
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.993626931815044,
+      "learning_rate": 6.122710266260344e-06,
+      "loss": 0.5387,
+      "step": 3874
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3113266302870112,
+      "learning_rate": 6.120896795462392e-06,
+      "loss": 0.5443,
+      "step": 3875
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0324540793999923,
+      "learning_rate": 6.1190831693980104e-06,
+      "loss": 0.5958,
+      "step": 3876
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8781463257660689,
+      "learning_rate": 6.117269388318423e-06,
+      "loss": 0.4241,
+      "step": 3877
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.766528617889605,
+      "learning_rate": 6.115455452474874e-06,
+      "loss": 0.5326,
+      "step": 3878
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.08696011168214,
+      "learning_rate": 6.113641362118627e-06,
+      "loss": 0.4967,
+      "step": 3879
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.617142562480076,
+      "learning_rate": 6.111827117500974e-06,
+      "loss": 0.6535,
+      "step": 3880
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4597446578833404,
+      "learning_rate": 6.110012718873219e-06,
+      "loss": 0.5694,
+      "step": 3881
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.338425793615208,
+      "learning_rate": 6.1081981664866955e-06,
+      "loss": 0.5326,
+      "step": 3882
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9114017477925207,
+      "learning_rate": 6.106383460592753e-06,
+      "loss": 0.4553,
+      "step": 3883
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.556580985057182,
+      "learning_rate": 6.104568601442765e-06,
+      "loss": 0.5088,
+      "step": 3884
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.033415219717344,
+      "learning_rate": 6.102753589288126e-06,
+      "loss": 0.5349,
+      "step": 3885
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9303393930953598,
+      "learning_rate": 6.1009384243802525e-06,
+      "loss": 0.5533,
+      "step": 3886
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7165561324441685,
+      "learning_rate": 6.099123106970578e-06,
+      "loss": 0.5292,
+      "step": 3887
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8476463618228964,
+      "learning_rate": 6.097307637310561e-06,
+      "loss": 0.4928,
+      "step": 3888
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2064613316135904,
+      "learning_rate": 6.095492015651681e-06,
+      "loss": 0.512,
+      "step": 3889
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7661190702478455,
+      "learning_rate": 6.093676242245435e-06,
+      "loss": 0.5571,
+      "step": 3890
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.52997232569591,
+      "learning_rate": 6.0918603173433465e-06,
+      "loss": 0.454,
+      "step": 3891
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2811725413680155,
+      "learning_rate": 6.0900442411969555e-06,
+      "loss": 0.5974,
+      "step": 3892
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7531346178433234,
+      "learning_rate": 6.088228014057825e-06,
+      "loss": 0.5644,
+      "step": 3893
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8584947561895628,
+      "learning_rate": 6.086411636177538e-06,
+      "loss": 0.4847,
+      "step": 3894
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8414385310343264,
+      "learning_rate": 6.0845951078077004e-06,
+      "loss": 0.5338,
+      "step": 3895
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7514750689193765,
+      "learning_rate": 6.082778429199937e-06,
+      "loss": 0.5172,
+      "step": 3896
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.213959314104493,
+      "learning_rate": 6.0809616006058915e-06,
+      "loss": 0.4999,
+      "step": 3897
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2684304661600034,
+      "learning_rate": 6.079144622277233e-06,
+      "loss": 0.5539,
+      "step": 3898
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.308793211415469,
+      "learning_rate": 6.077327494465648e-06,
+      "loss": 0.4763,
+      "step": 3899
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.542300597424697,
+      "learning_rate": 6.075510217422845e-06,
+      "loss": 0.5467,
+      "step": 3900
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.232535642634879,
+      "learning_rate": 6.073692791400553e-06,
+      "loss": 0.5943,
+      "step": 3901
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9565365197481344,
+      "learning_rate": 6.0718752166505214e-06,
+      "loss": 0.5765,
+      "step": 3902
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9536192156103716,
+      "learning_rate": 6.0700574934245215e-06,
+      "loss": 0.5209,
+      "step": 3903
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.329778383103884,
+      "learning_rate": 6.068239621974341e-06,
+      "loss": 0.5588,
+      "step": 3904
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.1984475232309357,
+      "learning_rate": 6.066421602551796e-06,
+      "loss": 0.5479,
+      "step": 3905
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6636663151806315,
+      "learning_rate": 6.064603435408714e-06,
+      "loss": 0.4796,
+      "step": 3906
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8595379909838388,
+      "learning_rate": 6.062785120796947e-06,
+      "loss": 0.7073,
+      "step": 3907
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8501763659093946,
+      "learning_rate": 6.0609666589683705e-06,
+      "loss": 0.7138,
+      "step": 3908
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8105520953965444,
+      "learning_rate": 6.0591480501748765e-06,
+      "loss": 0.5576,
+      "step": 3909
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.7219410778256607,
+      "learning_rate": 6.057329294668377e-06,
+      "loss": 0.5004,
+      "step": 3910
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0664917917040366,
+      "learning_rate": 6.055510392700807e-06,
+      "loss": 0.5547,
+      "step": 3911
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7630791616942876,
+      "learning_rate": 6.05369134452412e-06,
+      "loss": 0.4017,
+      "step": 3912
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.781440318369817,
+      "learning_rate": 6.051872150390293e-06,
+      "loss": 0.4903,
+      "step": 3913
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0591461032170564,
+      "learning_rate": 6.0500528105513156e-06,
+      "loss": 0.5188,
+      "step": 3914
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0386776643555526,
+      "learning_rate": 6.048233325259205e-06,
+      "loss": 0.5399,
+      "step": 3915
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4872150763238103,
+      "learning_rate": 6.046413694765996e-06,
+      "loss": 0.473,
+      "step": 3916
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8874152573187207,
+      "learning_rate": 6.044593919323742e-06,
+      "loss": 0.7141,
+      "step": 3917
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 13.933831816954205,
+      "learning_rate": 6.04277399918452e-06,
+      "loss": 0.5209,
+      "step": 3918
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0693774456743887,
+      "learning_rate": 6.040953934600425e-06,
+      "loss": 0.4891,
+      "step": 3919
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0545180849312867,
+      "learning_rate": 6.039133725823571e-06,
+      "loss": 0.4749,
+      "step": 3920
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0955142470465686,
+      "learning_rate": 6.037313373106092e-06,
+      "loss": 0.4954,
+      "step": 3921
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7350661425773517,
+      "learning_rate": 6.0354928767001465e-06,
+      "loss": 0.5401,
+      "step": 3922
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7204197049861876,
+      "learning_rate": 6.033672236857906e-06,
+      "loss": 0.5187,
+      "step": 3923
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.245124687924675,
+      "learning_rate": 6.031851453831565e-06,
+      "loss": 0.4353,
+      "step": 3924
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8770269777671564,
+      "learning_rate": 6.03003052787334e-06,
+      "loss": 0.4853,
+      "step": 3925
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.668819467282734,
+      "learning_rate": 6.028209459235466e-06,
+      "loss": 0.4585,
+      "step": 3926
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.3989676984328634,
+      "learning_rate": 6.0263882481701945e-06,
+      "loss": 0.5283,
+      "step": 3927
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.985824212374142,
+      "learning_rate": 6.024566894929801e-06,
+      "loss": 0.4527,
+      "step": 3928
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7689977369208556,
+      "learning_rate": 6.02274539976658e-06,
+      "loss": 0.4077,
+      "step": 3929
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3400806273556594,
+      "learning_rate": 6.020923762932841e-06,
+      "loss": 0.5154,
+      "step": 3930
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.318653649764081,
+      "learning_rate": 6.0191019846809206e-06,
+      "loss": 0.541,
+      "step": 3931
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.18393188501938,
+      "learning_rate": 6.0172800652631706e-06,
+      "loss": 0.6133,
+      "step": 3932
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7199326583339434,
+      "learning_rate": 6.015458004931961e-06,
+      "loss": 0.4393,
+      "step": 3933
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5962708448455993,
+      "learning_rate": 6.013635803939684e-06,
+      "loss": 0.5358,
+      "step": 3934
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2934882179298595,
+      "learning_rate": 6.011813462538752e-06,
+      "loss": 0.4911,
+      "step": 3935
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7604511215592078,
+      "learning_rate": 6.0099909809815925e-06,
+      "loss": 0.3932,
+      "step": 3936
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6621012690784056,
+      "learning_rate": 6.008168359520659e-06,
+      "loss": 0.4924,
+      "step": 3937
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0455475778685943,
+      "learning_rate": 6.006345598408418e-06,
+      "loss": 0.6394,
+      "step": 3938
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.741692364622363,
+      "learning_rate": 6.0045226978973614e-06,
+      "loss": 0.4205,
+      "step": 3939
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1803130784702707,
+      "learning_rate": 6.002699658239992e-06,
+      "loss": 0.4323,
+      "step": 3940
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.067762500213044,
+      "learning_rate": 6.0008764796888406e-06,
+      "loss": 0.5505,
+      "step": 3941
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1648373493138835,
+      "learning_rate": 5.999053162496453e-06,
+      "loss": 0.4269,
+      "step": 3942
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4000649531676244,
+      "learning_rate": 5.997229706915393e-06,
+      "loss": 0.5466,
+      "step": 3943
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9645038829853294,
+      "learning_rate": 5.995406113198248e-06,
+      "loss": 0.4737,
+      "step": 3944
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9625689774836608,
+      "learning_rate": 5.99358238159762e-06,
+      "loss": 0.481,
+      "step": 3945
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1628207279062837,
+      "learning_rate": 5.991758512366133e-06,
+      "loss": 0.5703,
+      "step": 3946
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1069189848738747,
+      "learning_rate": 5.989934505756429e-06,
+      "loss": 0.4677,
+      "step": 3947
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.265099512251831,
+      "learning_rate": 5.988110362021168e-06,
+      "loss": 0.6182,
+      "step": 3948
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.451871978042828,
+      "learning_rate": 5.986286081413031e-06,
+      "loss": 0.5586,
+      "step": 3949
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.328661493946251,
+      "learning_rate": 5.984461664184717e-06,
+      "loss": 0.5488,
+      "step": 3950
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.030697768858632,
+      "learning_rate": 5.982637110588945e-06,
+      "loss": 0.5583,
+      "step": 3951
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.457078791306872,
+      "learning_rate": 5.980812420878448e-06,
+      "loss": 0.4943,
+      "step": 3952
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.932424277482519,
+      "learning_rate": 5.978987595305985e-06,
+      "loss": 0.5248,
+      "step": 3953
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8820036008940983,
+      "learning_rate": 5.977162634124331e-06,
+      "loss": 0.4497,
+      "step": 3954
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4890278749172583,
+      "learning_rate": 5.975337537586278e-06,
+      "loss": 0.5321,
+      "step": 3955
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7744284105135804,
+      "learning_rate": 5.973512305944638e-06,
+      "loss": 0.4956,
+      "step": 3956
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1648847883550255,
+      "learning_rate": 5.971686939452241e-06,
+      "loss": 0.4549,
+      "step": 3957
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.812123623062125,
+      "learning_rate": 5.969861438361939e-06,
+      "loss": 0.4379,
+      "step": 3958
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9994075286171609,
+      "learning_rate": 5.968035802926598e-06,
+      "loss": 0.5259,
+      "step": 3959
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7781725784622178,
+      "learning_rate": 5.966210033399105e-06,
+      "loss": 0.5522,
+      "step": 3960
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0529129444429435,
+      "learning_rate": 5.964384130032366e-06,
+      "loss": 0.4067,
+      "step": 3961
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.328488968250573,
+      "learning_rate": 5.962558093079302e-06,
+      "loss": 0.4639,
+      "step": 3962
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.14096951540091,
+      "learning_rate": 5.960731922792861e-06,
+      "loss": 0.5085,
+      "step": 3963
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9016476470604728,
+      "learning_rate": 5.958905619426e-06,
+      "loss": 0.5246,
+      "step": 3964
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.629847096457984,
+      "learning_rate": 5.957079183231696e-06,
+      "loss": 0.4777,
+      "step": 3965
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7373107022831173,
+      "learning_rate": 5.955252614462952e-06,
+      "loss": 0.4338,
+      "step": 3966
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.046700541707747,
+      "learning_rate": 5.95342591337278e-06,
+      "loss": 0.5425,
+      "step": 3967
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2517165246456528,
+      "learning_rate": 5.951599080214216e-06,
+      "loss": 0.4851,
+      "step": 3968
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9236733675645439,
+      "learning_rate": 5.949772115240312e-06,
+      "loss": 0.5026,
+      "step": 3969
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4237342567495532,
+      "learning_rate": 5.94794501870414e-06,
+      "loss": 0.5535,
+      "step": 3970
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6671587830031098,
+      "learning_rate": 5.946117790858787e-06,
+      "loss": 0.5378,
+      "step": 3971
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8878139302420236,
+      "learning_rate": 5.944290431957362e-06,
+      "loss": 0.5116,
+      "step": 3972
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1318239916070567,
+      "learning_rate": 5.94246294225299e-06,
+      "loss": 0.5205,
+      "step": 3973
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8030828161053167,
+      "learning_rate": 5.940635321998815e-06,
+      "loss": 0.5034,
+      "step": 3974
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.794805107107148,
+      "learning_rate": 5.938807571447998e-06,
+      "loss": 0.5509,
+      "step": 3975
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.355952607966635,
+      "learning_rate": 5.9369796908537185e-06,
+      "loss": 0.5521,
+      "step": 3976
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5096814647070667,
+      "learning_rate": 5.9351516804691745e-06,
+      "loss": 0.5208,
+      "step": 3977
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2171385118497535,
+      "learning_rate": 5.933323540547581e-06,
+      "loss": 0.5488,
+      "step": 3978
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1007539885191306,
+      "learning_rate": 5.931495271342173e-06,
+      "loss": 0.4712,
+      "step": 3979
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.088343984488,
+      "learning_rate": 5.929666873106202e-06,
+      "loss": 0.4461,
+      "step": 3980
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8458501140332395,
+      "learning_rate": 5.927838346092936e-06,
+      "loss": 0.7471,
+      "step": 3981
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.948310419100518,
+      "learning_rate": 5.926009690555663e-06,
+      "loss": 0.6254,
+      "step": 3982
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8894552595928077,
+      "learning_rate": 5.924180906747688e-06,
+      "loss": 0.7081,
+      "step": 3983
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9463515431142673,
+      "learning_rate": 5.922351994922333e-06,
+      "loss": 0.4956,
+      "step": 3984
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1200639428898436,
+      "learning_rate": 5.9205229553329405e-06,
+      "loss": 0.4816,
+      "step": 3985
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5283295313298106,
+      "learning_rate": 5.918693788232868e-06,
+      "loss": 0.5632,
+      "step": 3986
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9344228822113,
+      "learning_rate": 5.91686449387549e-06,
+      "loss": 0.4958,
+      "step": 3987
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.8288477400796794,
+      "learning_rate": 5.915035072514202e-06,
+      "loss": 0.5446,
+      "step": 3988
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.33959458882891,
+      "learning_rate": 5.913205524402415e-06,
+      "loss": 0.5702,
+      "step": 3989
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5232575043546994,
+      "learning_rate": 5.9113758497935545e-06,
+      "loss": 0.4804,
+      "step": 3990
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.860874844124119,
+      "learning_rate": 5.909546048941071e-06,
+      "loss": 0.5178,
+      "step": 3991
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8720428303040475,
+      "learning_rate": 5.907716122098424e-06,
+      "loss": 0.7096,
+      "step": 3992
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.172881785234152,
+      "learning_rate": 5.9058860695191e-06,
+      "loss": 0.4496,
+      "step": 3993
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4980751016234626,
+      "learning_rate": 5.904055891456594e-06,
+      "loss": 0.4602,
+      "step": 3994
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.4277576863210166,
+      "learning_rate": 5.902225588164422e-06,
+      "loss": 0.4867,
+      "step": 3995
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.000118921657069,
+      "learning_rate": 5.900395159896117e-06,
+      "loss": 0.4512,
+      "step": 3996
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8354223754231285,
+      "learning_rate": 5.898564606905231e-06,
+      "loss": 0.5466,
+      "step": 3997
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9862896887531025,
+      "learning_rate": 5.896733929445333e-06,
+      "loss": 0.4749,
+      "step": 3998
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7885771469117557,
+      "learning_rate": 5.894903127770004e-06,
+      "loss": 0.4519,
+      "step": 3999
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.996893956192921,
+      "learning_rate": 5.8930722021328505e-06,
+      "loss": 0.539,
+      "step": 4000
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.40417944425607,
+      "learning_rate": 5.891241152787488e-06,
+      "loss": 0.5154,
+      "step": 4001
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.208862780268627,
+      "learning_rate": 5.889409979987557e-06,
+      "loss": 0.5757,
+      "step": 4002
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0610474524954556,
+      "learning_rate": 5.887578683986709e-06,
+      "loss": 0.5838,
+      "step": 4003
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0019750915908987,
+      "learning_rate": 5.885747265038616e-06,
+      "loss": 0.5483,
+      "step": 4004
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0403085030544705,
+      "learning_rate": 5.883915723396965e-06,
+      "loss": 0.5269,
+      "step": 4005
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.158196451499147,
+      "learning_rate": 5.882084059315461e-06,
+      "loss": 0.5354,
+      "step": 4006
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8841955024468213,
+      "learning_rate": 5.880252273047826e-06,
+      "loss": 0.4876,
+      "step": 4007
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.835501728419276,
+      "learning_rate": 5.878420364847799e-06,
+      "loss": 0.541,
+      "step": 4008
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2182259302627876,
+      "learning_rate": 5.8765883349691345e-06,
+      "loss": 0.4574,
+      "step": 4009
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.319627149811975,
+      "learning_rate": 5.874756183665605e-06,
+      "loss": 0.4995,
+      "step": 4010
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4219634948890376,
+      "learning_rate": 5.872923911191002e-06,
+      "loss": 0.4119,
+      "step": 4011
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1819224646555373,
+      "learning_rate": 5.871091517799129e-06,
+      "loss": 0.5976,
+      "step": 4012
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9772592227569705,
+      "learning_rate": 5.8692590037438105e-06,
+      "loss": 0.4904,
+      "step": 4013
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.044041404223672,
+      "learning_rate": 5.867426369278886e-06,
+      "loss": 0.4443,
+      "step": 4014
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.207962003226654,
+      "learning_rate": 5.865593614658209e-06,
+      "loss": 0.4282,
+      "step": 4015
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1994322056962803,
+      "learning_rate": 5.863760740135657e-06,
+      "loss": 0.5809,
+      "step": 4016
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2629985434405318,
+      "learning_rate": 5.861927745965116e-06,
+      "loss": 0.5015,
+      "step": 4017
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1454385547546098,
+      "learning_rate": 5.860094632400492e-06,
+      "loss": 0.4931,
+      "step": 4018
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.526214868914331,
+      "learning_rate": 5.8582613996957085e-06,
+      "loss": 0.5097,
+      "step": 4019
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1009662785195258,
+      "learning_rate": 5.856428048104706e-06,
+      "loss": 0.5104,
+      "step": 4020
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.459598459411013,
+      "learning_rate": 5.854594577881436e-06,
+      "loss": 0.4262,
+      "step": 4021
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.017125998301964,
+      "learning_rate": 5.852760989279874e-06,
+      "loss": 0.5284,
+      "step": 4022
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9910137659868243,
+      "learning_rate": 5.850927282554009e-06,
+      "loss": 0.5103,
+      "step": 4023
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3089119670440263,
+      "learning_rate": 5.849093457957844e-06,
+      "loss": 0.534,
+      "step": 4024
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4765252239383595,
+      "learning_rate": 5.8472595157454e-06,
+      "loss": 0.5392,
+      "step": 4025
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.058710930789199,
+      "learning_rate": 5.8454254561707135e-06,
+      "loss": 0.5104,
+      "step": 4026
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.89428021363449,
+      "learning_rate": 5.84359127948784e-06,
+      "loss": 0.5429,
+      "step": 4027
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9893551100894353,
+      "learning_rate": 5.8417569859508485e-06,
+      "loss": 0.3806,
+      "step": 4028
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.582631438874301,
+      "learning_rate": 5.839922575813824e-06,
+      "loss": 0.5774,
+      "step": 4029
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0311609250373075,
+      "learning_rate": 5.838088049330871e-06,
+      "loss": 0.4786,
+      "step": 4030
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2956484653685814,
+      "learning_rate": 5.836253406756108e-06,
+      "loss": 0.5011,
+      "step": 4031
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.42925988279104,
+      "learning_rate": 5.834418648343666e-06,
+      "loss": 0.5348,
+      "step": 4032
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7690973930786447,
+      "learning_rate": 5.832583774347697e-06,
+      "loss": 0.4958,
+      "step": 4033
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8680771943906376,
+      "learning_rate": 5.830748785022369e-06,
+      "loss": 0.4211,
+      "step": 4034
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7184966630315233,
+      "learning_rate": 5.828913680621863e-06,
+      "loss": 0.6152,
+      "step": 4035
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.768834612632359,
+      "learning_rate": 5.827078461400378e-06,
+      "loss": 0.5243,
+      "step": 4036
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8511709900862123,
+      "learning_rate": 5.825243127612127e-06,
+      "loss": 0.4675,
+      "step": 4037
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.615776631265324,
+      "learning_rate": 5.8234076795113415e-06,
+      "loss": 0.4936,
+      "step": 4038
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6588891901539942,
+      "learning_rate": 5.821572117352266e-06,
+      "loss": 0.4746,
+      "step": 4039
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6905596017734232,
+      "learning_rate": 5.819736441389164e-06,
+      "loss": 0.4343,
+      "step": 4040
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.049332279341107,
+      "learning_rate": 5.817900651876311e-06,
+      "loss": 0.5241,
+      "step": 4041
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4602133257384171,
+      "learning_rate": 5.816064749068002e-06,
+      "loss": 0.5168,
+      "step": 4042
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1539153478672253,
+      "learning_rate": 5.814228733218546e-06,
+      "loss": 0.5727,
+      "step": 4043
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9788387694478389,
+      "learning_rate": 5.812392604582265e-06,
+      "loss": 0.4259,
+      "step": 4044
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.188851240621248,
+      "learning_rate": 5.810556363413502e-06,
+      "loss": 0.4189,
+      "step": 4045
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8558530111245641,
+      "learning_rate": 5.808720009966613e-06,
+      "loss": 0.5138,
+      "step": 4046
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7967470049887913,
+      "learning_rate": 5.806883544495967e-06,
+      "loss": 0.4036,
+      "step": 4047
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8023200267365835,
+      "learning_rate": 5.805046967255954e-06,
+      "loss": 0.5527,
+      "step": 4048
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8675872581685433,
+      "learning_rate": 5.8032102785009725e-06,
+      "loss": 0.4392,
+      "step": 4049
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3813710263068186,
+      "learning_rate": 5.801373478485443e-06,
+      "loss": 0.5379,
+      "step": 4050
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7058208306681926,
+      "learning_rate": 5.799536567463798e-06,
+      "loss": 0.4979,
+      "step": 4051
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8880369058957132,
+      "learning_rate": 5.797699545690486e-06,
+      "loss": 0.511,
+      "step": 4052
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5543445120770203,
+      "learning_rate": 5.795862413419971e-06,
+      "loss": 0.5442,
+      "step": 4053
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8802568362331055,
+      "learning_rate": 5.794025170906733e-06,
+      "loss": 0.7211,
+      "step": 4054
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8601606667140826,
+      "learning_rate": 5.792187818405265e-06,
+      "loss": 0.498,
+      "step": 4055
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9314646790573065,
+      "learning_rate": 5.7903503561700795e-06,
+      "loss": 0.5596,
+      "step": 4056
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.9196911307073012,
+      "learning_rate": 5.788512784455697e-06,
+      "loss": 0.5733,
+      "step": 4057
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2386965796093476,
+      "learning_rate": 5.786675103516662e-06,
+      "loss": 0.5662,
+      "step": 4058
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.4356402046760577,
+      "learning_rate": 5.784837313607529e-06,
+      "loss": 0.5361,
+      "step": 4059
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.754407074130776,
+      "learning_rate": 5.782999414982865e-06,
+      "loss": 0.45,
+      "step": 4060
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.869665092733014,
+      "learning_rate": 5.781161407897258e-06,
+      "loss": 0.4561,
+      "step": 4061
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8045093924774072,
+      "learning_rate": 5.779323292605308e-06,
+      "loss": 0.3815,
+      "step": 4062
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7541340244559271,
+      "learning_rate": 5.77748506936163e-06,
+      "loss": 0.5327,
+      "step": 4063
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9308103687590457,
+      "learning_rate": 5.775646738420856e-06,
+      "loss": 0.5062,
+      "step": 4064
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1646673828172354,
+      "learning_rate": 5.773808300037631e-06,
+      "loss": 0.5159,
+      "step": 4065
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.038305543022963,
+      "learning_rate": 5.771969754466613e-06,
+      "loss": 0.5797,
+      "step": 4066
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.88308769477512,
+      "learning_rate": 5.7701311019624785e-06,
+      "loss": 0.5065,
+      "step": 4067
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.048867431426452,
+      "learning_rate": 5.7682923427799165e-06,
+      "loss": 0.5306,
+      "step": 4068
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.799904162119711,
+      "learning_rate": 5.766453477173633e-06,
+      "loss": 0.5325,
+      "step": 4069
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.078564736133288,
+      "learning_rate": 5.764614505398346e-06,
+      "loss": 0.5057,
+      "step": 4070
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9888277133899182,
+      "learning_rate": 5.7627754277087896e-06,
+      "loss": 0.5185,
+      "step": 4071
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.18474539742922,
+      "learning_rate": 5.760936244359715e-06,
+      "loss": 0.5304,
+      "step": 4072
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.06505800815184,
+      "learning_rate": 5.7590969556058815e-06,
+      "loss": 0.5176,
+      "step": 4073
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7829550663679807,
+      "learning_rate": 5.757257561702072e-06,
+      "loss": 0.4515,
+      "step": 4074
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5537998741685324,
+      "learning_rate": 5.755418062903074e-06,
+      "loss": 0.5228,
+      "step": 4075
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5822822561539227,
+      "learning_rate": 5.753578459463698e-06,
+      "loss": 0.4262,
+      "step": 4076
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.417564992377457,
+      "learning_rate": 5.751738751638763e-06,
+      "loss": 0.4744,
+      "step": 4077
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5911961687862044,
+      "learning_rate": 5.749898939683107e-06,
+      "loss": 0.4794,
+      "step": 4078
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.131929357369358,
+      "learning_rate": 5.748059023851581e-06,
+      "loss": 0.4574,
+      "step": 4079
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.708163014239714,
+      "learning_rate": 5.746219004399047e-06,
+      "loss": 0.5506,
+      "step": 4080
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.889848653330532,
+      "learning_rate": 5.744378881580386e-06,
+      "loss": 0.4513,
+      "step": 4081
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6772661482636169,
+      "learning_rate": 5.7425386556504915e-06,
+      "loss": 0.493,
+      "step": 4082
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.077401305791897,
+      "learning_rate": 5.740698326864271e-06,
+      "loss": 0.4488,
+      "step": 4083
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.452554381863501,
+      "learning_rate": 5.738857895476646e-06,
+      "loss": 0.5169,
+      "step": 4084
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7162223227451432,
+      "learning_rate": 5.737017361742554e-06,
+      "loss": 0.4311,
+      "step": 4085
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.527455321408591,
+      "learning_rate": 5.735176725916944e-06,
+      "loss": 0.4526,
+      "step": 4086
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5080522300312116,
+      "learning_rate": 5.733335988254782e-06,
+      "loss": 0.5442,
+      "step": 4087
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9202715090747422,
+      "learning_rate": 5.731495149011045e-06,
+      "loss": 0.4996,
+      "step": 4088
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7644955861942204,
+      "learning_rate": 5.729654208440727e-06,
+      "loss": 0.454,
+      "step": 4089
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.595297472188361,
+      "learning_rate": 5.727813166798836e-06,
+      "loss": 0.5285,
+      "step": 4090
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0251416974113816,
+      "learning_rate": 5.7259720243403896e-06,
+      "loss": 0.4626,
+      "step": 4091
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8700811030811115,
+      "learning_rate": 5.724130781320424e-06,
+      "loss": 0.4532,
+      "step": 4092
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0493133875302356,
+      "learning_rate": 5.722289437993989e-06,
+      "loss": 0.6167,
+      "step": 4093
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8185319240855067,
+      "learning_rate": 5.720447994616145e-06,
+      "loss": 0.4883,
+      "step": 4094
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3752442785658663,
+      "learning_rate": 5.71860645144197e-06,
+      "loss": 0.5547,
+      "step": 4095
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.219965061180549,
+      "learning_rate": 5.716764808726554e-06,
+      "loss": 0.4605,
+      "step": 4096
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1531086527738066,
+      "learning_rate": 5.714923066725e-06,
+      "loss": 0.5662,
+      "step": 4097
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4242933567956304,
+      "learning_rate": 5.713081225692429e-06,
+      "loss": 0.4251,
+      "step": 4098
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0755982859493183,
+      "learning_rate": 5.711239285883968e-06,
+      "loss": 0.5075,
+      "step": 4099
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.163603315260941,
+      "learning_rate": 5.709397247554764e-06,
+      "loss": 0.4874,
+      "step": 4100
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.819603495820722,
+      "learning_rate": 5.707555110959979e-06,
+      "loss": 0.6313,
+      "step": 4101
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9863060316053744,
+      "learning_rate": 5.7057128763547806e-06,
+      "loss": 0.4618,
+      "step": 4102
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.861698265235767,
+      "learning_rate": 5.703870543994357e-06,
+      "loss": 0.507,
+      "step": 4103
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8856624138825477,
+      "learning_rate": 5.7020281141339065e-06,
+      "loss": 0.5437,
+      "step": 4104
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6991631478477367,
+      "learning_rate": 5.700185587028644e-06,
+      "loss": 0.4918,
+      "step": 4105
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1232700414889787,
+      "learning_rate": 5.698342962933795e-06,
+      "loss": 0.5827,
+      "step": 4106
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9217145459342841,
+      "learning_rate": 5.696500242104601e-06,
+      "loss": 0.6902,
+      "step": 4107
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.102765786038736,
+      "learning_rate": 5.694657424796313e-06,
+      "loss": 0.4554,
+      "step": 4108
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6488490103289317,
+      "learning_rate": 5.692814511264199e-06,
+      "loss": 0.4874,
+      "step": 4109
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3196322521293373,
+      "learning_rate": 5.6909715017635385e-06,
+      "loss": 0.4901,
+      "step": 4110
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7787424806913932,
+      "learning_rate": 5.689128396549626e-06,
+      "loss": 0.511,
+      "step": 4111
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.572026086719596,
+      "learning_rate": 5.687285195877766e-06,
+      "loss": 0.4512,
+      "step": 4112
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9988195334038317,
+      "learning_rate": 5.685441900003281e-06,
+      "loss": 0.6307,
+      "step": 4113
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1184847284891357,
+      "learning_rate": 5.683598509181503e-06,
+      "loss": 0.5759,
+      "step": 4114
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.295700217844005,
+      "learning_rate": 5.681755023667777e-06,
+      "loss": 0.5746,
+      "step": 4115
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2836223464481344,
+      "learning_rate": 5.679911443717464e-06,
+      "loss": 0.5533,
+      "step": 4116
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0281870883366993,
+      "learning_rate": 5.678067769585935e-06,
+      "loss": 0.4295,
+      "step": 4117
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2329275839233045,
+      "learning_rate": 5.676224001528577e-06,
+      "loss": 0.5842,
+      "step": 4118
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.189577601090869,
+      "learning_rate": 5.674380139800786e-06,
+      "loss": 0.438,
+      "step": 4119
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.656414901277946,
+      "learning_rate": 5.6725361846579755e-06,
+      "loss": 0.5253,
+      "step": 4120
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7425987454629568,
+      "learning_rate": 5.670692136355569e-06,
+      "loss": 0.481,
+      "step": 4121
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6304100298108062,
+      "learning_rate": 5.668847995149005e-06,
+      "loss": 0.5675,
+      "step": 4122
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6019440227691,
+      "learning_rate": 5.667003761293731e-06,
+      "loss": 0.5271,
+      "step": 4123
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.041535111408343,
+      "learning_rate": 5.665159435045214e-06,
+      "loss": 0.4662,
+      "step": 4124
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8839017962855645,
+      "learning_rate": 5.663315016658925e-06,
+      "loss": 0.6767,
+      "step": 4125
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5897497689849227,
+      "learning_rate": 5.661470506390354e-06,
+      "loss": 0.4292,
+      "step": 4126
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0256400021780188,
+      "learning_rate": 5.659625904495004e-06,
+      "loss": 0.602,
+      "step": 4127
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.5325215587607035,
+      "learning_rate": 5.657781211228388e-06,
+      "loss": 0.5316,
+      "step": 4128
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.999266488784331,
+      "learning_rate": 5.655936426846033e-06,
+      "loss": 0.4817,
+      "step": 4129
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6807419991481787,
+      "learning_rate": 5.654091551603478e-06,
+      "loss": 0.4739,
+      "step": 4130
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.961561962990601,
+      "learning_rate": 5.652246585756274e-06,
+      "loss": 0.4727,
+      "step": 4131
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7214130324545978,
+      "learning_rate": 5.6504015295599846e-06,
+      "loss": 0.5621,
+      "step": 4132
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8168650473240665,
+      "learning_rate": 5.64855638327019e-06,
+      "loss": 0.6843,
+      "step": 4133
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.071989491518487,
+      "learning_rate": 5.646711147142477e-06,
+      "loss": 0.4794,
+      "step": 4134
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6043097567974187,
+      "learning_rate": 5.644865821432448e-06,
+      "loss": 0.4729,
+      "step": 4135
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.511174266566765,
+      "learning_rate": 5.643020406395716e-06,
+      "loss": 0.4514,
+      "step": 4136
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2220030768325048,
+      "learning_rate": 5.64117490228791e-06,
+      "loss": 0.4687,
+      "step": 4137
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.4090379789284713,
+      "learning_rate": 5.639329309364667e-06,
+      "loss": 0.5287,
+      "step": 4138
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.202049738729965,
+      "learning_rate": 5.637483627881639e-06,
+      "loss": 0.5152,
+      "step": 4139
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9827941804503668,
+      "learning_rate": 5.635637858094489e-06,
+      "loss": 0.591,
+      "step": 4140
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5627346021217887,
+      "learning_rate": 5.633792000258894e-06,
+      "loss": 0.5133,
+      "step": 4141
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8759455774815856,
+      "learning_rate": 5.6319460546305404e-06,
+      "loss": 0.469,
+      "step": 4142
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.509325023087706,
+      "learning_rate": 5.630100021465128e-06,
+      "loss": 0.5049,
+      "step": 4143
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3080639254461413,
+      "learning_rate": 5.628253901018371e-06,
+      "loss": 0.4792,
+      "step": 4144
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0027224557573473,
+      "learning_rate": 5.626407693545992e-06,
+      "loss": 0.5526,
+      "step": 4145
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7381523511728132,
+      "learning_rate": 5.624561399303727e-06,
+      "loss": 0.5085,
+      "step": 4146
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5077996319332847,
+      "learning_rate": 5.622715018547325e-06,
+      "loss": 0.4949,
+      "step": 4147
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8584742885020111,
+      "learning_rate": 5.620868551532548e-06,
+      "loss": 0.4701,
+      "step": 4148
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.317653547487619,
+      "learning_rate": 5.619021998515165e-06,
+      "loss": 0.5021,
+      "step": 4149
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9291006573565688,
+      "learning_rate": 5.617175359750964e-06,
+      "loss": 0.515,
+      "step": 4150
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9185328160460615,
+      "learning_rate": 5.615328635495738e-06,
+      "loss": 0.5502,
+      "step": 4151
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9276579937636906,
+      "learning_rate": 5.613481826005296e-06,
+      "loss": 0.4379,
+      "step": 4152
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6220967379404887,
+      "learning_rate": 5.6116349315354565e-06,
+      "loss": 0.5494,
+      "step": 4153
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1948035848535117,
+      "learning_rate": 5.609787952342054e-06,
+      "loss": 0.5596,
+      "step": 4154
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9424964079889804,
+      "learning_rate": 5.607940888680929e-06,
+      "loss": 0.5069,
+      "step": 4155
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8534538163941456,
+      "learning_rate": 5.6060937408079374e-06,
+      "loss": 0.4449,
+      "step": 4156
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.088534131861963,
+      "learning_rate": 5.604246508978947e-06,
+      "loss": 0.6058,
+      "step": 4157
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.366350784511451,
+      "learning_rate": 5.6023991934498345e-06,
+      "loss": 0.4441,
+      "step": 4158
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8860825535127888,
+      "learning_rate": 5.6005517944764895e-06,
+      "loss": 0.5204,
+      "step": 4159
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.851712126653901,
+      "learning_rate": 5.598704312314813e-06,
+      "loss": 0.4504,
+      "step": 4160
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.033320111303184,
+      "learning_rate": 5.59685674722072e-06,
+      "loss": 0.4864,
+      "step": 4161
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.238228005155648,
+      "learning_rate": 5.5950090994501335e-06,
+      "loss": 0.4531,
+      "step": 4162
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8149911279577937,
+      "learning_rate": 5.593161369258991e-06,
+      "loss": 0.4466,
+      "step": 4163
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.318340863744374,
+      "learning_rate": 5.591313556903238e-06,
+      "loss": 0.5283,
+      "step": 4164
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7543864291233446,
+      "learning_rate": 5.589465662638831e-06,
+      "loss": 0.5526,
+      "step": 4165
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7299705014569666,
+      "learning_rate": 5.587617686721745e-06,
+      "loss": 0.4967,
+      "step": 4166
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1934163214439772,
+      "learning_rate": 5.585769629407958e-06,
+      "loss": 0.4412,
+      "step": 4167
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.770499691107071,
+      "learning_rate": 5.583921490953463e-06,
+      "loss": 0.4855,
+      "step": 4168
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9553997281729163,
+      "learning_rate": 5.5820732716142645e-06,
+      "loss": 0.5158,
+      "step": 4169
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.5399882913403236,
+      "learning_rate": 5.580224971646377e-06,
+      "loss": 0.5656,
+      "step": 4170
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.88482970919887,
+      "learning_rate": 5.578376591305827e-06,
+      "loss": 0.5364,
+      "step": 4171
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5189968890444645,
+      "learning_rate": 5.576528130848652e-06,
+      "loss": 0.4761,
+      "step": 4172
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8175979832421028,
+      "learning_rate": 5.5746795905309e-06,
+      "loss": 0.6874,
+      "step": 4173
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3937920363676315,
+      "learning_rate": 5.57283097060863e-06,
+      "loss": 0.522,
+      "step": 4174
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1400877082978993,
+      "learning_rate": 5.570982271337916e-06,
+      "loss": 0.4829,
+      "step": 4175
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9870581805467729,
+      "learning_rate": 5.569133492974834e-06,
+      "loss": 0.4377,
+      "step": 4176
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.403721263461471,
+      "learning_rate": 5.567284635775479e-06,
+      "loss": 0.5694,
+      "step": 4177
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7361009979520146,
+      "learning_rate": 5.565435699995956e-06,
+      "loss": 0.4694,
+      "step": 4178
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.805783916485826,
+      "learning_rate": 5.5635866858923774e-06,
+      "loss": 0.3866,
+      "step": 4179
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9363467290094493,
+      "learning_rate": 5.561737593720867e-06,
+      "loss": 0.6802,
+      "step": 4180
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.75485757712712,
+      "learning_rate": 5.559888423737564e-06,
+      "loss": 0.436,
+      "step": 4181
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.06935680124501,
+      "learning_rate": 5.558039176198613e-06,
+      "loss": 0.4734,
+      "step": 4182
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.510465338613576,
+      "learning_rate": 5.556189851360173e-06,
+      "loss": 0.4779,
+      "step": 4183
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8589766636356018,
+      "learning_rate": 5.55434044947841e-06,
+      "loss": 0.4939,
+      "step": 4184
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.460997262452607,
+      "learning_rate": 5.552490970809504e-06,
+      "loss": 0.4721,
+      "step": 4185
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.077761661602961,
+      "learning_rate": 5.550641415609646e-06,
+      "loss": 0.4718,
+      "step": 4186
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.892015380372752,
+      "learning_rate": 5.548791784135034e-06,
+      "loss": 0.4632,
+      "step": 4187
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.183627881211844,
+      "learning_rate": 5.546942076641877e-06,
+      "loss": 0.5887,
+      "step": 4188
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.803741275940871,
+      "learning_rate": 5.545092293386399e-06,
+      "loss": 0.4608,
+      "step": 4189
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.468664610427685,
+      "learning_rate": 5.543242434624832e-06,
+      "loss": 0.533,
+      "step": 4190
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.137967659866638,
+      "learning_rate": 5.5413925006134165e-06,
+      "loss": 0.5465,
+      "step": 4191
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8090558621307973,
+      "learning_rate": 5.539542491608406e-06,
+      "loss": 0.4436,
+      "step": 4192
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8803895026643748,
+      "learning_rate": 5.537692407866063e-06,
+      "loss": 0.6918,
+      "step": 4193
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.333408820132182,
+      "learning_rate": 5.53584224964266e-06,
+      "loss": 0.4861,
+      "step": 4194
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.996113450128624,
+      "learning_rate": 5.533992017194481e-06,
+      "loss": 0.5363,
+      "step": 4195
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8025060082184459,
+      "learning_rate": 5.532141710777822e-06,
+      "loss": 0.5177,
+      "step": 4196
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.926863337854834,
+      "learning_rate": 5.530291330648985e-06,
+      "loss": 0.5616,
+      "step": 4197
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.420091851965088,
+      "learning_rate": 5.528440877064284e-06,
+      "loss": 0.4559,
+      "step": 4198
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9440919188221362,
+      "learning_rate": 5.526590350280043e-06,
+      "loss": 0.5827,
+      "step": 4199
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5707012384382986,
+      "learning_rate": 5.524739750552601e-06,
+      "loss": 0.4705,
+      "step": 4200
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.818261552427796,
+      "learning_rate": 5.522889078138298e-06,
+      "loss": 0.4518,
+      "step": 4201
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8722127777476045,
+      "learning_rate": 5.52103833329349e-06,
+      "loss": 0.4276,
+      "step": 4202
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9781903859733203,
+      "learning_rate": 5.5191875162745425e-06,
+      "loss": 0.575,
+      "step": 4203
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1610015067938004,
+      "learning_rate": 5.51733662733783e-06,
+      "loss": 0.5171,
+      "step": 4204
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7079685805933562,
+      "learning_rate": 5.515485666739739e-06,
+      "loss": 0.4369,
+      "step": 4205
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6951160336285662,
+      "learning_rate": 5.513634634736662e-06,
+      "loss": 0.5575,
+      "step": 4206
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.73643320295791,
+      "learning_rate": 5.511783531585004e-06,
+      "loss": 0.4252,
+      "step": 4207
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4095701143503256,
+      "learning_rate": 5.50993235754118e-06,
+      "loss": 0.5753,
+      "step": 4208
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.066039530427085,
+      "learning_rate": 5.508081112861614e-06,
+      "loss": 0.4593,
+      "step": 4209
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9531306094377258,
+      "learning_rate": 5.50622979780274e-06,
+      "loss": 0.7414,
+      "step": 4210
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8699006315609543,
+      "learning_rate": 5.504378412621003e-06,
+      "loss": 0.6138,
+      "step": 4211
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8251540803865427,
+      "learning_rate": 5.502526957572855e-06,
+      "loss": 0.555,
+      "step": 4212
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4978626745556893,
+      "learning_rate": 5.5006754329147595e-06,
+      "loss": 0.4954,
+      "step": 4213
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.804855865537222,
+      "learning_rate": 5.4988238389031904e-06,
+      "loss": 0.5597,
+      "step": 4214
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8936785388999215,
+      "learning_rate": 5.496972175794628e-06,
+      "loss": 0.4483,
+      "step": 4215
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8216465927204806,
+      "learning_rate": 5.4951204438455674e-06,
+      "loss": 0.4991,
+      "step": 4216
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.145256387930919,
+      "learning_rate": 5.49326864331251e-06,
+      "loss": 0.4945,
+      "step": 4217
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8095099173017117,
+      "learning_rate": 5.491416774451963e-06,
+      "loss": 0.5871,
+      "step": 4218
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0609005232464144,
+      "learning_rate": 5.489564837520451e-06,
+      "loss": 0.5,
+      "step": 4219
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7541152043472044,
+      "learning_rate": 5.487712832774502e-06,
+      "loss": 0.4983,
+      "step": 4220
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1514512289907914,
+      "learning_rate": 5.485860760470656e-06,
+      "loss": 0.5166,
+      "step": 4221
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9010564184976764,
+      "learning_rate": 5.484008620865461e-06,
+      "loss": 0.4977,
+      "step": 4222
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4878228814801493,
+      "learning_rate": 5.482156414215476e-06,
+      "loss": 0.479,
+      "step": 4223
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.377831540294386,
+      "learning_rate": 5.480304140777268e-06,
+      "loss": 0.4514,
+      "step": 4224
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9775334607957433,
+      "learning_rate": 5.478451800807414e-06,
+      "loss": 0.5268,
+      "step": 4225
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1322986645843716,
+      "learning_rate": 5.4765993945625e-06,
+      "loss": 0.5763,
+      "step": 4226
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6794139374447141,
+      "learning_rate": 5.474746922299119e-06,
+      "loss": 0.5109,
+      "step": 4227
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8470214797830506,
+      "learning_rate": 5.472894384273877e-06,
+      "loss": 0.4829,
+      "step": 4228
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.179447214431165,
+      "learning_rate": 5.471041780743388e-06,
+      "loss": 0.4743,
+      "step": 4229
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.07856986214627,
+      "learning_rate": 5.4691891119642725e-06,
+      "loss": 0.5479,
+      "step": 4230
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.096518014641802,
+      "learning_rate": 5.467336378193162e-06,
+      "loss": 0.4831,
+      "step": 4231
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7066353119078104,
+      "learning_rate": 5.465483579686697e-06,
+      "loss": 0.5222,
+      "step": 4232
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7466369247851958,
+      "learning_rate": 5.463630716701528e-06,
+      "loss": 0.5073,
+      "step": 4233
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7826965887395634,
+      "learning_rate": 5.461777789494312e-06,
+      "loss": 0.4965,
+      "step": 4234
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9315768243971463,
+      "learning_rate": 5.459924798321717e-06,
+      "loss": 0.502,
+      "step": 4235
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.061152777431662,
+      "learning_rate": 5.458071743440418e-06,
+      "loss": 0.5807,
+      "step": 4236
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9141693120915213,
+      "learning_rate": 5.4562186251071e-06,
+      "loss": 0.4165,
+      "step": 4237
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4984102978562386,
+      "learning_rate": 5.454365443578457e-06,
+      "loss": 0.5617,
+      "step": 4238
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7854418412148907,
+      "learning_rate": 5.452512199111193e-06,
+      "loss": 0.5487,
+      "step": 4239
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7118796507893006,
+      "learning_rate": 5.450658891962016e-06,
+      "loss": 0.4864,
+      "step": 4240
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8075983385958314,
+      "learning_rate": 5.4488055223876494e-06,
+      "loss": 0.3922,
+      "step": 4241
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.4561413500089495,
+      "learning_rate": 5.446952090644818e-06,
+      "loss": 0.4309,
+      "step": 4242
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5781385731931503,
+      "learning_rate": 5.445098596990261e-06,
+      "loss": 0.5029,
+      "step": 4243
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8081759585979522,
+      "learning_rate": 5.443245041680722e-06,
+      "loss": 0.5132,
+      "step": 4244
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5807753283071673,
+      "learning_rate": 5.441391424972958e-06,
+      "loss": 0.4887,
+      "step": 4245
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.773516632891921,
+      "learning_rate": 5.4395377471237295e-06,
+      "loss": 0.5751,
+      "step": 4246
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0648290134306038,
+      "learning_rate": 5.43768400838981e-06,
+      "loss": 0.474,
+      "step": 4247
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.224153987683954,
+      "learning_rate": 5.435830209027977e-06,
+      "loss": 0.5415,
+      "step": 4248
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6465132193170648,
+      "learning_rate": 5.433976349295018e-06,
+      "loss": 0.4725,
+      "step": 4249
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8740972767400428,
+      "learning_rate": 5.432122429447731e-06,
+      "loss": 0.6947,
+      "step": 4250
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0403575066044994,
+      "learning_rate": 5.43026844974292e-06,
+      "loss": 0.507,
+      "step": 4251
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.012382509718183,
+      "learning_rate": 5.428414410437397e-06,
+      "loss": 0.4546,
+      "step": 4252
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7782547283467982,
+      "learning_rate": 5.426560311787984e-06,
+      "loss": 0.4318,
+      "step": 4253
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1001178258887134,
+      "learning_rate": 5.4247061540515115e-06,
+      "loss": 0.526,
+      "step": 4254
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.785369659172782,
+      "learning_rate": 5.422851937484814e-06,
+      "loss": 0.6257,
+      "step": 4255
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.782298455372749,
+      "learning_rate": 5.420997662344741e-06,
+      "loss": 0.5945,
+      "step": 4256
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9192264034322384,
+      "learning_rate": 5.4191433288881445e-06,
+      "loss": 0.5392,
+      "step": 4257
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9576819276746926,
+      "learning_rate": 5.417288937371886e-06,
+      "loss": 0.5385,
+      "step": 4258
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3223863130300475,
+      "learning_rate": 5.415434488052838e-06,
+      "loss": 0.3732,
+      "step": 4259
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.259740039037036,
+      "learning_rate": 5.413579981187876e-06,
+      "loss": 0.5995,
+      "step": 4260
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9748927540697199,
+      "learning_rate": 5.411725417033886e-06,
+      "loss": 0.4326,
+      "step": 4261
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0168311808756862,
+      "learning_rate": 5.409870795847763e-06,
+      "loss": 0.4861,
+      "step": 4262
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.101034914947914,
+      "learning_rate": 5.408016117886408e-06,
+      "loss": 0.4539,
+      "step": 4263
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0662258097021082,
+      "learning_rate": 5.40616138340673e-06,
+      "loss": 0.4696,
+      "step": 4264
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7375359031607729,
+      "learning_rate": 5.404306592665649e-06,
+      "loss": 0.6386,
+      "step": 4265
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.701437154831304,
+      "learning_rate": 5.402451745920089e-06,
+      "loss": 0.4951,
+      "step": 4266
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7047353217265153,
+      "learning_rate": 5.400596843426982e-06,
+      "loss": 0.4027,
+      "step": 4267
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.950201280037983,
+      "learning_rate": 5.398741885443271e-06,
+      "loss": 0.5135,
+      "step": 4268
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9419896895584683,
+      "learning_rate": 5.396886872225902e-06,
+      "loss": 0.7379,
+      "step": 4269
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.285664711064563,
+      "learning_rate": 5.395031804031832e-06,
+      "loss": 0.5395,
+      "step": 4270
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8750682948924278,
+      "learning_rate": 5.3931766811180255e-06,
+      "loss": 0.452,
+      "step": 4271
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.179654062503672,
+      "learning_rate": 5.391321503741454e-06,
+      "loss": 0.6463,
+      "step": 4272
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9799662770085158,
+      "learning_rate": 5.3894662721590926e-06,
+      "loss": 0.4878,
+      "step": 4273
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9257453879214355,
+      "learning_rate": 5.387610986627933e-06,
+      "loss": 0.4606,
+      "step": 4274
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8355157905878297,
+      "learning_rate": 5.385755647404966e-06,
+      "loss": 0.4925,
+      "step": 4275
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3039730665079845,
+      "learning_rate": 5.383900254747195e-06,
+      "loss": 0.6112,
+      "step": 4276
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.217442509109483,
+      "learning_rate": 5.382044808911626e-06,
+      "loss": 0.487,
+      "step": 4277
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7726503949957795,
+      "learning_rate": 5.380189310155276e-06,
+      "loss": 0.4985,
+      "step": 4278
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9436139307883982,
+      "learning_rate": 5.378333758735168e-06,
+      "loss": 0.4206,
+      "step": 4279
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.068196128346733,
+      "learning_rate": 5.376478154908335e-06,
+      "loss": 0.5087,
+      "step": 4280
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9872965090234882,
+      "learning_rate": 5.374622498931812e-06,
+      "loss": 0.4316,
+      "step": 4281
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.01509428138601,
+      "learning_rate": 5.372766791062645e-06,
+      "loss": 0.4593,
+      "step": 4282
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8060045180271909,
+      "learning_rate": 5.370911031557887e-06,
+      "loss": 0.3792,
+      "step": 4283
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0466213177281234,
+      "learning_rate": 5.369055220674597e-06,
+      "loss": 0.4503,
+      "step": 4284
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.936093595553273,
+      "learning_rate": 5.3671993586698434e-06,
+      "loss": 0.4687,
+      "step": 4285
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9042011440650939,
+      "learning_rate": 5.365343445800697e-06,
+      "loss": 0.5608,
+      "step": 4286
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.195929676320687,
+      "learning_rate": 5.363487482324239e-06,
+      "loss": 0.4767,
+      "step": 4287
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.162688457490507,
+      "learning_rate": 5.361631468497559e-06,
+      "loss": 0.4512,
+      "step": 4288
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4730370416897656,
+      "learning_rate": 5.35977540457775e-06,
+      "loss": 0.4678,
+      "step": 4289
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9473156565727763,
+      "learning_rate": 5.3579192908219145e-06,
+      "loss": 0.4304,
+      "step": 4290
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0396690287040533,
+      "learning_rate": 5.3560631274871626e-06,
+      "loss": 0.3768,
+      "step": 4291
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.165263328611737,
+      "learning_rate": 5.354206914830605e-06,
+      "loss": 0.4551,
+      "step": 4292
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8923391706900512,
+      "learning_rate": 5.352350653109368e-06,
+      "loss": 0.4594,
+      "step": 4293
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2714876814163976,
+      "learning_rate": 5.350494342580581e-06,
+      "loss": 0.4843,
+      "step": 4294
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.235918414202698,
+      "learning_rate": 5.348637983501377e-06,
+      "loss": 0.4958,
+      "step": 4295
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9873927183824973,
+      "learning_rate": 5.3467815761289e-06,
+      "loss": 0.439,
+      "step": 4296
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7617756632461765,
+      "learning_rate": 5.344925120720299e-06,
+      "loss": 0.432,
+      "step": 4297
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3517273847692013,
+      "learning_rate": 5.343068617532729e-06,
+      "loss": 0.5884,
+      "step": 4298
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9316001360412404,
+      "learning_rate": 5.341212066823356e-06,
+      "loss": 0.7352,
+      "step": 4299
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.147101273486869,
+      "learning_rate": 5.339355468849344e-06,
+      "loss": 0.557,
+      "step": 4300
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0610002074838554,
+      "learning_rate": 5.337498823867872e-06,
+      "loss": 0.5003,
+      "step": 4301
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3205051367957763,
+      "learning_rate": 5.335642132136124e-06,
+      "loss": 0.5459,
+      "step": 4302
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8220091671274186,
+      "learning_rate": 5.333785393911284e-06,
+      "loss": 0.5629,
+      "step": 4303
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.628397747190286,
+      "learning_rate": 5.331928609450548e-06,
+      "loss": 0.5112,
+      "step": 4304
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8262453351010755,
+      "learning_rate": 5.33007177901112e-06,
+      "loss": 0.4596,
+      "step": 4305
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2186296865060164,
+      "learning_rate": 5.328214902850205e-06,
+      "loss": 0.389,
+      "step": 4306
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1208898499289632,
+      "learning_rate": 5.32635798122502e-06,
+      "loss": 0.4686,
+      "step": 4307
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.888215400637544,
+      "learning_rate": 5.324501014392782e-06,
+      "loss": 0.508,
+      "step": 4308
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9549221183415764,
+      "learning_rate": 5.322644002610722e-06,
+      "loss": 0.5682,
+      "step": 4309
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8053569026506293,
+      "learning_rate": 5.32078694613607e-06,
+      "loss": 0.447,
+      "step": 4310
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7138974462734162,
+      "learning_rate": 5.318929845226065e-06,
+      "loss": 0.495,
+      "step": 4311
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1456297340250887,
+      "learning_rate": 5.317072700137953e-06,
+      "loss": 0.4689,
+      "step": 4312
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8133898802325854,
+      "learning_rate": 5.3152155111289874e-06,
+      "loss": 0.6012,
+      "step": 4313
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.649796877364736,
+      "learning_rate": 5.313358278456422e-06,
+      "loss": 0.4731,
+      "step": 4314
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.084404569254779,
+      "learning_rate": 5.3115010023775225e-06,
+      "loss": 0.4889,
+      "step": 4315
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7574113251113324,
+      "learning_rate": 5.309643683149558e-06,
+      "loss": 0.4714,
+      "step": 4316
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.32321764253821,
+      "learning_rate": 5.307786321029804e-06,
+      "loss": 0.4894,
+      "step": 4317
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2052755098665786,
+      "learning_rate": 5.305928916275544e-06,
+      "loss": 0.5042,
+      "step": 4318
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3187308264612514,
+      "learning_rate": 5.304071469144061e-06,
+      "loss": 0.4639,
+      "step": 4319
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.765116291849073,
+      "learning_rate": 5.302213979892652e-06,
+      "loss": 0.3894,
+      "step": 4320
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.525920744131966,
+      "learning_rate": 5.300356448778614e-06,
+      "loss": 0.5832,
+      "step": 4321
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.237883535455513,
+      "learning_rate": 5.298498876059252e-06,
+      "loss": 0.5571,
+      "step": 4322
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.674111048425905,
+      "learning_rate": 5.29664126199188e-06,
+      "loss": 0.4623,
+      "step": 4323
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2680513703188048,
+      "learning_rate": 5.29478360683381e-06,
+      "loss": 0.4823,
+      "step": 4324
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.362862037665259,
+      "learning_rate": 5.292925910842366e-06,
+      "loss": 0.5677,
+      "step": 4325
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8700489467813168,
+      "learning_rate": 5.291068174274876e-06,
+      "loss": 0.5913,
+      "step": 4326
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9771333603192363,
+      "learning_rate": 5.289210397388673e-06,
+      "loss": 0.4329,
+      "step": 4327
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.4068578536091305,
+      "learning_rate": 5.287352580441095e-06,
+      "loss": 0.5226,
+      "step": 4328
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.08893941663053,
+      "learning_rate": 5.285494723689488e-06,
+      "loss": 0.4172,
+      "step": 4329
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.042582229866781,
+      "learning_rate": 5.283636827391201e-06,
+      "loss": 0.4304,
+      "step": 4330
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0324273903259087,
+      "learning_rate": 5.281778891803591e-06,
+      "loss": 0.4649,
+      "step": 4331
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.536997390166242,
+      "learning_rate": 5.279920917184016e-06,
+      "loss": 0.5185,
+      "step": 4332
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6478052513377677,
+      "learning_rate": 5.278062903789846e-06,
+      "loss": 0.5101,
+      "step": 4333
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0240575938795122,
+      "learning_rate": 5.276204851878448e-06,
+      "loss": 0.5014,
+      "step": 4334
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9664211056797702,
+      "learning_rate": 5.274346761707204e-06,
+      "loss": 0.5297,
+      "step": 4335
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8939815045988033,
+      "learning_rate": 5.272488633533493e-06,
+      "loss": 0.7381,
+      "step": 4336
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5707312262037825,
+      "learning_rate": 5.270630467614705e-06,
+      "loss": 0.5537,
+      "step": 4337
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.859466157561985,
+      "learning_rate": 5.268772264208231e-06,
+      "loss": 0.5047,
+      "step": 4338
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.972332579589227,
+      "learning_rate": 5.266914023571468e-06,
+      "loss": 0.4926,
+      "step": 4339
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0496259683604685,
+      "learning_rate": 5.265055745961821e-06,
+      "loss": 0.5593,
+      "step": 4340
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.112464638054792,
+      "learning_rate": 5.2631974316367e-06,
+      "loss": 0.4535,
+      "step": 4341
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.034860193759808,
+      "learning_rate": 5.261339080853514e-06,
+      "loss": 0.526,
+      "step": 4342
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6356161787167383,
+      "learning_rate": 5.2594806938696855e-06,
+      "loss": 0.4621,
+      "step": 4343
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3163697904653744,
+      "learning_rate": 5.257622270942636e-06,
+      "loss": 0.5368,
+      "step": 4344
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8011054415475338,
+      "learning_rate": 5.2557638123297924e-06,
+      "loss": 0.4875,
+      "step": 4345
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9260275417036818,
+      "learning_rate": 5.2539053182885916e-06,
+      "loss": 0.6972,
+      "step": 4346
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8487262902771333,
+      "learning_rate": 5.252046789076469e-06,
+      "loss": 0.5379,
+      "step": 4347
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8240406318622295,
+      "learning_rate": 5.25018822495087e-06,
+      "loss": 0.415,
+      "step": 4348
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9681706418100264,
+      "learning_rate": 5.2483296261692405e-06,
+      "loss": 0.5093,
+      "step": 4349
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0904278415737703,
+      "learning_rate": 5.246470992989034e-06,
+      "loss": 0.5288,
+      "step": 4350
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9818633877379053,
+      "learning_rate": 5.24461232566771e-06,
+      "loss": 0.4791,
+      "step": 4351
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7374506478528393,
+      "learning_rate": 5.242753624462728e-06,
+      "loss": 0.5259,
+      "step": 4352
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.422080367277146,
+      "learning_rate": 5.240894889631556e-06,
+      "loss": 0.4409,
+      "step": 4353
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.734159568443539,
+      "learning_rate": 5.239036121431664e-06,
+      "loss": 0.4472,
+      "step": 4354
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.871594824009396,
+      "learning_rate": 5.237177320120532e-06,
+      "loss": 0.4872,
+      "step": 4355
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.873872002614127,
+      "learning_rate": 5.235318485955638e-06,
+      "loss": 0.4124,
+      "step": 4356
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6794890196754846,
+      "learning_rate": 5.233459619194469e-06,
+      "loss": 0.5986,
+      "step": 4357
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8383798721512503,
+      "learning_rate": 5.231600720094513e-06,
+      "loss": 0.6959,
+      "step": 4358
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.870038946976061,
+      "learning_rate": 5.2297417889132655e-06,
+      "loss": 0.4866,
+      "step": 4359
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6334975065562425,
+      "learning_rate": 5.227882825908224e-06,
+      "loss": 0.5275,
+      "step": 4360
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.018515835667817,
+      "learning_rate": 5.226023831336895e-06,
+      "loss": 0.5197,
+      "step": 4361
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.683333857728174,
+      "learning_rate": 5.224164805456783e-06,
+      "loss": 0.4603,
+      "step": 4362
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.982486207758545,
+      "learning_rate": 5.222305748525401e-06,
+      "loss": 0.452,
+      "step": 4363
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.029233754403527,
+      "learning_rate": 5.220446660800264e-06,
+      "loss": 0.505,
+      "step": 4364
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0474856636766705,
+      "learning_rate": 5.218587542538895e-06,
+      "loss": 0.5432,
+      "step": 4365
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.684838802235383,
+      "learning_rate": 5.216728393998818e-06,
+      "loss": 0.4143,
+      "step": 4366
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9160241418602428,
+      "learning_rate": 5.214869215437562e-06,
+      "loss": 0.4501,
+      "step": 4367
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.873330938251454,
+      "learning_rate": 5.2130100071126565e-06,
+      "loss": 0.5433,
+      "step": 4368
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2401641636346383,
+      "learning_rate": 5.211150769281645e-06,
+      "loss": 0.555,
+      "step": 4369
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.038017172982198,
+      "learning_rate": 5.209291502202064e-06,
+      "loss": 0.5833,
+      "step": 4370
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8120940362954039,
+      "learning_rate": 5.20743220613146e-06,
+      "loss": 0.5407,
+      "step": 4371
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.065648516569301,
+      "learning_rate": 5.205572881327383e-06,
+      "loss": 0.4097,
+      "step": 4372
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.270095588239838,
+      "learning_rate": 5.203713528047386e-06,
+      "loss": 0.599,
+      "step": 4373
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1391595974591953,
+      "learning_rate": 5.201854146549027e-06,
+      "loss": 0.5998,
+      "step": 4374
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.437619006601976,
+      "learning_rate": 5.199994737089868e-06,
+      "loss": 0.5048,
+      "step": 4375
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.5846032549539664,
+      "learning_rate": 5.198135299927469e-06,
+      "loss": 0.4505,
+      "step": 4376
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.684985479226318,
+      "learning_rate": 5.196275835319405e-06,
+      "loss": 0.4815,
+      "step": 4377
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1384131635457635,
+      "learning_rate": 5.194416343523246e-06,
+      "loss": 0.4707,
+      "step": 4378
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0767712689451283,
+      "learning_rate": 5.1925568247965686e-06,
+      "loss": 0.5008,
+      "step": 4379
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1263901944187285,
+      "learning_rate": 5.190697279396954e-06,
+      "loss": 0.4936,
+      "step": 4380
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2794019538675903,
+      "learning_rate": 5.188837707581983e-06,
+      "loss": 0.5597,
+      "step": 4381
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.09279435308324,
+      "learning_rate": 5.186978109609248e-06,
+      "loss": 0.4557,
+      "step": 4382
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6592518939409033,
+      "learning_rate": 5.185118485736336e-06,
+      "loss": 0.4075,
+      "step": 4383
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3210002815613073,
+      "learning_rate": 5.183258836220844e-06,
+      "loss": 0.4643,
+      "step": 4384
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.044384672567464,
+      "learning_rate": 5.181399161320368e-06,
+      "loss": 0.5388,
+      "step": 4385
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1456808046138676,
+      "learning_rate": 5.179539461292514e-06,
+      "loss": 0.4865,
+      "step": 4386
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6872103740112088,
+      "learning_rate": 5.177679736394885e-06,
+      "loss": 0.3983,
+      "step": 4387
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9978209348756895,
+      "learning_rate": 5.1758199868850875e-06,
+      "loss": 0.4902,
+      "step": 4388
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1071930925392977,
+      "learning_rate": 5.173960213020737e-06,
+      "loss": 0.4628,
+      "step": 4389
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.074987885951766,
+      "learning_rate": 5.172100415059449e-06,
+      "loss": 0.5643,
+      "step": 4390
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9512771937207587,
+      "learning_rate": 5.170240593258839e-06,
+      "loss": 0.5357,
+      "step": 4391
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2705834412404644,
+      "learning_rate": 5.1683807478765335e-06,
+      "loss": 0.5121,
+      "step": 4392
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9471872284011937,
+      "learning_rate": 5.166520879170156e-06,
+      "loss": 0.4819,
+      "step": 4393
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3250527388533877,
+      "learning_rate": 5.1646609873973354e-06,
+      "loss": 0.5091,
+      "step": 4394
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1897222454917453,
+      "learning_rate": 5.162801072815702e-06,
+      "loss": 0.5601,
+      "step": 4395
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.26402763171055,
+      "learning_rate": 5.160941135682893e-06,
+      "loss": 0.6025,
+      "step": 4396
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1002163919247137,
+      "learning_rate": 5.159081176256545e-06,
+      "loss": 0.4894,
+      "step": 4397
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1251487603159185,
+      "learning_rate": 5.157221194794302e-06,
+      "loss": 0.5199,
+      "step": 4398
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8842010797318722,
+      "learning_rate": 5.155361191553804e-06,
+      "loss": 0.6349,
+      "step": 4399
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.810397302117786,
+      "learning_rate": 5.153501166792702e-06,
+      "loss": 0.4984,
+      "step": 4400
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3760559904731213,
+      "learning_rate": 5.1516411207686435e-06,
+      "loss": 0.431,
+      "step": 4401
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9991125693301748,
+      "learning_rate": 5.1497810537392844e-06,
+      "loss": 0.4583,
+      "step": 4402
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.240526245928486,
+      "learning_rate": 5.147920965962279e-06,
+      "loss": 0.6155,
+      "step": 4403
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.573512953126507,
+      "learning_rate": 5.146060857695288e-06,
+      "loss": 0.454,
+      "step": 4404
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9684184489817413,
+      "learning_rate": 5.1442007291959715e-06,
+      "loss": 0.5029,
+      "step": 4405
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.222177936080414,
+      "learning_rate": 5.142340580721995e-06,
+      "loss": 0.4521,
+      "step": 4406
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.084433030978843,
+      "learning_rate": 5.140480412531027e-06,
+      "loss": 0.5306,
+      "step": 4407
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8880798612677099,
+      "learning_rate": 5.138620224880735e-06,
+      "loss": 0.5455,
+      "step": 4408
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7852614397771662,
+      "learning_rate": 5.136760018028793e-06,
+      "loss": 0.5512,
+      "step": 4409
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0460389888499733,
+      "learning_rate": 5.134899792232879e-06,
+      "loss": 0.4369,
+      "step": 4410
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.953868119269436,
+      "learning_rate": 5.133039547750669e-06,
+      "loss": 0.424,
+      "step": 4411
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9390321726996727,
+      "learning_rate": 5.131179284839843e-06,
+      "loss": 0.483,
+      "step": 4412
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4585019776191643,
+      "learning_rate": 5.129319003758085e-06,
+      "loss": 0.4566,
+      "step": 4413
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9849129612851817,
+      "learning_rate": 5.1274587047630816e-06,
+      "loss": 0.4418,
+      "step": 4414
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7482859930494572,
+      "learning_rate": 5.12559838811252e-06,
+      "loss": 0.4723,
+      "step": 4415
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0225913574524164,
+      "learning_rate": 5.1237380540640915e-06,
+      "loss": 0.5704,
+      "step": 4416
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2120385249626775,
+      "learning_rate": 5.12187770287549e-06,
+      "loss": 0.4763,
+      "step": 4417
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.209407236134105,
+      "learning_rate": 5.12001733480441e-06,
+      "loss": 0.4632,
+      "step": 4418
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5682980079890456,
+      "learning_rate": 5.118156950108549e-06,
+      "loss": 0.4961,
+      "step": 4419
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.23644487557047,
+      "learning_rate": 5.116296549045607e-06,
+      "loss": 0.5267,
+      "step": 4420
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0703623510496376,
+      "learning_rate": 5.11443613187329e-06,
+      "loss": 0.5338,
+      "step": 4421
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.761164869980416,
+      "learning_rate": 5.1125756988492985e-06,
+      "loss": 0.5174,
+      "step": 4422
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9731425486949428,
+      "learning_rate": 5.110715250231338e-06,
+      "loss": 0.4494,
+      "step": 4423
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8234457846210086,
+      "learning_rate": 5.108854786277123e-06,
+      "loss": 0.4883,
+      "step": 4424
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.9658537361969675,
+      "learning_rate": 5.106994307244361e-06,
+      "loss": 0.476,
+      "step": 4425
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6050939769434147,
+      "learning_rate": 5.105133813390766e-06,
+      "loss": 0.4531,
+      "step": 4426
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.857408683426242,
+      "learning_rate": 5.103273304974054e-06,
+      "loss": 0.5616,
+      "step": 4427
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.746832964508357,
+      "learning_rate": 5.101412782251942e-06,
+      "loss": 0.5693,
+      "step": 4428
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1664387648996897,
+      "learning_rate": 5.099552245482148e-06,
+      "loss": 0.4356,
+      "step": 4429
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1036049183623673,
+      "learning_rate": 5.097691694922394e-06,
+      "loss": 0.5181,
+      "step": 4430
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4064135352355036,
+      "learning_rate": 5.0958311308304045e-06,
+      "loss": 0.4886,
+      "step": 4431
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7729229382258174,
+      "learning_rate": 5.093970553463901e-06,
+      "loss": 0.4196,
+      "step": 4432
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.097947705861859,
+      "learning_rate": 5.092109963080614e-06,
+      "loss": 0.4542,
+      "step": 4433
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.429005079460941,
+      "learning_rate": 5.090249359938273e-06,
+      "loss": 0.554,
+      "step": 4434
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.946827344150558,
+      "learning_rate": 5.088388744294603e-06,
+      "loss": 0.5755,
+      "step": 4435
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8314475235137133,
+      "learning_rate": 5.086528116407342e-06,
+      "loss": 0.5895,
+      "step": 4436
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0860068066893174,
+      "learning_rate": 5.084667476534221e-06,
+      "loss": 0.532,
+      "step": 4437
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4383616511966264,
+      "learning_rate": 5.0828068249329755e-06,
+      "loss": 0.485,
+      "step": 4438
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.766195864271854,
+      "learning_rate": 5.080946161861342e-06,
+      "loss": 0.6115,
+      "step": 4439
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6835757337439372,
+      "learning_rate": 5.0790854875770604e-06,
+      "loss": 0.4393,
+      "step": 4440
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7456475738059325,
+      "learning_rate": 5.077224802337872e-06,
+      "loss": 0.4675,
+      "step": 4441
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8717471764186547,
+      "learning_rate": 5.075364106401517e-06,
+      "loss": 0.7052,
+      "step": 4442
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2690095935046366,
+      "learning_rate": 5.073503400025737e-06,
+      "loss": 0.5163,
+      "step": 4443
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3592273330822313,
+      "learning_rate": 5.071642683468281e-06,
+      "loss": 0.6023,
+      "step": 4444
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.04600923860418,
+      "learning_rate": 5.069781956986894e-06,
+      "loss": 0.4787,
+      "step": 4445
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9667481473015458,
+      "learning_rate": 5.0679212208393196e-06,
+      "loss": 0.5252,
+      "step": 4446
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.644024291599729,
+      "learning_rate": 5.06606047528331e-06,
+      "loss": 0.5743,
+      "step": 4447
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.145402236935919,
+      "learning_rate": 5.064199720576615e-06,
+      "loss": 0.4606,
+      "step": 4448
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0620081524132967,
+      "learning_rate": 5.062338956976986e-06,
+      "loss": 0.5047,
+      "step": 4449
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8712706130409299,
+      "learning_rate": 5.060478184742176e-06,
+      "loss": 0.7158,
+      "step": 4450
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7213720806496773,
+      "learning_rate": 5.058617404129938e-06,
+      "loss": 0.4672,
+      "step": 4451
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4133541248976638,
+      "learning_rate": 5.056756615398026e-06,
+      "loss": 0.4419,
+      "step": 4452
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9803735824066313,
+      "learning_rate": 5.0548958188042e-06,
+      "loss": 0.461,
+      "step": 4453
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1501955369656502,
+      "learning_rate": 5.053035014606212e-06,
+      "loss": 0.4827,
+      "step": 4454
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.02400132473691,
+      "learning_rate": 5.051174203061825e-06,
+      "loss": 0.4974,
+      "step": 4455
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1407765667243654,
+      "learning_rate": 5.0493133844287955e-06,
+      "loss": 0.4706,
+      "step": 4456
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9997346287617235,
+      "learning_rate": 5.047452558964884e-06,
+      "loss": 0.5207,
+      "step": 4457
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.02911970575265,
+      "learning_rate": 5.045591726927853e-06,
+      "loss": 0.5419,
+      "step": 4458
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.389921304692081,
+      "learning_rate": 5.043730888575463e-06,
+      "loss": 0.5249,
+      "step": 4459
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6545975138723845,
+      "learning_rate": 5.04187004416548e-06,
+      "loss": 0.4582,
+      "step": 4460
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0461363535972277,
+      "learning_rate": 5.040009193955664e-06,
+      "loss": 0.5171,
+      "step": 4461
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9767063965321605,
+      "learning_rate": 5.0381483382037825e-06,
+      "loss": 0.5225,
+      "step": 4462
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.460285409521772,
+      "learning_rate": 5.0362874771676e-06,
+      "loss": 0.5281,
+      "step": 4463
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1709822502896583,
+      "learning_rate": 5.03442661110488e-06,
+      "loss": 0.5147,
+      "step": 4464
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.8122527028068007,
+      "learning_rate": 5.032565740273394e-06,
+      "loss": 0.4825,
+      "step": 4465
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.666213591095413,
+      "learning_rate": 5.030704864930907e-06,
+      "loss": 0.5902,
+      "step": 4466
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9960950637814083,
+      "learning_rate": 5.028843985335186e-06,
+      "loss": 0.4707,
+      "step": 4467
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7580022213779434,
+      "learning_rate": 5.0269831017440015e-06,
+      "loss": 0.6086,
+      "step": 4468
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.81004448569933,
+      "learning_rate": 5.025122214415123e-06,
+      "loss": 0.4495,
+      "step": 4469
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9941524771176407,
+      "learning_rate": 5.023261323606321e-06,
+      "loss": 0.5239,
+      "step": 4470
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.004711611200936,
+      "learning_rate": 5.021400429575363e-06,
+      "loss": 0.4855,
+      "step": 4471
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.380617890273587,
+      "learning_rate": 5.019539532580021e-06,
+      "loss": 0.5987,
+      "step": 4472
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3860292600304804,
+      "learning_rate": 5.017678632878067e-06,
+      "loss": 0.5407,
+      "step": 4473
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8716455927377267,
+      "learning_rate": 5.015817730727272e-06,
+      "loss": 0.4558,
+      "step": 4474
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9528274232852458,
+      "learning_rate": 5.013956826385406e-06,
+      "loss": 0.4614,
+      "step": 4475
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2529167601616287,
+      "learning_rate": 5.012095920110245e-06,
+      "loss": 0.5074,
+      "step": 4476
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.486852481886929,
+      "learning_rate": 5.0102350121595576e-06,
+      "loss": 0.5603,
+      "step": 4477
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.98223603844178,
+      "learning_rate": 5.008374102791119e-06,
+      "loss": 0.4898,
+      "step": 4478
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.1029245482487635,
