diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,41468 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 5918,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.829396931800074,
+      "learning_rate": 1.1235955056179776e-07,
+      "loss": 1.5548,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.516971777409553,
+      "learning_rate": 2.247191011235955e-07,
+      "loss": 1.8127,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.590770270958929,
+      "learning_rate": 3.3707865168539325e-07,
+      "loss": 1.4611,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.768343670691037,
+      "learning_rate": 4.49438202247191e-07,
+      "loss": 1.5073,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.78901644011542,
+      "learning_rate": 5.617977528089888e-07,
+      "loss": 1.4796,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.83975703444064,
+      "learning_rate": 6.741573033707865e-07,
+      "loss": 1.4958,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.317733030464851,
+      "learning_rate": 7.865168539325843e-07,
+      "loss": 1.4777,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.054361671835247,
+      "learning_rate": 8.98876404494382e-07,
+      "loss": 1.5439,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.52856266455307,
+      "learning_rate": 1.01123595505618e-06,
+      "loss": 1.5528,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.667682361111382,
+      "learning_rate": 1.1235955056179777e-06,
+      "loss": 1.5386,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.78154318653174,
+      "learning_rate": 1.2359550561797752e-06,
+      "loss": 1.793,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.507583771743644,
+      "learning_rate": 1.348314606741573e-06,
+      "loss": 1.4273,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.053017431282507,
+      "learning_rate": 1.4606741573033708e-06,
+      "loss": 1.417,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.42789838028838,
+      "learning_rate": 1.5730337078651686e-06,
+      "loss": 1.3666,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.172122863088601,
+      "learning_rate": 1.6853932584269663e-06,
+      "loss": 1.0806,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.247845821285938,
+      "learning_rate": 1.797752808988764e-06,
+      "loss": 1.7734,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.796109371614374,
+      "learning_rate": 1.910112359550562e-06,
+      "loss": 1.1297,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.052419129473575,
+      "learning_rate": 2.02247191011236e-06,
+      "loss": 1.32,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7001189171155815,
+      "learning_rate": 2.1348314606741574e-06,
+      "loss": 1.1795,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.21680267249169,
+      "learning_rate": 2.2471910112359554e-06,
+      "loss": 1.1083,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.929455577488571,
+      "learning_rate": 2.359550561797753e-06,
+      "loss": 1.5693,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.1963975230547588,
+      "learning_rate": 2.4719101123595505e-06,
+      "loss": 1.0585,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.094028137475137,
+      "learning_rate": 2.584269662921349e-06,
+      "loss": 1.2155,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.631304043212987,
+      "learning_rate": 2.696629213483146e-06,
+      "loss": 0.9481,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.1821779229847578,
+      "learning_rate": 2.8089887640449444e-06,
+      "loss": 1.569,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.725433023668504,
+      "learning_rate": 2.9213483146067416e-06,
+      "loss": 1.2071,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.725395169990188,
+      "learning_rate": 3.03370786516854e-06,
+      "loss": 1.199,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.9478138693190954,
+      "learning_rate": 3.146067415730337e-06,
+      "loss": 1.1506,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.384560570435504,
+      "learning_rate": 3.258426966292135e-06,
+      "loss": 1.1347,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2266735427226214,
+      "learning_rate": 3.3707865168539327e-06,
+      "loss": 1.4395,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.778435703466326,
+      "learning_rate": 3.4831460674157306e-06,
+      "loss": 1.4233,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.428564623038082,
+      "learning_rate": 3.595505617977528e-06,
+      "loss": 0.9917,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.933911470055859,
+      "learning_rate": 3.707865168539326e-06,
+      "loss": 1.3218,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7074510974448422,
+      "learning_rate": 3.820224719101124e-06,
+      "loss": 0.965,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2605300339075405,
+      "learning_rate": 3.932584269662922e-06,
+      "loss": 1.3953,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.943901447361913,
+      "learning_rate": 4.04494382022472e-06,
+      "loss": 1.1127,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9141178075776173,
+      "learning_rate": 4.157303370786518e-06,
+      "loss": 1.0154,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.861854976130239,
+      "learning_rate": 4.269662921348315e-06,
+      "loss": 1.1485,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6716105026854176,
+      "learning_rate": 4.382022471910113e-06,
+      "loss": 0.9676,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3939718290017167,
+      "learning_rate": 4.494382022471911e-06,
+      "loss": 1.4554,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.311235705308055,
+      "learning_rate": 4.606741573033709e-06,
+      "loss": 1.1512,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.38618805515742,
+      "learning_rate": 4.719101123595506e-06,
+      "loss": 0.9749,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9030722297997926,
+      "learning_rate": 4.831460674157304e-06,
+      "loss": 1.1892,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.845725894881137,
+      "learning_rate": 4.943820224719101e-06,
+      "loss": 0.9155,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4127423426422925,
+      "learning_rate": 5.0561797752809e-06,
+      "loss": 1.5184,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5258253743953207,
+      "learning_rate": 5.168539325842698e-06,
+      "loss": 0.8955,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.066442006134885,
+      "learning_rate": 5.280898876404494e-06,
+      "loss": 1.0385,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2960932254642423,
+      "learning_rate": 5.393258426966292e-06,
+      "loss": 1.0024,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5183391037652227,
+      "learning_rate": 5.50561797752809e-06,
+      "loss": 1.5389,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8301828937698474,
+      "learning_rate": 5.617977528089889e-06,
+      "loss": 1.0392,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8523808241078639,
+      "learning_rate": 5.730337078651685e-06,
+      "loss": 1.2283,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.932212324413878,
+      "learning_rate": 5.842696629213483e-06,
+      "loss": 1.0494,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4226274648453052,
+      "learning_rate": 5.955056179775281e-06,
+      "loss": 1.0451,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2469242429767275,
+      "learning_rate": 6.06741573033708e-06,
+      "loss": 1.3902,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4995842771420063,
+      "learning_rate": 6.179775280898876e-06,
+      "loss": 1.0139,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5520998798411278,
+      "learning_rate": 6.292134831460674e-06,
+      "loss": 1.0145,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0034841671962758,
+      "learning_rate": 6.404494382022472e-06,
+      "loss": 1.1227,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.076939043080104,
+      "learning_rate": 6.51685393258427e-06,
+      "loss": 1.1047,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.457743002175606,
+      "learning_rate": 6.629213483146067e-06,
+      "loss": 1.4766,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 17.49251958162293,
+      "learning_rate": 6.741573033707865e-06,
+      "loss": 1.0465,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.365376341178095,
+      "learning_rate": 6.853932584269663e-06,
+      "loss": 1.1408,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5964580331037521,
+      "learning_rate": 6.966292134831461e-06,
+      "loss": 0.8462,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4704355508101794,
+      "learning_rate": 7.078651685393258e-06,
+      "loss": 0.9103,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5555177485386862,
+      "learning_rate": 7.191011235955056e-06,
+      "loss": 1.4831,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8210194749546043,
+      "learning_rate": 7.303370786516854e-06,
+      "loss": 1.0916,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9611524740066,
+      "learning_rate": 7.415730337078652e-06,
+      "loss": 1.0493,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7226444270327106,
+      "learning_rate": 7.5280898876404495e-06,
+      "loss": 0.9902,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.847376469584565,
+      "learning_rate": 7.640449438202247e-06,
+      "loss": 1.3334,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3530378627809325,
+      "learning_rate": 7.752808988764046e-06,
+      "loss": 1.0025,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5358338022274638,
+      "learning_rate": 7.865168539325843e-06,
+      "loss": 0.9744,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8845017379575286,
+      "learning_rate": 7.97752808988764e-06,
+      "loss": 0.9376,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4449428212559188,
+      "learning_rate": 8.08988764044944e-06,
+      "loss": 0.9849,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9037420632109663,
+      "learning_rate": 8.202247191011237e-06,
+      "loss": 1.1507,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6045758697698378,
+      "learning_rate": 8.314606741573035e-06,
+      "loss": 1.1531,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2143670800533726,
+      "learning_rate": 8.426966292134832e-06,
+      "loss": 0.9114,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 32.096933020844055,
+      "learning_rate": 8.53932584269663e-06,
+      "loss": 2.3205,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.383765608323947,
+      "learning_rate": 8.651685393258428e-06,
+      "loss": 0.832,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6731199349917603,
+      "learning_rate": 8.764044943820226e-06,
+      "loss": 1.3254,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.389306508250511,
+      "learning_rate": 8.876404494382023e-06,
+      "loss": 0.9687,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6608026104635754,
+      "learning_rate": 8.988764044943822e-06,
+      "loss": 1.0614,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.385451729880962,
+      "learning_rate": 9.101123595505619e-06,
+      "loss": 1.1343,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.743414619800334,
+      "learning_rate": 9.213483146067417e-06,
+      "loss": 1.3608,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5457516196381953,
+      "learning_rate": 9.325842696629213e-06,
+      "loss": 1.0032,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8857684402182364,
+      "learning_rate": 9.438202247191012e-06,
+      "loss": 1.0283,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4511844969254628,
+      "learning_rate": 9.55056179775281e-06,
+      "loss": 1.0537,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2951630889643375,
+      "learning_rate": 9.662921348314608e-06,
+      "loss": 0.8699,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3123278230241777,
+      "learning_rate": 9.775280898876405e-06,
+      "loss": 1.3224,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3697170429954246,
+      "learning_rate": 9.887640449438202e-06,
+      "loss": 1.1539,
+      "step": 88
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.356025438692696,
+      "learning_rate": 1e-05,
+      "loss": 0.9567,
+      "step": 89
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0831689381700964,
+      "learning_rate": 1.01123595505618e-05,
+      "loss": 0.8364,
+      "step": 90
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5839706818515051,
+      "learning_rate": 1.0224719101123597e-05,
+      "loss": 0.8975,
+      "step": 91
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.281334762822879,
+      "learning_rate": 1.0337078651685396e-05,
+      "loss": 1.3108,
+      "step": 92
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6684359447285948,
+      "learning_rate": 1.0449438202247193e-05,
+      "loss": 0.9475,
+      "step": 93
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2965380879098833,
+      "learning_rate": 1.0561797752808988e-05,
+      "loss": 1.0208,
+      "step": 94
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5247896955401283,
+      "learning_rate": 1.0674157303370787e-05,
+      "loss": 1.2342,
+      "step": 95
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 18.81664314481584,
+      "learning_rate": 1.0786516853932584e-05,
+      "loss": 1.8062,
+      "step": 96
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5474554143824053,
+      "learning_rate": 1.0898876404494383e-05,
+      "loss": 1.3687,
+      "step": 97
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.031194602131902,
+      "learning_rate": 1.101123595505618e-05,
+      "loss": 0.8053,
+      "step": 98
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0225516193755158,
+      "learning_rate": 1.1123595505617979e-05,
+      "loss": 1.0457,
+      "step": 99
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.693373187338537,
+      "learning_rate": 1.1235955056179778e-05,
+      "loss": 0.8615,
+      "step": 100
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5874219911904255,
+      "learning_rate": 1.1348314606741575e-05,
+      "loss": 1.2479,
+      "step": 101
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.605449447268845,
+      "learning_rate": 1.146067415730337e-05,
+      "loss": 0.9936,
+      "step": 102
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7761669635500132,
+      "learning_rate": 1.157303370786517e-05,
+      "loss": 0.8811,
+      "step": 103
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.738769121056984,
+      "learning_rate": 1.1685393258426966e-05,
+      "loss": 0.9796,
+      "step": 104
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6508798803871119,
+      "learning_rate": 1.1797752808988765e-05,
+      "loss": 1.1147,
+      "step": 105
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3652623668206993,
+      "learning_rate": 1.1910112359550562e-05,
+      "loss": 1.2136,
+      "step": 106
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3671605068267123,
+      "learning_rate": 1.2022471910112361e-05,
+      "loss": 1.0412,
+      "step": 107
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2327810890406763,
+      "learning_rate": 1.213483146067416e-05,
+      "loss": 0.9264,
+      "step": 108
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7943512054587372,
+      "learning_rate": 1.2247191011235957e-05,
+      "loss": 1.0778,
+      "step": 109
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3989793914856785,
+      "learning_rate": 1.2359550561797752e-05,
+      "loss": 0.9234,
+      "step": 110
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.59189560497417,
+      "learning_rate": 1.2471910112359551e-05,
+      "loss": 1.3043,
+      "step": 111
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5653163217407942,
+      "learning_rate": 1.2584269662921348e-05,
+      "loss": 1.0049,
+      "step": 112
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7281754278673596,
+      "learning_rate": 1.2696629213483147e-05,
+      "loss": 1.1007,
+      "step": 113
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 28.985471287164113,
+      "learning_rate": 1.2808988764044944e-05,
+      "loss": 1.6577,
+      "step": 114
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5583720570274067,
+      "learning_rate": 1.2921348314606743e-05,
+      "loss": 1.1837,
+      "step": 115
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4721188832616692,
+      "learning_rate": 1.303370786516854e-05,
+      "loss": 1.2842,
+      "step": 116
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0606414394441224,
+      "learning_rate": 1.314606741573034e-05,
+      "loss": 0.7941,
+      "step": 117
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0305580498170954,
+      "learning_rate": 1.3258426966292135e-05,
+      "loss": 1.0888,
+      "step": 118
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4462457261518151,
+      "learning_rate": 1.3370786516853933e-05,
+      "loss": 0.9882,
+      "step": 119
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4580351513135303,
+      "learning_rate": 1.348314606741573e-05,
+      "loss": 1.0356,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.521484068750081,
+      "learning_rate": 1.359550561797753e-05,
+      "loss": 1.1753,
+      "step": 121
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5539984461692609,
+      "learning_rate": 1.3707865168539327e-05,
+      "loss": 0.7949,
+      "step": 122
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1771280927673604,
+      "learning_rate": 1.3820224719101125e-05,
+      "loss": 0.7838,
+      "step": 123
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.219353264161316,
+      "learning_rate": 1.3932584269662923e-05,
+      "loss": 0.9496,
+      "step": 124
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7794684464114083,
+      "learning_rate": 1.4044943820224721e-05,
+      "loss": 1.3437,
+      "step": 125
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8218363034167901,
+      "learning_rate": 1.4157303370786517e-05,
+      "loss": 1.0711,
+      "step": 126
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6287967727501695,
+      "learning_rate": 1.4269662921348316e-05,
+      "loss": 1.0789,
+      "step": 127
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8364712170846056,
+      "learning_rate": 1.4382022471910113e-05,
+      "loss": 0.9655,
+      "step": 128
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.87335076180381,
+      "learning_rate": 1.4494382022471912e-05,
+      "loss": 1.0812,
+      "step": 129
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.512686453127762,
+      "learning_rate": 1.4606741573033709e-05,
+      "loss": 1.0521,
+      "step": 130
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.735239976262144,
+      "learning_rate": 1.4719101123595508e-05,
+      "loss": 1.2121,
+      "step": 131
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4124870960378908,
+      "learning_rate": 1.4831460674157305e-05,
+      "loss": 1.0227,
+      "step": 132
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.255116049973962,
+      "learning_rate": 1.4943820224719104e-05,
+      "loss": 1.3034,
+      "step": 133
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.64247654089158,
+      "learning_rate": 1.5056179775280899e-05,
+      "loss": 0.797,
+      "step": 134
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5037639945966563,
+      "learning_rate": 1.5168539325842698e-05,
+      "loss": 1.2136,
+      "step": 135
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9648122479144003,
+      "learning_rate": 1.5280898876404495e-05,
+      "loss": 1.1044,
+      "step": 136
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9603286091116447,
+      "learning_rate": 1.5393258426966292e-05,
+      "loss": 0.726,
+      "step": 137
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4042005025076683,
+      "learning_rate": 1.5505617977528093e-05,
+      "loss": 1.1269,
+      "step": 138
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.506538280744306,
+      "learning_rate": 1.561797752808989e-05,
+      "loss": 0.9676,
+      "step": 139
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2448972270491265,
+      "learning_rate": 1.5730337078651687e-05,
+      "loss": 1.3881,
+      "step": 140
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4766835802384952,
+      "learning_rate": 1.5842696629213484e-05,
+      "loss": 0.9446,
+      "step": 141
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5013954289896487,
+      "learning_rate": 1.595505617977528e-05,
+      "loss": 0.9341,
+      "step": 142
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4560376426853254,
+      "learning_rate": 1.6067415730337078e-05,
+      "loss": 0.9022,
+      "step": 143
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7453704222413267,
+      "learning_rate": 1.617977528089888e-05,
+      "loss": 1.0848,
+      "step": 144
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5894656041792834,
+      "learning_rate": 1.6292134831460676e-05,
+      "loss": 1.2158,
+      "step": 145
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.67442496955442,
+      "learning_rate": 1.6404494382022473e-05,
+      "loss": 0.9961,
+      "step": 146
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.64538255059252,
+      "learning_rate": 1.651685393258427e-05,
+      "loss": 1.0368,
+      "step": 147
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.705528840931898,
+      "learning_rate": 1.662921348314607e-05,
+      "loss": 0.7592,
+      "step": 148
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5073974553616154,
+      "learning_rate": 1.6741573033707868e-05,
+      "loss": 1.2925,
+      "step": 149
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3925174774038767,
+      "learning_rate": 1.6853932584269665e-05,
+      "loss": 0.9366,
+      "step": 150
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 22.58387917126523,
+      "learning_rate": 1.6966292134831462e-05,
+      "loss": 1.9268,
+      "step": 151
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.209090018884432,
+      "learning_rate": 1.707865168539326e-05,
+      "loss": 0.7803,
+      "step": 152
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.258368047391059,
+      "learning_rate": 1.7191011235955056e-05,
+      "loss": 0.7738,
+      "step": 153
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3642366217441726,
+      "learning_rate": 1.7303370786516857e-05,
+      "loss": 1.0992,
+      "step": 154
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4853313571077384,
+      "learning_rate": 1.7415730337078654e-05,
+      "loss": 0.8179,
+      "step": 155
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.709625755245977,
+      "learning_rate": 1.752808988764045e-05,
+      "loss": 1.0694,
+      "step": 156
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5289352741398181,
+      "learning_rate": 1.7640449438202248e-05,
+      "loss": 1.0483,
+      "step": 157
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1115316735636456,
+      "learning_rate": 1.7752808988764045e-05,
+      "loss": 0.9681,
+      "step": 158
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4633817954558572,
+      "learning_rate": 1.7865168539325843e-05,
+      "loss": 1.2534,
+      "step": 159
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0109698514259686,
+      "learning_rate": 1.7977528089887643e-05,
+      "loss": 0.8203,
+      "step": 160
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9941767668344823,
+      "learning_rate": 1.808988764044944e-05,
+      "loss": 1.1229,
+      "step": 161
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.302989407505049,
+      "learning_rate": 1.8202247191011237e-05,
+      "loss": 0.9704,
+      "step": 162
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6186660824820323,
+      "learning_rate": 1.8314606741573034e-05,
+      "loss": 1.3981,
+      "step": 163
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2224890805183555,
+      "learning_rate": 1.8426966292134835e-05,
+      "loss": 0.9906,
+      "step": 164
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7737371602395169,
+      "learning_rate": 1.8539325842696632e-05,
+      "loss": 0.8477,
+      "step": 165
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5700545955401723,
+      "learning_rate": 1.8651685393258426e-05,
+      "loss": 0.8163,
+      "step": 166
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4229472568685633,
+      "learning_rate": 1.8764044943820226e-05,
+      "loss": 0.861,
+      "step": 167
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 24.652909847127408,
+      "learning_rate": 1.8876404494382024e-05,
+      "loss": 1.5394,
+      "step": 168
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.270866004395781,
+      "learning_rate": 1.898876404494382e-05,
+      "loss": 1.0602,
+      "step": 169
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7773585706627122,
+      "learning_rate": 1.910112359550562e-05,
+      "loss": 1.0997,
+      "step": 170
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.433636339501709,
+      "learning_rate": 1.921348314606742e-05,
+      "loss": 0.858,
+      "step": 171
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.208416039391502,
+      "learning_rate": 1.9325842696629215e-05,
+      "loss": 0.9658,
+      "step": 172
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7778086389289216,
+      "learning_rate": 1.9438202247191013e-05,
+      "loss": 1.3339,
+      "step": 173
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5906913393095496,
+      "learning_rate": 1.955056179775281e-05,
+      "loss": 0.8788,
+      "step": 174
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7610542776529152,
+      "learning_rate": 1.9662921348314607e-05,
+      "loss": 0.8365,
+      "step": 175
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0186395252027536,
+      "learning_rate": 1.9775280898876404e-05,
+      "loss": 1.0868,
+      "step": 176
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5268193484060335,
+      "learning_rate": 1.9887640449438205e-05,
+      "loss": 1.1386,
+      "step": 177
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3221773075747048,
+      "learning_rate": 2e-05,
+      "loss": 1.2162,
+      "step": 178
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6955506118206767,
+      "learning_rate": 1.9999998502227152e-05,
+      "loss": 0.7957,
+      "step": 179
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5549542091115767,
+      "learning_rate": 1.999999400890905e-05,
+      "loss": 1.176,
+      "step": 180
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4243693086233886,
+      "learning_rate": 1.999998652004704e-05,
+      "loss": 0.8635,
+      "step": 181
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7059121780742876,
+      "learning_rate": 1.9999976035643366e-05,
+      "loss": 0.903,
+      "step": 182
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.373187164813608,
+      "learning_rate": 1.999996255570117e-05,
+      "loss": 1.3579,
+      "step": 183
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.867590906603124,
+      "learning_rate": 1.9999946080224487e-05,
+      "loss": 1.0199,
+      "step": 184
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4972400866671742,
+      "learning_rate": 1.999992660921826e-05,
+      "loss": 0.9304,
+      "step": 185
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3705866797840702,
+      "learning_rate": 1.9999904142688313e-05,
+      "loss": 1.1003,
+      "step": 186
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 37.995141158752254,
+      "learning_rate": 1.999987868064138e-05,
+      "loss": 3.1558,
+      "step": 187
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2596711455269238,
+      "learning_rate": 1.9999850223085088e-05,
+      "loss": 1.0542,
+      "step": 188
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0994529767175214,
+      "learning_rate": 1.999981877002796e-05,
+      "loss": 1.0176,
+      "step": 189
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8712481356116821,
+      "learning_rate": 1.9999784321479416e-05,
+      "loss": 0.9255,
+      "step": 190
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7965291710594058,
+      "learning_rate": 1.9999746877449785e-05,
+      "loss": 0.9384,
+      "step": 191
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.724846605574325,
+      "learning_rate": 1.999970643795027e-05,
+      "loss": 1.2493,
+      "step": 192
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3098723098900862,
+      "learning_rate": 1.9999663002992994e-05,
+      "loss": 1.0143,
+      "step": 193
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5293653729556707,
+      "learning_rate": 1.999961657259097e-05,
+      "loss": 0.8454,
+      "step": 194
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4715252257456022,
+      "learning_rate": 1.9999567146758098e-05,
+      "loss": 0.852,
+      "step": 195
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7586193125138567,
+      "learning_rate": 1.9999514725509187e-05,
+      "loss": 1.0217,
+      "step": 196
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3769412129491618,
+      "learning_rate": 1.9999459308859944e-05,
+      "loss": 1.1751,
+      "step": 197
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5872133928979697,
+      "learning_rate": 1.9999400896826965e-05,
+      "loss": 0.8544,
+      "step": 198
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7239361707718608,
+      "learning_rate": 1.9999339489427746e-05,
+      "loss": 1.0572,
+      "step": 199
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7622725116094322,
+      "learning_rate": 1.9999275086680688e-05,
+      "loss": 1.0364,
+      "step": 200
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3907106420993856,
+      "learning_rate": 1.999920768860508e-05,
+      "loss": 0.9747,
+      "step": 201
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0848969907329264,
+      "learning_rate": 1.999913729522111e-05,
+      "loss": 1.0829,
+      "step": 202
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8672758533405887,
+      "learning_rate": 1.9999063906549868e-05,
+      "loss": 1.1763,
+      "step": 203
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.446479424998351,
+      "learning_rate": 1.9998987522613332e-05,
+      "loss": 1.271,
+      "step": 204
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1855792304439356,
+      "learning_rate": 1.999890814343439e-05,
+      "loss": 0.7418,
+      "step": 205
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8853056615279484,
+      "learning_rate": 1.999882576903682e-05,
+      "loss": 1.2966,
+      "step": 206
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4184491153888898,
+      "learning_rate": 1.999874039944529e-05,
+      "loss": 1.2179,
+      "step": 207
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6047489502884,
+      "learning_rate": 1.9998652034685377e-05,
+      "loss": 1.1141,
+      "step": 208
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.505908040781643,
+      "learning_rate": 1.9998560674783555e-05,
+      "loss": 1.0587,
+      "step": 209
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3119359720010848,
+      "learning_rate": 1.9998466319767188e-05,