+      "learning_rate": 5.006513192262702e-06,
+      "loss": 0.3911,
+      "step": 4479
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9095703801728892,
+      "learning_rate": 5.004652280832077e-06,
+      "loss": 0.5232,
+      "step": 4480
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1305763811877187,
+      "learning_rate": 5.002791368757019e-06,
+      "loss": 0.4502,
+      "step": 4481
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0869937591936347,
+      "learning_rate": 5.000930456295302e-06,
+      "loss": 0.5178,
+      "step": 4482
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9623864090465284,
+      "learning_rate": 4.9990695437046995e-06,
+      "loss": 0.5864,
+      "step": 4483
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.149820792131405,
+      "learning_rate": 4.997208631242981e-06,
+      "loss": 0.3687,
+      "step": 4484
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.207518213410718,
+      "learning_rate": 4.995347719167924e-06,
+      "loss": 0.5253,
+      "step": 4485
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.147841557923247,
+      "learning_rate": 4.9934868077373e-06,
+      "loss": 0.5243,
+      "step": 4486
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7359989898638204,
+      "learning_rate": 4.991625897208882e-06,
+      "loss": 0.4673,
+      "step": 4487
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8249811572087307,
+      "learning_rate": 4.989764987840442e-06,
+      "loss": 0.4294,
+      "step": 4488
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.797419057797103,
+      "learning_rate": 4.987904079889756e-06,
+      "loss": 0.5135,
+      "step": 4489
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.040783142534844,
+      "learning_rate": 4.9860431736145936e-06,
+      "loss": 0.4545,
+      "step": 4490
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9480582187031563,
+      "learning_rate": 4.984182269272731e-06,
+      "loss": 0.4824,
+      "step": 4491
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9628372042110296,
+      "learning_rate": 4.982321367121935e-06,
+      "loss": 0.5102,
+      "step": 4492
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7930549930982846,
+      "learning_rate": 4.980460467419982e-06,
+      "loss": 0.6359,
+      "step": 4493
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.301676637090723,
+      "learning_rate": 4.97859957042464e-06,
+      "loss": 0.4286,
+      "step": 4494
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.755334521032017,
+      "learning_rate": 4.976738676393682e-06,
+      "loss": 0.5428,
+      "step": 4495
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.805191950343552,
+      "learning_rate": 4.9748777855848786e-06,
+      "loss": 0.4783,
+      "step": 4496
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.255393928811061,
+      "learning_rate": 4.973016898255999e-06,
+      "loss": 0.4328,
+      "step": 4497
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.566190304225009,
+      "learning_rate": 4.971156014664816e-06,
+      "loss": 0.4419,
+      "step": 4498
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.355518475569609,
+      "learning_rate": 4.969295135069096e-06,
+      "loss": 0.4963,
+      "step": 4499
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8207446087422836,
+      "learning_rate": 4.967434259726608e-06,
+      "loss": 0.4437,
+      "step": 4500
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4606170727671937,
+      "learning_rate": 4.965573388895121e-06,
+      "loss": 0.501,
+      "step": 4501
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0731705456450755,
+      "learning_rate": 4.963712522832402e-06,
+      "loss": 0.5974,
+      "step": 4502
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.251682506217783,
+      "learning_rate": 4.96185166179622e-06,
+      "loss": 0.3679,
+      "step": 4503
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.113499774786899,
+      "learning_rate": 4.959990806044338e-06,
+      "loss": 0.4769,
+      "step": 4504
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7218646069601113,
+      "learning_rate": 4.958129955834522e-06,
+      "loss": 0.5516,
+      "step": 4505
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.4019887972656226,
+      "learning_rate": 4.956269111424537e-06,
+      "loss": 0.4634,
+      "step": 4506
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.906250384427298,
+      "learning_rate": 4.954408273072148e-06,
+      "loss": 0.5187,
+      "step": 4507
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7805073601184829,
+      "learning_rate": 4.952547441035117e-06,
+      "loss": 0.5198,
+      "step": 4508
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.480392726287118,
+      "learning_rate": 4.950686615571208e-06,
+      "loss": 0.4868,
+      "step": 4509
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0388878540692223,
+      "learning_rate": 4.948825796938178e-06,
+      "loss": 0.53,
+      "step": 4510
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0388311678181474,
+      "learning_rate": 4.94696498539379e-06,
+      "loss": 0.3548,
+      "step": 4511
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3788455393514436,
+      "learning_rate": 4.945104181195803e-06,
+      "loss": 0.4972,
+      "step": 4512
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1966462424171396,
+      "learning_rate": 4.9432433846019755e-06,
+      "loss": 0.4957,
+      "step": 4513
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0229378544039935,
+      "learning_rate": 4.941382595870065e-06,
+      "loss": 0.6074,
+      "step": 4514
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.963787557226285,
+      "learning_rate": 4.939521815257826e-06,
+      "loss": 0.5408,
+      "step": 4515
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7730372046872394,
+      "learning_rate": 4.937661043023015e-06,
+      "loss": 0.5389,
+      "step": 4516
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0767720330880635,
+      "learning_rate": 4.935800279423386e-06,
+      "loss": 0.5785,
+      "step": 4517
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.244031457382286,
+      "learning_rate": 4.933939524716692e-06,
+      "loss": 0.559,
+      "step": 4518
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2581850342872487,
+      "learning_rate": 4.932078779160682e-06,
+      "loss": 0.4575,
+      "step": 4519
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9445595648487,
+      "learning_rate": 4.930218043013109e-06,
+      "loss": 0.4918,
+      "step": 4520
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7615961433681504,
+      "learning_rate": 4.92835731653172e-06,
+      "loss": 0.5325,
+      "step": 4521
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.268311646347819,
+      "learning_rate": 4.9264965999742635e-06,
+      "loss": 0.5027,
+      "step": 4522
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6008891495791047,
+      "learning_rate": 4.9246358935984854e-06,
+      "loss": 0.4844,
+      "step": 4523
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8598120213966736,
+      "learning_rate": 4.922775197662129e-06,
+      "loss": 0.4984,
+      "step": 4524
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9445755112998133,
+      "learning_rate": 4.920914512422941e-06,
+      "loss": 0.4825,
+      "step": 4525
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.852023506608896,
+      "learning_rate": 4.91905383813866e-06,
+      "loss": 0.7103,
+      "step": 4526
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2574584322351874,
+      "learning_rate": 4.917193175067026e-06,
+      "loss": 0.5179,
+      "step": 4527
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1588110019499016,
+      "learning_rate": 4.915332523465781e-06,
+      "loss": 0.5071,
+      "step": 4528
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.490698435001959,
+      "learning_rate": 4.91347188359266e-06,
+      "loss": 0.4511,
+      "step": 4529
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9384555436414646,
+      "learning_rate": 4.9116112557053976e-06,
+      "loss": 0.4885,
+      "step": 4530
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0677818202028684,
+      "learning_rate": 4.909750640061729e-06,
+      "loss": 0.4725,
+      "step": 4531
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9844513033060593,
+      "learning_rate": 4.907890036919386e-06,
+      "loss": 0.4932,
+      "step": 4532
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.175757695171607,
+      "learning_rate": 4.9060294465360994e-06,
+      "loss": 0.4785,
+      "step": 4533
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3375492718169624,
+      "learning_rate": 4.904168869169597e-06,
+      "loss": 0.5052,
+      "step": 4534
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.324275346869292,
+      "learning_rate": 4.902308305077607e-06,
+      "loss": 0.5379,
+      "step": 4535
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9748633248700438,
+      "learning_rate": 4.900447754517854e-06,
+      "loss": 0.5484,
+      "step": 4536
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8697892369333078,
+      "learning_rate": 4.898587217748059e-06,
+      "loss": 0.4755,
+      "step": 4537
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.6475403723093165,
+      "learning_rate": 4.896726695025947e-06,
+      "loss": 0.5286,
+      "step": 4538
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.284885069141002,
+      "learning_rate": 4.894866186609234e-06,
+      "loss": 0.475,
+      "step": 4539
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.04512439950645,
+      "learning_rate": 4.893005692755639e-06,
+      "loss": 0.3489,
+      "step": 4540
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7069321309387255,
+      "learning_rate": 4.8911452137228775e-06,
+      "loss": 0.4078,
+      "step": 4541
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6585698781506637,
+      "learning_rate": 4.889284749768663e-06,
+      "loss": 0.5751,
+      "step": 4542
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.609727567666265,
+      "learning_rate": 4.887424301150705e-06,
+      "loss": 0.5077,
+      "step": 4543
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0803523414464182,
+      "learning_rate": 4.885563868126713e-06,
+      "loss": 0.4209,
+      "step": 4544
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2504738381789946,
+      "learning_rate": 4.8837034509543935e-06,
+      "loss": 0.5612,
+      "step": 4545
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8966215837902091,
+      "learning_rate": 4.881843049891452e-06,
+      "loss": 0.7617,
+      "step": 4546
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9006807432456274,
+      "learning_rate": 4.879982665195591e-06,
+      "loss": 0.5009,
+      "step": 4547
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.789301169646061,
+      "learning_rate": 4.878122297124512e-06,
+      "loss": 0.6044,
+      "step": 4548
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.56379560347693,
+      "learning_rate": 4.876261945935909e-06,
+      "loss": 0.4672,
+      "step": 4549
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7069983951925045,
+      "learning_rate": 4.874401611887481e-06,
+      "loss": 0.4882,
+      "step": 4550
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.906036924900144,
+      "learning_rate": 4.872541295236919e-06,
+      "loss": 0.5419,
+      "step": 4551
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.066403291549752,
+      "learning_rate": 4.870680996241916e-06,
+      "loss": 0.6192,
+      "step": 4552
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8361159966480445,
+      "learning_rate": 4.8688207151601576e-06,
+      "loss": 0.547,
+      "step": 4553
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8914070994458013,
+      "learning_rate": 4.866960452249332e-06,
+      "loss": 0.6832,
+      "step": 4554
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0316659519095306,
+      "learning_rate": 4.865100207767121e-06,
+      "loss": 0.4352,
+      "step": 4555
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7734491793695362,
+      "learning_rate": 4.863239981971206e-06,
+      "loss": 0.492,
+      "step": 4556
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2940421480125455,
+      "learning_rate": 4.861379775119265e-06,
+      "loss": 0.4796,
+      "step": 4557
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7205222568546756,
+      "learning_rate": 4.859519587468974e-06,
+      "loss": 0.4669,
+      "step": 4558
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.870807447629867,
+      "learning_rate": 4.857659419278007e-06,
+      "loss": 0.5142,
+      "step": 4559
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.166466914311534,
+      "learning_rate": 4.855799270804031e-06,
+      "loss": 0.5122,
+      "step": 4560
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1895022758410314,
+      "learning_rate": 4.853939142304714e-06,
+      "loss": 0.5253,
+      "step": 4561
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9920238045271101,
+      "learning_rate": 4.852079034037722e-06,
+      "loss": 0.5338,
+      "step": 4562
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8749201316861372,
+      "learning_rate": 4.850218946260717e-06,
+      "loss": 0.5309,
+      "step": 4563
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.209211279846637,
+      "learning_rate": 4.848358879231358e-06,
+      "loss": 0.4328,
+      "step": 4564
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.131699116704348,
+      "learning_rate": 4.8464988332073e-06,
+      "loss": 0.5443,
+      "step": 4565
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.368353932312741,
+      "learning_rate": 4.844638808446198e-06,
+      "loss": 0.5409,
+      "step": 4566
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.124976566493968,
+      "learning_rate": 4.8427788052057e-06,
+      "loss": 0.5216,
+      "step": 4567
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.044246560529135,
+      "learning_rate": 4.840918823743456e-06,
+      "loss": 0.5328,
+      "step": 4568
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.315500595247249,
+      "learning_rate": 4.839058864317109e-06,
+      "loss": 0.5294,
+      "step": 4569
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1159384416707363,
+      "learning_rate": 4.837198927184299e-06,
+      "loss": 0.4915,
+      "step": 4570
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1681512116907493,
+      "learning_rate": 4.835339012602666e-06,
+      "loss": 0.4837,
+      "step": 4571
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.784103190926599,
+      "learning_rate": 4.833479120829845e-06,
+      "loss": 0.4699,
+      "step": 4572
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.521894210548586,
+      "learning_rate": 4.8316192521234665e-06,
+      "loss": 0.4444,
+      "step": 4573
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5842375522375027,
+      "learning_rate": 4.82975940674116e-06,
+      "loss": 0.4297,
+      "step": 4574
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.232057758251001,
+      "learning_rate": 4.8278995849405515e-06,
+      "loss": 0.5342,
+      "step": 4575
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0055726262517743,
+      "learning_rate": 4.826039786979264e-06,
+      "loss": 0.4726,
+      "step": 4576
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6532190386300165,
+      "learning_rate": 4.824180013114914e-06,
+      "loss": 0.498,
+      "step": 4577
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3489841533554303,
+      "learning_rate": 4.822320263605118e-06,
+      "loss": 0.5791,
+      "step": 4578
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2678441188617304,
+      "learning_rate": 4.820460538707487e-06,
+      "loss": 0.5331,
+      "step": 4579
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7052086344595665,
+      "learning_rate": 4.818600838679633e-06,
+      "loss": 0.4619,
+      "step": 4580
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.093347669292747,
+      "learning_rate": 4.816741163779158e-06,
+      "loss": 0.5093,
+      "step": 4581
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6861824230104265,
+      "learning_rate": 4.814881514263666e-06,
+      "loss": 0.5033,
+      "step": 4582
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.067725485917308,
+      "learning_rate": 4.813021890390754e-06,
+      "loss": 0.4335,
+      "step": 4583
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4929502613591383,
+      "learning_rate": 4.811162292418018e-06,
+      "loss": 0.548,
+      "step": 4584
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6736121022145403,
+      "learning_rate": 4.809302720603049e-06,
+      "loss": 0.5388,
+      "step": 4585
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.495052710963959,
+      "learning_rate": 4.807443175203432e-06,
+      "loss": 0.479,
+      "step": 4586
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.014220067342264,
+      "learning_rate": 4.805583656476755e-06,
+      "loss": 0.4779,
+      "step": 4587
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.99278720798132,
+      "learning_rate": 4.803724164680596e-06,
+      "loss": 0.5119,
+      "step": 4588
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9095436280475346,
+      "learning_rate": 4.801864700072531e-06,
+      "loss": 0.521,
+      "step": 4589
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.433682315782996,
+      "learning_rate": 4.800005262910135e-06,
+      "loss": 0.5601,
+      "step": 4590
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.480523341064522,
+      "learning_rate": 4.798145853450973e-06,
+      "loss": 0.5557,
+      "step": 4591
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.596657736476968,
+      "learning_rate": 4.796286471952615e-06,
+      "loss": 0.5714,
+      "step": 4592
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.826378276861144,
+      "learning_rate": 4.794427118672619e-06,
+      "loss": 0.611,
+      "step": 4593
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.088002845855787,
+      "learning_rate": 4.7925677938685425e-06,
+      "loss": 0.5757,
+      "step": 4594
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.553614500472137,
+      "learning_rate": 4.7907084977979384e-06,
+      "loss": 0.4282,
+      "step": 4595
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8350135941282417,
+      "learning_rate": 4.7888492307183575e-06,
+      "loss": 0.5079,
+      "step": 4596
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1944624542917324,
+      "learning_rate": 4.786989992887344e-06,
+      "loss": 0.5607,
+      "step": 4597
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.218859580914836,
+      "learning_rate": 4.785130784562441e-06,
+      "loss": 0.4979,
+      "step": 4598
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0486123522148705,
+      "learning_rate": 4.783271606001183e-06,
+      "loss": 0.4847,
+      "step": 4599
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4057492714364312,
+      "learning_rate": 4.7814124574611055e-06,
+      "loss": 0.5351,
+      "step": 4600
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8034945482329836,
+      "learning_rate": 4.7795533391997365e-06,
+      "loss": 0.5282,
+      "step": 4601
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8616072470890517,
+      "learning_rate": 4.777694251474601e-06,
+      "loss": 0.4732,
+      "step": 4602
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7249869478548603,
+      "learning_rate": 4.775835194543219e-06,
+      "loss": 0.4924,
+      "step": 4603
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.925363412057559,
+      "learning_rate": 4.773976168663107e-06,
+      "loss": 0.4619,
+      "step": 4604
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.769554485785394,
+      "learning_rate": 4.772117174091776e-06,
+      "loss": 0.4432,
+      "step": 4605
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3072337862537524,
+      "learning_rate": 4.770258211086735e-06,
+      "loss": 0.456,
+      "step": 4606
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7009986533953696,
+      "learning_rate": 4.768399279905489e-06,
+      "loss": 0.5658,
+      "step": 4607
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0266423701437972,
+      "learning_rate": 4.766540380805533e-06,
+      "loss": 0.4838,
+      "step": 4608
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5523636747338307,
+      "learning_rate": 4.7646815140443625e-06,
+      "loss": 0.4747,
+      "step": 4609
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.544121167235555,
+      "learning_rate": 4.762822679879469e-06,
+      "loss": 0.6035,
+      "step": 4610
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.677293907848666,
+      "learning_rate": 4.7609638785683365e-06,
+      "loss": 0.5121,
+      "step": 4611
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9167508235507689,
+      "learning_rate": 4.759105110368446e-06,
+      "loss": 0.4309,
+      "step": 4612
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.09105511713776,
+      "learning_rate": 4.757246375537275e-06,
+      "loss": 0.5393,
+      "step": 4613
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8331872329457393,
+      "learning_rate": 4.755387674332292e-06,
+      "loss": 0.5051,
+      "step": 4614
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3796635600966725,
+      "learning_rate": 4.753529007010967e-06,
+      "loss": 0.6322,
+      "step": 4615
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5172924303763975,
+      "learning_rate": 4.75167037383076e-06,
+      "loss": 0.5493,
+      "step": 4616
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8632548689985227,
+      "learning_rate": 4.749811775049131e-06,
+      "loss": 0.4864,
+      "step": 4617
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1926319622923502,
+      "learning_rate": 4.7479532109235315e-06,
+      "loss": 0.5256,
+      "step": 4618
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9258503614453266,
+      "learning_rate": 4.746094681711409e-06,
+      "loss": 0.5179,
+      "step": 4619
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0129673265110566,
+      "learning_rate": 4.744236187670208e-06,
+      "loss": 0.4516,
+      "step": 4620
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7782092155735626,
+      "learning_rate": 4.742377729057366e-06,
+      "loss": 0.6714,
+      "step": 4621
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0880325364239374,
+      "learning_rate": 4.740519306130315e-06,
+      "loss": 0.519,
+      "step": 4622
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.445679293583082,
+      "learning_rate": 4.738660919146486e-06,
+      "loss": 0.5449,
+      "step": 4623
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9905400733509697,
+      "learning_rate": 4.736802568363301e-06,
+      "loss": 0.5069,
+      "step": 4624
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 19.65539353140791,
+      "learning_rate": 4.734944254038178e-06,
+      "loss": 0.4445,
+      "step": 4625
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.844515877516487,
+      "learning_rate": 4.733085976428535e-06,
+      "loss": 0.493,
+      "step": 4626
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4405149712545886,
+      "learning_rate": 4.7312277357917726e-06,
+      "loss": 0.5364,
+      "step": 4627
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8223939692917552,
+      "learning_rate": 4.7293695323852975e-06,
+      "loss": 0.4566,
+      "step": 4628
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9088404911423766,
+      "learning_rate": 4.7275113664665085e-06,
+      "loss": 0.4394,
+      "step": 4629
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8776388690910601,
+      "learning_rate": 4.725653238292798e-06,
+      "loss": 0.5273,
+      "step": 4630
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.820313119531497,
+      "learning_rate": 4.723795148121553e-06,
+      "loss": 0.5172,
+      "step": 4631
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4891082787747747,
+      "learning_rate": 4.721937096210156e-06,
+      "loss": 0.4589,
+      "step": 4632
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1170687152084224,
+      "learning_rate": 4.720079082815986e-06,
+      "loss": 0.4815,
+      "step": 4633
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7654492613414394,
+      "learning_rate": 4.718221108196412e-06,
+      "loss": 0.4667,
+      "step": 4634
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7966705051500982,
+      "learning_rate": 4.7163631726088e-06,
+      "loss": 0.4898,
+      "step": 4635
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9949175556003313,
+      "learning_rate": 4.714505276310513e-06,
+      "loss": 0.5029,
+      "step": 4636
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2710048799635425,
+      "learning_rate": 4.712647419558906e-06,
+      "loss": 0.4146,
+      "step": 4637
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1212367985990657,
+      "learning_rate": 4.710789602611328e-06,
+      "loss": 0.5525,
+      "step": 4638
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0497723041506144,
+      "learning_rate": 4.708931825725125e-06,
+      "loss": 0.5267,
+      "step": 4639
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.749457257831849,
+      "learning_rate": 4.707074089157634e-06,
+      "loss": 0.6715,
+      "step": 4640
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2923524231462693,
+      "learning_rate": 4.70521639316619e-06,
+      "loss": 0.5225,
+      "step": 4641
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.038078910233903,
+      "learning_rate": 4.703358738008121e-06,
+      "loss": 0.565,
+      "step": 4642
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.9779677378659004,
+      "learning_rate": 4.701501123940749e-06,
+      "loss": 0.4288,
+      "step": 4643
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.782312364187964,
+      "learning_rate": 4.699643551221388e-06,
+      "loss": 0.539,
+      "step": 4644
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.505084358588231,
+      "learning_rate": 4.697786020107351e-06,
+      "loss": 0.566,
+      "step": 4645
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9303263617525803,
+      "learning_rate": 4.695928530855941e-06,
+      "loss": 0.4211,
+      "step": 4646
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8939690580516948,
+      "learning_rate": 4.694071083724459e-06,
+      "loss": 0.5431,
+      "step": 4647
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0880005195890625,
+      "learning_rate": 4.692213678970197e-06,
+      "loss": 0.4942,
+      "step": 4648
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1711645610942836,
+      "learning_rate": 4.6903563168504435e-06,
+      "loss": 0.5048,
+      "step": 4649
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9410545037525402,
+      "learning_rate": 4.688498997622478e-06,
+      "loss": 0.429,
+      "step": 4650
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1782522782555915,
+      "learning_rate": 4.686641721543579e-06,
+      "loss": 0.4882,
+      "step": 4651
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.317031666319926,
+      "learning_rate": 4.684784488871014e-06,
+      "loss": 0.484,
+      "step": 4652
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.280647371521375,
+      "learning_rate": 4.682927299862048e-06,
+      "loss": 0.5494,
+      "step": 4653
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.252569718288382,
+      "learning_rate": 4.6810701547739364e-06,
+      "loss": 0.4461,
+      "step": 4654
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8265080809784993,
+      "learning_rate": 4.679213053863931e-06,
+      "loss": 0.5105,
+      "step": 4655
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8756012360122547,
+      "learning_rate": 4.677355997389279e-06,
+      "loss": 0.702,
+      "step": 4656
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.7162773996819185,
+      "learning_rate": 4.675498985607217e-06,
+      "loss": 0.5218,
+      "step": 4657
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6742250478018796,
+      "learning_rate": 4.673642018774981e-06,
+      "loss": 0.3734,
+      "step": 4658
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.946878799358248,
+      "learning_rate": 4.671785097149796e-06,
+      "loss": 0.4704,
+      "step": 4659
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4477476787025316,
+      "learning_rate": 4.669928220988883e-06,
+      "loss": 0.4783,
+      "step": 4660
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2596026477671045,
+      "learning_rate": 4.668071390549454e-06,
+      "loss": 0.5067,
+      "step": 4661
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.5088085501143538,
+      "learning_rate": 4.666214606088719e-06,
+      "loss": 0.5386,
+      "step": 4662
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1212458770516136,
+      "learning_rate": 4.664357867863879e-06,
+      "loss": 0.5175,
+      "step": 4663
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9348256038658231,
+      "learning_rate": 4.6625011761321285e-06,
+      "loss": 0.5458,
+      "step": 4664
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.226950920311374,
+      "learning_rate": 4.6606445311506564e-06,
+      "loss": 0.5777,
+      "step": 4665
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.3178438138762445,
+      "learning_rate": 4.6587879331766465e-06,
+      "loss": 0.5639,
+      "step": 4666
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4620681548092764,
+      "learning_rate": 4.656931382467272e-06,
+      "loss": 0.5633,
+      "step": 4667
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8010489685360302,
+      "learning_rate": 4.655074879279703e-06,
+      "loss": 0.5737,
+      "step": 4668
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8204130338659306,
+      "learning_rate": 4.653218423871102e-06,
+      "loss": 0.5318,
+      "step": 4669
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1273226926128994,
+      "learning_rate": 4.651362016498625e-06,
+      "loss": 0.5475,
+      "step": 4670
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3686236495776534,
+      "learning_rate": 4.64950565741942e-06,
+      "loss": 0.4927,
+      "step": 4671
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.8871602855563125,
+      "learning_rate": 4.647649346890633e-06,
+      "loss": 0.5704,
+      "step": 4672
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3473391920922544,
+      "learning_rate": 4.645793085169396e-06,
+      "loss": 0.5289,
+      "step": 4673
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2052704581082265,
+      "learning_rate": 4.64393687251284e-06,
+      "loss": 0.4617,
+      "step": 4674
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.461846910497991,
+      "learning_rate": 4.6420807091780855e-06,
+      "loss": 0.5598,
+      "step": 4675
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.026016128274953,
+      "learning_rate": 4.640224595422251e-06,
+      "loss": 0.469,
+      "step": 4676
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6290566132188997,
+      "learning_rate": 4.638368531502442e-06,
+      "loss": 0.4079,
+      "step": 4677
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.398560750017565,
+      "learning_rate": 4.636512517675763e-06,
+      "loss": 0.5602,
+      "step": 4678
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2580981286671573,
+      "learning_rate": 4.634656554199306e-06,
+      "loss": 0.5387,
+      "step": 4679
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.257152951874033,
+      "learning_rate": 4.632800641330159e-06,
+      "loss": 0.5057,
+      "step": 4680
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1334530992862666,
+      "learning_rate": 4.6309447793254046e-06,
+      "loss": 0.467,
+      "step": 4681
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.220644466957239,
+      "learning_rate": 4.629088968442115e-06,
+      "loss": 0.6066,
+      "step": 4682
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0690202270269302,
+      "learning_rate": 4.627233208937357e-06,
+      "loss": 0.4983,
+      "step": 4683
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8511330807711688,
+      "learning_rate": 4.62537750106819e-06,
+      "loss": 0.5837,
+      "step": 4684
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9143723198718514,
+      "learning_rate": 4.623521845091667e-06,
+      "loss": 0.5368,
+      "step": 4685
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.025459673779297,
+      "learning_rate": 4.6216662412648325e-06,
+      "loss": 0.5198,
+      "step": 4686
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9185796048543928,
+      "learning_rate": 4.619810689844726e-06,
+      "loss": 0.4716,
+      "step": 4687
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2758759579198133,
+      "learning_rate": 4.6179551910883755e-06,
+      "loss": 0.497,
+      "step": 4688
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0553296301338344,
+      "learning_rate": 4.616099745252806e-06,
+      "loss": 0.4888,
+      "step": 4689
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8616828258925835,
+      "learning_rate": 4.6142443525950345e-06,
+      "loss": 0.5117,
+      "step": 4690
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4366270225941506,
+      "learning_rate": 4.612389013372068e-06,
+      "loss": 0.4398,
+      "step": 4691
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.63250381770413,
+      "learning_rate": 4.610533727840908e-06,
+      "loss": 0.4308,
+      "step": 4692
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3914458142513424,
+      "learning_rate": 4.608678496258549e-06,
+      "loss": 0.5294,
+      "step": 4693
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.0856826128306705,
+      "learning_rate": 4.606823318881977e-06,
+      "loss": 0.5484,
+      "step": 4694
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8246530527529177,
+      "learning_rate": 4.6049681959681696e-06,
+      "loss": 0.5372,
+      "step": 4695
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1132533612609654,
+      "learning_rate": 4.6031131277741e-06,
+      "loss": 0.4991,
+      "step": 4696
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.155922554314053,
+      "learning_rate": 4.601258114556731e-06,
+      "loss": 0.3537,
+      "step": 4697
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.327295075291843,
+      "learning_rate": 4.59940315657302e-06,
+      "loss": 0.4251,
+      "step": 4698
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.314472832819074,
+      "learning_rate": 4.597548254079913e-06,
+      "loss": 0.509,
+      "step": 4699
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.976844568459057,
+      "learning_rate": 4.595693407334352e-06,
+      "loss": 0.4503,
+      "step": 4700
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.855313287648476,
+      "learning_rate": 4.593838616593271e-06,
+      "loss": 0.5359,
+      "step": 4701
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9107360006439185,
+      "learning_rate": 4.591983882113594e-06,
+      "loss": 0.5337,
+      "step": 4702
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.603074364611393,
+      "learning_rate": 4.590129204152239e-06,
+      "loss": 0.6203,
+      "step": 4703
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2543626661564793,
+      "learning_rate": 4.588274582966116e-06,
+      "loss": 0.4764,
+      "step": 4704
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.626338927292518,
+      "learning_rate": 4.586420018812125e-06,
+      "loss": 0.5262,
+      "step": 4705
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.078036401999381,
+      "learning_rate": 4.5845655119471625e-06,
+      "loss": 0.4957,
+      "step": 4706
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3990353860007314,
+      "learning_rate": 4.582711062628114e-06,
+      "loss": 0.5579,
+      "step": 4707
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.773348682252783,
+      "learning_rate": 4.5808566711118555e-06,
+      "loss": 0.4949,
+      "step": 4708
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8547930519693662,
+      "learning_rate": 4.57900233765526e-06,
+      "loss": 0.7306,
+      "step": 4709
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.1824132174941004,
+      "learning_rate": 4.577148062515186e-06,
+      "loss": 0.5513,
+      "step": 4710
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.523596564302978,
+      "learning_rate": 4.575293845948492e-06,
+      "loss": 0.5283,
+      "step": 4711
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.392486823652076,
+      "learning_rate": 4.573439688212018e-06,
+      "loss": 0.5184,
+      "step": 4712
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.503718370629469,
+      "learning_rate": 4.571585589562606e-06,
+      "loss": 0.5073,
+      "step": 4713
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.489185988165088,
+      "learning_rate": 4.569731550257083e-06,
+      "loss": 0.4307,
+      "step": 4714
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.121486689174363,
+      "learning_rate": 4.567877570552272e-06,
+      "loss": 0.4524,
+      "step": 4715
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9717820302544427,
+      "learning_rate": 4.566023650704984e-06,
+      "loss": 0.4249,
+      "step": 4716
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6455486557773136,
+      "learning_rate": 4.564169790972025e-06,
+      "loss": 0.5129,
+      "step": 4717
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9945622416665105,
+      "learning_rate": 4.562315991610192e-06,
+      "loss": 0.547,
+      "step": 4718
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.316971502978805,
+      "learning_rate": 4.560462252876271e-06,
+      "loss": 0.4596,
+      "step": 4719
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9399748608543934,
+      "learning_rate": 4.558608575027043e-06,
+      "loss": 0.4929,
+      "step": 4720
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0964507806084383,
+      "learning_rate": 4.5567549583192785e-06,
+      "loss": 0.4654,
+      "step": 4721
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2831027848607928,
+      "learning_rate": 4.55490140300974e-06,
+      "loss": 0.5135,
+      "step": 4722
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.56862221232208,
+      "learning_rate": 4.553047909355183e-06,
+      "loss": 0.465,
+      "step": 4723
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9180445005728384,
+      "learning_rate": 4.551194477612351e-06,
+      "loss": 0.6142,
+      "step": 4724
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3303072432971925,
+      "learning_rate": 4.549341108037984e-06,
+      "loss": 0.4185,
+      "step": 4725
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0118920473919855,
+      "learning_rate": 4.547487800888808e-06,
+      "loss": 0.4956,
+      "step": 4726
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8972623843282614,
+      "learning_rate": 4.545634556421542e-06,
+      "loss": 0.7837,
+      "step": 4727
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.0515014862573633,
+      "learning_rate": 4.543781374892902e-06,
+      "loss": 0.5154,
+      "step": 4728
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.4202463000809455,
+      "learning_rate": 4.541928256559584e-06,
+      "loss": 0.5046,
+      "step": 4729
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2805409575343534,
+      "learning_rate": 4.5400752016782854e-06,
+      "loss": 0.4745,
+      "step": 4730
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7229789282570054,
+      "learning_rate": 4.53822221050569e-06,
+      "loss": 0.5919,
+      "step": 4731
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.709486696510282,
+      "learning_rate": 4.536369283298474e-06,
+      "loss": 0.5141,
+      "step": 4732
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.374102442701293,
+      "learning_rate": 4.534516420313304e-06,
+      "loss": 0.5483,
+      "step": 4733
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.440598460649114,
+      "learning_rate": 4.5326636218068394e-06,
+      "loss": 0.4127,
+      "step": 4734
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.210959331333927,
+      "learning_rate": 4.530810888035729e-06,
+      "loss": 0.453,
+      "step": 4735
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0923357188717806,
+      "learning_rate": 4.528958219256613e-06,
+      "loss": 0.432,
+      "step": 4736
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.5166167372329564,
+      "learning_rate": 4.527105615726124e-06,
+      "loss": 0.4711,
+      "step": 4737
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1567139485835836,
+      "learning_rate": 4.525253077700882e-06,
+      "loss": 0.5106,
+      "step": 4738
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.1984622235676508,
+      "learning_rate": 4.523400605437501e-06,
+      "loss": 0.4556,
+      "step": 4739
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.065987980132794,
+      "learning_rate": 4.521548199192587e-06,
+      "loss": 0.4854,
+      "step": 4740
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1206497168478498,
+      "learning_rate": 4.519695859222733e-06,
+      "loss": 0.3951,
+      "step": 4741
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.15208434725664,
+      "learning_rate": 4.517843585784525e-06,
+      "loss": 0.4565,
+      "step": 4742
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7926666702326282,
+      "learning_rate": 4.515991379134539e-06,
+      "loss": 0.7006,
+      "step": 4743
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.770747463795314,
+      "learning_rate": 4.514139239529345e-06,
+      "loss": 0.5386,
+      "step": 4744
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.734926305865901,
+      "learning_rate": 4.512287167225501e-06,
+      "loss": 0.5087,
+      "step": 4745
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9792870976475687,
+      "learning_rate": 4.510435162479551e-06,
+      "loss": 0.5288,
+      "step": 4746
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1211661383539457,
+      "learning_rate": 4.508583225548039e-06,
+      "loss": 0.5292,
+      "step": 4747
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8406422465367296,
+      "learning_rate": 4.506731356687493e-06,
+      "loss": 0.5636,
+      "step": 4748
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.426633954291697,
+      "learning_rate": 4.504879556154433e-06,
+      "loss": 0.4705,
+      "step": 4749
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.951740043780009,
+      "learning_rate": 4.5030278242053725e-06,
+      "loss": 0.5418,
+      "step": 4750
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2739318165608915,
+      "learning_rate": 4.501176161096811e-06,
+      "loss": 0.5033,
+      "step": 4751
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.260449352853781,
+      "learning_rate": 4.499324567085242e-06,
+      "loss": 0.5161,
+      "step": 4752
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0218842622351834,
+      "learning_rate": 4.497473042427147e-06,
+      "loss": 0.5561,
+      "step": 4753
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.335807955174927,
+      "learning_rate": 4.495621587378998e-06,
+      "loss": 0.5447,
+      "step": 4754
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2378513670920257,
+      "learning_rate": 4.493770202197261e-06,
+      "loss": 0.5161,
+      "step": 4755
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.637832142203877,
+      "learning_rate": 4.491918887138387e-06,
+      "loss": 0.586,
+      "step": 4756
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6462058308685985,
+      "learning_rate": 4.490067642458822e-06,
+      "loss": 0.5833,
+      "step": 4757
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0283868772182996,
+      "learning_rate": 4.4882164684149975e-06,
+      "loss": 0.4125,
+      "step": 4758
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.866767667799671,
+      "learning_rate": 4.4863653652633396e-06,
+      "loss": 0.5244,
+      "step": 4759
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.178485740279844,
+      "learning_rate": 4.484514333260262e-06,
+      "loss": 0.5376,
+      "step": 4760
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1838489489424098,
+      "learning_rate": 4.48266337266217e-06,
+      "loss": 0.5047,
+      "step": 4761
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2007938235369613,
+      "learning_rate": 4.480812483725458e-06,
+      "loss": 0.5146,
+      "step": 4762
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4714701208204457,
+      "learning_rate": 4.478961666706512e-06,
+      "loss": 0.5874,
+      "step": 4763
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.579022381833512,
+      "learning_rate": 4.477110921861704e-06,
+      "loss": 0.5193,
+      "step": 4764
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9462746029049274,
+      "learning_rate": 4.475260249447401e-06,
+      "loss": 0.4654,
+      "step": 4765
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3376045446806373,
+      "learning_rate": 4.473409649719958e-06,
+      "loss": 0.4594,
+      "step": 4766
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.172075675309851,
+      "learning_rate": 4.471559122935718e-06,
+      "loss": 0.4822,
+      "step": 4767
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2756047681635656,
+      "learning_rate": 4.469708669351017e-06,
+      "loss": 0.5517,
+      "step": 4768
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8420447069086467,
+      "learning_rate": 4.467858289222179e-06,
+      "loss": 0.5175,
+      "step": 4769
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.4735135764737275,
+      "learning_rate": 4.4660079828055195e-06,
+      "loss": 0.4699,
+      "step": 4770
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2244726760047078,
+      "learning_rate": 4.464157750357341e-06,
+      "loss": 0.434,
+      "step": 4771
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5712301190342477,
+      "learning_rate": 4.462307592133938e-06,
+      "loss": 0.4684,
+      "step": 4772
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.871158163194741,
+      "learning_rate": 4.460457508391595e-06,
+      "loss": 0.5143,
+      "step": 4773
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8672411278905168,
+      "learning_rate": 4.458607499386584e-06,
+      "loss": 0.4884,
+      "step": 4774
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6948690898475525,
+      "learning_rate": 4.456757565375168e-06,
+      "loss": 0.527,
+      "step": 4775
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.242172624618669,
+      "learning_rate": 4.4549077066136e-06,
+      "loss": 0.5039,
+      "step": 4776
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.733169128846832,
+      "learning_rate": 4.4530579233581235e-06,
+      "loss": 0.5553,
+      "step": 4777
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.122524178534542,
+      "learning_rate": 4.451208215864969e-06,
+      "loss": 0.4604,
+      "step": 4778
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1732948956991884,
+      "learning_rate": 4.449358584390357e-06,
+      "loss": 0.4868,
+      "step": 4779
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.089888294581923,
+      "learning_rate": 4.4475090291904975e-06,
+      "loss": 0.5874,
+      "step": 4780
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6811187016555498,
+      "learning_rate": 4.445659550521591e-06,
+      "loss": 0.4389,
+      "step": 4781
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.954889852755012,
+      "learning_rate": 4.443810148639828e-06,
+      "loss": 0.483,
+      "step": 4782
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9926765029938955,
+      "learning_rate": 4.441960823801389e-06,
+      "loss": 0.471,
+      "step": 4783
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0117538880195527,
+      "learning_rate": 4.440111576262438e-06,
+      "loss": 0.5532,
+      "step": 4784
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.015242091584923,
+      "learning_rate": 4.438262406279134e-06,
+      "loss": 0.4926,
+      "step": 4785
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.268574524001044,
+      "learning_rate": 4.436413314107625e-06,
+      "loss": 0.5605,
+      "step": 4786
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.152610090676576,
+      "learning_rate": 4.434564300004046e-06,
+      "loss": 0.4431,
+      "step": 4787
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1769911205351855,
+      "learning_rate": 4.4327153642245215e-06,
+      "loss": 0.4975,
+      "step": 4788
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9946367912288085,
+      "learning_rate": 4.430866507025167e-06,
+      "loss": 0.4343,
+      "step": 4789
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9959800285746205,
+      "learning_rate": 4.429017728662086e-06,
+      "loss": 0.4843,
+      "step": 4790
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.585732296804081,
+      "learning_rate": 4.42716902939137e-06,
+      "loss": 0.4473,
+      "step": 4791
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.014648126681179,
+      "learning_rate": 4.425320409469101e-06,
+      "loss": 0.4318,
+      "step": 4792
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.038648033470926,
+      "learning_rate": 4.423471869151348e-06,
+      "loss": 0.4484,
+      "step": 4793
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1480448448808165,
+      "learning_rate": 4.421623408694173e-06,
+      "loss": 0.5539,
+      "step": 4794
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.778412486211518,
+      "learning_rate": 4.419775028353625e-06,
+      "loss": 0.447,
+      "step": 4795
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.074946855336135,
+      "learning_rate": 4.417926728385738e-06,
+      "loss": 0.499,
+      "step": 4796
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.386830510133657,
+      "learning_rate": 4.416078509046539e-06,
+      "loss": 0.4483,
+      "step": 4797
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.068768441262766,
+      "learning_rate": 4.4142303705920446e-06,
+      "loss": 0.4458,
+      "step": 4798
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3155004263275814,
+      "learning_rate": 4.412382313278257e-06,
+      "loss": 0.4361,
+      "step": 4799
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4602844358986777,
+      "learning_rate": 4.41053433736117e-06,
+      "loss": 0.5155,
+      "step": 4800
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.017909658723393,
+      "learning_rate": 4.408686443096765e-06,
+      "loss": 0.4961,
+      "step": 4801
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.227753870429378,
+      "learning_rate": 4.40683863074101e-06,
+      "loss": 0.486,
+      "step": 4802
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.720301005468441,
+      "learning_rate": 4.404990900549867e-06,
+      "loss": 0.4686,
+      "step": 4803
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.877168920792929,
+      "learning_rate": 4.403143252779281e-06,
+      "loss": 0.4592,
+      "step": 4804
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.702830868530728,
+      "learning_rate": 4.401295687685188e-06,
+      "loss": 0.4838,
+      "step": 4805
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2172314478687962,
+      "learning_rate": 4.399448205523512e-06,
+      "loss": 0.6032,
+      "step": 4806
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7116250228966026,
+      "learning_rate": 4.397600806550167e-06,
+      "loss": 0.4637,
+      "step": 4807
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8955644342042355,
+      "learning_rate": 4.395753491021053e-06,
+      "loss": 0.6679,
+      "step": 4808
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.968458394875596,
+      "learning_rate": 4.3939062591920625e-06,
+      "loss": 0.4265,
+      "step": 4809
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3346029719916306,
+      "learning_rate": 4.392059111319071e-06,
+      "loss": 0.5031,
+      "step": 4810
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.041066728892029,
+      "learning_rate": 4.3902120476579465e-06,
+      "loss": 0.4575,
+      "step": 4811
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.41872507370031,
+      "learning_rate": 4.388365068464545e-06,
+      "loss": 0.5538,
+      "step": 4812
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.498683019319059,
+      "learning_rate": 4.386518173994706e-06,
+      "loss": 0.504,
+      "step": 4813
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7280770441505457,
+      "learning_rate": 4.3846713645042646e-06,
+      "loss": 0.5821,
+      "step": 4814
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.14104253139992,
+      "learning_rate": 4.382824640249038e-06,
+      "loss": 0.4403,
+      "step": 4815
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1232688352886293,
+      "learning_rate": 4.380978001484836e-06,
+      "loss": 0.4647,
+      "step": 4816
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1331989964252442,
+      "learning_rate": 4.3791314484674545e-06,
+      "loss": 0.6034,
+      "step": 4817
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.904910913570794,
+      "learning_rate": 4.377284981452676e-06,
+      "loss": 0.4581,
+      "step": 4818
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6381239807927757,
+      "learning_rate": 4.375438600696274e-06,
+      "loss": 0.4858,
+      "step": 4819
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9451972938654336,
+      "learning_rate": 4.3735923064540094e-06,
+      "loss": 0.5459,
+      "step": 4820
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5070911339774806,
+      "learning_rate": 4.37174609898163e-06,
+      "loss": 0.5499,
+      "step": 4821
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.286754654617083,
+      "learning_rate": 4.369899978534873e-06,
+      "loss": 0.397,
+      "step": 4822
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9290064858255627,
+      "learning_rate": 4.368053945369461e-06,
+      "loss": 0.5227,
+      "step": 4823
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8849522859481973,
+      "learning_rate": 4.366207999741107e-06,
+      "loss": 0.4666,
+      "step": 4824
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8134678902539303,
+      "learning_rate": 4.364362141905512e-06,
+      "loss": 0.4364,
+      "step": 4825
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9787003476236429,
+      "learning_rate": 4.362516372118362e-06,
+      "loss": 0.5305,
+      "step": 4826
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8621097400659674,
+      "learning_rate": 4.360670690635334e-06,
+      "loss": 0.4453,
+      "step": 4827
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4536719025457447,
+      "learning_rate": 4.358825097712091e-06,
+      "loss": 0.4602,
+      "step": 4828
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.981973745978072,
+      "learning_rate": 4.356979593604286e-06,
+      "loss": 0.5247,
+      "step": 4829
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0018656131336146,
+      "learning_rate": 4.3551341785675546e-06,
+      "loss": 0.4626,
+      "step": 4830
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8068196070709868,
+      "learning_rate": 4.353288852857525e-06,
+      "loss": 0.4186,
+      "step": 4831
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8816722929981592,
+      "learning_rate": 4.351443616729812e-06,
+      "loss": 0.693,
+      "step": 4832
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8971294030527226,
+      "learning_rate": 4.349598470440016e-06,
+      "loss": 0.719,
+      "step": 4833
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8088481941079637,
+      "learning_rate": 4.3477534142437285e-06,
+      "loss": 0.5122,
+      "step": 4834
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8982260238549702,
+      "learning_rate": 4.345908448396524e-06,
+      "loss": 0.5487,
+      "step": 4835
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4202194473250307,
+      "learning_rate": 4.344063573153969e-06,
+      "loss": 0.5049,
+      "step": 4836
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.480662746767606,
+      "learning_rate": 4.342218788771614e-06,
+      "loss": 0.5098,
+      "step": 4837
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6324844766455513,
+      "learning_rate": 4.340374095504997e-06,
+      "loss": 0.4112,
+      "step": 4838
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0567558911709756,
+      "learning_rate": 4.338529493609647e-06,
+      "loss": 0.4768,
+      "step": 4839
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.926953251717545,
+      "learning_rate": 4.336684983341077e-06,
+      "loss": 0.4977,
+      "step": 4840
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0242083615998916,
+      "learning_rate": 4.334840564954789e-06,
+      "loss": 0.54,
+      "step": 4841
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.430326481099695,
+      "learning_rate": 4.3329962387062704e-06,
+      "loss": 0.4716,
+      "step": 4842
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6057779262389331,
+      "learning_rate": 4.331152004850997e-06,
+      "loss": 0.3888,
+      "step": 4843
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1191325288593066,
+      "learning_rate": 4.329307863644432e-06,
+      "loss": 0.5567,
+      "step": 4844
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0972236161890083,
+      "learning_rate": 4.327463815342025e-06,
+      "loss": 0.6165,
+      "step": 4845
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.04305921417872,
+      "learning_rate": 4.325619860199216e-06,
+      "loss": 0.5028,
+      "step": 4846
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.064445942496861,
+      "learning_rate": 4.323775998471426e-06,
+      "loss": 0.5568,
+      "step": 4847
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9356315393285697,
+      "learning_rate": 4.321932230414067e-06,
+      "loss": 0.4771,
+      "step": 4848
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8171768066015725,
+      "learning_rate": 4.320088556282539e-06,
+      "loss": 0.6486,
+      "step": 4849
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.354030690302065,
+      "learning_rate": 4.318244976332225e-06,
+      "loss": 0.5592,
+      "step": 4850
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8994471412906808,
+      "learning_rate": 4.316401490818499e-06,
+      "loss": 0.4806,
+      "step": 4851
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.957241280352434,
+      "learning_rate": 4.3145580999967205e-06,
+      "loss": 0.4902,
+      "step": 4852
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.109548970057433,
+      "learning_rate": 4.312714804122235e-06,
+      "loss": 0.4725,
+      "step": 4853
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.78084388797687,
+      "learning_rate": 4.310871603450376e-06,
+      "loss": 0.4891,
+      "step": 4854
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.24314626112775,
+      "learning_rate": 4.309028498236462e-06,
+      "loss": 0.5026,
+      "step": 4855
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9087017173502536,
+      "learning_rate": 4.307185488735802e-06,
+      "loss": 0.5205,
+      "step": 4856
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2605588684019096,
+      "learning_rate": 4.305342575203688e-06,
+      "loss": 0.4799,
+      "step": 4857
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3575709518161077,
+      "learning_rate": 4.3034997578954005e-06,
+      "loss": 0.5788,
+      "step": 4858
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.379705950307545,
+      "learning_rate": 4.3016570370662055e-06,
+      "loss": 0.5721,
+      "step": 4859
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8172598934029156,
+      "learning_rate": 4.299814412971356e-06,
+      "loss": 0.4542,
+      "step": 4860
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9075095940228246,
+      "learning_rate": 4.2979718858660935e-06,
+      "loss": 0.3934,
+      "step": 4861
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0111359194402008,
+      "learning_rate": 4.296129456005645e-06,
+      "loss": 0.4677,
+      "step": 4862
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.289069426300117,
+      "learning_rate": 4.294287123645222e-06,
+      "loss": 0.4561,
+      "step": 4863
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5489572400953158,
+      "learning_rate": 4.292444889040024e-06,
+      "loss": 0.6066,
+      "step": 4864
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6466856810242883,
+      "learning_rate": 4.290602752445237e-06,
+      "loss": 0.5328,
+      "step": 4865
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7841556348269962,
+      "learning_rate": 4.288760714116033e-06,
+      "loss": 0.5067,
+      "step": 4866
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8732661616526366,
+      "learning_rate": 4.286918774307572e-06,
+      "loss": 0.668,
+      "step": 4867
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1755388804886433,
+      "learning_rate": 4.285076933275001e-06,
+      "loss": 0.4578,
+      "step": 4868
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6093843805861137,
+      "learning_rate": 4.283235191273448e-06,
+      "loss": 0.4595,
+      "step": 4869
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8886196596728957,
+      "learning_rate": 4.281393548558031e-06,
+      "loss": 0.5814,
+      "step": 4870
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3165741518740535,
+      "learning_rate": 4.279552005383857e-06,
+      "loss": 0.4868,
+      "step": 4871
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.243643240126361,
+      "learning_rate": 4.277710562006013e-06,
+      "loss": 0.5029,
+      "step": 4872
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8190538095569726,
+      "learning_rate": 4.275869218679577e-06,
+      "loss": 0.4742,
+      "step": 4873
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6915922524522196,
+      "learning_rate": 4.274027975659611e-06,
+      "loss": 0.4619,
+      "step": 4874
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4115100271511696,
+      "learning_rate": 4.272186833201166e-06,
+      "loss": 0.5327,
+      "step": 4875
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.370105347238914,
+      "learning_rate": 4.270345791559272e-06,
+      "loss": 0.4557,
+      "step": 4876
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5742337229062744,
+      "learning_rate": 4.2685048509889545e-06,
+      "loss": 0.4928,
+      "step": 4877
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.59249573430252,
+      "learning_rate": 4.266664011745219e-06,
+      "loss": 0.4208,
+      "step": 4878
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.379763245748081,
+      "learning_rate": 4.264823274083056e-06,
+      "loss": 0.5031,
+      "step": 4879
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8292054580274524,
+      "learning_rate": 4.2629826382574485e-06,
+      "loss": 0.4508,
+      "step": 4880
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9603231027873116,
+      "learning_rate": 4.261142104523356e-06,
+      "loss": 0.5729,
+      "step": 4881
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.475036885292569,
+      "learning_rate": 4.259301673135732e-06,
+      "loss": 0.4342,
+      "step": 4882
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.874258557850412,
+      "learning_rate": 4.25746134434951e-06,
+      "loss": 0.5302,
+      "step": 4883
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3043173898004117,
+      "learning_rate": 4.255621118419616e-06,
+      "loss": 0.4859,
+      "step": 4884
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.417472425449803,
+      "learning_rate": 4.253780995600954e-06,
+      "loss": 0.5839,
+      "step": 4885
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1353327321355935,
+      "learning_rate": 4.251940976148421e-06,
+      "loss": 0.4422,
+      "step": 4886
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.131277589157522,
+      "learning_rate": 4.250101060316895e-06,
+      "loss": 0.4663,
+      "step": 4887
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.905176800252126,
+      "learning_rate": 4.248261248361238e-06,
+      "loss": 0.4874,
+      "step": 4888
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9391429980998005,
+      "learning_rate": 4.246421540536304e-06,
+      "loss": 0.586,
+      "step": 4889
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5998071869728867,
+      "learning_rate": 4.244581937096927e-06,
+      "loss": 0.4853,
+      "step": 4890
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.096220858236659,
+      "learning_rate": 4.2427424382979295e-06,
+      "loss": 0.5901,
+      "step": 4891
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.191849536420475,
+      "learning_rate": 4.240903044394118e-06,
+      "loss": 0.4185,
+      "step": 4892
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0076065766678504,
+      "learning_rate": 4.2390637556402855e-06,
+      "loss": 0.5511,
+      "step": 4893
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8609089774205958,
+      "learning_rate": 4.2372245722912096e-06,
+      "loss": 0.4869,
+      "step": 4894
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8921358984423517,
+      "learning_rate": 4.2353854946016545e-06,
+      "loss": 0.5514,
+      "step": 4895
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7325404169828718,
+      "learning_rate": 4.233546522826368e-06,
+      "loss": 0.5463,
+      "step": 4896
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3341226046259593,
+      "learning_rate": 4.231707657220086e-06,
+      "loss": 0.4918,
+      "step": 4897
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4579173255106137,
+      "learning_rate": 4.229868898037525e-06,
+      "loss": 0.5456,
+      "step": 4898
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.001502809621097,
+      "learning_rate": 4.22803024553339e-06,
+      "loss": 0.5462,
+      "step": 4899
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.014203429415695,
+      "learning_rate": 4.226191699962372e-06,
+      "loss": 0.5881,
+      "step": 4900
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.946634232605974,
+      "learning_rate": 4.224353261579145e-06,
+      "loss": 0.6741,
+      "step": 4901
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8804151613845406,
+      "learning_rate": 4.222514930638371e-06,
+      "loss": 0.5216,
+      "step": 4902
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2119805316163554,
+      "learning_rate": 4.220676707394693e-06,
+      "loss": 0.5555,
+      "step": 4903
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9959437903763895,
+      "learning_rate": 4.218838592102744e-06,
+      "loss": 0.4924,
+      "step": 4904
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7905737459839444,
+      "learning_rate": 4.217000585017137e-06,
+      "loss": 0.4355,
+      "step": 4905
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8971197224296756,
+      "learning_rate": 4.215162686392473e-06,
+      "loss": 0.4892,
+      "step": 4906
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9280343370076989,
+      "learning_rate": 4.2133248964833395e-06,
+      "loss": 0.508,
+      "step": 4907
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.529311727539601,
+      "learning_rate": 4.2114872155443035e-06,
+      "loss": 0.4692,
+      "step": 4908
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7326194097675334,
+      "learning_rate": 4.209649643829922e-06,
+      "loss": 0.5205,
+      "step": 4909
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7790444012482636,
+      "learning_rate": 4.207812181594735e-06,
+      "loss": 0.4721,
+      "step": 4910
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.349879977509902,
+      "learning_rate": 4.205974829093268e-06,
+      "loss": 0.5013,
+      "step": 4911
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6718859436935387,
+      "learning_rate": 4.204137586580029e-06,
+      "loss": 0.5034,
+      "step": 4912
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7281666435481362,
+      "learning_rate": 4.2023004543095166e-06,
+      "loss": 0.4819,
+      "step": 4913
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9850936560538484,
+      "learning_rate": 4.200463432536205e-06,
+      "loss": 0.4133,
+      "step": 4914
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.029263562374778,
+      "learning_rate": 4.19862652151456e-06,
+      "loss": 0.5032,
+      "step": 4915
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5993494665846297,
+      "learning_rate": 4.19678972149903e-06,
+      "loss": 0.4864,
+      "step": 4916
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9460637044325306,
+      "learning_rate": 4.194953032744049e-06,
+      "loss": 0.7168,
+      "step": 4917
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.196981884180265,
+      "learning_rate": 4.193116455504034e-06,
+      "loss": 0.5259,
+      "step": 4918
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1542055875153507,
+      "learning_rate": 4.191279990033389e-06,
+      "loss": 0.5454,
+      "step": 4919
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8785784731767823,
+      "learning_rate": 4.189443636586499e-06,
+      "loss": 0.5814,
+      "step": 4920
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.001989828028382,
+      "learning_rate": 4.187607395417736e-06,
+      "loss": 0.5132,
+      "step": 4921
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7824156873246804,
+      "learning_rate": 4.185771266781456e-06,
+      "loss": 0.4893,
+      "step": 4922
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7748320127775874,
+      "learning_rate": 4.183935250931999e-06,
+      "loss": 0.5937,
+      "step": 4923
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8522722545890007,
+      "learning_rate": 4.18209934812369e-06,
+      "loss": 0.4281,
+      "step": 4924
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8718730208534284,
+      "learning_rate": 4.1802635586108376e-06,
+      "loss": 0.4751,
+      "step": 4925
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.913702610122003,
+      "learning_rate": 4.178427882647735e-06,
+      "loss": 0.4585,
+      "step": 4926
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.223324439855877,
+      "learning_rate": 4.17659232048866e-06,
+      "loss": 0.5031,
+      "step": 4927
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2384172410151866,
+      "learning_rate": 4.174756872387874e-06,
+      "loss": 0.3995,
+      "step": 4928
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.065609620076174,
+      "learning_rate": 4.172921538599623e-06,
+      "loss": 0.4498,
+      "step": 4929
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8725773817573677,
+      "learning_rate": 4.171086319378138e-06,
+      "loss": 0.4205,
+      "step": 4930
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9859123812513846,
+      "learning_rate": 4.169251214977632e-06,
+      "loss": 0.5265,
+      "step": 4931
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.91293454616368,
+      "learning_rate": 4.1674162256523035e-06,
+      "loss": 0.5223,
+      "step": 4932
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8585811987141765,
+      "learning_rate": 4.1655813516563355e-06,
+      "loss": 0.4696,
+      "step": 4933
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9072509953141223,
+      "learning_rate": 4.163746593243895e-06,
+      "loss": 0.4282,
+      "step": 4934
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9718203896768194,
+      "learning_rate": 4.16191195066913e-06,
+      "loss": 0.579,
+      "step": 4935
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8897660720412377,
+      "learning_rate": 4.160077424186177e-06,
+      "loss": 0.534,
+      "step": 4936
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8441561837690763,
+      "learning_rate": 4.158243014049153e-06,
+      "loss": 0.4479,
+      "step": 4937
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9273697873552018,
+      "learning_rate": 4.156408720512162e-06,
+      "loss": 0.5085,
+      "step": 4938
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9204223135130989,
+      "learning_rate": 4.154574543829288e-06,
+      "loss": 0.5095,
+      "step": 4939
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7054908952069434,
+      "learning_rate": 4.152740484254602e-06,
+      "loss": 0.5133,
+      "step": 4940
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7949749662504497,
+      "learning_rate": 4.150906542042157e-06,
+      "loss": 0.4932,
+      "step": 4941
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9945027114465863,
+      "learning_rate": 4.1490727174459915e-06,
+      "loss": 0.5135,
+      "step": 4942
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1558234792945696,
+      "learning_rate": 4.147239010720125e-06,
+      "loss": 0.4807,
+      "step": 4943
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0298266155134237,
+      "learning_rate": 4.145405422118564e-06,
+      "loss": 0.5205,
+      "step": 4944
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6293622057820634,
+      "learning_rate": 4.143571951895295e-06,
+      "loss": 0.4909,
+      "step": 4945
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0428784154186506,
+      "learning_rate": 4.141738600304292e-06,
+      "loss": 0.4807,
+      "step": 4946
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2301659759375045,
+      "learning_rate": 4.13990536759951e-06,
+      "loss": 0.4982,
+      "step": 4947
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0425291356435835,
+      "learning_rate": 4.138072254034887e-06,
+      "loss": 0.5129,
+      "step": 4948
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5506930856377283,
+      "learning_rate": 4.136239259864345e-06,
+      "loss": 0.5271,
+      "step": 4949
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.551508088294162,
+      "learning_rate": 4.134406385341792e-06,
+      "loss": 0.4055,
+      "step": 4950
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7998341094436725,
+      "learning_rate": 4.132573630721116e-06,
+      "loss": 0.5095,
+      "step": 4951
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.039285514104016,
+      "learning_rate": 4.130740996256191e-06,
+      "loss": 0.5509,
+      "step": 4952
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9339287197276769,
+      "learning_rate": 4.128908482200873e-06,
+      "loss": 0.3957,
+      "step": 4953
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.237322072526917,
+      "learning_rate": 4.127076088809e-06,
+      "loss": 0.435,
+      "step": 4954
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.051578812904449,
+      "learning_rate": 4.125243816334396e-06,
+      "loss": 0.5125,
+      "step": 4955
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7685201898480072,
+      "learning_rate": 4.123411665030867e-06,
+      "loss": 0.5,
+      "step": 4956
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.491675411776145,
+      "learning_rate": 4.121579635152203e-06,
+      "loss": 0.4902,
+      "step": 4957
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1401442915416062,
+      "learning_rate": 4.119747726952175e-06,
+      "loss": 0.3505,
+      "step": 4958
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7495705293497226,
+      "learning_rate": 4.11791594068454e-06,
+      "loss": 0.4847,
+      "step": 4959
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7507459180621532,
+      "learning_rate": 4.116084276603036e-06,
+      "loss": 0.517,
+      "step": 4960
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8124322023785099,
+      "learning_rate": 4.1142527349613845e-06,
+      "loss": 0.6717,
+      "step": 4961
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4059524298386106,
+      "learning_rate": 4.112421316013291e-06,
+      "loss": 0.498,
+      "step": 4962
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.183244760037441,
+      "learning_rate": 4.110590020012444e-06,
+      "loss": 0.5584,
+      "step": 4963
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1503158973515872,
+      "learning_rate": 4.108758847212514e-06,
+      "loss": 0.5638,
+      "step": 4964
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9541332682791734,
+      "learning_rate": 4.106927797867153e-06,
+      "loss": 0.4394,
+      "step": 4965
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8696930627686397,
+      "learning_rate": 4.105096872229999e-06,
+      "loss": 0.7121,
+      "step": 4966
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0155083488013403,
+      "learning_rate": 4.10326607055467e-06,
+      "loss": 0.555,
+      "step": 4967
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8507965911239483,
+      "learning_rate": 4.1014353930947705e-06,
+      "loss": 0.5369,
+      "step": 4968
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.706019782779668,
+      "learning_rate": 4.099604840103884e-06,
+      "loss": 0.5027,
+      "step": 4969
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.832887168198852,
+      "learning_rate": 4.09777441183558e-06,
+      "loss": 0.5627,
+      "step": 4970
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6226289689159055,
+      "learning_rate": 4.095944108543407e-06,
+      "loss": 0.5838,
+      "step": 4971
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8795268547998067,
+      "learning_rate": 4.094113930480902e-06,
+      "loss": 0.4942,
+      "step": 4972
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.407372928687027,
+      "learning_rate": 4.092283877901576e-06,
+      "loss": 0.4334,
+      "step": 4973
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.138581029903175,
+      "learning_rate": 4.090453951058931e-06,
+      "loss": 0.5233,
+      "step": 4974
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.242861892293095,
+      "learning_rate": 4.088624150206446e-06,
+      "loss": 0.5421,
+      "step": 4975
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9590877057840173,
+      "learning_rate": 4.086794475597588e-06,
+      "loss": 0.5342,
+      "step": 4976
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9018599238666924,
+      "learning_rate": 4.084964927485799e-06,
+      "loss": 0.5087,
+      "step": 4977
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2444186842727274,
+      "learning_rate": 4.083135506124511e-06,
+      "loss": 0.4506,
+      "step": 4978
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8154180841638676,
+      "learning_rate": 4.081306211767133e-06,
+      "loss": 0.5088,
+      "step": 4979
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.941825819966256,
+      "learning_rate": 4.07947704466706e-06,
+      "loss": 0.493,
+      "step": 4980
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6202522639039056,
+      "learning_rate": 4.07764800507767e-06,
+      "loss": 0.4861,
+      "step": 4981
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.145563702525943,
+      "learning_rate": 4.075819093252315e-06,
+      "loss": 0.4918,
+      "step": 4982
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5851614081696654,
+      "learning_rate": 4.07399030944434e-06,
+      "loss": 0.5234,
+      "step": 4983
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1919231530265737,
+      "learning_rate": 4.072161653907067e-06,
+      "loss": 0.5362,
+      "step": 4984
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2291841144662365,
+      "learning_rate": 4.070333126893801e-06,
+      "loss": 0.5739,
+      "step": 4985
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.711840118797459,
+      "learning_rate": 4.068504728657829e-06,
+      "loss": 0.4767,
+      "step": 4986
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1098487339056544,
+      "learning_rate": 4.06667645945242e-06,
+      "loss": 0.5056,
+      "step": 4987
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.686138157850958,
+      "learning_rate": 4.064848319530827e-06,
+      "loss": 0.5041,
+      "step": 4988
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6779998089098929,
+      "learning_rate": 4.063020309146283e-06,
+      "loss": 0.4328,
+      "step": 4989
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7715543673413636,
+      "learning_rate": 4.061192428552003e-06,
+      "loss": 0.5868,
+      "step": 4990
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.87621677957861,
+      "learning_rate": 4.0593646780011855e-06,
+      "loss": 0.4041,
+      "step": 4991
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9799952214360097,
+      "learning_rate": 4.057537057747011e-06,
+      "loss": 0.3825,
+      "step": 4992
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.191962032564414,
+      "learning_rate": 4.055709568042639e-06,
+      "loss": 0.5182,
+      "step": 4993
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4056085504771247,
+      "learning_rate": 4.0538822091412135e-06,
+      "loss": 0.596,
+      "step": 4994
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1557029997561292,
+      "learning_rate": 4.052054981295861e-06,
+      "loss": 0.4718,
+      "step": 4995
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4745503699260496,
+      "learning_rate": 4.050227884759688e-06,
+      "loss": 0.4707,
+      "step": 4996
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8731063745074426,
+      "learning_rate": 4.0484009197857845e-06,
+      "loss": 0.4856,
+      "step": 4997
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3352027027644082,
+      "learning_rate": 4.0465740866272226e-06,
+      "loss": 0.4969,
+      "step": 4998
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1654976337617593,
+      "learning_rate": 4.044747385537051e-06,
+      "loss": 0.3864,
+      "step": 4999
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9222933691282333,
+      "learning_rate": 4.0429208167683055e-06,
+      "loss": 0.4549,
+      "step": 5000
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8384631792689854,
+      "learning_rate": 4.041094380574003e-06,
+      "loss": 0.4744,
+      "step": 5001
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7587667530725555,
+      "learning_rate": 4.039268077207142e-06,
+      "loss": 0.5799,
+      "step": 5002
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.443458710688763,
+      "learning_rate": 4.037441906920698e-06,
+      "loss": 0.534,
+      "step": 5003
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6869593214041998,
+      "learning_rate": 4.035615869967636e-06,
+      "loss": 0.4987,
+      "step": 5004
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5245552063171752,
+      "learning_rate": 4.033789966600897e-06,
+      "loss": 0.526,
+      "step": 5005
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1928208735367622,
+      "learning_rate": 4.031964197073403e-06,
+      "loss": 0.5846,
+      "step": 5006
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9780115053875704,
+      "learning_rate": 4.0301385616380625e-06,
+      "loss": 0.4264,
+      "step": 5007
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6959489344436074,
+      "learning_rate": 4.02831306054776e-06,
+      "loss": 0.5236,
+      "step": 5008
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.137343834748668,
+      "learning_rate": 4.026487694055363e-06,
+      "loss": 0.4965,
+      "step": 5009
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.063379617003464,
+      "learning_rate": 4.024662462413723e-06,
+      "loss": 0.5063,
+      "step": 5010
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.00919099193985,
+      "learning_rate": 4.022837365875669e-06,
+      "loss": 0.4789,
+      "step": 5011
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2186214029094358,
+      "learning_rate": 4.021012404694015e-06,
+      "loss": 0.501,
+      "step": 5012
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.164592089088275,
+      "learning_rate": 4.019187579121554e-06,
+      "loss": 0.4599,
+      "step": 5013
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4935651214366972,
+      "learning_rate": 4.017362889411057e-06,
+      "loss": 0.4958,
+      "step": 5014
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8270027641724713,
+      "learning_rate": 4.015538335815285e-06,
+      "loss": 0.5822,
+      "step": 5015
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9497195203509876,
+      "learning_rate": 4.01371391858697e-06,
+      "loss": 0.5634,
+      "step": 5016
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9640386680326911,
+      "learning_rate": 4.011889637978834e-06,
+      "loss": 0.5563,
+      "step": 5017
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7665206536814326,
+      "learning_rate": 4.010065494243573e-06,
+      "loss": 0.5306,
+      "step": 5018
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.119895955700955,
+      "learning_rate": 4.008241487633869e-06,
+      "loss": 0.5588,
+      "step": 5019
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8848738519774146,
+      "learning_rate": 4.006417618402382e-06,
+      "loss": 0.4787,
+      "step": 5020
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.159256972604456,
+      "learning_rate": 4.004593886801754e-06,
+      "loss": 0.5115,
+      "step": 5021
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.298096011661009,
+      "learning_rate": 4.002770293084608e-06,
+      "loss": 0.5072,
+      "step": 5022
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.058669829855097,
+      "learning_rate": 4.000946837503549e-06,
+      "loss": 0.5114,
+      "step": 5023
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1606690275260916,
+      "learning_rate": 3.99912352031116e-06,
+      "loss": 0.51,
+      "step": 5024
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8676412652791505,
+      "learning_rate": 3.997300341760009e-06,
+      "loss": 0.4045,
+      "step": 5025
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9506671275881255,
+      "learning_rate": 3.99547730210264e-06,
+      "loss": 0.5437,
+      "step": 5026
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8974971499162363,
+      "learning_rate": 3.993654401591582e-06,
+      "loss": 0.4975,
+      "step": 5027
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5773332696081204,
+      "learning_rate": 3.991831640479341e-06,
+      "loss": 0.4072,
+      "step": 5028
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1695677655377916,
+      "learning_rate": 3.990009019018407e-06,
+      "loss": 0.4043,
+      "step": 5029
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.535829142047127,
+      "learning_rate": 3.988186537461249e-06,
+      "loss": 0.4496,
+      "step": 5030
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1793178825187742,
+      "learning_rate": 3.986364196060317e-06,
+      "loss": 0.5623,
+      "step": 5031
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8726386936814023,
+      "learning_rate": 3.984541995068042e-06,
+      "loss": 0.4801,
+      "step": 5032
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.187100837158078,
+      "learning_rate": 3.982719934736832e-06,
+      "loss": 0.4795,
+      "step": 5033
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.053318075877081,
+      "learning_rate": 3.980898015319081e-06,
+      "loss": 0.469,
+      "step": 5034
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.750010143799455,
+      "learning_rate": 3.97907623706716e-06,
+      "loss": 0.436,
+      "step": 5035
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1557758426491014,
+      "learning_rate": 3.9772546002334225e-06,
+      "loss": 0.4734,
+      "step": 5036
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3709809724999213,
+      "learning_rate": 3.975433105070201e-06,
+      "loss": 0.4687,
+      "step": 5037
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2677680908771385,
+      "learning_rate": 3.973611751829806e-06,
+      "loss": 0.4006,
+      "step": 5038
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.050601411160466,
+      "learning_rate": 3.971790540764536e-06,
+      "loss": 0.4561,
+      "step": 5039
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.096496428736196,
+      "learning_rate": 3.9699694721266606e-06,
+      "loss": 0.5403,
+      "step": 5040
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7487509248770061,
+      "learning_rate": 3.968148546168436e-06,
+      "loss": 0.4529,
+      "step": 5041
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8886988088483981,
+      "learning_rate": 3.966327763142096e-06,
+      "loss": 0.7171,
+      "step": 5042
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.247997828679026,
+      "learning_rate": 3.964507123299855e-06,
+      "loss": 0.4773,
+      "step": 5043
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0062390587818424,
+      "learning_rate": 3.962686626893908e-06,
+      "loss": 0.3555,
+      "step": 5044
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6557572827983393,
+      "learning_rate": 3.96086627417643e-06,
+      "loss": 0.4223,
+      "step": 5045
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8692730725750724,
+      "learning_rate": 3.959046065399575e-06,
+      "loss": 0.4954,
+      "step": 5046
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4431973607265403,
+      "learning_rate": 3.95722600081548e-06,
+      "loss": 0.596,
+      "step": 5047
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9635399328161813,
+      "learning_rate": 3.955406080676259e-06,
+      "loss": 0.4882,
+      "step": 5048
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2757223190189193,
+      "learning_rate": 3.953586305234008e-06,
+      "loss": 0.5501,
+      "step": 5049
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0410664255160116,
+      "learning_rate": 3.951766674740798e-06,
+      "loss": 0.5016,
+      "step": 5050
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9191431290808638,
+      "learning_rate": 3.949947189448687e-06,
+      "loss": 0.5311,
+      "step": 5051
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.910937651993045,
+      "learning_rate": 3.94812784960971e-06,
+      "loss": 0.4287,
+      "step": 5052
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9264101675990217,
+      "learning_rate": 3.9463086554758804e-06,
+      "loss": 0.52,
+      "step": 5053
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7792582711392193,
+      "learning_rate": 3.944489607299193e-06,
+      "loss": 0.4696,
+      "step": 5054
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.305932155770469,
+      "learning_rate": 3.942670705331624e-06,
+      "loss": 0.5601,
+      "step": 5055
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.066615509475358,
+      "learning_rate": 3.940851949825124e-06,
+      "loss": 0.4613,
+      "step": 5056
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2295540689392266,
+      "learning_rate": 3.939033341031631e-06,
+      "loss": 0.504,
+      "step": 5057
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8611785447779898,
+      "learning_rate": 3.937214879203054e-06,
+      "loss": 0.5114,
+      "step": 5058
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1944821387216105,
+      "learning_rate": 3.935396564591289e-06,
+      "loss": 0.5118,
+      "step": 5059
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.294748910519658,
+      "learning_rate": 3.933578397448205e-06,
+      "loss": 0.5741,
+      "step": 5060
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.052986038411296,
+      "learning_rate": 3.931760378025659e-06,
+      "loss": 0.5571,
+      "step": 5061
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.215132614678951,
+      "learning_rate": 3.929942506575479e-06,
+      "loss": 0.4858,
+      "step": 5062
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6748661700030432,
+      "learning_rate": 3.9281247833494785e-06,
+      "loss": 0.3865,
+      "step": 5063
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.520971946681373,
+      "learning_rate": 3.926307208599447e-06,
+      "loss": 0.4741,
+      "step": 5064
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6837228010839616,
+      "learning_rate": 3.924489782577157e-06,
+      "loss": 0.4823,
+      "step": 5065
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8876522887850178,
+      "learning_rate": 3.922672505534354e-06,
+      "loss": 0.4355,
+      "step": 5066
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9093677652514591,
+      "learning_rate": 3.92085537772277e-06,
+      "loss": 0.7104,
+      "step": 5067
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7099539109042323,
+      "learning_rate": 3.91903839939411e-06,
+      "loss": 0.4875,
+      "step": 5068
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.928228535665067,
+      "learning_rate": 3.9172215708000655e-06,
+      "loss": 0.5067,
+      "step": 5069
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7177059068547176,
+      "learning_rate": 3.915404892192301e-06,
+      "loss": 0.5575,
+      "step": 5070
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.239370624976635,
+      "learning_rate": 3.9135883638224626e-06,
+      "loss": 0.5398,
+      "step": 5071
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8464339163562564,
+      "learning_rate": 3.911771985942177e-06,
+      "loss": 0.5605,
+      "step": 5072
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0069333490583983,
+      "learning_rate": 3.909955758803045e-06,
+      "loss": 0.5188,
+      "step": 5073
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5477613977058782,
+      "learning_rate": 3.908139682656655e-06,
+      "loss": 0.4431,
+      "step": 5074
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.257987285529607,
+      "learning_rate": 3.906323757754566e-06,
+      "loss": 0.5196,
+      "step": 5075
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.884613383799756,
+      "learning_rate": 3.90450798434832e-06,
+      "loss": 0.4711,
+      "step": 5076
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.818265529190095,
+      "learning_rate": 3.902692362689441e-06,
+      "loss": 0.5662,
+      "step": 5077
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9670194204144935,
+      "learning_rate": 3.9008768930294235e-06,
+      "loss": 0.5179,
+      "step": 5078
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.756112943138309,
+      "learning_rate": 3.899061575619748e-06,
+      "loss": 0.4144,
+      "step": 5079
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6131250867024352,
+      "learning_rate": 3.897246410711874e-06,
+      "loss": 0.4549,
+      "step": 5080
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0914356420400995,
+      "learning_rate": 3.895431398557235e-06,
+      "loss": 0.5001,
+      "step": 5081
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8408471056055373,
+      "learning_rate": 3.893616539407249e-06,
+      "loss": 0.5751,
+      "step": 5082
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9432495162021517,
+      "learning_rate": 3.891801833513308e-06,
+      "loss": 0.5183,
+      "step": 5083
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7052886461016197,
+      "learning_rate": 3.889987281126784e-06,
+      "loss": 0.5428,
+      "step": 5084
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7989533558028759,
+      "learning_rate": 3.8881728824990294e-06,
+      "loss": 0.4032,
+      "step": 5085
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9358715042143657,
+      "learning_rate": 3.886358637881375e-06,
+      "loss": 0.5367,
+      "step": 5086
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.096792394730582,
+      "learning_rate": 3.884544547525129e-06,
+      "loss": 0.5183,
+      "step": 5087
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1821248662723716,
+      "learning_rate": 3.882730611681579e-06,
+      "loss": 0.5262,
+      "step": 5088
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.778842648305525,
+      "learning_rate": 3.88091683060199e-06,
+      "loss": 0.5247,
+      "step": 5089
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.27926603008679,
+      "learning_rate": 3.87910320453761e-06,
+      "loss": 0.4334,
+      "step": 5090
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0475838239880115,
+      "learning_rate": 3.877289733739659e-06,
+      "loss": 0.5511,
+      "step": 5091
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.095930790642349,
+      "learning_rate": 3.875476418459339e-06,
+      "loss": 0.4956,
+      "step": 5092
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9226072577876656,
+      "learning_rate": 3.873663258947831e-06,
+      "loss": 0.5413,
+      "step": 5093
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7878089645494295,
+      "learning_rate": 3.871850255456294e-06,
+      "loss": 0.5152,
+      "step": 5094
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3609283129082823,
+      "learning_rate": 3.8700374082358635e-06,
+      "loss": 0.6077,
+      "step": 5095
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8463043222337554,
+      "learning_rate": 3.868224717537657e-06,
+      "loss": 0.4491,
+      "step": 5096
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.541036960538255,
+      "learning_rate": 3.8664121836127654e-06,
+      "loss": 0.4727,
+      "step": 5097
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7865942887703996,
+      "learning_rate": 3.864599806712261e-06,
+      "loss": 0.468,
+      "step": 5098
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.926914243018918,
+      "learning_rate": 3.8627875870871975e-06,
+      "loss": 0.4032,
+      "step": 5099
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9592095537899248,
+      "learning_rate": 3.860975524988598e-06,
+      "loss": 0.5282,
+      "step": 5100
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7466609514430285,
+      "learning_rate": 3.859163620667472e-06,
+      "loss": 0.4771,
+      "step": 5101
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5709428090125805,
+      "learning_rate": 3.857351874374805e-06,
+      "loss": 0.4547,
+      "step": 5102
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8515140593093564,
+      "learning_rate": 3.8555402863615564e-06,
+      "loss": 0.5148,
+      "step": 5103
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.211628102786537,
+      "learning_rate": 3.85372885687867e-06,
+      "loss": 0.4235,
+      "step": 5104
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.556255659978394,
+      "learning_rate": 3.851917586177063e-06,
+      "loss": 0.478,
+      "step": 5105
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.91936878903108,
+      "learning_rate": 3.850106474507635e-06,
+      "loss": 0.5042,
+      "step": 5106
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.2007554797552755,
+      "learning_rate": 3.848295522121257e-06,
+      "loss": 0.4346,
+      "step": 5107
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.1206728104682084,
+      "learning_rate": 3.846484729268784e-06,
+      "loss": 0.513,
+      "step": 5108
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7945190443348735,
+      "learning_rate": 3.844674096201047e-06,
+      "loss": 0.4905,
+      "step": 5109
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7014830190827321,
+      "learning_rate": 3.842863623168854e-06,
+      "loss": 0.4868,
+      "step": 5110
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6464996938385885,
+      "learning_rate": 3.841053310422992e-06,
+      "loss": 0.5061,
+      "step": 5111
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8463589866513597,
+      "learning_rate": 3.8392431582142245e-06,
+      "loss": 0.5389,
+      "step": 5112
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.4126482446103203,
+      "learning_rate": 3.837433166793293e-06,
+      "loss": 0.6167,
+      "step": 5113
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.2183611995152437,
+      "learning_rate": 3.835623336410919e-06,
+      "loss": 0.5473,
+      "step": 5114
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1560539760269832,
+      "learning_rate": 3.833813667317798e-06,
+      "loss": 0.476,
+      "step": 5115
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.362392335973571,
+      "learning_rate": 3.832004159764608e-06,
+      "loss": 0.5105,
+      "step": 5116
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.950244366474332,
+      "learning_rate": 3.830194814001997e-06,
+      "loss": 0.4924,
+      "step": 5117
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.11651555226258,
+      "learning_rate": 3.828385630280598e-06,
+      "loss": 0.489,
+      "step": 5118
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6478023428787254,
+      "learning_rate": 3.826576608851018e-06,
+      "loss": 0.4623,
+      "step": 5119
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.789802484321989,
+      "learning_rate": 3.824767749963844e-06,
+      "loss": 0.5006,
+      "step": 5120
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2723470621031883,
+      "learning_rate": 3.822959053869637e-06,
+      "loss": 0.4139,
+      "step": 5121
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.31402964176604,
+      "learning_rate": 3.8211505208189394e-06,
+      "loss": 0.5263,
+      "step": 5122
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.996856866642223,
+      "learning_rate": 3.819342151062266e-06,
+      "loss": 0.5374,
+      "step": 5123
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.313077275191408,
+      "learning_rate": 3.817533944850114e-06,
+      "loss": 0.3802,
+      "step": 5124
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8891070476286684,
+      "learning_rate": 3.815725902432955e-06,
+      "loss": 0.4929,
+      "step": 5125
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0824919662084085,
+      "learning_rate": 3.8139180240612386e-06,
+      "loss": 0.5195,
+      "step": 5126
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0232204013723156,
+      "learning_rate": 3.812110309985392e-06,
+      "loss": 0.3904,
+      "step": 5127
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6856058055890788,
+      "learning_rate": 3.81030276045582e-06,
+      "loss": 0.3562,
+      "step": 5128
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3789232962692366,
+      "learning_rate": 3.808495375722903e-06,
+      "loss": 0.4671,
+      "step": 5129
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.156310135622822,
+      "learning_rate": 3.8066881560370007e-06,
+      "loss": 0.4864,
+      "step": 5130
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7814220610235605,
+      "learning_rate": 3.804881101648448e-06,
+      "loss": 0.5279,
+      "step": 5131
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9565177372261564,
+      "learning_rate": 3.8030742128075584e-06,
+      "loss": 0.402,
+      "step": 5132
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.202462548608363,
+      "learning_rate": 3.801267489764623e-06,
+      "loss": 0.5782,
+      "step": 5133
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.749075307762249,
+      "learning_rate": 3.7994609327699055e-06,
+      "loss": 0.4144,
+      "step": 5134
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.904459324409021,
+      "learning_rate": 3.7976545420736518e-06,
+      "loss": 0.5036,
+      "step": 5135
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.5826028063315793,
+      "learning_rate": 3.7958483179260823e-06,
+      "loss": 0.4392,
+      "step": 5136
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.313782545564559,
+      "learning_rate": 3.7940422605773957e-06,
+      "loss": 0.4324,
+      "step": 5137
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5977554510193337,
+      "learning_rate": 3.7922363702777666e-06,
+      "loss": 0.4965,
+      "step": 5138
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5922671810192703,
+      "learning_rate": 3.7904306472773458e-06,
+      "loss": 0.4903,
+      "step": 5139
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6634004171854582,
+      "learning_rate": 3.7886250918262617e-06,
+      "loss": 0.4924,
+      "step": 5140
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6316297122755117,
+      "learning_rate": 3.7868197041746207e-06,
+      "loss": 0.38,
+      "step": 5141
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8601554709795787,
+      "learning_rate": 3.785014484572505e-06,
+      "loss": 0.4643,
+      "step": 5142
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.649172948373203,
+      "learning_rate": 3.783209433269972e-06,
+      "loss": 0.4407,
+      "step": 5143
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1188275988838012,
+      "learning_rate": 3.781404550517057e-06,
+      "loss": 0.5649,
+      "step": 5144
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.40227538812724,
+      "learning_rate": 3.7795998365637725e-06,
+      "loss": 0.4825,
+      "step": 5145
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6758135495213597,
+      "learning_rate": 3.777795291660107e-06,
+      "loss": 0.4683,
+      "step": 5146
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1592485223440843,
+      "learning_rate": 3.775990916056027e-06,
+      "loss": 0.5533,
+      "step": 5147
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8489728738101892,
+      "learning_rate": 3.7741867100014726e-06,
+      "loss": 0.4293,
+      "step": 5148
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8691819261941616,
+      "learning_rate": 3.7723826737463633e-06,
+      "loss": 0.5122,
+      "step": 5149
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.159434322061529,
+      "learning_rate": 3.770578807540595e-06,
+      "loss": 0.5283,
+      "step": 5150
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7440572091893427,
+      "learning_rate": 3.7687751116340353e-06,
+      "loss": 0.6008,
+      "step": 5151
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.052090379786352,
+      "learning_rate": 3.766971586276534e-06,
+      "loss": 0.4302,
+      "step": 5152
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.2211415403591745,
+      "learning_rate": 3.765168231717915e-06,
+      "loss": 0.6014,
+      "step": 5153
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2118118465653365,
+      "learning_rate": 3.7633650482079785e-06,
+      "loss": 0.4986,
+      "step": 5154
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8850217650313964,
+      "learning_rate": 3.761562035996502e-06,
+      "loss": 0.7244,
+      "step": 5155
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.233231746747674,
+      "learning_rate": 3.759759195333237e-06,
+      "loss": 0.4751,
+      "step": 5156
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8174492063917722,
+      "learning_rate": 3.7579565264679136e-06,
+      "loss": 0.4645,
+      "step": 5157
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.619666157434479,
+      "learning_rate": 3.756154029650237e-06,
+      "loss": 0.5249,
+      "step": 5158
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.381248703646977,
+      "learning_rate": 3.7543517051298887e-06,
+      "loss": 0.549,
+      "step": 5159
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.865945174285883,
+      "learning_rate": 3.7525495531565263e-06,
+      "loss": 0.5229,
+      "step": 5160
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5743063993222077,
+      "learning_rate": 3.7507475739797842e-06,
+      "loss": 0.5264,
+      "step": 5161
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1928359163704343,
+      "learning_rate": 3.748945767849273e-06,
+      "loss": 0.5894,
+      "step": 5162
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.154537743807666,
+      "learning_rate": 3.747144135014576e-06,
+      "loss": 0.4267,
+      "step": 5163
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.703758517326541,
+      "learning_rate": 3.7453426757252566e-06,
+      "loss": 0.4275,
+      "step": 5164
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7660343323383436,
+      "learning_rate": 3.743541390230852e-06,
+      "loss": 0.5147,
+      "step": 5165
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6829350294040817,
+      "learning_rate": 3.7417402787808766e-06,
+      "loss": 0.5379,
+      "step": 5166
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.83894653512329,
+      "learning_rate": 3.739939341624821e-06,
+      "loss": 0.5098,
+      "step": 5167
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8435996809835238,
+      "learning_rate": 3.7381385790121495e-06,
+      "loss": 0.4637,
+      "step": 5168
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.007474493390019,
+      "learning_rate": 3.7363379911923027e-06,
+      "loss": 0.5445,
+      "step": 5169
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8639610983991295,
+      "learning_rate": 3.7345375784146977e-06,
+      "loss": 0.4704,
+      "step": 5170
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8486174705138188,
+      "learning_rate": 3.7327373409287295e-06,
+      "loss": 0.4389,
+      "step": 5171
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.605898149332536,
+      "learning_rate": 3.730937278983764e-06,
+      "loss": 0.4896,
+      "step": 5172
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6322092583227494,
+      "learning_rate": 3.7291373928291475e-06,
+      "loss": 0.4128,
+      "step": 5173
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.2446505742004086,
+      "learning_rate": 3.7273376827141987e-06,
+      "loss": 0.5149,
+      "step": 5174
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8744305051259906,
+      "learning_rate": 3.7255381488882136e-06,
+      "loss": 0.578,
+      "step": 5175
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3539883885725783,
+      "learning_rate": 3.723738791600464e-06,
+      "loss": 0.5754,
+      "step": 5176
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.508821851660808,
+      "learning_rate": 3.721939611100196e-06,
+      "loss": 0.4888,
+      "step": 5177
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7946251169536143,
+      "learning_rate": 3.720140607636631e-06,
+      "loss": 0.485,
+      "step": 5178
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6746555907304834,
+      "learning_rate": 3.7183417814589685e-06,
+      "loss": 0.5244,
+      "step": 5179
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9031905011416685,
+      "learning_rate": 3.7165431328163793e-06,
+      "loss": 0.7693,
+      "step": 5180
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.265614167623082,
+      "learning_rate": 3.714744661958014e-06,
+      "loss": 0.5531,
+      "step": 5181
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.382694599613371,
+      "learning_rate": 3.712946369132995e-06,
+      "loss": 0.5047,
+      "step": 5182
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7571971210916035,
+      "learning_rate": 3.711148254590422e-06,
+      "loss": 0.4817,
+      "step": 5183
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.187704044167985,
+      "learning_rate": 3.709350318579371e-06,
+      "loss": 0.3888,
+      "step": 5184
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9806166480021807,
+      "learning_rate": 3.7075525613488887e-06,
+      "loss": 0.5675,
+      "step": 5185
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.767025064524522,
+      "learning_rate": 3.705754983148002e-06,
+      "loss": 0.4905,
+      "step": 5186
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7851244435095854,
+      "learning_rate": 3.7039575842257113e-06,
+      "loss": 0.4357,
+      "step": 5187
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1495033431958683,
+      "learning_rate": 3.702160364830991e-06,
+      "loss": 0.4497,
+      "step": 5188
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.742273447884854,
+      "learning_rate": 3.7003633252127925e-06,
+      "loss": 0.5312,
+      "step": 5189
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5965898641739713,
+      "learning_rate": 3.6985664656200402e-06,
+      "loss": 0.4468,
+      "step": 5190
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8302371158109296,
+      "learning_rate": 3.696769786301637e-06,
+      "loss": 0.5284,
+      "step": 5191
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.763988215639915,
+      "learning_rate": 3.6949732875064558e-06,
+      "loss": 0.5422,
+      "step": 5192
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7126095806810517,
+      "learning_rate": 3.69317696948335e-06,
+      "loss": 0.4968,
+      "step": 5193
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.819678163749761,
+      "learning_rate": 3.6913808324811434e-06,
+      "loss": 0.4307,
+      "step": 5194
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.355656312401466,
+      "learning_rate": 3.6895848767486374e-06,
+      "loss": 0.4794,
+      "step": 5195
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6798276162456356,
+      "learning_rate": 3.6877891025346067e-06,
+      "loss": 0.5185,
+      "step": 5196
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9482610910358236,
+      "learning_rate": 3.685993510087803e-06,
+      "loss": 0.5332,
+      "step": 5197
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.403382179608146,
+      "learning_rate": 3.6841980996569505e-06,
+      "loss": 0.4768,
+      "step": 5198
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1244089102150987,
+      "learning_rate": 3.6824028714907493e-06,
+      "loss": 0.4539,
+      "step": 5199
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7341090378199593,
+      "learning_rate": 3.6806078258378764e-06,
+      "loss": 0.4817,
+      "step": 5200
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6663169375882163,
+      "learning_rate": 3.678812962946977e-06,
+      "loss": 0.4364,
+      "step": 5201
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.824657750755704,
+      "learning_rate": 3.677018283066677e-06,
+      "loss": 0.4722,
+      "step": 5202
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2855630382986285,
+      "learning_rate": 3.675223786445574e-06,
+      "loss": 0.495,
+      "step": 5203
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1673325640774666,
+      "learning_rate": 3.673429473332244e-06,
+      "loss": 0.5388,
+      "step": 5204
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9498452062317948,
+      "learning_rate": 3.671635343975234e-06,
+      "loss": 0.4719,
+      "step": 5205
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6433333714114524,
+      "learning_rate": 3.669841398623065e-06,
+      "loss": 0.5031,
+      "step": 5206
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.628277434421528,
+      "learning_rate": 3.668047637524237e-06,
+      "loss": 0.3821,
+      "step": 5207
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8689921611061533,
+      "learning_rate": 3.6662540609272175e-06,
+      "loss": 0.4362,
+      "step": 5208
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9419601774502209,
+      "learning_rate": 3.664460669080455e-06,
+      "loss": 0.4884,
+      "step": 5209
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.584024743564064,
+      "learning_rate": 3.6626674622323687e-06,
+      "loss": 0.5127,
+      "step": 5210
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.714152450685024,
+      "learning_rate": 3.660874440631355e-06,
+      "loss": 0.4647,
+      "step": 5211
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6935775038699057,
+      "learning_rate": 3.6590816045257817e-06,
+      "loss": 0.4615,
+      "step": 5212
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1788554651812073,
+      "learning_rate": 3.657288954163991e-06,
+      "loss": 0.5538,
+      "step": 5213
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.910937791644827,
+      "learning_rate": 3.6554964897943033e-06,
+      "loss": 0.4315,
+      "step": 5214
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1319578237312133,
+      "learning_rate": 3.653704211665008e-06,
+      "loss": 0.4977,
+      "step": 5215
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.658594734948585,
+      "learning_rate": 3.651912120024372e-06,
+      "loss": 0.5164,
+      "step": 5216
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3427357787390792,
+      "learning_rate": 3.650120215120639e-06,
+      "loss": 0.4379,
+      "step": 5217
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9183660866668575,
+      "learning_rate": 3.648328497202017e-06,
+      "loss": 0.5347,
+      "step": 5218
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0174662668473893,
+      "learning_rate": 3.646536966516697e-06,
+      "loss": 0.504,
+      "step": 5219
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.407681155344638,
+      "learning_rate": 3.644745623312843e-06,
+      "loss": 0.5008,
+      "step": 5220
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9079389311400419,
+      "learning_rate": 3.64295446783859e-06,
+      "loss": 0.5657,
+      "step": 5221
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.158752138024685,
+      "learning_rate": 3.6411635003420494e-06,
+      "loss": 0.4678,
+      "step": 5222
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6414308534385946,
+      "learning_rate": 3.639372721071305e-06,
+      "loss": 0.4897,
+      "step": 5223
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2048509211163325,
+      "learning_rate": 3.6375821302744153e-06,
+      "loss": 0.4239,
+      "step": 5224
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3043329082204957,
+      "learning_rate": 3.6357917281994136e-06,
+      "loss": 0.461,
+      "step": 5225
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.787822226487705,
+      "learning_rate": 3.634001515094305e-06,
+      "loss": 0.5041,
+      "step": 5226
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.9267004856434613,
+      "learning_rate": 3.6322114912070716e-06,
+      "loss": 0.5209,
+      "step": 5227
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0120494929271424,
+      "learning_rate": 3.630421656785664e-06,
+      "loss": 0.404,
+      "step": 5228
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8675339089767384,
+      "learning_rate": 3.6286320120780113e-06,
+      "loss": 0.5213,
+      "step": 5229
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.100735100636779,
+      "learning_rate": 3.6268425573320143e-06,
+      "loss": 0.4335,
+      "step": 5230
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8252419539835898,
+      "learning_rate": 3.625053292795549e-06,
+      "loss": 0.3742,
+      "step": 5231
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7833609339060217,
+      "learning_rate": 3.6232642187164634e-06,
+      "loss": 0.4513,
+      "step": 5232
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8100465433653956,
+      "learning_rate": 3.6214753353425795e-06,
+      "loss": 0.4088,
+      "step": 5233
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.198763580969919,
+      "learning_rate": 3.619686642921696e-06,
+      "loss": 0.4848,
+      "step": 5234
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.411755264791243,
+      "learning_rate": 3.6178981417015767e-06,
+      "loss": 0.4337,
+      "step": 5235
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5973978927569226,
+      "learning_rate": 3.6161098319299682e-06,
+      "loss": 0.5213,
+      "step": 5236
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3144953478109858,
+      "learning_rate": 3.614321713854586e-06,
+      "loss": 0.4532,
+      "step": 5237
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3255429386702584,
+      "learning_rate": 3.6125337877231192e-06,
+      "loss": 0.5384,
+      "step": 5238
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1792170325168225,
+      "learning_rate": 3.610746053783233e-06,
+      "loss": 0.4802,
+      "step": 5239
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4945700096072567,
+      "learning_rate": 3.608958512282562e-06,
+      "loss": 0.5652,
+      "step": 5240
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.904857160561426,
+      "learning_rate": 3.607171163468717e-06,
+      "loss": 0.5058,
+      "step": 5241
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.152717183942851,
+      "learning_rate": 3.6053840075892816e-06,
+      "loss": 0.4417,
+      "step": 5242
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8536091996616813,
+      "learning_rate": 3.6035970448918117e-06,
+      "loss": 0.5222,
+      "step": 5243
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.784560161519531,
+      "learning_rate": 3.6018102756238373e-06,
+      "loss": 0.513,
+      "step": 5244
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.081735573007617,
+      "learning_rate": 3.600023700032861e-06,
+      "loss": 0.4797,
+      "step": 5245
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.559302015633876,
+      "learning_rate": 3.598237318366361e-06,
+      "loss": 0.4627,
+      "step": 5246
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.438376075944337,
+      "learning_rate": 3.596451130871783e-06,
+      "loss": 0.4577,
+      "step": 5247
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9018070366763367,
+      "learning_rate": 3.594665137796552e-06,
+      "loss": 0.5193,
+      "step": 5248
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0612480510945765,
+      "learning_rate": 3.5928793393880623e-06,
+      "loss": 0.527,
+      "step": 5249
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8937544089156397,
+      "learning_rate": 3.5910937358936825e-06,
+      "loss": 0.4952,
+      "step": 5250
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.478356072805228,
+      "learning_rate": 3.5893083275607565e-06,
+      "loss": 0.4835,
+      "step": 5251
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.0567687550395433,
+      "learning_rate": 3.5875231146365954e-06,
+      "loss": 0.5392,
+      "step": 5252
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2251454695912534,
+      "learning_rate": 3.5857380973684876e-06,
+      "loss": 0.4423,
+      "step": 5253
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7593047091888647,
+      "learning_rate": 3.5839532760036933e-06,
+      "loss": 0.3864,
+      "step": 5254
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.3671360678996054,
+      "learning_rate": 3.582168650789446e-06,
+      "loss": 0.6129,
+      "step": 5255
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.289354941784124,
+      "learning_rate": 3.580384221972951e-06,
+      "loss": 0.5094,
+      "step": 5256
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.816460514694996,
+      "learning_rate": 3.5785999898013887e-06,
+      "loss": 0.5085,
+      "step": 5257
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9020533848780438,
+      "learning_rate": 3.576815954521909e-06,
+      "loss": 0.7224,
+      "step": 5258
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.380320407323962,
+      "learning_rate": 3.575032116381637e-06,
+      "loss": 0.5264,
+      "step": 5259
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.683611385170924,
+      "learning_rate": 3.5732484756276693e-06,
+      "loss": 0.4612,
+      "step": 5260
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8986384435652586,
+      "learning_rate": 3.5714650325070752e-06,
+      "loss": 0.5239,
+      "step": 5261
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6789333148109358,
+      "learning_rate": 3.5696817872668984e-06,
+      "loss": 0.4833,
+      "step": 5262
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8765901497839967,
+      "learning_rate": 3.5678987401541522e-06,
+      "loss": 0.5072,
+      "step": 5263
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7807824253806515,
+      "learning_rate": 3.5661158914158243e-06,
+      "loss": 0.5973,
+      "step": 5264
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.577665132259173,
+      "learning_rate": 3.5643332412988753e-06,
+      "loss": 0.5137,
+      "step": 5265
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8805721493386138,
+      "learning_rate": 3.562550790050237e-06,
+      "loss": 0.7122,
+      "step": 5266
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9262208286072058,
+      "learning_rate": 3.5607685379168145e-06,
+      "loss": 0.5226,
+      "step": 5267
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9007790602720147,
+      "learning_rate": 3.558986485145485e-06,
+      "loss": 0.5135,
+      "step": 5268
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.859940978902418,
+      "learning_rate": 3.5572046319830973e-06,
+      "loss": 0.5489,
+      "step": 5269
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1312496578856273,
+      "learning_rate": 3.555422978676474e-06,
+      "loss": 0.345,
+      "step": 5270
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2400433393134884,
+      "learning_rate": 3.5536415254724092e-06,
+      "loss": 0.5206,
+      "step": 5271
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.442229092544587,
+      "learning_rate": 3.551860272617671e-06,
+      "loss": 0.5175,
+      "step": 5272
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.756505030843506,
+      "learning_rate": 3.5500792203589964e-06,
+      "loss": 0.4504,
+      "step": 5273
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1227232966055385,
+      "learning_rate": 3.548298368943097e-06,
+      "loss": 0.5371,
+      "step": 5274
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.800279877107523,
+      "learning_rate": 3.5465177186166556e-06,
+      "loss": 0.4504,
+      "step": 5275
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8861029929618702,
+      "learning_rate": 3.544737269626328e-06,
+      "loss": 0.4788,
+      "step": 5276
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.806270775090061,
+      "learning_rate": 3.5429570222187424e-06,
+      "loss": 0.5327,
+      "step": 5277
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.851348771162477,
+      "learning_rate": 3.5411769766404975e-06,
+      "loss": 0.5723,
+      "step": 5278
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9407529794864073,
+      "learning_rate": 3.539397133138165e-06,
+      "loss": 0.53,
+      "step": 5279
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.255678561509405,
+      "learning_rate": 3.5376174919582884e-06,
+      "loss": 0.4523,
+      "step": 5280
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.109333858003371,
+      "learning_rate": 3.5358380533473834e-06,
+      "loss": 0.6049,
+      "step": 5281
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9543463872371662,
+      "learning_rate": 3.5340588175519387e-06,
+      "loss": 0.5767,
+      "step": 5282
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.954789910891778,
+      "learning_rate": 3.532279784818412e-06,
+      "loss": 0.5222,
+      "step": 5283
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.217246768798719,
+      "learning_rate": 3.530500955393235e-06,
+      "loss": 0.4705,
+      "step": 5284
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.162742143026045,
+      "learning_rate": 3.5287223295228135e-06,
+      "loss": 0.496,
+      "step": 5285
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1952518784120705,
+      "learning_rate": 3.526943907453518e-06,
+      "loss": 0.5425,
+      "step": 5286
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0311327233338257,
+      "learning_rate": 3.5251656894316976e-06,
+      "loss": 0.5178,
+      "step": 5287
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1793184895728213,
+      "learning_rate": 3.52338767570367e-06,
+      "loss": 0.4764,
+      "step": 5288
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5171265720321623,
+      "learning_rate": 3.521609866515726e-06,
+      "loss": 0.4561,
+      "step": 5289
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0214141996089565,
+      "learning_rate": 3.5198322621141268e-06,
+      "loss": 0.4769,
+      "step": 5290
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.467404025753706,
+      "learning_rate": 3.518054862745107e-06,
+      "loss": 0.5324,
+      "step": 5291
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9587843756755257,
+      "learning_rate": 3.5162776686548717e-06,
+      "loss": 0.5348,
+      "step": 5292
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.727065902410251,
+      "learning_rate": 3.5145006800895952e-06,
+      "loss": 0.4488,
+      "step": 5293
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7731133394435948,
+      "learning_rate": 3.512723897295428e-06,
+      "loss": 0.6526,
+      "step": 5294
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0958243083660206,
+      "learning_rate": 3.5109473205184886e-06,
+      "loss": 0.6023,
+      "step": 5295
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5940262490219714,
+      "learning_rate": 3.509170950004869e-06,
+      "loss": 0.4935,
+      "step": 5296
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.631766107250355,
+      "learning_rate": 3.5073947860006298e-06,
+      "loss": 0.4375,
+      "step": 5297
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8206644892940258,
+      "learning_rate": 3.5056188287518074e-06,
+      "loss": 0.4782,
+      "step": 5298
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0115250956260056,
+      "learning_rate": 3.503843078504405e-06,
+      "loss": 0.5973,
+      "step": 5299
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0210062439668453,
+      "learning_rate": 3.5020675355044013e-06,
+      "loss": 0.5092,
+      "step": 5300
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.551098705009813,
+      "learning_rate": 3.500292199997743e-06,
+      "loss": 0.4806,
+      "step": 5301
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8530320403119205,
+      "learning_rate": 3.498517072230351e-06,
+      "loss": 0.5373,
+      "step": 5302
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3319728714856676,
+      "learning_rate": 3.4967421524481125e-06,
+      "loss": 0.4574,
+      "step": 5303
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.431112093270714,
+      "learning_rate": 3.49496744089689e-06,
+      "loss": 0.5058,
+      "step": 5304
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.143903065927962,
+      "learning_rate": 3.493192937822518e-06,
+      "loss": 0.4619,
+      "step": 5305
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.518626801519871,
+      "learning_rate": 3.491418643470799e-06,
+      "loss": 0.5322,
+      "step": 5306
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.915067079841346,
+      "learning_rate": 3.489644558087507e-06,
+      "loss": 0.4577,
+      "step": 5307
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4543087513451565,
+      "learning_rate": 3.4878706819183903e-06,
+      "loss": 0.4314,
+      "step": 5308
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5880688068754105,
+      "learning_rate": 3.4860970152091644e-06,
+      "loss": 0.5391,
+      "step": 5309
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5817385074194683,
+      "learning_rate": 3.484323558205518e-06,
+      "loss": 0.5179,
+      "step": 5310
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4559778435572803,
+      "learning_rate": 3.48255031115311e-06,
+      "loss": 0.4055,
+      "step": 5311
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.952043817900018,
+      "learning_rate": 3.480777274297571e-06,
+      "loss": 0.5661,
+      "step": 5312
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9926922834617962,
+      "learning_rate": 3.4790044478845e-06,
+      "loss": 0.4946,
+      "step": 5313
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7823876998327304,
+      "learning_rate": 3.4772318321594686e-06,
+      "loss": 0.403,
+      "step": 5314
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.339912615938851,
+      "learning_rate": 3.4754594273680205e-06,
+      "loss": 0.5297,
+      "step": 5315
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1285387966501754,
+      "learning_rate": 3.473687233755668e-06,
+      "loss": 0.5384,
+      "step": 5316
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.856271013534417,
+      "learning_rate": 3.4719152515678967e-06,
+      "loss": 0.5382,
+      "step": 5317
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1203329103856565,
+      "learning_rate": 3.470143481050159e-06,
+      "loss": 0.5993,
+      "step": 5318
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.864352373616673,
+      "learning_rate": 3.468371922447884e-06,
+      "loss": 0.4734,
+      "step": 5319
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4363359030816873,
+      "learning_rate": 3.4666005760064624e-06,
+      "loss": 0.5036,
+      "step": 5320
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.8305137052569864,
+      "learning_rate": 3.4648294419712637e-06,
+      "loss": 0.4562,
+      "step": 5321
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.893200774823598,
+      "learning_rate": 3.463058520587625e-06,
+      "loss": 0.464,
+      "step": 5322
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9173374060899253,
+      "learning_rate": 3.461287812100853e-06,
+      "loss": 0.4617,
+      "step": 5323
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.759629342310016,
+      "learning_rate": 3.459517316756228e-06,
+      "loss": 0.4925,
+      "step": 5324
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7412212971525658,
+      "learning_rate": 3.4577470347989965e-06,
+      "loss": 0.4454,
+      "step": 5325
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.8553309403186957,
+      "learning_rate": 3.4559769664743792e-06,
+      "loss": 0.6085,
+      "step": 5326
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.436516098349339,
+      "learning_rate": 3.4542071120275644e-06,
+      "loss": 0.4478,
+      "step": 5327
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.088911213070293,
+      "learning_rate": 3.4524374717037135e-06,
+      "loss": 0.462,
+      "step": 5328
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0056973989536324,
+      "learning_rate": 3.450668045747956e-06,
+      "loss": 0.5238,
+      "step": 5329
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7648725598491706,
+      "learning_rate": 3.448898834405392e-06,
+      "loss": 0.5179,
+      "step": 5330
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.358872644271964,
+      "learning_rate": 3.447129837921094e-06,
+      "loss": 0.5958,
+      "step": 5331
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1599831027891963,
+      "learning_rate": 3.4453610565401007e-06,
+      "loss": 0.4505,
+      "step": 5332
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7822481392512597,
+      "learning_rate": 3.443592490507425e-06,
+      "loss": 0.4951,
+      "step": 5333
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2730156458120083,
+      "learning_rate": 3.441824140068047e-06,
+      "loss": 0.5518,
+      "step": 5334
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8803813422595941,
+      "learning_rate": 3.4400560054669196e-06,
+      "loss": 0.7054,
+      "step": 5335
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6603821356390671,
+      "learning_rate": 3.4382880869489653e-06,
+      "loss": 0.4261,
+      "step": 5336
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9452875937859426,
+      "learning_rate": 3.436520384759075e-06,
+      "loss": 0.4445,
+      "step": 5337
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9081719436096778,
+      "learning_rate": 3.4347528991421085e-06,
+      "loss": 0.5139,
+      "step": 5338
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7761579161634535,
+      "learning_rate": 3.4329856303429e-06,
+      "loss": 0.5153,
+      "step": 5339
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8932842459481577,
+      "learning_rate": 3.43121857860625e-06,
+      "loss": 0.7029,
+      "step": 5340
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0125564829483054,
+      "learning_rate": 3.4294517441769314e-06,
+      "loss": 0.4078,
+      "step": 5341
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0091839556871025,
+      "learning_rate": 3.4276851272996847e-06,
+      "loss": 0.5321,
+      "step": 5342
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.444473233674903,
+      "learning_rate": 3.4259187282192217e-06,
+      "loss": 0.4928,
+      "step": 5343
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1310783503963164,
+      "learning_rate": 3.424152547180225e-06,
+      "loss": 0.5343,
+      "step": 5344
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.6403547022212046,
+      "learning_rate": 3.422386584427343e-06,
+      "loss": 0.4685,
+      "step": 5345
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9191705462152533,
+      "learning_rate": 3.420620840205199e-06,
+      "loss": 0.4971,
+      "step": 5346
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 12.429433660315318,
+      "learning_rate": 3.4188553147583824e-06,
+      "loss": 0.5284,
+      "step": 5347
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.040754745324801,
+      "learning_rate": 3.417090008331454e-06,
+      "loss": 0.4996,
+      "step": 5348
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.627370388501293,
+      "learning_rate": 3.4153249211689426e-06,
+      "loss": 0.5271,
+      "step": 5349
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.474877128720193,
+      "learning_rate": 3.413560053515349e-06,
+      "loss": 0.469,
+      "step": 5350
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7253322722524558,
+      "learning_rate": 3.4117954056151435e-06,
+      "loss": 0.5338,
+      "step": 5351
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.978972707055065,
+      "learning_rate": 3.4100309777127634e-06,
+      "loss": 0.4693,
+      "step": 5352
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0641287041823566,
+      "learning_rate": 3.408266770052615e-06,
+      "loss": 0.5592,
+      "step": 5353
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2327455245286902,
+      "learning_rate": 3.406502782879079e-06,
+      "loss": 0.5313,
+      "step": 5354
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2216321827365433,
+      "learning_rate": 3.404739016436501e-06,
+      "loss": 0.5033,
+      "step": 5355
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6715385416114652,
+      "learning_rate": 3.4029754709691976e-06,
+      "loss": 0.5335,
+      "step": 5356
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7420113626740266,
+      "learning_rate": 3.401212146721457e-06,
+      "loss": 0.5183,
+      "step": 5357
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9849758206579096,
+      "learning_rate": 3.3994490439375318e-06,
+      "loss": 0.4563,
+      "step": 5358
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.256120193531315,
+      "learning_rate": 3.3976861628616477e-06,
+      "loss": 0.4245,
+      "step": 5359
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7240707967092326,
+      "learning_rate": 3.395923503737999e-06,
+      "loss": 0.4165,
+      "step": 5360
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.940443166826805,
+      "learning_rate": 3.3941610668107482e-06,
+      "loss": 0.4334,
+      "step": 5361
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0212916783727173,
+      "learning_rate": 3.392398852324029e-06,
+      "loss": 0.5134,
+      "step": 5362
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.388627719039449,
+      "learning_rate": 3.3906368605219418e-06,
+      "loss": 0.4939,
+      "step": 5363
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3388150878217133,
+      "learning_rate": 3.3888750916485585e-06,
+      "loss": 0.4605,
+      "step": 5364
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9339088214998041,
+      "learning_rate": 3.3871135459479176e-06,
+      "loss": 0.5394,
+      "step": 5365
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.7420362780740306,
+      "learning_rate": 3.38535222366403e-06,
+      "loss": 0.5325,
+      "step": 5366
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9279599503585507,
+      "learning_rate": 3.383591125040872e-06,
+      "loss": 0.4081,
+      "step": 5367
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0373659589629542,
+      "learning_rate": 3.3818302503223915e-06,
+      "loss": 0.453,
+      "step": 5368
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0723744168466314,
+      "learning_rate": 3.380069599752507e-06,
+      "loss": 0.5241,
+      "step": 5369
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0650209590184887,
+      "learning_rate": 3.378309173575098e-06,
+      "loss": 0.4903,
+      "step": 5370
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.913256572438015,
+      "learning_rate": 3.3765489720340215e-06,
+      "loss": 0.5924,
+      "step": 5371
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.861917212430981,
+      "learning_rate": 3.374788995373101e-06,
+      "loss": 0.5935,
+      "step": 5372
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2134220949662993,
+      "learning_rate": 3.3730292438361266e-06,
+      "loss": 0.5764,
+      "step": 5373
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.404813075531006,
+      "learning_rate": 3.37126971766686e-06,
+      "loss": 0.4597,
+      "step": 5374
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8568576969483848,
+      "learning_rate": 3.3695104171090297e-06,
+      "loss": 0.7125,
+      "step": 5375
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9052219258192944,
+      "learning_rate": 3.3677513424063345e-06,
+      "loss": 0.5126,
+      "step": 5376
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8837848044143082,
+      "learning_rate": 3.365992493802441e-06,
+      "loss": 0.5549,
+      "step": 5377
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4224271200865855,
+      "learning_rate": 3.364233871540984e-06,
+      "loss": 0.5573,
+      "step": 5378
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2066407073879364,
+      "learning_rate": 3.3624754758655674e-06,
+      "loss": 0.4866,
+      "step": 5379
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8542094062435068,
+      "learning_rate": 3.360717307019764e-06,
+      "loss": 0.4013,
+      "step": 5380
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6822937202408952,
+      "learning_rate": 3.3589593652471153e-06,
+      "loss": 0.411,
+      "step": 5381
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7544209945581823,
+      "learning_rate": 3.3572016507911314e-06,
+      "loss": 0.6129,
+      "step": 5382
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.99018500358122,
+      "learning_rate": 3.3554441638952904e-06,
+      "loss": 0.5524,
+      "step": 5383
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2243994422778273,
+      "learning_rate": 3.3536869048030386e-06,
+      "loss": 0.4495,
+      "step": 5384
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9126584395974655,
+      "learning_rate": 3.351929873757792e-06,
+      "loss": 0.4654,
+      "step": 5385
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.976247615458194,
+      "learning_rate": 3.3501730710029358e-06,
+      "loss": 0.4626,
+      "step": 5386
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8031343232412036,
+      "learning_rate": 3.348416496781818e-06,
+      "loss": 0.4519,
+      "step": 5387
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.160944353821932,
+      "learning_rate": 3.3466601513377604e-06,
+      "loss": 0.5272,
+      "step": 5388
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6072434873495018,
+      "learning_rate": 3.3449040349140527e-06,
+      "loss": 0.4441,
+      "step": 5389
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8543581586058269,
+      "learning_rate": 3.3431481477539513e-06,
+      "loss": 0.546,
+      "step": 5390
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0003116429734575,
+      "learning_rate": 3.3413924901006817e-06,
+      "loss": 0.4957,
+      "step": 5391
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8716364435227555,
+      "learning_rate": 3.339637062197437e-06,
+      "loss": 0.5058,
+      "step": 5392
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6204509686606114,
+      "learning_rate": 3.337881864287379e-06,
+      "loss": 0.5113,
+      "step": 5393
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2798535147091936,
+      "learning_rate": 3.3361268966136367e-06,
+      "loss": 0.5756,
+      "step": 5394
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.828169098382473,
+      "learning_rate": 3.334372159419309e-06,
+      "loss": 0.4471,
+      "step": 5395
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4188946170754604,
+      "learning_rate": 3.3326176529474624e-06,
+      "loss": 0.4624,
+      "step": 5396
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.444168748635672,
+      "learning_rate": 3.330863377441128e-06,
+      "loss": 0.6145,
+      "step": 5397
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9246093398897894,
+      "learning_rate": 3.3291093331433107e-06,
+      "loss": 0.465,
+      "step": 5398
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4755913948672434,
+      "learning_rate": 3.3273555202969783e-06,
+      "loss": 0.502,
+      "step": 5399
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6170016059231933,
+      "learning_rate": 3.3256019391450696e-06,
+      "loss": 0.4998,
+      "step": 5400
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8409408123480376,
+      "learning_rate": 3.32384858993049e-06,
+      "loss": 0.7317,
+      "step": 5401
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0786923399199586,
+      "learning_rate": 3.3220954728961143e-06,
+      "loss": 0.5258,
+      "step": 5402
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4926398827495904,
+      "learning_rate": 3.320342588284784e-06,
+      "loss": 0.4232,
+      "step": 5403
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2739164111784627,
+      "learning_rate": 3.318589936339306e-06,
+      "loss": 0.5064,
+      "step": 5404
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5387373643599718,
+      "learning_rate": 3.316837517302459e-06,
+      "loss": 0.499,
+      "step": 5405
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4477125787975433,
+      "learning_rate": 3.315085331416987e-06,
+      "loss": 0.4711,
+      "step": 5406
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.078543627241337,
+      "learning_rate": 3.313333378925604e-06,
+      "loss": 0.541,
+      "step": 5407
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0096122942064905,
+      "learning_rate": 3.3115816600709883e-06,
+      "loss": 0.5223,
+      "step": 5408
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.134034070691248,
+      "learning_rate": 3.3098301750957883e-06,
+      "loss": 0.5498,
+      "step": 5409
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.814331441803216,
+      "learning_rate": 3.3080789242426202e-06,
+      "loss": 0.5152,
+      "step": 5410
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5947620033916907,
+      "learning_rate": 3.3063279077540674e-06,
+      "loss": 0.4965,
+      "step": 5411
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.670422635695194,
+      "learning_rate": 3.304577125872678e-06,
+      "loss": 0.5252,
+      "step": 5412
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8481309416108033,
+      "learning_rate": 3.3028265788409724e-06,
+      "loss": 0.7209,
+      "step": 5413
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3275015351397403,
+      "learning_rate": 3.301076266901435e-06,
+      "loss": 0.5774,
+      "step": 5414
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7723760058240716,
+      "learning_rate": 3.2993261902965185e-06,
+      "loss": 0.4969,
+      "step": 5415
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6288110863670269,
+      "learning_rate": 3.2975763492686446e-06,
+      "loss": 0.3498,
+      "step": 5416
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.3159070163206104,
+      "learning_rate": 3.295826744060199e-06,
+      "loss": 0.5038,
+      "step": 5417
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9223193865584487,
+      "learning_rate": 3.294077374913538e-06,
+      "loss": 0.5378,
+      "step": 5418
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.504274870200266,
+      "learning_rate": 3.2923282420709834e-06,
+      "loss": 0.5196,
+      "step": 5419
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8891047086202426,
+      "learning_rate": 3.2905793457748257e-06,
+      "loss": 0.4703,
+      "step": 5420
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4163546049434963,
+      "learning_rate": 3.2888306862673197e-06,
+      "loss": 0.4626,
+      "step": 5421
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2921702959739036,
+      "learning_rate": 3.2870822637906917e-06,
+      "loss": 0.5418,
+      "step": 5422
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.044643133074909,
+      "learning_rate": 3.2853340785871313e-06,
+      "loss": 0.4955,
+      "step": 5423
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0524347023059963,
+      "learning_rate": 3.283586130898797e-06,
+      "loss": 0.5119,
+      "step": 5424
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1986270563403023,
+      "learning_rate": 3.281838420967815e-06,
+      "loss": 0.4728,
+      "step": 5425
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7030788540366442,
+      "learning_rate": 3.280090949036277e-06,
+      "loss": 0.6014,
+      "step": 5426
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2389674800226254,
+      "learning_rate": 3.278343715346243e-06,
+      "loss": 0.5127,
+      "step": 5427
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9624577198445174,
+      "learning_rate": 3.2765967201397393e-06,
+      "loss": 0.53,
+      "step": 5428
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.060024978832597,
+      "learning_rate": 3.2748499636587596e-06,
+      "loss": 0.521,
+      "step": 5429
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8716538441438582,
+      "learning_rate": 3.2731034461452637e-06,
+      "loss": 0.4534,
+      "step": 5430
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0004141489829204,
+      "learning_rate": 3.2713571678411794e-06,
+      "loss": 0.534,
+      "step": 5431
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0127201001254247,
+      "learning_rate": 3.2696111289884008e-06,
+      "loss": 0.5803,
+      "step": 5432
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8434622285235553,
+      "learning_rate": 3.2678653298287887e-06,
+      "loss": 0.4429,
+      "step": 5433
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1851690966358803,
+      "learning_rate": 3.2661197706041715e-06,
+      "loss": 0.5494,
+      "step": 5434
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9239965932129126,
+      "learning_rate": 3.2643744515563437e-06,
+      "loss": 0.4243,
+      "step": 5435
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9364864861194138,
+      "learning_rate": 3.2626293729270663e-06,
+      "loss": 0.5061,
+      "step": 5436
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8256759623551713,
+      "learning_rate": 3.260884534958068e-06,
+      "loss": 0.4751,
+      "step": 5437
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.163303035184231,
+      "learning_rate": 3.2591399378910416e-06,
+      "loss": 0.5195,
+      "step": 5438
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.419613486368604,
+      "learning_rate": 3.2573955819676495e-06,
+      "loss": 0.49,
+      "step": 5439
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.492563208030891,
+      "learning_rate": 3.2556514674295192e-06,
+      "loss": 0.4921,
+      "step": 5440
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.151594633538163,
+      "learning_rate": 3.2539075945182458e-06,
+      "loss": 0.6302,
+      "step": 5441
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.222809386283672,
+      "learning_rate": 3.252163963475391e-06,
+      "loss": 0.4145,
+      "step": 5442
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8629987240356548,
+      "learning_rate": 3.25042057454248e-06,
+      "loss": 0.7121,
+      "step": 5443
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5957943659851723,
+      "learning_rate": 3.248677427961008e-06,
+      "loss": 0.4833,
+      "step": 5444
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6865892741349249,
+      "learning_rate": 3.246934523972436e-06,
+      "loss": 0.4822,
+      "step": 5445
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.032940037575656,
+      "learning_rate": 3.2451918628181887e-06,
+      "loss": 0.5021,
+      "step": 5446
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9625136227564548,
+      "learning_rate": 3.2434494447396613e-06,
+      "loss": 0.4151,
+      "step": 5447
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0234879948969127,
+      "learning_rate": 3.241707269978213e-06,
+      "loss": 0.4309,
+      "step": 5448
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9984934808801522,
+      "learning_rate": 3.239965338775169e-06,
+      "loss": 0.5347,
+      "step": 5449
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8865959966878423,
+      "learning_rate": 3.2382236513718213e-06,
+      "loss": 0.5553,
+      "step": 5450
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.917491796949907,
+      "learning_rate": 3.2364822080094284e-06,
+      "loss": 0.5164,
+      "step": 5451
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8386424263150054,
+      "learning_rate": 3.234741008929214e-06,
+      "loss": 0.4726,
+      "step": 5452
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.101511843486334,
+      "learning_rate": 3.23300005437237e-06,
+      "loss": 0.4987,
+      "step": 5453
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5834371313578064,
+      "learning_rate": 3.2312593445800543e-06,
+      "loss": 0.4833,
+      "step": 5454
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.074124067488697,
+      "learning_rate": 3.229518879793385e-06,
+      "loss": 0.5185,
+      "step": 5455
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6164772418194375,
+      "learning_rate": 3.227778660253455e-06,
+      "loss": 0.4007,
+      "step": 5456
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5970426868171055,
+      "learning_rate": 3.226038686201317e-06,
+      "loss": 0.454,
+      "step": 5457
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.530440871466244,
+      "learning_rate": 3.224298957877994e-06,
+      "loss": 0.481,
+      "step": 5458
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.872181662058484,
+      "learning_rate": 3.222559475524471e-06,
+      "loss": 0.5199,
+      "step": 5459
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.107434746615305,
+      "learning_rate": 3.2208202393817022e-06,
+      "loss": 0.4578,
+      "step": 5460
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.067788759016398,
+      "learning_rate": 3.219081249690606e-06,
+      "loss": 0.5462,
+      "step": 5461
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1036400735618215,
+      "learning_rate": 3.217342506692066e-06,
+      "loss": 0.4587,
+      "step": 5462
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0479744856530036,
+      "learning_rate": 3.2156040106269332e-06,
+      "loss": 0.4521,
+      "step": 5463
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7140449125220787,
+      "learning_rate": 3.2138657617360234e-06,
+      "loss": 0.4796,
+      "step": 5464
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2439638484013273,
+      "learning_rate": 3.2121277602601196e-06,
+      "loss": 0.4848,
+      "step": 5465
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.796180878793987,
+      "learning_rate": 3.2103900064399686e-06,
+      "loss": 0.5593,
+      "step": 5466
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4914069756075645,
+      "learning_rate": 3.2086525005162835e-06,
+      "loss": 0.5893,
+      "step": 5467
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.18116161977915,
+      "learning_rate": 3.206915242729744e-06,
+      "loss": 0.495,
+      "step": 5468
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9842344808766266,
+      "learning_rate": 3.2051782333209945e-06,
+      "loss": 0.5285,
+      "step": 5469
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2217710507703954,
+      "learning_rate": 3.203441472530645e-06,
+      "loss": 0.4712,
+      "step": 5470
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2199659190950634,
+      "learning_rate": 3.201704960599274e-06,
+      "loss": 0.4495,
+      "step": 5471
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9866354678362226,
+      "learning_rate": 3.1999686977674183e-06,
+      "loss": 0.4999,
+      "step": 5472
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.094767689638826,
+      "learning_rate": 3.198232684275586e-06,
+      "loss": 0.5136,
+      "step": 5473
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.696260652387171,
+      "learning_rate": 3.1964969203642513e-06,
+      "loss": 0.3798,
+      "step": 5474
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.743684191845175,
+      "learning_rate": 3.1947614062738507e-06,
+      "loss": 0.5076,
+      "step": 5475
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1477505059896025,
+      "learning_rate": 3.1930261422447874e-06,
+      "loss": 0.4909,
+      "step": 5476
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.110614112493166,
+      "learning_rate": 3.191291128517429e-06,
+      "loss": 0.5725,
+      "step": 5477
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.793577344726605,
+      "learning_rate": 3.189556365332111e-06,
+      "loss": 0.5389,
+      "step": 5478
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.48563357490089,
+      "learning_rate": 3.1878218529291315e-06,
+      "loss": 0.5038,
+      "step": 5479
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1961424574153425,
+      "learning_rate": 3.1860875915487557e-06,
+      "loss": 0.4658,
+      "step": 5480
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.176353334602592,
+      "learning_rate": 3.1843535814312122e-06,
+      "loss": 0.5073,
+      "step": 5481
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3908775853091595,
+      "learning_rate": 3.182619822816696e-06,
+      "loss": 0.5216,
+      "step": 5482
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9661136077323,
+      "learning_rate": 3.1808863159453675e-06,
+      "loss": 0.4825,
+      "step": 5483
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.047933580453743,
+      "learning_rate": 3.1791530610573508e-06,
+      "loss": 0.5584,
+      "step": 5484
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0209465887783704,
+      "learning_rate": 3.1774200583927365e-06,
+      "loss": 0.5056,
+      "step": 5485
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5529231465982036,
+      "learning_rate": 3.1756873081915807e-06,
+      "loss": 0.4111,
+      "step": 5486
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.189266039835674,
+      "learning_rate": 3.1739548106939044e-06,
+      "loss": 0.4383,
+      "step": 5487
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2626596865526483,
+      "learning_rate": 3.17222256613969e-06,
+      "loss": 0.5876,
+      "step": 5488
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5902035126375864,
+      "learning_rate": 3.1704905747688885e-06,
+      "loss": 0.5448,
+      "step": 5489
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.089935356967814,
+      "learning_rate": 3.1687588368214164e-06,
+      "loss": 0.4845,
+      "step": 5490
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8981239175060673,
+      "learning_rate": 3.1670273525371528e-06,
+      "loss": 0.4418,
+      "step": 5491
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0005416692033977,
+      "learning_rate": 3.1652961221559427e-06,
+      "loss": 0.5027,
+      "step": 5492
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3380197393451994,
+      "learning_rate": 3.163565145917596e-06,
+      "loss": 0.463,
+      "step": 5493
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8669324105633256,
+      "learning_rate": 3.161834424061887e-06,
+      "loss": 0.6778,
+      "step": 5494
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7693676977422323,
+      "learning_rate": 3.1601039568285553e-06,
+      "loss": 0.5203,
+      "step": 5495
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.494471524574695,
+      "learning_rate": 3.1583737444573048e-06,
+      "loss": 0.4885,
+      "step": 5496
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.986629870835382,
+      "learning_rate": 3.1566437871878047e-06,
+      "loss": 0.4557,
+      "step": 5497
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.90571627829019,
+      "learning_rate": 3.154914085259688e-06,
+      "loss": 0.4967,
+      "step": 5498
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8365470487215445,
+      "learning_rate": 3.153184638912552e-06,
+      "loss": 0.4354,
+      "step": 5499
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3045297144048993,
+      "learning_rate": 3.1514554483859607e-06,
+      "loss": 0.5078,
+      "step": 5500
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7201320928651003,
+      "learning_rate": 3.1497265139194403e-06,
+      "loss": 0.4696,
+      "step": 5501
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.475811120035794,
+      "learning_rate": 3.1479978357524825e-06,
+      "loss": 0.5005,
+      "step": 5502
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9022446174107563,
+      "learning_rate": 3.1462694141245436e-06,
+      "loss": 0.5551,
+      "step": 5503
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1188792851230835,
+      "learning_rate": 3.1445412492750453e-06,
+      "loss": 0.4868,
+      "step": 5504
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.697497688205146,
+      "learning_rate": 3.1428133414433716e-06,
+      "loss": 0.4967,
+      "step": 5505
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9011482899226602,
+      "learning_rate": 3.141085690868871e-06,
+      "loss": 0.5391,
+      "step": 5506
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.726026376328595,
+      "learning_rate": 3.13935829779086e-06,
+      "loss": 0.547,
+      "step": 5507
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8517259292572459,
+      "learning_rate": 3.1376311624486145e-06,
+      "loss": 0.6902,
+      "step": 5508
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0326054610223574,
+      "learning_rate": 3.135904285081377e-06,
+      "loss": 0.5106,
+      "step": 5509
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.845098460560559,
+      "learning_rate": 3.1341776659283563e-06,
+      "loss": 0.5586,
+      "step": 5510
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9079263731783191,
+      "learning_rate": 3.132451305228721e-06,
+      "loss": 0.5593,
+      "step": 5511
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8264977194364973,
+      "learning_rate": 3.1307252032216084e-06,
+      "loss": 0.5123,
+      "step": 5512
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1434414483742126,
+      "learning_rate": 3.1289993601461164e-06,
+      "loss": 0.563,
+      "step": 5513
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.018507874856822,
+      "learning_rate": 3.1272737762413085e-06,
+      "loss": 0.4767,
+      "step": 5514
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8679659914785318,
+      "learning_rate": 3.1255484517462132e-06,
+      "loss": 0.7118,
+      "step": 5515
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8612713908822534,
+      "learning_rate": 3.1238233868998226e-06,
+      "loss": 0.5348,
+      "step": 5516
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5154200542282614,
+      "learning_rate": 3.12209858194109e-06,
+      "loss": 0.3655,
+      "step": 5517
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0114303528931625,
+      "learning_rate": 3.1203740371089373e-06,
+      "loss": 0.4145,
+      "step": 5518
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.079473163891034,
+      "learning_rate": 3.1186497526422476e-06,
+      "loss": 0.4273,
+      "step": 5519
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9035999038090468,
+      "learning_rate": 3.116925728779868e-06,
+      "loss": 0.542,
+      "step": 5520
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7937425295291928,
+      "learning_rate": 3.115201965760612e-06,
+      "loss": 0.6775,
+      "step": 5521
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9570194616629277,
+      "learning_rate": 3.113478463823252e-06,
+      "loss": 0.4436,
+      "step": 5522
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.878348273581029,
+      "learning_rate": 3.1117552232065273e-06,
+      "loss": 0.5439,
+      "step": 5523
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1837151272103306,
+      "learning_rate": 3.1100322441491425e-06,
+      "loss": 0.5855,
+      "step": 5524
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.741340177795247,
+      "learning_rate": 3.1083095268897645e-06,
+      "loss": 0.4565,
+      "step": 5525
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.87273493201481,
+      "learning_rate": 3.106587071667023e-06,
+      "loss": 0.5266,
+      "step": 5526
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4403967139063805,
+      "learning_rate": 3.104864878719513e-06,
+      "loss": 0.4985,
+      "step": 5527
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8949647271768586,
+      "learning_rate": 3.103142948285791e-06,
+      "loss": 0.3935,
+      "step": 5528
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9168976559173336,
+      "learning_rate": 3.1014212806043794e-06,
+      "loss": 0.4998,
+      "step": 5529
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7620393697609043,
+      "learning_rate": 3.0996998759137624e-06,
+      "loss": 0.5804,
+      "step": 5530
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.777334719009669,
+      "learning_rate": 3.0979787344523905e-06,
+      "loss": 0.4908,
+      "step": 5531
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9867553883862528,
+      "learning_rate": 3.096257856458674e-06,
+      "loss": 0.4528,
+      "step": 5532
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7590566537287158,
+      "learning_rate": 3.0945372421709897e-06,
+      "loss": 0.547,
+      "step": 5533
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9427973146101325,
+      "learning_rate": 3.092816891827677e-06,
+      "loss": 0.4105,
+      "step": 5534
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8576271553816364,
+      "learning_rate": 3.0910968056670377e-06,
+      "loss": 0.5031,
+      "step": 5535
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.230890943788785,
+      "learning_rate": 3.0893769839273385e-06,
+      "loss": 0.5431,
+      "step": 5536
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0470094683375395,
+      "learning_rate": 3.0876574268468085e-06,
+      "loss": 0.4813,
+      "step": 5537
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1381721249714833,
+      "learning_rate": 3.0859381346636423e-06,
+      "loss": 0.4614,
+      "step": 5538
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9202944937013042,
+      "learning_rate": 3.084219107615992e-06,
+      "loss": 0.5541,
+      "step": 5539
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.946420394170314,
+      "learning_rate": 3.08250034594198e-06,
+      "loss": 0.4418,
+      "step": 5540
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2691653473809463,
+      "learning_rate": 3.0807818498796873e-06,
+      "loss": 0.41,
+      "step": 5541
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.557680871722288,
+      "learning_rate": 3.07906361966716e-06,
+      "loss": 0.4663,
+      "step": 5542
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7339254295367075,
+      "learning_rate": 3.077345655542408e-06,
+      "loss": 0.5292,
+      "step": 5543
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1464965626885126,
+      "learning_rate": 3.075627957743402e-06,
+      "loss": 0.4082,
+      "step": 5544
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.092898928589037,
+      "learning_rate": 3.0739105265080793e-06,
+      "loss": 0.4897,
+      "step": 5545
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.531116004392681,
+      "learning_rate": 3.072193362074337e-06,
+      "loss": 0.5057,
+      "step": 5546
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8261922761823381,
+      "learning_rate": 3.0704764646800356e-06,
+      "loss": 0.475,
+      "step": 5547
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.061763038437076,
+      "learning_rate": 3.068759834563e-06,
+      "loss": 0.4514,
+      "step": 5548
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9244023425591734,
+      "learning_rate": 3.067043471961017e-06,
+      "loss": 0.5022,
+      "step": 5549
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7487493527579148,
+      "learning_rate": 3.065327377111838e-06,
+      "loss": 0.4487,
+      "step": 5550
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8113799972949425,
+      "learning_rate": 3.0636115502531756e-06,
+      "loss": 0.4591,
+      "step": 5551
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.996205997864592,
+      "learning_rate": 3.0618959916227054e-06,
+      "loss": 0.505,
+      "step": 5552
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.275813189925835,
+      "learning_rate": 3.0601807014580675e-06,
+      "loss": 0.5045,
+      "step": 5553
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7863640020421998,
+      "learning_rate": 3.0584656799968626e-06,
+      "loss": 0.4414,
+      "step": 5554
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5195021084538376,
+      "learning_rate": 3.0567509274766573e-06,
+      "loss": 0.5951,
+      "step": 5555
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0794278893549683,
+      "learning_rate": 3.055036444134975e-06,
+      "loss": 0.561,
+      "step": 5556
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7128727653221711,
+      "learning_rate": 3.0533222302093078e-06,
+      "loss": 0.4713,
+      "step": 5557
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.105368716929038,
+      "learning_rate": 3.0516082859371077e-06,
+      "loss": 0.5605,
+      "step": 5558
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1726823942615194,
+      "learning_rate": 3.0498946115557902e-06,
+      "loss": 0.5954,
+      "step": 5559
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.78515968943789,
+      "learning_rate": 3.048181207302734e-06,
+      "loss": 0.5501,
+      "step": 5560
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.548231730586014,
+      "learning_rate": 3.0464680734152783e-06,
+      "loss": 0.5291,
+      "step": 5561
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9965825456504864,
+      "learning_rate": 3.0447552101307277e-06,
+      "loss": 0.4596,
+      "step": 5562
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.905289898316301,
+      "learning_rate": 3.043042617686346e-06,
+      "loss": 0.5355,
+      "step": 5563
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.413282011317972,
+      "learning_rate": 3.0413302963193613e-06,
+      "loss": 0.534,
+      "step": 5564
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.489726013349605,
+      "learning_rate": 3.0396182462669653e-06,
+      "loss": 0.5205,
+      "step": 5565
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9336462050810599,
+      "learning_rate": 3.0379064677663116e-06,
+      "loss": 0.5173,
+      "step": 5566
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.387608144900776,
+      "learning_rate": 3.0361949610545134e-06,
+      "loss": 0.5101,
+      "step": 5567
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6786748747638582,
+      "learning_rate": 3.034483726368648e-06,
+      "loss": 0.4581,
+      "step": 5568
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.190461940130666,
+      "learning_rate": 3.032772763945757e-06,
+      "loss": 0.4302,
+      "step": 5569
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.112801727325627,
+      "learning_rate": 3.0310620740228423e-06,
+      "loss": 0.4769,
+      "step": 5570
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.152726871749323,
+      "learning_rate": 3.0293516568368674e-06,
+      "loss": 0.4952,
+      "step": 5571
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4045654305167803,
+      "learning_rate": 3.027641512624763e-06,
+      "loss": 0.5409,
+      "step": 5572
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9276201081550612,
+      "learning_rate": 3.0259316416234108e-06,
+      "loss": 0.5344,
+      "step": 5573
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.052616622214022,
+      "learning_rate": 3.024222044069667e-06,
+      "loss": 0.472,
+      "step": 5574
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4088133171919877,
+      "learning_rate": 3.022512720200342e-06,
+      "loss": 0.5486,
+      "step": 5575
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9674143461359455,
+      "learning_rate": 3.0208036702522137e-06,
+      "loss": 0.57,
+      "step": 5576
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8536634249450537,
+      "learning_rate": 3.019094894462018e-06,
+      "loss": 0.499,
+      "step": 5577
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9017673700005961,
+      "learning_rate": 3.017386393066455e-06,
+      "loss": 0.5281,
+      "step": 5578
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6885242358406094,
+      "learning_rate": 3.0156781663021845e-06,
+      "loss": 0.6116,
+      "step": 5579
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.3147518412845445,
+      "learning_rate": 3.013970214405831e-06,
+      "loss": 0.5254,
+      "step": 5580
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9401791217959017,
+      "learning_rate": 3.0122625376139793e-06,
+      "loss": 0.4231,
+      "step": 5581
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7291908519215604,
+      "learning_rate": 3.010555136163177e-06,
+      "loss": 0.4613,
+      "step": 5582
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7713284558622675,
+      "learning_rate": 3.008848010289932e-06,
+      "loss": 0.4499,
+      "step": 5583
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7822423304715316,
+      "learning_rate": 3.0071411602307167e-06,
+      "loss": 0.496,
+      "step": 5584
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2463531897993114,
+      "learning_rate": 3.0054345862219615e-06,
+      "loss": 0.5358,
+      "step": 5585
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8413978291311452,
+      "learning_rate": 3.003728288500064e-06,
+      "loss": 0.513,
+      "step": 5586
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.792180429678406,
+      "learning_rate": 3.0020222673013767e-06,
+      "loss": 0.4822,
+      "step": 5587
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7628663491262724,
+      "learning_rate": 3.000316522862219e-06,
+      "loss": 0.4933,
+      "step": 5588
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.015449978990122,
+      "learning_rate": 2.998611055418871e-06,
+      "loss": 0.4533,
+      "step": 5589
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.002256252184459,
+      "learning_rate": 2.9969058652075722e-06,
+      "loss": 0.5612,
+      "step": 5590
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2630874976622684,
+      "learning_rate": 2.9952009524645254e-06,
+      "loss": 0.453,
+      "step": 5591
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.433846682130279,
+      "learning_rate": 2.993496317425897e-06,
+      "loss": 0.5015,
+      "step": 5592
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.810111399259158,
+      "learning_rate": 2.99179196032781e-06,
+      "loss": 0.4463,
+      "step": 5593
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.426628072379558,
+      "learning_rate": 2.9900878814063526e-06,
+      "loss": 0.4677,
+      "step": 5594
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4928502799682812,
+      "learning_rate": 2.9883840808975745e-06,
+      "loss": 0.4738,
+      "step": 5595
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7808878450193655,
+      "learning_rate": 2.9866805590374846e-06,
+      "loss": 0.5381,
+      "step": 5596
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8043546142564383,
+      "learning_rate": 2.9849773160620554e-06,
+      "loss": 0.6734,
+      "step": 5597
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6821684515836117,
+      "learning_rate": 2.983274352207218e-06,
+      "loss": 0.5298,
+      "step": 5598
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.529727386334369,
+      "learning_rate": 2.98157166770887e-06,
+      "loss": 0.4758,
+      "step": 5599
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.159918653984131,
+      "learning_rate": 2.9798692628028637e-06,
+      "loss": 0.545,
+      "step": 5600
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.192039667581039,
+      "learning_rate": 2.9781671377250177e-06,
+      "loss": 0.5305,
+      "step": 5601
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7636665877775353,
+      "learning_rate": 2.9764652927111092e-06,
+      "loss": 0.5106,
+      "step": 5602
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9573987250484117,
+      "learning_rate": 2.974763727996878e-06,
+      "loss": 0.5559,
+      "step": 5603
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.014230716464861,
+      "learning_rate": 2.9730624438180244e-06,
+      "loss": 0.5054,
+      "step": 5604
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1315577173684854,
+      "learning_rate": 2.9713614404102108e-06,
+      "loss": 0.5395,
+      "step": 5605
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6365137926792894,
+      "learning_rate": 2.9696607180090596e-06,
+      "loss": 0.5651,
+      "step": 5606
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.968610861132133,
+      "learning_rate": 2.9679602768501527e-06,
+      "loss": 0.4986,
+      "step": 5607
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8180774672669044,
+      "learning_rate": 2.9662601171690364e-06,
+      "loss": 0.5251,
+      "step": 5608
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.55405375533643,
+      "learning_rate": 2.9645602392012167e-06,
+      "loss": 0.4422,
+      "step": 5609
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8200873166140532,
+      "learning_rate": 2.96286064318216e-06,
+      "loss": 0.5504,
+      "step": 5610
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.277965607657865,
+      "learning_rate": 2.9611613293472953e-06,
+      "loss": 0.504,
+      "step": 5611
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3345107016301174,
+      "learning_rate": 2.9594622979320087e-06,
+      "loss": 0.508,
+      "step": 5612
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.383492905719264,
+      "learning_rate": 2.957763549171651e-06,
+      "loss": 0.507,
+      "step": 5613
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7011307169589607,
+      "learning_rate": 2.956065083301533e-06,
+      "loss": 0.53,
+      "step": 5614
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8740358657998757,
+      "learning_rate": 2.9543669005569257e-06,
+      "loss": 0.5109,
+      "step": 5615
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3754266751026205,
+      "learning_rate": 2.952669001173061e-06,
+      "loss": 0.4666,
+      "step": 5616
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.194253463705244,
+      "learning_rate": 2.950971385385132e-06,
+      "loss": 0.5572,
+      "step": 5617
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8873998754943746,
+      "learning_rate": 2.9492740534282917e-06,
+      "loss": 0.543,
+      "step": 5618
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.055232485591827,
+      "learning_rate": 2.947577005537654e-06,
+      "loss": 0.4085,
+      "step": 5619
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9962789172501518,
+      "learning_rate": 2.9458802419482934e-06,
+      "loss": 0.4775,
+      "step": 5620
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8066721073113896,
+      "learning_rate": 2.9441837628952468e-06,
+      "loss": 0.553,
+      "step": 5621
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2527767843167266,
+      "learning_rate": 2.942487568613509e-06,
+      "loss": 0.5608,
+      "step": 5622
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8920301331205227,
+      "learning_rate": 2.940791659338039e-06,
+      "loss": 0.4596,
+      "step": 5623
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2021626508148815,
+      "learning_rate": 2.93909603530375e-06,
+      "loss": 0.5382,
+      "step": 5624
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2409667496750187,
+      "learning_rate": 2.9374006967455203e-06,
+      "loss": 0.4685,
+      "step": 5625
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.839341710079866,
+      "learning_rate": 2.9357056438981894e-06,
+      "loss": 0.5012,
+      "step": 5626
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0694968623110483,
+      "learning_rate": 2.9340108769965553e-06,
+      "loss": 0.5146,
+      "step": 5627
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0164962760058835,
+      "learning_rate": 2.9323163962753764e-06,
+      "loss": 0.4064,
+      "step": 5628
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.082777147941806,
+      "learning_rate": 2.930622201969372e-06,
+      "loss": 0.506,
+      "step": 5629
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.886903302637016,
+      "learning_rate": 2.9289282943132226e-06,
+      "loss": 0.5085,
+      "step": 5630
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.361370364914034,
+      "learning_rate": 2.927234673541567e-06,
+      "loss": 0.5499,
+      "step": 5631
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.760155113786717,
+      "learning_rate": 2.9255413398890055e-06,
+      "loss": 0.4053,
+      "step": 5632
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2837585068520148,
+      "learning_rate": 2.9238482935900974e-06,
+      "loss": 0.5029,
+      "step": 5633
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6378144102170837,
+      "learning_rate": 2.9221555348793646e-06,
+      "loss": 0.4746,
+      "step": 5634
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7546234545372479,
+      "learning_rate": 2.9204630639912867e-06,
+      "loss": 0.4296,
+      "step": 5635
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9472238640172752,
+      "learning_rate": 2.9187708811603044e-06,
+      "loss": 0.5231,
+      "step": 5636
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4235417462581457,
+      "learning_rate": 2.9170789866208216e-06,
+      "loss": 0.4472,
+      "step": 5637
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4986276900734743,
+      "learning_rate": 2.915387380607196e-06,
+      "loss": 0.5465,
+      "step": 5638
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.478941897051696,
+      "learning_rate": 2.9136960633537493e-06,
+      "loss": 0.5031,
+      "step": 5639
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.238295223625105,
+      "learning_rate": 2.9120050350947614e-06,
+      "loss": 0.4651,
+      "step": 5640
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7928643623108445,
+      "learning_rate": 2.910314296064476e-06,
+      "loss": 0.5198,
+      "step": 5641
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3034484913734308,
+      "learning_rate": 2.9086238464970896e-06,
+      "loss": 0.5545,
+      "step": 5642
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.353874870908824,
+      "learning_rate": 2.9069336866267685e-06,
+      "loss": 0.5515,
+      "step": 5643
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7518219792267016,
+      "learning_rate": 2.9052438166876305e-06,
+      "loss": 0.3649,
+      "step": 5644
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.897320106620009,
+      "learning_rate": 2.903554236913754e-06,
+      "loss": 0.4353,
+      "step": 5645
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8003529443714816,
+      "learning_rate": 2.901864947539184e-06,
+      "loss": 0.6799,
+      "step": 5646
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.048362992868039,
+      "learning_rate": 2.900175948797916e-06,
+      "loss": 0.4855,
+      "step": 5647
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.684062112393004,
+      "learning_rate": 2.8984872409239136e-06,
+      "loss": 0.4983,
+      "step": 5648
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6400390166018495,
+      "learning_rate": 2.8967988241510924e-06,
+      "loss": 0.5321,
+      "step": 5649
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3501465331481017,
+      "learning_rate": 2.895110698713336e-06,
+      "loss": 0.4653,
+      "step": 5650
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8975275483326586,
+      "learning_rate": 2.893422864844479e-06,
+      "loss": 0.4664,
+      "step": 5651
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.126082003963139,
+      "learning_rate": 2.891735322778324e-06,
+      "loss": 0.4617,
+      "step": 5652
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7357050787203474,
+      "learning_rate": 2.890048072748625e-06,
+      "loss": 0.5149,
+      "step": 5653
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.4739901044199195,
+      "learning_rate": 2.888361114989103e-06,
+      "loss": 0.4183,
+      "step": 5654
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1827995866571017,
+      "learning_rate": 2.886674449733432e-06,
+      "loss": 0.4719,
+      "step": 5655
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8730794251334473,
+      "learning_rate": 2.8849880772152537e-06,
+      "loss": 0.4317,
+      "step": 5656
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4079443838520653,
+      "learning_rate": 2.8833019976681577e-06,
+      "loss": 0.4901,
+      "step": 5657
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9250668492759944,
+      "learning_rate": 2.8816162113257033e-06,
+      "loss": 0.5425,
+      "step": 5658
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7391656383410103,
+      "learning_rate": 2.879930718421403e-06,
+      "loss": 0.5909,
+      "step": 5659
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.205475936238161,
+      "learning_rate": 2.8782455191887345e-06,
+      "loss": 0.5035,
+      "step": 5660
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1627633028304825,
+      "learning_rate": 2.876560613861127e-06,
+      "loss": 0.4877,
+      "step": 5661
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0939309266324004,
+      "learning_rate": 2.874876002671977e-06,
+      "loss": 0.4919,
+      "step": 5662
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7273308552452442,
+      "learning_rate": 2.8731916858546335e-06,
+      "loss": 0.5061,
+      "step": 5663
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.621336402765989,
+      "learning_rate": 2.871507663642411e-06,
+      "loss": 0.5696,
+      "step": 5664
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9998646592161322,
+      "learning_rate": 2.8698239362685785e-06,
+      "loss": 0.5937,
+      "step": 5665
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.614144241895868,
+      "learning_rate": 2.8681405039663636e-06,
+      "loss": 0.4512,
+      "step": 5666
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4178270945943963,
+      "learning_rate": 2.8664573669689584e-06,
+      "loss": 0.5197,
+      "step": 5667
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8032245847710764,
+      "learning_rate": 2.864774525509507e-06,
+      "loss": 0.6664,
+      "step": 5668
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8471000327629293,
+      "learning_rate": 2.863091979821121e-06,
+      "loss": 0.7335,
+      "step": 5669
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8761578957606613,
+      "learning_rate": 2.8614097301368616e-06,
+      "loss": 0.4083,
+      "step": 5670
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8753151967331183,
+      "learning_rate": 2.859727776689758e-06,
+      "loss": 0.5161,
+      "step": 5671
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.515420821785121,
+      "learning_rate": 2.85804611971279e-06,
+      "loss": 0.3771,
+      "step": 5672
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2830445624264146,
+      "learning_rate": 2.8563647594389067e-06,
+      "loss": 0.4898,
+      "step": 5673
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.2151272323820685,
+      "learning_rate": 2.8546836961010016e-06,
+      "loss": 0.5167,
+      "step": 5674
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9515475982274781,
+      "learning_rate": 2.8530029299319416e-06,
+      "loss": 0.5451,
+      "step": 5675
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.823961450737385,
+      "learning_rate": 2.8513224611645414e-06,
+      "loss": 0.412,
+      "step": 5676
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4296216065732197,
+      "learning_rate": 2.8496422900315834e-06,
+      "loss": 0.4368,
+      "step": 5677
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8785117590328633,
+      "learning_rate": 2.8479624167658013e-06,
+      "loss": 0.4788,
+      "step": 5678
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.170129258520768,
+      "learning_rate": 2.846282841599895e-06,
+      "loss": 0.5471,
+      "step": 5679
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.674894837068322,
+      "learning_rate": 2.8446035647665136e-06,
+      "loss": 0.4356,
+      "step": 5680
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7638860126417175,
+      "learning_rate": 2.8429245864982756e-06,
+      "loss": 0.5654,
+      "step": 5681
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9252802047097994,
+      "learning_rate": 2.8412459070277486e-06,
+      "loss": 0.4529,
+      "step": 5682
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7797247844760844,
+      "learning_rate": 2.839567526587466e-06,
+      "loss": 0.5436,
+      "step": 5683
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.198085281001469,
+      "learning_rate": 2.8378894454099158e-06,
+      "loss": 0.4086,
+      "step": 5684
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1621371642707015,
+      "learning_rate": 2.8362116637275425e-06,
+      "loss": 0.4674,
+      "step": 5685
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6402653158014333,
+      "learning_rate": 2.834534181772758e-06,
+      "loss": 0.4756,
+      "step": 5686
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.23170526549326,
+      "learning_rate": 2.832856999777921e-06,
+      "loss": 0.4468,
+      "step": 5687