+      "loss": 1.135,
+      "step": 210
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4757463309050862,
+      "learning_rate": 1.9998368969664538e-05,
+      "loss": 0.7186,
+      "step": 211
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4515512962731818,
+      "learning_rate": 1.999826862450477e-05,
+      "loss": 1.2565,
+      "step": 212
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2755421815061525,
+      "learning_rate": 1.9998165284317944e-05,
+      "loss": 0.9036,
+      "step": 213
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.256622779672448,
+      "learning_rate": 1.9998058949135016e-05,
+      "loss": 0.6448,
+      "step": 214
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6161020148000083,
+      "learning_rate": 1.9997949618987833e-05,
+      "loss": 1.0446,
+      "step": 215
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.128382383730108,
+      "learning_rate": 1.999783729390915e-05,
+      "loss": 1.215,
+      "step": 216
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1528860594137853,
+      "learning_rate": 1.999772197393262e-05,
+      "loss": 0.9004,
+      "step": 217
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2662517764806975,
+      "learning_rate": 1.9997603659092773e-05,
+      "loss": 0.8128,
+      "step": 218
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.02470402027771,
+      "learning_rate": 1.999748234942507e-05,
+      "loss": 0.7855,
+      "step": 219
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3233179875784584,
+      "learning_rate": 1.9997358044965833e-05,
+      "loss": 0.9379,
+      "step": 220
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1300727386010214,
+      "learning_rate": 1.9997230745752304e-05,
+      "loss": 1.1749,
+      "step": 221
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 14.98341124398601,
+      "learning_rate": 1.9997100451822616e-05,
+      "loss": 1.2896,
+      "step": 222
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5242101029721562,
+      "learning_rate": 1.99969671632158e-05,
+      "loss": 1.1588,
+      "step": 223
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0041522781363719,
+      "learning_rate": 1.999683087997179e-05,
+      "loss": 0.8305,
+      "step": 224
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.088985549772884,
+      "learning_rate": 1.9996691602131392e-05,
+      "loss": 1.2569,
+      "step": 225
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1680386696046339,
+      "learning_rate": 1.9996549329736345e-05,
+      "loss": 1.1073,
+      "step": 226
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1990966874270717,
+      "learning_rate": 1.999640406282926e-05,
+      "loss": 0.902,
+      "step": 227
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.052002891453125,
+      "learning_rate": 1.999625580145365e-05,
+      "loss": 0.8165,
+      "step": 228
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3636527986254094,
+      "learning_rate": 1.9996104545653936e-05,
+      "loss": 1.0738,
+      "step": 229
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2038106086094114,
+      "learning_rate": 1.9995950295475422e-05,
+      "loss": 1.2311,
+      "step": 230
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2106658265745807,
+      "learning_rate": 1.999579305096431e-05,
+      "loss": 0.9729,
+      "step": 231
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1820544947224139,
+      "learning_rate": 1.9995632812167705e-05,
+      "loss": 0.9009,
+      "step": 232
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.269277567656275,
+      "learning_rate": 1.9995469579133618e-05,
+      "loss": 1.1069,
+      "step": 233
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0307297597007754,
+      "learning_rate": 1.9995303351910934e-05,
+      "loss": 1.0899,
+      "step": 234
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.277949527506337,
+      "learning_rate": 1.999513413054945e-05,
+      "loss": 1.2014,
+      "step": 235
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1405574389609026,
+      "learning_rate": 1.999496191509986e-05,
+      "loss": 0.8923,
+      "step": 236
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1752932860051108,
+      "learning_rate": 1.9994786705613748e-05,
+      "loss": 0.9511,
+      "step": 237
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.355326771615186,
+      "learning_rate": 1.9994608502143603e-05,
+      "loss": 0.8583,
+      "step": 238
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2865851543658442,
+      "learning_rate": 1.99944273047428e-05,
+      "loss": 0.9502,
+      "step": 239
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 187.88282556328667,
+      "learning_rate": 1.9994243113465627e-05,
+      "loss": 6.3926,
+      "step": 240
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.617404294236305,
+      "learning_rate": 1.9994055928367256e-05,
+      "loss": 0.9502,
+      "step": 241
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5955823067739574,
+      "learning_rate": 1.9993865749503754e-05,
+      "loss": 0.981,
+      "step": 242
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8043636295439134,
+      "learning_rate": 1.9993672576932092e-05,
+      "loss": 0.8493,
+      "step": 243
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5337421434223057,
+      "learning_rate": 1.999347641071014e-05,
+      "loss": 0.9803,
+      "step": 244
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4763937452011506,
+      "learning_rate": 1.9993277250896655e-05,
+      "loss": 1.2279,
+      "step": 245
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.193262872272579,
+      "learning_rate": 1.9993075097551302e-05,
+      "loss": 0.9787,
+      "step": 246
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3110368690622125,
+      "learning_rate": 1.9992869950734633e-05,
+      "loss": 0.9518,
+      "step": 247
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.599380770725851,
+      "learning_rate": 1.99926618105081e-05,
+      "loss": 1.0319,
+      "step": 248
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.216089824400922,
+      "learning_rate": 1.999245067693406e-05,
+      "loss": 1.3898,
+      "step": 249
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4825333105507696,
+      "learning_rate": 1.9992236550075746e-05,
+      "loss": 1.0832,
+      "step": 250
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5521661783344707,
+      "learning_rate": 1.999201942999731e-05,
+      "loss": 0.762,
+      "step": 251
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3244152912115512,
+      "learning_rate": 1.9991799316763788e-05,
+      "loss": 0.9244,
+      "step": 252
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4968385096824235,
+      "learning_rate": 1.999157621044112e-05,
+      "loss": 0.929,
+      "step": 253
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.021730591952234,
+      "learning_rate": 1.9991350111096133e-05,
+      "loss": 0.818,
+      "step": 254
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4836279702051351,
+      "learning_rate": 1.999112101879656e-05,
+      "loss": 1.3846,
+      "step": 255
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1596396782707874,
+      "learning_rate": 1.999088893361102e-05,
+      "loss": 0.8763,
+      "step": 256
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2004945210900315,
+      "learning_rate": 1.9990653855609047e-05,
+      "loss": 1.0246,
+      "step": 257
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 26.295485397324157,
+      "learning_rate": 1.999041578486105e-05,
+      "loss": 1.9499,
+      "step": 258
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1597212298811908,
+      "learning_rate": 1.9990174721438344e-05,
+      "loss": 1.1629,
+      "step": 259
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.373235008383188,
+      "learning_rate": 1.9989930665413148e-05,
+      "loss": 1.0532,
+      "step": 260
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2793496373483033,
+      "learning_rate": 1.9989683616858567e-05,
+      "loss": 0.914,
+      "step": 261
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9439700283704964,
+      "learning_rate": 1.99894335758486e-05,
+      "loss": 1.0676,
+      "step": 262
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.971791703112418,
+      "learning_rate": 1.9989180542458155e-05,
+      "loss": 0.8099,
+      "step": 263
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5744645051744408,
+      "learning_rate": 1.9988924516763027e-05,
+      "loss": 1.2531,
+      "step": 264
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9636576512482757,
+      "learning_rate": 1.998866549883991e-05,
+      "loss": 0.6997,
+      "step": 265
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.653812301211742,
+      "learning_rate": 1.998840348876639e-05,
+      "loss": 0.9258,
+      "step": 266
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.331578300039459,
+      "learning_rate": 1.998813848662096e-05,
+      "loss": 0.9923,
+      "step": 267
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.115920713744814,
+      "learning_rate": 1.9987870492483e-05,
+      "loss": 0.9065,
+      "step": 268
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2176700476740965,
+      "learning_rate": 1.9987599506432785e-05,
+      "loss": 1.2162,
+      "step": 269
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.542894798463883,
+      "learning_rate": 1.9987325528551492e-05,
+      "loss": 1.1439,
+      "step": 270
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4793895525952916,
+      "learning_rate": 1.9987048558921197e-05,
+      "loss": 0.8953,
+      "step": 271
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5944552933763054,
+      "learning_rate": 1.9986768597624864e-05,
+      "loss": 1.1627,
+      "step": 272
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1686845757865625,
+      "learning_rate": 1.9986485644746356e-05,
+      "loss": 1.0505,
+      "step": 273
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1784982898373324,
+      "learning_rate": 1.9986199700370435e-05,
+      "loss": 1.1498,
+      "step": 274
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3568765116693577,
+      "learning_rate": 1.9985910764582752e-05,
+      "loss": 1.0208,
+      "step": 275
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 64.26357910810243,
+      "learning_rate": 1.9985618837469864e-05,
+      "loss": 2.1076,
+      "step": 276
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3547082555837229,
+      "learning_rate": 1.9985323919119218e-05,
+      "loss": 0.8317,
+      "step": 277
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.488343976020008,
+      "learning_rate": 1.998502600961916e-05,
+      "loss": 1.1062,
+      "step": 278
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.64182954754031,
+      "learning_rate": 1.9984725109058928e-05,
+      "loss": 1.17,
+      "step": 279
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4019133387932097,
+      "learning_rate": 1.9984421217528654e-05,
+      "loss": 0.8603,
+      "step": 280
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8295550057717342,
+      "learning_rate": 1.998411433511938e-05,
+      "loss": 1.0859,
+      "step": 281
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6910243680727952,
+      "learning_rate": 1.9983804461923022e-05,
+      "loss": 0.869,
+      "step": 282
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.542214806025653,
+      "learning_rate": 1.998349159803241e-05,
+      "loss": 1.1486,
+      "step": 283
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.411733649459252,
+      "learning_rate": 1.998317574354127e-05,
+      "loss": 0.771,
+      "step": 284
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3604472181008787,
+      "learning_rate": 1.998285689854421e-05,
+      "loss": 0.9958,
+      "step": 285
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1462801972455663,
+      "learning_rate": 1.998253506313674e-05,
+      "loss": 0.9544,
+      "step": 286
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4238363798830878,
+      "learning_rate": 1.9982210237415272e-05,
+      "loss": 0.9432,
+      "step": 287
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6369916718310393,
+      "learning_rate": 1.9981882421477106e-05,
+      "loss": 1.3881,
+      "step": 288
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0682777353059096,
+      "learning_rate": 1.9981551615420446e-05,
+      "loss": 1.0973,
+      "step": 289
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1626760955242172,
+      "learning_rate": 1.998121781934438e-05,
+      "loss": 0.8515,
+      "step": 290
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2921380068219466,
+      "learning_rate": 1.9980881033348902e-05,
+      "loss": 1.2304,
+      "step": 291
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2367352039213346,
+      "learning_rate": 1.9980541257534894e-05,
+      "loss": 0.8595,
+      "step": 292
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4702906353499796,
+      "learning_rate": 1.9980198492004143e-05,
+      "loss": 1.2759,
+      "step": 293
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3918661085311506,
+      "learning_rate": 1.997985273685932e-05,
+      "loss": 0.8257,
+      "step": 294
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 48.79083498141921,
+      "learning_rate": 1.9979503992204002e-05,
+      "loss": 2.3253,
+      "step": 295
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.354700458625128,
+      "learning_rate": 1.9979152258142654e-05,
+      "loss": 1.0589,
+      "step": 296
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.954678340307447,
+      "learning_rate": 1.9978797534780646e-05,
+      "loss": 1.0719,
+      "step": 297
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.566276191433797,
+      "learning_rate": 1.9978439822224228e-05,
+      "loss": 1.19,
+      "step": 298
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7794397617787696,
+      "learning_rate": 1.997807912058056e-05,
+      "loss": 1.0805,
+      "step": 299
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3998078943851717,
+      "learning_rate": 1.997771542995769e-05,
+      "loss": 0.8571,
+      "step": 300
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2960565418835175,
+      "learning_rate": 1.997734875046456e-05,
+      "loss": 0.9686,
+      "step": 301
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5760965307134902,
+      "learning_rate": 1.997697908221102e-05,
+      "loss": 1.2054,
+      "step": 302
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9734897466292087,
+      "learning_rate": 1.9976606425307797e-05,
+      "loss": 0.8479,
+      "step": 303
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.709506105904352,
+      "learning_rate": 1.9976230779866527e-05,
+      "loss": 0.8967,
+      "step": 304
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2362190831776534,
+      "learning_rate": 1.997585214599973e-05,
+      "loss": 0.9185,
+      "step": 305
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4339519537496384,
+      "learning_rate": 1.9975470523820835e-05,
+      "loss": 0.9088,
+      "step": 306
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2076324284974118,
+      "learning_rate": 1.9975085913444152e-05,
+      "loss": 1.1328,
+      "step": 307
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9629430608982645,
+      "learning_rate": 1.99746983149849e-05,
+      "loss": 1.1269,
+      "step": 308
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6075308969873534,
+      "learning_rate": 1.9974307728559183e-05,
+      "loss": 0.9286,
+      "step": 309
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1605410367579245,
+      "learning_rate": 1.9973914154284e-05,
+      "loss": 0.9997,
+      "step": 310
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1317413243819772,
+      "learning_rate": 1.997351759227725e-05,
+      "loss": 0.7738,
+      "step": 311
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1489213853253362,
+      "learning_rate": 1.997311804265772e-05,
+      "loss": 1.3082,
+      "step": 312
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 34.92754662685241,
+      "learning_rate": 1.997271550554511e-05,
+      "loss": 1.7633,
+      "step": 313
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3495381515649465,
+      "learning_rate": 1.997230998105999e-05,
+      "loss": 1.3209,
+      "step": 314
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9983562505994962,
+      "learning_rate": 1.997190146932384e-05,
+      "loss": 0.8537,
+      "step": 315
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.241373952451485,
+      "learning_rate": 1.9971489970459034e-05,
+      "loss": 1.1973,
+      "step": 316
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4584462985594804,
+      "learning_rate": 1.9971075484588834e-05,
+      "loss": 0.9072,
+      "step": 317
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3322176547353186,
+      "learning_rate": 1.9970658011837404e-05,
+      "loss": 0.9479,
+      "step": 318
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5559988047760245,
+      "learning_rate": 1.9970237552329804e-05,
+      "loss": 1.1248,
+      "step": 319
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0740826972732813,
+      "learning_rate": 1.9969814106191973e-05,
+      "loss": 0.9359,
+      "step": 320
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0500406671173068,
+      "learning_rate": 1.9969387673550767e-05,
+      "loss": 1.2207,
+      "step": 321
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3294441104091472,
+      "learning_rate": 1.9968958254533922e-05,
+      "loss": 0.9256,
+      "step": 322
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1781754780453382,
+      "learning_rate": 1.9968525849270072e-05,
+      "loss": 1.1166,
+      "step": 323
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.342427970380265,
+      "learning_rate": 1.9968090457888748e-05,
+      "loss": 1.1046,
+      "step": 324
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1929059023775896,
+      "learning_rate": 1.996765208052037e-05,
+      "loss": 0.8626,
+      "step": 325
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1835607008593194,
+      "learning_rate": 1.9967210717296256e-05,
+      "loss": 1.2091,
+      "step": 326
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.775571089233626,
+      "learning_rate": 1.9966766368348622e-05,
+      "loss": 1.0572,
+      "step": 327
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0621298524349847,
+      "learning_rate": 1.9966319033810575e-05,
+      "loss": 0.8873,
+      "step": 328
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2355661688331485,
+      "learning_rate": 1.9965868713816112e-05,
+      "loss": 0.7819,
+      "step": 329
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.04340833904324,
+      "learning_rate": 1.996541540850013e-05,
+      "loss": 0.8238,
+      "step": 330
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0331870031903496,
+      "learning_rate": 1.996495911799842e-05,
+      "loss": 1.2059,
+      "step": 331
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 56.012992367883534,
+      "learning_rate": 1.9964499842447665e-05,
+      "loss": 2.891,
+      "step": 332
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9456671530822627,
+      "learning_rate": 1.9964037581985444e-05,
+      "loss": 1.0837,
+      "step": 333
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1546655004134045,
+      "learning_rate": 1.9963572336750227e-05,
+      "loss": 1.189,
+      "step": 334
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.20509353756158,
+      "learning_rate": 1.996310410688138e-05,
+      "loss": 0.7572,
+      "step": 335
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0293401701901028,
+      "learning_rate": 1.996263289251917e-05,
+      "loss": 1.277,
+      "step": 336
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1863317040513086,
+      "learning_rate": 1.9962158693804743e-05,
+      "loss": 0.867,
+      "step": 337
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6507254850274387,
+      "learning_rate": 1.996168151088015e-05,
+      "loss": 1.1006,
+      "step": 338
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.153010953674402,
+      "learning_rate": 1.996120134388834e-05,
+      "loss": 0.8967,
+      "step": 339
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2447264621546499,
+      "learning_rate": 1.996071819297314e-05,
+      "loss": 0.9864,
+      "step": 340
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7831892172540922,
+      "learning_rate": 1.9960232058279282e-05,
+      "loss": 0.9479,
+      "step": 341
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.093705324182066,
+      "learning_rate": 1.9959742939952393e-05,
+      "loss": 0.7944,
+      "step": 342
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.04956442447689,
+      "learning_rate": 1.995925083813899e-05,
+      "loss": 1.0613,
+      "step": 343
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0112084318317056,
+      "learning_rate": 1.995875575298648e-05,
+      "loss": 0.7157,
+      "step": 344
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.271337462564803,
+      "learning_rate": 1.9958257684643173e-05,
+      "loss": 1.2757,
+      "step": 345
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0950117991590314,
+      "learning_rate": 1.9957756633258264e-05,
+      "loss": 0.8981,
+      "step": 346
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.287731498132071,
+      "learning_rate": 1.995725259898185e-05,
+      "loss": 1.0272,
+      "step": 347
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.256478695437002,
+      "learning_rate": 1.9956745581964914e-05,
+      "loss": 1.1214,
+      "step": 348
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 26.609240750149247,
+      "learning_rate": 1.995623558235934e-05,
+      "loss": 1.9134,
+      "step": 349
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0316669673461139,
+      "learning_rate": 1.9955722600317887e-05,
+      "loss": 1.1246,
+      "step": 350
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5873692772868653,
+      "learning_rate": 1.9955206635994236e-05,
+      "loss": 0.8479,
+      "step": 351
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.051210404287832,
+      "learning_rate": 1.995468768954294e-05,
+      "loss": 1.044,
+      "step": 352
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6024093137514965,
+      "learning_rate": 1.995416576111945e-05,
+      "loss": 1.0312,
+      "step": 353
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6861585859735846,
+      "learning_rate": 1.9953640850880118e-05,
+      "loss": 0.8382,
+      "step": 354
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.457143030124105,
+      "learning_rate": 1.9953112958982178e-05,
+      "loss": 1.4734,
+      "step": 355
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6656996253343241,
+      "learning_rate": 1.9952582085583762e-05,
+      "loss": 0.8053,
+      "step": 356
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2813286951716827,
+      "learning_rate": 1.9952048230843898e-05,
+      "loss": 0.9799,
+      "step": 357
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3085823266033267,
+      "learning_rate": 1.995151139492251e-05,
+      "loss": 0.8599,
+      "step": 358
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9407776896936704,
+      "learning_rate": 1.99509715779804e-05,
+      "loss": 1.1317,
+      "step": 359
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0927547312214658,
+      "learning_rate": 1.9950428780179274e-05,
+      "loss": 0.9201,
+      "step": 360
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9202914653060243,
+      "learning_rate": 1.9949883001681733e-05,
+      "loss": 0.9099,
+      "step": 361
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3079387969923775,
+      "learning_rate": 1.994933424265127e-05,
+      "loss": 0.9013,
+      "step": 362
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.381006928621904,
+      "learning_rate": 1.9948782503252264e-05,
+      "loss": 1.0063,
+      "step": 363
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.609673232955171,
+      "learning_rate": 1.994822778364999e-05,
+      "loss": 1.2777,
+      "step": 364
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1427128102601014,
+      "learning_rate": 1.994767008401062e-05,
+      "loss": 0.8659,
+      "step": 365
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 32.3791031837414,
+      "learning_rate": 1.9947109404501216e-05,
+      "loss": 1.4085,
+      "step": 366
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3823587683685556,
+      "learning_rate": 1.9946545745289727e-05,
+      "loss": 1.0669,
+      "step": 367
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1785283779023612,
+      "learning_rate": 1.9945979106545003e-05,
+      "loss": 0.8125,
+      "step": 368
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4472014064776217,
+      "learning_rate": 1.9945409488436787e-05,
+      "loss": 1.2895,
+      "step": 369
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6805258542898696,
+      "learning_rate": 1.99448368911357e-05,
+      "loss": 0.9667,
+      "step": 370
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5098208167153122,
+      "learning_rate": 1.994426131481328e-05,
+      "loss": 1.0176,
+      "step": 371
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2617546488293478,
+      "learning_rate": 1.9943682759641937e-05,
+      "loss": 0.9706,
+      "step": 372
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2165458595304373,
+      "learning_rate": 1.9943101225794974e-05,
+      "loss": 1.0733,
+      "step": 373
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.238944135185444,
+      "learning_rate": 1.99425167134466e-05,
+      "loss": 1.1948,
+      "step": 374
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.204624026209821,
+      "learning_rate": 1.9941929222771907e-05,
+      "loss": 0.9973,
+      "step": 375
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9738674778430124,
+      "learning_rate": 1.9941338753946878e-05,
+      "loss": 0.8371,
+      "step": 376
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2337878184164965,
+      "learning_rate": 1.9940745307148394e-05,
+      "loss": 0.9973,
+      "step": 377
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0962851602919654,
+      "learning_rate": 1.9940148882554223e-05,
+      "loss": 1.3639,
+      "step": 378
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3925924685724334,
+      "learning_rate": 1.9939549480343025e-05,
+      "loss": 0.9546,
+      "step": 379
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1489175934706102,
+      "learning_rate": 1.9938947100694354e-05,
+      "loss": 0.8732,
+      "step": 380
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2941283088552242,
+      "learning_rate": 1.9938341743788658e-05,
+      "loss": 0.9474,
+      "step": 381
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9728318152349684,
+      "learning_rate": 1.9937733409807273e-05,
+      "loss": 1.0828,
+      "step": 382
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1662105629723933,
+      "learning_rate": 1.9937122098932428e-05,
+      "loss": 1.0754,
+      "step": 383
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0970540522022538,
+      "learning_rate": 1.9936507811347243e-05,
+      "loss": 0.8967,
+      "step": 384
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.91629872385282,
+      "learning_rate": 1.9935890547235735e-05,
+      "loss": 1.3075,
+      "step": 385
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3658553050957074,
+      "learning_rate": 1.9935270306782804e-05,
+      "loss": 0.9906,
+      "step": 386
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.188579393620228,
+      "learning_rate": 1.993464709017425e-05,
+      "loss": 0.9892,
+      "step": 387
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.226006927286577,
+      "learning_rate": 1.9934020897596752e-05,
+      "loss": 1.286,
+      "step": 388
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.131406817207597,
+      "learning_rate": 1.9933391729237898e-05,
+      "loss": 0.6906,
+      "step": 389
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0659826392600857,
+      "learning_rate": 1.9932759585286154e-05,
+      "loss": 0.7663,
+      "step": 390
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9013699816100185,
+      "learning_rate": 1.9932124465930884e-05,
+      "loss": 1.0505,
+      "step": 391
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1381300949896263,
+      "learning_rate": 1.9931486371362333e-05,
+      "loss": 0.9088,
+      "step": 392
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2217628215993273,
+      "learning_rate": 1.9930845301771658e-05,
+      "loss": 1.2801,
+      "step": 393
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9679444780082233,
+      "learning_rate": 1.9930201257350883e-05,
+      "loss": 0.8561,
+      "step": 394
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0514473084184017,
+      "learning_rate": 1.9929554238292944e-05,
+      "loss": 0.8245,
+      "step": 395
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8604154641195592,
+      "learning_rate": 1.992890424479165e-05,
+      "loss": 1.0081,
+      "step": 396
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2396228953559016,
+      "learning_rate": 1.9928251277041712e-05,
+      "loss": 1.3106,
+      "step": 397
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1154169329840353,
+      "learning_rate": 1.9927595335238736e-05,
+      "loss": 0.8084,
+      "step": 398
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1036624819044427,
+      "learning_rate": 1.9926936419579203e-05,
+      "loss": 0.902,
+      "step": 399
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1362050984653262,
+      "learning_rate": 1.99262745302605e-05,
+      "loss": 1.2938,
+      "step": 400
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1728829199932305,
+      "learning_rate": 1.9925609667480894e-05,
+      "loss": 0.8493,
+      "step": 401
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 15.221847787101918,
+      "learning_rate": 1.992494183143955e-05,
+      "loss": 1.4088,
+      "step": 402
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2407164624000968,
+      "learning_rate": 1.9924271022336525e-05,
+      "loss": 1.2135,
+      "step": 403
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4477557428281107,
+      "learning_rate": 1.992359724037276e-05,
+      "loss": 0.9057,
+      "step": 404
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1384564926578986,
+      "learning_rate": 1.992292048575009e-05,
+      "loss": 1.0428,
+      "step": 405
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1740559211726858,
+      "learning_rate": 1.992224075867124e-05,
+      "loss": 0.567,
+      "step": 406
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7789012756423233,
+      "learning_rate": 1.9921558059339822e-05,
+      "loss": 1.1478,
+      "step": 407
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2970936385781409,
+      "learning_rate": 1.992087238796035e-05,
+      "loss": 0.8959,
+      "step": 408
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9220184996892271,
+      "learning_rate": 1.9920183744738208e-05,