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8353874480073944,
+      "learning_rate": 2.8311801179753583e-06,
+      "loss": 0.3524,
+      "step": 5688
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4148347914274706,
+      "learning_rate": 2.8295035365973487e-06,
+      "loss": 0.4478,
+      "step": 5689
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.125699964299632,
+      "learning_rate": 2.827827255876137e-06,
+      "loss": 0.5036,
+      "step": 5690
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9577728957092018,
+      "learning_rate": 2.8261512760439136e-06,
+      "loss": 0.508,
+      "step": 5691
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0656096645437483,
+      "learning_rate": 2.8244755973328413e-06,
+      "loss": 0.5216,
+      "step": 5692
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.170529465923274,
+      "learning_rate": 2.8228002199750306e-06,
+      "loss": 0.5317,
+      "step": 5693
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.296225017803758,
+      "learning_rate": 2.821125144202558e-06,
+      "loss": 0.5199,
+      "step": 5694
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.701663214273712,
+      "learning_rate": 2.8194503702474505e-06,
+      "loss": 0.4798,
+      "step": 5695
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.292155764430784,
+      "learning_rate": 2.8177758983417024e-06,
+      "loss": 0.5405,
+      "step": 5696
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.339171185156427,
+      "learning_rate": 2.8161017287172573e-06,
+      "loss": 0.5379,
+      "step": 5697
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0378792943286688,
+      "learning_rate": 2.814427861606024e-06,
+      "loss": 0.4655,
+      "step": 5698
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7277773840510053,
+      "learning_rate": 2.8127542972398625e-06,
+      "loss": 0.457,
+      "step": 5699
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9781406220520874,
+      "learning_rate": 2.811081035850599e-06,
+      "loss": 0.5602,
+      "step": 5700
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.231285710850122,
+      "learning_rate": 2.80940807767001e-06,
+      "loss": 0.5333,
+      "step": 5701
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6892566653724403,
+      "learning_rate": 2.807735422929836e-06,
+      "loss": 0.5036,
+      "step": 5702
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1808149104142007,
+      "learning_rate": 2.8060630718617723e-06,
+      "loss": 0.6168,
+      "step": 5703
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6602727075015502,
+      "learning_rate": 2.804391024697469e-06,
+      "loss": 0.4828,
+      "step": 5704
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6278307415698396,
+      "learning_rate": 2.8027192816685434e-06,
+      "loss": 0.4313,
+      "step": 5705
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5519094177781656,
+      "learning_rate": 2.80104784300656e-06,
+      "loss": 0.38,
+      "step": 5706
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1714243849884087,
+      "learning_rate": 2.79937670894305e-06,
+      "loss": 0.4862,
+      "step": 5707
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.070083245800295,
+      "learning_rate": 2.7977058797094974e-06,
+      "loss": 0.5087,
+      "step": 5708
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.652779830037731,
+      "learning_rate": 2.7960353555373454e-06,
+      "loss": 0.5049,
+      "step": 5709
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.928166747132015,
+      "learning_rate": 2.7943651366579917e-06,
+      "loss": 0.5735,
+      "step": 5710
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.025659986101139,
+      "learning_rate": 2.792695223302799e-06,
+      "loss": 0.5095,
+      "step": 5711
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6655407474206225,
+      "learning_rate": 2.79102561570308e-06,
+      "loss": 0.4599,
+      "step": 5712
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.1855641239998915,
+      "learning_rate": 2.7893563140901125e-06,
+      "loss": 0.4193,
+      "step": 5713
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.079525184061289,
+      "learning_rate": 2.787687318695123e-06,
+      "loss": 0.4483,
+      "step": 5714
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2751085126428605,
+      "learning_rate": 2.786018629749305e-06,
+      "loss": 0.4702,
+      "step": 5715
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.1180719366246534,
+      "learning_rate": 2.7843502474838015e-06,
+      "loss": 0.4647,
+      "step": 5716
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9350993030459924,
+      "learning_rate": 2.782682172129719e-06,
+      "loss": 0.4894,
+      "step": 5717
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.106591020195427,
+      "learning_rate": 2.7810144039181164e-06,
+      "loss": 0.4359,
+      "step": 5718
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.426518038918427,
+      "learning_rate": 2.779346943080016e-06,
+      "loss": 0.5052,
+      "step": 5719
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.83520680278635,
+      "learning_rate": 2.777679789846391e-06,
+      "loss": 0.4917,
+      "step": 5720
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8041865998521727,
+      "learning_rate": 2.776012944448178e-06,
+      "loss": 0.381,
+      "step": 5721
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4271895550513658,
+      "learning_rate": 2.774346407116265e-06,
+      "loss": 0.4369,
+      "step": 5722
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0407899555614772,
+      "learning_rate": 2.7726801780815045e-06,
+      "loss": 0.627,
+      "step": 5723
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7715934730528438,
+      "learning_rate": 2.7710142575746998e-06,
+      "loss": 0.48,
+      "step": 5724
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0530119823439854,
+      "learning_rate": 2.769348645826614e-06,
+      "loss": 0.4594,
+      "step": 5725
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9831302503387818,
+      "learning_rate": 2.767683343067965e-06,
+      "loss": 0.519,
+      "step": 5726
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3778984137966988,
+      "learning_rate": 2.766018349529435e-06,
+      "loss": 0.5877,
+      "step": 5727
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4819785207870035,
+      "learning_rate": 2.764353665441654e-06,
+      "loss": 0.5585,
+      "step": 5728
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.46865970749228,
+      "learning_rate": 2.762689291035219e-06,
+      "loss": 0.4429,
+      "step": 5729
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6748284038011523,
+      "learning_rate": 2.7610252265406744e-06,
+      "loss": 0.5222,
+      "step": 5730
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.140217051413417,
+      "learning_rate": 2.7593614721885265e-06,
+      "loss": 0.5149,
+      "step": 5731
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0383365110845224,
+      "learning_rate": 2.757698028209241e-06,
+      "loss": 0.4583,
+      "step": 5732
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8334414182055687,
+      "learning_rate": 2.756034894833235e-06,
+      "loss": 0.6696,
+      "step": 5733
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7127542373334697,
+      "learning_rate": 2.7543720722908882e-06,
+      "loss": 0.4362,
+      "step": 5734
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.65288183957263,
+      "learning_rate": 2.752709560812531e-06,
+      "loss": 0.4362,
+      "step": 5735
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9265345807581933,
+      "learning_rate": 2.751047360628458e-06,
+      "loss": 0.5227,
+      "step": 5736
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.851985849118492,
+      "learning_rate": 2.749385471968914e-06,
+      "loss": 0.5566,
+      "step": 5737
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1845187670788384,
+      "learning_rate": 2.7477238950641055e-06,
+      "loss": 0.3884,
+      "step": 5738
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.569301862306714,
+      "learning_rate": 2.7460626301441917e-06,
+      "loss": 0.5418,
+      "step": 5739
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8375097797805147,
+      "learning_rate": 2.7444016774392933e-06,
+      "loss": 0.4643,
+      "step": 5740
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.616077363807139,
+      "learning_rate": 2.742741037179484e-06,
+      "loss": 0.4949,
+      "step": 5741
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.403794839636427,
+      "learning_rate": 2.7410807095947955e-06,
+      "loss": 0.5082,
+      "step": 5742
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8905499016080287,
+      "learning_rate": 2.7394206949152135e-06,
+      "loss": 0.4654,
+      "step": 5743
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7470326377205445,
+      "learning_rate": 2.7377609933706884e-06,
+      "loss": 0.4923,
+      "step": 5744
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2042882951814877,
+      "learning_rate": 2.7361016051911154e-06,
+      "loss": 0.4332,
+      "step": 5745
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.133535347566426,
+      "learning_rate": 2.734442530606358e-06,
+      "loss": 0.4246,
+      "step": 5746
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1860478311475227,
+      "learning_rate": 2.7327837698462276e-06,
+      "loss": 0.458,
+      "step": 5747
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6519785226670036,
+      "learning_rate": 2.7311253231404987e-06,
+      "loss": 0.5615,
+      "step": 5748
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8087182123627396,
+      "learning_rate": 2.7294671907188964e-06,
+      "loss": 0.6561,
+      "step": 5749
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.049457017965674,
+      "learning_rate": 2.7278093728111044e-06,
+      "loss": 0.5714,
+      "step": 5750
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0572112988414655,
+      "learning_rate": 2.726151869646766e-06,
+      "loss": 0.4624,
+      "step": 5751
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4465945015305666,
+      "learning_rate": 2.7244946814554746e-06,
+      "loss": 0.5179,
+      "step": 5752
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.386503856819927,
+      "learning_rate": 2.722837808466788e-06,
+      "loss": 0.446,
+      "step": 5753
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.1214577234598035,
+      "learning_rate": 2.7211812509102113e-06,
+      "loss": 0.4592,
+      "step": 5754
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6921632151016615,
+      "learning_rate": 2.719525009015216e-06,
+      "loss": 0.5305,
+      "step": 5755
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6616699300950832,
+      "learning_rate": 2.7178690830112186e-06,
+      "loss": 0.5036,
+      "step": 5756
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.728119427691811,
+      "learning_rate": 2.716213473127603e-06,
+      "loss": 0.6117,
+      "step": 5757
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9705482522973417,
+      "learning_rate": 2.7145581795937013e-06,
+      "loss": 0.6019,
+      "step": 5758
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.075590138977285,
+      "learning_rate": 2.7129032026388046e-06,
+      "loss": 0.4676,
+      "step": 5759
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4937973271565976,
+      "learning_rate": 2.7112485424921584e-06,
+      "loss": 0.583,
+      "step": 5760
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3486013301836492,
+      "learning_rate": 2.70959419938297e-06,
+      "loss": 0.5287,
+      "step": 5761
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8692078848441247,
+      "learning_rate": 2.7079401735403938e-06,
+      "loss": 0.4708,
+      "step": 5762
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.171377816131872,
+      "learning_rate": 2.70628646519355e-06,
+      "loss": 0.5051,
+      "step": 5763
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1408636203571176,
+      "learning_rate": 2.7046330745715056e-06,
+      "loss": 0.5066,
+      "step": 5764
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6010355322494831,
+      "learning_rate": 2.7029800019032924e-06,
+      "loss": 0.3531,
+      "step": 5765
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.678454681219228,
+      "learning_rate": 2.7013272474178898e-06,
+      "loss": 0.4946,
+      "step": 5766
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9396844145546188,
+      "learning_rate": 2.6996748113442397e-06,
+      "loss": 0.5542,
+      "step": 5767
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.729213259829574,
+      "learning_rate": 2.698022693911237e-06,
+      "loss": 0.5462,
+      "step": 5768
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6647278650685282,
+      "learning_rate": 2.6963708953477296e-06,
+      "loss": 0.4884,
+      "step": 5769
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8153585174880043,
+      "learning_rate": 2.6947194158825296e-06,
+      "loss": 0.6277,
+      "step": 5770
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3980437202311533,
+      "learning_rate": 2.6930682557443944e-06,
+      "loss": 0.5091,
+      "step": 5771
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1409167777760354,
+      "learning_rate": 2.6914174151620467e-06,
+      "loss": 0.5113,
+      "step": 5772
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5051745171781656,
+      "learning_rate": 2.6897668943641564e-06,
+      "loss": 0.4089,
+      "step": 5773
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.898037815837164,
+      "learning_rate": 2.688116693579361e-06,
+      "loss": 0.5209,
+      "step": 5774
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.092589102386741,
+      "learning_rate": 2.686466813036236e-06,
+      "loss": 0.4397,
+      "step": 5775
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.889973876038591,
+      "learning_rate": 2.684817252963329e-06,
+      "loss": 0.4695,
+      "step": 5776
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6749614680492515,
+      "learning_rate": 2.6831680135891338e-06,
+      "loss": 0.4861,
+      "step": 5777
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7676565528575214,
+      "learning_rate": 2.6815190951421054e-06,
+      "loss": 0.5502,
+      "step": 5778
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7419340407442663,
+      "learning_rate": 2.6798704978506485e-06,
+      "loss": 0.472,
+      "step": 5779
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.061560569618951,
+      "learning_rate": 2.678222221943131e-06,
+      "loss": 0.5639,
+      "step": 5780
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8470544800731576,
+      "learning_rate": 2.6765742676478666e-06,
+      "loss": 0.6728,
+      "step": 5781
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6921886413157088,
+      "learning_rate": 2.6749266351931343e-06,
+      "loss": 0.5165,
+      "step": 5782
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0420298872441762,
+      "learning_rate": 2.67327932480716e-06,
+      "loss": 0.4866,
+      "step": 5783
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.232205138935863,
+      "learning_rate": 2.671632336718132e-06,
+      "loss": 0.4507,
+      "step": 5784
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.525972939570713,
+      "learning_rate": 2.6699856711541884e-06,
+      "loss": 0.5002,
+      "step": 5785
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.004982462687212,
+      "learning_rate": 2.6683393283434273e-06,
+      "loss": 0.469,
+      "step": 5786
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.226837945624228,
+      "learning_rate": 2.6666933085138967e-06,
+      "loss": 0.4925,
+      "step": 5787
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.09694828725871,
+      "learning_rate": 2.665047611893607e-06,
+      "loss": 0.5697,
+      "step": 5788
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9859202613885023,
+      "learning_rate": 2.6634022387105176e-06,
+      "loss": 0.4209,
+      "step": 5789
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.373936364814647,
+      "learning_rate": 2.661757189192543e-06,
+      "loss": 0.5267,
+      "step": 5790
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8852222380156823,
+      "learning_rate": 2.66011246356756e-06,
+      "loss": 0.6659,
+      "step": 5791
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0026968299210233,
+      "learning_rate": 2.658468062063394e-06,
+      "loss": 0.6033,
+      "step": 5792
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.018272361637807,
+      "learning_rate": 2.6568239849078235e-06,
+      "loss": 0.4932,
+      "step": 5793
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5919971444446572,
+      "learning_rate": 2.6551802323285915e-06,
+      "loss": 0.5486,
+      "step": 5794
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8108697082396865,
+      "learning_rate": 2.6535368045533867e-06,
+      "loss": 0.5117,
+      "step": 5795
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9686476832251607,
+      "learning_rate": 2.6518937018098557e-06,
+      "loss": 0.4797,
+      "step": 5796
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.431416297345606,
+      "learning_rate": 2.6502509243256047e-06,
+      "loss": 0.4511,
+      "step": 5797
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.469366440860353,
+      "learning_rate": 2.6486084723281856e-06,
+      "loss": 0.495,
+      "step": 5798
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7812298332147831,
+      "learning_rate": 2.6469663460451167e-06,
+      "loss": 0.5248,
+      "step": 5799
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3021248792317532,
+      "learning_rate": 2.64532454570386e-06,
+      "loss": 0.4883,
+      "step": 5800
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2069537398883745,
+      "learning_rate": 2.643683071531841e-06,
+      "loss": 0.554,
+      "step": 5801
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.302472339346388,
+      "learning_rate": 2.642041923756433e-06,
+      "loss": 0.4753,
+      "step": 5802
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.095700946502289,
+      "learning_rate": 2.6404011026049714e-06,
+      "loss": 0.4385,
+      "step": 5803
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6754531588505395,
+      "learning_rate": 2.6387606083047378e-06,
+      "loss": 0.4341,
+      "step": 5804
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3009526407201433,
+      "learning_rate": 2.637120441082979e-06,
+      "loss": 0.4312,
+      "step": 5805
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6545528519463906,
+      "learning_rate": 2.6354806011668844e-06,
+      "loss": 0.3979,
+      "step": 5806
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.562064889601809,
+      "learning_rate": 2.6338410887836093e-06,
+      "loss": 0.4854,
+      "step": 5807
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8047182126155183,
+      "learning_rate": 2.632201904160257e-06,
+      "loss": 0.5709,
+      "step": 5808
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.83132552107066,
+      "learning_rate": 2.630563047523887e-06,
+      "loss": 0.4505,
+      "step": 5809
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8236078185137223,
+      "learning_rate": 2.628924519101511e-06,
+      "loss": 0.5214,
+      "step": 5810
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8599243557362735,
+      "learning_rate": 2.627286319120102e-06,
+      "loss": 0.5781,
+      "step": 5811
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.134736704623458,
+      "learning_rate": 2.625648447806579e-06,
+      "loss": 0.4123,
+      "step": 5812
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4878765522092148,
+      "learning_rate": 2.6240109053878236e-06,
+      "loss": 0.4699,
+      "step": 5813
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.236279819987923,
+      "learning_rate": 2.6223736920906668e-06,
+      "loss": 0.5377,
+      "step": 5814
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7918725670018099,
+      "learning_rate": 2.6207368081418914e-06,
+      "loss": 0.468,
+      "step": 5815
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6946205351266566,
+      "learning_rate": 2.619100253768244e-06,
+      "loss": 0.4462,
+      "step": 5816
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1603469870523693,
+      "learning_rate": 2.617464029196415e-06,
+      "loss": 0.4745,
+      "step": 5817
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.036179575006548,
+      "learning_rate": 2.615828134653058e-06,
+      "loss": 0.4957,
+      "step": 5818
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5471789419346162,
+      "learning_rate": 2.6141925703647725e-06,
+      "loss": 0.5005,
+      "step": 5819
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6142049129505565,
+      "learning_rate": 2.6125573365581215e-06,
+      "loss": 0.4582,
+      "step": 5820
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5907906281432374,
+      "learning_rate": 2.6109224334596117e-06,
+      "loss": 0.431,
+      "step": 5821
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.902475491157895,
+      "learning_rate": 2.609287861295715e-06,
+      "loss": 0.5234,
+      "step": 5822
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8977290341967217,
+      "learning_rate": 2.607653620292848e-06,
+      "loss": 0.5185,
+      "step": 5823
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.283802264046408,
+      "learning_rate": 2.6060197106773894e-06,
+      "loss": 0.4382,
+      "step": 5824
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9793385396153687,
+      "learning_rate": 2.604386132675666e-06,
+      "loss": 0.5457,
+      "step": 5825
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1103824934844484,
+      "learning_rate": 2.6027528865139596e-06,
+      "loss": 0.5486,
+      "step": 5826
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8337591122847308,
+      "learning_rate": 2.6011199724185067e-06,
+      "loss": 0.4055,
+      "step": 5827
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6484087504820424,
+      "learning_rate": 2.5994873906155016e-06,
+      "loss": 0.454,
+      "step": 5828
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.411906422774842,
+      "learning_rate": 2.597855141331085e-06,
+      "loss": 0.4936,
+      "step": 5829
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.028874857680586,
+      "learning_rate": 2.596223224791361e-06,
+      "loss": 0.5427,
+      "step": 5830
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8702025733168324,
+      "learning_rate": 2.5945916412223772e-06,
+      "loss": 0.4425,
+      "step": 5831
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.054842161873053,
+      "learning_rate": 2.592960390850144e-06,
+      "loss": 0.4847,
+      "step": 5832
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9077131100751072,
+      "learning_rate": 2.591329473900621e-06,
+      "loss": 0.5182,
+      "step": 5833
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.305056716342883,
+      "learning_rate": 2.58969889059972e-06,
+      "loss": 0.479,
+      "step": 5834
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7160853774053524,
+      "learning_rate": 2.5880686411733134e-06,
+      "loss": 0.5297,
+      "step": 5835
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4322941771733007,
+      "learning_rate": 2.586438725847219e-06,
+      "loss": 0.4996,
+      "step": 5836
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.317377722869986,
+      "learning_rate": 2.584809144847216e-06,
+      "loss": 0.5384,
+      "step": 5837
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8337430252487481,
+      "learning_rate": 2.58317989839903e-06,
+      "loss": 0.6971,
+      "step": 5838
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.5563754249765176,
+      "learning_rate": 2.581550986728348e-06,
+      "loss": 0.5495,
+      "step": 5839
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.0247023880593114,
+      "learning_rate": 2.5799224100608024e-06,
+      "loss": 0.5733,
+      "step": 5840
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.596171253109812,
+      "learning_rate": 2.578294168621987e-06,
+      "loss": 0.4468,
+      "step": 5841
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5054261032423866,
+      "learning_rate": 2.576666262637444e-06,
+      "loss": 0.5975,
+      "step": 5842
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4848492886658264,
+      "learning_rate": 2.575038692332671e-06,
+      "loss": 0.4821,
+      "step": 5843
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.455585066117292,
+      "learning_rate": 2.573411457933116e-06,
+      "loss": 0.471,
+      "step": 5844
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.1585447765360137,
+      "learning_rate": 2.571784559664188e-06,
+      "loss": 0.5836,
+      "step": 5845
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8686966842097241,
+      "learning_rate": 2.570157997751239e-06,
+      "loss": 0.462,
+      "step": 5846
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9284973952385376,
+      "learning_rate": 2.5685317724195868e-06,
+      "loss": 0.5161,
+      "step": 5847
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8089116556951226,
+      "learning_rate": 2.5669058838944894e-06,
+      "loss": 0.4898,
+      "step": 5848
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.373931924280234,
+      "learning_rate": 2.56528033240117e-06,
+      "loss": 0.4507,
+      "step": 5849
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.019465234607856,
+      "learning_rate": 2.563655118164795e-06,
+      "loss": 0.4243,
+      "step": 5850
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.852767570279682,
+      "learning_rate": 2.562030241410493e-06,
+      "loss": 0.4221,
+      "step": 5851
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.931578519999569,
+      "learning_rate": 2.5604057023633376e-06,
+      "loss": 0.5612,
+      "step": 5852
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.695620491370757,
+      "learning_rate": 2.558781501248364e-06,
+      "loss": 0.4485,
+      "step": 5853
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6636104918798624,
+      "learning_rate": 2.557157638290554e-06,
+      "loss": 0.4845,
+      "step": 5854
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7239997659845414,
+      "learning_rate": 2.555534113714843e-06,
+      "loss": 0.5544,
+      "step": 5855
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9063928477068948,
+      "learning_rate": 2.553910927746125e-06,
+      "loss": 0.4993,
+      "step": 5856
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.330906258320501,
+      "learning_rate": 2.55228808060924e-06,
+      "loss": 0.5308,
+      "step": 5857
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8683035618913877,
+      "learning_rate": 2.5506655725289874e-06,
+      "loss": 0.3372,
+      "step": 5858
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9377851895707632,
+      "learning_rate": 2.549043403730116e-06,
+      "loss": 0.4379,
+      "step": 5859
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8151606804033869,
+      "learning_rate": 2.547421574437327e-06,
+      "loss": 0.4818,
+      "step": 5860
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8300130293808814,
+      "learning_rate": 2.5458000848752753e-06,
+      "loss": 0.5119,
+      "step": 5861
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8444637999850237,
+      "learning_rate": 2.544178935268573e-06,
+      "loss": 0.5109,
+      "step": 5862
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8568923950329312,
+      "learning_rate": 2.542558125841777e-06,
+      "loss": 0.6191,
+      "step": 5863
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0394325888300284,
+      "learning_rate": 2.5409376568194056e-06,
+      "loss": 0.378,
+      "step": 5864
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8811242041766203,
+      "learning_rate": 2.539317528425922e-06,
+      "loss": 0.5135,
+      "step": 5865
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6879271225914863,
+      "learning_rate": 2.5376977408857505e-06,
+      "loss": 0.5758,
+      "step": 5866
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2391407212159407,
+      "learning_rate": 2.5360782944232594e-06,
+      "loss": 0.5673,
+      "step": 5867
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.111829363391716,
+      "learning_rate": 2.5344591892627777e-06,
+      "loss": 0.4914,
+      "step": 5868
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7997134113210094,
+      "learning_rate": 2.532840425628581e-06,
+      "loss": 0.4689,
+      "step": 5869
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.965281696844387,
+      "learning_rate": 2.531222003744902e-06,
+      "loss": 0.4955,
+      "step": 5870
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.149144986415558,
+      "learning_rate": 2.5296039238359227e-06,
+      "loss": 0.5781,
+      "step": 5871
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3657996245802226,
+      "learning_rate": 2.527986186125781e-06,
+      "loss": 0.5862,
+      "step": 5872
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.571738393780249,
+      "learning_rate": 2.526368790838565e-06,
+      "loss": 0.4651,
+      "step": 5873
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.112110816037857,
+      "learning_rate": 2.5247517381983137e-06,
+      "loss": 0.505,
+      "step": 5874
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8384796208281766,
+      "learning_rate": 2.5231350284290248e-06,
+      "loss": 0.4904,
+      "step": 5875
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4710005304435163,
+      "learning_rate": 2.5215186617546426e-06,
+      "loss": 0.5747,
+      "step": 5876
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7930812546481028,
+      "learning_rate": 2.519902638399064e-06,
+      "loss": 0.577,
+      "step": 5877
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7924746273497554,
+      "learning_rate": 2.518286958586145e-06,
+      "loss": 0.4963,
+      "step": 5878
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8607580234887013,
+      "learning_rate": 2.5166716225396864e-06,
+      "loss": 0.4719,
+      "step": 5879
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.758336164022484,
+      "learning_rate": 2.5150566304834422e-06,
+      "loss": 0.5133,
+      "step": 5880
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.107707681859021,
+      "learning_rate": 2.513441982641126e-06,
+      "loss": 0.5154,
+      "step": 5881
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7445000378886628,
+      "learning_rate": 2.511827679236393e-06,
+      "loss": 0.5194,
+      "step": 5882
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.374957739467125,
+      "learning_rate": 2.5102137204928604e-06,
+      "loss": 0.5384,
+      "step": 5883
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.583577895043586,
+      "learning_rate": 2.5086001066340907e-06,
+      "loss": 0.4747,
+      "step": 5884
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0775569342002007,
+      "learning_rate": 2.5069868378836048e-06,
+      "loss": 0.5145,
+      "step": 5885
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.924575383942415,
+      "learning_rate": 2.505373914464868e-06,
+      "loss": 0.4899,
+      "step": 5886
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.318661465013642,
+      "learning_rate": 2.5037613366013066e-06,
+      "loss": 0.52,
+      "step": 5887
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1068496097615985,
+      "learning_rate": 2.5021491045162903e-06,
+      "loss": 0.491,
+      "step": 5888
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.357556737274467,
+      "learning_rate": 2.50053721843315e-06,
+      "loss": 0.5215,
+      "step": 5889
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.342593945072089,
+      "learning_rate": 2.4989256785751588e-06,
+      "loss": 0.3865,
+      "step": 5890
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.660316939250707,
+      "learning_rate": 2.497314485165551e-06,
+      "loss": 0.5969,
+      "step": 5891
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.01227860389893,
+      "learning_rate": 2.495703638427508e-06,
+      "loss": 0.4278,
+      "step": 5892
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0886376121392596,
+      "learning_rate": 2.4940931385841625e-06,
+      "loss": 0.4287,
+      "step": 5893
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1251409755619926,
+      "learning_rate": 2.4924829858586e-06,
+      "loss": 0.4535,
+      "step": 5894
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.380658304002521,
+      "learning_rate": 2.490873180473862e-06,
+      "loss": 0.6059,
+      "step": 5895
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8390872948657713,
+      "learning_rate": 2.4892637226529344e-06,
+      "loss": 0.6989,
+      "step": 5896
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6574633646961692,
+      "learning_rate": 2.4876546126187622e-06,
+      "loss": 0.4482,
+      "step": 5897
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.727237372904544,
+      "learning_rate": 2.486045850594239e-06,
+      "loss": 0.5493,
+      "step": 5898
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8959039023599484,
+      "learning_rate": 2.484437436802207e-06,
+      "loss": 0.5254,
+      "step": 5899
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.699775917034189,
+      "learning_rate": 2.482829371465467e-06,
+      "loss": 0.3697,
+      "step": 5900
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.976082662185539,
+      "learning_rate": 2.4812216548067646e-06,
+      "loss": 0.4379,
+      "step": 5901
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2099332356892103,
+      "learning_rate": 2.479614287048805e-06,
+      "loss": 0.5683,
+      "step": 5902
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8170557459671952,
+      "learning_rate": 2.4780072684142355e-06,
+      "loss": 0.507,
+      "step": 5903
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1623949665197424,
+      "learning_rate": 2.476400599125664e-06,
+      "loss": 0.4321,
+      "step": 5904
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9566598946561105,
+      "learning_rate": 2.4747942794056425e-06,
+      "loss": 0.5124,
+      "step": 5905
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9278617931931055,
+      "learning_rate": 2.473188309476682e-06,
+      "loss": 0.5001,
+      "step": 5906
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2009734404072088,
+      "learning_rate": 2.4715826895612372e-06,
+      "loss": 0.4426,
+      "step": 5907
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0569089499298148,
+      "learning_rate": 2.4699774198817228e-06,
+      "loss": 0.5334,
+      "step": 5908
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4085875062925863,
+      "learning_rate": 2.4683725006604953e-06,
+      "loss": 0.4501,
+      "step": 5909
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.275463064603917,
+      "learning_rate": 2.466767932119875e-06,
+      "loss": 0.4302,
+      "step": 5910
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4495464204722053,
+      "learning_rate": 2.4651637144821176e-06,
+      "loss": 0.5357,
+      "step": 5911
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2745231301499715,
+      "learning_rate": 2.4635598479694455e-06,
+      "loss": 0.4778,
+      "step": 5912
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.95387042349627,
+      "learning_rate": 2.4619563328040226e-06,
+      "loss": 0.4717,
+      "step": 5913
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.338000430687956,
+      "learning_rate": 2.4603531692079703e-06,
+      "loss": 0.4823,
+      "step": 5914
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.828970515055067,
+      "learning_rate": 2.458750357403355e-06,
+      "loss": 0.6028,
+      "step": 5915
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.030293625647091,
+      "learning_rate": 2.457147897612202e-06,
+      "loss": 0.4697,
+      "step": 5916
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.457324888279669,
+      "learning_rate": 2.45554579005648e-06,
+      "loss": 0.4897,
+      "step": 5917
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.5149018114247985,
+      "learning_rate": 2.453944034958117e-06,
+      "loss": 0.5906,
+      "step": 5918
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.111930671468288,
+      "learning_rate": 2.4523426325389843e-06,
+      "loss": 0.5763,
+      "step": 5919
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8272807194201066,
+      "learning_rate": 2.4507415830209076e-06,
+      "loss": 0.681,
+      "step": 5920
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.19870236704104,
+      "learning_rate": 2.4491408866256667e-06,
+      "loss": 0.4728,
+      "step": 5921
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1619373629382297,
+      "learning_rate": 2.447540543574987e-06,
+      "loss": 0.5149,
+      "step": 5922
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8416824118970133,
+      "learning_rate": 2.4459405540905505e-06,
+      "loss": 0.7002,
+      "step": 5923
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.787171133351888,
+      "learning_rate": 2.4443409183939843e-06,
+      "loss": 0.5861,
+      "step": 5924
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3168382277292356,
+      "learning_rate": 2.4427416367068733e-06,
+      "loss": 0.4131,
+      "step": 5925
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8244781585766352,
+      "learning_rate": 2.4411427092507485e-06,
+      "loss": 0.4414,
+      "step": 5926
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6917434416431836,
+      "learning_rate": 2.439544136247092e-06,
+      "loss": 0.479,
+      "step": 5927
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9822450490727068,
+      "learning_rate": 2.437945917917336e-06,
+      "loss": 0.5427,
+      "step": 5928
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3105509495920047,
+      "learning_rate": 2.4363480544828692e-06,
+      "loss": 0.3691,
+      "step": 5929
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7290824275493988,
+      "learning_rate": 2.4347505461650252e-06,
+      "loss": 0.4979,
+      "step": 5930
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.058153629669663,
+      "learning_rate": 2.433153393185092e-06,
+      "loss": 0.4774,
+      "step": 5931
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8708556205538653,
+      "learning_rate": 2.431556595764305e-06,
+      "loss": 0.4148,
+      "step": 5932
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.534797694711912,
+      "learning_rate": 2.429960154123855e-06,
+      "loss": 0.5013,
+      "step": 5933
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8311032803583079,
+      "learning_rate": 2.4283640684848777e-06,
+      "loss": 0.6469,
+      "step": 5934
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8227949287226102,
+      "learning_rate": 2.426768339068466e-06,
+      "loss": 0.4366,
+      "step": 5935
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0716741185958862,
+      "learning_rate": 2.4251729660956563e-06,
+      "loss": 0.5072,
+      "step": 5936
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0368618310824407,
+      "learning_rate": 2.4235779497874433e-06,
+      "loss": 0.4663,
+      "step": 5937
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9303653734625565,
+      "learning_rate": 2.4219832903647667e-06,
+      "loss": 0.4678,
+      "step": 5938
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8610390800924805,
+      "learning_rate": 2.420388988048516e-06,
+      "loss": 0.5365,
+      "step": 5939
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2736576976171077,
+      "learning_rate": 2.418795043059538e-06,
+      "loss": 0.5545,
+      "step": 5940
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2571788289142782,
+      "learning_rate": 2.4172014556186214e-06,
+      "loss": 0.4829,
+      "step": 5941
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4508535760060868,
+      "learning_rate": 2.4156082259465135e-06,
+      "loss": 0.5832,
+      "step": 5942
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9465605839816635,
+      "learning_rate": 2.414015354263907e-06,
+      "loss": 0.4729,
+      "step": 5943
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.198921615696182,
+      "learning_rate": 2.412422840791446e-06,
+      "loss": 0.424,
+      "step": 5944
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5284007151253816,
+      "learning_rate": 2.410830685749722e-06,
+      "loss": 0.5808,
+      "step": 5945
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0027188132878093,
+      "learning_rate": 2.4092388893592856e-06,
+      "loss": 0.4738,
+      "step": 5946
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.387721754980416,
+      "learning_rate": 2.407647451840627e-06,
+      "loss": 0.3963,
+      "step": 5947
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9071194702160827,
+      "learning_rate": 2.406056373414197e-06,
+      "loss": 0.4922,
+      "step": 5948
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9357560730619225,
+      "learning_rate": 2.404465654300387e-06,
+      "loss": 0.5345,
+      "step": 5949
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4331672644261437,
+      "learning_rate": 2.402875294719546e-06,
+      "loss": 0.424,
+      "step": 5950
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9960314425091281,
+      "learning_rate": 2.4012852948919685e-06,
+      "loss": 0.4036,
+      "step": 5951
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9478962258208887,
+      "learning_rate": 2.399695655037903e-06,
+      "loss": 0.5203,
+      "step": 5952
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0279117417337145,
+      "learning_rate": 2.3981063753775437e-06,
+      "loss": 0.5042,
+      "step": 5953
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5617023860747241,
+      "learning_rate": 2.39651745613104e-06,
+      "loss": 0.4255,
+      "step": 5954
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9410015725332481,
+      "learning_rate": 2.3949288975184852e-06,
+      "loss": 0.5141,
+      "step": 5955
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8888283057345092,
+      "learning_rate": 2.393340699759931e-06,
+      "loss": 0.4686,
+      "step": 5956
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9654893684530157,
+      "learning_rate": 2.391752863075369e-06,
+      "loss": 0.5255,
+      "step": 5957
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6293294524419162,
+      "learning_rate": 2.3901653876847507e-06,
+      "loss": 0.4412,
+      "step": 5958
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3156004153884533,
+      "learning_rate": 2.388578273807971e-06,
+      "loss": 0.5249,
+      "step": 5959
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7606753747130122,
+      "learning_rate": 2.3869915216648766e-06,
+      "loss": 0.501,
+      "step": 5960
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.401625191080599,
+      "learning_rate": 2.385405131475262e-06,
+      "loss": 0.5061,
+      "step": 5961
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4407343178014203,
+      "learning_rate": 2.3838191034588774e-06,
+      "loss": 0.5863,
+      "step": 5962
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.331471916347791,
+      "learning_rate": 2.382233437835418e-06,
+      "loss": 0.3616,
+      "step": 5963
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9093641900961436,
+      "learning_rate": 2.380648134824527e-06,
+      "loss": 0.5257,
+      "step": 5964
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1927703336562963,
+      "learning_rate": 2.379063194645805e-06,
+      "loss": 0.5113,
+      "step": 5965
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5516660636358113,
+      "learning_rate": 2.3774786175187932e-06,
+      "loss": 0.5541,
+      "step": 5966
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.671485647554803,
+      "learning_rate": 2.3758944036629906e-06,
+      "loss": 0.4763,
+      "step": 5967
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2866112898900117,
+      "learning_rate": 2.3743105532978396e-06,
+      "loss": 0.5615,
+      "step": 5968
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8392508435778185,
+      "learning_rate": 2.372727066642737e-06,
+      "loss": 0.5083,
+      "step": 5969
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2012259284346043,
+      "learning_rate": 2.371143943917025e-06,
+      "loss": 0.5082,
+      "step": 5970
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.122949056670715,
+      "learning_rate": 2.3695611853399997e-06,
+      "loss": 0.3565,
+      "step": 5971
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.172474839746245,
+      "learning_rate": 2.3679787911309016e-06,
+      "loss": 0.4987,
+      "step": 5972
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.180674593448937,
+      "learning_rate": 2.366396761508928e-06,
+      "loss": 0.5726,
+      "step": 5973
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9020073263583017,
+      "learning_rate": 2.3648150966932163e-06,
+      "loss": 0.5499,
+      "step": 5974
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9741929707998185,
+      "learning_rate": 2.363233796902863e-06,
+      "loss": 0.4759,
+      "step": 5975
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5876454403560585,
+      "learning_rate": 2.361652862356906e-06,
+      "loss": 0.4912,
+      "step": 5976
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.071676875640921,
+      "learning_rate": 2.3600722932743407e-06,
+      "loss": 0.5367,
+      "step": 5977
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0376433727572096,
+      "learning_rate": 2.3584920898741003e-06,
+      "loss": 0.4038,
+      "step": 5978
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3827649805043736,
+      "learning_rate": 2.3569122523750804e-06,
+      "loss": 0.4669,
+      "step": 5979
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9950054070716192,
+      "learning_rate": 2.355332780996116e-06,
+      "loss": 0.5847,
+      "step": 5980
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9782160013929824,
+      "learning_rate": 2.3537536759559974e-06,
+      "loss": 0.499,
+      "step": 5981
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8530391521390959,
+      "learning_rate": 2.3521749374734602e-06,
+      "loss": 0.5523,
+      "step": 5982
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3967293067775794,
+      "learning_rate": 2.3505965657671943e-06,
+      "loss": 0.5352,
+      "step": 5983
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4430489469983714,
+      "learning_rate": 2.3490185610558324e-06,
+      "loss": 0.4885,
+      "step": 5984
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3367744425538715,
+      "learning_rate": 2.347440923557959e-06,
+      "loss": 0.3917,
+      "step": 5985
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3621241387968226,
+      "learning_rate": 2.345863653492111e-06,
+      "loss": 0.5136,
+      "step": 5986
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.778825982792816,
+      "learning_rate": 2.344286751076768e-06,
+      "loss": 0.3885,
+      "step": 5987
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4161980008752937,
+      "learning_rate": 2.342710216530366e-06,
+      "loss": 0.5154,
+      "step": 5988
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.397030237078764,
+      "learning_rate": 2.3411340500712833e-06,
+      "loss": 0.513,
+      "step": 5989
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1082244913607036,
+      "learning_rate": 2.339558251917853e-06,
+      "loss": 0.4425,
+      "step": 5990
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6383116268003521,
+      "learning_rate": 2.3379828222883504e-06,
+      "loss": 0.4463,
+      "step": 5991
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5284598906639957,
+      "learning_rate": 2.336407761401009e-06,
+      "loss": 0.5159,
+      "step": 5992
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.601763888594823,
+      "learning_rate": 2.3348330694740006e-06,
+      "loss": 0.4759,
+      "step": 5993
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.830866888235839,
+      "learning_rate": 2.333258746725458e-06,
+      "loss": 0.6617,
+      "step": 5994
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3371317127458706,
+      "learning_rate": 2.3316847933734478e-06,
+      "loss": 0.5051,
+      "step": 5995
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7351471305002308,
+      "learning_rate": 2.3301112096359996e-06,
+      "loss": 0.5742,
+      "step": 5996
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5085720921505628,
+      "learning_rate": 2.3285379957310827e-06,
+      "loss": 0.4256,
+      "step": 5997
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.493315681001551,
+      "learning_rate": 2.3269651518766217e-06,
+      "loss": 0.5224,
+      "step": 5998
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3946817070817255,
+      "learning_rate": 2.3253926782904833e-06,
+      "loss": 0.4506,
+      "step": 5999
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0483383656849172,
+      "learning_rate": 2.323820575190489e-06,
+      "loss": 0.5543,
+      "step": 6000
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.4177176319511386,
+      "learning_rate": 2.322248842794404e-06,
+      "loss": 0.5128,
+      "step": 6001
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.767185215870721,
+      "learning_rate": 2.320677481319947e-06,
+      "loss": 0.4102,
+      "step": 6002
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8607950468720453,
+      "learning_rate": 2.319106490984781e-06,
+      "loss": 0.4601,
+      "step": 6003
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.916011155192851,
+      "learning_rate": 2.3175358720065183e-06,
+      "loss": 0.4539,
+      "step": 6004
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9652054173281073,
+      "learning_rate": 2.3159656246027234e-06,
+      "loss": 0.4496,
+      "step": 6005
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7717552441787747,
+      "learning_rate": 2.3143957489909037e-06,
+      "loss": 0.4218,
+      "step": 6006
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1200393700223628,
+      "learning_rate": 2.312826245388521e-06,
+      "loss": 0.5103,
+      "step": 6007
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2541996338483004,
+      "learning_rate": 2.31125711401298e-06,
+      "loss": 0.5237,
+      "step": 6008
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7658195017258035,
+      "learning_rate": 2.3096883550816395e-06,
+      "loss": 0.649,
+      "step": 6009
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3785627947208035,
+      "learning_rate": 2.3081199688118e-06,
+      "loss": 0.4651,
+      "step": 6010
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7022018774681955,
+      "learning_rate": 2.3065519554207204e-06,
+      "loss": 0.4521,
+      "step": 6011
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2525920271046274,
+      "learning_rate": 2.3049843151255933e-06,
+      "loss": 0.5859,
+      "step": 6012
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8242512190761943,
+      "learning_rate": 2.303417048143574e-06,
+      "loss": 0.6502,
+      "step": 6013
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.426969402258358,
+      "learning_rate": 2.3018501546917567e-06,
+      "loss": 0.4548,
+      "step": 6014
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.280137146410003,
+      "learning_rate": 2.3002836349871897e-06,
+      "loss": 0.4785,
+      "step": 6015
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8456208434060648,
+      "learning_rate": 2.298717489246865e-06,
+      "loss": 0.523,
+      "step": 6016
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8362729219772587,
+      "learning_rate": 2.297151717687727e-06,
+      "loss": 0.6838,
+      "step": 6017
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9127212543216146,
+      "learning_rate": 2.295586320526663e-06,
+      "loss": 0.5422,
+      "step": 6018
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8800372750483636,
+      "learning_rate": 2.294021297980516e-06,
+      "loss": 0.4279,
+      "step": 6019
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9383113619130719,
+      "learning_rate": 2.2924566502660676e-06,
+      "loss": 0.4367,
+      "step": 6020
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1918159048233896,
+      "learning_rate": 2.2908923776000573e-06,
+      "loss": 0.5104,
+      "step": 6021
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.84951228496165,
+      "learning_rate": 2.289328480199164e-06,
+      "loss": 0.5931,
+      "step": 6022
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3453592692779086,
+      "learning_rate": 2.2877649582800216e-06,
+      "loss": 0.5014,
+      "step": 6023
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.2790843728307495,
+      "learning_rate": 2.2862018120592072e-06,
+      "loss": 0.4943,
+      "step": 6024
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.684177182594862,
+      "learning_rate": 2.284639041753246e-06,
+      "loss": 0.6272,
+      "step": 6025
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3795482414524782,
+      "learning_rate": 2.2830766475786166e-06,
+      "loss": 0.4736,
+      "step": 6026
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7121226588457445,
+      "learning_rate": 2.281514629751737e-06,
+      "loss": 0.4134,
+      "step": 6027
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5526451557591985,
+      "learning_rate": 2.2799529884889827e-06,
+      "loss": 0.5957,
+      "step": 6028
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0480610650375564,
+      "learning_rate": 2.278391724006669e-06,
+      "loss": 0.4511,
+      "step": 6029
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.297968189657318,
+      "learning_rate": 2.2768308365210616e-06,
+      "loss": 0.4794,
+      "step": 6030
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2946117563911375,
+      "learning_rate": 2.275270326248374e-06,
+      "loss": 0.5546,
+      "step": 6031
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9467766965516904,
+      "learning_rate": 2.2737101934047707e-06,
+      "loss": 0.5186,
+      "step": 6032
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7311912367817985,
+      "learning_rate": 2.2721504382063567e-06,
+      "loss": 0.515,
+      "step": 6033
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.915806268101038,
+      "learning_rate": 2.270591060869194e-06,
+      "loss": 0.3993,
+      "step": 6034
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6898429659122811,
+      "learning_rate": 2.2690320616092826e-06,
+      "loss": 0.5236,
+      "step": 6035
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0261210795281093,
+      "learning_rate": 2.267473440642579e-06,
+      "loss": 0.6032,
+      "step": 6036
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8855461015696047,
+      "learning_rate": 2.2659151981849793e-06,
+      "loss": 0.4586,
+      "step": 6037
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7773264561500792,
+      "learning_rate": 2.2643573344523345e-06,
+      "loss": 0.4123,
+      "step": 6038
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7245042829614439,
+      "learning_rate": 2.2627998496604366e-06,
+      "loss": 0.4524,
+      "step": 6039
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.025832286791638,
+      "learning_rate": 2.2612427440250308e-06,
+      "loss": 0.5347,
+      "step": 6040
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.343866946337479,
+      "learning_rate": 2.2596860177618034e-06,
+      "loss": 0.4964,
+      "step": 6041
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6663491910356623,
+      "learning_rate": 2.2581296710863963e-06,
+      "loss": 0.4953,
+      "step": 6042
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9073048903471526,
+      "learning_rate": 2.256573704214393e-06,
+      "loss": 0.474,
+      "step": 6043
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8172922955327782,
+      "learning_rate": 2.2550181173613226e-06,
+      "loss": 0.4244,
+      "step": 6044
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.39065109584252,
+      "learning_rate": 2.253462910742669e-06,
+      "loss": 0.5052,
+      "step": 6045
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.210920749014501,
+      "learning_rate": 2.2519080845738573e-06,
+      "loss": 0.4114,
+      "step": 6046
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.951174338119239,
+      "learning_rate": 2.2503536390702603e-06,
+      "loss": 0.5665,
+      "step": 6047
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8670668256813343,
+      "learning_rate": 2.248799574447202e-06,
+      "loss": 0.525,
+      "step": 6048
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6621342417821054,
+      "learning_rate": 2.2472458909199507e-06,
+      "loss": 0.5673,
+      "step": 6049
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.554834483288534,
+      "learning_rate": 2.2456925887037194e-06,
+      "loss": 0.4641,
+      "step": 6050
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2584445192667904,
+      "learning_rate": 2.2441396680136763e-06,
+      "loss": 0.5412,
+      "step": 6051
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3865652259470047,
+      "learning_rate": 2.242587129064927e-06,
+      "loss": 0.5597,
+      "step": 6052
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8739800679639163,
+      "learning_rate": 2.2410349720725327e-06,
+      "loss": 0.6751,
+      "step": 6053
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.412909229055911,
+      "learning_rate": 2.239483197251494e-06,
+      "loss": 0.5377,
+      "step": 6054
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.09882819213224,
+      "learning_rate": 2.237931804816767e-06,
+      "loss": 0.5065,
+      "step": 6055
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.267491590171007,
+      "learning_rate": 2.2363807949832463e-06,
+      "loss": 0.4412,
+      "step": 6056
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.120182952040126,
+      "learning_rate": 2.2348301679657802e-06,
+      "loss": 0.4912,
+      "step": 6057
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.938282558794569,
+      "learning_rate": 2.233279923979159e-06,
+      "loss": 0.4771,
+      "step": 6058
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7519178395675385,
+      "learning_rate": 2.2317300632381256e-06,
+      "loss": 0.4349,
+      "step": 6059
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9733519960816648,
+      "learning_rate": 2.230180585957362e-06,
+      "loss": 0.523,
+      "step": 6060
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1321645463668544,
+      "learning_rate": 2.2286314923515077e-06,
+      "loss": 0.5003,
+      "step": 6061
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.64313885793828,
+      "learning_rate": 2.227082782635136e-06,
+      "loss": 0.3769,
+      "step": 6062
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.967640964137952,
+      "learning_rate": 2.225534457022778e-06,
+      "loss": 0.558,
+      "step": 6063
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3766480462337616,
+      "learning_rate": 2.2239865157289046e-06,
+      "loss": 0.5446,
+      "step": 6064
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.050222391588646,
+      "learning_rate": 2.2224389589679407e-06,
+      "loss": 0.4218,
+      "step": 6065
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.055357287815045,
+      "learning_rate": 2.220891786954249e-06,
+      "loss": 0.577,
+      "step": 6066
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8057753645341201,
+      "learning_rate": 2.2193449999021476e-06,
+      "loss": 0.4971,
+      "step": 6067
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.304950218425839,
+      "learning_rate": 2.2177985980258946e-06,
+      "loss": 0.5512,
+      "step": 6068
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0380778208482693,
+      "learning_rate": 2.216252581539697e-06,
+      "loss": 0.5207,
+      "step": 6069
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7706933171083448,
+      "learning_rate": 2.2147069506577107e-06,
+      "loss": 0.4179,
+      "step": 6070
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4421290502312956,
+      "learning_rate": 2.2131617055940337e-06,
+      "loss": 0.5206,
+      "step": 6071
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.537186205080983,
+      "learning_rate": 2.2116168465627162e-06,
+      "loss": 0.4471,
+      "step": 6072
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0813377832526645,
+      "learning_rate": 2.2100723737777485e-06,
+      "loss": 0.4374,
+      "step": 6073
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2356801087220806,
+      "learning_rate": 2.208528287453074e-06,
+      "loss": 0.4669,
+      "step": 6074
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2001266916823154,
+      "learning_rate": 2.206984587802576e-06,
+      "loss": 0.5854,
+      "step": 6075
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3760486093599322,
+      "learning_rate": 2.205441275040091e-06,
+      "loss": 0.5073,
+      "step": 6076
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.174556509685616,
+      "learning_rate": 2.203898349379394e-06,
+      "loss": 0.4316,
+      "step": 6077
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0495525453349313,
+      "learning_rate": 2.202355811034218e-06,
+      "loss": 0.4381,
+      "step": 6078
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.127073549575823,
+      "learning_rate": 2.2008136602182264e-06,
+      "loss": 0.4969,
+      "step": 6079
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8252337778207794,
+      "learning_rate": 2.1992718971450432e-06,
+      "loss": 0.5506,
+      "step": 6080
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1426027074195924,
+      "learning_rate": 2.197730522028231e-06,
+      "loss": 0.4884,
+      "step": 6081
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9614088403641654,
+      "learning_rate": 2.196189535081302e-06,
+      "loss": 0.4805,
+      "step": 6082
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0584289502178366,
+      "learning_rate": 2.1946489365177122e-06,
+      "loss": 0.5545,
+      "step": 6083
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5501003820303723,
+      "learning_rate": 2.1931087265508674e-06,
+      "loss": 0.4662,
+      "step": 6084
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8905175041864024,
+      "learning_rate": 2.191568905394113e-06,
+      "loss": 0.5319,
+      "step": 6085
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4364584831944076,
+      "learning_rate": 2.19002947326075e-06,
+      "loss": 0.4951,
+      "step": 6086
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.558004712466623,
+      "learning_rate": 2.1884904303640155e-06,
+      "loss": 0.5673,
+      "step": 6087
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4434517789479515,
+      "learning_rate": 2.1869517769171016e-06,
+      "loss": 0.5213,
+      "step": 6088
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8107592346694084,
+      "learning_rate": 2.1854135131331405e-06,
+      "loss": 0.4808,
+      "step": 6089
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3228733520741947,
+      "learning_rate": 2.1838756392252098e-06,
+      "loss": 0.4606,
+      "step": 6090
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8932009453838858,
+      "learning_rate": 2.1823381554063398e-06,
+      "loss": 0.4332,
+      "step": 6091
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8639192485031266,
+      "learning_rate": 2.180801061889499e-06,
+      "loss": 0.5313,
+      "step": 6092
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3291430238693276,
+      "learning_rate": 2.1792643588876085e-06,
+      "loss": 0.5016,
+      "step": 6093
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.010188865865139,
+      "learning_rate": 2.177728046613528e-06,
+      "loss": 0.5202,
+      "step": 6094
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.530577327571859,
+      "learning_rate": 2.1761921252800737e-06,
+      "loss": 0.535,
+      "step": 6095
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8917624567800624,
+      "learning_rate": 2.174656595099994e-06,
+      "loss": 0.4239,
+      "step": 6096
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3008669801466186,
+      "learning_rate": 2.1731214562859942e-06,
+      "loss": 0.4451,
+      "step": 6097
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8878740422483733,
+      "learning_rate": 2.17158670905072e-06,
+      "loss": 0.4722,
+      "step": 6098
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.883008924914656,
+      "learning_rate": 2.1700523536067657e-06,
+      "loss": 0.4749,
+      "step": 6099
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0653937254471395,
+      "learning_rate": 2.168518390166668e-06,
+      "loss": 0.5465,
+      "step": 6100
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.792381014415643,
+      "learning_rate": 2.1669848189429136e-06,
+      "loss": 0.4474,
+      "step": 6101
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9723262229697558,
+      "learning_rate": 2.1654516401479303e-06,
+      "loss": 0.5592,
+      "step": 6102
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5850032701320735,
+      "learning_rate": 2.1639188539940968e-06,
+      "loss": 0.5041,
+      "step": 6103
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6468278591580296,
+      "learning_rate": 2.16238646069373e-06,
+      "loss": 0.5583,
+      "step": 6104
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2249743222432357,
+      "learning_rate": 2.1608544604591015e-06,
+      "loss": 0.3791,
+      "step": 6105
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8510480382730115,
+      "learning_rate": 2.1593228535024193e-06,
+      "loss": 0.4955,
+      "step": 6106
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.390802424236188,
+      "learning_rate": 2.1577916400358452e-06,
+      "loss": 0.4436,
+      "step": 6107
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.038227746960908,
+      "learning_rate": 2.1562608202714806e-06,
+      "loss": 0.5755,
+      "step": 6108
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.877853801657469,
+      "learning_rate": 2.1547303944213733e-06,
+      "loss": 0.549,
+      "step": 6109
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0373242384895027,
+      "learning_rate": 2.1532003626975204e-06,
+      "loss": 0.5007,
+      "step": 6110
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.946137803721126,
+      "learning_rate": 2.1516707253118586e-06,
+      "loss": 0.5437,
+      "step": 6111
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9595302840442244,
+      "learning_rate": 2.1501414824762763e-06,
+      "loss": 0.4117,
+      "step": 6112
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0512688963064463,
+      "learning_rate": 2.1486126344026027e-06,
+      "loss": 0.5599,
+      "step": 6113
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.162042152817863,
+      "learning_rate": 2.147084181302612e-06,
+      "loss": 0.4656,
+      "step": 6114
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0010939703518167,
+      "learning_rate": 2.145556123388026e-06,
+      "loss": 0.5207,
+      "step": 6115
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8299497088832897,
+      "learning_rate": 2.144028460870512e-06,
+      "loss": 0.5228,
+      "step": 6116
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.93564423568438,
+      "learning_rate": 2.1425011939616795e-06,
+      "loss": 0.7195,
+      "step": 6117
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1355027857381965,
+      "learning_rate": 2.1409743228730883e-06,
+      "loss": 0.5015,
+      "step": 6118
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4675885401992184,
+      "learning_rate": 2.139447847816237e-06,
+      "loss": 0.552,
+      "step": 6119
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9722422866494538,
+      "learning_rate": 2.1379217690025767e-06,
+      "loss": 0.5809,
+      "step": 6120
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7901429663762329,
+      "learning_rate": 2.1363960866434947e-06,
+      "loss": 0.6383,
+      "step": 6121
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.058591554757942,
+      "learning_rate": 2.1348708009503333e-06,
+      "loss": 0.5086,
+      "step": 6122
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.415292567200039,
+      "learning_rate": 2.1333459121343696e-06,
+      "loss": 0.5396,
+      "step": 6123
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1241612785901927,
+      "learning_rate": 2.131821420406836e-06,
+      "loss": 0.5066,
+      "step": 6124
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9311360003306395,
+      "learning_rate": 2.1302973259789004e-06,
+      "loss": 0.4979,
+      "step": 6125
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.552352369215579,
+      "learning_rate": 2.1287736290616845e-06,
+      "loss": 0.4483,
+      "step": 6126
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.340873814636964,
+      "learning_rate": 2.127250329866248e-06,
+      "loss": 0.4862,
+      "step": 6127
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.498350236841117,
+      "learning_rate": 2.1257274286035963e-06,
+      "loss": 0.4683,
+      "step": 6128
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3032079196746844,
+      "learning_rate": 2.124204925484685e-06,
+      "loss": 0.5819,
+      "step": 6129
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.317560894290641,
+      "learning_rate": 2.122682820720409e-06,
+      "loss": 0.5879,
+      "step": 6130
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0970471619186832,
+      "learning_rate": 2.121161114521609e-06,
+      "loss": 0.46,
+      "step": 6131
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.667528489541149,
+      "learning_rate": 2.119639807099075e-06,
+      "loss": 0.4275,
+      "step": 6132
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2976533226879177,
+      "learning_rate": 2.1181188986635354e-06,
+      "loss": 0.5435,
+      "step": 6133
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1458132577211506,
+      "learning_rate": 2.1165983894256647e-06,
+      "loss": 0.5011,
+      "step": 6134
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0309115832312896,
+      "learning_rate": 2.1150782795960884e-06,
+      "loss": 0.5232,
+      "step": 6135
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9981742421359077,
+      "learning_rate": 2.1135585693853665e-06,
+      "loss": 0.5973,
+      "step": 6136
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7767445531609898,
+      "learning_rate": 2.112039259004014e-06,
+      "loss": 0.4985,
+      "step": 6137
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6687932679965702,
+      "learning_rate": 2.110520348662481e-06,
+      "loss": 0.4441,
+      "step": 6138
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7915572571460514,
+      "learning_rate": 2.109001838571171e-06,
+      "loss": 0.4352,
+      "step": 6139
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8190769478857969,
+      "learning_rate": 2.107483728940423e-06,
+      "loss": 0.4775,
+      "step": 6140
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7629953896006052,
+      "learning_rate": 2.1059660199805303e-06,
+      "loss": 0.4981,
+      "step": 6141
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1676216541481166,
+      "learning_rate": 2.1044487119017215e-06,
+      "loss": 0.4625,
+      "step": 6142
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6228635225367642,
+      "learning_rate": 2.1029318049141772e-06,
+      "loss": 0.4988,
+      "step": 6143
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.280088133970521,
+      "learning_rate": 2.101415299228016e-06,
+      "loss": 0.4868,
+      "step": 6144
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8920051943358118,
+      "learning_rate": 2.0998991950533065e-06,
+      "loss": 0.5052,
+      "step": 6145
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6644437251131567,
+      "learning_rate": 2.098383492600059e-06,
+      "loss": 0.4929,
+      "step": 6146
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.6837673804815196,
+      "learning_rate": 2.0968681920782273e-06,
+      "loss": 0.3802,
+      "step": 6147
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.61637877115338,
+      "learning_rate": 2.095353293697709e-06,
+      "loss": 0.3433,
+      "step": 6148
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6086714137986147,
+      "learning_rate": 2.093838797668351e-06,
+      "loss": 0.5591,
+      "step": 6149
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7556559996098247,
+      "learning_rate": 2.092324704199938e-06,
+      "loss": 0.3852,
+      "step": 6150
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8223384893108303,
+      "learning_rate": 2.0908110135022046e-06,
+      "loss": 0.4682,
+      "step": 6151
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0452727998168636,
+      "learning_rate": 2.089297725784824e-06,
+      "loss": 0.4124,
+      "step": 6152
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8032088942780174,
+      "learning_rate": 2.08778484125742e-06,
+      "loss": 0.6352,
+      "step": 6153
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7910021896167811,
+      "learning_rate": 2.0862723601295557e-06,
+      "loss": 0.494,
+      "step": 6154
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9810359064632312,
+      "learning_rate": 2.084760282610738e-06,
+      "loss": 0.4643,
+      "step": 6155
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.0203997226600245,
+      "learning_rate": 2.083248608910422e-06,
+      "loss": 0.5457,
+      "step": 6156
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6991263500533322,
+      "learning_rate": 2.081737339238002e-06,
+      "loss": 0.4716,
+      "step": 6157
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.138702172592572,
+      "learning_rate": 2.0802264738028223e-06,
+      "loss": 0.4939,
+      "step": 6158
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2168756925803073,
+      "learning_rate": 2.0787160128141636e-06,
+      "loss": 0.5763,
+      "step": 6159
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4799559328896272,
+      "learning_rate": 2.077205956481259e-06,
+      "loss": 0.4167,
+      "step": 6160
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4093469195180695,
+      "learning_rate": 2.075696305013277e-06,
+      "loss": 0.4457,
+      "step": 6161
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7399593041718693,
+      "learning_rate": 2.074187058619338e-06,
+      "loss": 0.4775,
+      "step": 6162
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2001979250913606,
+      "learning_rate": 2.0726782175085016e-06,
+      "loss": 0.4942,
+      "step": 6163
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.61454728741997,
+      "learning_rate": 2.071169781889771e-06,
+      "loss": 0.5114,
+      "step": 6164
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.919899000271575,
+      "learning_rate": 2.069661751972093e-06,
+      "loss": 0.4536,
+      "step": 6165
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.741206933866748,
+      "learning_rate": 2.068154127964363e-06,
+      "loss": 0.5249,
+      "step": 6166
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3320957648737783,
+      "learning_rate": 2.0666469100754143e-06,
+      "loss": 0.4258,
+      "step": 6167
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.077486490704608,
+      "learning_rate": 2.065140098514029e-06,
+      "loss": 0.4958,
+      "step": 6168
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.32855634515467,
+      "learning_rate": 2.063633693488927e-06,
+      "loss": 0.429,
+      "step": 6169
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.743553259848041,
+      "learning_rate": 2.062127695208779e-06,
+      "loss": 0.5308,
+      "step": 6170
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.476298738517536,
+      "learning_rate": 2.060622103882192e-06,
+      "loss": 0.5192,
+      "step": 6171
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.889008293451934,
+      "learning_rate": 2.0591169197177244e-06,
+      "loss": 0.4916,
+      "step": 6172
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.4069098636298296,
+      "learning_rate": 2.0576121429238718e-06,
+      "loss": 0.4604,
+      "step": 6173
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8822815695214004,
+      "learning_rate": 2.0561077737090727e-06,
+      "loss": 0.4753,
+      "step": 6174
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.368589933780417,
+      "learning_rate": 2.0546038122817173e-06,
+      "loss": 0.5475,
+      "step": 6175
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9100669407726696,
+      "learning_rate": 2.05310025885013e-06,
+      "loss": 0.4698,
+      "step": 6176
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.997181869328449,
+      "learning_rate": 2.051597113622586e-06,
+      "loss": 0.3963,
+      "step": 6177
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7204518723701758,
+      "learning_rate": 2.0500943768072974e-06,
+      "loss": 0.5208,
+      "step": 6178
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6637508339897304,
+      "learning_rate": 2.0485920486124265e-06,
+      "loss": 0.4078,
+      "step": 6179
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.007682168111539,
+      "learning_rate": 2.0470901292460736e-06,
+      "loss": 0.4717,
+      "step": 6180
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1329703553079677,
+      "learning_rate": 2.045588618916285e-06,
+      "loss": 0.407,
+      "step": 6181
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1834981646381526,
+      "learning_rate": 2.0440875178310473e-06,
+      "loss": 0.5337,
+      "step": 6182
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.648299323472207,
+      "learning_rate": 2.0425868261982963e-06,
+      "loss": 0.4546,
+      "step": 6183
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7842332931902416,
+      "learning_rate": 2.0410865442259042e-06,
+      "loss": 0.5344,
+      "step": 6184
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7609632770206933,
+      "learning_rate": 2.0395866721216935e-06,
+      "loss": 0.5313,
+      "step": 6185
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4779027234914652,
+      "learning_rate": 2.038087210093422e-06,
+      "loss": 0.4022,
+      "step": 6186
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.038887924103739,
+      "learning_rate": 2.036588158348799e-06,
+      "loss": 0.4693,
+      "step": 6187
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7939348539677558,
+      "learning_rate": 2.0350895170954693e-06,
+      "loss": 0.442,
+      "step": 6188
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2302041573320897,
+      "learning_rate": 2.0335912865410277e-06,
+      "loss": 0.5002,
+      "step": 6189
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1776709688108418,
+      "learning_rate": 2.032093466893006e-06,
+      "loss": 0.4584,
+      "step": 6190
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.114152902373883,
+      "learning_rate": 2.0305960583588853e-06,
+      "loss": 0.5551,
+      "step": 6191
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8550707282687924,
+      "learning_rate": 2.0290990611460836e-06,
+      "loss": 0.5224,
+      "step": 6192
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6579874462889062,
+      "learning_rate": 2.0276024754619634e-06,
+      "loss": 0.4969,
+      "step": 6193
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.224697023168764,
+      "learning_rate": 2.026106301513836e-06,
+      "loss": 0.5534,
+      "step": 6194
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6969206057109427,
+      "learning_rate": 2.024610539508946e-06,
+      "loss": 0.4905,
+      "step": 6195
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0992038091722183,
+      "learning_rate": 2.023115189654491e-06,
+      "loss": 0.4442,
+      "step": 6196
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2469119318632256,
+      "learning_rate": 2.0216202521576045e-06,
+      "loss": 0.431,
+      "step": 6197
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.413811938273658,
+      "learning_rate": 2.0201257272253643e-06,
+      "loss": 0.4102,
+      "step": 6198
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.778002199721246,
+      "learning_rate": 2.0186316150647913e-06,
+      "loss": 0.4312,
+      "step": 6199
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1068346712180666,
+      "learning_rate": 2.017137915882851e-06,
+      "loss": 0.5234,
+      "step": 6200
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1603520531305187,
+      "learning_rate": 2.015644629886449e-06,
+      "loss": 0.4398,
+      "step": 6201
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8462567148752724,
+      "learning_rate": 2.014151757282438e-06,
+      "loss": 0.4742,
+      "step": 6202
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8544134266578187,
+      "learning_rate": 2.012659298277606e-06,
+      "loss": 0.435,
+      "step": 6203
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.421708215378694,
+      "learning_rate": 2.011167253078693e-06,
+      "loss": 0.5405,
+      "step": 6204
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1184987231673573,
+      "learning_rate": 2.0096756218923725e-06,
+      "loss": 0.4399,
+      "step": 6205
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9447532301703763,
+      "learning_rate": 2.0081844049252686e-06,
+      "loss": 0.48,
+      "step": 6206
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6929067632712549,
+      "learning_rate": 2.0066936023839406e-06,
+      "loss": 0.4173,
+      "step": 6207
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8269653347921048,
+      "learning_rate": 2.0052032144748982e-06,
+      "loss": 0.6632,
+      "step": 6208
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7998500735895029,
+      "learning_rate": 2.003713241404586e-06,
+      "loss": 0.4976,
+      "step": 6209
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.801614498341954,
+      "learning_rate": 2.002223683379399e-06,
+      "loss": 0.5235,
+      "step": 6210
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8173602610340023,
+      "learning_rate": 2.000734540605666e-06,
+      "loss": 0.5476,
+      "step": 6211
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7996769122619494,
+      "learning_rate": 1.999245813289667e-06,
+      "loss": 0.4999,
+      "step": 6212
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8060480864988084,
+      "learning_rate": 1.9977575016376177e-06,
+      "loss": 0.6504,
+      "step": 6213
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.722484353457214,
+      "learning_rate": 1.9962696058556795e-06,
+      "loss": 0.3862,
+      "step": 6214
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.006774140409978,
+      "learning_rate": 1.9947821261499533e-06,
+      "loss": 0.4603,
+      "step": 6215
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.27598932734707,
+      "learning_rate": 1.9932950627264884e-06,
+      "loss": 0.5607,
+      "step": 6216
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.355590951741181,
+      "learning_rate": 1.991808415791269e-06,
+      "loss": 0.4253,
+      "step": 6217
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3720431276795333,
+      "learning_rate": 1.9903221855502285e-06,
+      "loss": 0.5401,
+      "step": 6218
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.053626225769318,
+      "learning_rate": 1.9888363722092376e-06,
+      "loss": 0.5096,
+      "step": 6219
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2082417304700144,
+      "learning_rate": 1.987350975974109e-06,
+      "loss": 0.4925,
+      "step": 6220
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.49802811561495,
+      "learning_rate": 1.9858659970506027e-06,
+      "loss": 0.5357,
+      "step": 6221
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6328310617516402,
+      "learning_rate": 1.984381435644415e-06,
+      "loss": 0.4345,
+      "step": 6222
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4672107895408186,
+      "learning_rate": 1.982897291961191e-06,
+      "loss": 0.5537,
+      "step": 6223
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6759273335238465,
+      "learning_rate": 1.9814135662065093e-06,
+      "loss": 0.5496,
+      "step": 6224
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5134969914874206,
+      "learning_rate": 1.9799302585858988e-06,
+      "loss": 0.4352,
+      "step": 6225
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9424304226522964,
+      "learning_rate": 1.9784473693048245e-06,
+      "loss": 0.4412,
+      "step": 6226
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8404046902826372,
+      "learning_rate": 1.976964898568699e-06,
+      "loss": 0.6816,
+      "step": 6227
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9337040045585845,
+      "learning_rate": 1.9754828465828703e-06,
+      "loss": 0.4588,
+      "step": 6228
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.7013419410950563,
+      "learning_rate": 1.9740012135526358e-06,
+      "loss": 0.4941,
+      "step": 6229
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0858433522882223,
+      "learning_rate": 1.972519999683229e-06,
+      "loss": 0.5073,
+      "step": 6230
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.677424612388146,
+      "learning_rate": 1.9710392051798273e-06,
+      "loss": 0.4737,
+      "step": 6231
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.234617325019333,
+      "learning_rate": 1.969558830247549e-06,
+      "loss": 0.3999,
+      "step": 6232
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9047489057037716,
+      "learning_rate": 1.9680788750914575e-06,
+      "loss": 0.4204,
+      "step": 6233
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9018951529660804,
+      "learning_rate": 1.966599339916554e-06,
+      "loss": 0.4785,
+      "step": 6234
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8087597407239835,
+      "learning_rate": 1.9651202249277862e-06,
+      "loss": 0.6718,
+      "step": 6235
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.356844274535127,
+      "learning_rate": 1.9636415303300373e-06,
+      "loss": 0.4949,
+      "step": 6236
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.788468189459655,
+      "learning_rate": 1.9621632563281394e-06,
+      "loss": 0.6597,
+      "step": 6237
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4429863758106203,
+      "learning_rate": 1.960685403126861e-06,
+      "loss": 0.4895,
+      "step": 6238
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8907398797635535,
+      "learning_rate": 1.9592079709309126e-06,
+      "loss": 0.5394,
+      "step": 6239
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3376738180249794,
+      "learning_rate": 1.9577309599449513e-06,
+      "loss": 0.5212,
+      "step": 6240
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.9821480341518827,
+      "learning_rate": 1.9562543703735683e-06,
+      "loss": 0.4829,
+      "step": 6241
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3338223294549216,
+      "learning_rate": 1.9547782024213047e-06,
+      "loss": 0.4895,
+      "step": 6242
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.153309713192778,
+      "learning_rate": 1.9533024562926355e-06,
+      "loss": 0.4806,
+      "step": 6243
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.448066731413163,
+      "learning_rate": 1.9518271321919837e-06,
+      "loss": 0.4947,
+      "step": 6244
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5103778473413,
+      "learning_rate": 1.950352230323708e-06,
+      "loss": 0.4347,
+      "step": 6245
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0725462049302474,
+      "learning_rate": 1.9488777508921155e-06,
+      "loss": 0.5094,
+      "step": 6246
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1041108523691614,
+      "learning_rate": 1.9474036941014473e-06,
+      "loss": 0.4277,
+      "step": 6247
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.862483725065435,
+      "learning_rate": 1.945930060155892e-06,
+      "loss": 0.4024,
+      "step": 6248
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.007937062358391,
+      "learning_rate": 1.9444568492595727e-06,
+      "loss": 0.5861,
+      "step": 6249
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.254252500077309,
+      "learning_rate": 1.942984061616564e-06,
+      "loss": 0.4892,
+      "step": 6250
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.553542929158215,
+      "learning_rate": 1.941511697430871e-06,
+      "loss": 0.422,
+      "step": 6251
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9596478313299146,
+      "learning_rate": 1.9400397569064505e-06,
+      "loss": 0.4766,
+      "step": 6252
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4194313290806257,
+      "learning_rate": 1.9385682402471913e-06,
+      "loss": 0.5291,
+      "step": 6253
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7095678423697396,
+      "learning_rate": 1.9370971476569308e-06,
+      "loss": 0.4047,
+      "step": 6254
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.252417469180771,
+      "learning_rate": 1.93562647933944e-06,
+      "loss": 0.5313,
+      "step": 6255
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.119473766716319,
+      "learning_rate": 1.934156235498442e-06,
+      "loss": 0.4964,
+      "step": 6256
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9912461949435232,
+      "learning_rate": 1.93268641633759e-06,
+      "loss": 0.527,
+      "step": 6257
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1668886299740957,
+      "learning_rate": 1.931217022060483e-06,
+      "loss": 0.4919,
+      "step": 6258
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.024780992690607,
+      "learning_rate": 1.929748052870664e-06,
+      "loss": 0.4308,
+      "step": 6259
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.5129359053767404,
+      "learning_rate": 1.9282795089716116e-06,
+      "loss": 0.3991,
+      "step": 6260
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.513409482021169,
+      "learning_rate": 1.9268113905667514e-06,
+      "loss": 0.437,
+      "step": 6261
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.251324999915131,
+      "learning_rate": 1.9253436978594433e-06,
+      "loss": 0.5511,
+      "step": 6262
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0221517695317925,
+      "learning_rate": 1.923876431052995e-06,
+      "loss": 0.5427,
+      "step": 6263
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7938132473863428,
+      "learning_rate": 1.922409590350651e-06,
+      "loss": 0.4492,
+      "step": 6264
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5448970597681058,
+      "learning_rate": 1.9209431759555973e-06,
+      "loss": 0.4506,
+      "step": 6265
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2834376005333827,
+      "learning_rate": 1.91947718807096e-06,
+      "loss": 0.4852,
+      "step": 6266
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.5338523539634,
+      "learning_rate": 1.9180116268998104e-06,
+      "loss": 0.5126,
+      "step": 6267
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7748821023872656,
+      "learning_rate": 1.9165464926451556e-06,
+      "loss": 0.4952,
+      "step": 6268
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3581543930093014,
+      "learning_rate": 1.9150817855099473e-06,
+      "loss": 0.5251,
+      "step": 6269
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.22389756467088,
+      "learning_rate": 1.9136175056970747e-06,
+      "loss": 0.5504,
+      "step": 6270
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2950426306359484,
+      "learning_rate": 1.9121536534093723e-06,
+      "loss": 0.5849,
+      "step": 6271
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.760639573534165,
+      "learning_rate": 1.9106902288496087e-06,
+      "loss": 0.4468,
+      "step": 6272
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8493484806422544,
+      "learning_rate": 1.9092272322205013e-06,
+      "loss": 0.4984,
+      "step": 6273
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.6911902223139634,
+      "learning_rate": 1.907764663724701e-06,
+      "loss": 0.4536,
+      "step": 6274
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8380775182716391,
+      "learning_rate": 1.9063025235648058e-06,
+      "loss": 0.6931,
+      "step": 6275
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.987491301459409,
+      "learning_rate": 1.904840811943347e-06,
+      "loss": 0.4554,
+      "step": 6276
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2733647509690442,
+      "learning_rate": 1.903379529062805e-06,
+      "loss": 0.5086,
+      "step": 6277
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8319642662456037,
+      "learning_rate": 1.901918675125594e-06,
+      "loss": 0.48,
+      "step": 6278
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.252590812641215,
+      "learning_rate": 1.9004582503340696e-06,
+      "loss": 0.5294,
+      "step": 6279
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.981232683136987,
+      "learning_rate": 1.8989982548905333e-06,
+      "loss": 0.4432,
+      "step": 6280
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.869064168158839,
+      "learning_rate": 1.8975386889972218e-06,
+      "loss": 0.4783,
+      "step": 6281
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7857437200727784,
+      "learning_rate": 1.8960795528563125e-06,
+      "loss": 0.6173,
+      "step": 6282
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7624447796642495,
+      "learning_rate": 1.8946208466699267e-06,
+      "loss": 0.5138,
+      "step": 6283
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9750768174968976,
+      "learning_rate": 1.893162570640124e-06,
+      "loss": 0.4459,
+      "step": 6284
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.046137694778906,
+      "learning_rate": 1.891704724968902e-06,
+      "loss": 0.5025,
+      "step": 6285
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2206913667324106,
+      "learning_rate": 1.8902473098582048e-06,
+      "loss": 0.4478,
+      "step": 6286
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9557844491245842,
+      "learning_rate": 1.88879032550991e-06,
+      "loss": 0.4773,
+      "step": 6287
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2490218103674127,
+      "learning_rate": 1.8873337721258416e-06,
+      "loss": 0.5222,
+      "step": 6288
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9227280781160936,
+      "learning_rate": 1.8858776499077592e-06,
+      "loss": 0.5256,
+      "step": 6289
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.713613151354472,
+      "learning_rate": 1.8844219590573664e-06,
+      "loss": 0.4946,
+      "step": 6290
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7918477944864535,
+      "learning_rate": 1.8829666997763023e-06,