+      "loss": 0.7374,
+      "step": 409
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.158832902158099,
+      "learning_rate": 1.991949212987969e-05,
+      "loss": 0.8014,
+      "step": 410
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1820081962887254,
+      "learning_rate": 1.9918797543591974e-05,
+      "loss": 0.8722,
+      "step": 411
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2535376942844296,
+      "learning_rate": 1.9918099986083125e-05,
+      "loss": 1.4273,
+      "step": 412
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.042904048195446,
+      "learning_rate": 1.9917399457562093e-05,
+      "loss": 0.8198,
+      "step": 413
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1067429639913329,
+      "learning_rate": 1.9916695958238736e-05,
+      "loss": 1.0267,
+      "step": 414
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0906978255095643,
+      "learning_rate": 1.9915989488323778e-05,
+      "loss": 1.032,
+      "step": 415
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.872205931147537,
+      "learning_rate": 1.9915280048028853e-05,
+      "loss": 0.6962,
+      "step": 416
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2315230042962557,
+      "learning_rate": 1.9914567637566477e-05,
+      "loss": 1.1467,
+      "step": 417
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0609832477883663,
+      "learning_rate": 1.9913852257150053e-05,
+      "loss": 0.8772,
+      "step": 418
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.189402093404012,
+      "learning_rate": 1.991313390699388e-05,
+      "loss": 1.0334,
+      "step": 419
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.04692221778637,
+      "learning_rate": 1.991241258731314e-05,
+      "loss": 0.8094,
+      "step": 420
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2577816309142003,
+      "learning_rate": 1.9911688298323905e-05,
+      "loss": 1.2138,
+      "step": 421
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.962024559469412,
+      "learning_rate": 1.9910961040243148e-05,
+      "loss": 1.0566,
+      "step": 422
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0005849522670887,
+      "learning_rate": 1.9910230813288713e-05,
+      "loss": 0.8756,
+      "step": 423
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2721751233234033,
+      "learning_rate": 1.990949761767935e-05,
+      "loss": 0.8882,
+      "step": 424
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4646519628960513,
+      "learning_rate": 1.9908761453634687e-05,
+      "loss": 1.0449,
+      "step": 425
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.257503435994227,
+      "learning_rate": 1.9908022321375244e-05,
+      "loss": 1.2569,
+      "step": 426
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.113148399184072,
+      "learning_rate": 1.9907280221122437e-05,
+      "loss": 0.8684,
+      "step": 427
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3445544380224066,
+      "learning_rate": 1.9906535153098558e-05,
+      "loss": 1.049,
+      "step": 428
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4422807237638158,
+      "learning_rate": 1.9905787117526803e-05,
+      "loss": 0.9612,
+      "step": 429
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1891486328287892,
+      "learning_rate": 1.9905036114631247e-05,
+      "loss": 0.9666,
+      "step": 430
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0244669946069103,
+      "learning_rate": 1.990428214463685e-05,
+      "loss": 1.095,
+      "step": 431
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7078354064204718,
+      "learning_rate": 1.990352520776948e-05,
+      "loss": 1.019,
+      "step": 432
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2379831506488108,
+      "learning_rate": 1.990276530425587e-05,
+      "loss": 0.9274,
+      "step": 433
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.960457337189499,
+      "learning_rate": 1.9902002434323658e-05,
+      "loss": 0.8511,
+      "step": 434
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2056613223986556,
+      "learning_rate": 1.9901236598201363e-05,
+      "loss": 1.0038,
+      "step": 435
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2364553969314112,
+      "learning_rate": 1.9900467796118393e-05,
+      "loss": 1.183,
+      "step": 436
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0359728831463237,
+      "learning_rate": 1.989969602830505e-05,
+      "loss": 1.008,
+      "step": 437
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9446637021370143,
+      "learning_rate": 1.9898921294992516e-05,
+      "loss": 1.1322,
+      "step": 438
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8586812601799279,
+      "learning_rate": 1.9898143596412872e-05,
+      "loss": 0.7616,
+      "step": 439
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.296614953660857,
+      "learning_rate": 1.9897362932799078e-05,
+      "loss": 0.813,
+      "step": 440
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3684615719277307,
+      "learning_rate": 1.9896579304384984e-05,
+      "loss": 1.3035,
+      "step": 441
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2583805056895796,
+      "learning_rate": 1.9895792711405332e-05,
+      "loss": 0.9698,
+      "step": 442
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4418017973491841,
+      "learning_rate": 1.9895003154095747e-05,
+      "loss": 0.8692,
+      "step": 443
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.350176934862449,
+      "learning_rate": 1.9894210632692745e-05,
+      "loss": 1.1148,
+      "step": 444
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2373791421652622,
+      "learning_rate": 1.989341514743373e-05,
+      "loss": 1.283,
+      "step": 445
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0552398757211394,
+      "learning_rate": 1.9892616698556996e-05,
+      "loss": 0.8249,
+      "step": 446
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.206777457993399,
+      "learning_rate": 1.989181528630172e-05,
+      "loss": 1.0174,
+      "step": 447
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1463583480940882,
+      "learning_rate": 1.9891010910907967e-05,
+      "loss": 0.7888,
+      "step": 448
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8841815064520075,
+      "learning_rate": 1.9890203572616688e-05,
+      "loss": 0.7369,
+      "step": 449
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.506488586066231,
+      "learning_rate": 1.9889393271669735e-05,
+      "loss": 1.0857,
+      "step": 450
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.008027846907909,
+      "learning_rate": 1.988858000830983e-05,
+      "loss": 0.8064,
+      "step": 451
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.187817538821626,
+      "learning_rate": 1.988776378278059e-05,
+      "loss": 0.9258,
+      "step": 452
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2082146031016943,
+      "learning_rate": 1.9886944595326524e-05,
+      "loss": 0.8732,
+      "step": 453
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3001313559399355,
+      "learning_rate": 1.9886122446193015e-05,
+      "loss": 0.9408,
+      "step": 454
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7614064223722202,
+      "learning_rate": 1.988529733562635e-05,
+      "loss": 1.1128,
+      "step": 455
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1257183383830704,
+      "learning_rate": 1.988446926387369e-05,
+      "loss": 0.677,
+      "step": 456
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1141492446644663,
+      "learning_rate": 1.988363823118309e-05,
+      "loss": 0.9292,
+      "step": 457
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0085951215543723,
+      "learning_rate": 1.9882804237803487e-05,
+      "loss": 0.8015,
+      "step": 458
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.568340656817881,
+      "learning_rate": 1.988196728398471e-05,
+      "loss": 1.0118,
+      "step": 459
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0136250092133101,
+      "learning_rate": 1.988112736997747e-05,
+      "loss": 1.1425,
+      "step": 460
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0470647771510826,
+      "learning_rate": 1.988028449603337e-05,
+      "loss": 1.0653,
+      "step": 461
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3438987406839171,
+      "learning_rate": 1.9879438662404893e-05,
+      "loss": 0.7118,
+      "step": 462
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0587433077962647,
+      "learning_rate": 1.9878589869345414e-05,
+      "loss": 0.9888,
+      "step": 463
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1354520271532695,
+      "learning_rate": 1.9877738117109195e-05,
+      "loss": 1.1367,
+      "step": 464
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9963486878458707,
+      "learning_rate": 1.9876883405951378e-05,
+      "loss": 1.093,
+      "step": 465
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5105612737068392,
+      "learning_rate": 1.9876025736128e-05,
+      "loss": 0.8297,
+      "step": 466
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1677790260032999,
+      "learning_rate": 1.987516510789598e-05,
+      "loss": 0.9695,
+      "step": 467
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0929960344552727,
+      "learning_rate": 1.987430152151312e-05,
+      "loss": 0.9165,
+      "step": 468
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0260918869576994,
+      "learning_rate": 1.9873434977238108e-05,
+      "loss": 1.1095,
+      "step": 469
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1140755712842842,
+      "learning_rate": 1.987256547533053e-05,
+      "loss": 0.6902,
+      "step": 470
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.16484107721168,
+      "learning_rate": 1.9871693016050843e-05,
+      "loss": 0.9932,
+      "step": 471
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1959377307183807,
+      "learning_rate": 1.98708175996604e-05,
+      "loss": 0.8362,
+      "step": 472
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8168635971619997,
+      "learning_rate": 1.9869939226421435e-05,
+      "loss": 0.9681,
+      "step": 473
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.073145900812803,
+      "learning_rate": 1.9869057896597068e-05,
+      "loss": 1.1857,
+      "step": 474
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.867078347646907,
+      "learning_rate": 1.9868173610451302e-05,
+      "loss": 0.7992,
+      "step": 475
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.05703321197932,
+      "learning_rate": 1.9867286368249034e-05,
+      "loss": 0.966,
+      "step": 476
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.999800534335202,
+      "learning_rate": 1.986639617025604e-05,
+      "loss": 0.795,
+      "step": 477
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9819398571110847,
+      "learning_rate": 1.9865503016738983e-05,
+      "loss": 0.9797,
+      "step": 478
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1907549089791754,
+      "learning_rate": 1.9864606907965407e-05,
+      "loss": 1.1134,
+      "step": 479
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4758908128787274,
+      "learning_rate": 1.9863707844203756e-05,
+      "loss": 1.0626,
+      "step": 480
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0384483939037152,
+      "learning_rate": 1.9862805825723336e-05,
+      "loss": 0.7752,
+      "step": 481
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8378790760923875,
+      "learning_rate": 1.9861900852794357e-05,
+      "loss": 1.0201,
+      "step": 482
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9525784585674378,
+      "learning_rate": 1.9860992925687908e-05,
+      "loss": 0.7738,
+      "step": 483
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0746994801785679,
+      "learning_rate": 1.9860082044675964e-05,
+      "loss": 1.1989,
+      "step": 484
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0580459243074904,
+      "learning_rate": 1.985916821003138e-05,
+      "loss": 0.8937,
+      "step": 485
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0605049914311215,
+      "learning_rate": 1.9858251422027903e-05,
+      "loss": 1.2133,
+      "step": 486
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9689295707871675,
+      "learning_rate": 1.9857331680940156e-05,
+      "loss": 0.9665,
+      "step": 487
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1260135355690888,
+      "learning_rate": 1.985640898704366e-05,
+      "loss": 1.2554,
+      "step": 488
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9586539470636446,
+      "learning_rate": 1.98554833406148e-05,
+      "loss": 1.0559,
+      "step": 489
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9117374809862491,
+      "learning_rate": 1.985455474193087e-05,
+      "loss": 0.8992,
+      "step": 490
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9671235628768305,
+      "learning_rate": 1.9853623191270026e-05,
+      "loss": 0.7049,
+      "step": 491
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3296873355227528,
+      "learning_rate": 1.9852688688911325e-05,
+      "loss": 0.7829,
+      "step": 492
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0939243523673399,
+      "learning_rate": 1.98517512351347e-05,
+      "loss": 1.1829,
+      "step": 493
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9080190212662427,
+      "learning_rate": 1.985081083022097e-05,
+      "loss": 0.911,
+      "step": 494
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8464835864794619,
+      "learning_rate": 1.9849867474451835e-05,
+      "loss": 0.7188,
+      "step": 495
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0827057007447267,
+      "learning_rate": 1.984892116810988e-05,
+      "loss": 0.7961,
+      "step": 496
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9528094506277993,
+      "learning_rate": 1.984797191147858e-05,
+      "loss": 0.7174,
+      "step": 497
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9785171644437913,
+      "learning_rate": 1.984701970484229e-05,
+      "loss": 1.118,
+      "step": 498
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2097697218222294,
+      "learning_rate": 1.9846064548486242e-05,
+      "loss": 0.9069,
+      "step": 499
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8858947236289725,
+      "learning_rate": 1.9845106442696563e-05,
+      "loss": 0.9133,
+      "step": 500
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9605001415752112,
+      "learning_rate": 1.9844145387760253e-05,
+      "loss": 0.7985,
+      "step": 501
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0403036870565467,
+      "learning_rate": 1.9843181383965207e-05,
+      "loss": 1.2946,
+      "step": 502
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1595186123086805,
+      "learning_rate": 1.984221443160019e-05,
+      "loss": 0.9041,
+      "step": 503
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1321217231159466,
+      "learning_rate": 1.984124453095486e-05,
+      "loss": 0.908,
+      "step": 504
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.813878789447278,
+      "learning_rate": 1.9840271682319755e-05,
+      "loss": 1.041,
+      "step": 505
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0636255277036508,
+      "learning_rate": 1.98392958859863e-05,
+      "loss": 0.7655,
+      "step": 506
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0571512685928035,
+      "learning_rate": 1.983831714224679e-05,
+      "loss": 1.1018,
+      "step": 507
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8909171658252358,
+      "learning_rate": 1.983733545139442e-05,
+      "loss": 1.0095,
+      "step": 508
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.134648886340577,
+      "learning_rate": 1.983635081372326e-05,
+      "loss": 0.9489,
+      "step": 509
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.906909474190707,
+      "learning_rate": 1.9835363229528255e-05,
+      "loss": 1.1179,
+      "step": 510
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2070406193010854,
+      "learning_rate": 1.983437269910525e-05,
+      "loss": 0.7606,
+      "step": 511
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.753170895408428,
+      "learning_rate": 1.9833379222750957e-05,
+      "loss": 0.8368,
+      "step": 512
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.671044891260849,
+      "learning_rate": 1.983238280076298e-05,
+      "loss": 0.957,
+      "step": 513
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4395208236833323,
+      "learning_rate": 1.9831383433439798e-05,
+      "loss": 0.7997,
+      "step": 514
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9866173078845144,
+      "learning_rate": 1.983038112108078e-05,
+      "loss": 0.9289,
+      "step": 515
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4008225918143338,
+      "learning_rate": 1.9829375863986168e-05,
+      "loss": 0.914,
+      "step": 516
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0763595618639281,
+      "learning_rate": 1.9828367662457097e-05,
+      "loss": 1.1754,
+      "step": 517
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4520349430965602,
+      "learning_rate": 1.982735651679557e-05,
+      "loss": 0.9971,
+      "step": 518
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5407352570546937,
+      "learning_rate": 1.982634242730449e-05,
+      "loss": 0.7476,
+      "step": 519
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9655466560956543,
+      "learning_rate": 1.982532539428763e-05,
+      "loss": 0.8788,
+      "step": 520
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0105969852434864,
+      "learning_rate": 1.9824305418049645e-05,
+      "loss": 0.9334,
+      "step": 521
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.243446652956385,
+      "learning_rate": 1.982328249889607e-05,
+      "loss": 1.2129,
+      "step": 522
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9893964630824509,
+      "learning_rate": 1.9822256637133336e-05,
+      "loss": 0.9037,
+      "step": 523
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1142930876084902,
+      "learning_rate": 1.982122783306873e-05,
+      "loss": 0.7987,
+      "step": 524
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7046821468390176,
+      "learning_rate": 1.9820196087010446e-05,
+      "loss": 0.8504,
+      "step": 525
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0827280870654252,
+      "learning_rate": 1.9819161399267545e-05,
+      "loss": 0.7961,
+      "step": 526
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2092199045180412,
+      "learning_rate": 1.981812377014997e-05,
+      "loss": 1.2767,
+      "step": 527
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.09403908704094,
+      "learning_rate": 1.9817083199968552e-05,
+      "loss": 1.0068,
+      "step": 528
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6937727030304057,
+      "learning_rate": 1.9816039689034995e-05,
+      "loss": 0.6729,
+      "step": 529
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9218492502754294,
+      "learning_rate": 1.9814993237661887e-05,
+      "loss": 1.1565,
+      "step": 530
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8220768516875354,
+      "learning_rate": 1.98139438461627e-05,
+      "loss": 1.073,
+      "step": 531
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9728079767986487,
+      "learning_rate": 1.9812891514851782e-05,
+      "loss": 0.9641,
+      "step": 532
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5581772483261247,
+      "learning_rate": 1.9811836244044365e-05,
+      "loss": 0.9034,
+      "step": 533
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1959197583761465,
+      "learning_rate": 1.9810778034056562e-05,
+      "loss": 0.8817,
+      "step": 534
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0043057328028187,
+      "learning_rate": 1.9809716885205363e-05,
+      "loss": 1.0648,
+      "step": 535
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1231312559222413,
+      "learning_rate": 1.9808652797808635e-05,
+      "loss": 1.2197,
+      "step": 536
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.233488327468138,
+      "learning_rate": 1.9807585772185137e-05,
+      "loss": 1.0037,
+      "step": 537
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.32307573919409,
+      "learning_rate": 1.98065158086545e-05,
+      "loss": 0.7834,
+      "step": 538
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1921046659138717,
+      "learning_rate": 1.9805442907537232e-05,
+      "loss": 1.0643,
+      "step": 539
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0416922894734766,
+      "learning_rate": 1.980436706915473e-05,
+      "loss": 1.168,
+      "step": 540
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.156401261002975,
+      "learning_rate": 1.980328829382927e-05,
+      "loss": 0.8343,
+      "step": 541
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9741312429487489,
+      "learning_rate": 1.9802206581883992e-05,
+      "loss": 0.9285,
+      "step": 542
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0149459334112225,
+      "learning_rate": 1.980112193364294e-05,
+      "loss": 0.9435,
+      "step": 543
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.701388744605041,
+      "learning_rate": 1.980003434943102e-05,
+      "loss": 0.8644,
+      "step": 544
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9919485742682616,
+      "learning_rate": 1.979894382957402e-05,
+      "loss": 0.7085,
+      "step": 545
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2155956937557044,
+      "learning_rate": 1.9797850374398617e-05,
+      "loss": 1.259,
+      "step": 546
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8142339884916636,
+      "learning_rate": 1.9796753984232357e-05,
+      "loss": 0.6821,
+      "step": 547
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4890135357216874,
+      "learning_rate": 1.979565465940367e-05,
+      "loss": 0.8953,
+      "step": 548
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0761431861191848,
+      "learning_rate": 1.979455240024186e-05,
+      "loss": 0.9034,
+      "step": 549
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0902017309231902,
+      "learning_rate": 1.9793447207077114e-05,
+      "loss": 0.9129,
+      "step": 550
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0395089278212704,
+      "learning_rate": 1.9792339080240502e-05,
+      "loss": 1.4452,
+      "step": 551
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9423298871236481,
+      "learning_rate": 1.9791228020063965e-05,
+      "loss": 0.9177,
+      "step": 552
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0198006298301394,
+      "learning_rate": 1.9790114026880327e-05,
+      "loss": 0.8948,
+      "step": 553
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.136531823687823,
+      "learning_rate": 1.9788997101023294e-05,
+      "loss": 0.8323,
+      "step": 554
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1125538577585241,
+      "learning_rate": 1.9787877242827437e-05,
+      "loss": 1.1565,
+      "step": 555
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1028294196850077,
+      "learning_rate": 1.9786754452628226e-05,
+      "loss": 0.8757,
+      "step": 556
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.871166877134957,
+      "learning_rate": 1.9785628730761987e-05,
+      "loss": 1.0066,
+      "step": 557
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2800946126043202,
+      "learning_rate": 1.9784500077565944e-05,
+      "loss": 0.946,
+      "step": 558
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1373665849653194,
+      "learning_rate": 1.9783368493378186e-05,
+      "loss": 1.0651,
+      "step": 559
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0628013349213334,
+      "learning_rate": 1.978223397853768e-05,
+      "loss": 1.1274,
+      "step": 560
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0063460802125936,
+      "learning_rate": 1.9781096533384286e-05,
+      "loss": 0.5605,
+      "step": 561
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1874315553062746,
+      "learning_rate": 1.977995615825872e-05,
+      "loss": 0.8464,
+      "step": 562
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2579957840941105,
+      "learning_rate": 1.9778812853502592e-05,
+      "loss": 1.012,
+      "step": 563
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9393434873369301,
+      "learning_rate": 1.9777666619458385e-05,
+      "loss": 0.8294,
+      "step": 564
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.966531053306293,
+      "learning_rate": 1.9776517456469456e-05,
+      "loss": 0.9,
+      "step": 565
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0848023397326898,
+      "learning_rate": 1.9775365364880043e-05,
+      "loss": 0.9412,
+      "step": 566
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4853827736601393,
+      "learning_rate": 1.977421034503526e-05,
+      "loss": 0.9411,
+      "step": 567
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0159440787054508,
+      "learning_rate": 1.97730523972811e-05,
+      "loss": 0.827,
+      "step": 568
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1872140806834026,
+      "learning_rate": 1.9771891521964428e-05,
+      "loss": 0.9076,
+      "step": 569
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0071380529409069,
+      "learning_rate": 1.9770727719432994e-05,
+      "loss": 0.9496,
+      "step": 570
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.147287129990345,
+      "learning_rate": 1.9769560990035418e-05,
+      "loss": 0.9115,
+      "step": 571
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.246503392390965,
+      "learning_rate": 1.9768391334121196e-05,
+      "loss": 0.8894,
+      "step": 572
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9811489147685969,
+      "learning_rate": 1.976721875204071e-05,
+      "loss": 1.149,
+      "step": 573
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.04817377451563,
+      "learning_rate": 1.976604324414521e-05,
+      "loss": 1.1299,
+      "step": 574
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0358105373925968,
+      "learning_rate": 1.9764864810786824e-05,
+      "loss": 0.8322,
+      "step": 575
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2015752737133545,
+      "learning_rate": 1.9763683452318557e-05,
+      "loss": 0.8567,
+      "step": 576
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4662547601581837,
+      "learning_rate": 1.9762499169094288e-05,
+      "loss": 0.916,
+      "step": 577
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8560897610219825,
+      "learning_rate": 1.9761311961468782e-05,
+      "loss": 1.0858,
+      "step": 578
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1144675315885182,
+      "learning_rate": 1.976012182979766e-05,
+      "loss": 1.1778,
+      "step": 579
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.522438324197729,
+      "learning_rate": 1.9758928774437444e-05,
+      "loss": 0.7745,
+      "step": 580
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.357793340009838,
+      "learning_rate": 1.9757732795745515e-05,
+      "loss": 0.8411,
+      "step": 581
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.953960260896789,
+      "learning_rate": 1.9756533894080132e-05,
+      "loss": 1.0367,
+      "step": 582
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4133312392346062,
+      "learning_rate": 1.9755332069800432e-05,
+      "loss": 0.5947,
+      "step": 583
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1093407652297895,
+      "learning_rate": 1.9754127323266426e-05,
+      "loss": 1.2092,
+      "step": 584
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1637859458240973,
+      "learning_rate": 1.9752919654839004e-05,
+      "loss": 0.7238,
+      "step": 585
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.135547562407171,
+      "learning_rate": 1.975170906487993e-05,
+      "loss": 0.8923,
+      "step": 586
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1902838838869694,
+      "learning_rate": 1.9750495553751838e-05,
+      "loss": 0.8165,
+      "step": 587
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3094383574509707,
+      "learning_rate": 1.9749279121818235e-05,
+      "loss": 1.0309,
+      "step": 588
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7979739304717968,
+      "learning_rate": 1.9748059769443522e-05,
+      "loss": 1.1884,
+      "step": 589
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1763594195193923,
+      "learning_rate": 1.9746837496992953e-05,
+      "loss": 0.7888,
+      "step": 590
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3826921686151625,
+      "learning_rate": 1.9745612304832672e-05,
+      "loss": 1.0932,
+      "step": 591
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1038849952317655,
+      "learning_rate": 1.9744384193329683e-05,
+      "loss": 0.8413,
+      "step": 592
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3585309298366672,
+      "learning_rate": 1.9743153162851875e-05,
+      "loss": 1.1636,
+      "step": 593
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.230137175084341,
+      "learning_rate": 1.974191921376801e-05,
+      "loss": 0.8354,
+      "step": 594
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2572411386020548,
+      "learning_rate": 1.9740682346447724e-05,
+      "loss": 1.0312,
+      "step": 595
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2402931172613536,
+      "learning_rate": 1.9739442561261525e-05,
+      "loss": 0.7352,
+      "step": 596
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1477458792880981,
+      "learning_rate": 1.9738199858580796e-05,
+      "loss": 1.0713,
+      "step": 597
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1469734062089518,
+      "learning_rate": 1.9736954238777793e-05,
+      "loss": 1.1227,
+      "step": 598
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1821649088462298,
+      "learning_rate": 1.973570570222565e-05,
+      "loss": 0.7504,
+      "step": 599
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.283217002347211,
+      "learning_rate": 1.9734454249298367e-05,
+      "loss": 0.8129,
+      "step": 600
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8245914902440055,
+      "learning_rate": 1.973319988037083e-05,
+      "loss": 0.6897,
+      "step": 601
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1826050268628896,
+      "learning_rate": 1.9731942595818786e-05,
+      "loss": 0.8977,
+      "step": 602
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3098766754973783,
+      "learning_rate": 1.973068239601886e-05,
+      "loss": 1.403,
+      "step": 603
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.774219693915334,
+      "learning_rate": 1.9729419281348553e-05,
+      "loss": 1.0831,
+      "step": 604
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1821690114268,
+      "learning_rate": 1.972815325218624e-05,
+      "loss": 0.8871,
+      "step": 605
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0752058366369566,
+      "learning_rate": 1.9726884308911153e-05,
+      "loss": 0.7128,
+      "step": 606
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0436513379174401,
+      "learning_rate": 1.9725612451903423e-05,
+      "loss": 0.7633,
+      "step": 607
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.169040139012969,
+      "learning_rate": 1.9724337681544036e-05,
+      "loss": 1.2996,
+      "step": 608
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.051820524713669,
+      "learning_rate": 1.9723059998214852e-05,
+      "loss": 0.8148,
+      "step": 609
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.875443272208857,
+      "learning_rate": 1.9721779402298614e-05,
+      "loss": 1.1487,
+      "step": 610
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1410347291971419,
+      "learning_rate": 1.972049589417892e-05,
+      "loss": 0.8474,
+      "step": 611
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3564489002506386,
+      "learning_rate": 1.9719209474240263e-05,
+      "loss": 0.8299,
+      "step": 612
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8746068029029166,
+      "learning_rate": 1.9717920142867988e-05,
+      "loss": 1.1156,
+      "step": 613
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1799279265012412,
+      "learning_rate": 1.9716627900448323e-05,
+      "loss": 0.9333,
+      "step": 614
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0616173321315296,
+      "learning_rate": 1.9715332747368362e-05,
+      "loss": 1.0612,
+      "step": 615
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9688325279428881,
+      "learning_rate": 1.971403468401608e-05,
+      "loss": 0.9492,
+      "step": 616
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.391312210006241,
+      "learning_rate": 1.971273371078031e-05,
+      "loss": 0.7747,
+      "step": 617
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1630266628977075,
+      "learning_rate": 1.971142982805077e-05,
+      "loss": 1.2953,
+      "step": 618
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.265702731053522,
+      "learning_rate": 1.9710123036218044e-05,
+      "loss": 0.846,
+      "step": 619
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5283196070978808,
+      "learning_rate": 1.9708813335673582e-05,
+      "loss": 0.5645,
+      "step": 620
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5194651089799394,
+      "learning_rate": 1.970750072680972e-05,
+      "loss": 0.8206,
+      "step": 621
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.303827215345261,
+      "learning_rate": 1.9706185210019648e-05,
+      "loss": 1.1383,
+      "step": 622
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1478440318951875,
+      "learning_rate": 1.9704866785697437e-05,
+      "loss": 0.8564,
+      "step": 623
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3891426930854838,
+      "learning_rate": 1.970354545423803e-05,
+      "loss": 0.9424,
+      "step": 624
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0603769198109512,
+      "learning_rate": 1.9702221216037234e-05,
+      "loss": 0.7823,
+      "step": 625
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8578429196248756,
+      "learning_rate": 1.9700894071491736e-05,
+      "loss": 1.0263,
+      "step": 626
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4227838952662184,
+      "learning_rate": 1.9699564020999077e-05,
+      "loss": 1.2226,
+      "step": 627
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0451593894907176,
+      "learning_rate": 1.9698231064957695e-05,
+      "loss": 0.8664,
+      "step": 628
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9888276290514061,
+      "learning_rate": 1.969689520376687e-05,
+      "loss": 1.1054,
+      "step": 629
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0306791454785607,
+      "learning_rate": 1.969555643782677e-05,
+      "loss": 0.6669,
+      "step": 630
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9625704591588373,
+      "learning_rate": 1.9694214767538434e-05,
+      "loss": 1.1126,
+      "step": 631
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9175515917308579,
+      "learning_rate": 1.9692870193303758e-05,
+      "loss": 0.7274,
+      "step": 632
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0972588263046745,
+      "learning_rate": 1.969152271552552e-05,
+      "loss": 0.802,
+      "step": 633
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8212046458530384,
+      "learning_rate": 1.969017233460736e-05,
+      "loss": 1.1291,
+      "step": 634
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2563778843837863,
+      "learning_rate": 1.968881905095379e-05,
+      "loss": 0.84,
+      "step": 635
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1434827205087603,
+      "learning_rate": 1.9687462864970198e-05,
+      "loss": 0.8279,
+      "step": 636
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1720494913389898,
+      "learning_rate": 1.9686103777062828e-05,
+      "loss": 0.988,
+      "step": 637
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1558020679782732,
+      "learning_rate": 1.9684741787638808e-05,
+      "loss": 0.8066,
+      "step": 638
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0240785313771403,
+      "learning_rate": 1.9683376897106123e-05,
+      "loss": 0.7792,
+      "step": 639
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0214271955092709,
+      "learning_rate": 1.9682009105873633e-05,
+      "loss": 1.1244,
+      "step": 640
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9232907187807675,
+      "learning_rate": 1.968063841435107e-05,
+      "loss": 1.1428,
+      "step": 641
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4073266777413644,
+      "learning_rate": 1.9679264822949027e-05,
+      "loss": 0.9692,
+      "step": 642
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4488644227918874,
+      "learning_rate": 1.9677888332078973e-05,
+      "loss": 0.8452,
+      "step": 643
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1765555256479354,
+      "learning_rate": 1.9676508942153235e-05,
+      "loss": 0.8233,
+      "step": 644
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1903387781501324,
+      "learning_rate": 1.9675126653585024e-05,
+      "loss": 0.7285,
+      "step": 645
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0480525258002429,
+      "learning_rate": 1.967374146678841e-05,
+      "loss": 1.1486,
+      "step": 646
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1638693780443699,
+      "learning_rate": 1.967235338217832e-05,
+      "loss": 0.8701,
+      "step": 647
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2668476812501832,
+      "learning_rate": 1.967096240017058e-05,
+      "loss": 0.9441,
+      "step": 648
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1789113265093503,
+      "learning_rate": 1.966956852118185e-05,
+      "loss": 0.8227,
+      "step": 649
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0477747971116116,
+      "learning_rate": 1.966817174562968e-05,
+      "loss": 1.092,
+      "step": 650
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3600070063535914,
+      "learning_rate": 1.966677207393248e-05,
+      "loss": 0.858,
+      "step": 651
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.31693275670863,
+      "learning_rate": 1.9665369506509526e-05,
+      "loss": 0.7364,
+      "step": 652
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0703709484498183,
+      "learning_rate": 1.9663964043780966e-05,
+      "loss": 1.0651,
+      "step": 653
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7711131514435945,
+      "learning_rate": 1.9662555686167808e-05,
+      "loss": 1.057,
+      "step": 654
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0265293766523427,
+      "learning_rate": 1.9661144434091935e-05,
+      "loss": 0.9077,
+      "step": 655
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0211075718216838,
+      "learning_rate": 1.9659730287976097e-05,
+      "loss": 0.8355,
+      "step": 656
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1569504647763447,
+      "learning_rate": 1.9658313248243907e-05,
+      "loss": 0.7514,
+      "step": 657
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.168424942876194,
+      "learning_rate": 1.9656893315319838e-05,
+      "loss": 0.8016,
+      "step": 658
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.039360527132193,
+      "learning_rate": 1.9655470489629248e-05,
+      "loss": 0.8975,
+      "step": 659
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3046848581798707,
+      "learning_rate": 1.9654044771598343e-05,
+      "loss": 1.2567,
+      "step": 660
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4352026183590068,
+      "learning_rate": 1.9652616161654204e-05,
+      "loss": 0.87,
+      "step": 661
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1197914258598476,
+      "learning_rate": 1.9651184660224786e-05,
+      "loss": 1.11,
+      "step": 662
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8469839671547741,
+      "learning_rate": 1.9649750267738894e-05,
+      "loss": 0.5848,
+      "step": 663
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3336096354802143,
+      "learning_rate": 1.9648312984626208e-05,
+      "loss": 0.9413,
+      "step": 664
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2306445464807212,
+      "learning_rate": 1.9646872811317274e-05,
+      "loss": 1.2615,
+      "step": 665
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2713498500733575,
+      "learning_rate": 1.96454297482435e-05,
+      "loss": 0.8057,
+      "step": 666
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2213742889842474,
+      "learning_rate": 1.964398379583717e-05,
+      "loss": 1.1089,
+      "step": 667
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2128581959427758,
+      "learning_rate": 1.964253495453141e-05,
+      "loss": 0.8664,
+      "step": 668
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0049232984548877,
+      "learning_rate": 1.964108322476024e-05,
+      "loss": 0.9144,
+      "step": 669
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3221385204524048,
+      "learning_rate": 1.9639628606958535e-05,
+      "loss": 1.2025,
+      "step": 670
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.58075379282766,
+      "learning_rate": 1.9638171101562024e-05,
+      "loss": 0.8289,
+      "step": 671
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0415052913752936,
+      "learning_rate": 1.963671070900731e-05,
+      "loss": 0.9049,
+      "step": 672
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3353395062210662,
+      "learning_rate": 1.9635247429731862e-05,
+      "loss": 0.9855,
+      "step": 673
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1280631868338349,
+      "learning_rate": 1.963378126417401e-05,
+      "loss": 0.7269,
+      "step": 674
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.191909993128642,
+      "learning_rate": 1.9632312212772956e-05,
+      "loss": 0.8596,
+      "step": 675
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0983153034751856,
+      "learning_rate": 1.963084027596876e-05,
+      "loss": 1.0214,
+      "step": 676
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.762744754844499,
+      "learning_rate": 1.962936545420234e-05,
+      "loss": 1.0488,
+      "step": 677
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1712593460016212,
+      "learning_rate": 1.9627887747915496e-05,
+      "loss": 0.8453,
+      "step": 678
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0666616159610733,
+      "learning_rate": 1.962640715755087e-05,
+      "loss": 1.0987,
+      "step": 679
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0949084090456407,
+      "learning_rate": 1.9624923683551992e-05,
+      "loss": 0.8427,
+      "step": 680
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1273152846272645,
+      "learning_rate": 1.9623437326363235e-05,
+      "loss": 1.0333,
+      "step": 681
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9114093177552463,
+      "learning_rate": 1.9621948086429847e-05,
+      "loss": 0.8303,
+      "step": 682
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.073947424184198,
+      "learning_rate": 1.9620455964197936e-05,
+      "loss": 1.145,
+      "step": 683
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0896904796995355,
+      "learning_rate": 1.9618960960114472e-05,
+      "loss": 1.164,
+      "step": 684
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2482158862167327,
+      "learning_rate": 1.9617463074627294e-05,
+      "loss": 0.8359,
+      "step": 685
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9658386691998154,
+      "learning_rate": 1.96159623081851e-05,
+      "loss": 0.8683,
+      "step": 686
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2572105898942787,
+      "learning_rate": 1.961445866123745e-05,
+      "loss": 0.9744,
+      "step": 687
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.463368568611212,
+      "learning_rate": 1.9612952134234768e-05,
+      "loss": 0.8677,
+      "step": 688
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4584872853200639,
+      "learning_rate": 1.9611442727628344e-05,
+      "loss": 1.2369,
+      "step": 689
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0154614901416181,
+      "learning_rate": 1.960993044187032e-05,
+      "loss": 0.9179,
+      "step": 690
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.400130874426054,
+      "learning_rate": 1.960841527741372e-05,
+      "loss": 0.8249,
+      "step": 691
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1942912729416553,
+      "learning_rate": 1.9606897234712408e-05,
+      "loss": 0.5839,
+      "step": 692
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2402339213150047,
+      "learning_rate": 1.9605376314221127e-05,
+      "loss": 1.1118,
+      "step": 693
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0346365096941326,
+      "learning_rate": 1.9603852516395472e-05,
+      "loss": 1.0133,
+      "step": 694
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1130564626751125,
+      "learning_rate": 1.9602325841691905e-05,
+      "loss": 0.7208,
+      "step": 695
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.328382061629369,
+      "learning_rate": 1.9600796290567747e-05,
+      "loss": 0.9223,
+      "step": 696
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.117422974348816,
+      "learning_rate": 1.9599263863481185e-05,
+      "loss": 0.7209,
+      "step": 697
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.043889772889515,
+      "learning_rate": 1.9597728560891266e-05,
+      "loss": 1.1162,
+      "step": 698
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2287967299693303,
+      "learning_rate": 1.959619038325789e-05,
+      "loss": 0.9639,
+      "step": 699
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1107331831610987,
+      "learning_rate": 1.9594649331041826e-05,
+      "loss": 1.0615,
+      "step": 700
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0238222514275965,
+      "learning_rate": 1.959310540470471e-05,
+      "loss": 0.8206,
+      "step": 701
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.104191832015569,
+      "learning_rate": 1.9591558604709028e-05,
+      "loss": 0.7244,
+      "step": 702
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1459177077420915,
+      "learning_rate": 1.9590008931518133e-05,
+      "loss": 1.2524,
+      "step": 703
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.085348934749462,
+      "learning_rate": 1.9588456385596232e-05,
+      "loss": 1.095,
+      "step": 704
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7846872093292252,
+      "learning_rate": 1.95869009674084e-05,
+      "loss": 1.0106,
+      "step": 705
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9007718832975623,
+      "learning_rate": 1.958534267742057e-05,
+      "loss": 1.0251,
+      "step": 706
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2847711100113748,
+      "learning_rate": 1.9583781516099538e-05,
+      "loss": 0.8897,
+      "step": 707
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1256142026478833,
+      "learning_rate": 1.9582217483912952e-05,
+      "loss": 1.1468,
+      "step": 708
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.931656433723545,
+      "learning_rate": 1.9580650581329324e-05,
+      "loss": 0.6196,
+      "step": 709
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9030057321729272,
+      "learning_rate": 1.9579080808818035e-05,
+      "loss": 0.9649,
+      "step": 710
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9360929885932352,
+      "learning_rate": 1.9577508166849308e-05,
+      "loss": 0.9246,
+      "step": 711
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1091665506932824,
+      "learning_rate": 1.9575932655894238e-05,
+      "loss": 0.8757,
+      "step": 712
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9534794455205482,
+      "learning_rate": 1.9574354276424782e-05,
+      "loss": 1.0326,
+      "step": 713
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8865683989073171,
+      "learning_rate": 1.9572773028913744e-05,
+      "loss": 0.7296,
+      "step": 714
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8571025895498565,
+      "learning_rate": 1.9571188913834794e-05,
+      "loss": 0.7468,
+      "step": 715
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1255239131564296,
+      "learning_rate": 1.9569601931662468e-05,
+      "loss": 0.9408,
+      "step": 716
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2115045950006742,
+      "learning_rate": 1.9568012082872148e-05,
+      "loss": 1.248,
+      "step": 717
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.091431415484773,
+      "learning_rate": 1.956641936794008e-05,
+      "loss": 0.7845,
+      "step": 718
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0590703951824743,
+      "learning_rate": 1.956482378734337e-05,
+      "loss": 1.0315,
+      "step": 719
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.200688812276531,
+      "learning_rate": 1.9563225341559982e-05,
+      "loss": 0.8267,
+      "step": 720
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0122188188518428,
+      "learning_rate": 1.9561624031068743e-05,
+      "loss": 0.8679,
+      "step": 721
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.310113197048741,
+      "learning_rate": 1.956001985634932e-05,
+      "loss": 1.1601,
+      "step": 722
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9570133212281858,
+      "learning_rate": 1.9558412817882263e-05,
+      "loss": 0.7562,
+      "step": 723
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3052639788259397,
+      "learning_rate": 1.9556802916148963e-05,
+      "loss": 0.9294,
+      "step": 724
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.03816540728872,
+      "learning_rate": 1.9555190151631672e-05,
+      "loss": 1.1824,
+      "step": 725
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3098335168383637,
+      "learning_rate": 1.9553574524813507e-05,
+      "loss": 0.9546,
+      "step": 726
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9291398739752258,
+      "learning_rate": 1.955195603617843e-05,
+      "loss": 1.0914,
+      "step": 727
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.294782171182801,
+      "learning_rate": 1.955033468621126e-05,
+      "loss": 0.6264,
+      "step": 728
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0288351270939442,
+      "learning_rate": 1.95487104753977e-05,
+      "loss": 0.8954,
+      "step": 729
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1476875318181434,
+      "learning_rate": 1.9547083404224273e-05,
+      "loss": 0.7054,
+      "step": 730
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9103873438655581,
+      "learning_rate": 1.9545453473178384e-05,
+      "loss": 1.091,
+      "step": 731
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9634508630770229,
+      "learning_rate": 1.954382068274828e-05,
+      "loss": 1.1327,
+      "step": 732
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1246066638969752,
+      "learning_rate": 1.9542185033423076e-05,
+      "loss": 0.6889,
+      "step": 733
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.098045696777537,
+      "learning_rate": 1.9540546525692736e-05,
+      "loss": 0.8304,
+      "step": 734
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0127346912511326,
+      "learning_rate": 1.9538905160048082e-05,
+      "loss": 0.8252,
+      "step": 735
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.269093380984089,
+      "learning_rate": 1.9537260936980793e-05,
+      "loss": 1.3468,
+      "step": 736
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0737877536687457,
+      "learning_rate": 1.9535613856983408e-05,
+      "loss": 0.7515,
+      "step": 737
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.254303188425808,
+      "learning_rate": 1.9533963920549307e-05,
+      "loss": 0.9569,
+      "step": 738
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8714043684137062,
+      "learning_rate": 1.9532311128172744e-05,
+      "loss": 0.6541,
+      "step": 739
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.851708185618675,
+      "learning_rate": 1.9530655480348823e-05,
+      "loss": 0.7475,
+      "step": 740
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.048119258997242,
+      "learning_rate": 1.952899697757349e-05,
+      "loss": 0.7368,
+      "step": 741
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0276678518821705,
+      "learning_rate": 1.9527335620343567e-05,
+      "loss": 1.0289,
+      "step": 742
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1920563142361915,
+      "learning_rate": 1.952567140915672e-05,
+      "loss": 0.9675,
+      "step": 743
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0715659552312737,
+      "learning_rate": 1.9524004344511463e-05,
+      "loss": 0.8697,
+      "step": 744
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1991356127902424,
+      "learning_rate": 1.9522334426907185e-05,
+      "loss": 1.0615,
+      "step": 745
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4141812752173493,
+      "learning_rate": 1.9520661656844107e-05,
+      "loss": 1.0416,
+      "step": 746
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4247136453622244,
+      "learning_rate": 1.951898603482332e-05,
+      "loss": 0.6111,
+      "step": 747
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.523847179914613,
+      "learning_rate": 1.9517307561346762e-05,
+      "loss": 0.9368,
+      "step": 748
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.079955406683968,
+      "learning_rate": 1.951562623691723e-05,
+      "loss": 0.7764,
+      "step": 749
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1097167621046584,
+      "learning_rate": 1.9513942062038375e-05,
+      "loss": 1.1544,
+      "step": 750
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9965533019689123,
+      "learning_rate": 1.951225503721469e-05,
+      "loss": 1.189,
+      "step": 751
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3220476762340716,
+      "learning_rate": 1.9510565162951538e-05,
+      "loss": 0.9875,
+      "step": 752
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7072854807767459,
+      "learning_rate": 1.9508872439755128e-05,
+      "loss": 1.0234,
+      "step": 753
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.968944248577686,
+      "learning_rate": 1.9507176868132517e-05,
+      "loss": 0.8964,
+      "step": 754
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0188772425840784,
+      "learning_rate": 1.9505478448591632e-05,
+      "loss": 1.0198,
+      "step": 755
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1614955983886734,
+      "learning_rate": 1.9503777181641233e-05,
+      "loss": 0.7379,
+      "step": 756
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1068404120037179,
+      "learning_rate": 1.950207306779095e-05,
+      "loss": 1.0268,
+      "step": 757
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2581750183542166,
+      "learning_rate": 1.950036610755125e-05,
+      "loss": 0.8322,
+      "step": 758
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0675451853888347,
+      "learning_rate": 1.9498656301433466e-05,
+      "loss": 0.8777,
+      "step": 759
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9479105591068522,
+      "learning_rate": 1.9496943649949777e-05,
+      "loss": 0.7943,
+      "step": 760
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1955279764307862,
+      "learning_rate": 1.9495228153613216e-05,
+      "loss": 1.1979,
+      "step": 761
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3315967811982208,
+      "learning_rate": 1.9493509812937666e-05,
+      "loss": 0.9112,
+      "step": 762
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8123626698127167,
+      "learning_rate": 1.9491788628437866e-05,
+      "loss": 0.6024,
+      "step": 763
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1716085216203933,
+      "learning_rate": 1.9490064600629403e-05,
+      "loss": 0.9911,
+      "step": 764
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5472847332969857,
+      "learning_rate": 1.9488337730028717e-05,
+      "loss": 1.1101,
+      "step": 765
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.701586001177953,
+      "learning_rate": 1.94866080171531e-05,
+      "loss": 1.0049,
+      "step": 766
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.041919757771477,
+      "learning_rate": 1.94848754625207e-05,
+      "loss": 0.7513,
+      "step": 767
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0752413515491206,
+      "learning_rate": 1.9483140066650508e-05,
+      "loss": 0.8135,
+      "step": 768
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1521335180824868,
+      "learning_rate": 1.9481401830062367e-05,
+      "loss": 1.049,
+      "step": 769
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9629560997833431,
+      "learning_rate": 1.9479660753276978e-05,
+      "loss": 1.2039,
+      "step": 770
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3474058751364635,
+      "learning_rate": 1.947791683681589e-05,
+      "loss": 1.0857,
+      "step": 771
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9935357767184485,
+      "learning_rate": 1.9476170081201494e-05,
+      "loss": 0.7588,
+      "step": 772
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0758177942969178,
+      "learning_rate": 1.9474420486957045e-05,
+      "loss": 1.0463,
+      "step": 773
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9851979435361444,
+      "learning_rate": 1.947266805460664e-05,
+      "loss": 0.8418,
+      "step": 774
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.026495346800711,
+      "learning_rate": 1.9470912784675225e-05,
+      "loss": 1.0933,
+      "step": 775
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.268647231321864,
+      "learning_rate": 1.9469154677688605e-05,
+      "loss": 1.0346,
+      "step": 776
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8857156851818576,
+      "learning_rate": 1.9467393734173424e-05,
+      "loss": 1.0235,
+      "step": 777
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0296803214506773,
+      "learning_rate": 1.9465629954657185e-05,
+      "loss": 0.79,
+      "step": 778
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9646834260327546,
+      "learning_rate": 1.9463863339668232e-05,
+      "loss": 0.7725,
+      "step": 779
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0804213904818853,
+      "learning_rate": 1.9462093889735766e-05,
+      "loss": 1.1787,
+      "step": 780
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1994403902967359,
+      "learning_rate": 1.9460321605389834e-05,
+      "loss": 0.8588,
+      "step": 781
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8072446374179754,
+      "learning_rate": 1.9458546487161326e-05,
+      "loss": 0.6523,
+      "step": 782
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2577527275209799,
+      "learning_rate": 1.9456768535581994e-05,
+      "loss": 0.8274,
+      "step": 783
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0621911570504574,
+      "learning_rate": 1.9454987751184427e-05,
+      "loss": 1.1835,
+      "step": 784
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1635133952858785,
+      "learning_rate": 1.945320413450207e-05,
+      "loss": 1.0124,
+      "step": 785
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.814173628562641,
+      "learning_rate": 1.945141768606921e-05,
+      "loss": 0.6129,
+      "step": 786
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0132059951578736,
+      "learning_rate": 1.944962840642099e-05,
+      "loss": 0.7543,
+      "step": 787
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3026316618001434,
+      "learning_rate": 1.9447836296093393e-05,
+      "loss": 0.9425,
+      "step": 788
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0470119923547943,
+      "learning_rate": 1.944604135562326e-05,
+      "loss": 1.1274,
+      "step": 789
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2377073487052468,
+      "learning_rate": 1.9444243585548266e-05,
+      "loss": 0.8388,
+      "step": 790
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5615229947773357,
+      "learning_rate": 1.9442442986406944e-05,
+      "loss": 1.2069,
+      "step": 791
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1022668053409692,
+      "learning_rate": 1.944063955873867e-05,
+      "loss": 0.9134,
+      "step": 792
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.097773034093608,
+      "learning_rate": 1.9438833303083677e-05,
+      "loss": 0.8053,
+      "step": 793
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.077127312236201,
+      "learning_rate": 1.943702421998303e-05,
+      "loss": 1.06,
+      "step": 794
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0613858131115734,
+      "learning_rate": 1.9435212309978645e-05,
+      "loss": 0.9768,
+      "step": 795
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1280300626303834,
+      "learning_rate": 1.9433397573613298e-05,
+      "loss": 0.8301,
+      "step": 796
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9165308474048411,
+      "learning_rate": 1.9431580011430595e-05,
+      "loss": 0.8527,
+      "step": 797
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.194197466084327,
+      "learning_rate": 1.9429759623974992e-05,
+      "loss": 0.9686,
+      "step": 798
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.92415966991595,
+      "learning_rate": 1.9427936411791802e-05,
+      "loss": 0.8104,
+      "step": 799
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2007438111248574,
+      "learning_rate": 1.9426110375427175e-05,
+      "loss": 0.7328,
+      "step": 800
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.09918114405879,
+      "learning_rate": 1.94242815154281e-05,
+      "loss": 1.0308,
+      "step": 801
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0793233547783136,
+      "learning_rate": 1.942244983234243e-05,
+      "loss": 0.6786,
+      "step": 802
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0479325218349096,
+      "learning_rate": 1.9420615326718852e-05,
+      "loss": 1.0343,
+      "step": 803
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2265522655566636,
+      "learning_rate": 1.9418777999106903e-05,
+      "loss": 0.989,
+      "step": 804
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.578529480364538,
+      "learning_rate": 1.941693785005695e-05,
+      "loss": 0.9087,
+      "step": 805
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0836967590821378,
+      "learning_rate": 1.9415094880120236e-05,
+      "loss": 0.7251,
+      "step": 806
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2814165261572854,
+      "learning_rate": 1.941324908984882e-05,
+      "loss": 0.9596,
+      "step": 807