+      "loss": 0.4691,
+      "step": 6291
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.125522885475883,
+      "learning_rate": 1.8815118722661534e-06,
+      "loss": 0.5789,
+      "step": 6292
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8642218016353034,
+      "learning_rate": 1.8800574767284379e-06,
+      "loss": 0.5137,
+      "step": 6293
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.622094681514146,
+      "learning_rate": 1.8786035133646219e-06,
+      "loss": 0.47,
+      "step": 6294
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.656413240169926,
+      "learning_rate": 1.8771499823761047e-06,
+      "loss": 0.5391,
+      "step": 6295
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4128879127927987,
+      "learning_rate": 1.8756968839642332e-06,
+      "loss": 0.5205,
+      "step": 6296
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.853746632585124,
+      "learning_rate": 1.8742442183302879e-06,
+      "loss": 0.4701,
+      "step": 6297
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9206821482355807,
+      "learning_rate": 1.8727919856754922e-06,
+      "loss": 0.4757,
+      "step": 6298
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0283327894468535,
+      "learning_rate": 1.8713401862010071e-06,
+      "loss": 0.5549,
+      "step": 6299
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.738924086024506,
+      "learning_rate": 1.8698888201079395e-06,
+      "loss": 0.6512,
+      "step": 6300
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0175739596668265,
+      "learning_rate": 1.8684378875973286e-06,
+      "loss": 0.5435,
+      "step": 6301
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8987495873792826,
+      "learning_rate": 1.8669873888701606e-06,
+      "loss": 0.4949,
+      "step": 6302
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5561852490344634,
+      "learning_rate": 1.8655373241273572e-06,
+      "loss": 0.5716,
+      "step": 6303
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.29288659667296,
+      "learning_rate": 1.8640876935697787e-06,
+      "loss": 0.4937,
+      "step": 6304
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0551766403121445,
+      "learning_rate": 1.8626384973982314e-06,
+      "loss": 0.4862,
+      "step": 6305
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7148499191159217,
+      "learning_rate": 1.861189735813455e-06,
+      "loss": 0.444,
+      "step": 6306
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.622374401215246,
+      "learning_rate": 1.8597414090161336e-06,
+      "loss": 0.4473,
+      "step": 6307
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.664376767481218,
+      "learning_rate": 1.8582935172068873e-06,
+      "loss": 0.4907,
+      "step": 6308
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.017427987994815,
+      "learning_rate": 1.8568460605862797e-06,
+      "loss": 0.5515,
+      "step": 6309
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8821712065861482,
+      "learning_rate": 1.8553990393548105e-06,
+      "loss": 0.4799,
+      "step": 6310
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6390397294263601,
+      "learning_rate": 1.8539524537129232e-06,
+      "loss": 0.5953,
+      "step": 6311
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.422146080469555,
+      "learning_rate": 1.8525063038609954e-06,
+      "loss": 0.5155,
+      "step": 6312
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.262202152337742,
+      "learning_rate": 1.8510605899993505e-06,
+      "loss": 0.5299,
+      "step": 6313
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1216627723796337,
+      "learning_rate": 1.8496153123282461e-06,
+      "loss": 0.4619,
+      "step": 6314
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8825292551519075,
+      "learning_rate": 1.848170471047886e-06,
+      "loss": 0.4712,
+      "step": 6315
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.957892167108236,
+      "learning_rate": 1.846726066358403e-06,
+      "loss": 0.4486,
+      "step": 6316
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7580557316199963,
+      "learning_rate": 1.8452820984598813e-06,
+      "loss": 0.5351,
+      "step": 6317
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.287818176431958,
+      "learning_rate": 1.8438385675523346e-06,
+      "loss": 0.4909,
+      "step": 6318
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0894758133740403,
+      "learning_rate": 1.8423954738357248e-06,
+      "loss": 0.5343,
+      "step": 6319
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.031781029852434,
+      "learning_rate": 1.840952817509945e-06,
+      "loss": 0.5428,
+      "step": 6320
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7938253995116438,
+      "learning_rate": 1.8395105987748357e-06,
+      "loss": 0.483,
+      "step": 6321
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.783295159369498,
+      "learning_rate": 1.8380688178301693e-06,
+      "loss": 0.5477,
+      "step": 6322
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.478841883175993,
+      "learning_rate": 1.8366274748756646e-06,
+      "loss": 0.4507,
+      "step": 6323
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3859199827465707,
+      "learning_rate": 1.8351865701109734e-06,
+      "loss": 0.4569,
+      "step": 6324
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.015573209330615,
+      "learning_rate": 1.8337461037356892e-06,
+      "loss": 0.3931,
+      "step": 6325
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7337011326667826,
+      "learning_rate": 1.8323060759493477e-06,
+      "loss": 0.5,
+      "step": 6326
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7056545177008253,
+      "learning_rate": 1.8308664869514186e-06,
+      "loss": 0.4095,
+      "step": 6327
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1783797001360603,
+      "learning_rate": 1.8294273369413163e-06,
+      "loss": 0.487,
+      "step": 6328
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.984539197433747,
+      "learning_rate": 1.8279886261183883e-06,
+      "loss": 0.5071,
+      "step": 6329
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7634494874272522,
+      "learning_rate": 1.826550354681928e-06,
+      "loss": 0.5173,
+      "step": 6330
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.021276001508078,
+      "learning_rate": 1.825112522831161e-06,
+      "loss": 0.4555,
+      "step": 6331
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9082756263584135,
+      "learning_rate": 1.8236751307652617e-06,
+      "loss": 0.4669,
+      "step": 6332
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.738171703966473,
+      "learning_rate": 1.8222381786833293e-06,
+      "loss": 0.5502,
+      "step": 6333
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8385368408201282,
+      "learning_rate": 1.8208016667844153e-06,
+      "loss": 0.6684,
+      "step": 6334
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7184776174031064,
+      "learning_rate": 1.8193655952675027e-06,
+      "loss": 0.5249,
+      "step": 6335
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7550181437076597,
+      "learning_rate": 1.8179299643315184e-06,
+      "loss": 0.5674,
+      "step": 6336
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.050740032125593,
+      "learning_rate": 1.8164947741753225e-06,
+      "loss": 0.44,
+      "step": 6337
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1031751866405166,
+      "learning_rate": 1.8150600249977208e-06,
+      "loss": 0.4578,
+      "step": 6338
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8950367821048322,
+      "learning_rate": 1.8136257169974507e-06,
+      "loss": 0.458,
+      "step": 6339
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8715140553921894,
+      "learning_rate": 1.8121918503731966e-06,
+      "loss": 0.4117,
+      "step": 6340
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7623193219146276,
+      "learning_rate": 1.8107584253235733e-06,
+      "loss": 0.4534,
+      "step": 6341
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.162854321557444,
+      "learning_rate": 1.8093254420471424e-06,
+      "loss": 0.5155,
+      "step": 6342
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.131349867676294,
+      "learning_rate": 1.8078929007423985e-06,
+      "loss": 0.538,
+      "step": 6343
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3171937676212875,
+      "learning_rate": 1.8064608016077756e-06,
+      "loss": 0.4898,
+      "step": 6344
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1941811052445366,
+      "learning_rate": 1.8050291448416506e-06,
+      "loss": 0.5508,
+      "step": 6345
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4249831999730684,
+      "learning_rate": 1.803597930642334e-06,
+      "loss": 0.505,
+      "step": 6346
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.831641128393604,
+      "learning_rate": 1.8021671592080796e-06,
+      "loss": 0.4834,
+      "step": 6347
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.83024681700238,
+      "learning_rate": 1.800736830737077e-06,
+      "loss": 0.591,
+      "step": 6348
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.093736525473045,
+      "learning_rate": 1.7993069454274537e-06,
+      "loss": 0.5379,
+      "step": 6349
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.226649914206589,
+      "learning_rate": 1.7978775034772766e-06,
+      "loss": 0.5277,
+      "step": 6350
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9358468896290768,
+      "learning_rate": 1.7964485050845548e-06,
+      "loss": 0.4821,
+      "step": 6351
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.354623166387629,
+      "learning_rate": 1.79501995044723e-06,
+      "loss": 0.4964,
+      "step": 6352
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.601164648841618,
+      "learning_rate": 1.7935918397631875e-06,
+      "loss": 0.4614,
+      "step": 6353
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9745288360551739,
+      "learning_rate": 1.7921641732302463e-06,
+      "loss": 0.7706,
+      "step": 6354
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6957725828212997,
+      "learning_rate": 1.7907369510461702e-06,
+      "loss": 0.5397,
+      "step": 6355
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6100211400013387,
+      "learning_rate": 1.7893101734086543e-06,
+      "loss": 0.4746,
+      "step": 6356
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8447170499866847,
+      "learning_rate": 1.7878838405153388e-06,
+      "loss": 0.4802,
+      "step": 6357
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3255522690529595,
+      "learning_rate": 1.7864579525637948e-06,
+      "loss": 0.5432,
+      "step": 6358
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9970097266407991,
+      "learning_rate": 1.785032509751541e-06,
+      "loss": 0.4358,
+      "step": 6359
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.858462670711407,
+      "learning_rate": 1.7836075122760255e-06,
+      "loss": 0.5524,
+      "step": 6360
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.765228617380252,
+      "learning_rate": 1.7821829603346418e-06,
+      "loss": 0.4894,
+      "step": 6361
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9915575129798142,
+      "learning_rate": 1.7807588541247167e-06,
+      "loss": 0.5333,
+      "step": 6362
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.466567466442859,
+      "learning_rate": 1.7793351938435166e-06,
+      "loss": 0.504,
+      "step": 6363
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1604306099873614,
+      "learning_rate": 1.7779119796882489e-06,
+      "loss": 0.5605,
+      "step": 6364
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.972826586225186,
+      "learning_rate": 1.7764892118560555e-06,
+      "loss": 0.5054,
+      "step": 6365
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6839004446481611,
+      "learning_rate": 1.7750668905440166e-06,
+      "loss": 0.4431,
+      "step": 6366
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.914727198177321,
+      "learning_rate": 1.7736450159491552e-06,
+      "loss": 0.5,
+      "step": 6367
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.377258711989198,
+      "learning_rate": 1.7722235882684275e-06,
+      "loss": 0.4811,
+      "step": 6368
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2865486391080005,
+      "learning_rate": 1.7708026076987273e-06,
+      "loss": 0.5003,
+      "step": 6369
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9638668258879712,
+      "learning_rate": 1.7693820744368928e-06,
+      "loss": 0.5083,
+      "step": 6370
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.048879096489691,
+      "learning_rate": 1.7679619886796917e-06,
+      "loss": 0.5189,
+      "step": 6371
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8554372389125955,
+      "learning_rate": 1.7665423506238377e-06,
+      "loss": 0.6866,
+      "step": 6372
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8075358909319172,
+      "learning_rate": 1.7651231604659757e-06,
+      "loss": 0.5364,
+      "step": 6373
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.686032530221486,
+      "learning_rate": 1.7637044184026946e-06,
+      "loss": 0.502,
+      "step": 6374
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7045689232556733,
+      "learning_rate": 1.7622861246305156e-06,
+      "loss": 0.3874,
+      "step": 6375
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.522599926988805,
+      "learning_rate": 1.7608682793459037e-06,
+      "loss": 0.5155,
+      "step": 6376
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0948567119521013,
+      "learning_rate": 1.7594508827452545e-06,
+      "loss": 0.5264,
+      "step": 6377
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8183544113194867,
+      "learning_rate": 1.7580339350249099e-06,
+      "loss": 0.5063,
+      "step": 6378
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9572714333319212,
+      "learning_rate": 1.756617436381141e-06,
+      "loss": 0.4382,
+      "step": 6379
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8614628196136365,
+      "learning_rate": 1.7552013870101652e-06,
+      "loss": 0.4981,
+      "step": 6380
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7257071472198622,
+      "learning_rate": 1.7537857871081293e-06,
+      "loss": 0.4084,
+      "step": 6381
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.586278891113459,
+      "learning_rate": 1.752370636871127e-06,
+      "loss": 0.4939,
+      "step": 6382
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.3354946446510554,
+      "learning_rate": 1.7509559364951783e-06,
+      "loss": 0.5147,
+      "step": 6383
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9165456395701335,
+      "learning_rate": 1.7495416861762527e-06,
+      "loss": 0.4639,
+      "step": 6384
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8607136472667773,
+      "learning_rate": 1.7481278861102475e-06,
+      "loss": 0.5229,
+      "step": 6385
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9296542064715891,
+      "learning_rate": 1.7467145364930066e-06,
+      "loss": 0.421,
+      "step": 6386
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5224945288851752,
+      "learning_rate": 1.7453016375203024e-06,
+      "loss": 0.5117,
+      "step": 6387
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7633416445115957,
+      "learning_rate": 1.7438891893878534e-06,
+      "loss": 0.4943,
+      "step": 6388
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2711841442929446,
+      "learning_rate": 1.7424771922913098e-06,
+      "loss": 0.5057,
+      "step": 6389
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8239839286176018,
+      "learning_rate": 1.7410656464262598e-06,
+      "loss": 0.4879,
+      "step": 6390
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.55260194675326,
+      "learning_rate": 1.7396545519882336e-06,
+      "loss": 0.537,
+      "step": 6391
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0085794227469926,
+      "learning_rate": 1.7382439091726927e-06,
+      "loss": 0.4738,
+      "step": 6392
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.872407716447733,
+      "learning_rate": 1.7368337181750423e-06,
+      "loss": 0.4844,
+      "step": 6393
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3833575908885747,
+      "learning_rate": 1.735423979190618e-06,
+      "loss": 0.4874,
+      "step": 6394
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.909705183011182,
+      "learning_rate": 1.7340146924147005e-06,
+      "loss": 0.4619,
+      "step": 6395
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.345725128941785,
+      "learning_rate": 1.7326058580425003e-06,
+      "loss": 0.5255,
+      "step": 6396
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.045014748527228,
+      "learning_rate": 1.7311974762691725e-06,
+      "loss": 0.5067,
+      "step": 6397
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.954162767205877,
+      "learning_rate": 1.7297895472898024e-06,
+      "loss": 0.5367,
+      "step": 6398
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6307621442838125,
+      "learning_rate": 1.7283820712994214e-06,
+      "loss": 0.4839,
+      "step": 6399
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2618544388805404,
+      "learning_rate": 1.7269750484929853e-06,
+      "loss": 0.4794,
+      "step": 6400
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.058868031489827,
+      "learning_rate": 1.7255684790654008e-06,
+      "loss": 0.5088,
+      "step": 6401
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9609008204360812,
+      "learning_rate": 1.7241623632115017e-06,
+      "loss": 0.482,
+      "step": 6402
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7412906977922125,
+      "learning_rate": 1.722756701126066e-06,
+      "loss": 0.4823,
+      "step": 6403
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2391597397466776,
+      "learning_rate": 1.7213514930038028e-06,
+      "loss": 0.5442,
+      "step": 6404
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8119908409784025,
+      "learning_rate": 1.7199467390393649e-06,
+      "loss": 0.4962,
+      "step": 6405
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.048155774082454,
+      "learning_rate": 1.7185424394273347e-06,
+      "loss": 0.515,
+      "step": 6406
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1866790125883577,
+      "learning_rate": 1.7171385943622392e-06,
+      "loss": 0.4608,
+      "step": 6407
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9690426812406312,
+      "learning_rate": 1.7157352040385372e-06,
+      "loss": 0.5462,
+      "step": 6408
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8117968748463251,
+      "learning_rate": 1.7143322686506236e-06,
+      "loss": 0.4495,
+      "step": 6409
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.694165654878887,
+      "learning_rate": 1.712929788392837e-06,
+      "loss": 0.5052,
+      "step": 6410
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5453272644743388,
+      "learning_rate": 1.7115277634594451e-06,
+      "loss": 0.4388,
+      "step": 6411
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3359065796656604,
+      "learning_rate": 1.7101261940446601e-06,
+      "loss": 0.4996,
+      "step": 6412
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6751353193935272,
+      "learning_rate": 1.7087250803426225e-06,
+      "loss": 0.4569,
+      "step": 6413
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.302166197764301,
+      "learning_rate": 1.7073244225474184e-06,
+      "loss": 0.5813,
+      "step": 6414
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.877642489009678,
+      "learning_rate": 1.7059242208530634e-06,
+      "loss": 0.5277,
+      "step": 6415
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1077683577988853,
+      "learning_rate": 1.7045244754535185e-06,
+      "loss": 0.4862,
+      "step": 6416
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6034975711186212,
+      "learning_rate": 1.7031251865426685e-06,
+      "loss": 0.4744,
+      "step": 6417
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.778621279359413,
+      "learning_rate": 1.7017263543143486e-06,
+      "loss": 0.4952,
+      "step": 6418
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4732778910217057,
+      "learning_rate": 1.7003279789623212e-06,
+      "loss": 0.4811,
+      "step": 6419
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7266084062224527,
+      "learning_rate": 1.6989300606802921e-06,
+      "loss": 0.4854,
+      "step": 6420
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.787029287262787,
+      "learning_rate": 1.697532599661898e-06,
+      "loss": 0.4182,
+      "step": 6421
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9997075024957107,
+      "learning_rate": 1.6961355961007176e-06,
+      "loss": 0.5135,
+      "step": 6422
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.7935876039895264,
+      "learning_rate": 1.6947390501902606e-06,
+      "loss": 0.5237,
+      "step": 6423
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.067730841617121,
+      "learning_rate": 1.6933429621239795e-06,
+      "loss": 0.4889,
+      "step": 6424
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.991052640745627,
+      "learning_rate": 1.6919473320952567e-06,
+      "loss": 0.5638,
+      "step": 6425
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7047905990763956,
+      "learning_rate": 1.6905521602974183e-06,
+      "loss": 0.5754,
+      "step": 6426
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.04502082485513,
+      "learning_rate": 1.6891574469237215e-06,
+      "loss": 0.5033,
+      "step": 6427
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7168090564711342,
+      "learning_rate": 1.6877631921673598e-06,
+      "loss": 0.502,
+      "step": 6428
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3544269510995246,
+      "learning_rate": 1.6863693962214688e-06,
+      "loss": 0.4742,
+      "step": 6429
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.565972907655242,
+      "learning_rate": 1.6849760592791137e-06,
+      "loss": 0.467,
+      "step": 6430
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8749970376187846,
+      "learning_rate": 1.683583181533302e-06,
+      "loss": 0.4901,
+      "step": 6431
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8284467855369886,
+      "learning_rate": 1.6821907631769719e-06,
+      "loss": 0.4852,
+      "step": 6432
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1949053716752425,
+      "learning_rate": 1.680798804403006e-06,
+      "loss": 0.4531,
+      "step": 6433
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.156298320854908,
+      "learning_rate": 1.679407305404212e-06,
+      "loss": 0.5589,
+      "step": 6434
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.059707928309894,
+      "learning_rate": 1.678016266373344e-06,
+      "loss": 0.5108,
+      "step": 6435
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9102415546504543,
+      "learning_rate": 1.6766256875030856e-06,
+      "loss": 0.4959,
+      "step": 6436
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6636396383603198,
+      "learning_rate": 1.6752355689860634e-06,
+      "loss": 0.3945,
+      "step": 6437
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9708050477105261,
+      "learning_rate": 1.6738459110148326e-06,
+      "loss": 0.5654,
+      "step": 6438
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.247662399858331,
+      "learning_rate": 1.672456713781892e-06,
+      "loss": 0.5184,
+      "step": 6439
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.737273447721273,
+      "learning_rate": 1.671067977479669e-06,
+      "loss": 0.4076,
+      "step": 6440
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7819924861146742,
+      "learning_rate": 1.6696797023005346e-06,
+      "loss": 0.6929,
+      "step": 6441
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7843069943623606,
+      "learning_rate": 1.6682918884367899e-06,
+      "loss": 0.4926,
+      "step": 6442
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6272574211800417,
+      "learning_rate": 1.6669045360806774e-06,
+      "loss": 0.4138,
+      "step": 6443
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.284159297519438,
+      "learning_rate": 1.6655176454243694e-06,
+      "loss": 0.497,
+      "step": 6444
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.05139120874132,
+      "learning_rate": 1.664131216659981e-06,
+      "loss": 0.4619,
+      "step": 6445
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.777785856776366,
+      "learning_rate": 1.662745249979557e-06,
+      "loss": 0.4772,
+      "step": 6446
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2672654060794244,
+      "learning_rate": 1.6613597455750853e-06,
+      "loss": 0.5381,
+      "step": 6447
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8903462946285106,
+      "learning_rate": 1.6599747036384829e-06,
+      "loss": 0.5699,
+      "step": 6448
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0403232238840023,
+      "learning_rate": 1.6585901243616044e-06,
+      "loss": 0.4962,
+      "step": 6449
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.21318523121185,
+      "learning_rate": 1.6572060079362444e-06,
+      "loss": 0.4519,
+      "step": 6450
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.817783628267319,
+      "learning_rate": 1.6558223545541297e-06,
+      "loss": 0.4784,
+      "step": 6451
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6839627314004544,
+      "learning_rate": 1.6544391644069218e-06,
+      "loss": 0.5107,
+      "step": 6452
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5923458401925887,
+      "learning_rate": 1.6530564376862224e-06,
+      "loss": 0.5001,
+      "step": 6453
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.27004056592196,
+      "learning_rate": 1.6516741745835658e-06,
+      "loss": 0.5522,
+      "step": 6454
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.345695794352978,
+      "learning_rate": 1.650292375290421e-06,
+      "loss": 0.528,
+      "step": 6455
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6512121069329335,
+      "learning_rate": 1.6489110399981978e-06,
+      "loss": 0.4474,
+      "step": 6456
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2856920148305546,
+      "learning_rate": 1.6475301688982353e-06,
+      "loss": 0.4873,
+      "step": 6457
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.861781930697599,
+      "learning_rate": 1.646149762181815e-06,
+      "loss": 0.4651,
+      "step": 6458
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8717502600985453,
+      "learning_rate": 1.644769820040147e-06,
+      "loss": 0.6936,
+      "step": 6459
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.394106193048894,
+      "learning_rate": 1.6433903426643838e-06,
+      "loss": 0.5062,
+      "step": 6460
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3339287744302526,
+      "learning_rate": 1.642011330245607e-06,
+      "loss": 0.459,
+      "step": 6461
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9225696643733123,
+      "learning_rate": 1.6406327829748415e-06,
+      "loss": 0.5465,
+      "step": 6462
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1577958070996566,
+      "learning_rate": 1.6392547010430388e-06,
+      "loss": 0.398,
+      "step": 6463
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2035995038236735,
+      "learning_rate": 1.6378770846410946e-06,
+      "loss": 0.4616,
+      "step": 6464
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8851861953740503,
+      "learning_rate": 1.6364999339598332e-06,
+      "loss": 0.666,
+      "step": 6465
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0225110605363,
+      "learning_rate": 1.6351232491900193e-06,
+      "loss": 0.5437,
+      "step": 6466
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2434118631157975,
+      "learning_rate": 1.6337470305223502e-06,
+      "loss": 0.4068,
+      "step": 6467
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8505707557928814,
+      "learning_rate": 1.6323712781474598e-06,
+      "loss": 0.6022,
+      "step": 6468
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8327671581593565,
+      "learning_rate": 1.630995992255915e-06,
+      "loss": 0.4561,
+      "step": 6469
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3938361714602876,
+      "learning_rate": 1.6296211730382229e-06,
+      "loss": 0.5554,
+      "step": 6470
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.839348439254524,
+      "learning_rate": 1.6282468206848206e-06,
+      "loss": 0.6787,
+      "step": 6471
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8063425840487671,
+      "learning_rate": 1.6268729353860867e-06,
+      "loss": 0.6805,
+      "step": 6472
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6225881314619857,
+      "learning_rate": 1.6254995173323296e-06,
+      "loss": 0.4509,
+      "step": 6473
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.065901225285745,
+      "learning_rate": 1.6241265667137928e-06,
+      "loss": 0.4508,
+      "step": 6474
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8554665486141357,
+      "learning_rate": 1.6227540837206613e-06,
+      "loss": 0.5044,
+      "step": 6475
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1188290272697006,
+      "learning_rate": 1.6213820685430477e-06,
+      "loss": 0.5207,
+      "step": 6476
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8769755451955927,
+      "learning_rate": 1.620010521371006e-06,
+      "loss": 0.4482,
+      "step": 6477
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.235845942195978,
+      "learning_rate": 1.6186394423945196e-06,
+      "loss": 0.5343,
+      "step": 6478
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.038567248580995,
+      "learning_rate": 1.617268831803514e-06,
+      "loss": 0.4926,
+      "step": 6479
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.338115106820033,
+      "learning_rate": 1.615898689787842e-06,
+      "loss": 0.4469,
+      "step": 6480
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.151616747188723,
+      "learning_rate": 1.6145290165372994e-06,
+      "loss": 0.4721,
+      "step": 6481
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8494268864176064,
+      "learning_rate": 1.6131598122416091e-06,
+      "loss": 0.6913,
+      "step": 6482
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.0869226272740136,
+      "learning_rate": 1.6117910770904365e-06,
+      "loss": 0.5689,
+      "step": 6483
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.241384501135865,
+      "learning_rate": 1.6104228112733777e-06,
+      "loss": 0.4783,
+      "step": 6484
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.303813222093908,
+      "learning_rate": 1.6090550149799634e-06,
+      "loss": 0.4778,
+      "step": 6485
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.837938147458567,
+      "learning_rate": 1.6076876883996595e-06,
+      "loss": 0.4836,
+      "step": 6486
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7654473971557707,
+      "learning_rate": 1.6063208317218714e-06,
+      "loss": 0.4906,
+      "step": 6487
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6045997003348882,
+      "learning_rate": 1.6049544451359318e-06,
+      "loss": 0.4385,
+      "step": 6488
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9455510774500184,
+      "learning_rate": 1.6035885288311164e-06,
+      "loss": 0.4142,
+      "step": 6489
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.220625328969181,
+      "learning_rate": 1.6022230829966278e-06,
+      "loss": 0.4961,
+      "step": 6490
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0445280253362896,
+      "learning_rate": 1.6008581078216111e-06,
+      "loss": 0.5092,
+      "step": 6491
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9874898014858842,
+      "learning_rate": 1.5994936034951402e-06,
+      "loss": 0.4813,
+      "step": 6492
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.520446653323418,
+      "learning_rate": 1.5981295702062255e-06,
+      "loss": 0.5548,
+      "step": 6493
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0567385458596945,
+      "learning_rate": 1.5967660081438146e-06,
+      "loss": 0.4403,
+      "step": 6494
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.533717328334331,
+      "learning_rate": 1.595402917496785e-06,
+      "loss": 0.4295,
+      "step": 6495
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.350417094086101,
+      "learning_rate": 1.5940402984539554e-06,
+      "loss": 0.5359,
+      "step": 6496
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6200789870942718,
+      "learning_rate": 1.5926781512040723e-06,
+      "loss": 0.4778,
+      "step": 6497
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.375887839362443,
+      "learning_rate": 1.591316475935823e-06,
+      "loss": 0.4472,
+      "step": 6498
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.840483246365416,
+      "learning_rate": 1.5899552728378231e-06,
+      "loss": 0.4863,
+      "step": 6499
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.565592707319838,
+      "learning_rate": 1.5885945420986321e-06,
+      "loss": 0.5255,
+      "step": 6500
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3243601300857955,
+      "learning_rate": 1.5872342839067305e-06,
+      "loss": 0.4858,
+      "step": 6501
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7428290057231206,
+      "learning_rate": 1.5858744984505465e-06,
+      "loss": 0.3655,
+      "step": 6502
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.004284766770113,
+      "learning_rate": 1.5845151859184338e-06,
+      "loss": 0.5335,
+      "step": 6503
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9590778431250877,
+      "learning_rate": 1.5831563464986883e-06,
+      "loss": 0.5577,
+      "step": 6504
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9471470521338405,
+      "learning_rate": 1.5817979803795314e-06,
+      "loss": 0.4656,
+      "step": 6505
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0539476460015305,
+      "learning_rate": 1.5804400877491282e-06,
+      "loss": 0.5194,
+      "step": 6506
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9044183481952217,
+      "learning_rate": 1.57908266879557e-06,
+      "loss": 0.6776,
+      "step": 6507
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.009004259154785,
+      "learning_rate": 1.5777257237068898e-06,
+      "loss": 0.5251,
+      "step": 6508
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.61949569174443,
+      "learning_rate": 1.5763692526710484e-06,
+      "loss": 0.4259,
+      "step": 6509
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8368468941528928,
+      "learning_rate": 1.5750132558759463e-06,
+      "loss": 0.4993,
+      "step": 6510
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7350478970366157,
+      "learning_rate": 1.5736577335094128e-06,
+      "loss": 0.5492,
+      "step": 6511
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9835185449775052,
+      "learning_rate": 1.5723026857592184e-06,
+      "loss": 0.466,
+      "step": 6512
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1710787715551576,
+      "learning_rate": 1.5709481128130628e-06,
+      "loss": 0.4599,
+      "step": 6513
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.17593040766446,
+      "learning_rate": 1.5695940148585787e-06,
+      "loss": 0.4722,
+      "step": 6514
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8489899581094549,
+      "learning_rate": 1.5682403920833388e-06,
+      "loss": 0.5329,
+      "step": 6515
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8639757309254745,
+      "learning_rate": 1.566887244674844e-06,
+      "loss": 0.4902,
+      "step": 6516
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7582749214117077,
+      "learning_rate": 1.5655345728205351e-06,
+      "loss": 0.4425,
+      "step": 6517
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9459301081776175,
+      "learning_rate": 1.5641823767077824e-06,
+      "loss": 0.5635,
+      "step": 6518
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0018333615394157,
+      "learning_rate": 1.5628306565238915e-06,
+      "loss": 0.5627,
+      "step": 6519
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.270063467558181,
+      "learning_rate": 1.5614794124561017e-06,
+      "loss": 0.4462,
+      "step": 6520
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0825871551195156,
+      "learning_rate": 1.560128644691589e-06,
+      "loss": 0.5102,
+      "step": 6521
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.773889060479605,
+      "learning_rate": 1.5587783534174595e-06,
+      "loss": 0.433,
+      "step": 6522
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.184627689609748,
+      "learning_rate": 1.5574285388207576e-06,
+      "loss": 0.401,
+      "step": 6523
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9654382227985172,
+      "learning_rate": 1.5560792010884574e-06,
+      "loss": 0.4702,
+      "step": 6524
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7981755577564958,
+      "learning_rate": 1.554730340407471e-06,
+      "loss": 0.6933,
+      "step": 6525
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.254894237436918,
+      "learning_rate": 1.553381956964639e-06,
+      "loss": 0.4319,
+      "step": 6526
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1722644940021274,
+      "learning_rate": 1.5520340509467435e-06,
+      "loss": 0.465,
+      "step": 6527
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.059492478884718,
+      "learning_rate": 1.5506866225404926e-06,
+      "loss": 0.5407,
+      "step": 6528
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8130527865409911,
+      "learning_rate": 1.5493396719325343e-06,
+      "loss": 0.5493,
+      "step": 6529
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.2083034783499094,
+      "learning_rate": 1.547993199309446e-06,
+      "loss": 0.6175,
+      "step": 6530
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9651822619407406,
+      "learning_rate": 1.546647204857743e-06,
+      "loss": 0.4722,
+      "step": 6531
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.629206748747898,
+      "learning_rate": 1.545301688763871e-06,
+      "loss": 0.4409,
+      "step": 6532
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.05629937912357,
+      "learning_rate": 1.5439566512142095e-06,
+      "loss": 0.5608,
+      "step": 6533
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2203633035077437,
+      "learning_rate": 1.5426120923950755e-06,
+      "loss": 0.5351,
+      "step": 6534
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9629628700751978,
+      "learning_rate": 1.5412680124927154e-06,
+      "loss": 0.4183,
+      "step": 6535
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1655307757075035,
+      "learning_rate": 1.5399244116933098e-06,
+      "loss": 0.5619,
+      "step": 6536
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7060701668836673,
+      "learning_rate": 1.5385812901829766e-06,
+      "loss": 0.5226,
+      "step": 6537
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7364926782371983,
+      "learning_rate": 1.5372386481477641e-06,
+      "loss": 0.4513,
+      "step": 6538
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.63487142639888,
+      "learning_rate": 1.5358964857736524e-06,
+      "loss": 0.5477,
+      "step": 6539
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1386593216876606,
+      "learning_rate": 1.5345548032465613e-06,
+      "loss": 0.5046,
+      "step": 6540
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8196775447189655,
+      "learning_rate": 1.5332136007523368e-06,
+      "loss": 0.4471,
+      "step": 6541
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.523946785359498,
+      "learning_rate": 1.5318728784767656e-06,
+      "loss": 0.453,
+      "step": 6542
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9099040637123323,
+      "learning_rate": 1.5305326366055606e-06,
+      "loss": 0.4545,
+      "step": 6543
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9811563678089645,
+      "learning_rate": 1.5291928753243757e-06,
+      "loss": 0.4474,
+      "step": 6544
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8091770421846807,
+      "learning_rate": 1.5278535948187912e-06,
+      "loss": 0.5124,
+      "step": 6545
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.721423777899966,
+      "learning_rate": 1.5265147952743263e-06,
+      "loss": 0.3936,
+      "step": 6546
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9753044279944842,
+      "learning_rate": 1.5251764768764293e-06,
+      "loss": 0.4866,
+      "step": 6547
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3621269206789646,
+      "learning_rate": 1.5238386398104864e-06,
+      "loss": 0.4053,
+      "step": 6548
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.732112313838319,
+      "learning_rate": 1.5225012842618114e-06,
+      "loss": 0.3963,
+      "step": 6549
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6981184351366223,
+      "learning_rate": 1.5211644104156575e-06,
+      "loss": 0.576,
+      "step": 6550
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.407810505562715,
+      "learning_rate": 1.5198280184572072e-06,
+      "loss": 0.5632,
+      "step": 6551
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.879663918914485,
+      "learning_rate": 1.518492108571577e-06,
+      "loss": 0.5241,
+      "step": 6552
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0734233824369928,
+      "learning_rate": 1.5171566809438154e-06,
+      "loss": 0.5232,
+      "step": 6553
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.02367506864186,
+      "learning_rate": 1.5158217357589084e-06,
+      "loss": 0.5619,
+      "step": 6554
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.321301776742757,
+      "learning_rate": 1.5144872732017696e-06,
+      "loss": 0.555,
+      "step": 6555
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9950068405223962,
+      "learning_rate": 1.5131532934572517e-06,
+      "loss": 0.4192,
+      "step": 6556
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.917789848600706,
+      "learning_rate": 1.5118197967101356e-06,
+      "loss": 0.4867,
+      "step": 6557
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.553987917661604,
+      "learning_rate": 1.5104867831451353e-06,
+      "loss": 0.4546,
+      "step": 6558
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2052232438341757,
+      "learning_rate": 1.5091542529469034e-06,
+      "loss": 0.5113,
+      "step": 6559
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.976599209711734,
+      "learning_rate": 1.507822206300018e-06,
+      "loss": 0.4657,
+      "step": 6560
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.113744940036828,
+      "learning_rate": 1.506490643388997e-06,
+      "loss": 0.5725,
+      "step": 6561
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.447207902052716,
+      "learning_rate": 1.5051595643982858e-06,
+      "loss": 0.5257,
+      "step": 6562
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0297719082331636,
+      "learning_rate": 1.5038289695122676e-06,
+      "loss": 0.4612,
+      "step": 6563
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.840962244723054,
+      "learning_rate": 1.502498858915254e-06,
+      "loss": 0.4838,
+      "step": 6564
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.074981335749734,
+      "learning_rate": 1.501169232791494e-06,
+      "loss": 0.4652,
+      "step": 6565
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7213026150012902,
+      "learning_rate": 1.4998400913251637e-06,
+      "loss": 0.3766,
+      "step": 6566
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.373812447332412,
+      "learning_rate": 1.4985114347003799e-06,
+      "loss": 0.4282,
+      "step": 6567
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.146567323360134,
+      "learning_rate": 1.4971832631011857e-06,
+      "loss": 0.5062,
+      "step": 6568
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9561716934132185,
+      "learning_rate": 1.495855576711559e-06,
+      "loss": 0.5109,
+      "step": 6569
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.180869293050752,
+      "learning_rate": 1.4945283757154095e-06,
+      "loss": 0.4952,
+      "step": 6570
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8989991772210126,
+      "learning_rate": 1.4932016602965838e-06,
+      "loss": 0.5181,
+      "step": 6571
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9704282840750442,
+      "learning_rate": 1.4918754306388544e-06,
+      "loss": 0.446,
+      "step": 6572
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1743100052283038,
+      "learning_rate": 1.4905496869259351e-06,
+      "loss": 0.385,
+      "step": 6573
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.896113551308374,
+      "learning_rate": 1.4892244293414636e-06,
+      "loss": 0.5528,
+      "step": 6574
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.0867008139008565,
+      "learning_rate": 1.4878996580690175e-06,
+      "loss": 0.3734,
+      "step": 6575
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0302456674221756,
+      "learning_rate": 1.4865753732921012e-06,
+      "loss": 0.5046,
+      "step": 6576
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3610913084914906,
+      "learning_rate": 1.4852515751941565e-06,
+      "loss": 0.5336,
+      "step": 6577
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.109537909584165,
+      "learning_rate": 1.4839282639585557e-06,
+      "loss": 0.4624,
+      "step": 6578
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9397038821135526,
+      "learning_rate": 1.4826054397686008e-06,
+      "loss": 0.4036,
+      "step": 6579
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2767227633423626,
+      "learning_rate": 1.4812831028075324e-06,
+      "loss": 0.4076,
+      "step": 6580
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8132435721433877,
+      "learning_rate": 1.4799612532585178e-06,
+      "loss": 0.4231,
+      "step": 6581
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.083181386480729,
+      "learning_rate": 1.4786398913046628e-06,
+      "loss": 0.4911,
+      "step": 6582
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8136039211205544,
+      "learning_rate": 1.4773190171289981e-06,
+      "loss": 0.4411,
+      "step": 6583
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.413975949078423,
+      "learning_rate": 1.4759986309144947e-06,
+      "loss": 0.4696,
+      "step": 6584
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5380519490564257,
+      "learning_rate": 1.4746787328440503e-06,
+      "loss": 0.4677,
+      "step": 6585
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5327498755294644,
+      "learning_rate": 1.4733593231004972e-06,
+      "loss": 0.4937,
+      "step": 6586
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.834001152876658,
+      "learning_rate": 1.4720404018665985e-06,
+      "loss": 0.4485,
+      "step": 6587
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.656975754038567,
+      "learning_rate": 1.4707219693250541e-06,
+      "loss": 0.4647,
+      "step": 6588
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.985685645394257,
+      "learning_rate": 1.469404025658489e-06,
+      "loss": 0.534,
+      "step": 6589
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2108764104319367,
+      "learning_rate": 1.4680865710494691e-06,
+      "loss": 0.4688,
+      "step": 6590
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.929567640616009,
+      "learning_rate": 1.466769605680483e-06,
+      "loss": 0.447,
+      "step": 6591
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8315123547880507,
+      "learning_rate": 1.465453129733962e-06,
+      "loss": 0.6603,
+      "step": 6592
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7953323448257033,
+      "learning_rate": 1.4641371433922585e-06,
+      "loss": 0.4477,
+      "step": 6593
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.396355983041216,
+      "learning_rate": 1.4628216468376677e-06,
+      "loss": 0.5773,
+      "step": 6594
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8604037469684871,
+      "learning_rate": 1.461506640252408e-06,
+      "loss": 0.6612,
+      "step": 6595
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8318393397870936,
+      "learning_rate": 1.4601921238186374e-06,
+      "loss": 0.5352,
+      "step": 6596
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8227475262603838,
+      "learning_rate": 1.4588780977184402e-06,
+      "loss": 0.3912,
+      "step": 6597
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1349264150938163,
+      "learning_rate": 1.4575645621338346e-06,
+      "loss": 0.4929,
+      "step": 6598
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2500825300809826,
+      "learning_rate": 1.4562515172467734e-06,
+      "loss": 0.4307,
+      "step": 6599
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4081264347155016,
+      "learning_rate": 1.454938963239137e-06,
+      "loss": 0.539,
+      "step": 6600
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4067927275539147,
+      "learning_rate": 1.4536269002927427e-06,
+      "loss": 0.4989,
+      "step": 6601
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8227725375285402,
+      "learning_rate": 1.452315328589336e-06,
+      "loss": 0.481,
+      "step": 6602
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7991728316623232,
+      "learning_rate": 1.4510042483105957e-06,
+      "loss": 0.6959,
+      "step": 6603
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.145257616019355,
+      "learning_rate": 1.44969365963813e-06,
+      "loss": 0.4675,
+      "step": 6604
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8316462068486614,
+      "learning_rate": 1.4483835627534858e-06,
+      "loss": 0.5692,
+      "step": 6605
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.531275864182249,
+      "learning_rate": 1.4470739578381338e-06,
+      "loss": 0.4897,
+      "step": 6606
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0105876107711502,
+      "learning_rate": 1.445764845073483e-06,
+      "loss": 0.5321,
+      "step": 6607
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.220177639510594,
+      "learning_rate": 1.4444562246408689e-06,
+      "loss": 0.4963,
+      "step": 6608
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6561210678092055,
+      "learning_rate": 1.4431480967215645e-06,
+      "loss": 0.4517,
+      "step": 6609
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.850792347980298,
+      "learning_rate": 1.441840461496768e-06,
+      "loss": 0.3954,
+      "step": 6610
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.527090126991325,
+      "learning_rate": 1.4405333191476157e-06,
+      "loss": 0.5334,
+      "step": 6611
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8827810382458436,
+      "learning_rate": 1.43922666985517e-06,
+      "loss": 0.6992,
+      "step": 6612
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1592274873138466,
+      "learning_rate": 1.437920513800431e-06,
+      "loss": 0.5163,
+      "step": 6613
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7869913810243843,
+      "learning_rate": 1.4366148511643235e-06,
+      "loss": 0.4591,
+      "step": 6614
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1662034409806004,
+      "learning_rate": 1.4353096821277118e-06,
+      "loss": 0.4818,
+      "step": 6615
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6523858417216504,
+      "learning_rate": 1.434005006871385e-06,
+      "loss": 0.4172,
+      "step": 6616
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7986419634734306,
+      "learning_rate": 1.432700825576066e-06,
+      "loss": 0.6827,
+      "step": 6617
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.564660496633666,
+      "learning_rate": 1.4313971384224117e-06,
+      "loss": 0.4548,
+      "step": 6618
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.606948956125746,
+      "learning_rate": 1.4300939455910084e-06,
+      "loss": 0.4268,
+      "step": 6619
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.007475510505929,
+      "learning_rate": 1.428791247262371e-06,
+      "loss": 0.4881,
+      "step": 6620
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.915019256243825,
+      "learning_rate": 1.4274890436169537e-06,
+      "loss": 0.5175,
+      "step": 6621
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 13.23176047380593,
+      "learning_rate": 1.426187334835135e-06,
+      "loss": 0.4908,
+      "step": 6622
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1606317416565974,
+      "learning_rate": 1.4248861210972265e-06,
+      "loss": 0.4951,
+      "step": 6623
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7357673208513777,
+      "learning_rate": 1.4235854025834744e-06,
+      "loss": 0.5471,
+      "step": 6624
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.916909522410285,
+      "learning_rate": 1.4222851794740516e-06,
+      "loss": 0.4742,
+      "step": 6625
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.002212867463506,
+      "learning_rate": 1.4209854519490673e-06,
+      "loss": 0.522,
+      "step": 6626
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.200180532008721,
+      "learning_rate": 1.4196862201885569e-06,
+      "loss": 0.4447,
+      "step": 6627
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.285593349131978,
+      "learning_rate": 1.4183874843724927e-06,
+      "loss": 0.5197,
+      "step": 6628
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8269835896463489,
+      "learning_rate": 1.4170892446807721e-06,
+      "loss": 0.5648,
+      "step": 6629
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.131528002473121,
+      "learning_rate": 1.4157915012932306e-06,
+      "loss": 0.5684,
+      "step": 6630
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9152060192451006,
+      "learning_rate": 1.4144942543896278e-06,
+      "loss": 0.4851,
+      "step": 6631
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.325849778014444,
+      "learning_rate": 1.4131975041496615e-06,
+      "loss": 0.5064,
+      "step": 6632
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6409304012787898,
+      "learning_rate": 1.4119012507529546e-06,
+      "loss": 0.4878,
+      "step": 6633
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9891175117290074,
+      "learning_rate": 1.410605494379066e-06,
+      "loss": 0.4987,
+      "step": 6634
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8524344984751682,
+      "learning_rate": 1.4093102352074822e-06,
+      "loss": 0.5424,
+      "step": 6635
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7056084386475636,
+      "learning_rate": 1.4080154734176233e-06,
+      "loss": 0.4137,
+      "step": 6636
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.483467293651881,
+      "learning_rate": 1.406721209188837e-06,
+      "loss": 0.5231,
+      "step": 6637
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8730483855541245,
+      "learning_rate": 1.4054274427004083e-06,
+      "loss": 0.4751,
+      "step": 6638
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.884111771177903,
+      "learning_rate": 1.4041341741315456e-06,
+      "loss": 0.6868,
+      "step": 6639
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.866258976394583,
+      "learning_rate": 1.4028414036613962e-06,
+      "loss": 0.4424,
+      "step": 6640
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.224037300154087,
+      "learning_rate": 1.401549131469031e-06,
+      "loss": 0.5474,
+      "step": 6641
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.282468745165187,
+      "learning_rate": 1.4002573577334583e-06,
+      "loss": 0.5153,
+      "step": 6642
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4825000530042605,
+      "learning_rate": 1.3989660826336133e-06,
+      "loss": 0.4304,
+      "step": 6643
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7171878760574766,
+      "learning_rate": 1.3976753063483605e-06,
+      "loss": 0.5521,
+      "step": 6644
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.092815378631865,
+      "learning_rate": 1.3963850290565023e-06,
+      "loss": 0.4809,
+      "step": 6645
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.598373402938719,
+      "learning_rate": 1.3950952509367644e-06,
+      "loss": 0.4018,
+      "step": 6646
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1026657426709794,
+      "learning_rate": 1.3938059721678088e-06,
+      "loss": 0.5078,
+      "step": 6647
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9541351430320089,
+      "learning_rate": 1.3925171929282243e-06,
+      "loss": 0.4653,
+      "step": 6648
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.131037081244338,
+      "learning_rate": 1.3912289133965345e-06,
+      "loss": 0.5085,
+      "step": 6649
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.11137653844253,
+      "learning_rate": 1.3899411337511897e-06,
+      "loss": 0.3772,
+      "step": 6650
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.877443913788874,
+      "learning_rate": 1.3886538541705751e-06,
+      "loss": 0.4421,
+      "step": 6651
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8294019185900634,
+      "learning_rate": 1.3873670748330025e-06,
+      "loss": 0.5097,
+      "step": 6652
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4641861458968948,
+      "learning_rate": 1.3860807959167178e-06,
+      "loss": 0.3986,
+      "step": 6653
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6126462954653091,
+      "learning_rate": 1.3847950175998932e-06,
+      "loss": 0.4066,
+      "step": 6654
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4857846844518905,
+      "learning_rate": 1.3835097400606384e-06,
+      "loss": 0.4137,
+      "step": 6655
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.295171109763108,
+      "learning_rate": 1.3822249634769864e-06,
+      "loss": 0.3941,
+      "step": 6656
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.335849644499583,
+      "learning_rate": 1.3809406880269073e-06,
+      "loss": 0.4476,
+      "step": 6657
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9184579543243196,
+      "learning_rate": 1.379656913888296e-06,
+      "loss": 0.5073,
+      "step": 6658
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4596407911503575,
+      "learning_rate": 1.3783736412389831e-06,
+      "loss": 0.4856,
+      "step": 6659
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4712794136793605,
+      "learning_rate": 1.377090870256725e-06,
+      "loss": 0.5209,
+      "step": 6660
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.374465108561623,
+      "learning_rate": 1.3758086011192135e-06,
+      "loss": 0.427,
+      "step": 6661
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8257027227663676,
+      "learning_rate": 1.3745268340040663e-06,
+      "loss": 0.4491,
+      "step": 6662
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.991417850680566,
+      "learning_rate": 1.373245569088833e-06,
+      "loss": 0.5223,
+      "step": 6663
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1221172568523756,
+      "learning_rate": 1.3719648065509966e-06,
+      "loss": 0.3985,
+      "step": 6664
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5704922343822907,
+      "learning_rate": 1.3706845465679658e-06,
+      "loss": 0.5162,
+      "step": 6665
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7160631421619812,
+      "learning_rate": 1.3694047893170841e-06,
+      "loss": 0.4201,
+      "step": 6666
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8697161248998917,
+      "learning_rate": 1.3681255349756206e-06,
+      "loss": 0.5343,
+      "step": 6667
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.02145798006618,
+      "learning_rate": 1.3668467837207805e-06,
+      "loss": 0.584,
+      "step": 6668
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7749988908637504,
+      "learning_rate": 1.365568535729695e-06,
+      "loss": 0.5244,
+      "step": 6669
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6333339003228142,
+      "learning_rate": 1.3642907911794257e-06,
+      "loss": 0.3855,
+      "step": 6670
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1050267200370243,
+      "learning_rate": 1.3630135502469655e-06,
+      "loss": 0.5186,
+      "step": 6671
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.171583801043754,
+      "learning_rate": 1.3617368131092396e-06,
+      "loss": 0.5437,
+      "step": 6672
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.972909725672133,
+      "learning_rate": 1.3604605799430987e-06,
+      "loss": 0.4793,
+      "step": 6673
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0314852965674035,
+      "learning_rate": 1.3591848509253292e-06,
+      "loss": 0.409,
+      "step": 6674
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1976797939625277,
+      "learning_rate": 1.357909626232642e-06,
+      "loss": 0.5483,
+      "step": 6675
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7451618155841675,
+      "learning_rate": 1.3566349060416845e-06,
+      "loss": 0.4602,
+      "step": 6676
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.271396155610624,
+      "learning_rate": 1.3553606905290268e-06,
+      "loss": 0.5312,
+      "step": 6677
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7799878881217612,
+      "learning_rate": 1.3540869798711765e-06,
+      "loss": 0.5145,
+      "step": 6678
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0740888029896687,
+      "learning_rate": 1.352813774244565e-06,
+      "loss": 0.4343,
+      "step": 6679
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1783691349729897,
+      "learning_rate": 1.351541073825559e-06,
+      "loss": 0.5254,
+      "step": 6680
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.115331027551632,
+      "learning_rate": 1.3502688787904505e-06,
+      "loss": 0.4274,
+      "step": 6681
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.180658267193341,
+      "learning_rate": 1.3489971893154658e-06,
+      "loss": 0.4842,
+      "step": 6682
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.200878522626764,
+      "learning_rate": 1.3477260055767583e-06,
+      "loss": 0.5514,
+      "step": 6683
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.141751692830821,
+      "learning_rate": 1.34645532775041e-06,
+      "loss": 0.4969,
+      "step": 6684
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.542104361263541,
+      "learning_rate": 1.345185156012439e-06,
+      "loss": 0.4858,
+      "step": 6685
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.199353934404737,
+      "learning_rate": 1.3439154905387869e-06,
+      "loss": 0.5359,
+      "step": 6686
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0536693787132463,
+      "learning_rate": 1.3426463315053279e-06,
+      "loss": 0.4807,
+      "step": 6687
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8629775979087793,
+      "learning_rate": 1.341377679087864e-06,
+      "loss": 0.4624,
+      "step": 6688
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8700245662012343,
+      "learning_rate": 1.3401095334621317e-06,
+      "loss": 0.4581,
+      "step": 6689
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9597611887978124,
+      "learning_rate": 1.3388418948037912e-06,
+      "loss": 0.5094,
+      "step": 6690
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.408663139893484,
+      "learning_rate": 1.3375747632884394e-06,
+      "loss": 0.4206,
+      "step": 6691
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.233119843177849,
+      "learning_rate": 1.336308139091595e-06,
+      "loss": 0.5235,
+      "step": 6692
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7859186690040665,
+      "learning_rate": 1.3350420223887145e-06,
+      "loss": 0.421,
+      "step": 6693
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.845046863870964,
+      "learning_rate": 1.3337764133551767e-06,
+      "loss": 0.5359,
+      "step": 6694
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8607617862151697,
+      "learning_rate": 1.3325113121662964e-06,
+      "loss": 0.43,
+      "step": 6695
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9771253330996137,
+      "learning_rate": 1.3312467189973122e-06,
+      "loss": 0.5041,
+      "step": 6696
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8774328281013428,
+      "learning_rate": 1.329982634023399e-06,
+      "loss": 0.3976,
+      "step": 6697
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7853459653958383,
+      "learning_rate": 1.328719057419654e-06,
+      "loss": 0.5299,
+      "step": 6698
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7385518925394607,
+      "learning_rate": 1.327455989361111e-06,
+      "loss": 0.4364,
+      "step": 6699
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9275880876634222,
+      "learning_rate": 1.3261934300227263e-06,
+      "loss": 0.6018,
+      "step": 6700
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9209156302563102,
+      "learning_rate": 1.3249313795793934e-06,
+      "loss": 0.5259,
+      "step": 6701
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8747399048716538,
+      "learning_rate": 1.3236698382059287e-06,
+      "loss": 0.4488,
+      "step": 6702
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.046195900023158,
+      "learning_rate": 1.3224088060770817e-06,
+      "loss": 0.5061,
+      "step": 6703
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8702143119137158,
+      "learning_rate": 1.3211482833675283e-06,
+      "loss": 0.4952,
+      "step": 6704
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8926954428513714,
+      "learning_rate": 1.3198882702518789e-06,
+      "loss": 0.499,
+      "step": 6705
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3506148361316903,
+      "learning_rate": 1.318628766904667e-06,
+      "loss": 0.4748,
+      "step": 6706
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9381458265235822,
+      "learning_rate": 1.3173697735003627e-06,
+      "loss": 0.5442,
+      "step": 6707
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.45641902634138,
+      "learning_rate": 1.3161112902133594e-06,
+      "loss": 0.4845,
+      "step": 6708
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.911917760111031,
+      "learning_rate": 1.3148533172179806e-06,
+      "loss": 0.5563,
+      "step": 6709
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.898658814367033,
+      "learning_rate": 1.3135958546884836e-06,
+      "loss": 0.4194,
+      "step": 6710
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1850544090484014,
+      "learning_rate": 1.3123389027990492e-06,
+      "loss": 0.4339,
+      "step": 6711
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.319310540817452,
+      "learning_rate": 1.3110824617237922e-06,
+      "loss": 0.4327,
+      "step": 6712
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2309585501826974,
+      "learning_rate": 1.309826531636752e-06,
+      "loss": 0.567,
+      "step": 6713
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8230232281134822,
+      "learning_rate": 1.3085711127119033e-06,
+      "loss": 0.4394,
+      "step": 6714
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.105475986335523,
+      "learning_rate": 1.3073162051231431e-06,
+      "loss": 0.5651,
+      "step": 6715
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8639335527642217,
+      "learning_rate": 1.306061809044304e-06,
+      "loss": 0.6957,
+      "step": 6716
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0500317119988685,
+      "learning_rate": 1.3048079246491418e-06,
+      "loss": 0.4819,
+      "step": 6717
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.792999354237437,
+      "learning_rate": 1.3035545521113473e-06,
+      "loss": 0.5142,
+      "step": 6718
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.580358152576347,
+      "learning_rate": 1.302301691604535e-06,
+      "loss": 0.4886,
+      "step": 6719
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8364984487164793,
+      "learning_rate": 1.3010493433022543e-06,
+      "loss": 0.3732,
+      "step": 6720
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7054045130686233,
+      "learning_rate": 1.299797507377975e-06,
+      "loss": 0.4411,
+      "step": 6721
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2869154653238772,
+      "learning_rate": 1.298546184005106e-06,
+      "loss": 0.4411,
+      "step": 6722
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5432253021972209,
+      "learning_rate": 1.2972953733569764e-06,
+      "loss": 0.408,
+      "step": 6723
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7813140598872401,
+      "learning_rate": 1.2960450756068526e-06,
+      "loss": 0.461,
+      "step": 6724
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.4014320847964417,
+      "learning_rate": 1.2947952909279216e-06,
+      "loss": 0.4893,
+      "step": 6725
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8590359594943828,
+      "learning_rate": 1.2935460194933064e-06,
+      "loss": 0.5247,
+      "step": 6726
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.357576698208126,
+      "learning_rate": 1.2922972614760548e-06,
+      "loss": 0.5521,
+      "step": 6727
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2039762027947756,
+      "learning_rate": 1.2910490170491424e-06,
+      "loss": 0.4539,
+      "step": 6728
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6844148982337923,
+      "learning_rate": 1.2898012863854797e-06,
+      "loss": 0.4719,
+      "step": 6729
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9000770987316664,
+      "learning_rate": 1.2885540696578985e-06,
+      "loss": 0.4771,
+      "step": 6730
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3177356337333626,
+      "learning_rate": 1.287307367039166e-06,
+      "loss": 0.4729,
+      "step": 6731
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.475186446131532,
+      "learning_rate": 1.2860611787019733e-06,
+      "loss": 0.5234,
+      "step": 6732
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.49160385672363,
+      "learning_rate": 1.2848155048189437e-06,
+      "loss": 0.5278,
+      "step": 6733
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.18391066145214,
+      "learning_rate": 1.2835703455626253e-06,
+      "loss": 0.4879,
+      "step": 6734
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7488137517806368,
+      "learning_rate": 1.2823257011055006e-06,
+      "loss": 0.4936,
+      "step": 6735
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.976012220543827,
+      "learning_rate": 1.2810815716199748e-06,
+      "loss": 0.4757,
+      "step": 6736
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8172464920094802,
+      "learning_rate": 1.2798379572783881e-06,
+      "loss": 0.5045,
+      "step": 6737
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9450642090119168,
+      "learning_rate": 1.278594858253e-06,
+      "loss": 0.4171,
+      "step": 6738
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6726112677378306,
+      "learning_rate": 1.2773522747160094e-06,
+      "loss": 0.4592,
+      "step": 6739
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.8267315663769557,
+      "learning_rate": 1.2761102068395353e-06,
+      "loss": 0.4235,
+      "step": 6740
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.955101897526041,
+      "learning_rate": 1.2748686547956319e-06,
+      "loss": 0.5214,
+      "step": 6741
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7940065315901625,
+      "learning_rate": 1.2736276187562757e-06,
+      "loss": 0.4251,
+      "step": 6742
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.267550615914824,
+      "learning_rate": 1.2723870988933778e-06,
+      "loss": 0.5193,
+      "step": 6743
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.878695188811749,
+      "learning_rate": 1.271147095378772e-06,
+      "loss": 0.4575,
+      "step": 6744
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8649535172535148,
+      "learning_rate": 1.269907608384226e-06,
+      "loss": 0.4974,
+      "step": 6745
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9159280970011907,
+      "learning_rate": 1.2686686380814305e-06,
+      "loss": 0.5062,
+      "step": 6746
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.133537628417536,
+      "learning_rate": 1.2674301846420107e-06,
+      "loss": 0.5272,
+      "step": 6747
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5401004355469747,
+      "learning_rate": 1.266192248237515e-06,
+      "loss": 0.4476,
+      "step": 6748
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9157370839975103,
+      "learning_rate": 1.2649548290394208e-06,
+      "loss": 0.4047,
+      "step": 6749
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.247453756391527,
+      "learning_rate": 1.2637179272191386e-06,
+      "loss": 0.5311,
+      "step": 6750
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1663503576833767,
+      "learning_rate": 1.2624815429480003e-06,
+      "loss": 0.547,
+      "step": 6751
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5433187204903076,
+      "learning_rate": 1.2612456763972724e-06,
+      "loss": 0.504,
+      "step": 6752
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9063523641437436,
+      "learning_rate": 1.2600103277381448e-06,
+      "loss": 0.523,
+      "step": 6753
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9030750621005672,
+      "learning_rate": 1.2587754971417421e-06,
+      "loss": 0.5281,
+      "step": 6754
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7819758777386262,
+      "learning_rate": 1.257541184779106e-06,
+      "loss": 0.6673,
+      "step": 6755
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.668046341226919,
+      "learning_rate": 1.2563073908212182e-06,
+      "loss": 0.4839,
+      "step": 6756
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.627001906787616,
+      "learning_rate": 1.2550741154389813e-06,
+      "loss": 0.5136,
+      "step": 6757
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.802223119575306,
+      "learning_rate": 1.25384135880323e-06,
+      "loss": 0.5239,
+      "step": 6758
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2767557894425052,
+      "learning_rate": 1.252609121084724e-06,
+      "loss": 0.4705,
+      "step": 6759
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.263001748897031,
+      "learning_rate": 1.2513774024541547e-06,
+      "loss": 0.5737,
+      "step": 6760
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.990291190075724,
+      "learning_rate": 1.2501462030821365e-06,
+      "loss": 0.5995,
+      "step": 6761
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.255397059676647,
+      "learning_rate": 1.2489155231392187e-06,
+      "loss": 0.4798,
+      "step": 6762
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3781526185115225,
+      "learning_rate": 1.2476853627958713e-06,
+      "loss": 0.4888,
+      "step": 6763
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.310530150693741,
+      "learning_rate": 1.2464557222224994e-06,
+      "loss": 0.5195,
+      "step": 6764
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.014776980740316,
+      "learning_rate": 1.2452266015894288e-06,
+      "loss": 0.4814,
+      "step": 6765
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.058618529572185,
+      "learning_rate": 1.2439980010669206e-06,
+      "loss": 0.4204,
+      "step": 6766
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2861074665934713,
+      "learning_rate": 1.2427699208251587e-06,
+      "loss": 0.5205,
+      "step": 6767
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2437829666207953,
+      "learning_rate": 1.2415423610342548e-06,
+      "loss": 0.5345,
+      "step": 6768
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7977842659437432,
+      "learning_rate": 1.2403153218642538e-06,
+      "loss": 0.6579,
+      "step": 6769
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.47932444235642,
+      "learning_rate": 1.2390888034851223e-06,
+      "loss": 0.5622,
+      "step": 6770
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2955424000696927,
+      "learning_rate": 1.2378628060667591e-06,
+      "loss": 0.5141,
+      "step": 6771
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6490586482640284,
+      "learning_rate": 1.2366373297789886e-06,
+      "loss": 0.4911,
+      "step": 6772
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.196992848784864,
+      "learning_rate": 1.2354123747915631e-06,
+      "loss": 0.4335,
+      "step": 6773
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.842891679313481,
+      "learning_rate": 1.2341879412741625e-06,
+      "loss": 0.432,
+      "step": 6774
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.309189463218756,
+      "learning_rate": 1.2329640293963968e-06,
+      "loss": 0.4566,
+      "step": 6775
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.478285387005094,
+      "learning_rate": 1.2317406393278004e-06,
+      "loss": 0.5637,
+      "step": 6776
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7932254696937506,
+      "learning_rate": 1.2305177712378391e-06,
+      "loss": 0.5194,
+      "step": 6777
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.120509430267194,
+      "learning_rate": 1.2292954252959017e-06,
+      "loss": 0.5533,
+      "step": 6778
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1515692253505447,
+      "learning_rate": 1.2280736016713107e-06,
+      "loss": 0.4646,
+      "step": 6779
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.224594169473721,
+      "learning_rate": 1.2268523005333093e-06,
+      "loss": 0.5846,
+      "step": 6780
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0420015161311666,
+      "learning_rate": 1.225631522051075e-06,
+      "loss": 0.5074,
+      "step": 6781
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6798174356296443,
+      "learning_rate": 1.2244112663937073e-06,
+      "loss": 0.5912,
+      "step": 6782
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.356405234873417,
+      "learning_rate": 1.2231915337302386e-06,
+      "loss": 0.5078,
+      "step": 6783
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1991395936904867,
+      "learning_rate": 1.2219723242296233e-06,
+      "loss": 0.4549,
+      "step": 6784
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.158126118939817,
+      "learning_rate": 1.2207536380607481e-06,
+      "loss": 0.4861,
+      "step": 6785
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.083604177977479,
+      "learning_rate": 1.2195354753924248e-06,
+      "loss": 0.4929,
+      "step": 6786
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1246398835578613,
+      "learning_rate": 1.2183178363933923e-06,
+      "loss": 0.3935,
+      "step": 6787
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.45148569611535,
+      "learning_rate": 1.2171007212323171e-06,
+      "loss": 0.4775,
+      "step": 6788
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.638307679309626,
+      "learning_rate": 1.2158841300777956e-06,
+      "loss": 0.3952,
+      "step": 6789
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9360154899302933,
+      "learning_rate": 1.2146680630983482e-06,
+      "loss": 0.4473,
+      "step": 6790
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.148329940155619,
+      "learning_rate": 1.2134525204624265e-06,
+      "loss": 0.4843,
+      "step": 6791
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.638775584220251,
+      "learning_rate": 1.2122375023384052e-06,
+      "loss": 0.4817,
+      "step": 6792
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4414395286657804,
+      "learning_rate": 1.2110230088945879e-06,
+      "loss": 0.5892,
+      "step": 6793
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4227825154114564,
+      "learning_rate": 1.2098090402992085e-06,
+      "loss": 0.5097,
+      "step": 6794
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8358335621606696,
+      "learning_rate": 1.2085955967204232e-06,
+      "loss": 0.6773,
+      "step": 6795
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7419927315017967,
+      "learning_rate": 1.2073826783263204e-06,
+      "loss": 0.5391,
+      "step": 6796
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.282668667576598,
+      "learning_rate": 1.2061702852849106e-06,
+      "loss": 0.3795,
+      "step": 6797
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.2122987091039255,
+      "learning_rate": 1.2049584177641372e-06,
+      "loss": 0.5424,
+      "step": 6798
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4038558945441806,
+      "learning_rate": 1.2037470759318648e-06,
+      "loss": 0.4995,
+      "step": 6799
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2346334215149954,
+      "learning_rate": 1.2025362599558916e-06,
+      "loss": 0.4915,
+      "step": 6800
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.310228137016606,
+      "learning_rate": 1.2013259700039364e-06,
+      "loss": 0.5251,
+      "step": 6801
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8931384870779213,
+      "learning_rate": 1.2001162062436517e-06,
+      "loss": 0.3893,
+      "step": 6802
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.320937755375561,
+      "learning_rate": 1.1989069688426103e-06,
+      "loss": 0.4368,
+      "step": 6803
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8312779568191377,
+      "learning_rate": 1.19769825796832e-06,
+      "loss": 0.4747,
+      "step": 6804
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4786233530509816,
+      "learning_rate": 1.1964900737882057e-06,
+      "loss": 0.5685,
+      "step": 6805
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.761824017245682,
+      "learning_rate": 1.1952824164696287e-06,
+      "loss": 0.5017,
+      "step": 6806
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9162395943008483,
+      "learning_rate": 1.1940752861798711e-06,
+      "loss": 0.4936,
+      "step": 6807
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3902552252150273,
+      "learning_rate": 1.192868683086147e-06,
+      "loss": 0.4478,
+      "step": 6808
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.187967322113359,
+      "learning_rate": 1.1916626073555926e-06,
+      "loss": 0.5396,
+      "step": 6809
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0656061700738793,
+      "learning_rate": 1.190457059155275e-06,
+      "loss": 0.5102,
+      "step": 6810
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7194725515543636,
+      "learning_rate": 1.1892520386521843e-06,
+      "loss": 0.4628,
+      "step": 6811
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9011110535247093,
+      "learning_rate": 1.188047546013243e-06,
+      "loss": 0.4164,
+      "step": 6812
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.485173749848713,
+      "learning_rate": 1.1868435814052953e-06,
+      "loss": 0.454,
+      "step": 6813
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.962450487849109,
+      "learning_rate": 1.1856401449951128e-06,
+      "loss": 0.4923,
+      "step": 6814
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.422970149099892,
+      "learning_rate": 1.1844372369493977e-06,
+      "loss": 0.4826,
+      "step": 6815
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.875345394226263,
+      "learning_rate": 1.1832348574347747e-06,
+      "loss": 0.5334,
+      "step": 6816
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.746154329079848,
+      "learning_rate": 1.1820330066178e-06,
+      "loss": 0.5129,
+      "step": 6817
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2608347470323045,
+      "learning_rate": 1.1808316846649498e-06,
+      "loss": 0.4999,
+      "step": 6818
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1486211474879475,
+      "learning_rate": 1.1796308917426347e-06,
+      "loss": 0.497,
+      "step": 6819
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.80592640680791,
+      "learning_rate": 1.1784306280171854e-06,
+      "loss": 0.478,
+      "step": 6820
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.6989225133514743,
+      "learning_rate": 1.1772308936548664e-06,
+      "loss": 0.4468,
+      "step": 6821
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.856001956477636,
+      "learning_rate": 1.1760316888218594e-06,
+      "loss": 0.475,
+      "step": 6822
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8653365562727315,
+      "learning_rate": 1.1748330136842817e-06,
+      "loss": 0.4231,
+      "step": 6823
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9782036404673067,
+      "learning_rate": 1.1736348684081705e-06,
+      "loss": 0.3855,
+      "step": 6824
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2989993737188033,
+      "learning_rate": 1.1724372531594969e-06,
+      "loss": 0.5216,
+      "step": 6825
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.027458946354153,
+      "learning_rate": 1.1712401681041502e-06,
+      "loss": 0.444,
+      "step": 6826
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.337849639508019,
+      "learning_rate": 1.1700436134079535e-06,
+      "loss": 0.5159,
+      "step": 6827
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3860164571926723,
+      "learning_rate": 1.1688475892366507e-06,
+      "loss": 0.5347,
+      "step": 6828
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.379605159909446,
+      "learning_rate": 1.1676520957559179e-06,
+      "loss": 0.5115,
+      "step": 6829
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.270479711868883,
+      "learning_rate": 1.1664571331313513e-06,
+      "loss": 0.4693,
+      "step": 6830
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8932128764483234,
+      "learning_rate": 1.1652627015284806e-06,
+      "loss": 0.5513,
+      "step": 6831
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1853336048844705,
+      "learning_rate": 1.1640688011127553e-06,
+      "loss": 0.5341,
+      "step": 6832
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.333743872207169,
+      "learning_rate": 1.162875432049555e-06,
+      "loss": 0.5411,
+      "step": 6833
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.756120974849827,
+      "learning_rate": 1.161682594504186e-06,
+      "loss": 0.5759,
+      "step": 6834
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1276041720392986,
+      "learning_rate": 1.1604902886418778e-06,
+      "loss": 0.3545,
+      "step": 6835
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9222165337346127,
+      "learning_rate": 1.1592985146277914e-06,
+      "loss": 0.5175,
+      "step": 6836
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.147927014709979,
+      "learning_rate": 1.1581072726270082e-06,
+      "loss": 0.6192,
+      "step": 6837
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.081538766656623,
+      "learning_rate": 1.1569165628045426e-06,
+      "loss": 0.3708,
+      "step": 6838
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3135169599905336,
+      "learning_rate": 1.1557263853253264e-06,
+      "loss": 0.4965,
+      "step": 6839
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.960732314892969,
+      "learning_rate": 1.1545367403542273e-06,
+      "loss": 0.5104,
+      "step": 6840
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.690353255836699,
+      "learning_rate": 1.1533476280560308e-06,
+      "loss": 0.5303,
+      "step": 6841
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7897476631418416,
+      "learning_rate": 1.1521590485954564e-06,
+      "loss": 0.5265,
+      "step": 6842
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.027332174096245,
+      "learning_rate": 1.1509710021371428e-06,
+      "loss": 0.4271,
+      "step": 6843
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.045655772207243,
+      "learning_rate": 1.1497834888456606e-06,
+      "loss": 0.4733,
+      "step": 6844
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7947768129361377,
+      "learning_rate": 1.148596508885501e-06,
+      "loss": 0.4679,
+      "step": 6845
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1073656403914875,
+      "learning_rate": 1.1474100624210877e-06,
+      "loss": 0.3981,
+      "step": 6846
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.920288539829552,
+      "learning_rate": 1.1462241496167636e-06,
+      "loss": 0.4917,
+      "step": 6847
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.145477458701994,
+      "learning_rate": 1.1450387706368043e-06,
+      "loss": 0.4516,
+      "step": 6848
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.819790441987797,
+      "learning_rate": 1.143853925645405e-06,
+      "loss": 0.7107,
+      "step": 6849
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8391440810504196,
+      "learning_rate": 1.142669614806694e-06,
+      "loss": 0.5009,
+      "step": 6850
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7812400360517238,
+      "learning_rate": 1.1414858382847198e-06,
+      "loss": 0.4872,
+      "step": 6851
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7902226741195288,
+      "learning_rate": 1.1403025962434572e-06,
+      "loss": 0.4706,
+      "step": 6852
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0661529522006967,
+      "learning_rate": 1.1391198888468118e-06,
+      "loss": 0.4459,
+      "step": 6853
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8511199811439056,
+      "learning_rate": 1.1379377162586097e-06,
+      "loss": 0.4471,
+      "step": 6854
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.722957225090292,
+      "learning_rate": 1.1367560786426075e-06,
+      "loss": 0.4583,
+      "step": 6855
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4733571939489982,
+      "learning_rate": 1.135574976162484e-06,
+      "loss": 0.4524,
+      "step": 6856
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.911453829308716,
+      "learning_rate": 1.134394408981846e-06,
+      "loss": 0.4235,
+      "step": 6857
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3720338080075316,
+      "learning_rate": 1.133214377264223e-06,
+      "loss": 0.4963,
+      "step": 6858
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.133757253736453,
+      "learning_rate": 1.1320348811730759e-06,
+      "loss": 0.4636,
+      "step": 6859
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.663119311581652,
+      "learning_rate": 1.1308559208717862e-06,
+      "loss": 0.5076,
+      "step": 6860
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.020318347017507,
+      "learning_rate": 1.129677496523665e-06,
+      "loss": 0.5991,
+      "step": 6861
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0212951357422853,
+      "learning_rate": 1.128499608291946e-06,
+      "loss": 0.498,
+      "step": 6862
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.527734612746698,
+      "learning_rate": 1.1273222563397911e-06,
+      "loss": 0.5276,
+      "step": 6863
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.710120500290196,
+      "learning_rate": 1.1261454408302858e-06,
+      "loss": 0.4525,
+      "step": 6864
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6499088438348073,
+      "learning_rate": 1.1249691619264447e-06,
+      "loss": 0.4487,
+      "step": 6865
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4159326361833844,
+      "learning_rate": 1.1237934197912021e-06,
+      "loss": 0.4697,
+      "step": 6866
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1854973490643403,
+      "learning_rate": 1.1226182145874255e-06,
+      "loss": 0.4908,
+      "step": 6867
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1868808233124906,
+      "learning_rate": 1.1214435464779006e-06,
+      "loss": 0.4471,
+      "step": 6868
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5998775298666703,
+      "learning_rate": 1.1202694156253452e-06,
+      "loss": 0.4518,
+      "step": 6869
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.797325286171102,
+      "learning_rate": 1.119095822192397e-06,
+      "loss": 0.484,
+      "step": 6870
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5402210327121946,
+      "learning_rate": 1.1179227663416248e-06,
+      "loss": 0.4556,
+      "step": 6871
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.156653301030024,
+      "learning_rate": 1.1167502482355186e-06,
+      "loss": 0.5674,
+      "step": 6872
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9492662540895498,
+      "learning_rate": 1.1155782680364952e-06,
+      "loss": 0.5253,
+      "step": 6873
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9536336601098874,
+      "learning_rate": 1.1144068259068957e-06,
+      "loss": 0.4616,
+      "step": 6874
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.357512342482728,
+      "learning_rate": 1.1132359220089917e-06,
+      "loss": 0.4609,
+      "step": 6875
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6604133194759714,
+      "learning_rate": 1.1120655565049726e-06,
+      "loss": 0.4053,
+      "step": 6876
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.534929117448477,
+      "learning_rate": 1.1108957295569611e-06,
+      "loss": 0.3737,
+      "step": 6877
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1165745640875997,
+      "learning_rate": 1.1097264413269992e-06,
+      "loss": 0.4306,
+      "step": 6878
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7134970914880534,
+      "learning_rate": 1.1085576919770557e-06,
+      "loss": 0.4518,
+      "step": 6879
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9628344112929053,
+      "learning_rate": 1.1073894816690277e-06,
+      "loss": 0.4285,
+      "step": 6880
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8714986414464168,
+      "learning_rate": 1.1062218105647338e-06,
+      "loss": 0.4372,
+      "step": 6881
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.924438013178842,
+      "learning_rate": 1.1050546788259208e-06,
+      "loss": 0.481,
+      "step": 6882
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.980846120156529,
+      "learning_rate": 1.1038880866142582e-06,
+      "loss": 0.5062,
+      "step": 6883
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.026980720851859,
+      "learning_rate": 1.1027220340913448e-06,
+      "loss": 0.4165,
+      "step": 6884
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7940762222690434,
+      "learning_rate": 1.101556521418698e-06,
+      "loss": 0.4906,
+      "step": 6885
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1970407044966542,
+      "learning_rate": 1.1003915487577683e-06,
+      "loss": 0.4923,
+      "step": 6886
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.981984242282057,
+      "learning_rate": 1.099227116269924e-06,
+      "loss": 0.5154,
+      "step": 6887
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.8357228272358075,
+      "learning_rate": 1.0980632241164663e-06,
+      "loss": 0.495,
+      "step": 6888
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.928662829113934,
+      "learning_rate": 1.0968998724586143e-06,
+      "loss": 0.4377,
+      "step": 6889
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8823113009848835,
+      "learning_rate": 1.0957370614575158e-06,
+      "loss": 0.5572,
+      "step": 6890
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4705444252994853,
+      "learning_rate": 1.0945747912742428e-06,
+      "loss": 0.527,
+      "step": 6891
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.899860569424478,
+      "learning_rate": 1.0934130620697935e-06,
+      "loss": 0.4854,
+      "step": 6892
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6260547508544922,
+      "learning_rate": 1.0922518740050896e-06,
+      "loss": 0.464,
+      "step": 6893