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0315759672285518,
+      "learning_rate": 1.9411400479795618e-05,
+      "loss": 1.1479,
+      "step": 808
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2866532739915941,
+      "learning_rate": 1.940954905051439e-05,
+      "loss": 0.9305,
+      "step": 809
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1054777146026176,
+      "learning_rate": 1.940769480255974e-05,
+      "loss": 0.8625,
+      "step": 810
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.247600586518675,
+      "learning_rate": 1.9405837736487126e-05,
+      "loss": 0.7291,
+      "step": 811
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9987510549079747,
+      "learning_rate": 1.9403977852852826e-05,
+      "loss": 0.6986,
+      "step": 812
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3074422774855123,
+      "learning_rate": 1.9402115152213982e-05,
+      "loss": 1.0651,
+      "step": 813
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2213217041870008,
+      "learning_rate": 1.9400249635128575e-05,
+      "loss": 1.135,
+      "step": 814
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.04973285010659,
+      "learning_rate": 1.9398381302155435e-05,
+      "loss": 0.7378,
+      "step": 815
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0115298090775997,
+      "learning_rate": 1.9396510153854215e-05,
+      "loss": 0.7066,
+      "step": 816
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1998688663816468,
+      "learning_rate": 1.939463619078544e-05,
+      "loss": 0.8787,
+      "step": 817
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0950299331994804,
+      "learning_rate": 1.939275941351046e-05,
+      "loss": 1.1235,
+      "step": 818
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.086052533496378,
+      "learning_rate": 1.9390879822591473e-05,
+      "loss": 0.797,
+      "step": 819
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1726981196748167,
+      "learning_rate": 1.9388997418591518e-05,
+      "loss": 0.9172,
+      "step": 820
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4480850420371854,
+      "learning_rate": 1.938711220207447e-05,
+      "loss": 1.1676,
+      "step": 821
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.109577123571445,
+      "learning_rate": 1.9385224173605072e-05,
+      "loss": 1.1523,
+      "step": 822
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0763090646682287,
+      "learning_rate": 1.9383333333748875e-05,
+      "loss": 1.0056,
+      "step": 823
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4368942539000493,
+      "learning_rate": 1.9381439683072296e-05,
+      "loss": 0.8269,
+      "step": 824
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.220211326538106,
+      "learning_rate": 1.9379543222142584e-05,
+      "loss": 0.8977,
+      "step": 825
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.229396758960435,
+      "learning_rate": 1.9377643951527834e-05,
+      "loss": 0.7843,
+      "step": 826
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6177665232424525,
+      "learning_rate": 1.9375741871796982e-05,
+      "loss": 1.1114,
+      "step": 827
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1089679405274946,
+      "learning_rate": 1.9373836983519807e-05,
+      "loss": 0.9373,
+      "step": 828
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0778270439129907,
+      "learning_rate": 1.937192928726692e-05,
+      "loss": 0.6777,
+      "step": 829
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.470076344675518,
+      "learning_rate": 1.937001878360979e-05,
+      "loss": 0.9831,
+      "step": 830
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1528881653247323,
+      "learning_rate": 1.9368105473120704e-05,
+      "loss": 0.7977,
+      "step": 831
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3393975935609193,
+      "learning_rate": 1.9366189356372813e-05,
+      "loss": 1.142,
+      "step": 832
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2832753724394799,
+      "learning_rate": 1.9364270433940096e-05,
+      "loss": 1.0228,
+      "step": 833
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.128369542508823,
+      "learning_rate": 1.9362348706397374e-05,
+      "loss": 0.8443,
+      "step": 834
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2243301780994544,
+      "learning_rate": 1.936042417432031e-05,
+      "loss": 0.987,
+      "step": 835
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9706951888327247,
+      "learning_rate": 1.9358496838285408e-05,
+      "loss": 0.7666,
+      "step": 836
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3801350519834585,
+      "learning_rate": 1.9356566698870006e-05,
+      "loss": 0.9307,
+      "step": 837
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7824477890098847,
+      "learning_rate": 1.935463375665229e-05,
+      "loss": 0.5985,
+      "step": 838
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9938523365821846,
+      "learning_rate": 1.9352698012211277e-05,
+      "loss": 0.8855,
+      "step": 839
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5791263719958613,
+      "learning_rate": 1.9350759466126838e-05,
+      "loss": 0.8136,
+      "step": 840
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1935857683673243,
+      "learning_rate": 1.934881811897966e-05,
+      "loss": 1.1712,
+      "step": 841
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2605933178636184,
+      "learning_rate": 1.934687397135129e-05,
+      "loss": 0.9375,
+      "step": 842
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9789265093310145,
+      "learning_rate": 1.9344927023824112e-05,
+      "loss": 0.7789,
+      "step": 843
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.127326387907762,
+      "learning_rate": 1.934297727698133e-05,
+      "loss": 0.9382,
+      "step": 844
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0356387703363552,
+      "learning_rate": 1.934102473140701e-05,
+      "loss": 0.969,
+      "step": 845
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0763286027536738,
+      "learning_rate": 1.933906938768604e-05,
+      "loss": 1.1073,
+      "step": 846
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2879634296518978,
+      "learning_rate": 1.9337111246404156e-05,
+      "loss": 1.0617,
+      "step": 847
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9921837301333384,
+      "learning_rate": 1.9335150308147924e-05,
+      "loss": 0.8167,
+      "step": 848
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2394982825616039,
+      "learning_rate": 1.9333186573504753e-05,
+      "loss": 0.9813,
+      "step": 849
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.084467918257116,
+      "learning_rate": 1.9331220043062894e-05,
+      "loss": 0.8895,
+      "step": 850
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0702652605170329,
+      "learning_rate": 1.9329250717411426e-05,
+      "loss": 1.1414,
+      "step": 851
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.908940345153813,
+      "learning_rate": 1.9327278597140268e-05,
+      "loss": 1.034,
+      "step": 852
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0758753426003205,
+      "learning_rate": 1.932530368284018e-05,
+      "loss": 0.7396,
+      "step": 853
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0495747988493886,
+      "learning_rate": 1.9323325975102757e-05,
+      "loss": 0.5801,
+      "step": 854
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0748581767815162,
+      "learning_rate": 1.9321345474520424e-05,
+      "loss": 1.1047,
+      "step": 855
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1336537064864165,
+      "learning_rate": 1.931936218168646e-05,
+      "loss": 1.2148,
+      "step": 856
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3481216973906989,
+      "learning_rate": 1.9317376097194964e-05,
+      "loss": 0.723,
+      "step": 857
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1368882681266756,
+      "learning_rate": 1.9315387221640874e-05,
+      "loss": 0.7529,
+      "step": 858
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2008553626629712,
+      "learning_rate": 1.931339555561997e-05,
+      "loss": 1.0402,
+      "step": 859
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9180801755572948,
+      "learning_rate": 1.9311401099728865e-05,
+      "loss": 0.873,
+      "step": 860
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1984306026369431,
+      "learning_rate": 1.9309403854565e-05,
+      "loss": 1.2818,
+      "step": 861
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2163320649095284,
+      "learning_rate": 1.9307403820726675e-05,
+      "loss": 0.6152,
+      "step": 862
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9890130283518421,
+      "learning_rate": 1.9305400998812994e-05,
+      "loss": 0.8167,
+      "step": 863
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2113675131243966,
+      "learning_rate": 1.9303395389423918e-05,
+      "loss": 0.8258,
+      "step": 864
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.216059186169061,
+      "learning_rate": 1.9301386993160235e-05,
+      "loss": 1.2184,
+      "step": 865
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.622509157336192,
+      "learning_rate": 1.929937581062357e-05,
+      "loss": 0.9742,
+      "step": 866
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.082963217402141,
+      "learning_rate": 1.9297361842416383e-05,
+      "loss": 0.8391,
+      "step": 867
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5733514091073721,
+      "learning_rate": 1.9295345089141962e-05,
+      "loss": 0.8504,
+      "step": 868
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0523467068027776,
+      "learning_rate": 1.9293325551404444e-05,
+      "loss": 0.9777,
+      "step": 869
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1022198933996765,
+      "learning_rate": 1.929130322980878e-05,
+      "loss": 1.2074,
+      "step": 870
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.059997181895445,
+      "learning_rate": 1.9289278124960777e-05,
+      "loss": 0.6891,
+      "step": 871
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1970326498407753,
+      "learning_rate": 1.9287250237467053e-05,
+      "loss": 1.0162,
+      "step": 872
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8742740262004398,
+      "learning_rate": 1.9285219567935078e-05,
+      "loss": 0.588,
+      "step": 873
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2186467732075255,
+      "learning_rate": 1.9283186116973146e-05,
+      "loss": 0.7101,
+      "step": 874
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.08760508590206,
+      "learning_rate": 1.928114988519039e-05,
+      "loss": 1.1299,
+      "step": 875
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.105815123538476,
+      "learning_rate": 1.927911087319677e-05,
+      "loss": 1.0084,
+      "step": 876
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3598955897972629,
+      "learning_rate": 1.927706908160308e-05,
+      "loss": 0.8405,
+      "step": 877
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8991511625612272,
+      "learning_rate": 1.927502451102095e-05,
+      "loss": 0.948,
+      "step": 878
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1700404272297198,
+      "learning_rate": 1.9272977162062835e-05,
+      "loss": 0.8369,
+      "step": 879
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1496917140312206,
+      "learning_rate": 1.927092703534204e-05,
+      "loss": 1.2317,
+      "step": 880
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0232187320357744,
+      "learning_rate": 1.9268874131472676e-05,
+      "loss": 0.7511,
+      "step": 881
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1470732867223499,
+      "learning_rate": 1.926681845106971e-05,
+      "loss": 0.9248,
+      "step": 882
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1520409269969814,
+      "learning_rate": 1.9264759994748924e-05,
+      "loss": 0.8487,
+      "step": 883
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0111374541543867,
+      "learning_rate": 1.926269876312694e-05,
+      "loss": 1.1358,
+      "step": 884
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5926472615814236,
+      "learning_rate": 1.926063475682121e-05,
+      "loss": 1.0058,
+      "step": 885
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2810491008572462,
+      "learning_rate": 1.925856797645002e-05,
+      "loss": 0.9369,
+      "step": 886
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8864676479244558,
+      "learning_rate": 1.925649842263248e-05,
+      "loss": 0.9845,
+      "step": 887
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4117479125132777,
+      "learning_rate": 1.925442609598853e-05,
+      "loss": 0.6471,
+      "step": 888
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.030980963415441,
+      "learning_rate": 1.9252350997138948e-05,
+      "loss": 1.1935,
+      "step": 889
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1717508359305526,
+      "learning_rate": 1.925027312670534e-05,
+      "loss": 0.6419,
+      "step": 890
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2174200622512739,
+      "learning_rate": 1.9248192485310147e-05,
+      "loss": 0.9703,
+      "step": 891
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1678296300719504,
+      "learning_rate": 1.924610907357663e-05,
+      "loss": 0.8223,
+      "step": 892
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0411770154101894,
+      "learning_rate": 1.9244022892128877e-05,
+      "loss": 0.8438,
+      "step": 893
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8553825527748675,
+      "learning_rate": 1.9241933941591826e-05,
+      "loss": 1.0155,
+      "step": 894
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.126029847228734,
+      "learning_rate": 1.9239842222591226e-05,
+      "loss": 0.9201,
+      "step": 895
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.249613612923545,
+      "learning_rate": 1.923774773575366e-05,
+      "loss": 0.8652,
+      "step": 896
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9564775981788904,
+      "learning_rate": 1.9235650481706542e-05,
+      "loss": 0.9073,
+      "step": 897
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5417580147728243,
+      "learning_rate": 1.9233550461078114e-05,
+      "loss": 0.7103,
+      "step": 898
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.123938316399307,
+      "learning_rate": 1.9231447674497444e-05,
+      "loss": 1.1188,
+      "step": 899
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9876401825839051,
+      "learning_rate": 1.922934212259444e-05,
+      "loss": 1.1252,
+      "step": 900
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9978542273519061,
+      "learning_rate": 1.922723380599982e-05,
+      "loss": 1.0028,
+      "step": 901
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.002962096581876,
+      "learning_rate": 1.9225122725345145e-05,
+      "loss": 0.7931,
+      "step": 902
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1295593284136838,
+      "learning_rate": 1.9223008881262796e-05,
+      "loss": 1.1974,
+      "step": 903
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2521535862315538,
+      "learning_rate": 1.922089227438599e-05,
+      "loss": 0.9104,
+      "step": 904
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.007599133521679,
+      "learning_rate": 1.9218772905348756e-05,
+      "loss": 0.8228,
+      "step": 905
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.134929318663513,
+      "learning_rate": 1.9216650774785975e-05,
+      "loss": 0.7867,
+      "step": 906
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0623707665957145,
+      "learning_rate": 1.921452588333333e-05,
+      "loss": 0.771,
+      "step": 907
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9730383211337779,
+      "learning_rate": 1.9212398231627343e-05,
+      "loss": 0.8033,
+      "step": 908
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0056728102282324,
+      "learning_rate": 1.9210267820305366e-05,
+      "loss": 0.9513,
+      "step": 909
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1435686199189532,
+      "learning_rate": 1.920813465000557e-05,
+      "loss": 0.7943,
+      "step": 910
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7446358355965481,
+      "learning_rate": 1.920599872136696e-05,
+      "loss": 0.9128,
+      "step": 911
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.152644516307872,
+      "learning_rate": 1.9203860035029357e-05,
+      "loss": 0.8799,
+      "step": 912
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0788788305721375,
+      "learning_rate": 1.9201718591633417e-05,
+      "loss": 1.1743,
+      "step": 913
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0575388988469738,
+      "learning_rate": 1.9199574391820627e-05,
+      "loss": 0.6493,
+      "step": 914
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.323779889091937,
+      "learning_rate": 1.9197427436233278e-05,
+      "loss": 0.9469,
+      "step": 915
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0704369381407295,
+      "learning_rate": 1.919527772551451e-05,
+      "loss": 1.0181,
+      "step": 916
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2558210099279834,
+      "learning_rate": 1.919312526030827e-05,
+      "loss": 1.0226,
+      "step": 917
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.241971307031594,
+      "learning_rate": 1.9190970041259352e-05,
+      "loss": 1.2125,
+      "step": 918
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0021983459682362,
+      "learning_rate": 1.9188812069013352e-05,
+      "loss": 0.8654,
+      "step": 919
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9428110376070552,
+      "learning_rate": 1.9186651344216703e-05,
+      "loss": 1.0566,
+      "step": 920
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.051120829184323,
+      "learning_rate": 1.9184487867516657e-05,
+      "loss": 0.802,
+      "step": 921
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1262053909058392,
+      "learning_rate": 1.91823216395613e-05,
+      "loss": 0.8741,
+      "step": 922
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9891742992858287,
+      "learning_rate": 1.9180152660999526e-05,
+      "loss": 1.2769,
+      "step": 923
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0710059068752988,
+      "learning_rate": 1.9177980932481072e-05,
+      "loss": 0.7502,
+      "step": 924
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8190779811765558,
+      "learning_rate": 1.9175806454656485e-05,
+      "loss": 1.0406,
+      "step": 925
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4526681442646736,
+      "learning_rate": 1.917362922817714e-05,
+      "loss": 0.7805,
+      "step": 926
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7929256846401522,
+      "learning_rate": 1.9171449253695233e-05,
+      "loss": 0.8282,
+      "step": 927
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9731739223089927,
+      "learning_rate": 1.916926653186379e-05,
+      "loss": 0.7022,
+      "step": 928
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3129972554511964,
+      "learning_rate": 1.916708106333665e-05,
+      "loss": 0.8765,
+      "step": 929
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2480373514449041,
+      "learning_rate": 1.9164892848768485e-05,
+      "loss": 0.8426,
+      "step": 930
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.268019368373636,
+      "learning_rate": 1.916270188881478e-05,
+      "loss": 0.7846,
+      "step": 931
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1581387380800177,
+      "learning_rate": 1.9160508184131853e-05,
+      "loss": 1.2224,
+      "step": 932
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0361599549282958,
+      "learning_rate": 1.915831173537683e-05,
+      "loss": 0.8255,
+      "step": 933
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9268106366219236,
+      "learning_rate": 1.9156112543207674e-05,
+      "loss": 0.691,
+      "step": 934
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.113487462703542,
+      "learning_rate": 1.9153910608283162e-05,
+      "loss": 0.9488,
+      "step": 935
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.080425713723176,
+      "learning_rate": 1.9151705931262895e-05,
+      "loss": 0.9866,
+      "step": 936
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9200110649247928,
+      "learning_rate": 1.9149498512807287e-05,
+      "loss": 0.9398,
+      "step": 937
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1416045679463873,
+      "learning_rate": 1.914728835357759e-05,
+      "loss": 0.87,
+      "step": 938
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0604805481949224,
+      "learning_rate": 1.914507545423586e-05,
+      "loss": 0.9432,
+      "step": 939
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9098130581641932,
+      "learning_rate": 1.9142859815444982e-05,
+      "loss": 1.0149,
+      "step": 940
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1540948079193658,
+      "learning_rate": 1.9140641437868664e-05,
+      "loss": 0.9089,
+      "step": 941
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8573684047155641,
+      "learning_rate": 1.913842032217143e-05,
+      "loss": 1.0396,
+      "step": 942
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1384359461026197,
+      "learning_rate": 1.9136196469018623e-05,
+      "loss": 0.9073,
+      "step": 943
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7549324641622942,
+      "learning_rate": 1.9133969879076412e-05,
+      "loss": 0.6609,
+      "step": 944
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3555746036841736,
+      "learning_rate": 1.913174055301178e-05,
+      "loss": 0.8083,
+      "step": 945
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0827424273399109,
+      "learning_rate": 1.912950849149253e-05,
+      "loss": 1.3551,
+      "step": 946
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.227584077857669,
+      "learning_rate": 1.912727369518729e-05,
+      "loss": 0.8706,
+      "step": 947
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.277907487556059,
+      "learning_rate": 1.9125036164765502e-05,
+      "loss": 1.0984,
+      "step": 948
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9239580798224117,
+      "learning_rate": 1.9122795900897428e-05,
+      "loss": 1.0195,
+      "step": 949
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1680151874150235,
+      "learning_rate": 1.912055290425415e-05,
+      "loss": 0.9248,
+      "step": 950
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0077961263941935,
+      "learning_rate": 1.9118307175507567e-05,
+      "loss": 1.0239,
+      "step": 951
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0618672429764717,
+      "learning_rate": 1.9116058715330396e-05,
+      "loss": 0.7845,
+      "step": 952
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9397926949492307,
+      "learning_rate": 1.911380752439618e-05,
+      "loss": 0.5994,
+      "step": 953
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0041478180233663,
+      "learning_rate": 1.911155360337926e-05,
+      "loss": 0.9404,
+      "step": 954
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1833540600154666,
+      "learning_rate": 1.9109296952954826e-05,
+      "loss": 0.8132,
+      "step": 955
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6021569953233414,
+      "learning_rate": 1.9107037573798858e-05,
+      "loss": 1.1581,
+      "step": 956
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9619936307308521,
+      "learning_rate": 1.9104775466588162e-05,
+      "loss": 0.8528,
+      "step": 957
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2362747365053157,
+      "learning_rate": 1.9102510632000365e-05,
+      "loss": 0.9078,
+      "step": 958
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0803128440593195,
+      "learning_rate": 1.910024307071391e-05,
+      "loss": 0.8572,
+      "step": 959
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3429724365788427,
+      "learning_rate": 1.909797278340805e-05,
+      "loss": 0.8367,
+      "step": 960
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0766569337509049,
+      "learning_rate": 1.9095699770762866e-05,
+      "loss": 1.1283,
+      "step": 961
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5885335681732502,
+      "learning_rate": 1.909342403345925e-05,
+      "loss": 0.5523,
+      "step": 962
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3180697367969543,
+      "learning_rate": 1.9091145572178905e-05,
+      "loss": 1.0823,
+      "step": 963
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9635004674815335,
+      "learning_rate": 1.9088864387604355e-05,
+      "loss": 0.7518,
+      "step": 964
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.947173004959862,
+      "learning_rate": 1.908658048041894e-05,
+      "loss": 1.3422,
+      "step": 965
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.997828549063244,
+      "learning_rate": 1.9084293851306815e-05,
+      "loss": 0.6988,
+      "step": 966
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9978094417112805,
+      "learning_rate": 1.9082004500952953e-05,
+      "loss": 0.886,
+      "step": 967
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2594581966949676,
+      "learning_rate": 1.9079712430043136e-05,
+      "loss": 0.89,
+      "step": 968
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0498228412789732,
+      "learning_rate": 1.9077417639263966e-05,
+      "loss": 0.784,
+      "step": 969
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.119397282313796,
+      "learning_rate": 1.9075120129302857e-05,
+      "loss": 1.1724,
+      "step": 970
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1193417493121265,
+      "learning_rate": 1.9072819900848037e-05,
+      "loss": 0.916,
+      "step": 971
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1735092221870416,
+      "learning_rate": 1.907051695458855e-05,
+      "loss": 0.8211,
+      "step": 972
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2247788956168748,
+      "learning_rate": 1.906821129121426e-05,
+      "loss": 0.9543,
+      "step": 973
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.861137437561905,
+      "learning_rate": 1.9065902911415836e-05,
+      "loss": 1.0422,
+      "step": 974
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9956292987509492,
+      "learning_rate": 1.9063591815884758e-05,
+      "loss": 0.9776,
+      "step": 975
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1350533552167947,
+      "learning_rate": 1.906127800531333e-05,
+      "loss": 0.8451,
+      "step": 976
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0755190431468742,
+      "learning_rate": 1.905896148039467e-05,
+      "loss": 0.9053,
+      "step": 977
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.141925522181747,
+      "learning_rate": 1.905664224182269e-05,
+      "loss": 0.8416,
+      "step": 978
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6638248103500659,
+      "learning_rate": 1.9054320290292143e-05,
+      "loss": 0.5885,
+      "step": 979
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1621612723359842,
+      "learning_rate": 1.905199562649857e-05,
+      "loss": 1.1976,
+      "step": 980
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9002696788471642,
+      "learning_rate": 1.904966825113834e-05,
+      "loss": 0.8022,
+      "step": 981
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.030281726599579,
+      "learning_rate": 1.9047338164908628e-05,
+      "loss": 0.8595,
+      "step": 982
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1016761957679935,
+      "learning_rate": 1.9045005368507418e-05,
+      "loss": 0.8495,
+      "step": 983
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1723723004282887,
+      "learning_rate": 1.9042669862633514e-05,
+      "loss": 1.1391,
+      "step": 984
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2056694598242421,
+      "learning_rate": 1.9040331647986526e-05,
+      "loss": 0.8843,
+      "step": 985
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0338409800487465,
+      "learning_rate": 1.903799072526688e-05,
+      "loss": 0.8034,
+      "step": 986
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9241825915486948,
+      "learning_rate": 1.9035647095175807e-05,
+      "loss": 0.8811,
+      "step": 987
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.159265445275835,
+      "learning_rate": 1.9033300758415354e-05,
+      "loss": 0.961,
+      "step": 988
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.360492479646211,
+      "learning_rate": 1.903095171568837e-05,
+      "loss": 1.3685,
+      "step": 989
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0285710785855295,
+      "learning_rate": 1.9028599967698533e-05,
+      "loss": 0.7939,
+      "step": 990
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.96549048003609,
+      "learning_rate": 1.902624551515031e-05,
+      "loss": 0.6779,
+      "step": 991
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0784082925106355,
+      "learning_rate": 1.9023888358748994e-05,
+      "loss": 0.8037,
+      "step": 992
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.158368038955191,
+      "learning_rate": 1.9021528499200677e-05,
+      "loss": 0.8727,
+      "step": 993
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0515308907450014,
+      "learning_rate": 1.9019165937212273e-05,
+      "loss": 0.786,
+      "step": 994
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0911872518174823,
+      "learning_rate": 1.901680067349149e-05,
+      "loss": 1.1321,
+      "step": 995
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.925143868230278,
+      "learning_rate": 1.9014432708746856e-05,
+      "loss": 0.8995,
+      "step": 996
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1283679480169355,
+      "learning_rate": 1.9012062043687713e-05,
+      "loss": 0.8448,
+      "step": 997
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0286296983413823,
+      "learning_rate": 1.900968867902419e-05,
+      "loss": 0.8911,
+      "step": 998
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0137752004247202,
+      "learning_rate": 1.9007312615467258e-05,
+      "loss": 1.0527,
+      "step": 999
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0253530526023122,
+      "learning_rate": 1.900493385372866e-05,
+      "loss": 0.8644,
+      "step": 1000
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0315537751195052,
+      "learning_rate": 1.9002552394520976e-05,
+      "loss": 0.903,
+      "step": 1001
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1103709309395349,
+      "learning_rate": 1.9000168238557574e-05,
+      "loss": 0.809,
+      "step": 1002
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.219304195819454,
+      "learning_rate": 1.899778138655265e-05,
+      "loss": 0.8083,
+      "step": 1003
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0958014717257258,
+      "learning_rate": 1.899539183922119e-05,
+      "loss": 1.1506,
+      "step": 1004
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9245548791625859,
+      "learning_rate": 1.899299959727899e-05,
+      "loss": 0.5802,
+      "step": 1005
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9966838888574404,
+      "learning_rate": 1.899060466144267e-05,
+      "loss": 0.7665,
+      "step": 1006