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3701246864996928,
+      "learning_rate": 1.09109122724098e-06,
+      "loss": 0.5483,
+      "step": 6894
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.026589501504713,
+      "learning_rate": 1.0899311219382358e-06,
+      "loss": 0.4764,
+      "step": 6895
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8643949362639638,
+      "learning_rate": 1.0887715582575565e-06,
+      "loss": 0.4558,
+      "step": 6896
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.466005920746009,
+      "learning_rate": 1.0876125363595635e-06,
+      "loss": 0.5115,
+      "step": 6897
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4048158990542214,
+      "learning_rate": 1.086454056404803e-06,
+      "loss": 0.4889,
+      "step": 6898
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0050197628860555,
+      "learning_rate": 1.0852961185537502e-06,
+      "loss": 0.5716,
+      "step": 6899
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.157050295518263,
+      "learning_rate": 1.0841387229667994e-06,
+      "loss": 0.3877,
+      "step": 6900
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7425829884828348,
+      "learning_rate": 1.082981869804276e-06,
+      "loss": 0.4453,
+      "step": 6901
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1438184694566167,
+      "learning_rate": 1.0818255592264242e-06,
+      "loss": 0.4935,
+      "step": 6902
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6676341778685355,
+      "learning_rate": 1.0806697913934183e-06,
+      "loss": 0.4217,
+      "step": 6903
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.848714942702512,
+      "learning_rate": 1.0795145664653534e-06,
+      "loss": 0.4233,
+      "step": 6904
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8788119623056538,
+      "learning_rate": 1.0783598846022526e-06,
+      "loss": 0.5946,
+      "step": 6905
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 54.542465475210705,
+      "learning_rate": 1.0772057459640612e-06,
+      "loss": 0.4921,
+      "step": 6906
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7726760480017982,
+      "learning_rate": 1.076052150710651e-06,
+      "loss": 0.545,
+      "step": 6907
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1422865062916396,
+      "learning_rate": 1.0748990990018149e-06,
+      "loss": 0.5227,
+      "step": 6908
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0815910853850275,
+      "learning_rate": 1.0737465909972778e-06,
+      "loss": 0.5556,
+      "step": 6909
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.346965877095207,
+      "learning_rate": 1.0725946268566812e-06,
+      "loss": 0.4844,
+      "step": 6910
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1619557154686233,
+      "learning_rate": 1.0714432067395985e-06,
+      "loss": 0.4785,
+      "step": 6911
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8547340764421911,
+      "learning_rate": 1.0702923308055208e-06,
+      "loss": 0.498,
+      "step": 6912
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.8896719820628562,
+      "learning_rate": 1.0691419992138697e-06,
+      "loss": 0.5503,
+      "step": 6913
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8106425014328813,
+      "learning_rate": 1.067992212123987e-06,
+      "loss": 0.4525,
+      "step": 6914
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.426918064900197,
+      "learning_rate": 1.0668429696951432e-06,
+      "loss": 0.5103,
+      "step": 6915
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.180739630841152,
+      "learning_rate": 1.0656942720865303e-06,
+      "loss": 0.3896,
+      "step": 6916
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7871294050976116,
+      "learning_rate": 1.0645461194572642e-06,
+      "loss": 0.5765,
+      "step": 6917
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.507997146290191,
+      "learning_rate": 1.0633985119663886e-06,
+      "loss": 0.4662,
+      "step": 6918
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8084917282912344,
+      "learning_rate": 1.0622514497728686e-06,
+      "loss": 0.6919,
+      "step": 6919
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9346300743401115,
+      "learning_rate": 1.061104933035597e-06,
+      "loss": 0.5561,
+      "step": 6920
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.002055042925914,
+      "learning_rate": 1.0599589619133865e-06,
+      "loss": 0.4863,
+      "step": 6921
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8461546089783812,
+      "learning_rate": 1.0588135365649804e-06,
+      "loss": 0.7013,
+      "step": 6922
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2195006398791386,
+      "learning_rate": 1.0576686571490386e-06,
+      "loss": 0.468,
+      "step": 6923
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8848874804638946,
+      "learning_rate": 1.0565243238241525e-06,
+      "loss": 0.392,
+      "step": 6924
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9736907108445945,
+      "learning_rate": 1.0553805367488324e-06,
+      "loss": 0.6129,
+      "step": 6925
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9402108565671452,
+      "learning_rate": 1.0542372960815189e-06,
+      "loss": 0.4796,
+      "step": 6926
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5723012541646164,
+      "learning_rate": 1.0530946019805704e-06,
+      "loss": 0.4899,
+      "step": 6927
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.317020650700444,
+      "learning_rate": 1.0519524546042754e-06,
+      "loss": 0.4573,
+      "step": 6928
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.153141278992208,
+      "learning_rate": 1.0508108541108408e-06,
+      "loss": 0.4583,
+      "step": 6929
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8824673691468163,
+      "learning_rate": 1.0496698006584044e-06,
+      "loss": 0.4798,
+      "step": 6930
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3708191379497645,
+      "learning_rate": 1.0485292944050213e-06,
+      "loss": 0.4244,
+      "step": 6931
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.505625473071346,
+      "learning_rate": 1.0473893355086773e-06,
+      "loss": 0.4772,
+      "step": 6932
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.245740870319385,
+      "learning_rate": 1.0462499241272767e-06,
+      "loss": 0.5423,
+      "step": 6933
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4594798626658765,
+      "learning_rate": 1.045111060418653e-06,
+      "loss": 0.594,
+      "step": 6934
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.072492417941293,
+      "learning_rate": 1.0439727445405596e-06,
+      "loss": 0.5704,
+      "step": 6935
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1201752812378274,
+      "learning_rate": 1.0428349766506768e-06,
+      "loss": 0.4473,
+      "step": 6936
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.596679220699931,
+      "learning_rate": 1.0416977569066084e-06,
+      "loss": 0.5047,
+      "step": 6937
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.016111942319278,
+      "learning_rate": 1.0405610854658794e-06,
+      "loss": 0.5053,
+      "step": 6938
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.973109727123369,
+      "learning_rate": 1.0394249624859444e-06,
+      "loss": 0.4764,
+      "step": 6939
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0665456318696362,
+      "learning_rate": 1.0382893881241773e-06,
+      "loss": 0.5783,
+      "step": 6940
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9007858580361008,
+      "learning_rate": 1.0371543625378772e-06,
+      "loss": 0.4804,
+      "step": 6941
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3689725617246,
+      "learning_rate": 1.0360198858842695e-06,
+      "loss": 0.5323,
+      "step": 6942
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.786753858207934,
+      "learning_rate": 1.0348859583205007e-06,
+      "loss": 0.4684,
+      "step": 6943
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3795322644822012,
+      "learning_rate": 1.0337525800036409e-06,
+      "loss": 0.452,
+      "step": 6944
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.887280243484445,
+      "learning_rate": 1.0326197510906876e-06,
+      "loss": 0.4558,
+      "step": 6945
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.285407813230962,
+      "learning_rate": 1.031487471738558e-06,
+      "loss": 0.435,
+      "step": 6946
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.5980685336818787,
+      "learning_rate": 1.0303557421040983e-06,
+      "loss": 0.4886,
+      "step": 6947
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9678972838336406,
+      "learning_rate": 1.0292245623440722e-06,
+      "loss": 0.4215,
+      "step": 6948
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2110063860834526,
+      "learning_rate": 1.0280939326151729e-06,
+      "loss": 0.5736,
+      "step": 6949
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.990120625443097,
+      "learning_rate": 1.0269638530740128e-06,
+      "loss": 0.4683,
+      "step": 6950
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7992209540986273,
+      "learning_rate": 1.0258343238771334e-06,
+      "loss": 0.4856,
+      "step": 6951
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.049716779880409,
+      "learning_rate": 1.0247053451809935e-06,
+      "loss": 0.4853,
+      "step": 6952
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.930729803373553,
+      "learning_rate": 1.0235769171419818e-06,
+      "loss": 0.4806,
+      "step": 6953
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.161785309286897,
+      "learning_rate": 1.0224490399164056e-06,
+      "loss": 0.4474,
+      "step": 6954
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3264028387010347,
+      "learning_rate": 1.0213217136605008e-06,
+      "loss": 0.5402,
+      "step": 6955
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1014553382553105,
+      "learning_rate": 1.0201949385304233e-06,
+      "loss": 0.5393,
+      "step": 6956
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6406130800531504,
+      "learning_rate": 1.0190687146822536e-06,
+      "loss": 0.4486,
+      "step": 6957
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4837414601393397,
+      "learning_rate": 1.0179430422719944e-06,
+      "loss": 0.5196,
+      "step": 6958
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2713310772466926,
+      "learning_rate": 1.0168179214555767e-06,
+      "loss": 0.5546,
+      "step": 6959
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.139404299533773,
+      "learning_rate": 1.015693352388849e-06,
+      "loss": 0.5079,
+      "step": 6960
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3206529771845754,
+      "learning_rate": 1.0145693352275897e-06,
+      "loss": 0.5122,
+      "step": 6961
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.141156822566217,
+      "learning_rate": 1.0134458701274957e-06,
+      "loss": 0.431,
+      "step": 6962
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.543361889658675,
+      "learning_rate": 1.0123229572441884e-06,
+      "loss": 0.4217,
+      "step": 6963
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.810841062441158,
+      "learning_rate": 1.011200596733215e-06,
+      "loss": 0.4445,
+      "step": 6964
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6999187249403787,
+      "learning_rate": 1.010078788750043e-06,
+      "loss": 0.5488,
+      "step": 6965
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4450110206668136,
+      "learning_rate": 1.0089575334500674e-06,
+      "loss": 0.5814,
+      "step": 6966
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3807295950666747,
+      "learning_rate": 1.0078368309886017e-06,
+      "loss": 0.4198,
+      "step": 6967
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9905215404513177,
+      "learning_rate": 1.0067166815208885e-06,
+      "loss": 0.4378,
+      "step": 6968
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.978963913982134,
+      "learning_rate": 1.0055970852020869e-06,
+      "loss": 0.5111,
+      "step": 6969
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5712813181336114,
+      "learning_rate": 1.0044780421872869e-06,
+      "loss": 0.5221,
+      "step": 6970
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4163569095718094,
+      "learning_rate": 1.0033595526314948e-06,
+      "loss": 0.497,
+      "step": 6971
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.441038785615688,
+      "learning_rate": 1.0022416166896471e-06,
+      "loss": 0.5116,
+      "step": 6972
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.230760761112982,
+      "learning_rate": 1.0011242345165978e-06,
+      "loss": 0.4375,
+      "step": 6973
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.571087343361732,
+      "learning_rate": 1.0000074062671266e-06,
+      "loss": 0.4441,
+      "step": 6974
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2293860604637015,
+      "learning_rate": 9.988911320959361e-07,
+      "loss": 0.442,
+      "step": 6975
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8853905051124316,
+      "learning_rate": 9.977754121576538e-07,
+      "loss": 0.5029,
+      "step": 6976
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9576030434705964,
+      "learning_rate": 9.96660246606827e-07,
+      "loss": 0.5137,
+      "step": 6977
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8489505181429395,
+      "learning_rate": 9.95545635597931e-07,
+      "loss": 0.5345,
+      "step": 6978
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7472132666535427,
+      "learning_rate": 9.944315792853583e-07,
+      "loss": 0.4273,
+      "step": 6979
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3817454598977292,
+      "learning_rate": 9.93318077823431e-07,
+      "loss": 0.4812,
+      "step": 6980
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8356873481223022,
+      "learning_rate": 9.922051313663895e-07,
+      "loss": 0.4546,
+      "step": 6981
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.430939947794239,
+      "learning_rate": 9.910927400683973e-07,
+      "loss": 0.5812,
+      "step": 6982
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6871584678850673,
+      "learning_rate": 9.899809040835463e-07,
+      "loss": 0.4872,
+      "step": 6983
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.175770247155596,
+      "learning_rate": 9.88869623565844e-07,
+      "loss": 0.437,
+      "step": 6984
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.838049926944601,
+      "learning_rate": 9.877588986692287e-07,
+      "loss": 0.541,
+      "step": 6985
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.230130952393589,
+      "learning_rate": 9.866487295475541e-07,
+      "loss": 0.4317,
+      "step": 6986
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8681483162207415,
+      "learning_rate": 9.855391163546041e-07,
+      "loss": 0.5667,
+      "step": 6987
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9918095690909632,
+      "learning_rate": 9.844300592440786e-07,
+      "loss": 0.4849,
+      "step": 6988
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.566977259145139,
+      "learning_rate": 9.83321558369607e-07,
+      "loss": 0.4822,
+      "step": 6989
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4108322474721455,
+      "learning_rate": 9.822136138847376e-07,
+      "loss": 0.3626,
+      "step": 6990
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.24716410499373,
+      "learning_rate": 9.811062259429427e-07,
+      "loss": 0.5606,
+      "step": 6991
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.139841354911367,
+      "learning_rate": 9.799993946976156e-07,
+      "loss": 0.4896,
+      "step": 6992
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7821758112873707,
+      "learning_rate": 9.788931203020779e-07,
+      "loss": 0.4456,
+      "step": 6993
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.551450790531179,
+      "learning_rate": 9.777874029095669e-07,
+      "loss": 0.5511,
+      "step": 6994
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7574118054142498,
+      "learning_rate": 9.766822426732498e-07,
+      "loss": 0.5174,
+      "step": 6995
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.689577713743871,
+      "learning_rate": 9.755776397462097e-07,
+      "loss": 0.4808,
+      "step": 6996
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.090351730834388,
+      "learning_rate": 9.744735942814598e-07,
+      "loss": 0.4548,
+      "step": 6997
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.851570535551841,
+      "learning_rate": 9.73370106431929e-07,
+      "loss": 0.4751,
+      "step": 6998
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.685757237897816,
+      "learning_rate": 9.722671763504748e-07,
+      "loss": 0.5521,
+      "step": 6999
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3329880438269983,
+      "learning_rate": 9.711648041898725e-07,
+      "loss": 0.4475,
+      "step": 7000
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8211388058320978,
+      "learning_rate": 9.700629901028248e-07,
+      "loss": 0.6808,
+      "step": 7001
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.7025022924990716,
+      "learning_rate": 9.689617342419537e-07,
+      "loss": 0.498,
+      "step": 7002
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8618205336584218,
+      "learning_rate": 9.67861036759804e-07,
+      "loss": 0.6715,
+      "step": 7003
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.757651123589481,
+      "learning_rate": 9.66760897808846e-07,
+      "loss": 0.4526,
+      "step": 7004
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0239934404780215,
+      "learning_rate": 9.65661317541469e-07,
+      "loss": 0.5025,
+      "step": 7005
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1214484633625643,
+      "learning_rate": 9.645622961099888e-07,
+      "loss": 0.5062,
+      "step": 7006
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1421659946204272,
+      "learning_rate": 9.634638336666403e-07,
+      "loss": 0.4379,
+      "step": 7007
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8710571435972807,
+      "learning_rate": 9.623659303635819e-07,
+      "loss": 0.6844,
+      "step": 7008
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2186878990820427,
+      "learning_rate": 9.612685863528949e-07,
+      "loss": 0.5013,
+      "step": 7009
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1765719997345463,
+      "learning_rate": 9.601718017865847e-07,
+      "loss": 0.449,
+      "step": 7010
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.95680934889087,
+      "learning_rate": 9.590755768165755e-07,
+      "loss": 0.5009,
+      "step": 7011
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.94023325599711,
+      "learning_rate": 9.579799115947193e-07,
+      "loss": 0.5847,
+      "step": 7012
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.92846928151658,
+      "learning_rate": 9.568848062727836e-07,
+      "loss": 0.4829,
+      "step": 7013
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.238418984263425,
+      "learning_rate": 9.557902610024655e-07,
+      "loss": 0.4819,
+      "step": 7014
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9366319011502207,
+      "learning_rate": 9.546962759353794e-07,
+      "loss": 0.3804,
+      "step": 7015
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.388475621553339,
+      "learning_rate": 9.536028512230655e-07,
+      "loss": 0.4742,
+      "step": 7016
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1441527452603784,
+      "learning_rate": 9.525099870169824e-07,
+      "loss": 0.5296,
+      "step": 7017
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3951285106336235,
+      "learning_rate": 9.514176834685157e-07,
+      "loss": 0.4536,
+      "step": 7018
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5776643327558262,
+      "learning_rate": 9.503259407289695e-07,
+      "loss": 0.4895,
+      "step": 7019
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.62008075437621,
+      "learning_rate": 9.492347589495737e-07,
+      "loss": 0.4446,
+      "step": 7020
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.031754906439386,
+      "learning_rate": 9.481441382814776e-07,
+      "loss": 0.4304,
+      "step": 7021
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.9209253818117356,
+      "learning_rate": 9.470540788757526e-07,
+      "loss": 0.5019,
+      "step": 7022
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9689679585386908,
+      "learning_rate": 9.459645808833956e-07,
+      "loss": 0.5793,
+      "step": 7023
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9025320177411236,
+      "learning_rate": 9.448756444553226e-07,
+      "loss": 0.5115,
+      "step": 7024
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0321541978501743,
+      "learning_rate": 9.437872697423717e-07,
+      "loss": 0.5404,
+      "step": 7025
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3897363933159563,
+      "learning_rate": 9.426994568953069e-07,
+      "loss": 0.5554,
+      "step": 7026
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.076949630484575,
+      "learning_rate": 9.416122060648109e-07,
+      "loss": 0.4384,
+      "step": 7027
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8868886055212564,
+      "learning_rate": 9.405255174014876e-07,
+      "loss": 0.3976,
+      "step": 7028
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.451106531534848,
+      "learning_rate": 9.39439391055868e-07,
+      "loss": 0.5219,
+      "step": 7029
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.342789745505471,
+      "learning_rate": 9.383538271783993e-07,
+      "loss": 0.5348,
+      "step": 7030
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8863303992556801,
+      "learning_rate": 9.372688259194556e-07,
+      "loss": 0.6751,
+      "step": 7031
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.871688100187901,
+      "learning_rate": 9.361843874293302e-07,
+      "loss": 0.4663,
+      "step": 7032
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.726703894177428,
+      "learning_rate": 9.351005118582401e-07,
+      "loss": 0.438,
+      "step": 7033
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1699527537512284,
+      "learning_rate": 9.340171993563224e-07,
+      "loss": 0.5172,
+      "step": 7034
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8219710895995123,
+      "learning_rate": 9.329344500736387e-07,
+      "loss": 0.4662,
+      "step": 7035
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9764108806576621,
+      "learning_rate": 9.318522641601696e-07,
+      "loss": 0.5037,
+      "step": 7036
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1248778128600723,
+      "learning_rate": 9.307706417658213e-07,
+      "loss": 0.5511,
+      "step": 7037
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6427497621219986,
+      "learning_rate": 9.296895830404178e-07,
+      "loss": 0.6188,
+      "step": 7038
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1166235041367365,
+      "learning_rate": 9.286090881337096e-07,
+      "loss": 0.5768,
+      "step": 7039
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6982839443888422,
+      "learning_rate": 9.275291571953637e-07,
+      "loss": 0.4529,
+      "step": 7040
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.541298453763416,
+      "learning_rate": 9.264497903749764e-07,
+      "loss": 0.4675,
+      "step": 7041
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.703669809130345,
+      "learning_rate": 9.253709878220557e-07,
+      "loss": 0.4815,
+      "step": 7042
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.590531413012473,
+      "learning_rate": 9.242927496860416e-07,
+      "loss": 0.4211,
+      "step": 7043
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.773865878715323,
+      "learning_rate": 9.232150761162884e-07,
+      "loss": 0.4711,
+      "step": 7044
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.8889318578347654,
+      "learning_rate": 9.221379672620784e-07,
+      "loss": 0.472,
+      "step": 7045
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1391464765388766,
+      "learning_rate": 9.210614232726107e-07,
+      "loss": 0.5312,
+      "step": 7046
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5623251622112444,
+      "learning_rate": 9.199854442970068e-07,
+      "loss": 0.4753,
+      "step": 7047
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.373675449431209,
+      "learning_rate": 9.189100304843135e-07,
+      "loss": 0.5195,
+      "step": 7048
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.817825588650336,
+      "learning_rate": 9.178351819834952e-07,
+      "loss": 0.5063,
+      "step": 7049
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0391025133179412,
+      "learning_rate": 9.167608989434413e-07,
+      "loss": 0.5285,
+      "step": 7050
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8905690937983424,
+      "learning_rate": 9.156871815129592e-07,
+      "loss": 0.6918,
+      "step": 7051
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.331994320352603,
+      "learning_rate": 9.146140298407824e-07,
+      "loss": 0.4236,
+      "step": 7052
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2558961752035853,
+      "learning_rate": 9.135414440755619e-07,
+      "loss": 0.5603,
+      "step": 7053
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8970930755125217,
+      "learning_rate": 9.124694243658732e-07,
+      "loss": 0.4776,
+      "step": 7054
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.742581628557631,
+      "learning_rate": 9.113979708602111e-07,
+      "loss": 0.5159,
+      "step": 7055
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8577410099208607,
+      "learning_rate": 9.103270837069955e-07,
+      "loss": 0.4904,
+      "step": 7056
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.375162775843596,
+      "learning_rate": 9.092567630545624e-07,
+      "loss": 0.4795,
+      "step": 7057
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.688642271705832,
+      "learning_rate": 9.081870090511763e-07,
+      "loss": 0.4817,
+      "step": 7058
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8957659174873532,
+      "learning_rate": 9.071178218450144e-07,
+      "loss": 0.7194,
+      "step": 7059
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0578758123620093,
+      "learning_rate": 9.060492015841843e-07,
+      "loss": 0.5513,
+      "step": 7060
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5262024679770978,
+      "learning_rate": 9.049811484167082e-07,
+      "loss": 0.469,
+      "step": 7061
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3349400012848576,
+      "learning_rate": 9.039136624905359e-07,
+      "loss": 0.3991,
+      "step": 7062
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.6870298734876092,
+      "learning_rate": 9.028467439535321e-07,
+      "loss": 0.4558,
+      "step": 7063
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2816171306392543,
+      "learning_rate": 9.017803929534885e-07,
+      "loss": 0.2933,
+      "step": 7064
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.21499383593897,
+      "learning_rate": 9.007146096381142e-07,
+      "loss": 0.4905,
+      "step": 7065
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9759448784554605,
+      "learning_rate": 8.996493941550438e-07,
+      "loss": 0.4567,
+      "step": 7066
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.720270389015571,
+      "learning_rate": 8.985847466518288e-07,
+      "loss": 0.4695,
+      "step": 7067
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9371179415351363,
+      "learning_rate": 8.975206672759429e-07,
+      "loss": 0.5705,
+      "step": 7068
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9784798254939324,
+      "learning_rate": 8.964571561747847e-07,
+      "loss": 0.5348,
+      "step": 7069
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1355817136758852,
+      "learning_rate": 8.953942134956695e-07,
+      "loss": 0.4473,
+      "step": 7070
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9644382723443754,
+      "learning_rate": 8.943318393858381e-07,
+      "loss": 0.5142,
+      "step": 7071
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.249466319865713,
+      "learning_rate": 8.932700339924477e-07,
+      "loss": 0.4784,
+      "step": 7072
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.345899129478599,
+      "learning_rate": 8.922087974625826e-07,
+      "loss": 0.5234,
+      "step": 7073
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6984425009761908,
+      "learning_rate": 8.911481299432434e-07,
+      "loss": 0.4393,
+      "step": 7074
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7600225076890579,
+      "learning_rate": 8.900880315813532e-07,
+      "loss": 0.4763,
+      "step": 7075
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3842609592823973,
+      "learning_rate": 8.890285025237561e-07,
+      "loss": 0.5632,
+      "step": 7076
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6873067275351232,
+      "learning_rate": 8.879695429172197e-07,
+      "loss": 0.5483,
+      "step": 7077
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0682223147057854,
+      "learning_rate": 8.869111529084285e-07,
+      "loss": 0.6063,
+      "step": 7078
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.783422168271318,
+      "learning_rate": 8.858533326439938e-07,
+      "loss": 0.5584,
+      "step": 7079
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.356151228943602,
+      "learning_rate": 8.847960822704416e-07,
+      "loss": 0.4072,
+      "step": 7080
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8749885815061054,
+      "learning_rate": 8.837394019342244e-07,
+      "loss": 0.5363,
+      "step": 7081
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.425604513229108,
+      "learning_rate": 8.826832917817107e-07,
+      "loss": 0.5061,
+      "step": 7082
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8149863183424358,
+      "learning_rate": 8.816277519591959e-07,
+      "loss": 0.6774,
+      "step": 7083
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7749192638716431,
+      "learning_rate": 8.805727826128901e-07,
+      "loss": 0.4582,
+      "step": 7084
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7441954094730574,
+      "learning_rate": 8.795183838889304e-07,
+      "loss": 0.4721,
+      "step": 7085
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9559520341536387,
+      "learning_rate": 8.784645559333705e-07,
+      "loss": 0.5098,
+      "step": 7086
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7421057039736922,
+      "learning_rate": 8.774112988921852e-07,
+      "loss": 0.4674,
+      "step": 7087
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3574221597582334,
+      "learning_rate": 8.763586129112739e-07,
+      "loss": 0.442,
+      "step": 7088
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6381455464775856,
+      "learning_rate": 8.753064981364523e-07,
+      "loss": 0.5638,
+      "step": 7089
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7092716554515601,
+      "learning_rate": 8.74254954713461e-07,
+      "loss": 0.5396,
+      "step": 7090
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7658416380613209,
+      "learning_rate": 8.732039827879591e-07,
+      "loss": 0.4619,
+      "step": 7091
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6604205410467268,
+      "learning_rate": 8.721535825055266e-07,
+      "loss": 0.4715,
+      "step": 7092
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.325193798997003,
+      "learning_rate": 8.711037540116635e-07,
+      "loss": 0.469,
+      "step": 7093
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9721404510586076,
+      "learning_rate": 8.700544974517944e-07,
+      "loss": 0.4353,
+      "step": 7094
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.067629913134902,
+      "learning_rate": 8.690058129712603e-07,
+      "loss": 0.4741,
+      "step": 7095
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2141461427416207,
+      "learning_rate": 8.679577007153262e-07,
+      "loss": 0.5423,
+      "step": 7096
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.426286036112623,
+      "learning_rate": 8.669101608291747e-07,
+      "loss": 0.4429,
+      "step": 7097
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.186779115355316,
+      "learning_rate": 8.658631934579126e-07,
+      "loss": 0.5433,
+      "step": 7098
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8603791509333492,
+      "learning_rate": 8.648167987465645e-07,
+      "loss": 0.4491,
+      "step": 7099
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7274928311168423,
+      "learning_rate": 8.637709768400776e-07,
+      "loss": 0.4572,
+      "step": 7100
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9439760164781315,
+      "learning_rate": 8.627257278833179e-07,
+      "loss": 0.5436,
+      "step": 7101
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.738983886181228,
+      "learning_rate": 8.61681052021075e-07,
+      "loss": 0.5225,
+      "step": 7102
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1759089941098884,
+      "learning_rate": 8.606369493980543e-07,
+      "loss": 0.4136,
+      "step": 7103
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7673926337731398,
+      "learning_rate": 8.595934201588879e-07,
+      "loss": 0.4672,
+      "step": 7104
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2140059409297845,
+      "learning_rate": 8.58550464448123e-07,
+      "loss": 0.487,
+      "step": 7105
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6931790570549277,
+      "learning_rate": 8.575080824102311e-07,
+      "loss": 0.4321,
+      "step": 7106
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.079819975287847,
+      "learning_rate": 8.564662741896024e-07,
+      "loss": 0.4554,
+      "step": 7107
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5216672131939792,
+      "learning_rate": 8.554250399305475e-07,
+      "loss": 0.531,
+      "step": 7108
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9759464116717769,
+      "learning_rate": 8.543843797772972e-07,
+      "loss": 0.4573,
+      "step": 7109
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0123788135809666,
+      "learning_rate": 8.533442938740055e-07,
+      "loss": 0.4851,
+      "step": 7110
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.10217092208774,
+      "learning_rate": 8.523047823647429e-07,
+      "loss": 0.5369,
+      "step": 7111
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6482496956753625,
+      "learning_rate": 8.512658453935052e-07,
+      "loss": 0.535,
+      "step": 7112
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6951791794413231,
+      "learning_rate": 8.502274831042035e-07,
+      "loss": 0.4698,
+      "step": 7113
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1419048283401603,
+      "learning_rate": 8.491896956406709e-07,
+      "loss": 0.3742,
+      "step": 7114
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.647288674381827,
+      "learning_rate": 8.481524831466636e-07,
+      "loss": 0.4663,
+      "step": 7115
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.212816205352019,
+      "learning_rate": 8.471158457658546e-07,
+      "loss": 0.4788,
+      "step": 7116
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.226411240350637,
+      "learning_rate": 8.460797836418406e-07,
+      "loss": 0.4566,
+      "step": 7117
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0047558874321316,
+      "learning_rate": 8.450442969181339e-07,
+      "loss": 0.4682,
+      "step": 7118
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.928286400598645,
+      "learning_rate": 8.440093857381726e-07,
+      "loss": 0.4678,
+      "step": 7119
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.049124996554422,
+      "learning_rate": 8.429750502453104e-07,
+      "loss": 0.3416,
+      "step": 7120
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.167257519010175,
+      "learning_rate": 8.419412905828256e-07,
+      "loss": 0.5378,
+      "step": 7121
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0033286249349964,
+      "learning_rate": 8.409081068939112e-07,
+      "loss": 0.4851,
+      "step": 7122
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0456320652895945,
+      "learning_rate": 8.398754993216873e-07,
+      "loss": 0.474,
+      "step": 7123
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2988666583937296,
+      "learning_rate": 8.388434680091873e-07,
+      "loss": 0.4066,
+      "step": 7124
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0193507375383803,
+      "learning_rate": 8.378120130993717e-07,
+      "loss": 0.5461,
+      "step": 7125
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8785091086583843,
+      "learning_rate": 8.367811347351129e-07,
+      "loss": 0.4863,
+      "step": 7126
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7100357221126279,
+      "learning_rate": 8.357508330592113e-07,
+      "loss": 0.4289,
+      "step": 7127
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.248096663910823,
+      "learning_rate": 8.347211082143813e-07,
+      "loss": 0.5027,
+      "step": 7128
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8806878421355433,
+      "learning_rate": 8.336919603432641e-07,
+      "loss": 0.671,
+      "step": 7129
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8075767460880735,
+      "learning_rate": 8.326633895884129e-07,
+      "loss": 0.6814,
+      "step": 7130
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8092762032053218,
+      "learning_rate": 8.31635396092309e-07,
+      "loss": 0.5631,
+      "step": 7131
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7939449235489127,
+      "learning_rate": 8.306079799973477e-07,
+      "loss": 0.5455,
+      "step": 7132
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5401624836141803,
+      "learning_rate": 8.29581141445846e-07,
+      "loss": 0.5001,
+      "step": 7133
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.253319121927762,
+      "learning_rate": 8.285548805800431e-07,
+      "loss": 0.4524,
+      "step": 7134
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.734909214092982,
+      "learning_rate": 8.275291975420946e-07,
+      "loss": 0.5165,
+      "step": 7135
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.773533349665385,
+      "learning_rate": 8.265040924740798e-07,
+      "loss": 0.4372,
+      "step": 7136
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.18771972982896,
+      "learning_rate": 8.254795655179943e-07,
+      "loss": 0.4934,
+      "step": 7137
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8559613934026883,
+      "learning_rate": 8.24455616815758e-07,
+      "loss": 0.6579,
+      "step": 7138
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9152957395454688,
+      "learning_rate": 8.234322465092049e-07,
+      "loss": 0.551,
+      "step": 7139
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7113550761355243,
+      "learning_rate": 8.224094547400946e-07,
+      "loss": 0.3392,
+      "step": 7140
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9014598729376788,
+      "learning_rate": 8.213872416501018e-07,
+      "loss": 0.5268,
+      "step": 7141
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8982393778091111,
+      "learning_rate": 8.203656073808269e-07,
+      "loss": 0.4163,
+      "step": 7142
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8096820995286262,
+      "learning_rate": 8.193445520737819e-07,
+      "loss": 0.6212,
+      "step": 7143
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.834729205758222,
+      "learning_rate": 8.18324075870407e-07,
+      "loss": 0.4669,
+      "step": 7144
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9465965060091415,
+      "learning_rate": 8.17304178912055e-07,
+      "loss": 0.5516,
+      "step": 7145
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5377106618394305,
+      "learning_rate": 8.162848613400054e-07,
+      "loss": 0.5824,
+      "step": 7146
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.453241493699161,
+      "learning_rate": 8.152661232954506e-07,
+      "loss": 0.5529,
+      "step": 7147
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8181703412781312,
+      "learning_rate": 8.14247964919509e-07,
+      "loss": 0.5294,
+      "step": 7148
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7323284929595972,
+      "learning_rate": 8.132303863532126e-07,
+      "loss": 0.4541,
+      "step": 7149
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.103415859691071,
+      "learning_rate": 8.1221338773752e-07,
+      "loss": 0.5308,
+      "step": 7150
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.613934354221605,
+      "learning_rate": 8.111969692133032e-07,
+      "loss": 0.4561,
+      "step": 7151
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7704158575285973,
+      "learning_rate": 8.101811309213553e-07,
+      "loss": 0.5544,
+      "step": 7152
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8445106765520443,
+      "learning_rate": 8.091658730023927e-07,
+      "loss": 0.4831,
+      "step": 7153
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0784167410747862,
+      "learning_rate": 8.081511955970461e-07,
+      "loss": 0.4512,
+      "step": 7154
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.992705775146944,
+      "learning_rate": 8.071370988458715e-07,
+      "loss": 0.4659,
+      "step": 7155
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9903867730686569,
+      "learning_rate": 8.061235828893382e-07,
+      "loss": 0.4863,
+      "step": 7156
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8442058279685288,
+      "learning_rate": 8.051106478678411e-07,
+      "loss": 0.6752,
+      "step": 7157
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0109709247903855,
+      "learning_rate": 8.040982939216891e-07,
+      "loss": 0.5485,
+      "step": 7158
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 12.34895072254103,
+      "learning_rate": 8.030865211911176e-07,
+      "loss": 0.4745,
+      "step": 7159
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.122180840698105,
+      "learning_rate": 8.020753298162715e-07,
+      "loss": 0.445,
+      "step": 7160
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5367150801455722,
+      "learning_rate": 8.010647199372246e-07,
+      "loss": 0.3822,
+      "step": 7161
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.090481456205168,
+      "learning_rate": 8.000546916939644e-07,
+      "loss": 0.4591,
+      "step": 7162
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8667318194323672,
+      "learning_rate": 7.990452452264025e-07,
+      "loss": 0.6822,
+      "step": 7163
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.922746520559312,
+      "learning_rate": 7.980363806743641e-07,
+      "loss": 0.4893,
+      "step": 7164
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7570605638722427,
+      "learning_rate": 7.970280981775991e-07,
+      "loss": 0.533,
+      "step": 7165
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.180705708402237,
+      "learning_rate": 7.960203978757736e-07,
+      "loss": 0.4946,
+      "step": 7166
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8792795687230016,
+      "learning_rate": 7.950132799084753e-07,
+      "loss": 0.6833,
+      "step": 7167
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8132059701644931,
+      "learning_rate": 7.940067444152078e-07,
+      "loss": 0.5378,
+      "step": 7168
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.8548728903877825,
+      "learning_rate": 7.930007915353988e-07,
+      "loss": 0.4892,
+      "step": 7169
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.305215098458179,
+      "learning_rate": 7.919954214083903e-07,
+      "loss": 0.5132,
+      "step": 7170
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.941541227634566,
+      "learning_rate": 7.909906341734486e-07,
+      "loss": 0.4447,
+      "step": 7171
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.102629072062652,
+      "learning_rate": 7.899864299697546e-07,
+      "loss": 0.5014,
+      "step": 7172
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.2349095162524355,
+      "learning_rate": 7.889828089364105e-07,
+      "loss": 0.5379,
+      "step": 7173
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1837705866064567,
+      "learning_rate": 7.879797712124399e-07,
+      "loss": 0.4881,
+      "step": 7174
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9113937243569958,
+      "learning_rate": 7.8697731693678e-07,
+      "loss": 0.5547,
+      "step": 7175
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1675205577960854,
+      "learning_rate": 7.859754462482938e-07,
+      "loss": 0.476,
+      "step": 7176
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.069727601843752,
+      "learning_rate": 7.849741592857585e-07,
+      "loss": 0.4878,
+      "step": 7177
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1014293597331877,
+      "learning_rate": 7.839734561878732e-07,
+      "loss": 0.5298,
+      "step": 7178
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2550857355073965,
+      "learning_rate": 7.829733370932529e-07,
+      "loss": 0.5246,
+      "step": 7179
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2002130063956766,
+      "learning_rate": 7.819738021404371e-07,
+      "loss": 0.5188,
+      "step": 7180
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.955232269767143,
+      "learning_rate": 7.809748514678783e-07,
+      "loss": 0.4627,
+      "step": 7181
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.017760987003644,
+      "learning_rate": 7.799764852139535e-07,
+      "loss": 0.4037,
+      "step": 7182
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.151292242865054,
+      "learning_rate": 7.789787035169539e-07,
+      "loss": 0.546,
+      "step": 7183
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6367685862030645,
+      "learning_rate": 7.779815065150942e-07,
+      "loss": 0.464,
+      "step": 7184
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2718581227359955,
+      "learning_rate": 7.769848943465041e-07,
+      "loss": 0.4998,
+      "step": 7185
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.133329482062187,
+      "learning_rate": 7.75988867149236e-07,
+      "loss": 0.4041,
+      "step": 7186
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0147540243557125,
+      "learning_rate": 7.749934250612568e-07,
+      "loss": 0.5469,
+      "step": 7187
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.195133399334986,
+      "learning_rate": 7.739985682204581e-07,
+      "loss": 0.4322,
+      "step": 7188
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3687806953539097,
+      "learning_rate": 7.730042967646451e-07,
+      "loss": 0.4531,
+      "step": 7189
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.559193039701547,
+      "learning_rate": 7.720106108315451e-07,
+      "loss": 0.5108,
+      "step": 7190
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.069987156628196,
+      "learning_rate": 7.710175105588036e-07,
+      "loss": 0.4847,
+      "step": 7191
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.056257570932104,
+      "learning_rate": 7.700249960839823e-07,
+      "loss": 0.4405,
+      "step": 7192
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1160321978693313,
+      "learning_rate": 7.690330675445673e-07,
+      "loss": 0.5208,
+      "step": 7193
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7549145362097103,
+      "learning_rate": 7.680417250779593e-07,
+      "loss": 0.4808,
+      "step": 7194
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9161957836107297,
+      "learning_rate": 7.670509688214766e-07,
+      "loss": 0.4607,
+      "step": 7195
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8418102874712863,
+      "learning_rate": 7.660607989123625e-07,
+      "loss": 0.5038,
+      "step": 7196
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1732809153605315,
+      "learning_rate": 7.650712154877732e-07,
+      "loss": 0.4705,
+      "step": 7197
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1347844942114245,
+      "learning_rate": 7.640822186847841e-07,
+      "loss": 0.519,
+      "step": 7198
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9066774776618884,
+      "learning_rate": 7.630938086403933e-07,
+      "loss": 0.5416,
+      "step": 7199
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8626974855215608,
+      "learning_rate": 7.621059854915136e-07,
+      "loss": 0.542,
+      "step": 7200
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.527681097273383,
+      "learning_rate": 7.611187493749794e-07,
+      "loss": 0.5229,
+      "step": 7201
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.498535550130743,
+      "learning_rate": 7.601321004275414e-07,
+      "loss": 0.4809,
+      "step": 7202
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2208369051575825,
+      "learning_rate": 7.591460387858712e-07,
+      "loss": 0.5338,
+      "step": 7203
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4686269944689885,
+      "learning_rate": 7.581605645865558e-07,
+      "loss": 0.5613,
+      "step": 7204
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7285385220055365,
+      "learning_rate": 7.57175677966106e-07,
+      "loss": 0.5189,
+      "step": 7205
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.913520529194837,
+      "learning_rate": 7.561913790609449e-07,
+      "loss": 0.4224,
+      "step": 7206
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8510649091136632,
+      "learning_rate": 7.5520766800742e-07,
+      "loss": 0.6425,
+      "step": 7207
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8318864509980899,
+      "learning_rate": 7.542245449417929e-07,
+      "loss": 0.4831,
+      "step": 7208
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6374017847328486,
+      "learning_rate": 7.532420100002486e-07,
+      "loss": 0.5372,
+      "step": 7209
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9842241519757968,
+      "learning_rate": 7.522600633188831e-07,
+      "loss": 0.3895,
+      "step": 7210
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2831521182222896,
+      "learning_rate": 7.51278705033719e-07,
+      "loss": 0.4943,
+      "step": 7211
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.187244868998728,
+      "learning_rate": 7.502979352806916e-07,
+      "loss": 0.5258,
+      "step": 7212
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0148017872374226,
+      "learning_rate": 7.493177541956592e-07,
+      "loss": 0.403,
+      "step": 7213
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.6706479359769095,
+      "learning_rate": 7.483381619143942e-07,
+      "loss": 0.4627,
+      "step": 7214
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8363730324012142,
+      "learning_rate": 7.473591585725921e-07,
+      "loss": 0.4652,
+      "step": 7215
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9130070966900155,
+      "learning_rate": 7.463807443058624e-07,
+      "loss": 0.4329,
+      "step": 7216
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5028289626930813,
+      "learning_rate": 7.454029192497342e-07,
+      "loss": 0.4683,
+      "step": 7217
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.997286108115056,
+      "learning_rate": 7.444256835396579e-07,
+      "loss": 0.5105,
+      "step": 7218
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.143459082216822,
+      "learning_rate": 7.434490373109976e-07,
+      "loss": 0.5333,
+      "step": 7219
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.981983029765865,
+      "learning_rate": 7.424729806990411e-07,
+      "loss": 0.5497,
+      "step": 7220
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8421357048375944,
+      "learning_rate": 7.414975138389879e-07,
+      "loss": 0.4144,
+      "step": 7221
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.076303628192361,
+      "learning_rate": 7.40522636865963e-07,
+      "loss": 0.5216,
+      "step": 7222
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.2948685931315005,
+      "learning_rate": 7.395483499150036e-07,
+      "loss": 0.3816,
+      "step": 7223
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.843025641190779,
+      "learning_rate": 7.385746531210697e-07,
+      "loss": 0.5038,
+      "step": 7224
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8560779780918644,
+      "learning_rate": 7.376015466190362e-07,
+      "loss": 0.4859,
+      "step": 7225
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0367798672986273,
+      "learning_rate": 7.366290305436996e-07,
+      "loss": 0.4701,
+      "step": 7226
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6686742076763643,
+      "learning_rate": 7.356571050297695e-07,
+      "loss": 0.4786,
+      "step": 7227
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8883007968056043,
+      "learning_rate": 7.346857702118798e-07,
+      "loss": 0.4542,
+      "step": 7228
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6796377307076744,
+      "learning_rate": 7.337150262245774e-07,
+      "loss": 0.4212,
+      "step": 7229
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3615960097503157,
+      "learning_rate": 7.327448732023312e-07,
+      "loss": 0.5169,
+      "step": 7230
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2561572660804856,
+      "learning_rate": 7.317753112795256e-07,
+      "loss": 0.6223,
+      "step": 7231
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.827493591725782,
+      "learning_rate": 7.308063405904653e-07,
+      "loss": 0.4697,
+      "step": 7232
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2371019032096555,
+      "learning_rate": 7.298379612693701e-07,
+      "loss": 0.4984,
+      "step": 7233
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.1185151856443065,
+      "learning_rate": 7.288701734503823e-07,
+      "loss": 0.5263,
+      "step": 7234
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8242167454328136,
+      "learning_rate": 7.279029772675572e-07,
+      "loss": 0.6658,
+      "step": 7235
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.333117371326312,
+      "learning_rate": 7.269363728548723e-07,
+      "loss": 0.3986,
+      "step": 7236
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.330393626976881,
+      "learning_rate": 7.259703603462215e-07,
+      "loss": 0.4969,
+      "step": 7237
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2417854671646,
+      "learning_rate": 7.250049398754144e-07,
+      "loss": 0.4614,
+      "step": 7238
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8014093345280997,
+      "learning_rate": 7.240401115761841e-07,
+      "loss": 0.5091,
+      "step": 7239
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9211338336842472,
+      "learning_rate": 7.23075875582176e-07,
+      "loss": 0.4664,
+      "step": 7240
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.973058076861141,
+      "learning_rate": 7.22112232026958e-07,
+      "loss": 0.5227,
+      "step": 7241
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.128538434917711,
+      "learning_rate": 7.211491810440107e-07,
+      "loss": 0.495,
+      "step": 7242
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0774347077688655,
+      "learning_rate": 7.201867227667408e-07,
+      "loss": 0.3879,
+      "step": 7243
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.903181520863112,
+      "learning_rate": 7.192248573284621e-07,
+      "loss": 0.5501,
+      "step": 7244
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.973632568988977,
+      "learning_rate": 7.182635848624164e-07,
+      "loss": 0.5582,
+      "step": 7245
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1575937046995155,
+      "learning_rate": 7.173029055017555e-07,
+      "loss": 0.4276,
+      "step": 7246
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.4822813595951945,
+      "learning_rate": 7.163428193795557e-07,
+      "loss": 0.4861,
+      "step": 7247
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9677855053376065,
+      "learning_rate": 7.153833266288057e-07,
+      "loss": 0.4927,
+      "step": 7248
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9754368478743356,
+      "learning_rate": 7.144244273824164e-07,
+      "loss": 0.4683,
+      "step": 7249
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.036520067071005,
+      "learning_rate": 7.134661217732114e-07,
+      "loss": 0.4501,
+      "step": 7250
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.310057510790527,
+      "learning_rate": 7.12508409933938e-07,
+      "loss": 0.467,
+      "step": 7251
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1583248267490194,
+      "learning_rate": 7.115512919972562e-07,
+      "loss": 0.5565,
+      "step": 7252
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1177560350843114,
+      "learning_rate": 7.105947680957481e-07,
+      "loss": 0.3885,
+      "step": 7253
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7597421935917266,
+      "learning_rate": 7.09638838361908e-07,
+      "loss": 0.4723,
+      "step": 7254
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4650762097638905,
+      "learning_rate": 7.086835029281541e-07,
+      "loss": 0.4598,
+      "step": 7255
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.182764556918169,
+      "learning_rate": 7.077287619268186e-07,
+      "loss": 0.513,
+      "step": 7256
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1153158883086816,
+      "learning_rate": 7.0677461549015e-07,
+      "loss": 0.4066,
+      "step": 7257
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8832279179047732,
+      "learning_rate": 7.058210637503193e-07,
+      "loss": 0.7148,
+      "step": 7258
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.357671654469918,
+      "learning_rate": 7.048681068394098e-07,
+      "loss": 0.408,
+      "step": 7259
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.335062610667309,
+      "learning_rate": 7.039157448894279e-07,
+      "loss": 0.5368,
+      "step": 7260
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2434962596324306,
+      "learning_rate": 7.029639780322923e-07,
+      "loss": 0.4847,
+      "step": 7261
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0793097349689256,
+      "learning_rate": 7.020128063998421e-07,
+      "loss": 0.4318,
+      "step": 7262
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8175615690636726,
+      "learning_rate": 7.010622301238329e-07,
+      "loss": 0.4899,
+      "step": 7263
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3933248161458263,
+      "learning_rate": 7.001122493359397e-07,
+      "loss": 0.5284,
+      "step": 7264
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8818318174902324,
+      "learning_rate": 6.991628641677522e-07,
+      "loss": 0.5069,
+      "step": 7265
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8001121898212702,
+      "learning_rate": 6.98214074750781e-07,
+      "loss": 0.4734,
+      "step": 7266
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8642097293664586,
+      "learning_rate": 6.972658812164495e-07,
+      "loss": 0.5144,
+      "step": 7267
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.120147867239031,
+      "learning_rate": 6.963182836961041e-07,
+      "loss": 0.433,
+      "step": 7268
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.036972781559473,
+      "learning_rate": 6.953712823210035e-07,
+      "loss": 0.4622,
+      "step": 7269
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.291862339002079,
+      "learning_rate": 6.944248772223289e-07,
+      "loss": 0.5592,
+      "step": 7270
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8802091831343655,
+      "learning_rate": 6.934790685311732e-07,
+      "loss": 0.4569,
+      "step": 7271
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.384779987683665,
+      "learning_rate": 6.925338563785517e-07,
+      "loss": 0.4725,
+      "step": 7272
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9031285359657009,
+      "learning_rate": 6.915892408953934e-07,
+      "loss": 0.7345,
+      "step": 7273
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9233502096819546,
+      "learning_rate": 6.906452222125487e-07,
+      "loss": 0.4081,
+      "step": 7274
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7567318430030023,
+      "learning_rate": 6.897018004607814e-07,
+      "loss": 0.5376,
+      "step": 7275
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.1502653565000953,
+      "learning_rate": 6.887589757707725e-07,
+      "loss": 0.5263,
+      "step": 7276
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7787776445745294,
+      "learning_rate": 6.878167482731251e-07,
+      "loss": 0.5127,
+      "step": 7277
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7885508151509206,
+      "learning_rate": 6.868751180983546e-07,
+      "loss": 0.5484,
+      "step": 7278
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4907070317549804,
+      "learning_rate": 6.859340853768948e-07,
+      "loss": 0.481,
+      "step": 7279
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7385753883349415,
+      "learning_rate": 6.849936502390991e-07,
+      "loss": 0.5362,
+      "step": 7280
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5130809901777207,
+      "learning_rate": 6.840538128152358e-07,
+      "loss": 0.4787,
+      "step": 7281
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.203891098524167,
+      "learning_rate": 6.831145732354893e-07,
+      "loss": 0.5576,
+      "step": 7282
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1083472086145525,
+      "learning_rate": 6.821759316299659e-07,
+      "loss": 0.4538,
+      "step": 7283
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9808281837846622,
+      "learning_rate": 6.812378881286835e-07,
+      "loss": 0.4825,
+      "step": 7284
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1055819656301473,
+      "learning_rate": 6.803004428615817e-07,
+      "loss": 0.5048,
+      "step": 7285
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8562845569871673,
+      "learning_rate": 6.793635959585138e-07,
+      "loss": 0.5021,
+      "step": 7286
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0677075964091864,
+      "learning_rate": 6.784273475492525e-07,
+      "loss": 0.4679,
+      "step": 7287
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.868580076636986,
+      "learning_rate": 6.77491697763486e-07,
+      "loss": 0.5486,
+      "step": 7288
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7362013947820045,
+      "learning_rate": 6.765566467308216e-07,
+      "loss": 0.3757,
+      "step": 7289
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5948824632769025,
+      "learning_rate": 6.756221945807806e-07,
+      "loss": 0.5286,
+      "step": 7290
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5621107448840275,
+      "learning_rate": 6.746883414428057e-07,
+      "loss": 0.5916,
+      "step": 7291
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.074388585901026,
+      "learning_rate": 6.737550874462512e-07,
+      "loss": 0.4832,
+      "step": 7292
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5517872052451054,
+      "learning_rate": 6.728224327203936e-07,
+      "loss": 0.4132,
+      "step": 7293
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0665227334637226,
+      "learning_rate": 6.718903773944235e-07,
+      "loss": 0.507,
+      "step": 7294
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0197844704331724,
+      "learning_rate": 6.709589215974482e-07,
+      "loss": 0.4449,
+      "step": 7295
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.142513687937387,
+      "learning_rate": 6.700280654584923e-07,
+      "loss": 0.4468,
+      "step": 7296
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.428367670189275,
+      "learning_rate": 6.690978091065004e-07,
+      "loss": 0.5605,
+      "step": 7297
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.2580809457365927,
+      "learning_rate": 6.681681526703282e-07,
+      "loss": 0.3997,
+      "step": 7298
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9850901016927027,
+      "learning_rate": 6.672390962787545e-07,
+      "loss": 0.5436,
+      "step": 7299
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.935373496211033,
+      "learning_rate": 6.663106400604696e-07,
+      "loss": 0.4635,
+      "step": 7300
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8979026651528956,
+      "learning_rate": 6.653827841440852e-07,
+      "loss": 0.5728,
+      "step": 7301
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.085432877011784,
+      "learning_rate": 6.644555286581267e-07,
+      "loss": 0.5542,
+      "step": 7302
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3719931022451353,
+      "learning_rate": 6.635288737310369e-07,
+      "loss": 0.4327,
+      "step": 7303
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.9702345049432926,
+      "learning_rate": 6.626028194911771e-07,
+      "loss": 0.4512,
+      "step": 7304
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.892402874265305,
+      "learning_rate": 6.616773660668224e-07,
+      "loss": 0.6806,
+      "step": 7305
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7963512602223803,
+      "learning_rate": 6.607525135861686e-07,
+      "loss": 0.3657,
+      "step": 7306
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.033086489718355,
+      "learning_rate": 6.598282621773239e-07,
+      "loss": 0.504,
+      "step": 7307
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8160602966284962,
+      "learning_rate": 6.589046119683178e-07,
+      "loss": 0.3991,
+      "step": 7308
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8391129506107051,
+      "learning_rate": 6.579815630870917e-07,
+      "loss": 0.5244,
+      "step": 7309
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9279276846684756,
+      "learning_rate": 6.570591156615086e-07,
+      "loss": 0.5446,
+      "step": 7310
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9682734346201853,
+      "learning_rate": 6.561372698193446e-07,
+      "loss": 0.4732,
+      "step": 7311
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.967342666929636,
+      "learning_rate": 6.552160256882934e-07,
+      "loss": 0.4819,
+      "step": 7312
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9117496544383945,
+      "learning_rate": 6.542953833959647e-07,
+      "loss": 0.5036,
+      "step": 7313
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.984548968619642,
+      "learning_rate": 6.533753430698886e-07,
+      "loss": 0.4913,
+      "step": 7314
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9382272509102525,
+      "learning_rate": 6.524559048375051e-07,
+      "loss": 0.5107,
+      "step": 7315
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3080913144905724,
+      "learning_rate": 6.515370688261785e-07,
+      "loss": 0.4449,
+      "step": 7316
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9016757030546967,
+      "learning_rate": 6.506188351631826e-07,
+      "loss": 0.4996,
+      "step": 7317
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.144887325334697,
+      "learning_rate": 6.497012039757128e-07,
+      "loss": 0.4779,
+      "step": 7318
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1351754650895503,
+      "learning_rate": 6.487841753908785e-07,
+      "loss": 0.5263,
+      "step": 7319
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8555589757768867,
+      "learning_rate": 6.478677495357072e-07,
+      "loss": 0.5147,
+      "step": 7320
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.943888422074758,
+      "learning_rate": 6.469519265371416e-07,
+      "loss": 0.5326,
+      "step": 7321
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0927064029019484,
+      "learning_rate": 6.460367065220396e-07,
+      "loss": 0.4357,
+      "step": 7322
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8481601143699864,
+      "learning_rate": 6.451220896171806e-07,
+      "loss": 0.5587,
+      "step": 7323
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0063529725808857,
+      "learning_rate": 6.442080759492541e-07,
+      "loss": 0.561,
+      "step": 7324
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.161625122387795,
+      "learning_rate": 6.432946656448719e-07,
+      "loss": 0.4135,
+      "step": 7325
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0998797580917614,
+      "learning_rate": 6.423818588305564e-07,
+      "loss": 0.5521,
+      "step": 7326
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.812299048864969,
+      "learning_rate": 6.414696556327526e-07,
+      "loss": 0.4679,
+      "step": 7327
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1679924072304866,
+      "learning_rate": 6.405580561778168e-07,
+      "loss": 0.4392,
+      "step": 7328
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2402366981500133,
+      "learning_rate": 6.396470605920241e-07,
+      "loss": 0.591,
+      "step": 7329
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9692912998689012,
+      "learning_rate": 6.387366690015645e-07,
+      "loss": 0.5103,
+      "step": 7330
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7143980078176757,
+      "learning_rate": 6.378268815325467e-07,
+      "loss": 0.4535,
+      "step": 7331
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4849000453090535,
+      "learning_rate": 6.369176983109932e-07,
+      "loss": 0.4627,
+      "step": 7332
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.044473031082685,
+      "learning_rate": 6.360091194628448e-07,
+      "loss": 0.4489,
+      "step": 7333
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6810682903497514,
+      "learning_rate": 6.351011451139566e-07,
+      "loss": 0.4535,
+      "step": 7334
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.850868256136373,
+      "learning_rate": 6.341937753901029e-07,
+      "loss": 0.4973,
+      "step": 7335
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8522292941520664,
+      "learning_rate": 6.332870104169703e-07,
+      "loss": 0.5491,
+      "step": 7336
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.349820739891909,
+      "learning_rate": 6.323808503201656e-07,
+      "loss": 0.5185,
+      "step": 7337
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.998877229402378,
+      "learning_rate": 6.314752952252079e-07,
+      "loss": 0.4089,
+      "step": 7338
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1606313186159842,
+      "learning_rate": 6.305703452575368e-07,
+      "loss": 0.4719,
+      "step": 7339
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1783792503470907,
+      "learning_rate": 6.296660005425048e-07,
+      "loss": 0.5149,
+      "step": 7340
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0250373912882496,
+      "learning_rate": 6.287622612053801e-07,
+      "loss": 0.5143,
+      "step": 7341
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1402418014763884,
+      "learning_rate": 6.278591273713508e-07,
+      "loss": 0.4493,
+      "step": 7342
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0539604830382006,
+      "learning_rate": 6.269565991655174e-07,
+      "loss": 0.395,
+      "step": 7343
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.823952932675824,
+      "learning_rate": 6.260546767128989e-07,
+      "loss": 0.4575,
+      "step": 7344
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8877956434440728,
+      "learning_rate": 6.251533601384296e-07,
+      "loss": 0.4795,
+      "step": 7345
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8681202966325348,
+      "learning_rate": 6.242526495669587e-07,
+      "loss": 0.4477,
+      "step": 7346
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.350969096037664,
+      "learning_rate": 6.233525451232519e-07,
+      "loss": 0.532,
+      "step": 7347
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5970827149402904,
+      "learning_rate": 6.224530469319934e-07,
+      "loss": 0.5717,
+      "step": 7348
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.730334697204506,
+      "learning_rate": 6.215541551177795e-07,
+      "loss": 0.4345,
+      "step": 7349
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.760807513290828,
+      "learning_rate": 6.206558698051274e-07,
+      "loss": 0.5329,
+      "step": 7350
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.899477012690792,
+      "learning_rate": 6.197581911184642e-07,
+      "loss": 0.4678,
+      "step": 7351
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1284397859555195,
+      "learning_rate": 6.188611191821387e-07,
+      "loss": 0.481,
+      "step": 7352
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7093903886383504,
+      "learning_rate": 6.17964654120411e-07,
+      "loss": 0.441,
+      "step": 7353
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.36866324503014,
+      "learning_rate": 6.170687960574612e-07,
+      "loss": 0.493,
+      "step": 7354
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.344742754319983,
+      "learning_rate": 6.161735451173822e-07,
+      "loss": 0.4349,
+      "step": 7355
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0564823050814267,
+      "learning_rate": 6.15278901424185e-07,
+      "loss": 0.5004,
+      "step": 7356
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.7087505489007735,
+      "learning_rate": 6.143848651017941e-07,
+      "loss": 0.4614,
+      "step": 7357
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8809648301136048,
+      "learning_rate": 6.134914362740535e-07,
+      "loss": 0.4934,
+      "step": 7358
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1639950624237,
+      "learning_rate": 6.125986150647179e-07,
+      "loss": 0.4779,
+      "step": 7359
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.826944463172608,
+      "learning_rate": 6.117064015974633e-07,
+      "loss": 0.4459,
+      "step": 7360
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.04297174490235,
+      "learning_rate": 6.108147959958776e-07,
+      "loss": 0.5707,
+      "step": 7361
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9468530383110887,
+      "learning_rate": 6.099237983834666e-07,
+      "loss": 0.4852,
+      "step": 7362
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4549057830870034,
+      "learning_rate": 6.090334088836492e-07,
+      "loss": 0.5002,
+      "step": 7363
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.255534548097752,
+      "learning_rate": 6.08143627619765e-07,
+      "loss": 0.4671,
+      "step": 7364
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.866377773184388,
+      "learning_rate": 6.072544547150633e-07,
+      "loss": 0.4448,
+      "step": 7365
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6404253234514976,
+      "learning_rate": 6.06365890292715e-07,
+      "loss": 0.4381,
+      "step": 7366
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1536274923910272,
+      "learning_rate": 6.054779344758027e-07,
+      "loss": 0.5694,
+      "step": 7367
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.123731391328768,
+      "learning_rate": 6.045905873873242e-07,
+      "loss": 0.4282,
+      "step": 7368
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.71583646228802,
+      "learning_rate": 6.037038491501978e-07,
+      "loss": 0.522,
+      "step": 7369
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.6626470094922468,
+      "learning_rate": 6.028177198872514e-07,
+      "loss": 0.4022,
+      "step": 7370
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3966502341835882,
+      "learning_rate": 6.019321997212341e-07,
+      "loss": 0.5581,
+      "step": 7371
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1022482974203394,
+      "learning_rate": 6.010472887748053e-07,
+      "loss": 0.5295,
+      "step": 7372
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.133049330434335,
+      "learning_rate": 6.001629871705455e-07,
+      "loss": 0.5329,
+      "step": 7373
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1318766430647753,
+      "learning_rate": 5.992792950309456e-07,
+      "loss": 0.4958,
+      "step": 7374
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.2795485747545445,
+      "learning_rate": 5.983962124784165e-07,
+      "loss": 0.4428,
+      "step": 7375
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.840424922034586,
+      "learning_rate": 5.975137396352809e-07,
+      "loss": 0.5365,
+      "step": 7376
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4705495884182205,
+      "learning_rate": 5.9663187662378e-07,
+      "loss": 0.4392,
+      "step": 7377
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7647193124232938,
+      "learning_rate": 5.957506235660693e-07,
+      "loss": 0.4282,
+      "step": 7378
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3122739260326632,
+      "learning_rate": 5.948699805842195e-07,
+      "loss": 0.5017,
+      "step": 7379
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0337310633161976,
+      "learning_rate": 5.939899478002153e-07,
+      "loss": 0.4441,
+      "step": 7380
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 13.309177810807729,
+      "learning_rate": 5.93110525335962e-07,
+      "loss": 0.4713,
+      "step": 7381
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.275022713390884,
+      "learning_rate": 5.922317133132744e-07,
+      "loss": 0.4914,
+      "step": 7382
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.258148905905298,
+      "learning_rate": 5.913535118538871e-07,
+      "loss": 0.5342,
+      "step": 7383
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8281947079945777,
+      "learning_rate": 5.904759210794469e-07,
+      "loss": 0.3753,
+      "step": 7384
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7727812750868706,
+      "learning_rate": 5.895989411115194e-07,
+      "loss": 0.5151,
+      "step": 7385
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.750507187125224,
+      "learning_rate": 5.887225720715828e-07,
+      "loss": 0.4906,
+      "step": 7386
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3939165498474915,
+      "learning_rate": 5.8784681408103e-07,
+      "loss": 0.5329,
+      "step": 7387
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.847794750890957,
+      "learning_rate": 5.869716672611741e-07,
+      "loss": 0.5703,
+      "step": 7388
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1626650434342456,
+      "learning_rate": 5.86097131733237e-07,
+      "loss": 0.4765,
+      "step": 7389
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.950007852604291,
+      "learning_rate": 5.852232076183617e-07,
+      "loss": 0.4654,
+      "step": 7390
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0174853234801198,
+      "learning_rate": 5.84349895037602e-07,
+      "loss": 0.409,
+      "step": 7391
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8636683998457557,
+      "learning_rate": 5.83477194111931e-07,
+      "loss": 0.5184,
+      "step": 7392
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9209244208404415,
+      "learning_rate": 5.826051049622334e-07,
+      "loss": 0.4648,
+      "step": 7393
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.961140893703432,
+      "learning_rate": 5.817336277093121e-07,
+      "loss": 0.5048,
+      "step": 7394
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8272880542945591,
+      "learning_rate": 5.808627624738838e-07,
+      "loss": 0.6636,
+      "step": 7395
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.9045614838281164,
+      "learning_rate": 5.799925093765801e-07,
+      "loss": 0.5106,
+      "step": 7396
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9259639263360695,
+      "learning_rate": 5.791228685379474e-07,
+      "loss": 0.4476,
+      "step": 7397
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9933237910892496,
+      "learning_rate": 5.782538400784498e-07,
+      "loss": 0.3991,
+      "step": 7398
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9932924191212764,
+      "learning_rate": 5.773854241184639e-07,
+      "loss": 0.4832,
+      "step": 7399
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9699709420561649,
+      "learning_rate": 5.765176207782841e-07,
+      "loss": 0.485,
+      "step": 7400
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.26833355998099,
+      "learning_rate": 5.756504301781163e-07,
+      "loss": 0.5202,
+      "step": 7401
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.278133729047044,
+      "learning_rate": 5.747838524380855e-07,
+      "loss": 0.4762,
+      "step": 7402
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.426927383761171,
+      "learning_rate": 5.739178876782275e-07,
+      "loss": 0.4968,
+      "step": 7403
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.414839559479861,
+      "learning_rate": 5.730525360184985e-07,
+      "loss": 0.5137,
+      "step": 7404
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.072038642652952,
+      "learning_rate": 5.72187797578766e-07,
+      "loss": 0.4782,
+      "step": 7405
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8059920066218003,
+      "learning_rate": 5.713236724788113e-07,
+      "loss": 0.4751,
+      "step": 7406
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.428150506744794,
+      "learning_rate": 5.704601608383353e-07,
+      "loss": 0.4594,
+      "step": 7407
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.640569917083776,
+      "learning_rate": 5.695972627769497e-07,
+      "loss": 0.5418,
+      "step": 7408
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1321981168922832,
+      "learning_rate": 5.687349784141849e-07,
+      "loss": 0.4839,
+      "step": 7409
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0342254512892333,
+      "learning_rate": 5.678733078694825e-07,
+      "loss": 0.4413,
+      "step": 7410
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4091287003184116,
+      "learning_rate": 5.670122512622022e-07,
+      "loss": 0.4554,
+      "step": 7411
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7794283360263499,
+      "learning_rate": 5.66151808711618e-07,
+      "loss": 0.4437,
+      "step": 7412
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.307850046768339,
+      "learning_rate": 5.652919803369167e-07,
+      "loss": 0.4991,
+      "step": 7413
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3830578502811623,
+      "learning_rate": 5.644327662572014e-07,
+      "loss": 0.4406,
+      "step": 7414
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9691020748667285,
+      "learning_rate": 5.635741665914918e-07,
+      "loss": 0.515,
+      "step": 7415
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6626636922513374,
+      "learning_rate": 5.627161814587195e-07,
+      "loss": 0.4041,
+      "step": 7416
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.974654177270855,
+      "learning_rate": 5.61858810977734e-07,
+      "loss": 0.4101,
+      "step": 7417
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.001690882768793,
+      "learning_rate": 5.610020552672968e-07,
+      "loss": 0.4966,
+      "step": 7418
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7832428661847315,
+      "learning_rate": 5.601459144460864e-07,
+      "loss": 0.4855,
+      "step": 7419
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3001500946569484,
+      "learning_rate": 5.59290388632694e-07,
+      "loss": 0.4874,
+      "step": 7420
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8782861835794988,
+      "learning_rate": 5.584354779456291e-07,
+      "loss": 0.4775,
+      "step": 7421
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9786489134352634,
+      "learning_rate": 5.575811825033112e-07,
+      "loss": 0.5153,
+      "step": 7422
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.910591231207547,
+      "learning_rate": 5.567275024240798e-07,
+      "loss": 0.4797,
+      "step": 7423
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.221850131757607,
+      "learning_rate": 5.558744378261838e-07,
+      "loss": 0.4794,
+      "step": 7424
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2519760106994378,
+      "learning_rate": 5.550219888277925e-07,
+      "loss": 0.5467,
+      "step": 7425
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.274203674030171,
+      "learning_rate": 5.541701555469847e-07,
+      "loss": 0.4564,
+      "step": 7426
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.03904845205746,
+      "learning_rate": 5.533189381017562e-07,
+      "loss": 0.4055,
+      "step": 7427
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.791902391858443,
+      "learning_rate": 5.524683366100192e-07,
+      "loss": 0.4861,
+      "step": 7428
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.553883234470799,
+      "learning_rate": 5.516183511895979e-07,
+      "loss": 0.506,
+      "step": 7429
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.927828753670472,
+      "learning_rate": 5.507689819582312e-07,
+      "loss": 0.5527,
+      "step": 7430
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.14131738300678,
+      "learning_rate": 5.499202290335754e-07,
+      "loss": 0.5234,
+      "step": 7431
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.317922969490147,
+      "learning_rate": 5.490720925331988e-07,
+      "loss": 0.4588,
+      "step": 7432
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8989339617481684,
+      "learning_rate": 5.482245725745838e-07,
+      "loss": 0.5264,
+      "step": 7433
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8727741966306468,
+      "learning_rate": 5.473776692751315e-07,
+      "loss": 0.4525,
+      "step": 7434
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.201912153173782,
+      "learning_rate": 5.465313827521518e-07,
+      "loss": 0.5183,
+      "step": 7435
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.910531175184344,
+      "learning_rate": 5.456857131228743e-07,
+      "loss": 0.5138,
+      "step": 7436
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0405328944223204,
+      "learning_rate": 5.448406605044398e-07,
+      "loss": 0.5227,
+      "step": 7437
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.990494301689452,
+      "learning_rate": 5.439962250139058e-07,
+      "loss": 0.5237,
+      "step": 7438
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.412908144764464,
+      "learning_rate": 5.431524067682426e-07,
+      "loss": 0.5102,
+      "step": 7439
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9537980283123841,
+      "learning_rate": 5.423092058843365e-07,
+      "loss": 0.4347,
+      "step": 7440
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4212470704707916,
+      "learning_rate": 5.414666224789866e-07,
+      "loss": 0.4625,
+      "step": 7441
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8890819629395847,
+      "learning_rate": 5.406246566689083e-07,
+      "loss": 0.4552,
+      "step": 7442
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.052124001674223,
+      "learning_rate": 5.397833085707299e-07,
+      "loss": 0.4724,
+      "step": 7443
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8986847731790975,
+      "learning_rate": 5.389425783009955e-07,
+      "loss": 0.4344,
+      "step": 7444
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3410407172732035,
+      "learning_rate": 5.381024659761624e-07,
+      "loss": 0.5271,
+      "step": 7445
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.656594892094784,
+      "learning_rate": 5.372629717126033e-07,
+      "loss": 0.5087,
+      "step": 7446
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.288793186539535,
+      "learning_rate": 5.364240956266031e-07,
+      "loss": 0.5156,
+      "step": 7447
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.008679332605421,
+      "learning_rate": 5.355858378343653e-07,
+      "loss": 0.4764,
+      "step": 7448
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.96839773703762,
+      "learning_rate": 5.347481984520031e-07,
+      "loss": 0.4203,
+      "step": 7449
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.275713169463398,
+      "learning_rate": 5.339111775955475e-07,
+      "loss": 0.5153,
+      "step": 7450
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.48068613428696,
+      "learning_rate": 5.330747753809423e-07,
+      "loss": 0.481,
+      "step": 7451
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3287475387582415,
+      "learning_rate": 5.32238991924045e-07,
+      "loss": 0.4605,
+      "step": 7452
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5777994765805736,
+      "learning_rate": 5.314038273406291e-07,
+      "loss": 0.4057,
+      "step": 7453
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3979543114221613,
+      "learning_rate": 5.305692817463803e-07,
+      "loss": 0.5007,
+      "step": 7454
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.353294508241976,
+      "learning_rate": 5.297353552569012e-07,
+      "loss": 0.5123,
+      "step": 7455
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1625963337692955,
+      "learning_rate": 5.289020479877055e-07,
+      "loss": 0.5288,
+      "step": 7456
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8236788696298336,
+      "learning_rate": 5.280693600542247e-07,
+      "loss": 0.4559,
+      "step": 7457
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.601304463244207,
+      "learning_rate": 5.27237291571801e-07,
+      "loss": 0.5463,
+      "step": 7458
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.695240440914216,
+      "learning_rate": 5.264058426556934e-07,
+      "loss": 0.4991,
+      "step": 7459
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2838188417449214,
+      "learning_rate": 5.25575013421073e-07,
+      "loss": 0.5237,
+      "step": 7460
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8865845075275451,
+      "learning_rate": 5.247448039830277e-07,
+      "loss": 0.4752,
+      "step": 7461
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.817657850117066,
+      "learning_rate": 5.239152144565557e-07,
+      "loss": 0.4576,
+      "step": 7462
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0936638824475673,
+      "learning_rate": 5.230862449565755e-07,
+      "loss": 0.4935,
+      "step": 7463
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.189477321752696,
+      "learning_rate": 5.222578955979107e-07,
+      "loss": 0.4369,
+      "step": 7464
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.247748327507094,
+      "learning_rate": 5.214301664953075e-07,
+      "loss": 0.4046,
+      "step": 7465
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.436868510031934,
+      "learning_rate": 5.206030577634214e-07,
+      "loss": 0.5248,
+      "step": 7466
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.539791197896571,
+      "learning_rate": 5.19776569516825e-07,
+      "loss": 0.4151,
+      "step": 7467
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2760368345960984,
+      "learning_rate": 5.189507018700007e-07,
+      "loss": 0.4938,
+      "step": 7468
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.430810431601701,
+      "learning_rate": 5.181254549373505e-07,
+      "loss": 0.4728,
+      "step": 7469
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.245732223229314,
+      "learning_rate": 5.173008288331843e-07,
+      "loss": 0.517,
+      "step": 7470
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8906805604225645,
+      "learning_rate": 5.164768236717326e-07,
+      "loss": 0.5882,
+      "step": 7471
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.858442766536668,
+      "learning_rate": 5.156534395671342e-07,
+      "loss": 0.5328,
+      "step": 7472
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9210649022042006,
+      "learning_rate": 5.148306766334438e-07,
+      "loss": 0.4368,
+      "step": 7473
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.602497516123075,
+      "learning_rate": 5.140085349846324e-07,
+      "loss": 0.4218,
+      "step": 7474
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8326532634441488,
+      "learning_rate": 5.131870147345808e-07,
+      "loss": 0.536,
+      "step": 7475
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1422459040202435,
+      "learning_rate": 5.123661159970872e-07,
+      "loss": 0.4156,
+      "step": 7476
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.288292535626195,
+      "learning_rate": 5.115458388858613e-07,
+      "loss": 0.5201,
+      "step": 7477
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.280316333386202,
+      "learning_rate": 5.107261835145294e-07,
+      "loss": 0.3821,
+      "step": 7478
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6011649528916359,
+      "learning_rate": 5.099071499966279e-07,
+      "loss": 0.4292,
+      "step": 7479
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9634467598732879,
+      "learning_rate": 5.090887384456127e-07,
+      "loss": 0.5263,
+      "step": 7480
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.854036052821305,
+      "learning_rate": 5.08270948974845e-07,
+      "loss": 0.5598,
+      "step": 7481
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8054213757972821,
+      "learning_rate": 5.074537816976089e-07,
+      "loss": 0.69,
+      "step": 7482
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.832278853643356,
+      "learning_rate": 5.066372367270955e-07,
+      "loss": 0.4698,
+      "step": 7483
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.668743812932087,
+      "learning_rate": 5.058213141764151e-07,
+      "loss": 0.3872,
+      "step": 7484
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.292126066413465,
+      "learning_rate": 5.050060141585866e-07,
+      "loss": 0.5002,
+      "step": 7485
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.637549916453231,
+      "learning_rate": 5.041913367865475e-07,
+      "loss": 0.414,
+      "step": 7486
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3914156546030223,
+      "learning_rate": 5.033772821731447e-07,
+      "loss": 0.6529,
+      "step": 7487
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.004763845281532,
+      "learning_rate": 5.025638504311431e-07,
+      "loss": 0.5122,
+      "step": 7488
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.40200742298774,
+      "learning_rate": 5.017510416732169e-07,
+      "loss": 0.5485,
+      "step": 7489
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7873289842726132,
+      "learning_rate": 5.009388560119583e-07,
+      "loss": 0.5399,
+      "step": 7490
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.251080739973806,
+      "learning_rate": 5.0012729355987e-07,
+      "loss": 0.4036,
+      "step": 7491
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8710136111710667,
+      "learning_rate": 4.993163544293689e-07,
+      "loss": 0.3984,
+      "step": 7492
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.4801780620277,
+      "learning_rate": 4.985060387327872e-07,
+      "loss": 0.4531,
+      "step": 7493
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.337404458610312,
+      "learning_rate": 4.976963465823686e-07,
+      "loss": 0.5495,
+      "step": 7494
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9616018509901116,
+      "learning_rate": 4.968872780902739e-07,
+      "loss": 0.4503,
+      "step": 7495
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1278768332354505,
+      "learning_rate": 4.960788333685729e-07,
+      "loss": 0.5188,
+      "step": 7496
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8193805983553766,
+      "learning_rate": 4.952710125292515e-07,
+      "loss": 0.5065,
+      "step": 7497
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7420640235334077,
+      "learning_rate": 4.944638156842086e-07,
+      "loss": 0.4362,
+      "step": 7498
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.715360849479292,