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1612903463628366,
+      "learning_rate": 1.898820703242963e-05,
+      "loss": 0.9749,
+      "step": 1007
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9779906341249063,
+      "learning_rate": 1.8985806710958094e-05,
+      "loss": 0.8098,
+      "step": 1008
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1812334922685412,
+      "learning_rate": 1.8983403697747094e-05,
+      "loss": 1.1516,
+      "step": 1009
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0354382087457255,
+      "learning_rate": 1.8980997993516462e-05,
+      "loss": 1.1319,
+      "step": 1010
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9695272164554363,
+      "learning_rate": 1.8978589598986838e-05,
+      "loss": 0.899,
+      "step": 1011
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.425048191027762,
+      "learning_rate": 1.8976178514879663e-05,
+      "loss": 1.0705,
+      "step": 1012
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.12327801958672,
+      "learning_rate": 1.897376474191719e-05,
+      "loss": 1.1869,
+      "step": 1013
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0950284176613745,
+      "learning_rate": 1.897134828082248e-05,
+      "loss": 0.745,
+      "step": 1014
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5606282971637848,
+      "learning_rate": 1.8968929132319394e-05,
+      "loss": 0.5246,
+      "step": 1015
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1423115882119608,
+      "learning_rate": 1.8966507297132596e-05,
+      "loss": 0.8501,
+      "step": 1016
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0529980378563393,
+      "learning_rate": 1.896408277598756e-05,
+      "loss": 0.9345,
+      "step": 1017
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9702105187252507,
+      "learning_rate": 1.8961655569610557e-05,
+      "loss": 1.0935,
+      "step": 1018
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2695316848257663,
+      "learning_rate": 1.8959225678728678e-05,
+      "loss": 0.8314,
+      "step": 1019
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1072889526027592,
+      "learning_rate": 1.8956793104069797e-05,
+      "loss": 0.8172,
+      "step": 1020
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8992978608441612,
+      "learning_rate": 1.895435784636261e-05,
+      "loss": 0.696,
+      "step": 1021
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0935769613923192,
+      "learning_rate": 1.895191990633661e-05,
+      "loss": 1.0573,
+      "step": 1022
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8709624092602829,
+      "learning_rate": 1.8949479284722087e-05,
+      "loss": 0.9627,
+      "step": 1023
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0808486237938455,
+      "learning_rate": 1.8947035982250145e-05,
+      "loss": 0.7928,
+      "step": 1024
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9276873900358711,
+      "learning_rate": 1.8944589999652687e-05,
+      "loss": 0.9226,
+      "step": 1025
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0119516728624458,
+      "learning_rate": 1.8942141337662415e-05,
+      "loss": 0.9465,
+      "step": 1026
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2176070882008119,
+      "learning_rate": 1.893968999701284e-05,
+      "loss": 0.8454,
+      "step": 1027
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0049430775123818,
+      "learning_rate": 1.8937235978438272e-05,
+      "loss": 1.1136,
+      "step": 1028
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.051650339058944,
+      "learning_rate": 1.893477928267382e-05,
+      "loss": 0.9784,
+      "step": 1029
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0402141331567962,
+      "learning_rate": 1.89323199104554e-05,
+      "loss": 1.0739,
+      "step": 1030
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3524518171340263,
+      "learning_rate": 1.892985786251973e-05,
+      "loss": 0.9164,
+      "step": 1031
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.091373296882657,
+      "learning_rate": 1.8927393139604327e-05,
+      "loss": 1.1114,
+      "step": 1032
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2653753321231713,
+      "learning_rate": 1.892492574244751e-05,
+      "loss": 0.843,
+      "step": 1033
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2067413679430383,
+      "learning_rate": 1.89224556717884e-05,
+      "loss": 0.8099,
+      "step": 1034
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7556619802076903,
+      "learning_rate": 1.8919982928366917e-05,
+      "loss": 0.5174,
+      "step": 1035
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1017905785189521,
+      "learning_rate": 1.8917507512923778e-05,
+      "loss": 1.0501,
+      "step": 1036
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9886383073310552,
+      "learning_rate": 1.891502942620051e-05,
+      "loss": 1.1506,
+      "step": 1037
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1358658076566626,
+      "learning_rate": 1.891254866893944e-05,
+      "loss": 0.8632,
+      "step": 1038
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2256691821010295,
+      "learning_rate": 1.891006524188368e-05,
+      "loss": 1.0571,
+      "step": 1039
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0033815786495668,
+      "learning_rate": 1.8907579145777156e-05,
+      "loss": 0.6774,
+      "step": 1040
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9899581256065064,
+      "learning_rate": 1.8905090381364596e-05,
+      "loss": 0.8874,
+      "step": 1041
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9820446887688561,
+      "learning_rate": 1.8902598949391507e-05,
+      "loss": 1.0649,
+      "step": 1042
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2436911588064876,
+      "learning_rate": 1.8900104850604224e-05,
+      "loss": 0.7969,
+      "step": 1043
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0581789886023611,
+      "learning_rate": 1.8897608085749853e-05,
+      "loss": 0.7604,
+      "step": 1044
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2357918208258396,
+      "learning_rate": 1.8895108655576318e-05,
+      "loss": 0.7711,
+      "step": 1045
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4139585784717148,
+      "learning_rate": 1.8892606560832335e-05,
+      "loss": 1.047,
+      "step": 1046
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4172217826112297,
+      "learning_rate": 1.8890101802267415e-05,
+      "loss": 1.1237,
+      "step": 1047
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.138618562421369,
+      "learning_rate": 1.8887594380631875e-05,
+      "loss": 1.0746,
+      "step": 1048
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9455773279317367,
+      "learning_rate": 1.8885084296676816e-05,
+      "loss": 0.792,
+      "step": 1049
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2786808164807801,
+      "learning_rate": 1.8882571551154152e-05,
+      "loss": 0.8336,
+      "step": 1050
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0806462710193578,
+      "learning_rate": 1.8880056144816585e-05,
+      "loss": 1.0677,
+      "step": 1051
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1216536780506112,
+      "learning_rate": 1.8877538078417616e-05,
+      "loss": 1.0131,
+      "step": 1052
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1240212247658539,
+      "learning_rate": 1.8875017352711547e-05,
+      "loss": 0.6349,
+      "step": 1053
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0760929540316113,
+      "learning_rate": 1.8872493968453467e-05,
+      "loss": 1.0519,
+      "step": 1054
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2416254284886579,
+      "learning_rate": 1.886996792639927e-05,
+      "loss": 0.7866,
+      "step": 1055
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2127318498784134,
+      "learning_rate": 1.8867439227305647e-05,
+      "loss": 1.0868,
+      "step": 1056
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1549177271331346,
+      "learning_rate": 1.8864907871930077e-05,
+      "loss": 0.8467,
+      "step": 1057
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3265992474043025,
+      "learning_rate": 1.8862373861030838e-05,
+      "loss": 0.8503,
+      "step": 1058
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1386166295849485,
+      "learning_rate": 1.885983719536701e-05,
+      "loss": 1.088,
+      "step": 1059
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0465626984942606,
+      "learning_rate": 1.8857297875698455e-05,
+      "loss": 0.9504,
+      "step": 1060
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.127449172066196,
+      "learning_rate": 1.8854755902785846e-05,
+      "loss": 1.0484,
+      "step": 1061
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3093020944153684,
+      "learning_rate": 1.885221127739064e-05,
+      "loss": 0.9159,
+      "step": 1062
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.415796838207878,
+      "learning_rate": 1.8849664000275084e-05,
+      "loss": 0.9807,
+      "step": 1063
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2258993480615195,
+      "learning_rate": 1.884711407220224e-05,
+      "loss": 0.9581,
+      "step": 1064
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0232724139235787,
+      "learning_rate": 1.884456149393594e-05,
+      "loss": 1.1581,
+      "step": 1065
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2228845915539313,
+      "learning_rate": 1.8842006266240826e-05,
+      "loss": 0.8723,
+      "step": 1066
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2545107724135214,
+      "learning_rate": 1.883944838988232e-05,
+      "loss": 0.8773,
+      "step": 1067
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1416599579215436,
+      "learning_rate": 1.8836887865626656e-05,
+      "loss": 0.7268,
+      "step": 1068
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1268006564236401,
+      "learning_rate": 1.8834324694240847e-05,
+      "loss": 0.8868,
+      "step": 1069
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.391691984146576,
+      "learning_rate": 1.88317588764927e-05,
+      "loss": 0.7833,
+      "step": 1070
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.145687094288742,
+      "learning_rate": 1.882919041315082e-05,
+      "loss": 1.0757,
+      "step": 1071
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4482062810749683,
+      "learning_rate": 1.88266193049846e-05,
+      "loss": 0.6076,
+      "step": 1072
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7198326803140827,
+      "learning_rate": 1.8824045552764226e-05,
+      "loss": 1.0243,
+      "step": 1073
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0898033479795874,
+      "learning_rate": 1.8821469157260687e-05,
+      "loss": 1.1538,
+      "step": 1074
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9563907975209561,
+      "learning_rate": 1.8818890119245742e-05,
+      "loss": 0.9823,
+      "step": 1075
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5271788650616012,
+      "learning_rate": 1.8816308439491955e-05,
+      "loss": 0.9739,
+      "step": 1076
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.425291983523779,
+      "learning_rate": 1.8813724118772686e-05,
+      "loss": 0.7719,
+      "step": 1077
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2782587683240039,
+      "learning_rate": 1.8811137157862084e-05,
+      "loss": 1.0256,
+      "step": 1078
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3351162703030905,
+      "learning_rate": 1.880854755753507e-05,
+      "loss": 0.769,
+      "step": 1079
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8306566095127473,
+      "learning_rate": 1.880595531856738e-05,
+      "loss": 0.7516,
+      "step": 1080
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.573710841924402,
+      "learning_rate": 1.8803360441735533e-05,
+      "loss": 0.9402,
+      "step": 1081
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.422157790814467,
+      "learning_rate": 1.8800762927816835e-05,
+      "loss": 0.9946,
+      "step": 1082
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.125004019571279,
+      "learning_rate": 1.8798162777589376e-05,
+      "loss": 0.8552,
+      "step": 1083
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0860375137652702,
+      "learning_rate": 1.8795559991832052e-05,
+      "loss": 0.6932,
+      "step": 1084
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.26495907412361,
+      "learning_rate": 1.8792954571324535e-05,
+      "loss": 1.0888,
+      "step": 1085
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.643328718946265,
+      "learning_rate": 1.8790346516847288e-05,
+      "loss": 0.8967,
+      "step": 1086
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0886958015553834,
+      "learning_rate": 1.878773582918157e-05,
+      "loss": 0.7573,
+      "step": 1087
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.520084740219217,
+      "learning_rate": 1.8785122509109425e-05,
+      "loss": 1.095,
+      "step": 1088
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9358652649489996,
+      "learning_rate": 1.8782506557413685e-05,
+      "loss": 0.5787,
+      "step": 1089
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.375696796416792,
+      "learning_rate": 1.8779887974877965e-05,
+      "loss": 1.0637,
+      "step": 1090
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6714094499258103,
+      "learning_rate": 1.877726676228668e-05,
+      "loss": 0.9914,
+      "step": 1091
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0228681416157486,
+      "learning_rate": 1.8774642920425018e-05,
+      "loss": 0.7817,
+      "step": 1092
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.302690566480895,
+      "learning_rate": 1.8772016450078972e-05,
+      "loss": 0.9328,
+      "step": 1093
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9688331646649241,
+      "learning_rate": 1.876938735203531e-05,
+      "loss": 1.0062,
+      "step": 1094
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2772451623097973,
+      "learning_rate": 1.8766755627081586e-05,
+      "loss": 1.1534,
+      "step": 1095
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.246502905432544,
+      "learning_rate": 1.8764121276006153e-05,
+      "loss": 0.811,
+      "step": 1096
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2613348274043468,
+      "learning_rate": 1.8761484299598135e-05,
+      "loss": 0.8602,
+      "step": 1097
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9366800682605562,
+      "learning_rate": 1.8758844698647457e-05,
+      "loss": 1.0511,
+      "step": 1098
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.215946157826358,
+      "learning_rate": 1.8756202473944818e-05,
+      "loss": 1.0626,
+      "step": 1099
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.681172452642843,
+      "learning_rate": 1.875355762628171e-05,
+      "loss": 0.8709,
+      "step": 1100
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1807170413658687,
+      "learning_rate": 1.8750910156450412e-05,
+      "loss": 0.8178,
+      "step": 1101
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.395224634660174,
+      "learning_rate": 1.8748260065243985e-05,
+      "loss": 1.0156,
+      "step": 1102
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3512792295527476,
+      "learning_rate": 1.8745607353456276e-05,
+      "loss": 0.6829,
+      "step": 1103
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6795291494488593,
+      "learning_rate": 1.8742952021881912e-05,
+      "loss": 1.2228,
+      "step": 1104
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5665016499811162,
+      "learning_rate": 1.8740294071316313e-05,
+      "loss": 0.8805,
+      "step": 1105
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.647784913029632,
+      "learning_rate": 1.8737633502555685e-05,
+      "loss": 0.7788,
+      "step": 1106
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.18923865971618,
+      "learning_rate": 1.8734970316397005e-05,
+      "loss": 0.857,
+      "step": 1107
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.455906404046145,
+      "learning_rate": 1.873230451363805e-05,
+      "loss": 0.8445,
+      "step": 1108
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4743846972741128,
+      "learning_rate": 1.8729636095077368e-05,
+      "loss": 1.2243,
+      "step": 1109
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4204507272324542,
+      "learning_rate": 1.8726965061514297e-05,
+      "loss": 0.8284,
+      "step": 1110
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2725768054559106,
+      "learning_rate": 1.8724291413748962e-05,
+      "loss": 0.9145,
+      "step": 1111
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4704684463569297,
+      "learning_rate": 1.8721615152582264e-05,
+      "loss": 0.9625,
+      "step": 1112
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2978823981125516,
+      "learning_rate": 1.8718936278815883e-05,
+      "loss": 1.0797,
+      "step": 1113
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5867592512892137,
+      "learning_rate": 1.8716254793252294e-05,
+      "loss": 1.0372,
+      "step": 1114
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1044644613374055,
+      "learning_rate": 1.871357069669475e-05,
+      "loss": 0.7523,
+      "step": 1115
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9689177371406539,
+      "learning_rate": 1.8710883989947278e-05,
+      "loss": 0.7815,
+      "step": 1116
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6870409960819883,
+      "learning_rate": 1.87081946738147e-05,
+      "loss": 0.9772,
+      "step": 1117
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2357875925899344,
+      "learning_rate": 1.870550274910261e-05,
+      "loss": 1.2773,
+      "step": 1118
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.358892867063466,
+      "learning_rate": 1.870280821661739e-05,
+      "loss": 0.8459,
+      "step": 1119
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2713732931413513,
+      "learning_rate": 1.8700111077166186e-05,
+      "loss": 0.8397,
+      "step": 1120
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.344205708356864,
+      "learning_rate": 1.8697411331556958e-05,
+      "loss": 0.8898,
+      "step": 1121
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8707145235838232,
+      "learning_rate": 1.869470898059841e-05,
+      "loss": 1.0128,
+      "step": 1122
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1813775761348804,
+      "learning_rate": 1.8692004025100054e-05,
+      "loss": 1.0246,
+      "step": 1123
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.108717503089494,
+      "learning_rate": 1.868929646587217e-05,
+      "loss": 0.8373,
+      "step": 1124
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1645050359622984,
+      "learning_rate": 1.8686586303725815e-05,
+      "loss": 0.5488,
+      "step": 1125
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1516754144273416,
+      "learning_rate": 1.868387353947284e-05,
+      "loss": 0.8668,
+      "step": 1126
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.387142081932942,
+      "learning_rate": 1.8681158173925854e-05,
+      "loss": 0.9691,
+      "step": 1127
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2517364657593335,
+      "learning_rate": 1.8678440207898264e-05,
+      "loss": 1.2146,
+      "step": 1128
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8609209690034908,
+      "learning_rate": 1.867571964220425e-05,
+      "loss": 0.521,
+      "step": 1129
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.009880416702178,
+      "learning_rate": 1.8672996477658767e-05,
+      "loss": 0.7442,
+      "step": 1130
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0145688226050105,
+      "learning_rate": 1.867027071507755e-05,
+      "loss": 1.0591,
+      "step": 1131
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.231975306893098,
+      "learning_rate": 1.866754235527712e-05,
+      "loss": 1.1295,
+      "step": 1132
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.072491475283073,
+      "learning_rate": 1.8664811399074763e-05,
+      "loss": 0.919,
+      "step": 1133
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1465540407650667,
+      "learning_rate": 1.8662077847288554e-05,
+      "loss": 0.7477,
+      "step": 1134
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4600865053921008,
+      "learning_rate": 1.8659341700737338e-05,
+      "loss": 0.7699,
+      "step": 1135
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.165657791596246,
+      "learning_rate": 1.8656602960240743e-05,
+      "loss": 0.7854,
+      "step": 1136
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2989419537877505,
+      "learning_rate": 1.8653861626619166e-05,
+      "loss": 1.2574,
+      "step": 1137
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3096411987497005,
+      "learning_rate": 1.8651117700693794e-05,
+      "loss": 0.8431,
+      "step": 1138
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9845723261704172,
+      "learning_rate": 1.8648371183286573e-05,
+      "loss": 1.1567,
+      "step": 1139
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1322549649574827,
+      "learning_rate": 1.8645622075220246e-05,
+      "loss": 0.8768,
+      "step": 1140
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0524065686654662,
+      "learning_rate": 1.8642870377318306e-05,
+      "loss": 0.8046,
+      "step": 1141
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1626129996445005,
+      "learning_rate": 1.8640116090405055e-05,
+      "loss": 0.5872,
+      "step": 1142
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2595096722425325,
+      "learning_rate": 1.8637359215305536e-05,
+      "loss": 1.2451,
+      "step": 1143
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.149522882119887,
+      "learning_rate": 1.8634599752845594e-05,
+      "loss": 0.8638,
+      "step": 1144
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0774851380788124,
+      "learning_rate": 1.8631837703851834e-05,
+      "loss": 0.894,
+      "step": 1145
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1491521052170068,
+      "learning_rate": 1.8629073069151638e-05,
+      "loss": 0.9952,
+      "step": 1146
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1322090506415985,
+      "learning_rate": 1.862630584957317e-05,
+      "loss": 1.1585,
+      "step": 1147
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1076204085605315,
+      "learning_rate": 1.8623536045945365e-05,
+      "loss": 0.7333,
+      "step": 1148
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0612510603804504,
+      "learning_rate": 1.8620763659097924e-05,
+      "loss": 0.7566,
+      "step": 1149
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0813066307293486,
+      "learning_rate": 1.8617988689861333e-05,
+      "loss": 0.887,
+      "step": 1150
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2627490993360617,
+      "learning_rate": 1.861521113906684e-05,
+      "loss": 0.8437,
+      "step": 1151
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4742185629491806,
+      "learning_rate": 1.8612431007546484e-05,
+      "loss": 1.2613,
+      "step": 1152
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.015896728826535,
+      "learning_rate": 1.8609648296133052e-05,
+      "loss": 0.7045,
+      "step": 1153
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.109399318377247,
+      "learning_rate": 1.8606863005660133e-05,
+      "loss": 1.0702,
+      "step": 1154
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0039263588328753,
+      "learning_rate": 1.860407513696206e-05,
+      "loss": 0.6892,
+      "step": 1155
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7987966572652913,
+      "learning_rate": 1.860128469087396e-05,
+      "loss": 0.6533,
+      "step": 1156
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0517139374513749,
+      "learning_rate": 1.8598491668231723e-05,
+      "loss": 1.1832,
+      "step": 1157
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0628982133640759,
+      "learning_rate": 1.8595696069872013e-05,
+      "loss": 0.8367,
+      "step": 1158
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1009779904380903,
+      "learning_rate": 1.859289789663226e-05,
+      "loss": 0.8372,
+      "step": 1159
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9177720218022811,
+      "learning_rate": 1.859009714935067e-05,
+      "loss": 0.6456,
+      "step": 1160
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2179734577529298,
+      "learning_rate": 1.8587293828866223e-05,
+      "loss": 1.1602,
+      "step": 1161
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2843488133136014,
+      "learning_rate": 1.8584487936018663e-05,
+      "loss": 0.7144,
+      "step": 1162
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0652914083648641,
+      "learning_rate": 1.858167947164851e-05,
+      "loss": 0.778,
+      "step": 1163
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2512695385489618,
+      "learning_rate": 1.857886843659705e-05,
+      "loss": 1.0103,
+      "step": 1164
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2003958168758584,
+      "learning_rate": 1.8576054831706348e-05,
+      "loss": 0.9441,
+      "step": 1165
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1721033325236403,
+      "learning_rate": 1.8573238657819225e-05,
+      "loss": 1.0943,
+      "step": 1166
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1927780960872798,
+      "learning_rate": 1.857041991577928e-05,
+      "loss": 0.9755,
+      "step": 1167
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0184259240639164,
+      "learning_rate": 1.8567598606430883e-05,
+      "loss": 0.7924,
+      "step": 1168
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4037978560271371,
+      "learning_rate": 1.856477473061917e-05,
+      "loss": 0.9189,
+      "step": 1169
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1736624054493778,
+      "learning_rate": 1.8561948289190035e-05,
+      "loss": 0.8214,
+      "step": 1170
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1996838344156362,
+      "learning_rate": 1.8559119282990172e-05,
+      "loss": 1.1161,
+      "step": 1171
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9488386404864834,
+      "learning_rate": 1.8556287712867006e-05,
+      "loss": 0.9841,
+      "step": 1172
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1579803327255422,
+      "learning_rate": 1.8553453579668755e-05,
+      "loss": 0.8288,
+      "step": 1173
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0827666206617064,
+      "learning_rate": 1.855061688424439e-05,
+      "loss": 0.8524,
+      "step": 1174
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3792170275609885,
+      "learning_rate": 1.8547777627443663e-05,
+      "loss": 0.8417,
+      "step": 1175
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0577101236771234,
+      "learning_rate": 1.8544935810117085e-05,
+      "loss": 1.1855,
+      "step": 1176
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9415938162377266,
+      "learning_rate": 1.854209143311593e-05,
+      "loss": 0.8756,
+      "step": 1177
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0185690348614527,
+      "learning_rate": 1.8539244497292248e-05,
+      "loss": 1.0507,
+      "step": 1178
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0087355199417065,
+      "learning_rate": 1.8536395003498857e-05,
+      "loss": 0.5123,
+      "step": 1179
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1168781049821879,
+      "learning_rate": 1.8533542952589322e-05,
+      "loss": 1.0765,
+      "step": 1180
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3330049553569185,
+      "learning_rate": 1.8530688345418002e-05,
+      "loss": 0.8092,
+      "step": 1181
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.862014597986358,
+      "learning_rate": 1.8527831182839997e-05,
+      "loss": 0.8531,
+      "step": 1182
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.122805081347554,
+      "learning_rate": 1.8524971465711193e-05,
+      "loss": 0.9225,
+      "step": 1183
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.265825933597711,
+      "learning_rate": 1.8522109194888225e-05,
+      "loss": 1.1219,
+      "step": 1184
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3476442026675897,
+      "learning_rate": 1.8519244371228497e-05,
+      "loss": 1.1424,
+      "step": 1185
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4168980518444692,
+      "learning_rate": 1.8516376995590185e-05,
+      "loss": 0.7907,
+      "step": 1186
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2479556448160924,
+      "learning_rate": 1.8513507068832226e-05,
+      "loss": 0.978,
+      "step": 1187
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.10144537181972,
+      "learning_rate": 1.8510634591814316e-05,
+      "loss": 0.8108,
+      "step": 1188
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9274695187434725,
+      "learning_rate": 1.8507759565396915e-05,
+      "loss": 0.6408,
+      "step": 1189
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.122104662433347,
+      "learning_rate": 1.850488199044126e-05,
+      "loss": 0.9951,
+      "step": 1190
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1182914017337404,
+      "learning_rate": 1.8502001867809333e-05,
+      "loss": 0.8832,
+      "step": 1191
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.662942536089653,
+      "learning_rate": 1.8499119198363894e-05,
+      "loss": 0.9845,
+      "step": 1192
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1092682930010687,
+      "learning_rate": 1.849623398296846e-05,
+      "loss": 0.8132,
+      "step": 1193
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1442636178774166,
+      "learning_rate": 1.84933462224873e-05,
+      "loss": 0.7386,
+      "step": 1194
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.160720032387123,
+      "learning_rate": 1.8490455917785466e-05,
+      "loss": 1.1141,
+      "step": 1195
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.730109706962241,
+      "learning_rate": 1.8487563069728756e-05,
+      "loss": 0.5694,
+      "step": 1196
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1737518009446104,
+      "learning_rate": 1.8484667679183746e-05,
+      "loss": 0.8689,
+      "step": 1197
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.171199710204817,
+      "learning_rate": 1.848176974701775e-05,
+      "loss": 0.7012,
+      "step": 1198
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6796690187016456,
+      "learning_rate": 1.847886927409887e-05,
+      "loss": 1.2631,
+      "step": 1199
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2901495844896242,
+      "learning_rate": 1.8475966261295947e-05,
+      "loss": 0.7523,
+      "step": 1200
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2103394888236971,
+      "learning_rate": 1.8473060709478593e-05,
+      "loss": 1.0027,
+      "step": 1201
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.196346046792923,
+      "learning_rate": 1.847015261951718e-05,
+      "loss": 0.9124,
+      "step": 1202
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3315374454122542,
+      "learning_rate": 1.8467241992282842e-05,
+      "loss": 0.9624,
+      "step": 1203
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2478380292988014,
+      "learning_rate": 1.846432882864747e-05,