+      "learning_rate": 4.936572429452585e-07,
+      "loss": 0.3728,
+      "step": 7499
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.35047211516371,
+      "learning_rate": 4.928512944241259e-07,
+      "loss": 0.602,
+      "step": 7500
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2032144104208675,
+      "learning_rate": 4.92045970232452e-07,
+      "loss": 0.5692,
+      "step": 7501
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.315436215093921,
+      "learning_rate": 4.912412704817882e-07,
+      "loss": 0.4412,
+      "step": 7502
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.561208254704693,
+      "learning_rate": 4.90437195283604e-07,
+      "loss": 0.5435,
+      "step": 7503
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1285488672474813,
+      "learning_rate": 4.896337447492777e-07,
+      "loss": 0.4126,
+      "step": 7504
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.418404871475747,
+      "learning_rate": 4.888309189901047e-07,
+      "loss": 0.4225,
+      "step": 7505
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8156476880753913,
+      "learning_rate": 4.880287181172905e-07,
+      "loss": 0.4272,
+      "step": 7506
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.6184234218254083,
+      "learning_rate": 4.872271422419572e-07,
+      "loss": 0.5367,
+      "step": 7507
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4102223705954655,
+      "learning_rate": 4.864261914751384e-07,
+      "loss": 0.4896,
+      "step": 7508
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.607189392234605,
+      "learning_rate": 4.856258659277818e-07,
+      "loss": 0.5388,
+      "step": 7509
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.894098615547395,
+      "learning_rate": 4.848261657107489e-07,
+      "loss": 0.4778,
+      "step": 7510
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.711126982510492,
+      "learning_rate": 4.840270909348127e-07,
+      "loss": 0.5282,
+      "step": 7511
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9900422783623928,
+      "learning_rate": 4.832286417106625e-07,
+      "loss": 0.4204,
+      "step": 7512
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1723320276684794,
+      "learning_rate": 4.824308181488979e-07,
+      "loss": 0.453,
+      "step": 7513
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6545563238851657,
+      "learning_rate": 4.816336203600335e-07,
+      "loss": 0.5992,
+      "step": 7514
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9543991706579598,
+      "learning_rate": 4.808370484544983e-07,
+      "loss": 0.5535,
+      "step": 7515
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7896131603545111,
+      "learning_rate": 4.800411025426327e-07,
+      "loss": 0.515,
+      "step": 7516
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.467694598012592,
+      "learning_rate": 4.792457827346891e-07,
+      "loss": 0.4574,
+      "step": 7517
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6959573048144598,
+      "learning_rate": 4.784510891408384e-07,
+      "loss": 0.4071,
+      "step": 7518
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.538551124852352,
+      "learning_rate": 4.776570218711579e-07,
+      "loss": 0.603,
+      "step": 7519
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3882086650261427,
+      "learning_rate": 4.768635810356448e-07,
+      "loss": 0.5168,
+      "step": 7520
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.287792532683033,
+      "learning_rate": 4.760707667442044e-07,
+      "loss": 0.5717,
+      "step": 7521
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4629981307283093,
+      "learning_rate": 4.752785791066583e-07,
+      "loss": 0.475,
+      "step": 7522
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8116538011486973,
+      "learning_rate": 4.744870182327388e-07,
+      "loss": 0.5354,
+      "step": 7523
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5116871383925488,
+      "learning_rate": 4.736960842320948e-07,
+      "loss": 0.5967,
+      "step": 7524
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.36716815419743,
+      "learning_rate": 4.729057772142842e-07,
+      "loss": 0.5885,
+      "step": 7525
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.4475358017803974,
+      "learning_rate": 4.721160972887823e-07,
+      "loss": 0.4357,
+      "step": 7526
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1782520953726636,
+      "learning_rate": 4.7132704456497314e-07,
+      "loss": 0.4876,
+      "step": 7527
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7500905113878773,
+      "learning_rate": 4.705386191521588e-07,
+      "loss": 0.5218,
+      "step": 7528
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0629892189182644,
+      "learning_rate": 4.697508211595492e-07,
+      "loss": 0.4484,
+      "step": 7529
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.321121311455933,
+      "learning_rate": 4.6896365069627246e-07,
+      "loss": 0.5005,
+      "step": 7530
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.498448567084177,
+      "learning_rate": 4.6817710787136486e-07,
+      "loss": 0.4248,
+      "step": 7531
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.321289687938903,
+      "learning_rate": 4.673911927937802e-07,
+      "loss": 0.502,
+      "step": 7532
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5052951072668868,
+      "learning_rate": 4.666059055723815e-07,
+      "loss": 0.4723,
+      "step": 7533
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8895375386432036,
+      "learning_rate": 4.6582124631594836e-07,
+      "loss": 0.4204,
+      "step": 7534
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1003799561355723,
+      "learning_rate": 4.6503721513317004e-07,
+      "loss": 0.4149,
+      "step": 7535
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.010101496357928,
+      "learning_rate": 4.6425381213265177e-07,
+      "loss": 0.4661,
+      "step": 7536
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.4539598171914836,
+      "learning_rate": 4.634710374229101e-07,
+      "loss": 0.3747,
+      "step": 7537
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2007487055175696,
+      "learning_rate": 4.626888911123739e-07,
+      "loss": 0.5049,
+      "step": 7538
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6540120031111214,
+      "learning_rate": 4.619073733093871e-07,
+      "loss": 0.4898,
+      "step": 7539
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7305761254661562,
+      "learning_rate": 4.6112648412220404e-07,
+      "loss": 0.4862,
+      "step": 7540
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4111604569080427,
+      "learning_rate": 4.6034622365899563e-07,
+      "loss": 0.5019,
+      "step": 7541
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0405162086909336,
+      "learning_rate": 4.595665920278408e-07,
+      "loss": 0.5327,
+      "step": 7542
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9608567758775484,
+      "learning_rate": 4.587875893367361e-07,
+      "loss": 0.5201,
+      "step": 7543
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.471351893342806,
+      "learning_rate": 4.5800921569358723e-07,
+      "loss": 0.4711,
+      "step": 7544
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2555620054457632,
+      "learning_rate": 4.572314712062159e-07,
+      "loss": 0.5299,
+      "step": 7545
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8620946231082607,
+      "learning_rate": 4.564543559823531e-07,
+      "loss": 0.5069,
+      "step": 7546
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9170212126994397,
+      "learning_rate": 4.5567787012964826e-07,
+      "loss": 0.4502,
+      "step": 7547
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5994814476165717,
+      "learning_rate": 4.549020137556559e-07,
+      "loss": 0.4379,
+      "step": 7548
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9795954128202704,
+      "learning_rate": 4.5412678696785007e-07,
+      "loss": 0.4387,
+      "step": 7549
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.018285653542061,
+      "learning_rate": 4.533521898736132e-07,
+      "loss": 0.5368,
+      "step": 7550
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0308004086066225,
+      "learning_rate": 4.52578222580245e-07,
+      "loss": 0.4131,
+      "step": 7551
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6287344952768934,
+      "learning_rate": 4.5180488519495246e-07,
+      "loss": 0.5289,
+      "step": 7552
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.009431965692164,
+      "learning_rate": 4.5103217782486053e-07,
+      "loss": 0.4215,
+      "step": 7553
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.83465563681277,
+      "learning_rate": 4.5026010057700186e-07,
+      "loss": 0.5095,
+      "step": 7554
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8910496617226145,
+      "learning_rate": 4.494886535583276e-07,
+      "loss": 0.4467,
+      "step": 7555
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4201526145443437,
+      "learning_rate": 4.487178368756967e-07,
+      "loss": 0.5583,
+      "step": 7556
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3835953142310613,
+      "learning_rate": 4.4794765063588207e-07,
+      "loss": 0.4061,
+      "step": 7557
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1120525521200597,
+      "learning_rate": 4.471780949455712e-07,
+      "loss": 0.4898,
+      "step": 7558
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.323746980456667,
+      "learning_rate": 4.4640916991136096e-07,
+      "loss": 0.4664,
+      "step": 7559
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.807541042737632,
+      "learning_rate": 4.456408756397651e-07,
+      "loss": 0.6671,
+      "step": 7560
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9930287038756664,
+      "learning_rate": 4.4487321223720516e-07,
+      "loss": 0.5729,
+      "step": 7561
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8427141621996793,
+      "learning_rate": 4.4410617981001934e-07,
+      "loss": 0.6786,
+      "step": 7562
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9801169210275944,
+      "learning_rate": 4.433397784644561e-07,
+      "loss": 0.4996,
+      "step": 7563
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8534931025692385,
+      "learning_rate": 4.425740083066793e-07,
+      "loss": 0.4843,
+      "step": 7564
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8228869320205519,
+      "learning_rate": 4.4180886944275914e-07,
+      "loss": 0.5617,
+      "step": 7565
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4970343715073633,
+      "learning_rate": 4.410443619786864e-07,
+      "loss": 0.526,
+      "step": 7566
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3130563545694187,
+      "learning_rate": 4.40280486020358e-07,
+      "loss": 0.5599,
+      "step": 7567
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.485456864375333,
+      "learning_rate": 4.3951724167358766e-07,
+      "loss": 0.4832,
+      "step": 7568
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5240161688697342,
+      "learning_rate": 4.3875462904409806e-07,
+      "loss": 0.495,
+      "step": 7569
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.282509741593416,
+      "learning_rate": 4.379926482375285e-07,
+      "loss": 0.4779,
+      "step": 7570
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6271917099069408,
+      "learning_rate": 4.372312993594258e-07,
+      "loss": 0.3866,
+      "step": 7571
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8691207948266293,
+      "learning_rate": 4.364705825152543e-07,
+      "loss": 0.4998,
+      "step": 7572
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9627580119973405,
+      "learning_rate": 4.357104978103865e-07,
+      "loss": 0.4986,
+      "step": 7573
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.500034605990953,
+      "learning_rate": 4.34951045350111e-07,
+      "loss": 0.556,
+      "step": 7574
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.12064812133031,
+      "learning_rate": 4.341922252396258e-07,
+      "loss": 0.5121,
+      "step": 7575
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.830750219903513,
+      "learning_rate": 4.334340375840418e-07,
+      "loss": 0.4013,
+      "step": 7576
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1664226726545976,
+      "learning_rate": 4.3267648248838446e-07,
+      "loss": 0.5316,
+      "step": 7577
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.055989796070457,
+      "learning_rate": 4.319195600575893e-07,
+      "loss": 0.4916,
+      "step": 7578
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1064678834924364,
+      "learning_rate": 4.311632703965063e-07,
+      "loss": 0.5412,
+      "step": 7579
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9017183348267375,
+      "learning_rate": 4.3040761360989503e-07,
+      "loss": 0.4888,
+      "step": 7580
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8082831017060668,
+      "learning_rate": 4.2965258980243116e-07,
+      "loss": 0.4364,
+      "step": 7581
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8632897058631996,
+      "learning_rate": 4.288981990786972e-07,
+      "loss": 0.4631,
+      "step": 7582
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4351204003025693,
+      "learning_rate": 4.28144441543194e-07,
+      "loss": 0.5043,
+      "step": 7583
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0871884040065516,
+      "learning_rate": 4.273913173003297e-07,
+      "loss": 0.4692,
+      "step": 7584
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0828053825434742,
+      "learning_rate": 4.266388264544291e-07,
+      "loss": 0.506,
+      "step": 7585
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7879559282966846,
+      "learning_rate": 4.258869691097256e-07,
+      "loss": 0.4153,
+      "step": 7586
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.185177368114945,
+      "learning_rate": 4.251357453703675e-07,
+      "loss": 0.4763,
+      "step": 7587
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.199721893337713,
+      "learning_rate": 4.243851553404127e-07,
+      "loss": 0.5243,
+      "step": 7588
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.581961493489355,
+      "learning_rate": 4.236351991238347e-07,
+      "loss": 0.4515,
+      "step": 7589
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2269169342791253,
+      "learning_rate": 4.2288587682451534e-07,
+      "loss": 0.5182,
+      "step": 7590
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.677518784819435,
+      "learning_rate": 4.221371885462522e-07,
+      "loss": 0.4909,
+      "step": 7591
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.065194191043958,
+      "learning_rate": 4.2138913439275173e-07,
+      "loss": 0.4637,
+      "step": 7592
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8198682163949804,
+      "learning_rate": 4.206417144676367e-07,
+      "loss": 0.6692,
+      "step": 7593
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1448505436570664,
+      "learning_rate": 4.1989492887443697e-07,
+      "loss": 0.5272,
+      "step": 7594
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9841489831245294,
+      "learning_rate": 4.1914877771659925e-07,
+      "loss": 0.4991,
+      "step": 7595
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.878147052110389,
+      "learning_rate": 4.1840326109747974e-07,
+      "loss": 0.424,
+      "step": 7596
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6150963184659404,
+      "learning_rate": 4.176583791203459e-07,
+      "loss": 0.5105,
+      "step": 7597
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9396595405128974,
+      "learning_rate": 4.169141318883807e-07,
+      "loss": 0.4205,
+      "step": 7598
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.520451829596819,
+      "learning_rate": 4.1617051950467613e-07,
+      "loss": 0.4737,
+      "step": 7599
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9721295550496296,
+      "learning_rate": 4.154275420722359e-07,
+      "loss": 0.4234,
+      "step": 7600
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1640747745046762,
+      "learning_rate": 4.1468519969397993e-07,
+      "loss": 0.5124,
+      "step": 7601
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4308831681560097,
+      "learning_rate": 4.139434924727359e-07,
+      "loss": 0.469,
+      "step": 7602
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0390990258898096,
+      "learning_rate": 4.1320242051124395e-07,
+      "loss": 0.5201,
+      "step": 7603
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.704325546544485,
+      "learning_rate": 4.1246198391215853e-07,
+      "loss": 0.4716,
+      "step": 7604
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7961369562525837,
+      "learning_rate": 4.117221827780443e-07,
+      "loss": 0.4738,
+      "step": 7605
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.136810796271047,
+      "learning_rate": 4.109830172113793e-07,
+      "loss": 0.4365,
+      "step": 7606
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.920080851974724,
+      "learning_rate": 4.102444873145511e-07,
+      "loss": 0.4449,
+      "step": 7607
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9729734470636426,
+      "learning_rate": 4.095065931898623e-07,
+      "loss": 0.5439,
+      "step": 7608
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7844033086701856,
+      "learning_rate": 4.0876933493952444e-07,
+      "loss": 0.5091,
+      "step": 7609
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2201202945661884,
+      "learning_rate": 4.080327126656641e-07,
+      "loss": 0.5655,
+      "step": 7610
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4503506877360866,
+      "learning_rate": 4.0729672647031593e-07,
+      "loss": 0.4812,
+      "step": 7611
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8543311802120197,
+      "learning_rate": 4.065613764554305e-07,
+      "loss": 0.4991,
+      "step": 7612
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.210093465820383,
+      "learning_rate": 4.0582666272286685e-07,
+      "loss": 0.5658,
+      "step": 7613
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3728375979821306,
+      "learning_rate": 4.0509258537439866e-07,
+      "loss": 0.5008,
+      "step": 7614
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.632928450192723,
+      "learning_rate": 4.043591445117101e-07,
+      "loss": 0.5199,
+      "step": 7615
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9584567907168937,
+      "learning_rate": 4.0362634023639713e-07,
+      "loss": 0.4008,
+      "step": 7616
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.414719967964127,
+      "learning_rate": 4.028941726499658e-07,
+      "loss": 0.5683,
+      "step": 7617
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7871151914851655,
+      "learning_rate": 4.021626418538388e-07,
+      "loss": 0.4876,
+      "step": 7618
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.315510466823666,
+      "learning_rate": 4.0143174794934516e-07,
+      "loss": 0.4824,
+      "step": 7619
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.550212452180025,
+      "learning_rate": 4.007014910377305e-07,
+      "loss": 0.517,
+      "step": 7620
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9865535921344637,
+      "learning_rate": 3.999718712201484e-07,
+      "loss": 0.5212,
+      "step": 7621
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5375720164956674,
+      "learning_rate": 3.9924288859766514e-07,
+      "loss": 0.4375,
+      "step": 7622
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9771022495100916,
+      "learning_rate": 3.985145432712606e-07,
+      "loss": 0.4561,
+      "step": 7623
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2274032834331448,
+      "learning_rate": 3.9778683534182403e-07,
+      "loss": 0.4892,
+      "step": 7624
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4149666658440734,
+      "learning_rate": 3.9705976491015874e-07,
+      "loss": 0.5119,
+      "step": 7625
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7399396363204704,
+      "learning_rate": 3.963333320769769e-07,
+      "loss": 0.537,
+      "step": 7626
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.586987277356396,
+      "learning_rate": 3.956075369429052e-07,
+      "loss": 0.5208,
+      "step": 7627
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8703693874369917,
+      "learning_rate": 3.948823796084789e-07,
+      "loss": 0.6761,
+      "step": 7628
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5742361972194594,
+      "learning_rate": 3.941578601741491e-07,
+      "loss": 0.5332,
+      "step": 7629
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7847256195736128,
+      "learning_rate": 3.934339787402736e-07,
+      "loss": 0.4645,
+      "step": 7630
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9681946171498312,
+      "learning_rate": 3.927107354071269e-07,
+      "loss": 0.5269,
+      "step": 7631
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.177223105812872,
+      "learning_rate": 3.9198813027488956e-07,
+      "loss": 0.4951,
+      "step": 7632
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.667181161705969,
+      "learning_rate": 3.9126616344365933e-07,
+      "loss": 0.4384,
+      "step": 7633
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8102901722703293,
+      "learning_rate": 3.905448350134411e-07,
+      "loss": 0.516,
+      "step": 7634
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0195748339134885,
+      "learning_rate": 3.8982414508415445e-07,
+      "loss": 0.5234,
+      "step": 7635
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.0240742798109825,
+      "learning_rate": 3.891040937556284e-07,
+      "loss": 0.4229,
+      "step": 7636
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9735427922366915,
+      "learning_rate": 3.8838468112760485e-07,
+      "loss": 0.4627,
+      "step": 7637
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.63888436315548,
+      "learning_rate": 3.876659072997363e-07,
+      "loss": 0.5075,
+      "step": 7638
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8804080373674346,
+      "learning_rate": 3.869477723715881e-07,
+      "loss": 0.6041,
+      "step": 7639
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.775846574354268,
+      "learning_rate": 3.862302764426351e-07,
+      "loss": 0.5119,
+      "step": 7640
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.211944554532105,
+      "learning_rate": 3.855134196122645e-07,
+      "loss": 0.4835,
+      "step": 7641
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.39585833870452,
+      "learning_rate": 3.847972019797763e-07,
+      "loss": 0.4659,
+      "step": 7642
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.177169041259043,
+      "learning_rate": 3.8408162364438004e-07,
+      "loss": 0.5575,
+      "step": 7643
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4226263043834115,
+      "learning_rate": 3.833666847051981e-07,
+      "loss": 0.432,
+      "step": 7644
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9584017361848907,
+      "learning_rate": 3.826523852612629e-07,
+      "loss": 0.5794,
+      "step": 7645
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.676550378382369,
+      "learning_rate": 3.8193872541151976e-07,
+      "loss": 0.461,
+      "step": 7646
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.290950066290489,
+      "learning_rate": 3.812257052548246e-07,
+      "loss": 0.5006,
+      "step": 7647
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6514198033085647,
+      "learning_rate": 3.8051332488994565e-07,
+      "loss": 0.4542,
+      "step": 7648
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0979813018291282,
+      "learning_rate": 3.798015844155595e-07,
+      "loss": 0.4694,
+      "step": 7649
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8896330434950261,
+      "learning_rate": 3.790904839302584e-07,
+      "loss": 0.4557,
+      "step": 7650
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0573358495363783,
+      "learning_rate": 3.7838002353254243e-07,
+      "loss": 0.5122,
+      "step": 7651
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.006059148635446,
+      "learning_rate": 3.776702033208257e-07,
+      "loss": 0.5127,
+      "step": 7652
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.773230512060994,
+      "learning_rate": 3.7696102339343067e-07,
+      "loss": 0.4881,
+      "step": 7653
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.011246293340883,
+      "learning_rate": 3.7625248384859536e-07,
+      "loss": 0.5343,
+      "step": 7654
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.736556812428555,
+      "learning_rate": 3.755445847844641e-07,
+      "loss": 0.3857,
+      "step": 7655
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.613376102658403,
+      "learning_rate": 3.7483732629909673e-07,
+      "loss": 0.5046,
+      "step": 7656
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3114825902498857,
+      "learning_rate": 3.741307084904611e-07,
+      "loss": 0.535,
+      "step": 7657
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.906164401445148,
+      "learning_rate": 3.734247314564393e-07,
+      "loss": 0.4092,
+      "step": 7658
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.622250686792475,
+      "learning_rate": 3.727193952948216e-07,
+      "loss": 0.5649,
+      "step": 7659
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2921323813166135,
+      "learning_rate": 3.720147001033125e-07,
+      "loss": 0.451,
+      "step": 7660
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6287354790537265,
+      "learning_rate": 3.7131064597952517e-07,
+      "loss": 0.4905,
+      "step": 7661
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0666547816300813,
+      "learning_rate": 3.706072330209853e-07,
+      "loss": 0.5872,
+      "step": 7662
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9262064370297467,
+      "learning_rate": 3.6990446132513014e-07,
+      "loss": 0.5343,
+      "step": 7663
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8463670618413531,
+      "learning_rate": 3.6920233098930614e-07,
+      "loss": 0.6583,
+      "step": 7664
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.013682512529784,
+      "learning_rate": 3.685008421107744e-07,
+      "loss": 0.3735,
+      "step": 7665
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6217176621340392,
+      "learning_rate": 3.6779999478670337e-07,
+      "loss": 0.3438,
+      "step": 7666
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1155802754632727,
+      "learning_rate": 3.670997891141753e-07,
+      "loss": 0.4704,
+      "step": 7667
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1504878098407323,
+      "learning_rate": 3.6640022519018046e-07,
+      "loss": 0.5195,
+      "step": 7668
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.191136715546454,
+      "learning_rate": 3.657013031116252e-07,
+      "loss": 0.5064,
+      "step": 7669
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9404872179957902,
+      "learning_rate": 3.6500302297532154e-07,
+      "loss": 0.5422,
+      "step": 7670
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.160099085520927,
+      "learning_rate": 3.643053848779976e-07,
+      "loss": 0.4684,
+      "step": 7671
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.122939551053653,
+      "learning_rate": 3.636083889162878e-07,
+      "loss": 0.341,
+      "step": 7672
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6812121227885597,
+      "learning_rate": 3.629120351867416e-07,
+      "loss": 0.5076,
+      "step": 7673
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3797288627740376,
+      "learning_rate": 3.6221632378581616e-07,
+      "loss": 0.5253,
+      "step": 7674
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0319780553520372,
+      "learning_rate": 3.615212548098834e-07,
+      "loss": 0.4276,
+      "step": 7675
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.868804732116299,
+      "learning_rate": 3.6082682835522245e-07,
+      "loss": 0.5022,
+      "step": 7676
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2819979632788727,
+      "learning_rate": 3.601330445180262e-07,
+      "loss": 0.4387,
+      "step": 7677
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3076599046546082,
+      "learning_rate": 3.594399033943963e-07,
+      "loss": 0.4153,
+      "step": 7678
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9975086667441697,
+      "learning_rate": 3.5874740508034744e-07,
+      "loss": 0.4853,
+      "step": 7679
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.933047421896532,
+      "learning_rate": 3.5805554967180467e-07,
+      "loss": 0.3763,
+      "step": 7680
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1011014465582996,
+      "learning_rate": 3.5736433726460243e-07,
+      "loss": 0.4322,
+      "step": 7681
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4366798178867417,
+      "learning_rate": 3.566737679544885e-07,
+      "loss": 0.4687,
+      "step": 7682
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1359394526995357,
+      "learning_rate": 3.5598384183712033e-07,
+      "loss": 0.4901,
+      "step": 7683
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.835377871972897,
+      "learning_rate": 3.552945590080653e-07,
+      "loss": 0.5058,
+      "step": 7684
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8637498660216119,
+      "learning_rate": 3.546059195628038e-07,
+      "loss": 0.4098,
+      "step": 7685
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3482424971739007,
+      "learning_rate": 3.5391792359672605e-07,
+      "loss": 0.5955,
+      "step": 7686
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.918424979794592,
+      "learning_rate": 3.5323057120513206e-07,
+      "loss": 0.5,
+      "step": 7687
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1253241595733146,
+      "learning_rate": 3.5254386248323504e-07,
+      "loss": 0.3799,
+      "step": 7688
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5316001483898267,
+      "learning_rate": 3.5185779752615615e-07,
+      "loss": 0.5699,
+      "step": 7689
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9755734816467194,
+      "learning_rate": 3.5117237642893153e-07,
+      "loss": 0.4022,
+      "step": 7690
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5152519540655685,
+      "learning_rate": 3.504875992865031e-07,
+      "loss": 0.4126,
+      "step": 7691
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8459814563889863,
+      "learning_rate": 3.4980346619372776e-07,
+      "loss": 0.4908,
+      "step": 7692
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5237017988286428,
+      "learning_rate": 3.4911997724537016e-07,
+      "loss": 0.4812,
+      "step": 7693
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8863619885869625,
+      "learning_rate": 3.484371325361086e-07,
+      "loss": 0.5379,
+      "step": 7694
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8008544136479896,
+      "learning_rate": 3.477549321605289e-07,
+      "loss": 0.5266,
+      "step": 7695
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.095555357636055,
+      "learning_rate": 3.4707337621313066e-07,
+      "loss": 0.4798,
+      "step": 7696
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2290302447944317,
+      "learning_rate": 3.463924647883221e-07,
+      "loss": 0.3975,
+      "step": 7697
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9273922420198621,
+      "learning_rate": 3.45712197980424e-07,
+      "loss": 0.4769,
+      "step": 7698
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1630417183333215,
+      "learning_rate": 3.450325758836659e-07,
+      "loss": 0.5193,
+      "step": 7699
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7419965722192794,
+      "learning_rate": 3.443535985921892e-07,
+      "loss": 0.5145,
+      "step": 7700
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.525127605511081,
+      "learning_rate": 3.436752662000448e-07,
+      "loss": 0.458,
+      "step": 7701
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.840323738437548,
+      "learning_rate": 3.4299757880119686e-07,
+      "loss": 0.4776,
+      "step": 7702
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.878628845690861,
+      "learning_rate": 3.423205364895171e-07,
+      "loss": 0.4476,
+      "step": 7703
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6400156119136993,
+      "learning_rate": 3.4164413935879046e-07,
+      "loss": 0.3143,
+      "step": 7704
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7797403688488327,
+      "learning_rate": 3.4096838750271087e-07,
+      "loss": 0.5563,
+      "step": 7705
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8234686096849333,
+      "learning_rate": 3.4029328101488293e-07,
+      "loss": 0.5051,
+      "step": 7706
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.4755845756027597,
+      "learning_rate": 3.3961881998882285e-07,
+      "loss": 0.4257,
+      "step": 7707
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.522880705937556,
+      "learning_rate": 3.3894500451795597e-07,
+      "loss": 0.4489,
+      "step": 7708
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.548021454071435,
+      "learning_rate": 3.382718346956204e-07,
+      "loss": 0.4839,
+      "step": 7709
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0246695343609917,
+      "learning_rate": 3.375993106150621e-07,
+      "loss": 0.515,
+      "step": 7710
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.938437945090546,
+      "learning_rate": 3.369274323694405e-07,
+      "loss": 0.5125,
+      "step": 7711
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4190789947974656,
+      "learning_rate": 3.362562000518227e-07,
+      "loss": 0.4284,
+      "step": 7712
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1202512731788032,
+      "learning_rate": 3.355856137551888e-07,
+      "loss": 0.5927,
+      "step": 7713
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.912847185010895,
+      "learning_rate": 3.3491567357242736e-07,
+      "loss": 0.4627,
+      "step": 7714
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2200391498301246,
+      "learning_rate": 3.3424637959633964e-07,
+      "loss": 0.5148,
+      "step": 7715
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9414270599981116,
+      "learning_rate": 3.335777319196348e-07,
+      "loss": 0.4688,
+      "step": 7716
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8458331594935771,
+      "learning_rate": 3.3290973063493437e-07,
+      "loss": 0.6902,
+      "step": 7717
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0967144022055346,
+      "learning_rate": 3.3224237583476924e-07,
+      "loss": 0.433,
+      "step": 7718
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.388217349491806,
+      "learning_rate": 3.315756676115822e-07,
+      "loss": 0.5036,
+      "step": 7719
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2631294466185308,
+      "learning_rate": 3.309096060577244e-07,
+      "loss": 0.5007,
+      "step": 7720
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9035945867096333,
+      "learning_rate": 3.3024419126546026e-07,
+      "loss": 0.4711,
+      "step": 7721
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.113104448797073,
+      "learning_rate": 3.295794233269611e-07,
+      "loss": 0.4494,
+      "step": 7722
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.787173529468723,
+      "learning_rate": 3.289153023343117e-07,
+      "loss": 0.4853,
+      "step": 7723
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7812707297385089,
+      "learning_rate": 3.2825182837950496e-07,
+      "loss": 0.5351,
+      "step": 7724
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4471721222126113,
+      "learning_rate": 3.27589001554447e-07,
+      "loss": 0.4612,
+      "step": 7725
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8333826762653511,
+      "learning_rate": 3.269268219509508e-07,
+      "loss": 0.5878,
+      "step": 7726
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2738888541239364,
+      "learning_rate": 3.26265289660741e-07,
+      "loss": 0.4697,
+      "step": 7727
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.430661640591568,
+      "learning_rate": 3.2560440477545474e-07,
+      "loss": 0.5221,
+      "step": 7728
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1560119625114877,
+      "learning_rate": 3.2494416738663547e-07,
+      "loss": 0.5069,
+      "step": 7729
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.374307837213059,
+      "learning_rate": 3.2428457758574173e-07,
+      "loss": 0.4999,
+      "step": 7730
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4206911794475285,
+      "learning_rate": 3.236256354641376e-07,
+      "loss": 0.4911,
+      "step": 7731
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.859164167260816,
+      "learning_rate": 3.229673411131007e-07,
+      "loss": 0.4454,
+      "step": 7732
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.990919302956601,
+      "learning_rate": 3.22309694623818e-07,
+      "loss": 0.5464,
+      "step": 7733
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2963688270103266,
+      "learning_rate": 3.216526960873867e-07,
+      "loss": 0.5367,
+      "step": 7734
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9743035442087487,
+      "learning_rate": 3.209963455948123e-07,
+      "loss": 0.54,
+      "step": 7735
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7157558145288492,
+      "learning_rate": 3.203406432370143e-07,
+      "loss": 0.4946,
+      "step": 7736
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8511911782957249,
+      "learning_rate": 3.196855891048195e-07,
+      "loss": 0.5035,
+      "step": 7737
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0580455456919853,
+      "learning_rate": 3.19031183288967e-07,
+      "loss": 0.5628,
+      "step": 7738
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5973663622580487,
+      "learning_rate": 3.183774258801031e-07,
+      "loss": 0.4052,
+      "step": 7739
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.065618512979532,
+      "learning_rate": 3.1772431696878827e-07,
+      "loss": 0.5071,
+      "step": 7740
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.123382797318764,
+      "learning_rate": 3.170718566454889e-07,
+      "loss": 0.4078,
+      "step": 7741
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.7147147594558705,
+      "learning_rate": 3.164200450005861e-07,
+      "loss": 0.5081,
+      "step": 7742
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8441308521121934,
+      "learning_rate": 3.157688821243665e-07,
+      "loss": 0.6617,
+      "step": 7743
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4911383668396474,
+      "learning_rate": 3.151183681070308e-07,
+      "loss": 0.4469,
+      "step": 7744
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.418827584094532,
+      "learning_rate": 3.144685030386874e-07,
+      "loss": 0.5005,
+      "step": 7745
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9391508504504427,
+      "learning_rate": 3.1381928700935484e-07,
+      "loss": 0.4535,
+      "step": 7746
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1615594203551316,
+      "learning_rate": 3.1317072010896344e-07,
+      "loss": 0.3997,
+      "step": 7747
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.537062001952909,
+      "learning_rate": 3.1252280242735136e-07,
+      "loss": 0.4466,
+      "step": 7748
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.374381247419036,
+      "learning_rate": 3.118755340542695e-07,
+      "loss": 0.5366,
+      "step": 7749
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.4153640618899415,
+      "learning_rate": 3.112289150793768e-07,
+      "loss": 0.5229,
+      "step": 7750
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1033695379162634,
+      "learning_rate": 3.105829455922427e-07,
+      "loss": 0.5169,
+      "step": 7751
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7321611745321257,
+      "learning_rate": 3.0993762568234554e-07,
+      "loss": 0.4287,
+      "step": 7752
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7286133803739854,
+      "learning_rate": 3.092929554390772e-07,
+      "loss": 0.4292,
+      "step": 7753
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8643817002166243,
+      "learning_rate": 3.0864893495173575e-07,
+      "loss": 0.4611,
+      "step": 7754
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.929251815169608,
+      "learning_rate": 3.0800556430953143e-07,
+      "loss": 0.4239,
+      "step": 7755
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.33003959594801,
+      "learning_rate": 3.073628436015835e-07,
+      "loss": 0.4121,
+      "step": 7756
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5612847041073423,
+      "learning_rate": 3.0672077291692193e-07,
+      "loss": 0.5089,
+      "step": 7757
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9480081914045402,
+      "learning_rate": 3.0607935234448547e-07,
+      "loss": 0.5639,
+      "step": 7758
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5149635915060675,
+      "learning_rate": 3.054385819731248e-07,
+      "loss": 0.4471,
+      "step": 7759
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.0014668209435893,
+      "learning_rate": 3.047984618915978e-07,
+      "loss": 0.4622,
+      "step": 7760
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0587821562889888,
+      "learning_rate": 3.041589921885757e-07,
+      "loss": 0.549,
+      "step": 7761
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.212448212242227,
+      "learning_rate": 3.0352017295263603e-07,
+      "loss": 0.4772,
+      "step": 7762
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0854718038892646,
+      "learning_rate": 3.0288200427226856e-07,
+      "loss": 0.4436,
+      "step": 7763
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.7174992735536927,
+      "learning_rate": 3.02244486235872e-07,
+      "loss": 0.5528,
+      "step": 7764
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9997415711691486,
+      "learning_rate": 3.0160761893175625e-07,
+      "loss": 0.4978,
+      "step": 7765
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.346855821794742,
+      "learning_rate": 3.009714024481397e-07,
+      "loss": 0.4614,
+      "step": 7766
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3959352863048253,
+      "learning_rate": 3.0033583687315013e-07,
+      "loss": 0.4974,
+      "step": 7767
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5043022999854316,
+      "learning_rate": 2.997009222948255e-07,
+      "loss": 0.4509,
+      "step": 7768
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8837659006289813,
+      "learning_rate": 2.990666588011165e-07,
+      "loss": 0.4205,
+      "step": 7769
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4228718694801596,
+      "learning_rate": 2.9843304647987904e-07,
+      "loss": 0.5316,
+      "step": 7770
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7235001140768424,
+      "learning_rate": 2.9780008541888116e-07,
+      "loss": 0.5393,
+      "step": 7771
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.074344524395402,
+      "learning_rate": 2.971677757058017e-07,
+      "loss": 0.4203,
+      "step": 7772
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9926817095464413,
+      "learning_rate": 2.9653611742822664e-07,
+      "loss": 0.5198,
+      "step": 7773
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2100538465240613,
+      "learning_rate": 2.9590511067365436e-07,
+      "loss": 0.4147,
+      "step": 7774
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.615202009297096,
+      "learning_rate": 2.9527475552949106e-07,
+      "loss": 0.4574,
+      "step": 7775
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6596655606151223,
+      "learning_rate": 2.946450520830546e-07,
+      "loss": 0.5196,
+      "step": 7776
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.770903818471776,
+      "learning_rate": 2.940160004215692e-07,
+      "loss": 0.5432,
+      "step": 7777
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.402518018017984,
+      "learning_rate": 2.9338760063217344e-07,
+      "loss": 0.5162,
+      "step": 7778
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7615263462726805,
+      "learning_rate": 2.9275985280191153e-07,
+      "loss": 0.5337,
+      "step": 7779
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9233575331746176,
+      "learning_rate": 2.921327570177396e-07,
+      "loss": 0.5657,
+      "step": 7780
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9936759571930225,
+      "learning_rate": 2.9150631336652245e-07,
+      "loss": 0.4473,
+      "step": 7781
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9427183915317798,
+      "learning_rate": 2.908805219350358e-07,
+      "loss": 0.4751,
+      "step": 7782
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.005507008673433,
+      "learning_rate": 2.902553828099636e-07,
+      "loss": 0.5409,
+      "step": 7783
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.17015499026607,
+      "learning_rate": 2.896308960779004e-07,
+      "loss": 0.5925,
+      "step": 7784
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.0320142624396254,
+      "learning_rate": 2.8900706182534874e-07,
+      "loss": 0.5564,
+      "step": 7785
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7341321731178418,
+      "learning_rate": 2.8838388013872344e-07,
+      "loss": 0.449,
+      "step": 7786
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7317662668692784,
+      "learning_rate": 2.877613511043459e-07,
+      "loss": 0.5358,
+      "step": 7787
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7777550529279151,
+      "learning_rate": 2.8713947480845103e-07,
+      "loss": 0.5052,
+      "step": 7788
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8631783602341898,
+      "learning_rate": 2.8651825133717894e-07,
+      "loss": 0.444,
+      "step": 7789
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.238934633838346,
+      "learning_rate": 2.858976807765834e-07,
+      "loss": 0.4943,
+      "step": 7790
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1387758587545447,
+      "learning_rate": 2.852777632126241e-07,
+      "loss": 0.5509,
+      "step": 7791
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5868507757215116,
+      "learning_rate": 2.8465849873117177e-07,
+      "loss": 0.5663,
+      "step": 7792
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.263252273686556,
+      "learning_rate": 2.840398874180084e-07,
+      "loss": 0.5056,
+      "step": 7793
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3541751331306076,
+      "learning_rate": 2.834219293588225e-07,
+      "loss": 0.4596,
+      "step": 7794
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.6048330471009375,
+      "learning_rate": 2.828046246392141e-07,
+      "loss": 0.4431,
+      "step": 7795
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0595725321667384,
+      "learning_rate": 2.821879733446919e-07,
+      "loss": 0.5222,
+      "step": 7796
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5483177348497805,
+      "learning_rate": 2.8157197556067495e-07,
+      "loss": 0.4616,
+      "step": 7797
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.728568095472027,
+      "learning_rate": 2.8095663137248984e-07,
+      "loss": 0.4933,
+      "step": 7798
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.185031737150029,
+      "learning_rate": 2.803419408653757e-07,
+      "loss": 0.4876,
+      "step": 7799
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.172349202060047,
+      "learning_rate": 2.7972790412447827e-07,
+      "loss": 0.5167,
+      "step": 7800
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.976727912092295,
+      "learning_rate": 2.791145212348539e-07,
+      "loss": 0.4744,
+      "step": 7801
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2534991986493904,
+      "learning_rate": 2.785017922814681e-07,
+      "loss": 0.5464,
+      "step": 7802
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3510773459806895,
+      "learning_rate": 2.778897173491968e-07,
+      "loss": 0.5094,
+      "step": 7803
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.97130035466109,
+      "learning_rate": 2.772782965228232e-07,
+      "loss": 0.4587,
+      "step": 7804
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.006764167875854,
+      "learning_rate": 2.766675298870425e-07,
+      "loss": 0.4666,
+      "step": 7805
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.938958638616173,
+      "learning_rate": 2.7605741752645686e-07,
+      "loss": 0.4192,
+      "step": 7806
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7852334010547655,
+      "learning_rate": 2.7544795952558045e-07,
+      "loss": 0.3891,
+      "step": 7807
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.025615234693282,
+      "learning_rate": 2.748391559688335e-07,
+      "loss": 0.5499,
+      "step": 7808
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.297855721679448,
+      "learning_rate": 2.742310069405485e-07,
+      "loss": 0.5158,
+      "step": 7809
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8545338457820344,
+      "learning_rate": 2.736235125249664e-07,
+      "loss": 0.4255,
+      "step": 7810
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.07817933727314,
+      "learning_rate": 2.730166728062361e-07,
+      "loss": 0.4137,
+      "step": 7811
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4258874537778103,
+      "learning_rate": 2.7241048786841805e-07,
+      "loss": 0.5015,
+      "step": 7812
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9924916361666247,
+      "learning_rate": 2.718049577954796e-07,
+      "loss": 0.4822,
+      "step": 7813
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5208362841367475,
+      "learning_rate": 2.7120008267130016e-07,
+      "loss": 0.5691,
+      "step": 7814
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.224433965261639,
+      "learning_rate": 2.7059586257966565e-07,
+      "loss": 0.4665,
+      "step": 7815
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9370963190426809,
+      "learning_rate": 2.699922976042735e-07,
+      "loss": 0.4578,
+      "step": 7816
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.124564876365193,
+      "learning_rate": 2.693893878287296e-07,
+      "loss": 0.514,
+      "step": 7817
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7393942931598239,
+      "learning_rate": 2.687871333365477e-07,
+      "loss": 0.5096,
+      "step": 7818
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.621501975933147,
+      "learning_rate": 2.6818553421115226e-07,
+      "loss": 0.5663,
+      "step": 7819
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.355057755956894,
+      "learning_rate": 2.675845905358776e-07,
+      "loss": 0.4714,
+      "step": 7820
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0595378830192024,
+      "learning_rate": 2.6698430239396557e-07,
+      "loss": 0.4783,
+      "step": 7821
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9258348031331385,
+      "learning_rate": 2.6638466986856847e-07,
+      "loss": 0.6038,
+      "step": 7822
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7988468341200656,
+      "learning_rate": 2.6578569304274604e-07,
+      "loss": 0.5295,
+      "step": 7823
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8827985673437198,
+      "learning_rate": 2.6518737199947077e-07,
+      "loss": 0.4528,
+      "step": 7824
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9847250216835213,
+      "learning_rate": 2.6458970682161964e-07,
+      "loss": 0.4569,
+      "step": 7825
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8663020254704306,
+      "learning_rate": 2.6399269759198266e-07,
+      "loss": 0.5017,
+      "step": 7826
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.740935843545659,
+      "learning_rate": 2.6339634439325634e-07,
+      "loss": 0.4507,
+      "step": 7827
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7323149699664946,
+      "learning_rate": 2.6280064730804853e-07,
+      "loss": 0.5093,
+      "step": 7828
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9615762734735587,
+      "learning_rate": 2.6220560641887385e-07,
+      "loss": 0.432,
+      "step": 7829
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.099720551705267,
+      "learning_rate": 2.61611221808159e-07,
+      "loss": 0.4206,
+      "step": 7830
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9990147881943885,
+      "learning_rate": 2.610174935582366e-07,
+      "loss": 0.5038,
+      "step": 7831
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.32052907609017,
+      "learning_rate": 2.604244217513496e-07,
+      "loss": 0.4088,
+      "step": 7832
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9795931261880877,
+      "learning_rate": 2.598320064696519e-07,
+      "loss": 0.579,
+      "step": 7833
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.641042032224614,
+      "learning_rate": 2.592402477952033e-07,
+      "loss": 0.4506,
+      "step": 7834
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0658239217940304,
+      "learning_rate": 2.5864914580997327e-07,
+      "loss": 0.4741,
+      "step": 7835
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7732258501777476,
+      "learning_rate": 2.580587005958435e-07,
+      "loss": 0.5072,
+      "step": 7836
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5403787781438294,
+      "learning_rate": 2.5746891223460135e-07,
+      "loss": 0.4934,
+      "step": 7837
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8413429257923315,
+      "learning_rate": 2.568797808079432e-07,
+      "loss": 0.4593,
+      "step": 7838
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2145572781900977,
+      "learning_rate": 2.56291306397477e-07,
+      "loss": 0.5592,
+      "step": 7839
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.527526104364451,
+      "learning_rate": 2.5570348908471653e-07,
+      "loss": 0.4337,
+      "step": 7840
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.208597016000098,
+      "learning_rate": 2.551163289510877e-07,
+      "loss": 0.4758,
+      "step": 7841
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.305936529770572,
+      "learning_rate": 2.5452982607792274e-07,
+      "loss": 0.4881,
+      "step": 7842
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7894452178924711,
+      "learning_rate": 2.5394398054646494e-07,
+      "loss": 0.44,
+      "step": 7843
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.959332415233702,
+      "learning_rate": 2.533587924378644e-07,
+      "loss": 0.513,
+      "step": 7844
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1575747688850315,
+      "learning_rate": 2.527742618331819e-07,
+      "loss": 0.563,
+      "step": 7845
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.785407727450857,
+      "learning_rate": 2.5219038881338643e-07,
+      "loss": 0.4688,
+      "step": 7846
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.16238964544465,
+      "learning_rate": 2.5160717345935616e-07,
+      "loss": 0.4166,
+      "step": 7847
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.056138899657702,
+      "learning_rate": 2.5102461585187696e-07,
+      "loss": 0.4324,
+      "step": 7848
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4313422238719458,
+      "learning_rate": 2.504427160716466e-07,
+      "loss": 0.6332,
+      "step": 7849
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.180746989358462,
+      "learning_rate": 2.498614741992683e-07,
+      "loss": 0.4537,
+      "step": 7850
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6698701662664577,
+      "learning_rate": 2.4928089031525605e-07,
+      "loss": 0.3459,
+      "step": 7851
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9421465948987424,
+      "learning_rate": 2.48700964500031e-07,
+      "loss": 0.4065,
+      "step": 7852
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8940166371692349,
+      "learning_rate": 2.4812169683392616e-07,
+      "loss": 0.4605,
+      "step": 7853
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2815199474804584,
+      "learning_rate": 2.4754308739718013e-07,
+      "loss": 0.5149,
+      "step": 7854
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.227105372137781,
+      "learning_rate": 2.4696513626994324e-07,
+      "loss": 0.5106,
+      "step": 7855
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.044729410403832,
+      "learning_rate": 2.463878435322725e-07,
+      "loss": 0.4892,
+      "step": 7856
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2222546792805877,
+      "learning_rate": 2.458112092641335e-07,
+      "loss": 0.4331,
+      "step": 7857
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.669096366875629,
+      "learning_rate": 2.4523523354540336e-07,
+      "loss": 0.5399,
+      "step": 7858
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9622700239994508,
+      "learning_rate": 2.4465991645586385e-07,
+      "loss": 0.4898,
+      "step": 7859
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5205071184664154,
+      "learning_rate": 2.440852580752101e-07,
+      "loss": 0.4699,
+      "step": 7860
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8207304980007373,
+      "learning_rate": 2.435112584830418e-07,
+      "loss": 0.5299,
+      "step": 7861
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7393924580442095,
+      "learning_rate": 2.429379177588709e-07,
+      "loss": 0.3485,
+      "step": 7862
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8259391689984361,
+      "learning_rate": 2.423652359821155e-07,
+      "loss": 0.4758,
+      "step": 7863
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.848383333352061,
+      "learning_rate": 2.417932132321038e-07,
+      "loss": 0.4283,
+      "step": 7864
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8182018735023306,
+      "learning_rate": 2.412218495880714e-07,
+      "loss": 0.6072,
+      "step": 7865
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9272712883611105,
+      "learning_rate": 2.406511451291643e-07,
+      "loss": 0.5115,
+      "step": 7866
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6451377115610355,
+      "learning_rate": 2.400810999344361e-07,
+      "loss": 0.4589,
+      "step": 7867
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.082948159586117,
+      "learning_rate": 2.3951171408285123e-07,
+      "loss": 0.5206,
+      "step": 7868
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4741513171645533,
+      "learning_rate": 2.3894298765327726e-07,
+      "loss": 0.6061,
+      "step": 7869
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.813540857185806,
+      "learning_rate": 2.3837492072449676e-07,
+      "loss": 0.4257,
+      "step": 7870
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.811492498915913,
+      "learning_rate": 2.378075133751967e-07,
+      "loss": 0.5986,
+      "step": 7871
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.052362304216388,
+      "learning_rate": 2.3724076568397592e-07,
+      "loss": 0.5656,
+      "step": 7872
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8146237474627175,
+      "learning_rate": 2.3667467772933884e-07,
+      "loss": 0.6639,
+      "step": 7873
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1638717006587993,
+      "learning_rate": 2.3610924958970105e-07,
+      "loss": 0.5056,
+      "step": 7874
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6855076845887367,
+      "learning_rate": 2.3554448134338436e-07,
+      "loss": 0.4699,
+      "step": 7875
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.986137450591559,
+      "learning_rate": 2.3498037306862066e-07,
+      "loss": 0.5402,
+      "step": 7876
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6888609446441045,
+      "learning_rate": 2.3441692484355073e-07,
+      "loss": 0.4808,
+      "step": 7877
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.16569576641359,
+      "learning_rate": 2.338541367462227e-07,
+      "loss": 0.5756,
+      "step": 7878
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.359891302067057,
+      "learning_rate": 2.3329200885459425e-07,
+      "loss": 0.4404,
+      "step": 7879
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1358706472129474,
+      "learning_rate": 2.3273054124653082e-07,
+      "loss": 0.3898,
+      "step": 7880
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2483282206129926,
+      "learning_rate": 2.3216973399980802e-07,
+      "loss": 0.4378,
+      "step": 7881
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5440668978290772,
+      "learning_rate": 2.3160958719210647e-07,
+      "loss": 0.3826,
+      "step": 7882
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.288074018557181,
+      "learning_rate": 2.310501009010202e-07,
+      "loss": 0.6138,
+      "step": 7883
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3860820512455287,
+      "learning_rate": 2.3049127520404723e-07,
+      "loss": 0.4967,
+      "step": 7884
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.4394393527133116,
+      "learning_rate": 2.2993311017859788e-07,
+      "loss": 0.4829,
+      "step": 7885
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4818134290848963,
+      "learning_rate": 2.2937560590198638e-07,
+      "loss": 0.463,
+      "step": 7886
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.965274911720076,
+      "learning_rate": 2.2881876245144097e-07,
+      "loss": 0.5334,
+      "step": 7887
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.456860246091764,
+      "learning_rate": 2.2826257990409273e-07,
+      "loss": 0.6101,
+      "step": 7888
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.784972688206685,
+      "learning_rate": 2.277070583369867e-07,
+      "loss": 0.5243,
+      "step": 7889
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8551967813119554,
+      "learning_rate": 2.2715219782707131e-07,
+      "loss": 0.4489,
+      "step": 7890
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.835141588016095,
+      "learning_rate": 2.2659799845120788e-07,
+      "loss": 0.6437,
+      "step": 7891
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.014836539433783,
+      "learning_rate": 2.2604446028616223e-07,
+      "loss": 0.4565,
+      "step": 7892
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0700200633516395,
+      "learning_rate": 2.2549158340861133e-07,
+      "loss": 0.4155,
+      "step": 7893
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1585587435393943,
+      "learning_rate": 2.2493936789513892e-07,
+      "loss": 0.4611,
+      "step": 7894
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.039401459473966,
+      "learning_rate": 2.2438781382223883e-07,
+      "loss": 0.5237,
+      "step": 7895
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5451554989707046,
+      "learning_rate": 2.2383692126631163e-07,
+      "loss": 0.5172,
+      "step": 7896
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8845787032282753,
+      "learning_rate": 2.2328669030366623e-07,
+      "loss": 0.5119,
+      "step": 7897
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.308412960245356,
+      "learning_rate": 2.227371210105217e-07,
+      "loss": 0.4585,
+      "step": 7898
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6656681667150586,
+      "learning_rate": 2.2218821346300267e-07,
+      "loss": 0.4303,
+      "step": 7899
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6855180751074736,
+      "learning_rate": 2.2163996773714612e-07,
+      "loss": 0.5139,
+      "step": 7900
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9463451324311563,
+      "learning_rate": 2.2109238390889242e-07,
+      "loss": 0.5371,
+      "step": 7901
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1599190921216116,
+      "learning_rate": 2.205454620540959e-07,
+      "loss": 0.5614,
+      "step": 7902
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.034484805269269,
+      "learning_rate": 2.19999202248512e-07,
+      "loss": 0.511,
+      "step": 7903
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7256945735437206,
+      "learning_rate": 2.1945360456781194e-07,
+      "loss": 0.4428,
+      "step": 7904
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1424955419217793,
+      "learning_rate": 2.189086690875697e-07,
+      "loss": 0.4546,
+      "step": 7905
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.808120872490734,
+      "learning_rate": 2.1836439588327152e-07,
+      "loss": 0.5006,
+      "step": 7906
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.99715505987209,
+      "learning_rate": 2.1782078503030768e-07,
+      "loss": 0.5577,
+      "step": 7907
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.196377525071023,
+      "learning_rate": 2.1727783660398183e-07,
+      "loss": 0.4787,
+      "step": 7908
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2807307949400792,
+      "learning_rate": 2.1673555067950047e-07,
+      "loss": 0.444,
+      "step": 7909
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7897977102757918,
+      "learning_rate": 2.1619392733198298e-07,
+      "loss": 0.3901,
+      "step": 7910
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9262104435629732,
+      "learning_rate": 2.1565296663645319e-07,
+      "loss": 0.6008,
+      "step": 7911
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7646105100648166,
+      "learning_rate": 2.1511266866784674e-07,
+      "loss": 0.5516,
+      "step": 7912
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0243533104616436,
+      "learning_rate": 2.1457303350100377e-07,
+      "loss": 0.4357,
+      "step": 7913
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4053809341574377,
+      "learning_rate": 2.1403406121067616e-07,
+      "loss": 0.4965,
+      "step": 7914
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4104894606764375,
+      "learning_rate": 2.1349575187152138e-07,
+      "loss": 0.4795,
+      "step": 7915
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.290855080983749,
+      "learning_rate": 2.1295810555810535e-07,
+      "loss": 0.5292,
+      "step": 7916
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1069852773780116,
+      "learning_rate": 2.1242112234490407e-07,
+      "loss": 0.5557,
+      "step": 7917
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0555822051899986,
+      "learning_rate": 2.118848023062997e-07,
+      "loss": 0.4313,
+      "step": 7918
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.08521513651175,
+      "learning_rate": 2.113491455165828e-07,
+      "loss": 0.4186,
+      "step": 7919
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7251146126476526,
+      "learning_rate": 2.1081415204995291e-07,
+      "loss": 0.4792,
+      "step": 7920
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7243842526200532,
+      "learning_rate": 2.1027982198051744e-07,
+      "loss": 0.4471,
+      "step": 7921
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9857146366430551,
+      "learning_rate": 2.0974615538229105e-07,
+      "loss": 0.381,
+      "step": 7922
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.0762951873324385,
+      "learning_rate": 2.0921315232919793e-07,
+      "loss": 0.5641,
+      "step": 7923
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0703247753060747,
+      "learning_rate": 2.0868081289506847e-07,
+      "loss": 0.4934,
+      "step": 7924
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9450966598596273,
+      "learning_rate": 2.0814913715364372e-07,
+      "loss": 0.4501,
+      "step": 7925
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7332638109404221,
+      "learning_rate": 2.0761812517856972e-07,
+      "loss": 0.5285,
+      "step": 7926
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.25691588406782,
+      "learning_rate": 2.0708777704340376e-07,
+      "loss": 0.4712,
+      "step": 7927
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5183246050024986,
+      "learning_rate": 2.0655809282160767e-07,
+      "loss": 0.5181,
+      "step": 7928
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.889796052070294,
+      "learning_rate": 2.0602907258655546e-07,
+      "loss": 0.7208,
+      "step": 7929
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3634521607194836,
+      "learning_rate": 2.0550071641152525e-07,
+      "loss": 0.5143,
+      "step": 7930
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0103287007531963,
+      "learning_rate": 2.049730243697057e-07,
+      "loss": 0.4362,
+      "step": 7931
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.063801226644467,
+      "learning_rate": 2.0444599653419161e-07,
+      "loss": 0.4527,
+      "step": 7932
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1762019662621235,
+      "learning_rate": 2.03919632977988e-07,
+      "loss": 0.4099,
+      "step": 7933
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9649189007307415,
+      "learning_rate": 2.033939337740065e-07,
+      "loss": 0.4776,
+      "step": 7934
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.231257507852725,
+      "learning_rate": 2.0286889899506613e-07,
+      "loss": 0.4694,
+      "step": 7935
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.022821007824068,
+      "learning_rate": 2.0234452871389476e-07,
+      "loss": 0.4836,
+      "step": 7936
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8361961659004296,
+      "learning_rate": 2.0182082300312877e-07,
+      "loss": 0.53,
+      "step": 7937
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.938551377997809,
+      "learning_rate": 2.0129778193531015e-07,
+      "loss": 0.3781,
+      "step": 7938
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.743316207098689,
+      "learning_rate": 2.007754055828931e-07,
+      "loss": 0.497,
+      "step": 7939
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9919900875456569,
+      "learning_rate": 2.002536940182348e-07,
+      "loss": 0.4792,
+      "step": 7940
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9104387245533985,
+      "learning_rate": 1.9973264731360354e-07,
+      "loss": 0.5489,
+      "step": 7941
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9562856943574627,
+      "learning_rate": 1.992122655411749e-07,
+      "loss": 0.4595,
+      "step": 7942
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.041512491814604,
+      "learning_rate": 1.986925487730307e-07,
+      "loss": 0.5324,
+      "step": 7943
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3490432269975887,
+      "learning_rate": 1.981734970811644e-07,
+      "loss": 0.5197,
+      "step": 7944
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.334460729098768,
+      "learning_rate": 1.9765511053747243e-07,
+      "loss": 0.4413,
+      "step": 7945
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.194681099494441,
+      "learning_rate": 1.9713738921376346e-07,
+      "loss": 0.5379,
+      "step": 7946
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.608421774278065,
+      "learning_rate": 1.9662033318175068e-07,
+      "loss": 0.4535,
+      "step": 7947
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.307409505118636,
+      "learning_rate": 1.961039425130584e-07,
+      "loss": 0.5116,
+      "step": 7948
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7055474374068202,
+      "learning_rate": 1.9558821727921508e-07,
+      "loss": 0.5067,
+      "step": 7949
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9317621673654353,
+      "learning_rate": 1.9507315755166068e-07,
+      "loss": 0.4693,
+      "step": 7950
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.024154011725036,
+      "learning_rate": 1.9455876340173929e-07,
+      "loss": 0.4662,
+      "step": 7951
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2753951416791844,
+      "learning_rate": 1.940450349007067e-07,
+      "loss": 0.4417,
+      "step": 7952
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2203302839309846,
+      "learning_rate": 1.9353197211972262e-07,
+      "loss": 0.5659,
+      "step": 7953
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2563893580384287,
+      "learning_rate": 1.9301957512985802e-07,
+      "loss": 0.5505,
+      "step": 7954
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8760583035835703,
+      "learning_rate": 1.9250784400208832e-07,
+      "loss": 0.4464,
+      "step": 7955
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9990966454915617,
+      "learning_rate": 1.9199677880730018e-07,
+      "loss": 0.6996,
+      "step": 7956
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9237599776195486,
+      "learning_rate": 1.9148637961628468e-07,
+      "loss": 0.6038,
+      "step": 7957
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6371339210292004,
+      "learning_rate": 1.9097664649974368e-07,
+      "loss": 0.4466,
+      "step": 7958
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7662190798850002,
+      "learning_rate": 1.9046757952828398e-07,
+      "loss": 0.4156,
+      "step": 7959
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3338822562752646,
+      "learning_rate": 1.8995917877242308e-07,
+      "loss": 0.4625,
+      "step": 7960
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2647587853794446,
+      "learning_rate": 1.8945144430258356e-07,
+      "loss": 0.4832,
+      "step": 7961
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.5206310993709504,
+      "learning_rate": 1.8894437618909578e-07,
+      "loss": 0.4999,
+      "step": 7962
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9137003315175436,
+      "learning_rate": 1.8843797450220024e-07,
+      "loss": 0.4108,
+      "step": 7963
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9533683001233906,
+      "learning_rate": 1.8793223931204308e-07,
+      "loss": 0.4483,
+      "step": 7964
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.188862087683525,
+      "learning_rate": 1.8742717068867877e-07,
+      "loss": 0.3995,
+      "step": 7965
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9665401382357417,
+      "learning_rate": 1.869227687020686e-07,
+      "loss": 0.4365,
+      "step": 7966
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3143608259203163,
+      "learning_rate": 1.8641903342208389e-07,
+      "loss": 0.5685,
+      "step": 7967
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9153070641244236,
+      "learning_rate": 1.8591596491849996e-07,
+      "loss": 0.4483,
+      "step": 7968
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.005173668831575,
+      "learning_rate": 1.8541356326100436e-07,
+      "loss": 0.5629,
+      "step": 7969
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.292940574260069,
+      "learning_rate": 1.8491182851918643e-07,
+      "loss": 0.4789,
+      "step": 7970
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.078937021704676,
+      "learning_rate": 1.8441076076254837e-07,
+      "loss": 0.574,
+      "step": 7971
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3798440783072388,
+      "learning_rate": 1.8391036006049744e-07,
+      "loss": 0.5465,
+      "step": 7972
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.833638060508819,
+      "learning_rate": 1.8341062648234987e-07,
+      "loss": 0.5143,
+      "step": 7973
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.191557370823667,
+      "learning_rate": 1.8291156009732746e-07,
+      "loss": 0.5249,
+      "step": 7974
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8036034276767148,
+      "learning_rate": 1.8241316097456218e-07,
+      "loss": 0.5334,
+      "step": 7975
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.155544703329232,
+      "learning_rate": 1.8191542918309045e-07,
+      "loss": 0.3753,
+      "step": 7976
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.004205508804817,
+      "learning_rate": 1.8141836479185993e-07,
+      "loss": 0.4891,
+      "step": 7977
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1882546701008163,
+      "learning_rate": 1.8092196786972215e-07,
+      "loss": 0.4722,
+      "step": 7978
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.996178721859881,
+      "learning_rate": 1.804262384854394e-07,
+      "loss": 0.4469,
+      "step": 7979
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7781120572922013,
+      "learning_rate": 1.799311767076789e-07,
+      "loss": 0.5628,
+      "step": 7980
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7522065403533245,
+      "learning_rate": 1.7943678260501641e-07,
+      "loss": 0.441,
+      "step": 7981
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0651646883159995,
+      "learning_rate": 1.7894305624593655e-07,
+      "loss": 0.5002,
+      "step": 7982
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.568981008222588,
+      "learning_rate": 1.784499976988291e-07,
+      "loss": 0.5103,
+      "step": 7983
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7278617113090513,
+      "learning_rate": 1.7795760703199327e-07,
+      "loss": 0.5113,
+      "step": 7984
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.472571841758644,
+      "learning_rate": 1.7746588431363397e-07,
+      "loss": 0.4739,
+      "step": 7985
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5581462119885905,
+      "learning_rate": 1.7697482961186674e-07,
+      "loss": 0.5417,
+      "step": 7986
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8780913172927785,
+      "learning_rate": 1.764844429947088e-07,
+      "loss": 0.531,
+      "step": 7987
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7570848361516738,
+      "learning_rate": 1.7599472453009136e-07,
+      "loss": 0.4832,
+      "step": 7988
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7991695453187564,
+      "learning_rate": 1.7550567428584852e-07,
+      "loss": 0.4902,
+      "step": 7989
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9105296270141725,
+      "learning_rate": 1.7501729232972442e-07,
+      "loss": 0.4739,
+      "step": 7990
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.824306785711125,
+      "learning_rate": 1.7452957872936881e-07,
+      "loss": 0.3892,
+      "step": 7991
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.166072161654502,
+      "learning_rate": 1.7404253355234102e-07,
+      "loss": 0.5336,
+      "step": 7992
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.247264251826484,
+      "learning_rate": 1.7355615686610427e-07,
+      "loss": 0.4501,
+      "step": 7993
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.267710077394332,
+      "learning_rate": 1.7307044873803414e-07,
+      "loss": 0.4626,
+      "step": 7994
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.076295760548668,
+      "learning_rate": 1.7258540923540846e-07,
+      "loss": 0.5351,
+      "step": 7995
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9948699624914832,
+      "learning_rate": 1.7210103842541626e-07,
+      "loss": 0.4632,
+      "step": 7996
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9105770342424637,
+      "learning_rate": 1.7161733637515166e-07,
+      "loss": 0.5645,
+      "step": 7997
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.287209472860106,
+      "learning_rate": 1.7113430315161772e-07,
+      "loss": 0.5097,
+      "step": 7998
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.957730057335225,
+      "learning_rate": 1.706519388217237e-07,
+      "loss": 0.4966,
+      "step": 7999
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.896980782921911,
+      "learning_rate": 1.701702434522856e-07,
+      "loss": 0.5705,
+      "step": 8000
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3015932772251526,
+      "learning_rate": 1.6968921711003005e-07,
+      "loss": 0.4411,
+      "step": 8001
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.687678852196777,
+      "learning_rate": 1.6920885986158707e-07,
+      "loss": 0.4408,
+      "step": 8002
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.78801883923739,
+      "learning_rate": 1.6872917177349623e-07,
+      "loss": 0.5845,
+      "step": 8003
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9501255451639143,
+      "learning_rate": 1.682501529122038e-07,
+      "loss": 0.45,
+      "step": 8004
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1386630897533303,
+      "learning_rate": 1.677718033440634e-07,
+      "loss": 0.4636,
+      "step": 8005
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.001747229819958,
+      "learning_rate": 1.6729412313533534e-07,
+      "loss": 0.4446,
+      "step": 8006
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.028711579311054,
+      "learning_rate": 1.668171123521889e-07,
+      "loss": 0.5,
+      "step": 8007
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9993125420999664,
+      "learning_rate": 1.6634077106069791e-07,
+      "loss": 0.4531,
+      "step": 8008
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.061783550149706,
+      "learning_rate": 1.6586509932684735e-07,
+      "loss": 0.466,
+      "step": 8009
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.792411794727479,
+      "learning_rate": 1.6539009721652455e-07,
+      "loss": 0.4588,
+      "step": 8010
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.105670370370146,
+      "learning_rate": 1.6491576479552962e-07,
+      "loss": 0.5492,
+      "step": 8011
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.179724434910707,
+      "learning_rate": 1.6444210212956392e-07,
+      "loss": 0.3773,
+      "step": 8012
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6165921712600508,
+      "learning_rate": 1.6396910928424216e-07,
+      "loss": 0.401,
+      "step": 8013
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5996996434133965,
+      "learning_rate": 1.634967863250808e-07,
+      "loss": 0.4079,
+      "step": 8014
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3590043042466773,
+      "learning_rate": 1.6302513331750702e-07,
+      "loss": 0.3117,
+      "step": 8015
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.958467029029524,
+      "learning_rate": 1.6255415032685406e-07,
+      "loss": 0.4985,
+      "step": 8016
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9379479335531633,
+      "learning_rate": 1.6208383741836254e-07,
+      "loss": 0.5138,
+      "step": 8017
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9022610537063376,
+      "learning_rate": 1.6161419465717975e-07,
+      "loss": 0.4968,
+      "step": 8018
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5953016404846974,
+      "learning_rate": 1.6114522210836091e-07,
+      "loss": 0.5201,
+      "step": 8019
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.965895843505926,
+      "learning_rate": 1.6067691983686794e-07,
+      "loss": 0.4837,
+      "step": 8020
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0828353285766883,
+      "learning_rate": 1.602092879075695e-07,
+      "loss": 0.43,
+      "step": 8021
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.901578612517592,
+      "learning_rate": 1.5974232638524212e-07,
+      "loss": 0.4287,
+      "step": 8022
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5527459468832463,
+      "learning_rate": 1.592760353345696e-07,
+      "loss": 0.43,
+      "step": 8023
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8200867181013363,
+      "learning_rate": 1.5881041482014192e-07,
+      "loss": 0.4889,
+      "step": 8024
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0920180572405096,
+      "learning_rate": 1.5834546490645696e-07,
+      "loss": 0.4931,
+      "step": 8025
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.255213561270281,
+      "learning_rate": 1.5788118565792042e-07,
+      "loss": 0.4483,
+      "step": 8026
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.0911784018832433,
+      "learning_rate": 1.5741757713884253e-07,
+      "loss": 0.4902,
+      "step": 8027
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7819304773351499,
+      "learning_rate": 1.569546394134436e-07,
+      "loss": 0.4614,
+      "step": 8028
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0487995875711533,
+      "learning_rate": 1.5649237254584838e-07,
+      "loss": 0.5201,
+      "step": 8029
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2793007325278176,
+      "learning_rate": 1.560307766000918e-07,
+      "loss": 0.5048,
+      "step": 8030
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.027387421472596,
+      "learning_rate": 1.5556985164011217e-07,
+      "loss": 0.5009,
+      "step": 8031
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9782535662943772,
+      "learning_rate": 1.5510959772975841e-07,
+      "loss": 0.4431,
+      "step": 8032
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7835028566349568,
+      "learning_rate": 1.546500149327834e-07,
+      "loss": 0.645,
+      "step": 8033
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9547264074324655,
+      "learning_rate": 1.5419110331284904e-07,
+      "loss": 0.5546,
+      "step": 8034
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5281659046810456,
+      "learning_rate": 1.5373286293352385e-07,
+      "loss": 0.4084,
+      "step": 8035
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9748924234422105,
+      "learning_rate": 1.5327529385828377e-07,
+      "loss": 0.4805,
+      "step": 8036
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.533086763430095,
+      "learning_rate": 1.5281839615050975e-07,
+      "loss": 0.3945,
+      "step": 8037
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1843738059619233,
+      "learning_rate": 1.5236216987349283e-07,
+      "loss": 0.5754,
+      "step": 8038
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.929045406871986,
+      "learning_rate": 1.5190661509042748e-07,
+      "loss": 0.5093,
+      "step": 8039
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0385363491766872,
+      "learning_rate": 1.5145173186441875e-07,
+      "loss": 0.4116,
+      "step": 8040
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6040494668183753,
+      "learning_rate": 1.509975202584757e-07,
+      "loss": 0.4406,
+      "step": 8041
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6057678082783395,
+      "learning_rate": 1.5054398033551688e-07,
+      "loss": 0.4755,
+      "step": 8042
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.911946231688503,
+      "learning_rate": 1.5009111215836536e-07,
+      "loss": 0.4923,
+      "step": 8043
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.079354177191774,
+      "learning_rate": 1.496389157897532e-07,
+      "loss": 0.5125,
+      "step": 8044
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.011808963488335,
+      "learning_rate": 1.4918739129231863e-07,
+      "loss": 0.5339,
+      "step": 8045
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2023340088146117,
+      "learning_rate": 1.4873653872860605e-07,
+      "loss": 0.4644,
+      "step": 8046
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.988131415038431,
+      "learning_rate": 1.482863581610683e-07,
+      "loss": 0.4792,
+      "step": 8047
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.708896182831012,
+      "learning_rate": 1.4783684965206323e-07,
+      "loss": 0.5109,
+      "step": 8048
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.145691622547765,
+      "learning_rate": 1.4738801326385777e-07,
+      "loss": 0.4751,
+      "step": 8049
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.436820363411749,
+      "learning_rate": 1.4693984905862378e-07,
+      "loss": 0.4558,
+      "step": 8050
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9115897861693334,
+      "learning_rate": 1.4649235709844168e-07,
+      "loss": 0.3913,
+      "step": 8051
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.253423538052999,
+      "learning_rate": 1.4604553744529737e-07,
+      "loss": 0.5442,
+      "step": 8052
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.3090020171053345,
+      "learning_rate": 1.4559939016108472e-07,
+      "loss": 0.3653,
+      "step": 8053
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7298989112674725,
+      "learning_rate": 1.4515391530760426e-07,
+      "loss": 0.5342,
+      "step": 8054
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.057603114024709,
+      "learning_rate": 1.4470911294656222e-07,
+      "loss": 0.5192,
+      "step": 8055
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5248043766875514,
+      "learning_rate": 1.442649831395726e-07,
+      "loss": 0.5517,
+      "step": 8056
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.284828305335089,
+      "learning_rate": 1.438215259481568e-07,
+      "loss": 0.4144,
+      "step": 8057
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8905914957021667,
+      "learning_rate": 1.433787414337412e-07,
+      "loss": 0.6678,
+      "step": 8058
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.71499600710676,
+      "learning_rate": 1.429366296576623e-07,
+      "loss": 0.5011,
+      "step": 8059
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0756106205858575,
+      "learning_rate": 1.4249519068115947e-07,
+      "loss": 0.5492,
+      "step": 8060
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.208101659783904,
+      "learning_rate": 1.420544245653821e-07,
+      "loss": 0.5102,
+      "step": 8061
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4389428200649936,
+      "learning_rate": 1.4161433137138358e-07,
+      "loss": 0.4644,
+      "step": 8062
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3045402830990773,
+      "learning_rate": 1.4117491116012683e-07,
+      "loss": 0.479,
+      "step": 8063
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9541189830213517,
+      "learning_rate": 1.4073616399248037e-07,
+      "loss": 0.42,
+      "step": 8064
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6130659098449072,
+      "learning_rate": 1.4029808992921778e-07,
+      "loss": 0.4834,
+      "step": 8065
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.147447428008245,
+      "learning_rate": 1.398606890310228e-07,
+      "loss": 0.4786,
+      "step": 8066
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3567534803800236,
+      "learning_rate": 1.3942396135848301e-07,
+      "loss": 0.5011,
+      "step": 8067
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.338880318588386,
+      "learning_rate": 1.3898790697209453e-07,
+      "loss": 0.5872,
+      "step": 8068
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.581096956853325,
+      "learning_rate": 1.3855252593225842e-07,
+      "loss": 0.3963,
+      "step": 8069
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.770299562492781,
+      "learning_rate": 1.3811781829928593e-07,
+      "loss": 0.5404,
+      "step": 8070
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9217966646673075,
+      "learning_rate": 1.3768378413339e-07,
+      "loss": 0.5116,
+      "step": 8071
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8189868338312154,
+      "learning_rate": 1.372504234946942e-07,
+      "loss": 0.4853,
+      "step": 8072
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9162263331461113,
+      "learning_rate": 1.3681773644322772e-07,
+      "loss": 0.5646,
+      "step": 8073
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.50214157945727,
+      "learning_rate": 1.3638572303892594e-07,
+      "loss": 0.4916,
+      "step": 8074
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0582244282782653,
+      "learning_rate": 1.3595438334163103e-07,
+      "loss": 0.4577,
+      "step": 8075
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8823412127574939,
+      "learning_rate": 1.355237174110935e-07,
+      "loss": 0.6555,
+      "step": 8076
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7736473569464835,
+      "learning_rate": 1.3509372530696674e-07,
+      "loss": 0.5195,
+      "step": 8077
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.328138835937736,
+      "learning_rate": 1.3466440708881534e-07,
+      "loss": 0.5302,
+      "step": 8078
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.225383681527746,
+      "learning_rate": 1.342357628161073e-07,
+      "loss": 0.417,
+      "step": 8079
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1613858444869454,
+      "learning_rate": 1.3380779254821896e-07,
+      "loss": 0.4258,
+      "step": 8080
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.587411829219413,
+      "learning_rate": 1.3338049634443183e-07,
+      "loss": 0.5498,
+      "step": 8081
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3632206368946655,
+      "learning_rate": 1.329538742639358e-07,
+      "loss": 0.4379,
+      "step": 8082
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.490719239649113,
+      "learning_rate": 1.3252792636582578e-07,
+      "loss": 0.4835,
+      "step": 8083
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9899899202423035,
+      "learning_rate": 1.3210265270910516e-07,
+      "loss": 0.4994,
+      "step": 8084
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8264529311252422,
+      "learning_rate": 1.3167805335268126e-07,
+      "loss": 0.5148,
+      "step": 8085
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6789998294119555,
+      "learning_rate": 1.3125412835537032e-07,
+      "loss": 0.556,
+      "step": 8086
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.4764406271377695,
+      "learning_rate": 1.3083087777589432e-07,
+      "loss": 0.532,
+      "step": 8087
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.312818347790458,
+      "learning_rate": 1.3040830167288188e-07,
+      "loss": 0.5922,
+      "step": 8088
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3705055475791075,
+      "learning_rate": 1.299864001048673e-07,
+      "loss": 0.4495,
+      "step": 8089
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6649292996569292,
+      "learning_rate": 1.295651731302938e-07,
+      "loss": 0.4586,
+      "step": 8090
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2257675478148604,
+      "learning_rate": 1.2914462080750923e-07,
+      "loss": 0.4869,
+      "step": 8091
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4165760911904846,
+      "learning_rate": 1.2872474319476747e-07,
+      "loss": 0.4218,
+      "step": 8092
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2242998201781967,
+      "learning_rate": 1.283055403502309e-07,
+      "loss": 0.4685,
+      "step": 8093
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0898630633447604,
+      "learning_rate": 1.2788701233196643e-07,
+      "loss": 0.5471,
+      "step": 8094
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2808461757542458,
+      "learning_rate": 1.274691591979499e-07,
+      "loss": 0.443,
+      "step": 8095
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.078243897977429,
+      "learning_rate": 1.2705198100606052e-07,
+      "loss": 0.4613,
+      "step": 8096
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3736714805211085,
+      "learning_rate": 1.2663547781408769e-07,
+      "loss": 0.3855,
+      "step": 8097
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0280938638639836,
+      "learning_rate": 1.262196496797241e-07,
+      "loss": 0.5124,
+      "step": 8098
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7620594821719144,
+      "learning_rate": 1.2580449666057038e-07,
+      "loss": 0.4486,
+      "step": 8099
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6836999326179645,
+      "learning_rate": 1.2539001881413326e-07,
+      "loss": 0.4806,
+      "step": 8100
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.16261666007082,
+      "learning_rate": 1.2497621619782686e-07,
+      "loss": 0.4673,
+      "step": 8101
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.128580757361907,
+      "learning_rate": 1.2456308886897028e-07,
+      "loss": 0.4954,