+      "loss": 1.0635,
+      "step": 1204
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1474524763211913,
+      "learning_rate": 1.8461413129483712e-05,
+      "loss": 0.9944,
+      "step": 1205
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1801937337056587,
+      "learning_rate": 1.845849489566498e-05,
+      "loss": 0.8202,
+      "step": 1206
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.091263099486394,
+      "learning_rate": 1.845557412806545e-05,
+      "loss": 0.9441,
+      "step": 1207
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1083745105560356,
+      "learning_rate": 1.8452650827560043e-05,
+      "loss": 0.8388,
+      "step": 1208
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0406439080807672,
+      "learning_rate": 1.8449724995024453e-05,
+      "loss": 1.104,
+      "step": 1209
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0349505208431566,
+      "learning_rate": 1.8446796631335123e-05,
+      "loss": 0.7388,
+      "step": 1210
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3666150980104497,
+      "learning_rate": 1.844386573736926e-05,
+      "loss": 0.8698,
+      "step": 1211
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.287317436948201,
+      "learning_rate": 1.8440932314004826e-05,
+      "loss": 0.8203,
+      "step": 1212
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0515587226352086,
+      "learning_rate": 1.8437996362120542e-05,
+      "loss": 0.7911,
+      "step": 1213
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0427131434565038,
+      "learning_rate": 1.8435057882595885e-05,
+      "loss": 0.7548,
+      "step": 1214
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1848289732413835,
+      "learning_rate": 1.8432116876311088e-05,
+      "loss": 1.0289,
+      "step": 1215
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9819856059748545,
+      "learning_rate": 1.8429173344147148e-05,
+      "loss": 0.9933,
+      "step": 1216
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1006719221028496,
+      "learning_rate": 1.8426227286985808e-05,
+      "loss": 0.6745,
+      "step": 1217
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.028743335173196,
+      "learning_rate": 1.8423278705709573e-05,
+      "loss": 0.9327,
+      "step": 1218
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.018549239432446,
+      "learning_rate": 1.842032760120171e-05,
+      "loss": 1.0776,
+      "step": 1219
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7489056722297471,
+      "learning_rate": 1.841737397434623e-05,
+      "loss": 0.786,
+      "step": 1220
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.117321622830377,
+      "learning_rate": 1.8414417826027907e-05,
+      "loss": 1.042,
+      "step": 1221
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1683534765328647,
+      "learning_rate": 1.841145915713227e-05,
+      "loss": 0.7987,
+      "step": 1222
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9723809266865888,
+      "learning_rate": 1.8408497968545604e-05,
+      "loss": 0.9792,
+      "step": 1223
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.041713021393351,
+      "learning_rate": 1.840553426115494e-05,
+      "loss": 0.7766,
+      "step": 1224
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4284480210535793,
+      "learning_rate": 1.840256803584808e-05,
+      "loss": 1.0673,
+      "step": 1225
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3592494359465996,
+      "learning_rate": 1.8399599293513555e-05,
+      "loss": 1.078,
+      "step": 1226
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2966912516807167,
+      "learning_rate": 1.839662803504068e-05,
+      "loss": 0.8437,
+      "step": 1227
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1315376614107806,
+      "learning_rate": 1.8393654261319504e-05,
+      "loss": 1.2719,
+      "step": 1228
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1509636295105325,
+      "learning_rate": 1.839067797324083e-05,
+      "loss": 1.0414,
+      "step": 1229
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9775851883109282,
+      "learning_rate": 1.8387699171696222e-05,
+      "loss": 0.7882,
+      "step": 1230
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0634923674877197,
+      "learning_rate": 1.8384717857578e-05,
+      "loss": 0.6752,
+      "step": 1231
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0044460135864264,
+      "learning_rate": 1.838173403177922e-05,
+      "loss": 0.7157,
+      "step": 1232
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.021128433312841,
+      "learning_rate": 1.8378747695193705e-05,
+      "loss": 1.0727,
+      "step": 1233
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0503473910350198,
+      "learning_rate": 1.837575884871603e-05,
+      "loss": 0.5894,
+      "step": 1234
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2630311649473998,
+      "learning_rate": 1.837276749324151e-05,
+      "loss": 0.9877,
+      "step": 1235
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.162962551678498,
+      "learning_rate": 1.8369773629666222e-05,
+      "loss": 0.7299,
+      "step": 1236
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8181917182751836,
+      "learning_rate": 1.8366777258886993e-05,
+      "loss": 0.606,
+      "step": 1237
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9439161507356242,
+      "learning_rate": 1.8363778381801404e-05,
+      "loss": 1.0262,
+      "step": 1238
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2453264018859627,
+      "learning_rate": 1.836077699930777e-05,
+      "loss": 0.8309,
+      "step": 1239
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8322354223721016,
+      "learning_rate": 1.8357773112305183e-05,
+      "loss": 0.7665,
+      "step": 1240
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0207512918810344,
+      "learning_rate": 1.835476672169346e-05,
+      "loss": 0.9825,
+      "step": 1241
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0859809315022844,
+      "learning_rate": 1.8351757828373183e-05,
+      "loss": 1.1791,
+      "step": 1242
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.135797750184544,
+      "learning_rate": 1.8348746433245682e-05,
+      "loss": 0.7652,
+      "step": 1243
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4847507583214756,
+      "learning_rate": 1.834573253721303e-05,
+      "loss": 0.8901,
+      "step": 1244
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3489595776068157,
+      "learning_rate": 1.8342716141178055e-05,
+      "loss": 1.02,
+      "step": 1245
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.024751071196933,
+      "learning_rate": 1.8339697246044333e-05,
+      "loss": 0.7078,
+      "step": 1246
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9068551898675749,
+      "learning_rate": 1.8336675852716188e-05,
+      "loss": 0.9702,
+      "step": 1247
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0165234116644677,
+      "learning_rate": 1.833365196209869e-05,
+      "loss": 1.0975,
+      "step": 1248
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.309450225993906,
+      "learning_rate": 1.8330625575097663e-05,
+      "loss": 1.0771,
+      "step": 1249
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1712488723896253,
+      "learning_rate": 1.8327596692619674e-05,
+      "loss": 0.9358,
+      "step": 1250
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1083019715068807,
+      "learning_rate": 1.8324565315572036e-05,
+      "loss": 0.7709,
+      "step": 1251
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9579608739675545,
+      "learning_rate": 1.8321531444862814e-05,
+      "loss": 0.716,
+      "step": 1252
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1903024670763802,
+      "learning_rate": 1.831849508140082e-05,
+      "loss": 0.8159,
+      "step": 1253
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1097250373500858,
+      "learning_rate": 1.83154562260956e-05,
+      "loss": 0.9103,
+      "step": 1254
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.150960958487696,
+      "learning_rate": 1.831241487985747e-05,
+      "loss": 0.8663,
+      "step": 1255
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1162605062581399,
+      "learning_rate": 1.8309371043597472e-05,
+      "loss": 0.8441,
+      "step": 1256
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9751262548224968,
+      "learning_rate": 1.8306324718227405e-05,
+      "loss": 1.0955,
+      "step": 1257
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5709029457994992,
+      "learning_rate": 1.8303275904659807e-05,
+      "loss": 0.9655,
+      "step": 1258
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2051165560211585,
+      "learning_rate": 1.8300224603807965e-05,
+      "loss": 0.7294,
+      "step": 1259
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9786235120956335,
+      "learning_rate": 1.829717081658591e-05,
+      "loss": 0.7974,
+      "step": 1260
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.034988909543753,
+      "learning_rate": 1.8294114543908417e-05,
+      "loss": 1.1595,
+      "step": 1261
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4462266703765039,
+      "learning_rate": 1.829105578669101e-05,
+      "loss": 0.7752,
+      "step": 1262
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0888417644370638,
+      "learning_rate": 1.8287994545849948e-05,
+      "loss": 0.8732,
+      "step": 1263
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9377219715830603,
+      "learning_rate": 1.8284930822302242e-05,
+      "loss": 0.6373,
+      "step": 1264
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1340162273234844,
+      "learning_rate": 1.8281864616965646e-05,
+      "loss": 0.8668,
+      "step": 1265
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1158593884683214,
+      "learning_rate": 1.8278795930758656e-05,
+      "loss": 1.1148,
+      "step": 1266
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.006196054589877,
+      "learning_rate": 1.8275724764600508e-05,
+      "loss": 0.8396,
+      "step": 1267
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0629180837064287,
+      "learning_rate": 1.8272651119411186e-05,
+      "loss": 1.0366,
+      "step": 1268
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9494870126414312,
+      "learning_rate": 1.8269574996111414e-05,
+      "loss": 0.6872,
+      "step": 1269
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2611521037582807,
+      "learning_rate": 1.826649639562266e-05,
+      "loss": 0.9992,
+      "step": 1270
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0279894935252953,
+      "learning_rate": 1.826341531886713e-05,
+      "loss": 1.0676,
+      "step": 1271
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.232503481236035,
+      "learning_rate": 1.8260331766767777e-05,
+      "loss": 0.8491,
+      "step": 1272
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8233830129214608,
+      "learning_rate": 1.8257245740248295e-05,
+      "loss": 0.8236,
+      "step": 1273
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2671542987267455,
+      "learning_rate": 1.8254157240233114e-05,
+      "loss": 0.8368,
+      "step": 1274
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2564873769626876,
+      "learning_rate": 1.8251066267647404e-05,
+      "loss": 1.0015,
+      "step": 1275
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1322577775134655,
+      "learning_rate": 1.824797282341709e-05,
+      "loss": 1.1558,
+      "step": 1276
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2013871511937098,
+      "learning_rate": 1.8244876908468826e-05,
+      "loss": 0.9325,
+      "step": 1277
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0206173973646284,
+      "learning_rate": 1.8241778523729997e-05,
+      "loss": 0.7234,
+      "step": 1278
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.09564070831006,
+      "learning_rate": 1.8238677670128747e-05,
+      "loss": 0.9428,
+      "step": 1279
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2409238920079189,
+      "learning_rate": 1.823557434859395e-05,
+      "loss": 1.2583,
+      "step": 1280
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.144215301982065,
+      "learning_rate": 1.8232468560055218e-05,
+      "loss": 0.851,
+      "step": 1281
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9389648647560807,
+      "learning_rate": 1.8229360305442906e-05,
+      "loss": 0.9636,
+      "step": 1282
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0924967148344082,
+      "learning_rate": 1.8226249585688107e-05,
+      "loss": 0.7623,
+      "step": 1283
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0912518554228485,
+      "learning_rate": 1.8223136401722648e-05,
+      "loss": 0.78,
+      "step": 1284
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0184802346492834,
+      "learning_rate": 1.8220020754479104e-05,
+      "loss": 1.0547,
+      "step": 1285
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.315549181138606,
+      "learning_rate": 1.8216902644890768e-05,
+      "loss": 0.8307,
+      "step": 1286
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1539329170192538,
+      "learning_rate": 1.8213782073891698e-05,
+      "loss": 0.6349,
+      "step": 1287
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0994240585982484,
+      "learning_rate": 1.821065904241667e-05,
+      "loss": 0.7517,
+      "step": 1288
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.223079949474518,
+      "learning_rate": 1.8207533551401198e-05,
+      "loss": 0.8723,
+      "step": 1289
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0746729478109536,
+      "learning_rate": 1.8204405601781542e-05,
+      "loss": 1.1205,
+      "step": 1290
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0373199238023678,
+      "learning_rate": 1.8201275194494695e-05,
+      "loss": 1.1755,
+      "step": 1291
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9508024825584168,
+      "learning_rate": 1.819814233047838e-05,
+      "loss": 0.8765,
+      "step": 1292
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.944678000544019,
+      "learning_rate": 1.8195007010671067e-05,
+      "loss": 0.8351,
+      "step": 1293
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1348409840942426,
+      "learning_rate": 1.819186923601195e-05,
+      "loss": 1.2215,
+      "step": 1294
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0367579551852038,
+      "learning_rate": 1.8188729007440968e-05,
+      "loss": 0.7721,
+      "step": 1295
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2470508918578174,
+      "learning_rate": 1.8185586325898782e-05,
+      "loss": 1.03,
+      "step": 1296
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.942734519973784,
+      "learning_rate": 1.8182441192326806e-05,
+      "loss": 0.6931,
+      "step": 1297
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0951748816377305,
+      "learning_rate": 1.8179293607667177e-05,
+      "loss": 1.068,
+      "step": 1298
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1902415155599155,
+      "learning_rate": 1.8176143572862766e-05,
+      "loss": 1.1978,
+      "step": 1299
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0226182553265508,
+      "learning_rate": 1.8172991088857187e-05,
+      "loss": 0.7975,
+      "step": 1300
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.854336826901316,
+      "learning_rate": 1.8169836156594767e-05,
+      "loss": 0.5543,
+      "step": 1301
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5273764125704379,
+      "learning_rate": 1.8166678777020595e-05,
+      "loss": 0.9157,
+      "step": 1302
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0516569169410004,
+      "learning_rate": 1.816351895108047e-05,
+      "loss": 0.8565,
+      "step": 1303
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1237909435288937,
+      "learning_rate": 1.8160356679720938e-05,
+      "loss": 0.567,
+      "step": 1304
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1116507461627432,
+      "learning_rate": 1.8157191963889265e-05,
+      "loss": 1.2251,
+      "step": 1305
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2270405705762457,
+      "learning_rate": 1.8154024804533464e-05,
+      "loss": 0.8695,
+      "step": 1306
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9455401148371151,
+      "learning_rate": 1.8150855202602264e-05,
+      "loss": 0.8768,
+      "step": 1307
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0403993360482366,
+      "learning_rate": 1.8147683159045143e-05,
+      "loss": 0.8357,
+      "step": 1308
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.277582847953879,
+      "learning_rate": 1.814450867481229e-05,
+      "loss": 1.1777,
+      "step": 1309
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3186577431766515,
+      "learning_rate": 1.8141331750854647e-05,
+      "loss": 0.7492,
+      "step": 1310
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.285926578681398,
+      "learning_rate": 1.8138152388123863e-05,
+      "loss": 0.8174,
+      "step": 1311
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.052215839769487,
+      "learning_rate": 1.8134970587572345e-05,
+      "loss": 0.7883,
+      "step": 1312
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8134682057519476,
+      "learning_rate": 1.813178635015321e-05,
+      "loss": 1.1592,
+      "step": 1313
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1468635352092824,
+      "learning_rate": 1.8128599676820306e-05,
+      "loss": 1.1018,
+      "step": 1314
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1249341206087062,
+      "learning_rate": 1.8125410568528225e-05,
+      "loss": 0.7584,
+      "step": 1315
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.04907395281351,
+      "learning_rate": 1.812221902623227e-05,
+      "loss": 0.8356,
+      "step": 1316
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8320090546541636,
+      "learning_rate": 1.811902505088849e-05,
+      "loss": 0.9857,
+      "step": 1317
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0081348157313152,
+      "learning_rate": 1.8115828643453647e-05,
+      "loss": 1.0463,
+      "step": 1318
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0349578555658845,
+      "learning_rate": 1.8112629804885248e-05,
+      "loss": 1.0506,
+      "step": 1319
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9196901524242892,
+      "learning_rate": 1.8109428536141515e-05,
+      "loss": 0.7917,
+      "step": 1320
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0521383279486682,
+      "learning_rate": 1.8106224838181397e-05,
+      "loss": 0.9909,
+      "step": 1321
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0967001946520123,
+      "learning_rate": 1.8103018711964588e-05,
+      "loss": 0.8187,
+      "step": 1322
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.198010315005603,
+      "learning_rate": 1.809981015845149e-05,
+      "loss": 0.5355,
+      "step": 1323
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0053704210807706,
+      "learning_rate": 1.8096599178603242e-05,
+      "loss": 1.1338,
+      "step": 1324
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.162403270123755,
+      "learning_rate": 1.8093385773381705e-05,
+      "loss": 0.81,
+      "step": 1325
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0204108660954747,
+      "learning_rate": 1.8090169943749477e-05,
+      "loss": 0.8577,
+      "step": 1326
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2469781581030894,
+      "learning_rate": 1.8086951690669866e-05,
+      "loss": 0.8317,
+      "step": 1327
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0646249080965842,
+      "learning_rate": 1.8083731015106916e-05,
+      "loss": 1.1383,
+      "step": 1328
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4453148071554143,
+      "learning_rate": 1.80805079180254e-05,
+      "loss": 0.8536,
+      "step": 1329
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1638302934236486,
+      "learning_rate": 1.807728240039081e-05,
+      "loss": 0.8046,
+      "step": 1330
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.062997003811436,
+      "learning_rate": 1.8074054463169357e-05,
+      "loss": 0.9376,
+      "step": 1331
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.198237691785956,
+      "learning_rate": 1.8070824107327994e-05,
+      "loss": 0.8667,
+      "step": 1332
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3290557797744025,
+      "learning_rate": 1.8067591333834382e-05,
+      "loss": 1.1716,
+      "step": 1333
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0289144562321662,
+      "learning_rate": 1.806435614365692e-05,
+      "loss": 0.7914,
+      "step": 1334
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0593617894965728,
+      "learning_rate": 1.806111853776472e-05,
+      "loss": 0.8758,
+      "step": 1335
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0407466801544267,
+      "learning_rate": 1.805787851712762e-05,
+      "loss": 0.9979,
+      "step": 1336
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.149578693821349,
+      "learning_rate": 1.8054636082716183e-05,
+      "loss": 0.8572,
+      "step": 1337
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0249249702126584,
+      "learning_rate": 1.8051391235501695e-05,
+      "loss": 1.0181,
+      "step": 1338
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2763070880833818,
+      "learning_rate": 1.8048143976456168e-05,
+      "loss": 0.7735,
+      "step": 1339
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0601362757642265,
+      "learning_rate": 1.8044894306552338e-05,
+      "loss": 0.8043,
+      "step": 1340
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1015633950441615,
+      "learning_rate": 1.8041642226763647e-05,
+      "loss": 0.5779,
+      "step": 1341
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2230433915602157,
+      "learning_rate": 1.8038387738064274e-05,
+      "loss": 0.8072,
+      "step": 1342
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1629468022861391,
+      "learning_rate": 1.803513084142912e-05,
+      "loss": 1.2678,
+      "step": 1343
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1138855476506444,
+      "learning_rate": 1.8031871537833802e-05,
+      "loss": 1.0634,
+      "step": 1344
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8984977608885524,
+      "learning_rate": 1.8028609828254654e-05,
+      "loss": 0.6784,
+      "step": 1345
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0369693346343685,
+      "learning_rate": 1.8025345713668746e-05,
+      "loss": 0.868,
+      "step": 1346
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7113905817785033,
+      "learning_rate": 1.802207919505385e-05,
+      "loss": 1.0948,
+      "step": 1347
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0659644216757564,
+      "learning_rate": 1.8018810273388468e-05,
+      "loss": 0.8915,
+      "step": 1348
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1627348841277587,
+      "learning_rate": 1.8015538949651823e-05,
+      "loss": 0.8131,
+      "step": 1349
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5091334003103474,
+      "learning_rate": 1.8012265224823853e-05,
+      "loss": 0.8493,
+      "step": 1350
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1162902782636555,
+      "learning_rate": 1.8008989099885218e-05,
+      "loss": 0.8869,
+      "step": 1351
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2103162362529922,
+      "learning_rate": 1.8005710575817297e-05,
+      "loss": 1.1914,
+      "step": 1352
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0551738103547437,
+      "learning_rate": 1.8002429653602184e-05,
+      "loss": 0.8133,
+      "step": 1353
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.081794578429614,
+      "learning_rate": 1.79991463342227e-05,
+      "loss": 0.8532,
+      "step": 1354
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0741669856842964,
+      "learning_rate": 1.7995860618662367e-05,
+      "loss": 1.0272,
+      "step": 1355
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.871259612907036,
+      "learning_rate": 1.799257250790545e-05,
+      "loss": 0.7836,
+      "step": 1356
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0906070124251348,
+      "learning_rate": 1.798928200293691e-05,
+      "loss": 1.0511,
+      "step": 1357
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0830818625575442,
+      "learning_rate": 1.7985989104742434e-05,
+      "loss": 1.2087,
+      "step": 1358
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1813697388729485,
+      "learning_rate": 1.7982693814308423e-05,
+      "loss": 0.8085,
+      "step": 1359
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0630121875635359,
+      "learning_rate": 1.7979396132621997e-05,
+      "loss": 0.4356,
+      "step": 1360
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9951005427403627,
+      "learning_rate": 1.7976096060671e-05,
+      "loss": 0.9297,
+      "step": 1361
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2140703587144115,
+      "learning_rate": 1.7972793599443966e-05,
+      "loss": 0.9273,
+      "step": 1362
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1443999565002534,
+      "learning_rate": 1.796948874993018e-05,
+      "loss": 0.9305,
+      "step": 1363
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1115403921488072,
+      "learning_rate": 1.7966181513119612e-05,
+      "loss": 0.7406,
+      "step": 1364
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0750245909474379,
+      "learning_rate": 1.7962871890002967e-05,
+      "loss": 0.865,
+      "step": 1365
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.035335705615154,
+      "learning_rate": 1.795955988157166e-05,
+      "loss": 0.8828,
+      "step": 1366
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0179472607765199,
+      "learning_rate": 1.795624548881781e-05,
+      "loss": 1.1514,
+      "step": 1367
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0736777827241206,
+      "learning_rate": 1.7952928712734266e-05,
+      "loss": 1.0317,
+      "step": 1368
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.030768194004018,
+      "learning_rate": 1.7949609554314576e-05,
+      "loss": 1.0495,
+      "step": 1369
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0048969601686886,
+      "learning_rate": 1.7946288014553017e-05,
+      "loss": 0.7454,
+      "step": 1370
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0521621257315446,
+      "learning_rate": 1.794296409444457e-05,
+      "loss": 1.05,
+      "step": 1371
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1119980244584298,
+      "learning_rate": 1.7939637794984922e-05,
+      "loss": 0.8441,
+      "step": 1372
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0671108033051728,
+      "learning_rate": 1.793630911717049e-05,
+      "loss": 0.7981,
+      "step": 1373
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1586329672601916,
+      "learning_rate": 1.7932978061998393e-05,
+      "loss": 0.9532,
+      "step": 1374
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.210123673976772,
+      "learning_rate": 1.792964463046646e-05,
+      "loss": 0.7022,
+      "step": 1375
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9050043873588368,
+      "learning_rate": 1.792630882357324e-05,
+      "loss": 0.6789,
+      "step": 1376
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3476066467957213,
+      "learning_rate": 1.7922970642317988e-05,
+      "loss": 0.9785,
+      "step": 1377
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0607732656594797,
+      "learning_rate": 1.7919630087700672e-05,
+      "loss": 0.8195,
+      "step": 1378
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0425416493758801,
+      "learning_rate": 1.791628716072197e-05,
+      "loss": 0.9102,
+      "step": 1379
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3789494640720092,
+      "learning_rate": 1.791294186238327e-05,
+      "loss": 1.154,
+      "step": 1380
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1572523882331955,
+      "learning_rate": 1.790959419368667e-05,
+      "loss": 1.1007,
+      "step": 1381
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3469123737477358,
+      "learning_rate": 1.790624415563498e-05,
+      "loss": 0.8233,
+      "step": 1382
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9447150637753287,
+      "learning_rate": 1.7902891749231724e-05,
+      "loss": 0.7437,
+      "step": 1383
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.326543713984935,
+      "learning_rate": 1.7899536975481125e-05,
+      "loss": 0.8564,
+      "step": 1384
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0992196117212427,
+      "learning_rate": 1.7896179835388127e-05,
+      "loss": 0.7587,
+      "step": 1385
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2182914262483597,
+      "learning_rate": 1.7892820329958368e-05,
+      "loss": 1.1691,
+      "step": 1386
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.109730077749496,
+      "learning_rate": 1.7889458460198212e-05,
+      "loss": 0.7612,
+      "step": 1387
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.91785081888439,
+      "learning_rate": 1.7886094227114715e-05,
+      "loss": 0.8491,
+      "step": 1388
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2213750316633452,
+      "learning_rate": 1.7882727631715655e-05,
+      "loss": 0.9197,
+      "step": 1389
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9198351492929044,
+      "learning_rate": 1.7879358675009506e-05,
+      "loss": 0.9189,
+      "step": 1390
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1057079134955352,
+      "learning_rate": 1.7875987358005455e-05,
+      "loss": 0.9502,
+      "step": 1391
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.068736265319549,
+      "learning_rate": 1.78726136817134e-05,
+      "loss": 0.8359,
+      "step": 1392
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9971019606391807,
+      "learning_rate": 1.7869237647143935e-05,
+      "loss": 1.063,
+      "step": 1393
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8516832525624646,
+      "learning_rate": 1.786585925530837e-05,
+      "loss": 1.0164,
+      "step": 1394
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8383808410138736,
+      "learning_rate": 1.7862478507218718e-05,
+      "loss": 0.7503,
+      "step": 1395
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1089871130757714,
+      "learning_rate": 1.7859095403887697e-05,
+      "loss": 0.7437,
+      "step": 1396
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1565478967125715,
+      "learning_rate": 1.785570994632873e-05,
+      "loss": 0.7748,
+      "step": 1397
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1159141002842163,
+      "learning_rate": 1.7852322135555946e-05,
+      "loss": 0.888,
+      "step": 1398
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1568054436379258,
+      "learning_rate": 1.784893197258418e-05,
+      "loss": 0.8738,
+      "step": 1399
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3120064532993743,
+      "learning_rate": 1.7845539458428973e-05,
+      "loss": 1.2899,
+      "step": 1400
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.172872561678439,
+      "learning_rate": 1.7842144594106565e-05,
+      "loss": 1.0563,
+      "step": 1401
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1485446984008076,
+      "learning_rate": 1.7838747380633904e-05,