+      "step": 8102
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4660150799955782,
+      "learning_rate": 1.2415063688479057e-07,
+      "loss": 0.5254,
+      "step": 8103
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.8492569745892475,
+      "learning_rate": 1.2373886030242032e-07,
+      "loss": 0.4658,
+      "step": 8104
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4469744950134027,
+      "learning_rate": 1.233277591788984e-07,
+      "loss": 0.428,
+      "step": 8105
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0688889668412185,
+      "learning_rate": 1.229173335711703e-07,
+      "loss": 0.4191,
+      "step": 8106
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6850460293375198,
+      "learning_rate": 1.225075835360884e-07,
+      "loss": 0.4412,
+      "step": 8107
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.027183857434624,
+      "learning_rate": 1.220985091304111e-07,
+      "loss": 0.494,
+      "step": 8108
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3193272264515987,
+      "learning_rate": 1.2169011041080426e-07,
+      "loss": 0.5424,
+      "step": 8109
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.948700376686977,
+      "learning_rate": 1.2128238743383758e-07,
+      "loss": 0.526,
+      "step": 8110
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8783381119927861,
+      "learning_rate": 1.2087534025598979e-07,
+      "loss": 0.6496,
+      "step": 8111
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9264507623716343,
+      "learning_rate": 1.2046896893364467e-07,
+      "loss": 0.3456,
+      "step": 8112
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7001400029963276,
+      "learning_rate": 1.2006327352309276e-07,
+      "loss": 0.5803,
+      "step": 8113
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.071528182286443,
+      "learning_rate": 1.1965825408053133e-07,
+      "loss": 0.4143,
+      "step": 8114
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.82997352598845,
+      "learning_rate": 1.1925391066206272e-07,
+      "loss": 0.4194,
+      "step": 8115
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8583410944710604,
+      "learning_rate": 1.1885024332369765e-07,
+      "loss": 0.6604,
+      "step": 8116
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7939437646106957,
+      "learning_rate": 1.1844725212135089e-07,
+      "loss": 0.5033,
+      "step": 8117
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1610446836265407,
+      "learning_rate": 1.1804493711084553e-07,
+      "loss": 0.5463,
+      "step": 8118
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6273086423574008,
+      "learning_rate": 1.1764329834790977e-07,
+      "loss": 0.5022,
+      "step": 8119
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.128296965963352,
+      "learning_rate": 1.1724233588817913e-07,
+      "loss": 0.489,
+      "step": 8120
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.439341377778607,
+      "learning_rate": 1.1684204978719416e-07,
+      "loss": 0.4551,
+      "step": 8121
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9258077805869753,
+      "learning_rate": 1.1644244010040273e-07,
+      "loss": 0.4802,
+      "step": 8122
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.041682990701367,
+      "learning_rate": 1.1604350688315836e-07,
+      "loss": 0.4408,
+      "step": 8123
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1558770680714976,
+      "learning_rate": 1.1564525019072181e-07,
+      "loss": 0.4288,
+      "step": 8124
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7986136346332064,
+      "learning_rate": 1.1524767007825843e-07,
+      "loss": 0.5222,
+      "step": 8125
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2118192927416396,
+      "learning_rate": 1.1485076660084249e-07,
+      "loss": 0.5244,
+      "step": 8126
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.576924720495921,
+      "learning_rate": 1.1445453981345167e-07,
+      "loss": 0.5054,
+      "step": 8127
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.4055570032563813,
+      "learning_rate": 1.1405898977097263e-07,
+      "loss": 0.5293,
+      "step": 8128
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1388486439285113,
+      "learning_rate": 1.1366411652819598e-07,
+      "loss": 0.5156,
+      "step": 8129
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.687929030296724,
+      "learning_rate": 1.1326992013981852e-07,
+      "loss": 0.4119,
+      "step": 8130
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.337838203617382,
+      "learning_rate": 1.1287640066044658e-07,
+      "loss": 0.4676,
+      "step": 8131
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7215105292591768,
+      "learning_rate": 1.1248355814458822e-07,
+      "loss": 0.5422,
+      "step": 8132
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.241602321479729,
+      "learning_rate": 1.1209139264666102e-07,
+      "loss": 0.4502,
+      "step": 8133
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.20939546163229,
+      "learning_rate": 1.1169990422098764e-07,
+      "loss": 0.5408,
+      "step": 8134
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4460914610358757,
+      "learning_rate": 1.1130909292179693e-07,
+      "loss": 0.4632,
+      "step": 8135
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9707081571418674,
+      "learning_rate": 1.1091895880322334e-07,
+      "loss": 0.4798,
+      "step": 8136
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.149942080882984,
+      "learning_rate": 1.1052950191930978e-07,
+      "loss": 0.4869,
+      "step": 8137
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0004859743975216,
+      "learning_rate": 1.1014072232400196e-07,
+      "loss": 0.4771,
+      "step": 8138
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0617511363898235,
+      "learning_rate": 1.0975262007115516e-07,
+      "loss": 0.494,
+      "step": 8139
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.096215152771438,
+      "learning_rate": 1.0936519521452748e-07,
+      "loss": 0.5197,
+      "step": 8140
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.1247173985396843,
+      "learning_rate": 1.0897844780778654e-07,
+      "loss": 0.5415,
+      "step": 8141
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0902380658116426,
+      "learning_rate": 1.0859237790450284e-07,
+      "loss": 0.4642,
+      "step": 8142
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3101410486832417,
+      "learning_rate": 1.082069855581569e-07,
+      "loss": 0.4522,
+      "step": 8143
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9710199907840242,
+      "learning_rate": 1.0782227082213104e-07,
+      "loss": 0.4392,
+      "step": 8144
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8184119849112796,
+      "learning_rate": 1.0743823374971762e-07,
+      "loss": 0.4452,
+      "step": 8145
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1703026282756555,
+      "learning_rate": 1.0705487439411243e-07,
+      "loss": 0.5118,
+      "step": 8146
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7661611310252745,
+      "learning_rate": 1.0667219280841857e-07,
+      "loss": 0.5821,
+      "step": 8147
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7427371760023855,
+      "learning_rate": 1.062901890456447e-07,
+      "loss": 0.4471,
+      "step": 8148
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8208886403659923,
+      "learning_rate": 1.0590886315870685e-07,
+      "loss": 0.4343,
+      "step": 8149
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 26.994696167644015,
+      "learning_rate": 1.0552821520042556e-07,
+      "loss": 0.4926,
+      "step": 8150
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8665526001866652,
+      "learning_rate": 1.051482452235275e-07,
+      "loss": 0.5555,
+      "step": 8151
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5980064434509742,
+      "learning_rate": 1.0476895328064729e-07,
+      "loss": 0.452,
+      "step": 8152
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4626612061107154,
+      "learning_rate": 1.0439033942432341e-07,
+      "loss": 0.493,
+      "step": 8153
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.008864849450951,
+      "learning_rate": 1.0401240370700228e-07,
+      "loss": 0.5863,
+      "step": 8154
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.10335587358992,
+      "learning_rate": 1.0363514618103477e-07,
+      "loss": 0.4546,
+      "step": 8155
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.307544911497827,
+      "learning_rate": 1.0325856689867853e-07,
+      "loss": 0.4889,
+      "step": 8156
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.192924263589598,
+      "learning_rate": 1.0288266591209738e-07,
+      "loss": 0.4186,
+      "step": 8157
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3445574228265977,
+      "learning_rate": 1.0250744327336138e-07,
+      "loss": 0.5138,
+      "step": 8158
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.706912809284124,
+      "learning_rate": 1.0213289903444557e-07,
+      "loss": 0.5092,
+      "step": 8159
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.820597525748506,
+      "learning_rate": 1.0175903324723291e-07,
+      "loss": 0.3952,
+      "step": 8160
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8781883247540678,
+      "learning_rate": 1.0138584596350976e-07,
+      "loss": 0.5263,
+      "step": 8161
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.769212397174568,
+      "learning_rate": 1.0101333723497198e-07,
+      "loss": 0.4713,
+      "step": 8162
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.706592328887155,
+      "learning_rate": 1.0064150711321718e-07,
+      "loss": 0.4013,
+      "step": 8163
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8304667641849134,
+      "learning_rate": 1.0027035564975252e-07,
+      "loss": 0.6854,
+      "step": 8164
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1183799627668582,
+      "learning_rate": 9.989988289598962e-08,
+      "loss": 0.4873,
+      "step": 8165
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8848187251062787,
+      "learning_rate": 9.953008890324634e-08,
+      "loss": 0.5102,
+      "step": 8166
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0322758501280087,
+      "learning_rate": 9.916097372274558e-08,
+      "loss": 0.5795,
+      "step": 8167
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1706580031313187,
+      "learning_rate": 9.879253740561867e-08,
+      "loss": 0.536,
+      "step": 8168
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0326705961764104,
+      "learning_rate": 9.84247800029009e-08,
+      "loss": 0.5295,
+      "step": 8169
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.032374106842836,
+      "learning_rate": 9.805770156553263e-08,
+      "loss": 0.5612,
+      "step": 8170
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.809415950512553,
+      "learning_rate": 9.769130214436318e-08,
+      "loss": 0.4451,
+      "step": 8171
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5016152767064646,
+      "learning_rate": 9.732558179014528e-08,
+      "loss": 0.4934,
+      "step": 8172
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0134033984882094,
+      "learning_rate": 9.696054055353843e-08,
+      "loss": 0.5059,
+      "step": 8173
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7037241977666748,
+      "learning_rate": 9.659617848510883e-08,
+      "loss": 0.3832,
+      "step": 8174
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.072702739979123,
+      "learning_rate": 9.623249563532666e-08,
+      "loss": 0.483,
+      "step": 8175
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0798594034074487,
+      "learning_rate": 9.586949205456997e-08,
+      "loss": 0.5168,
+      "step": 8176
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6667071021639077,
+      "learning_rate": 9.550716779312242e-08,
+      "loss": 0.5309,
+      "step": 8177
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9226690385661063,
+      "learning_rate": 9.514552290117162e-08,
+      "loss": 0.5141,
+      "step": 8178
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1610775118456247,
+      "learning_rate": 9.478455742881365e-08,
+      "loss": 0.5141,
+      "step": 8179
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5058743972699011,
+      "learning_rate": 9.44242714260496e-08,
+      "loss": 0.487,
+      "step": 8180
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.512104274286153,
+      "learning_rate": 9.406466494278566e-08,
+      "loss": 0.5097,
+      "step": 8181
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.038052809655248,
+      "learning_rate": 9.37057380288342e-08,
+      "loss": 0.6662,
+      "step": 8182
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.8717268926974873,
+      "learning_rate": 9.334749073391436e-08,
+      "loss": 0.4569,
+      "step": 8183
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.849468868144424,
+      "learning_rate": 9.298992310765032e-08,
+      "loss": 0.4643,
+      "step": 8184
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.556247108719828,
+      "learning_rate": 9.26330351995719e-08,
+      "loss": 0.4953,
+      "step": 8185
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9799756523816332,
+      "learning_rate": 9.227682705911567e-08,
+      "loss": 0.5309,
+      "step": 8186
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8543210327092232,
+      "learning_rate": 9.192129873562383e-08,
+      "loss": 0.4953,
+      "step": 8187
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.538752010183699,
+      "learning_rate": 9.156645027834255e-08,
+      "loss": 0.4215,
+      "step": 8188
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7522510004601333,
+      "learning_rate": 9.121228173642749e-08,
+      "loss": 0.4356,
+      "step": 8189
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.340639346117667,
+      "learning_rate": 9.085879315893664e-08,
+      "loss": 0.4889,
+      "step": 8190
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8741167195516106,
+      "learning_rate": 9.050598459483528e-08,
+      "loss": 0.4988,
+      "step": 8191
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.97533943183261,
+      "learning_rate": 9.015385609299431e-08,
+      "loss": 0.5377,
+      "step": 8192
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.672060008411488,
+      "learning_rate": 8.980240770219139e-08,
+      "loss": 0.5132,
+      "step": 8193
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1834326861067987,
+      "learning_rate": 8.945163947110758e-08,
+      "loss": 0.5576,
+      "step": 8194
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7101987298510644,
+      "learning_rate": 8.910155144833288e-08,
+      "loss": 0.3655,
+      "step": 8195
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.834269291635032,
+      "learning_rate": 8.875214368236073e-08,
+      "loss": 0.4482,
+      "step": 8196
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.762876428500223,
+      "learning_rate": 8.840341622159077e-08,
+      "loss": 0.4892,
+      "step": 8197
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.844746178482963,
+      "learning_rate": 8.805536911432877e-08,
+      "loss": 0.5435,
+      "step": 8198
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.1227393898759264,
+      "learning_rate": 8.770800240878619e-08,
+      "loss": 0.4752,
+      "step": 8199
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9073978175133313,
+      "learning_rate": 8.736131615308063e-08,
+      "loss": 0.3858,
+      "step": 8200
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.293273129720375,
+      "learning_rate": 8.701531039523426e-08,
+      "loss": 0.4441,
+      "step": 8201
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.222552394733756,
+      "learning_rate": 8.666998518317648e-08,
+      "loss": 0.4865,
+      "step": 8202
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6788054651142348,
+      "learning_rate": 8.632534056474129e-08,
+      "loss": 0.5124,
+      "step": 8203
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8581537318400115,
+      "learning_rate": 8.598137658766881e-08,
+      "loss": 0.6607,
+      "step": 8204
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1944355909695923,
+      "learning_rate": 8.563809329960481e-08,
+      "loss": 0.5083,
+      "step": 8205
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.112898432923045,
+      "learning_rate": 8.529549074810129e-08,
+      "loss": 0.5304,
+      "step": 8206
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7773799897234068,
+      "learning_rate": 8.495356898061469e-08,
+      "loss": 0.425,
+      "step": 8207
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0268704327700857,
+      "learning_rate": 8.46123280445088e-08,
+      "loss": 0.4676,
+      "step": 8208
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.343903693059684,
+      "learning_rate": 8.427176798705139e-08,
+      "loss": 0.5652,
+      "step": 8209
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1152987859795718,
+      "learning_rate": 8.393188885541748e-08,
+      "loss": 0.4278,
+      "step": 8210
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1574151463604943,
+      "learning_rate": 8.359269069668662e-08,
+      "loss": 0.4443,
+      "step": 8211
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0445715608288832,
+      "learning_rate": 8.325417355784516e-08,
+      "loss": 0.4074,
+      "step": 8212
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.811679910282104,
+      "learning_rate": 8.291633748578331e-08,
+      "loss": 0.495,
+      "step": 8213
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1170193438453664,
+      "learning_rate": 8.257918252729924e-08,
+      "loss": 0.4103,
+      "step": 8214
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.248550909235313,
+      "learning_rate": 8.224270872909556e-08,
+      "loss": 0.5475,
+      "step": 8215
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.590481319259768,
+      "learning_rate": 8.19069161377789e-08,
+      "loss": 0.4672,
+      "step": 8216
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9362351511910811,
+      "learning_rate": 8.157180479986538e-08,
+      "loss": 0.4249,
+      "step": 8217
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5470414448666432,
+      "learning_rate": 8.123737476177284e-08,
+      "loss": 0.4729,
+      "step": 8218
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6596123467268318,
+      "learning_rate": 8.09036260698276e-08,
+      "loss": 0.4436,
+      "step": 8219
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8623066206762264,
+      "learning_rate": 8.057055877025988e-08,
+      "loss": 0.4641,
+      "step": 8220
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1584336283905814,
+      "learning_rate": 8.023817290920666e-08,
+      "loss": 0.5372,
+      "step": 8221
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7393935408484684,
+      "learning_rate": 7.990646853270944e-08,
+      "loss": 0.4489,
+      "step": 8222
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.319954745166295,
+      "learning_rate": 7.957544568671593e-08,
+      "loss": 0.4723,
+      "step": 8223
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6144617123171034,
+      "learning_rate": 7.924510441707945e-08,
+      "loss": 0.4888,
+      "step": 8224
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6440783132425225,
+      "learning_rate": 7.891544476955892e-08,
+      "loss": 0.4689,
+      "step": 8225
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3108294817929007,
+      "learning_rate": 7.858646678981841e-08,
+      "loss": 0.5496,
+      "step": 8226
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6537288980063074,
+      "learning_rate": 7.82581705234281e-08,
+      "loss": 0.4277,
+      "step": 8227
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.077571046408205,
+      "learning_rate": 7.79305560158633e-08,
+      "loss": 0.5252,
+      "step": 8228
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0328674661781676,
+      "learning_rate": 7.760362331250604e-08,
+      "loss": 0.5546,
+      "step": 8229
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9392274085690162,
+      "learning_rate": 7.727737245864175e-08,
+      "loss": 0.5003,
+      "step": 8230
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.157190619506827,
+      "learning_rate": 7.695180349946318e-08,
+      "loss": 0.4555,
+      "step": 8231
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.1867975622490645,
+      "learning_rate": 7.662691648006815e-08,
+      "loss": 0.4839,
+      "step": 8232
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.369498163987176,
+      "learning_rate": 7.630271144546008e-08,
+      "loss": 0.4683,
+      "step": 8233
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.45293076996086,
+      "learning_rate": 7.597918844054753e-08,
+      "loss": 0.5565,
+      "step": 8234
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.053536479159425,
+      "learning_rate": 7.565634751014517e-08,
+      "loss": 0.4757,
+      "step": 8235
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2550532845876514,
+      "learning_rate": 7.533418869897225e-08,
+      "loss": 0.5403,
+      "step": 8236
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0463940998104366,
+      "learning_rate": 7.501271205165473e-08,
+      "loss": 0.5134,
+      "step": 8237
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.213458060128385,
+      "learning_rate": 7.46919176127231e-08,
+      "loss": 0.5811,
+      "step": 8238
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.093498383595694,
+      "learning_rate": 7.43718054266146e-08,
+      "loss": 0.4991,
+      "step": 8239
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7322372661029075,
+      "learning_rate": 7.405237553766986e-08,
+      "loss": 0.4705,
+      "step": 8240
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6661916541879334,
+      "learning_rate": 7.373362799013683e-08,
+      "loss": 0.4979,
+      "step": 8241
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6652458841755704,
+      "learning_rate": 7.341556282816853e-08,
+      "loss": 0.4535,
+      "step": 8242
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.136477538047666,
+      "learning_rate": 7.309818009582304e-08,
+      "loss": 0.4759,
+      "step": 8243
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5367530976434796,
+      "learning_rate": 7.278147983706463e-08,
+      "loss": 0.4927,
+      "step": 8244
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9602328214878317,
+      "learning_rate": 7.246546209576155e-08,
+      "loss": 0.4948,
+      "step": 8245
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.410380220729539,
+      "learning_rate": 7.21501269156899e-08,
+      "loss": 0.4527,
+      "step": 8246
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9721338354242834,
+      "learning_rate": 7.183547434052863e-08,
+      "loss": 0.5225,
+      "step": 8247
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2233943377118486,
+      "learning_rate": 7.152150441386452e-08,
+      "loss": 0.5162,
+      "step": 8248
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.389017145371587,
+      "learning_rate": 7.120821717918724e-08,
+      "loss": 0.464,
+      "step": 8249
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.579128130815504,
+      "learning_rate": 7.089561267989431e-08,
+      "loss": 0.5667,
+      "step": 8250
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9191728095600276,
+      "learning_rate": 7.058369095928719e-08,
+      "loss": 0.4732,
+      "step": 8251
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8138580393729101,
+      "learning_rate": 7.027245206057409e-08,
+      "loss": 0.7062,
+      "step": 8252
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5118539484844935,
+      "learning_rate": 6.996189602686609e-08,
+      "loss": 0.4177,
+      "step": 8253
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0104235382452162,
+      "learning_rate": 6.965202290118323e-08,
+      "loss": 0.5101,
+      "step": 8254
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.115350204816361,
+      "learning_rate": 6.934283272644782e-08,
+      "loss": 0.4891,
+      "step": 8255
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.272952748064892,
+      "learning_rate": 6.90343255454895e-08,
+      "loss": 0.4468,
+      "step": 8256
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.609129833312253,
+      "learning_rate": 6.872650140104186e-08,
+      "loss": 0.572,
+      "step": 8257
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5063754100614526,
+      "learning_rate": 6.841936033574526e-08,
+      "loss": 0.5694,
+      "step": 8258
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.279190487247249,
+      "learning_rate": 6.811290239214452e-08,
+      "loss": 0.4815,
+      "step": 8259
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.525446881570976,
+      "learning_rate": 6.78071276126907e-08,
+      "loss": 0.4325,
+      "step": 8260
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9271975965316115,
+      "learning_rate": 6.750203603973937e-08,
+      "loss": 0.4595,
+      "step": 8261
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8132687488716681,
+      "learning_rate": 6.719762771555116e-08,
+      "loss": 0.4607,
+      "step": 8262
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.806652710156521,
+      "learning_rate": 6.689390268229345e-08,
+      "loss": 0.3799,
+      "step": 8263
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0368999218445993,
+      "learning_rate": 6.65908609820376e-08,
+      "loss": 0.4306,
+      "step": 8264
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.511606902321926,
+      "learning_rate": 6.628850265676167e-08,
+      "loss": 0.478,
+      "step": 8265
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.924056142845482,
+      "learning_rate": 6.598682774834775e-08,
+      "loss": 0.5219,
+      "step": 8266
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.152242874660278,
+      "learning_rate": 6.568583629858349e-08,
+      "loss": 0.454,
+      "step": 8267
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.50656737652948,
+      "learning_rate": 6.538552834916278e-08,
+      "loss": 0.4911,
+      "step": 8268
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8896848370015829,
+      "learning_rate": 6.508590394168402e-08,
+      "loss": 0.407,
+      "step": 8269
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7199146605830398,
+      "learning_rate": 6.478696311765065e-08,
+      "loss": 0.4707,
+      "step": 8270
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7173928531057958,
+      "learning_rate": 6.448870591847289e-08,
+      "loss": 0.4485,
+      "step": 8271
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9465516788346606,
+      "learning_rate": 6.419113238546438e-08,
+      "loss": 0.4874,
+      "step": 8272
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.878305572550131,
+      "learning_rate": 6.3894242559846e-08,
+      "loss": 0.4274,
+      "step": 8273
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8513453368641881,
+      "learning_rate": 6.359803648274154e-08,
+      "loss": 0.5761,
+      "step": 8274
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6468128485972924,
+      "learning_rate": 6.330251419518263e-08,
+      "loss": 0.4787,
+      "step": 8275
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3711208574473503,
+      "learning_rate": 6.300767573810373e-08,
+      "loss": 0.4408,
+      "step": 8276
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7380019881402677,
+      "learning_rate": 6.271352115234663e-08,
+      "loss": 0.4383,
+      "step": 8277
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.400774104624241,
+      "learning_rate": 6.242005047865707e-08,
+      "loss": 0.4868,
+      "step": 8278
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4139888324158805,
+      "learning_rate": 6.212726375768751e-08,
+      "loss": 0.5151,
+      "step": 8279
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9479617748656484,
+      "learning_rate": 6.183516102999387e-08,
+      "loss": 0.4939,
+      "step": 8280
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9495927928379975,
+      "learning_rate": 6.154374233603876e-08,
+      "loss": 0.4245,
+      "step": 8281
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0464669028443323,
+      "learning_rate": 6.12530077161888e-08,
+      "loss": 0.4699,
+      "step": 8282
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1409418423140614,
+      "learning_rate": 6.09629572107162e-08,
+      "loss": 0.4791,
+      "step": 8283
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.130466763204626,
+      "learning_rate": 6.067359085980051e-08,
+      "loss": 0.4834,
+      "step": 8284
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.794358424495802,
+      "learning_rate": 6.038490870352242e-08,
+      "loss": 0.4945,
+      "step": 8285
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5082685382169236,
+      "learning_rate": 6.009691078187163e-08,
+      "loss": 0.4403,
+      "step": 8286
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2014644764366356,
+      "learning_rate": 5.980959713474122e-08,
+      "loss": 0.5277,
+      "step": 8287
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1501686601696033,
+      "learning_rate": 5.95229678019299e-08,
+      "loss": 0.5127,
+      "step": 8288
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.800698362247267,
+      "learning_rate": 5.9237022823140924e-08,
+      "loss": 0.372,
+      "step": 8289
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9283674952445404,
+      "learning_rate": 5.8951762237984264e-08,
+      "loss": 0.5348,
+      "step": 8290
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.529976429818939,
+      "learning_rate": 5.8667186085973326e-08,
+      "loss": 0.4561,
+      "step": 8291
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9901692254826935,
+      "learning_rate": 5.838329440652824e-08,
+      "loss": 0.4535,
+      "step": 8292
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4536320604389217,
+      "learning_rate": 5.8100087238972534e-08,
+      "loss": 0.3832,
+      "step": 8293
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8204082221724118,
+      "learning_rate": 5.7817564622537626e-08,
+      "loss": 0.5183,
+      "step": 8294
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6346098918642138,
+      "learning_rate": 5.75357265963572e-08,
+      "loss": 0.4043,
+      "step": 8295
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.341002831282058,
+      "learning_rate": 5.7254573199472254e-08,
+      "loss": 0.4451,
+      "step": 8296
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7549114316318337,
+      "learning_rate": 5.6974104470827186e-08,
+      "loss": 0.4879,
+      "step": 8297
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.574620651372921,
+      "learning_rate": 5.6694320449273145e-08,
+      "loss": 0.5027,
+      "step": 8298
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.599635536073557,
+      "learning_rate": 5.641522117356635e-08,
+      "loss": 0.5163,
+      "step": 8299
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7737757669358272,
+      "learning_rate": 5.613680668236588e-08,
+      "loss": 0.4186,
+      "step": 8300
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9342727566092899,
+      "learning_rate": 5.5859077014239205e-08,
+      "loss": 0.4512,
+      "step": 8301
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5314673892545008,
+      "learning_rate": 5.5582032207656676e-08,
+      "loss": 0.5119,
+      "step": 8302
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3780497467574446,
+      "learning_rate": 5.530567230099482e-08,
+      "loss": 0.4709,
+      "step": 8303
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.356193781021113,
+      "learning_rate": 5.502999733253411e-08,
+      "loss": 0.5361,
+      "step": 8304
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8040589578286474,
+      "learning_rate": 5.4755007340462354e-08,
+      "loss": 0.4968,
+      "step": 8305
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.295567632926318,
+      "learning_rate": 5.448070236287017e-08,
+      "loss": 0.4163,
+      "step": 8306
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.849799273526126,
+      "learning_rate": 5.420708243775497e-08,
+      "loss": 0.5169,
+      "step": 8307
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.964378658410538,
+      "learning_rate": 5.3934147603017514e-08,
+      "loss": 0.5743,
+      "step": 8308
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7737300496357848,
+      "learning_rate": 5.366189789646592e-08,
+      "loss": 0.4387,
+      "step": 8309
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0994933560696567,
+      "learning_rate": 5.339033335581056e-08,
+      "loss": 0.4968,
+      "step": 8310
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9293592193953497,
+      "learning_rate": 5.311945401867025e-08,
+      "loss": 0.5156,
+      "step": 8311
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7107945886635205,
+      "learning_rate": 5.284925992256551e-08,
+      "loss": 0.408,
+      "step": 8312
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0218748146858965,
+      "learning_rate": 5.25797511049253e-08,
+      "loss": 0.5117,
+      "step": 8313
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4401371717876486,
+      "learning_rate": 5.231092760308032e-08,
+      "loss": 0.5121,
+      "step": 8314
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.114197273563366,
+      "learning_rate": 5.2042789454269104e-08,
+      "loss": 0.5482,
+      "step": 8315
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.030794865734462,
+      "learning_rate": 5.1775336695633616e-08,
+      "loss": 0.4513,
+      "step": 8316
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0056748547593934,
+      "learning_rate": 5.150856936422144e-08,
+      "loss": 0.3973,
+      "step": 8317
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.319002857003995,
+      "learning_rate": 5.124248749698468e-08,
+      "loss": 0.4725,
+      "step": 8318
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3188070986369174,
+      "learning_rate": 5.09770911307822e-08,
+      "loss": 0.5389,
+      "step": 8319
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7582917085029484,
+      "learning_rate": 5.071238030237569e-08,
+      "loss": 0.444,
+      "step": 8320
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.012920319343171,
+      "learning_rate": 5.0448355048432484e-08,
+      "loss": 0.4383,
+      "step": 8321
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9170507531004495,
+      "learning_rate": 5.018501540552611e-08,
+      "loss": 0.5265,
+      "step": 8322
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.1193905185817394,
+      "learning_rate": 4.992236141013407e-08,
+      "loss": 0.3957,
+      "step": 8323
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7264590627593157,
+      "learning_rate": 4.966039309863946e-08,
+      "loss": 0.4861,
+      "step": 8324
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7713697854863821,
+      "learning_rate": 4.939911050732937e-08,
+      "loss": 0.5064,
+      "step": 8325
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.5448806095320884,
+      "learning_rate": 4.913851367239764e-08,
+      "loss": 0.4752,
+      "step": 8326
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.864661199703402,
+      "learning_rate": 4.887860262994038e-08,
+      "loss": 0.3706,
+      "step": 8327
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.215967204836886,
+      "learning_rate": 4.861937741596268e-08,
+      "loss": 0.5291,
+      "step": 8328
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.079927876863055,
+      "learning_rate": 4.8360838066370266e-08,
+      "loss": 0.5562,
+      "step": 8329
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.243969187991946,
+      "learning_rate": 4.810298461697727e-08,
+      "loss": 0.4789,
+      "step": 8330
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.014411034545867,
+      "learning_rate": 4.784581710350123e-08,
+      "loss": 0.4511,
+      "step": 8331
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5906508418840577,
+      "learning_rate": 4.758933556156475e-08,
+      "loss": 0.4787,
+      "step": 8332
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9918233939079075,
+      "learning_rate": 4.73335400266961e-08,
+      "loss": 0.5215,
+      "step": 8333
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7545592311875304,
+      "learning_rate": 4.707843053432748e-08,
+      "loss": 0.3953,
+      "step": 8334
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8255317064485896,
+      "learning_rate": 4.6824007119796176e-08,
+      "loss": 0.4647,
+      "step": 8335
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.65089495443273,
+      "learning_rate": 4.657026981834623e-08,
+      "loss": 0.4531,
+      "step": 8336
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.065329220433614,
+      "learning_rate": 4.6317218665123953e-08,
+      "loss": 0.4651,
+      "step": 8337
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8427960822306894,
+      "learning_rate": 4.606485369518354e-08,
+      "loss": 0.6722,
+      "step": 8338
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8565782117245647,
+      "learning_rate": 4.581317494348092e-08,
+      "loss": 0.6953,
+      "step": 8339
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9944112565467742,
+      "learning_rate": 4.556218244487876e-08,
+      "loss": 0.5364,
+      "step": 8340
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8492523932549663,
+      "learning_rate": 4.53118762341459e-08,
+      "loss": 0.4648,
+      "step": 8341
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9259408803126086,
+      "learning_rate": 4.506225634595296e-08,
+      "loss": 0.5278,
+      "step": 8342
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2960479323297487,
+      "learning_rate": 4.481332281487838e-08,
+      "loss": 0.4093,
+      "step": 8343
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8104865082487747,
+      "learning_rate": 4.4565075675404024e-08,
+      "loss": 0.5431,
+      "step": 8344
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.67742257981651,
+      "learning_rate": 4.4317514961917387e-08,
+      "loss": 0.5489,
+      "step": 8345
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.493866747800699,
+      "learning_rate": 4.407064070870992e-08,
+      "loss": 0.46,
+      "step": 8346
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.6316221402586075,
+      "learning_rate": 4.3824452949978725e-08,
+      "loss": 0.4902,
+      "step": 8347
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.052226605028183,
+      "learning_rate": 4.357895171982596e-08,
+      "loss": 0.5052,
+      "step": 8348
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8510981466546902,
+      "learning_rate": 4.333413705225886e-08,
+      "loss": 0.7161,
+      "step": 8349
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0769268074883893,
+      "learning_rate": 4.3090008981187534e-08,
+      "loss": 0.5383,
+      "step": 8350
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9229641820451802,
+      "learning_rate": 4.284656754043048e-08,
+      "loss": 0.4235,
+      "step": 8351
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.069245540581593,
+      "learning_rate": 4.2603812763707956e-08,
+      "loss": 0.4546,
+      "step": 8352
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7400006153394267,
+      "learning_rate": 4.2361744684646934e-08,
+      "loss": 0.4568,
+      "step": 8353
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0192758117837375,
+      "learning_rate": 4.212036333677783e-08,
+      "loss": 0.4641,
+      "step": 8354
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2915084454007686,
+      "learning_rate": 4.187966875353777e-08,
+      "loss": 0.5357,
+      "step": 8355
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3644591370553076,
+      "learning_rate": 4.163966096826677e-08,
+      "loss": 0.5057,
+      "step": 8356
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6612270854985094,
+      "learning_rate": 4.140034001421156e-08,
+      "loss": 0.4144,
+      "step": 8357
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6508439030693616,
+      "learning_rate": 4.116170592452229e-08,
+      "loss": 0.4143,
+      "step": 8358
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.2723744471142227,
+      "learning_rate": 4.0923758732254734e-08,
+      "loss": 0.4128,
+      "step": 8359
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2170461936941415,
+      "learning_rate": 4.068649847036865e-08,
+      "loss": 0.6117,
+      "step": 8360
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9644081176828636,
+      "learning_rate": 4.044992517173052e-08,
+      "loss": 0.4898,
+      "step": 8361
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1337027713314534,
+      "learning_rate": 4.021403886910913e-08,
+      "loss": 0.471,
+      "step": 8362
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.767649621758456,
+      "learning_rate": 3.997883959518001e-08,
+      "loss": 0.4579,
+      "step": 8363
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0886999861838875,
+      "learning_rate": 3.9744327382523206e-08,
+      "loss": 0.6288,
+      "step": 8364
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8882376352265282,
+      "learning_rate": 3.951050226362274e-08,
+      "loss": 0.4235,
+      "step": 8365
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.789232361878686,
+      "learning_rate": 3.9277364270868813e-08,
+      "loss": 0.483,
+      "step": 8366
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.406925349270383,
+      "learning_rate": 3.904491343655503e-08,
+      "loss": 0.5216,
+      "step": 8367
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8018086552200833,
+      "learning_rate": 3.8813149792880645e-08,
+      "loss": 0.6529,
+      "step": 8368
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7943811290522491,
+      "learning_rate": 3.858207337194886e-08,
+      "loss": 0.4787,
+      "step": 8369
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.187189507261587,
+      "learning_rate": 3.8351684205769625e-08,
+      "loss": 0.4737,
+      "step": 8370
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.198379545315124,
+      "learning_rate": 3.812198232625519e-08,
+      "loss": 0.4239,
+      "step": 8371
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.643911876224704,
+      "learning_rate": 3.789296776522455e-08,
+      "loss": 0.5484,
+      "step": 8372
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0109929193367604,
+      "learning_rate": 3.7664640554400664e-08,
+      "loss": 0.4673,
+      "step": 8373
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8846239807661482,
+      "learning_rate": 3.743700072541101e-08,
+      "loss": 0.4271,
+      "step": 8374
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9854813325344618,
+      "learning_rate": 3.721004830978869e-08,
+      "loss": 0.4916,
+      "step": 8375
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1284789018538683,
+      "learning_rate": 3.698378333897079e-08,
+      "loss": 0.5168,
+      "step": 8376
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.151850844664872,
+      "learning_rate": 3.675820584429945e-08,
+      "loss": 0.5084,
+      "step": 8377
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6475363117531057,
+      "learning_rate": 3.653331585702247e-08,
+      "loss": 0.5225,
+      "step": 8378
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6651793597962228,
+      "learning_rate": 3.630911340828991e-08,
+      "loss": 0.5083,
+      "step": 8379
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0134997382854154,
+      "learning_rate": 3.6085598529159714e-08,
+      "loss": 0.5645,
+      "step": 8380
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.128904667769331,
+      "learning_rate": 3.586277125059268e-08,
+      "loss": 0.4944,
+      "step": 8381
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6927035545953664,
+      "learning_rate": 3.5640631603454653e-08,
+      "loss": 0.466,
+      "step": 8382
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7987987792779336,
+      "learning_rate": 3.5419179618516596e-08,
+      "loss": 0.5473,
+      "step": 8383
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2235500287758265,
+      "learning_rate": 3.519841532645396e-08,
+      "loss": 0.524,
+      "step": 8384
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8052836410480801,
+      "learning_rate": 3.497833875784673e-08,
+      "loss": 0.4734,
+      "step": 8385
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6543562281863915,
+      "learning_rate": 3.475894994317996e-08,
+      "loss": 0.4936,
+      "step": 8386
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7577600683884453,
+      "learning_rate": 3.45402489128438e-08,
+      "loss": 0.6546,
+      "step": 8387
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4332431479677776,
+      "learning_rate": 3.432223569713178e-08,
+      "loss": 0.552,
+      "step": 8388
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.390866864673474,
+      "learning_rate": 3.410491032624419e-08,
+      "loss": 0.5447,
+      "step": 8389
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.17077927121182,
+      "learning_rate": 3.388827283028362e-08,
+      "loss": 0.4603,
+      "step": 8390
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7194717503043204,
+      "learning_rate": 3.367232323925995e-08,
+      "loss": 0.4518,
+      "step": 8391
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.816602315120856,
+      "learning_rate": 3.3457061583085374e-08,
+      "loss": 0.4718,
+      "step": 8392
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8277417618324248,
+      "learning_rate": 3.32424878915788e-08,
+      "loss": 0.4985,
+      "step": 8393
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5512317915937124,
+      "learning_rate": 3.3028602194462575e-08,
+      "loss": 0.4821,
+      "step": 8394
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8852016333211405,
+      "learning_rate": 3.281540452136356e-08,
+      "loss": 0.6756,
+      "step": 8395
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0603564883434085,
+      "learning_rate": 3.260289490181479e-08,
+      "loss": 0.5182,
+      "step": 8396
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1124459115636793,
+      "learning_rate": 3.239107336525271e-08,
+      "loss": 0.5522,
+      "step": 8397
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2708996199617846,
+      "learning_rate": 3.217993994101831e-08,
+      "loss": 0.454,
+      "step": 8398
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2983711027818856,
+      "learning_rate": 3.1969494658358746e-08,
+      "loss": 0.5477,
+      "step": 8399
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5255894993535972,
+      "learning_rate": 3.1759737546423476e-08,
+      "loss": 0.4896,
+      "step": 8400
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7655659795787515,
+      "learning_rate": 3.155066863426981e-08,
+      "loss": 0.5487,
+      "step": 8401
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.089434539574228,
+      "learning_rate": 3.1342287950856256e-08,
+      "loss": 0.3889,
+      "step": 8402
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.000821966318182,
+      "learning_rate": 3.1134595525048606e-08,
+      "loss": 0.5534,
+      "step": 8403
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6159325657035253,
+      "learning_rate": 3.092759138561607e-08,
+      "loss": 0.4524,
+      "step": 8404
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.823958876161883,
+      "learning_rate": 3.0721275561232946e-08,
+      "loss": 0.4518,
+      "step": 8405
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7726353103165573,
+      "learning_rate": 3.0515648080478023e-08,
+      "loss": 0.4692,
+      "step": 8406
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1349416848488594,
+      "learning_rate": 3.0310708971834635e-08,
+      "loss": 0.4878,
+      "step": 8407
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8783516817970018,
+      "learning_rate": 3.010645826369174e-08,
+      "loss": 0.4874,
+      "step": 8408
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.326212082852649,
+      "learning_rate": 2.990289598434115e-08,
+      "loss": 0.4293,
+      "step": 8409
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2717858746221826,
+      "learning_rate": 2.9700022161980312e-08,
+      "loss": 0.4886,
+      "step": 8410
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6908264024446573,
+      "learning_rate": 2.9497836824711192e-08,
+      "loss": 0.5352,
+      "step": 8411
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8391981957765515,
+      "learning_rate": 2.929634000054138e-08,
+      "loss": 0.5575,
+      "step": 8412
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3371403026086774,
+      "learning_rate": 2.909553171738133e-08,
+      "loss": 0.4999,
+      "step": 8413
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7504321700164074,
+      "learning_rate": 2.889541200304713e-08,
+      "loss": 0.4945,
+      "step": 8414
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1208162869757556,
+      "learning_rate": 2.8695980885259933e-08,
+      "loss": 0.4417,
+      "step": 8415
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.455106333105743,
+      "learning_rate": 2.8497238391643756e-08,
+      "loss": 0.5292,
+      "step": 8416
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5402633986817746,
+      "learning_rate": 2.8299184549729352e-08,
+      "loss": 0.4951,
+      "step": 8417
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8239994502568786,
+      "learning_rate": 2.810181938695089e-08,
+      "loss": 0.6507,
+      "step": 8418
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6206688399795766,
+      "learning_rate": 2.7905142930647057e-08,
+      "loss": 0.5395,
+      "step": 8419
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7201942730462734,
+      "learning_rate": 2.770915520806161e-08,
+      "loss": 0.4526,
+      "step": 8420
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2308190204431595,
+      "learning_rate": 2.751385624634284e-08,
+      "loss": 0.4954,
+      "step": 8421
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4642404555732598,
+      "learning_rate": 2.7319246072543548e-08,
+      "loss": 0.5399,
+      "step": 8422
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.937547888831758,
+      "learning_rate": 2.712532471362106e-08,
+      "loss": 0.4536,
+      "step": 8423
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.23933470458382,
+      "learning_rate": 2.693209219643722e-08,
+      "loss": 0.4931,
+      "step": 8424
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3441629634171455,
+      "learning_rate": 2.6739548547758398e-08,
+      "loss": 0.4677,
+      "step": 8425
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.242565192478841,
+      "learning_rate": 2.6547693794256037e-08,
+      "loss": 0.5458,
+      "step": 8426
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0644891175962727,
+      "learning_rate": 2.635652796250554e-08,
+      "loss": 0.5168,
+      "step": 8427
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0575660061453633,
+      "learning_rate": 2.6166051078987954e-08,
+      "loss": 0.4991,
+      "step": 8428
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.691731750873132,
+      "learning_rate": 2.597626317008717e-08,
+      "loss": 0.4864,
+      "step": 8429
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9080349391019298,
+      "learning_rate": 2.5787164262092158e-08,
+      "loss": 0.5258,
+      "step": 8430
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9531367460445592,
+      "learning_rate": 2.5598754381198632e-08,
+      "loss": 0.4467,
+      "step": 8431
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.822465814819731,
+      "learning_rate": 2.5411033553503495e-08,
+      "loss": 0.4288,
+      "step": 8432
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7390680119366433,
+      "learning_rate": 2.522400180501039e-08,
+      "loss": 0.4249,
+      "step": 8433
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.257068989349596,
+      "learning_rate": 2.5037659161626925e-08,
+      "loss": 0.5719,
+      "step": 8434
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8494445754465583,
+      "learning_rate": 2.4852005649165235e-08,
+      "loss": 0.54,
+      "step": 8435
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8188644235731277,
+      "learning_rate": 2.4667041293341964e-08,
+      "loss": 0.474,
+      "step": 8436
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.142761233091188,
+      "learning_rate": 2.448276611977829e-08,
+      "loss": 0.5026,
+      "step": 8437
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.160623722539833,
+      "learning_rate": 2.4299180154000456e-08,
+      "loss": 0.4253,
+      "step": 8438
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9475239563726947,
+      "learning_rate": 2.4116283421438126e-08,
+      "loss": 0.3693,
+      "step": 8439
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6578631355556057,
+      "learning_rate": 2.3934075947426027e-08,
+      "loss": 0.4346,
+      "step": 8440
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0678799984307927,
+      "learning_rate": 2.375255775720453e-08,
+      "loss": 0.4726,
+      "step": 8441
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.20676376509795,
+      "learning_rate": 2.35717288759163e-08,
+      "loss": 0.541,
+      "step": 8442
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9981210931030393,
+      "learning_rate": 2.3391589328610743e-08,
+      "loss": 0.494,
+      "step": 8443
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5814425396576608,
+      "learning_rate": 2.3212139140239563e-08,
+      "loss": 0.4527,
+      "step": 8444
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.018066009769416,
+      "learning_rate": 2.3033378335661217e-08,
+      "loss": 0.5379,
+      "step": 8445
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.175053052839372,
+      "learning_rate": 2.2855306939637557e-08,
+      "loss": 0.5147,
+      "step": 8446
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0260251094210053,
+      "learning_rate": 2.267792497683441e-08,
+      "loss": 0.4859,
+      "step": 8447
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8713588163861292,
+      "learning_rate": 2.2501232471822675e-08,
+      "loss": 0.4937,
+      "step": 8448
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.604649603649941,
+      "learning_rate": 2.232522944907889e-08,
+      "loss": 0.4605,
+      "step": 8449
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.965090530245984,
+      "learning_rate": 2.2149915932981327e-08,
+      "loss": 0.3869,
+      "step": 8450
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.147430737549078,
+      "learning_rate": 2.1975291947815558e-08,
+      "loss": 0.4408,
+      "step": 8451
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8104388605531028,
+      "learning_rate": 2.180135751777057e-08,
+      "loss": 0.3794,
+      "step": 8452
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.109809351145331,
+      "learning_rate": 2.162811266693876e-08,
+      "loss": 0.4269,
+      "step": 8453
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9526464006350546,
+      "learning_rate": 2.1455557419318705e-08,
+      "loss": 0.4702,
+      "step": 8454
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8359060230630577,
+      "learning_rate": 2.1283691798812402e-08,
+      "loss": 0.4791,
+      "step": 8455
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.25065276494736,
+      "learning_rate": 2.1112515829226908e-08,
+      "loss": 0.4716,
+      "step": 8456
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5567277534554544,
+      "learning_rate": 2.0942029534273267e-08,
+      "loss": 0.4761,
+      "step": 8457
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.631247685154881,
+      "learning_rate": 2.077223293756703e-08,
+      "loss": 0.4911,
+      "step": 8458
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0323116118688063,
+      "learning_rate": 2.0603126062629374e-08,
+      "loss": 0.4584,
+      "step": 8459
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3192401821762623,
+      "learning_rate": 2.0434708932883796e-08,
+      "loss": 0.4833,
+      "step": 8460
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.680906670255701,
+      "learning_rate": 2.0266981571659405e-08,
+      "loss": 0.5055,
+      "step": 8461
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8193953670570945,
+      "learning_rate": 2.0099944002190953e-08,
+      "loss": 0.7005,
+      "step": 8462
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8741689641931003,
+      "learning_rate": 1.9933596247615483e-08,
+      "loss": 0.5364,
+      "step": 8463
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1983409769176583,
+      "learning_rate": 1.9767938330975678e-08,
+      "loss": 0.4177,
+      "step": 8464
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.167007146401783,
+      "learning_rate": 1.960297027521818e-08,
+      "loss": 0.4965,
+      "step": 8465
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.870986641745271,
+      "learning_rate": 1.9438692103194713e-08,
+      "loss": 0.4802,
+      "step": 8466
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1619573963761702,
+      "learning_rate": 1.9275103837660956e-08,
+      "loss": 0.4939,
+      "step": 8467
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.054941735966446,
+      "learning_rate": 1.9112205501277124e-08,
+      "loss": 0.5166,
+      "step": 8468
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.139724149684847,
+      "learning_rate": 1.8949997116608498e-08,
+      "loss": 0.5103,
+      "step": 8469
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.349340462580863,
+      "learning_rate": 1.878847870612266e-08,
+      "loss": 0.4815,
+      "step": 8470
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.851937866277887,
+      "learning_rate": 1.8627650292194488e-08,
+      "loss": 0.4975,
+      "step": 8471
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.51334280521689,
+      "learning_rate": 1.8467511897101164e-08,
+      "loss": 0.4954,
+      "step": 8472
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.210048224824802,
+      "learning_rate": 1.83080635430255e-08,
+      "loss": 0.5644,
+      "step": 8473
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1642360409715606,
+      "learning_rate": 1.8149305252053716e-08,
+      "loss": 0.482,
+      "step": 8474
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.696092060331628,
+      "learning_rate": 1.7991237046177667e-08,
+      "loss": 0.494,
+      "step": 8475
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9634367009493712,
+      "learning_rate": 1.783385894729206e-08,
+      "loss": 0.5477,
+      "step": 8476
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9698477278841708,
+      "learning_rate": 1.767717097719779e-08,
+      "loss": 0.5269,
+      "step": 8477
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.321785312113312,
+      "learning_rate": 1.7521173157598603e-08,
+      "loss": 0.5042,
+      "step": 8478
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.765743303210131,
+      "learning_rate": 1.736586551010333e-08,
+      "loss": 0.4459,
+      "step": 8479
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1481328335478977,
+      "learning_rate": 1.7211248056225872e-08,
+      "loss": 0.4413,
+      "step": 8480
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1950317676732265,
+      "learning_rate": 1.7057320817382984e-08,
+      "loss": 0.4531,
+      "step": 8481
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8356280333577921,
+      "learning_rate": 1.6904083814897054e-08,
+      "loss": 0.4618,
+      "step": 8482
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.075591198535836,
+      "learning_rate": 1.675153706999444e-08,
+      "loss": 0.4331,
+      "step": 8483
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9990094901101156,
+      "learning_rate": 1.6599680603805458e-08,
+      "loss": 0.4703,
+      "step": 8484
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.127273916817755,
+      "learning_rate": 1.6448514437365503e-08,
+      "loss": 0.496,
+      "step": 8485
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9145256582137946,
+      "learning_rate": 1.6298038591614497e-08,
+      "loss": 0.3906,
+      "step": 8486
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0969988143908243,
+      "learning_rate": 1.614825308739576e-08,
+      "loss": 0.4669,
+      "step": 8487
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.917974179311906,
+      "learning_rate": 1.59991579454577e-08,
+      "loss": 0.416,
+      "step": 8488
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3076442465064444,
+      "learning_rate": 1.5850753186453238e-08,
+      "loss": 0.4845,
+      "step": 8489
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6636443427063503,
+      "learning_rate": 1.5703038830939266e-08,
+      "loss": 0.4655,
+      "step": 8490
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5570495942017426,
+      "learning_rate": 1.5556014899376637e-08,
+      "loss": 0.5515,
+      "step": 8491
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5162814625025276,
+      "learning_rate": 1.540968141213184e-08,
+      "loss": 0.4226,
+      "step": 8492
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1397169738795956,
+      "learning_rate": 1.5264038389474768e-08,
+      "loss": 0.5577,
+      "step": 8493
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.60614032565627,
+      "learning_rate": 1.5119085851579286e-08,
+      "loss": 0.4528,
+      "step": 8494
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.6643033549567723,
+      "learning_rate": 1.4974823818524885e-08,
+      "loss": 0.431,
+      "step": 8495
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2926451754578205,
+      "learning_rate": 1.4831252310294474e-08,
+      "loss": 0.5129,
+      "step": 8496
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0152383731863117,
+      "learning_rate": 1.4688371346775477e-08,
+      "loss": 0.4879,
+      "step": 8497
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.421600297366718,
+      "learning_rate": 1.4546180947759837e-08,
+      "loss": 0.4992,
+      "step": 8498
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.281790832238897,
+      "learning_rate": 1.4404681132943465e-08,
+      "loss": 0.4321,
+      "step": 8499
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7485509574901028,
+      "learning_rate": 1.4263871921927352e-08,
+      "loss": 0.5278,
+      "step": 8500
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.613976016237742,
+      "learning_rate": 1.4123753334215895e-08,
+      "loss": 0.4557,
+      "step": 8501
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4195857646804195,
+      "learning_rate": 1.3984325389219123e-08,
+      "loss": 0.4454,
+      "step": 8502
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.6673902316032474,
+      "learning_rate": 1.3845588106249364e-08,
+      "loss": 0.4604,
+      "step": 8503
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8462693168908872,
+      "learning_rate": 1.370754150452569e-08,
+      "loss": 0.4589,
+      "step": 8504
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1850044896747827,
+      "learning_rate": 1.3570185603168917e-08,
+      "loss": 0.5954,
+      "step": 8505
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3899324423913835,
+      "learning_rate": 1.3433520421207158e-08,
+      "loss": 0.4744,
+      "step": 8506
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0112272781875453,
+      "learning_rate": 1.3297545977569714e-08,
+      "loss": 0.5041,
+      "step": 8507
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8851436413315588,
+      "learning_rate": 1.3162262291093187e-08,
+      "loss": 0.4849,
+      "step": 8508
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.357141443423784,
+      "learning_rate": 1.3027669380515918e-08,
+      "loss": 0.4167,
+      "step": 8509
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9113891969660923,
+      "learning_rate": 1.2893767264482438e-08,
+      "loss": 0.5183,
+      "step": 8510
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5564075672629807,
+      "learning_rate": 1.2760555961540133e-08,
+      "loss": 0.3791,
+      "step": 8511
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7676913062937203,
+      "learning_rate": 1.2628035490142021e-08,
+      "loss": 0.4722,
+      "step": 8512
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.170345819527668,
+      "learning_rate": 1.2496205868644529e-08,
+      "loss": 0.4538,
+      "step": 8513
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.316262792178341,
+      "learning_rate": 1.236506711530916e-08,
+      "loss": 0.5178,
+      "step": 8514
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.002275652729228,
+      "learning_rate": 1.2234619248300273e-08,
+      "loss": 0.4668,
+      "step": 8515
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0921247522436763,
+      "learning_rate": 1.2104862285688413e-08,
+      "loss": 0.4325,
+      "step": 8516
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.982930861805802,
+      "learning_rate": 1.1975796245446425e-08,
+      "loss": 0.413,
+      "step": 8517
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.9727829082129404,
+      "learning_rate": 1.1847421145453896e-08,
+      "loss": 0.5543,
+      "step": 8518
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.009434385601694,
+      "learning_rate": 1.171973700349216e-08,
+      "loss": 0.5987,
+      "step": 8519
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7292943239194842,
+      "learning_rate": 1.1592743837248731e-08,
+      "loss": 0.4451,
+      "step": 8520
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.044725409114241,
+      "learning_rate": 1.1466441664314543e-08,
+      "loss": 0.4616,
+      "step": 8521
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.80333876280949,
+      "learning_rate": 1.1340830502184486e-08,
+      "loss": 0.3796,
+      "step": 8522
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.833313222283905,
+      "learning_rate": 1.1215910368258531e-08,
+      "loss": 0.3819,
+      "step": 8523
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3957836306616564,
+      "learning_rate": 1.1091681279840616e-08,
+      "loss": 0.5141,
+      "step": 8524
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7694226944397944,
+      "learning_rate": 1.0968143254139196e-08,
+      "loss": 0.4345,
+      "step": 8525
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.8815356097743163,
+      "learning_rate": 1.084529630826614e-08,
+      "loss": 0.5465,
+      "step": 8526
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8265132012478755,
+      "learning_rate": 1.072314045923839e-08,
+      "loss": 0.5159,
+      "step": 8527
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.150949089204457,
+      "learning_rate": 1.0601675723977412e-08,
+      "loss": 0.5081,
+      "step": 8528
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.483612270092897,
+      "learning_rate": 1.0480902119308078e-08,
+      "loss": 0.4364,
+      "step": 8529
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.752178594072906,
+      "learning_rate": 1.0360819661959786e-08,
+      "loss": 0.4205,
+      "step": 8530
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.1776093438044084,
+      "learning_rate": 1.0241428368566453e-08,
+      "loss": 0.5201,
+      "step": 8531
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8303775928671429,
+      "learning_rate": 1.0122728255666514e-08,
+      "loss": 0.5272,
+      "step": 8532
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.939028692573307,
+      "learning_rate": 1.0004719339701818e-08,
+      "loss": 0.5291,
+      "step": 8533
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.449724745337555,
+      "learning_rate": 9.887401637019289e-09,
+      "loss": 0.5051,
+      "step": 8534
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2797608774893727,
+      "learning_rate": 9.770775163869262e-09,
+      "loss": 0.5102,
+      "step": 8535
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4888090943252643,
+      "learning_rate": 9.654839936407145e-09,
+      "loss": 0.5772,
+      "step": 8536
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.457495641497708,
+      "learning_rate": 9.539595970692318e-09,
+      "loss": 0.4646,
+      "step": 8537
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8701443425665243,
+      "learning_rate": 9.425043282688118e-09,
+      "loss": 0.4945,
+      "step": 8538
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.029551704944981,
+      "learning_rate": 9.311181888262965e-09,
+      "loss": 0.5002,
+      "step": 8539
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.383623191173553,
+      "learning_rate": 9.198011803188134e-09,
+      "loss": 0.4711,
+      "step": 8540
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2909766674051615,
+      "learning_rate": 9.085533043140526e-09,
+      "loss": 0.3959,
+      "step": 8541
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2582704344039635,
+      "learning_rate": 8.973745623699904e-09,
+      "loss": 0.4068,
+      "step": 8542
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0371618658084167,
+      "learning_rate": 8.86264956035221e-09,
+      "loss": 0.4619,
+      "step": 8543
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1135467765767633,
+      "learning_rate": 8.752244868485692e-09,
+      "loss": 0.4541,
+      "step": 8544
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2003478395875833,
+      "learning_rate": 8.64253156339312e-09,
+      "loss": 0.5296,
+      "step": 8545
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8541614999106166,
+      "learning_rate": 8.533509660273443e-09,
+      "loss": 0.4841,
+      "step": 8546
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3349111512738374,
+      "learning_rate": 8.425179174226806e-09,
+      "loss": 0.4507,
+      "step": 8547
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0729501813777516,
+      "learning_rate": 8.317540120260093e-09,
+      "loss": 0.5069,
+      "step": 8548
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.304912080605692,
+      "learning_rate": 8.210592513283044e-09,
+      "loss": 0.4546,
+      "step": 8549
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8076757188099863,
+      "learning_rate": 8.104336368109922e-09,
+      "loss": 0.4525,
+      "step": 8550
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8024973954871126,
+      "learning_rate": 7.998771699459506e-09,
+      "loss": 0.548,
+      "step": 8551
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.897601398837334,
+      "learning_rate": 7.8938985219551e-09,
+      "loss": 0.5335,
+      "step": 8552
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0794020608788992,
+      "learning_rate": 7.789716850122863e-09,
+      "loss": 0.4895,
+      "step": 8553
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9681372173852025,
+      "learning_rate": 7.686226698394028e-09,
+      "loss": 0.4632,
+      "step": 8554
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8034034162526709,
+      "learning_rate": 7.583428081104905e-09,
+      "loss": 0.4651,
+      "step": 8555
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1226761239874907,
+      "learning_rate": 7.481321012494658e-09,
+      "loss": 0.392,
+      "step": 8556
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2948510184863076,
+      "learning_rate": 7.37990550670642e-09,
+      "loss": 0.5451,
+      "step": 8557
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6692194908255125,
+      "learning_rate": 7.279181577789507e-09,
+      "loss": 0.4258,
+      "step": 8558
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3350585851063483,
+      "learning_rate": 7.179149239695538e-09,
+      "loss": 0.3638,
+      "step": 8559
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.419721711929588,
+      "learning_rate": 7.079808506281205e-09,
+      "loss": 0.5438,
+      "step": 8560
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.375983809543111,
+      "learning_rate": 6.981159391306613e-09,
+      "loss": 0.4911,
+      "step": 8561
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.02420963976272,
+      "learning_rate": 6.883201908436943e-09,
+      "loss": 0.475,
+      "step": 8562
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8510713419896108,
+      "learning_rate": 6.7859360712418945e-09,
+      "loss": 0.6603,
+      "step": 8563
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8438141280400084,
+      "learning_rate": 6.689361893193469e-09,
+      "loss": 0.4068,
+      "step": 8564
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.456464460836008,
+      "learning_rate": 6.593479387669854e-09,
+      "loss": 0.4216,
+      "step": 8565
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.751682333494253,
+      "learning_rate": 6.498288567953204e-09,
+      "loss": 0.4706,
+      "step": 8566
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6375181445623097,
+      "learning_rate": 6.40378944722797e-09,
+      "loss": 0.4423,
+      "step": 8567
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0754917796511094,
+      "learning_rate": 6.309982038585349e-09,
+      "loss": 0.429,
+      "step": 8568
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.3267378812199415,
+      "learning_rate": 6.2168663550188356e-09,
+      "loss": 0.4595,
+      "step": 8569
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.3000545329164703,
+      "learning_rate": 6.124442409427001e-09,
+      "loss": 0.4661,
+      "step": 8570
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.595311015917194,
+      "learning_rate": 6.032710214612936e-09,
+      "loss": 0.5048,
+      "step": 8571
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.000507948445932,
+      "learning_rate": 5.941669783282589e-09,
+      "loss": 0.5218,
+      "step": 8572
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2721584672003172,
+      "learning_rate": 5.851321128046983e-09,
+      "loss": 0.5126,
+      "step": 8573
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6228493234912813,
+      "learning_rate": 5.761664261421662e-09,
+      "loss": 0.4873,
+      "step": 8574
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.456458132930476,
+      "learning_rate": 5.6726991958250265e-09,
+      "loss": 0.552,
+      "step": 8575
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4854409277569296,
+      "learning_rate": 5.5844259435816615e-09,
+      "loss": 0.4871,
+      "step": 8576
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1682074798640367,
+      "learning_rate": 5.496844516918454e-09,
+      "loss": 0.5335,
+      "step": 8577
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.846728805331797,
+      "learning_rate": 5.409954927966809e-09,
+      "loss": 0.5168,
+      "step": 8578
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9658518545458625,
+      "learning_rate": 5.323757188763768e-09,
+      "loss": 0.5128,
+      "step": 8579
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.682218920076119,
+      "learning_rate": 5.2382513112481105e-09,
+      "loss": 0.5317,
+      "step": 8580
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8767494447228479,
+      "learning_rate": 5.153437307265363e-09,
+      "loss": 0.539,
+      "step": 8581
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8510255968077208,
+      "learning_rate": 5.0693151885627955e-09,
+      "loss": 0.6957,
+      "step": 8582
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8085321647330515,
+      "learning_rate": 4.985884966793864e-09,
+      "loss": 0.4895,
+      "step": 8583
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0813163187646797,
+      "learning_rate": 4.903146653515434e-09,
+      "loss": 0.5301,
+      "step": 8584
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8625512875989088,
+      "learning_rate": 4.821100260187228e-09,
+      "loss": 0.6781,
+      "step": 8585
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8171125373989705,
+      "learning_rate": 4.739745798175155e-09,
+      "loss": 0.4937,
+      "step": 8586
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6599400124616426,
+      "learning_rate": 4.659083278748533e-09,
+      "loss": 0.3579,
+      "step": 8587
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.016290464523796,
+      "learning_rate": 4.5791127130800916e-09,
+      "loss": 0.4551,
+      "step": 8588
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2060162517301647,
+      "learning_rate": 4.499834112248191e-09,
+      "loss": 0.4552,
+      "step": 8589
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.1963431740166874,
+      "learning_rate": 4.421247487233493e-09,
+      "loss": 0.5383,
+      "step": 8590
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4032892068817984,
+      "learning_rate": 4.343352848922844e-09,
+      "loss": 0.4433,
+      "step": 8591
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.005096962045565,
+      "learning_rate": 4.2661502081053905e-09,
+      "loss": 0.5012,
+      "step": 8592
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.035971664649162,
+      "learning_rate": 4.189639575475912e-09,
+      "loss": 0.5677,
+      "step": 8593
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.152531079340269,
+      "learning_rate": 4.1138209616320426e-09,
+      "loss": 0.5031,
+      "step": 8594
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4093864588153306,
+      "learning_rate": 4.038694377075936e-09,
+      "loss": 0.5274,
+      "step": 8595
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5620957703645124,
+      "learning_rate": 3.964259832215378e-09,
+      "loss": 0.4775,
+      "step": 8596
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8963137399836518,
+      "learning_rate": 3.890517337359345e-09,
+      "loss": 0.4439,
+      "step": 8597
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.04489372373147,
+      "learning_rate": 3.8174669027241095e-09,
+      "loss": 0.5024,
+      "step": 8598
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.49376132145976,
+      "learning_rate": 3.745108538427688e-09,
+      "loss": 0.4608,
+      "step": 8599
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1156139131456584,
+      "learning_rate": 3.673442254493731e-09,
+      "loss": 0.4705,
+      "step": 8600
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.075661739913366,
+      "learning_rate": 3.6024680608487406e-09,
+      "loss": 0.526,
+      "step": 8601
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.953779536786496,
+      "learning_rate": 3.532185967324853e-09,
+      "loss": 0.5343,
+      "step": 8602
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.077554738644134,
+      "learning_rate": 3.4625959836570576e-09,
+      "loss": 0.4168,
+      "step": 8603
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.340869385240241,
+      "learning_rate": 3.3936981194848674e-09,
+      "loss": 0.4325,
+      "step": 8604
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8055745131885308,
+      "learning_rate": 3.3254923843523135e-09,
+      "loss": 0.4636,
+      "step": 8605
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.615977719319574,
+      "learning_rate": 3.257978787706839e-09,
+      "loss": 0.4568,
+      "step": 8606
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.022503086931025,
+      "learning_rate": 3.1911573389015183e-09,
+      "loss": 0.4728,
+      "step": 8607
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8038261347112179,
+      "learning_rate": 3.1250280471906148e-09,
+      "loss": 0.5022,
+      "step": 8608
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9005098232792654,
+      "learning_rate": 3.059590921735689e-09,
+      "loss": 0.4903,
+      "step": 8609
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7121184255497983,
+      "learning_rate": 2.994845971601157e-09,
+      "loss": 0.5057,
+      "step": 8610
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8829936499192288,
+      "learning_rate": 2.930793205754845e-09,
+      "loss": 0.4401,
+      "step": 8611
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2458740811960114,
+      "learning_rate": 2.8674326330691005e-09,
+      "loss": 0.4826,
+      "step": 8612
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5255337539533813,
+      "learning_rate": 2.804764262321347e-09,
+      "loss": 0.3954,
+      "step": 8613
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.073872778583016,
+      "learning_rate": 2.7427881021918624e-09,
+      "loss": 0.4575,
+      "step": 8614
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7573421244613352,
+      "learning_rate": 2.6815041612665573e-09,
+      "loss": 0.4923,
+      "step": 8615
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1072704686075476,
+      "learning_rate": 2.6209124480330863e-09,
+      "loss": 0.4179,
+      "step": 8616
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.339779565883222,
+      "learning_rate": 2.5610129708852903e-09,
+      "loss": 0.467,
+      "step": 8617
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.976507025797458,
+      "learning_rate": 2.501805738120422e-09,
+      "loss": 0.4968,
+      "step": 8618
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7842495301999894,
+      "learning_rate": 2.443290757940253e-09,
+      "loss": 0.4336,
+      "step": 8619
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0513403546478757,
+      "learning_rate": 2.3854680384494123e-09,
+      "loss": 0.5684,
+      "step": 8620
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0306113134590316,
+      "learning_rate": 2.3283375876581583e-09,
+      "loss": 0.471,
+      "step": 8621
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9422893301731392,
+      "learning_rate": 2.2718994134796056e-09,
+      "loss": 0.4196,
+      "step": 8622
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8288777836366124,
+      "learning_rate": 2.2161535237319453e-09,
+      "loss": 0.4639,
+      "step": 8623
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0518676659970074,
+      "learning_rate": 2.1610999261373335e-09,
+      "loss": 0.5486,
+      "step": 8624
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0143869195850015,
+      "learning_rate": 2.1067386283213366e-09,
+      "loss": 0.504,
+      "step": 8625
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1311116119542324,
+      "learning_rate": 2.0530696378145974e-09,
+      "loss": 0.5689,
+      "step": 8626
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.101204366982583,
+      "learning_rate": 2.000092962050615e-09,
+      "loss": 0.6446,
+      "step": 8627
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5468030815408875,
+      "learning_rate": 1.9478086083679627e-09,
+      "loss": 0.5374,
+      "step": 8628
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.361396273221522,
+      "learning_rate": 1.8962165840097356e-09,
+      "loss": 0.5044,
+      "step": 8629
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7665549740058983,
+      "learning_rate": 1.8453168961213297e-09,
+      "loss": 0.4604,
+      "step": 8630
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9173112353057813,
+      "learning_rate": 1.795109551754326e-09,
+      "loss": 0.4643,
+      "step": 8631
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8942448814934543,
+      "learning_rate": 1.7455945578626065e-09,
+      "loss": 0.6472,
+      "step": 8632
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.639079680908984,
+      "learning_rate": 1.6967719213056843e-09,
+      "loss": 0.4571,
+      "step": 8633
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.975131564195945,
+      "learning_rate": 1.6486416488459279e-09,
+      "loss": 0.5866,
+      "step": 8634
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8141308878069986,
+      "learning_rate": 1.6012037471507813e-09,
+      "loss": 0.4341,
+      "step": 8635
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.029796067609798,
+      "learning_rate": 1.554458222791655e-09,
+      "loss": 0.4774,
+      "step": 8636
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.590206471907889,
+      "learning_rate": 1.5084050822422591e-09,
+      "loss": 0.5284,
+      "step": 8637
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9410480869798183,
+      "learning_rate": 1.4630443318836007e-09,
+      "loss": 0.4055,
+      "step": 8638
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7601269585994348,
+      "learning_rate": 1.4183759779978768e-09,
+      "loss": 0.481,
+      "step": 8639
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.083915156751448,
+      "learning_rate": 1.3744000267729152e-09,
+      "loss": 0.4667,
+      "step": 8640
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0299591777780366,
+      "learning_rate": 1.3311164843005098e-09,
+      "loss": 0.5181,
+      "step": 8641
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8870151749360664,
+      "learning_rate": 1.2885253565758649e-09,
+      "loss": 0.5497,
+      "step": 8642
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0456848978654145,
+      "learning_rate": 1.2466266494987056e-09,
+      "loss": 0.5793,
+      "step": 8643
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.932323656091851,
+      "learning_rate": 1.205420368873278e-09,
+      "loss": 0.4592,
+      "step": 8644
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9983708576039143,
+      "learning_rate": 1.1649065204072385e-09,
+      "loss": 0.5058,
+      "step": 8645
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.364473602530412,
+      "learning_rate": 1.1250851097122096e-09,
+      "loss": 0.4603,
+      "step": 8646
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7044245168971675,
+      "learning_rate": 1.0859561423048892e-09,
+      "loss": 0.4447,
+      "step": 8647
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4729336592649127,
+      "learning_rate": 1.0475196236053863e-09,
+      "loss": 0.5199,
+      "step": 8648
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8966801209688295,
+      "learning_rate": 1.0097755589372204e-09,
+      "loss": 0.7169,
+      "step": 8649
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.161874786591972,
+      "learning_rate": 9.727239535289867e-10,
+      "loss": 0.5136,
+      "step": 8650
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9354472703661458,
+      "learning_rate": 9.363648125132462e-10,
+      "loss": 0.5284,
+      "step": 8651
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6412937863737416,
+      "learning_rate": 9.006981409265258e-10,
+      "loss": 0.5027,
+      "step": 8652
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.2178850678866944,
+      "learning_rate": 8.657239437087628e-10,
+      "loss": 0.4669,
+      "step": 8653
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.963727970450669,
+      "learning_rate": 8.314422257055255e-10,
+      "loss": 0.5062,
+      "step": 8654
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3062782219734808,
+      "learning_rate": 7.978529916646827e-10,
+      "loss": 0.4759,
+      "step": 8655
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8441119099354049,
+      "learning_rate": 7.649562462397342e-10,
+      "loss": 0.383,
+      "step": 8656
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.449801381717954,
+      "learning_rate": 7.327519939870353e-10,
+      "loss": 0.4437,
+      "step": 8657
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.127470098160178,
+      "learning_rate": 7.012402393674622e-10,
+      "loss": 0.4213,
+      "step": 8658
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.183668209647357,
+      "learning_rate": 6.704209867464117e-10,
+      "loss": 0.4882,
+      "step": 8659
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.789568882209908,
+      "learning_rate": 6.402942403926915e-10,
+      "loss": 0.5107,
+      "step": 8660
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9040750708214508,
+      "learning_rate": 6.108600044796298e-10,
+      "loss": 0.6958,
+      "step": 8661
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.732296919489568,
+      "learning_rate": 5.821182830839655e-10,
+      "loss": 0.4526,
+      "step": 8662
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9694283481010917,
+      "learning_rate": 5.540690801875137e-10,
+      "loss": 0.6215,
+      "step": 8663
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8392061804814037,
+      "learning_rate": 5.267123996754997e-10,
+      "loss": 0.399,
+      "step": 8664
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9209099817699458,
+      "learning_rate": 5.000482453376698e-10,
+      "loss": 0.4623,
+      "step": 8665
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.402210225928313,
+      "learning_rate": 4.740766208666259e-10,
+      "loss": 0.4203,
+      "step": 8666
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9846695534012289,
+      "learning_rate": 4.487975298606007e-10,
+      "loss": 0.5537,
+      "step": 8667
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.186052721466957,
+      "learning_rate": 4.242109758217927e-10,
+      "loss": 0.4282,
+      "step": 8668
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.9878835013958995,
+      "learning_rate": 4.0031696215470093e-10,
+      "loss": 0.4387,
+      "step": 8669
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9925442771888353,
+      "learning_rate": 3.771154921700104e-10,
+      "loss": 0.4824,
+      "step": 8670
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.021271178237182,
+      "learning_rate": 3.5460656908126166e-10,
+      "loss": 0.4705,
+      "step": 8671
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.3519819289165538,
+      "learning_rate": 3.327901960065161e-10,
+      "loss": 0.5725,
+      "step": 8672
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2437279941984687,
+      "learning_rate": 3.116663759678007e-10,
+      "loss": 0.4309,
+      "step": 8673
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1463365593084553,
+      "learning_rate": 2.9123511189110833e-10,
+      "loss": 0.5002,
+      "step": 8674
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.106139316077505,
+      "learning_rate": 2.714964066063974e-10,
+      "loss": 0.5088,
+      "step": 8675
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2066937703640463,
+      "learning_rate": 2.524502628475922e-10,
+      "loss": 0.55,
+      "step": 8676
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9069622265745285,
+      "learning_rate": 2.3409668325424795e-10,
+      "loss": 0.4362,
+      "step": 8677
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8819084356917204,
+      "learning_rate": 2.1643567036711e-10,
+      "loss": 0.4415,
+      "step": 8678
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.715773637521027,
+      "learning_rate": 1.9946722663366502e-10,
+      "loss": 0.4657,
+      "step": 8679
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7911372347873028,
+      "learning_rate": 1.8319135440425517e-10,
+      "loss": 0.6461,
+      "step": 8680
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.4677066927749842,
+      "learning_rate": 1.676080559326332e-10,
+      "loss": 0.4878,
+      "step": 8681
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.3928213531300107,
+      "learning_rate": 1.5271733337818284e-10,
+      "loss": 0.5157,
+      "step": 8682
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.4463703707380082,
+      "learning_rate": 1.3851918880369853e-10,
+      "loss": 0.4726,
+      "step": 8683
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.416832537824428,
+      "learning_rate": 1.250136241748301e-10,
+      "loss": 0.4753,
+      "step": 8684
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9745248537301268,
+      "learning_rate": 1.1220064136341358e-10,
+      "loss": 0.5421,
+      "step": 8685
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.559466389901249,
+      "learning_rate": 1.0008024214414048e-10,
+      "loss": 0.4686,
+      "step": 8686
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.4318864684333654,
+      "learning_rate": 8.865242819566799e-11,
+      "loss": 0.4927,
+      "step": 8687
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9268296621983754,
+      "learning_rate": 7.791720110117417e-11,
+      "loss": 0.5602,
+      "step": 8688
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.861678464172052,
+      "learning_rate": 6.787456234724765e-11,
+      "loss": 0.5594,
+      "step": 8689
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.290861302156981,
+      "learning_rate": 5.852451332555298e-11,
+      "loss": 0.4977,
+      "step": 8690
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.055235885134363,
+      "learning_rate": 4.98670553306102e-11,
+      "loss": 0.5084,
+      "step": 8691
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7596783179811035,
+      "learning_rate": 4.1902189562570416e-11,
+      "loss": 0.4557,
+      "step": 8692
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.408393892082615,
+      "learning_rate": 3.462991712388508e-11,
+      "loss": 0.575,
+      "step": 8693
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2439517753114413,
+      "learning_rate": 2.8050239022636704e-11,
+      "loss": 0.5415,
+      "step": 8694
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2654484288330323,
+      "learning_rate": 2.2163156169208167e-11,
+      "loss": 0.498,
+      "step": 8695
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8332978729276281,
+      "learning_rate": 1.69686693801685e-11,
+      "loss": 0.6542,
+      "step": 8696
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.189068784614712,
+      "learning_rate": 1.2466779374942228e-11,
+      "loss": 0.4562,
+      "step": 8697
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.9263971122915615,
+      "learning_rate": 8.657486776364465e-12,
+      "loss": 0.4561,
+      "step": 8698
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.4445107736609613,
+      "learning_rate": 5.5407921123462606e-12,
+      "loss": 0.4447,
+      "step": 8699
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.923642959262328,
+      "learning_rate": 3.1166958153194813e-12,
+      "loss": 0.537,
+      "step": 8700
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9357179395376982,
+      "learning_rate": 1.3851982200163705e-12,
+      "loss": 0.425,
+      "step": 8701
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.621611557708318,
+      "learning_rate": 3.4629956680021226e-13,
+      "loss": 0.4618,
+      "step": 8702
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3558951288903296,
+      "learning_rate": 0.0,
+      "loss": 0.6268,
+      "step": 8703
+    },
+    {
+      "epoch": 1.0,
+      "step": 8703,
+      "total_flos": 2814559681249280.0,
+      "train_loss": 0.5202090283186973,
+      "train_runtime": 33769.6613,
+      "train_samples_per_second": 32.988,
+      "train_steps_per_second": 0.258
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 8703,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 871,
+  "total_flos": 2814559681249280.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}