+      "loss": 0.694,
+      "step": 1402
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9474028825514861,
+      "learning_rate": 1.7835347819028642e-05,
+      "loss": 0.7295,
+      "step": 1403
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1109681660074737,
+      "learning_rate": 1.783194591030913e-05,
+      "loss": 1.0875,
+      "step": 1404
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3267714421717767,
+      "learning_rate": 1.782854165549443e-05,
+      "loss": 1.0268,
+      "step": 1405
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0955262182069048,
+      "learning_rate": 1.78251350556043e-05,
+      "loss": 0.7141,
+      "step": 1406
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.05111058102432,
+      "learning_rate": 1.78217261116592e-05,
+      "loss": 1.032,
+      "step": 1407
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8901940978425628,
+      "learning_rate": 1.78183148246803e-05,
+      "loss": 0.7479,
+      "step": 1408
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9935323337274103,
+      "learning_rate": 1.7814901195689463e-05,
+      "loss": 0.8477,
+      "step": 1409
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5420956590803656,
+      "learning_rate": 1.7811485225709255e-05,
+      "loss": 1.2981,
+      "step": 1410
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1123172769725003,
+      "learning_rate": 1.780806691576295e-05,
+      "loss": 0.8636,
+      "step": 1411
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.077930320955849,
+      "learning_rate": 1.7804646266874516e-05,
+      "loss": 0.9392,
+      "step": 1412
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9813592081209351,
+      "learning_rate": 1.7801223280068626e-05,
+      "loss": 0.7373,
+      "step": 1413
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7775379725222764,
+      "learning_rate": 1.7797797956370646e-05,
+      "loss": 0.6715,
+      "step": 1414
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1981266157995982,
+      "learning_rate": 1.7794370296806655e-05,
+      "loss": 0.871,
+      "step": 1415
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.84473393357144,
+      "learning_rate": 1.779094030240342e-05,
+      "loss": 1.0552,
+      "step": 1416
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1392260593562502,
+      "learning_rate": 1.778750797418841e-05,
+      "loss": 0.8319,
+      "step": 1417
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9827388404290185,
+      "learning_rate": 1.7784073313189796e-05,
+      "loss": 0.7073,
+      "step": 1418
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.038296204593829,
+      "learning_rate": 1.778063632043645e-05,
+      "loss": 1.2075,
+      "step": 1419
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2730843166251227,
+      "learning_rate": 1.7777196996957934e-05,
+      "loss": 0.6997,
+      "step": 1420
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9016996020351624,
+      "learning_rate": 1.7773755343784516e-05,
+      "loss": 0.8781,
+      "step": 1421
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9558221552126538,
+      "learning_rate": 1.7770311361947155e-05,
+      "loss": 1.0392,
+      "step": 1422
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1031202269611569,
+      "learning_rate": 1.7766865052477515e-05,
+      "loss": 0.7691,
+      "step": 1423
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1217145805750437,
+      "learning_rate": 1.7763416416407953e-05,
+      "loss": 1.1688,
+      "step": 1424
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.545581896018546,
+      "learning_rate": 1.7759965454771526e-05,
+      "loss": 0.8339,
+      "step": 1425
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2495622003579574,
+      "learning_rate": 1.775651216860198e-05,
+      "loss": 0.8661,
+      "step": 1426
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.207271927486595,
+      "learning_rate": 1.7753056558933766e-05,
+      "loss": 0.8555,
+      "step": 1427
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0659299250814893,
+      "learning_rate": 1.7749598626802028e-05,
+      "loss": 0.705,
+      "step": 1428
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2735306184496042,
+      "learning_rate": 1.774613837324261e-05,
+      "loss": 1.2069,
+      "step": 1429
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6091967212390537,
+      "learning_rate": 1.7742675799292033e-05,
+      "loss": 0.7447,
+      "step": 1430
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.182258968908205,
+      "learning_rate": 1.773921090598754e-05,
+      "loss": 0.8851,
+      "step": 1431
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9689395316456179,
+      "learning_rate": 1.773574369436705e-05,
+      "loss": 0.5309,
+      "step": 1432
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.105449716027601,
+      "learning_rate": 1.7732274165469182e-05,
+      "loss": 1.2239,
+      "step": 1433
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.064760192433155,
+      "learning_rate": 1.772880232033325e-05,
+      "loss": 0.6459,
+      "step": 1434
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0703095593600311,
+      "learning_rate": 1.7725328159999263e-05,
+      "loss": 0.7184,
+      "step": 1435
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1901560631739054,
+      "learning_rate": 1.772185168550792e-05,
+      "loss": 0.8796,
+      "step": 1436
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9274469861853726,
+      "learning_rate": 1.771837289790061e-05,
+      "loss": 0.6803,
+      "step": 1437
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1280954478793652,
+      "learning_rate": 1.7714891798219432e-05,
+      "loss": 1.199,
+      "step": 1438
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6321148439675766,
+      "learning_rate": 1.771140838750715e-05,
+      "loss": 0.8675,
+      "step": 1439
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.938139218259457,
+      "learning_rate": 1.770792266680725e-05,
+      "loss": 1.0914,
+      "step": 1440
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1731852106246947,
+      "learning_rate": 1.7704434637163888e-05,
+      "loss": 0.8068,
+      "step": 1441
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1023204916402591,
+      "learning_rate": 1.770094429962192e-05,
+      "loss": 1.0587,
+      "step": 1442
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9623524168081732,
+      "learning_rate": 1.769745165522689e-05,
+      "loss": 0.8915,
+      "step": 1443
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7823108941430823,
+      "learning_rate": 1.769395670502504e-05,
+      "loss": 1.0761,
+      "step": 1444
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1858759347448433,
+      "learning_rate": 1.7690459450063297e-05,
+      "loss": 0.7496,
+      "step": 1445
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3306820682468599,
+      "learning_rate": 1.768695989138928e-05,
+      "loss": 0.9868,
+      "step": 1446
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.152123086256411,
+      "learning_rate": 1.7683458030051292e-05,
+      "loss": 1.1395,
+      "step": 1447
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1676297955385164,
+      "learning_rate": 1.7679953867098346e-05,
+      "loss": 0.7885,
+      "step": 1448
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.092853504659759,
+      "learning_rate": 1.7676447403580114e-05,
+      "loss": 0.5577,
+      "step": 1449
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.422740499892565,
+      "learning_rate": 1.767293864054698e-05,
+      "loss": 0.9268,
+      "step": 1450
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2692267709926406,
+      "learning_rate": 1.7669427579050013e-05,
+      "loss": 0.9125,
+      "step": 1451
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0563616007403964,
+      "learning_rate": 1.7665914220140964e-05,
+      "loss": 1.0812,
+      "step": 1452
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3627034485137006,
+      "learning_rate": 1.7662398564872273e-05,
+      "loss": 0.6556,
+      "step": 1453
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2413815056065993,
+      "learning_rate": 1.7658880614297075e-05,
+      "loss": 0.7025,
+      "step": 1454
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3896116227658708,
+      "learning_rate": 1.765536036946919e-05,
+      "loss": 0.808,
+      "step": 1455
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6628204300272498,
+      "learning_rate": 1.7651837831443117e-05,
+      "loss": 0.9251,
+      "step": 1456
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1690143040298864,
+      "learning_rate": 1.7648313001274053e-05,
+      "loss": 1.1944,
+      "step": 1457
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9917140531208035,
+      "learning_rate": 1.7644785880017874e-05,
+      "loss": 0.7375,
+      "step": 1458
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6548956203500453,
+      "learning_rate": 1.764125646873115e-05,
+      "loss": 1.0795,
+      "step": 1459
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5162912534271742,
+      "learning_rate": 1.7637724768471127e-05,
+      "loss": 0.8145,
+      "step": 1460
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1331171803011988,
+      "learning_rate": 1.7634190780295744e-05,
+      "loss": 0.8027,
+      "step": 1461
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.170680792964419,
+      "learning_rate": 1.7630654505263628e-05,
+      "loss": 1.146,
+      "step": 1462
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0971319789612306,
+      "learning_rate": 1.7627115944434076e-05,
+      "loss": 0.7571,
+      "step": 1463
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1735906677578134,
+      "learning_rate": 1.762357509886709e-05,
+      "loss": 1.1356,
+      "step": 1464
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.260504356769815,
+      "learning_rate": 1.762003196962334e-05,
+      "loss": 1.0152,
+      "step": 1465
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4794777018493914,
+      "learning_rate": 1.7616486557764187e-05,
+      "loss": 0.891,
+      "step": 1466
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8484349812513912,
+      "learning_rate": 1.7612938864351684e-05,
+      "loss": 0.718,
+      "step": 1467
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.03823385591221,
+      "learning_rate": 1.7609388890448548e-05,
+      "loss": 0.7328,
+      "step": 1468
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2569228503964809,
+      "learning_rate": 1.760583663711819e-05,
+      "loss": 0.8959,
+      "step": 1469
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1662887151056405,
+      "learning_rate": 1.7602282105424712e-05,
+      "loss": 0.79,
+      "step": 1470
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1111100444631763,
+      "learning_rate": 1.7598725296432888e-05,
+      "loss": 1.1436,
+      "step": 1471
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0309418278091476,
+      "learning_rate": 1.759516621120817e-05,
+      "loss": 0.8426,
+      "step": 1472
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0143855753063644,
+      "learning_rate": 1.7591604850816705e-05,
+      "loss": 0.7691,
+      "step": 1473
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1733281278095145,
+      "learning_rate": 1.7588041216325314e-05,
+      "loss": 0.7606,
+      "step": 1474
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1580556000592672,
+      "learning_rate": 1.7584475308801496e-05,
+      "loss": 0.7695,
+      "step": 1475
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4875182750586682,
+      "learning_rate": 1.758090712931344e-05,
+      "loss": 1.3506,
+      "step": 1476
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1477596582103322,
+      "learning_rate": 1.7577336678930003e-05,
+      "loss": 0.8532,
+      "step": 1477
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.929212116501647,
+      "learning_rate": 1.7573763958720736e-05,
+      "loss": 0.5997,
+      "step": 1478
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.223740261263272,
+      "learning_rate": 1.7570188969755865e-05,
+      "loss": 0.7418,
+      "step": 1479
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9725234034421936,
+      "learning_rate": 1.7566611713106287e-05,
+      "loss": 0.6452,
+      "step": 1480
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3113602106753275,
+      "learning_rate": 1.7563032189843592e-05,
+      "loss": 1.0763,
+      "step": 1481
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0430759574851511,
+      "learning_rate": 1.755945040104004e-05,
+      "loss": 0.8409,
+      "step": 1482
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3641013539176445,
+      "learning_rate": 1.7555866347768576e-05,
+      "loss": 0.9714,
+      "step": 1483
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0475669105412213,
+      "learning_rate": 1.7552280031102813e-05,
+      "loss": 0.8472,
+      "step": 1484
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1640296875945004,
+      "learning_rate": 1.754869145211705e-05,
+      "loss": 0.9418,
+      "step": 1485
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0766548805727212,
+      "learning_rate": 1.754510061188627e-05,
+      "loss": 1.123,
+      "step": 1486
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8318530455231248,
+      "learning_rate": 1.7541507511486114e-05,
+      "loss": 0.4873,
+      "step": 1487
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4086760469782769,
+      "learning_rate": 1.753791215199292e-05,
+      "loss": 0.8869,
+      "step": 1488
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1606720703128892,
+      "learning_rate": 1.7534314534483693e-05,
+      "loss": 1.0068,
+      "step": 1489
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3416366426054485,
+      "learning_rate": 1.7530714660036112e-05,
+      "loss": 0.9951,
+      "step": 1490
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0490034287271945,
+      "learning_rate": 1.7527112529728537e-05,
+      "loss": 1.1724,
+      "step": 1491
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2918132163606264,
+      "learning_rate": 1.7523508144640008e-05,
+      "loss": 0.7173,
+      "step": 1492
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.739729946859766,
+      "learning_rate": 1.751990150585023e-05,
+      "loss": 1.0415,
+      "step": 1493
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0597961871351333,
+      "learning_rate": 1.7516292614439586e-05,
+      "loss": 0.862,
+      "step": 1494
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1022953167008516,
+      "learning_rate": 1.751268147148914e-05,
+      "loss": 1.0368,
+      "step": 1495
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.142618239366642,
+      "learning_rate": 1.7509068078080626e-05,
+      "loss": 0.8346,
+      "step": 1496
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1249133726207856,
+      "learning_rate": 1.750545243529645e-05,
+      "loss": 1.0056,
+      "step": 1497
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0390572804070328,
+      "learning_rate": 1.7501834544219697e-05,
+      "loss": 0.6871,
+      "step": 1498
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.292189028670805,
+      "learning_rate": 1.7498214405934126e-05,
+      "loss": 0.9135,
+      "step": 1499
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0695166143769046,
+      "learning_rate": 1.7494592021524156e-05,
+      "loss": 1.0098,
+      "step": 1500
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1815783127543076,
+      "learning_rate": 1.7490967392074897e-05,
+      "loss": 0.7931,
+      "step": 1501
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2173493820943635,
+      "learning_rate": 1.748734051867212e-05,
+      "loss": 0.531,
+      "step": 1502
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9759002665433754,
+      "learning_rate": 1.7483711402402277e-05,
+      "loss": 0.7526,
+      "step": 1503
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9974561409721036,
+      "learning_rate": 1.748008004435248e-05,
+      "loss": 0.9282,
+      "step": 1504
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0635482822962026,
+      "learning_rate": 1.747644644561052e-05,
+      "loss": 1.057,
+      "step": 1505
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2273411532054093,
+      "learning_rate": 1.7472810607264853e-05,
+      "loss": 0.9103,
+      "step": 1506
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.100390215652504,
+      "learning_rate": 1.7469172530404624e-05,
+      "loss": 0.7043,
+      "step": 1507
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0580596093968302,
+      "learning_rate": 1.7465532216119628e-05,
+      "loss": 0.8391,
+      "step": 1508
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.10549872271474,
+      "learning_rate": 1.7461889665500335e-05,
+      "loss": 1.2966,
+      "step": 1509
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9648835051091457,
+      "learning_rate": 1.7458244879637897e-05,
+      "loss": 1.0829,
+      "step": 1510
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.15465870038038,
+      "learning_rate": 1.7454597859624115e-05,
+      "loss": 0.9812,
+      "step": 1511
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0430960254887784,
+      "learning_rate": 1.7450948606551478e-05,
+      "loss": 0.989,
+      "step": 1512
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9876259950763493,
+      "learning_rate": 1.744729712151313e-05,
+      "loss": 0.7557,
+      "step": 1513
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9214316201768585,
+      "learning_rate": 1.74436434056029e-05,
+      "loss": 1.0392,
+      "step": 1514
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0213214521503584,
+      "learning_rate": 1.7439987459915265e-05,
+      "loss": 0.8056,
+      "step": 1515
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1153868778475193,
+      "learning_rate": 1.7436329285545385e-05,
+      "loss": 0.9698,
+      "step": 1516
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8861994607876665,
+      "learning_rate": 1.7432668883589086e-05,
+      "loss": 1.0192,
+      "step": 1517
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8483040058697056,
+      "learning_rate": 1.742900625514285e-05,
+      "loss": 0.6977,
+      "step": 1518
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0856104488990963,
+      "learning_rate": 1.7425341401303844e-05,
+      "loss": 1.096,
+      "step": 1519
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.178213399410515,
+      "learning_rate": 1.7421674323169885e-05,
+      "loss": 0.8005,
+      "step": 1520
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4682380639297938,
+      "learning_rate": 1.7418005021839462e-05,
+      "loss": 0.9337,
+      "step": 1521
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2811169384977399,
+      "learning_rate": 1.7414333498411734e-05,
+      "loss": 0.6217,
+      "step": 1522
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5852016054468212,
+      "learning_rate": 1.7410659753986524e-05,
+      "loss": 0.7736,
+      "step": 1523
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0907388056986707,
+      "learning_rate": 1.7406983789664315e-05,
+      "loss": 1.1084,
+      "step": 1524
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1507014084718639,
+      "learning_rate": 1.740330560654626e-05,
+      "loss": 0.8661,
+      "step": 1525
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7686252927988844,
+      "learning_rate": 1.739962520573418e-05,
+      "loss": 0.6251,
+      "step": 1526
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0867741821580297,
+      "learning_rate": 1.739594258833055e-05,
+      "loss": 0.8171,
+      "step": 1527
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.46982809165677,
+      "learning_rate": 1.7392257755438516e-05,
+      "loss": 0.9548,
+      "step": 1528
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9591764020646072,
+      "learning_rate": 1.7388570708161895e-05,
+      "loss": 0.9912,
+      "step": 1529
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.129865096825842,
+      "learning_rate": 1.7384881447605144e-05,
+      "loss": 0.9182,
+      "step": 1530
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0705760317907165,
+      "learning_rate": 1.738118997487341e-05,
+      "loss": 1.0645,
+      "step": 1531
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1225582007268364,
+      "learning_rate": 1.737749629107248e-05,
+      "loss": 0.763,
+      "step": 1532
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2608709202794488,
+      "learning_rate": 1.7373800397308827e-05,
+      "loss": 1.1583,
+      "step": 1533
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4688225732557405,
+      "learning_rate": 1.7370102294689563e-05,
+      "loss": 0.8167,
+      "step": 1534
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0372039598513896,
+      "learning_rate": 1.7366401984322473e-05,
+      "loss": 0.9802,
+      "step": 1535
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.222497656766016,
+      "learning_rate": 1.7362699467316005e-05,
+      "loss": 0.9568,
+      "step": 1536
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8804215355017379,
+      "learning_rate": 1.7358994744779263e-05,
+      "loss": 0.8845,
+      "step": 1537
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.358257417296836,
+      "learning_rate": 1.7355287817822014e-05,
+      "loss": 1.0897,
+      "step": 1538
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2963930020071222,
+      "learning_rate": 1.7351578687554683e-05,
+      "loss": 0.5189,
+      "step": 1539
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.136859351992714,
+      "learning_rate": 1.7347867355088358e-05,
+      "loss": 0.8842,
+      "step": 1540
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.939195986169097,
+      "learning_rate": 1.7344153821534786e-05,
+      "loss": 0.9981,
+      "step": 1541
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2140665122740268,
+      "learning_rate": 1.7340438088006376e-05,
+      "loss": 0.6859,
+      "step": 1542
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9848013719167101,
+      "learning_rate": 1.7336720155616186e-05,
+      "loss": 0.997,
+      "step": 1543
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4240438312023644,
+      "learning_rate": 1.733300002547795e-05,
+      "loss": 0.932,
+      "step": 1544
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1809802664829188,
+      "learning_rate": 1.7329277698706034e-05,
+      "loss": 0.7861,
+      "step": 1545
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1304166724511278,
+      "learning_rate": 1.7325553176415496e-05,
+      "loss": 0.8066,
+      "step": 1546
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0020042538512415,
+      "learning_rate": 1.732182645972202e-05,
+      "loss": 1.0315,
+      "step": 1547
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8972135116904365,
+      "learning_rate": 1.7318097549741972e-05,
+      "loss": 1.0022,
+      "step": 1548
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3693137250135972,
+      "learning_rate": 1.7314366447592354e-05,
+      "loss": 1.0022,
+      "step": 1549
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6812926717490708,
+      "learning_rate": 1.731063315439084e-05,
+      "loss": 0.6904,
+      "step": 1550
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.186011562150817,
+      "learning_rate": 1.7306897671255756e-05,
+      "loss": 0.8204,
+      "step": 1551
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9899452691138451,
+      "learning_rate": 1.7303159999306078e-05,
+      "loss": 0.9995,
+      "step": 1552
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9997991877767028,
+      "learning_rate": 1.7299420139661446e-05,
+      "loss": 1.0497,
+      "step": 1553
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.08810716679855,
+      "learning_rate": 1.7295678093442155e-05,
+      "loss": 0.7675,
+      "step": 1554
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0810891325324332,
+      "learning_rate": 1.7291933861769145e-05,
+      "loss": 0.8076,
+      "step": 1555
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9582782264946056,
+      "learning_rate": 1.7288187445764024e-05,
+      "loss": 0.6949,
+      "step": 1556
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0631674542014293,
+      "learning_rate": 1.7284438846549045e-05,
+      "loss": 1.0453,
+      "step": 1557
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8585683388505678,
+      "learning_rate": 1.728068806524712e-05,
+      "loss": 0.7258,
+      "step": 1558
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.877184805651686,
+      "learning_rate": 1.727693510298181e-05,
+      "loss": 0.861,
+      "step": 1559
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1000754609834664,
+      "learning_rate": 1.7273179960877335e-05,
+      "loss": 0.8802,
+      "step": 1560
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1453292419668448,
+      "learning_rate": 1.726942264005856e-05,
+      "loss": 1.1043,
+      "step": 1561
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.049866949491542,
+      "learning_rate": 1.7265663141651016e-05,
+      "loss": 1.1001,
+      "step": 1562
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.185247987082358,
+      "learning_rate": 1.7261901466780868e-05,
+      "loss": 0.9371,
+      "step": 1563
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2470726589435783,
+      "learning_rate": 1.725813761657495e-05,
+      "loss": 0.84,
+      "step": 1564
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0234770952904064,
+      "learning_rate": 1.7254371592160736e-05,
+      "loss": 0.7513,
+      "step": 1565
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6139660996023315,
+      "learning_rate": 1.7250603394666364e-05,
+      "loss": 0.9904,
+      "step": 1566
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2688856385835368,
+      "learning_rate": 1.7246833025220605e-05,
+      "loss": 1.0661,
+      "step": 1567
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9849719632844077,
+      "learning_rate": 1.7243060484952893e-05,
+      "loss": 0.942,
+      "step": 1568
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8987812902559401,
+      "learning_rate": 1.7239285774993315e-05,
+      "loss": 0.6193,
+      "step": 1569
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0560204929108923,
+      "learning_rate": 1.7235508896472593e-05,
+      "loss": 0.8459,
+      "step": 1570
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0472370844228391,
+      "learning_rate": 1.723172985052212e-05,
+      "loss": 0.7991,
+      "step": 1571
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3744004221628872,
+      "learning_rate": 1.7227948638273918e-05,
+      "loss": 1.0638,
+      "step": 1572
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9118096901050964,
+      "learning_rate": 1.7224165260860667e-05,
+      "loss": 0.9227,
+      "step": 1573
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.05365383605747,
+      "learning_rate": 1.72203797194157e-05,
+      "loss": 0.8438,
+      "step": 1574
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2905352139676254,
+      "learning_rate": 1.7216592015072987e-05,
+      "loss": 0.9606,
+      "step": 1575
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7314704279044274,
+      "learning_rate": 1.7212802148967157e-05,
+      "loss": 0.6835,
+      "step": 1576
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1771213351013652,
+      "learning_rate": 1.720901012223348e-05,
+      "loss": 0.7856,
+      "step": 1577
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7760104536797723,
+      "learning_rate": 1.720521593600787e-05,
+      "loss": 0.6998,
+      "step": 1578
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0041524146428897,
+      "learning_rate": 1.72014195914269e-05,
+      "loss": 0.7774,
+      "step": 1579
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0403397837535864,
+      "learning_rate": 1.7197621089627785e-05,
+      "loss": 0.7868,
+      "step": 1580
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9990006996932679,
+      "learning_rate": 1.7193820431748376e-05,
+      "loss": 1.1145,
+      "step": 1581
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4562340796231308,
+      "learning_rate": 1.7190017618927177e-05,
+      "loss": 0.9627,
+      "step": 1582
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0023611618339077,
+      "learning_rate": 1.7186212652303343e-05,
+      "loss": 0.6476,
+      "step": 1583
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0343692713751425,
+      "learning_rate": 1.7182405533016665e-05,
+      "loss": 0.8015,
+      "step": 1584
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9050894829733995,
+      "learning_rate": 1.717859626220759e-05,
+      "loss": 0.995,
+      "step": 1585
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9768403487521701,
+      "learning_rate": 1.7174784841017197e-05,
+      "loss": 1.1703,
+      "step": 1586
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1987700524027347,
+      "learning_rate": 1.7170971270587212e-05,
+      "loss": 0.8631,
+      "step": 1587
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8737036077581125,
+      "learning_rate": 1.716715555206001e-05,
+      "loss": 0.9132,
+      "step": 1588
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.09235249351304,
+      "learning_rate": 1.716333768657861e-05,
+      "loss": 0.8328,
+      "step": 1589
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9861432074347207,
+      "learning_rate": 1.7159517675286667e-05,
+      "loss": 0.739,
+      "step": 1590
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3126937623320385,
+      "learning_rate": 1.7155695519328486e-05,
+      "loss": 1.15,
+      "step": 1591
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8298014166882438,
+      "learning_rate": 1.715187121984901e-05,
+      "loss": 0.9948,
+      "step": 1592
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0493115037712375,
+      "learning_rate": 1.7148044777993826e-05,
+      "loss": 0.8064,
+      "step": 1593
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8226795551709406,
+      "learning_rate": 1.7144216194909157e-05,
+      "loss": 0.539,
+      "step": 1594
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9735433254772534,
+      "learning_rate": 1.714038547174188e-05,
+      "loss": 1.031,
+      "step": 1595
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1711133080763605,
+      "learning_rate": 1.71365526096395e-05,
+      "loss": 1.1529,
+      "step": 1596
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1693957124875298,
+      "learning_rate": 1.713271760975017e-05,
+      "loss": 0.8033,
+      "step": 1597
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0857354226843705,
+      "learning_rate": 1.7128880473222688e-05,
+      "loss": 0.7319,
+      "step": 1598
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1476871125151722,
+      "learning_rate": 1.7125041201206478e-05,
+      "loss": 1.0837,
+      "step": 1599
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0051354375279735,
+      "learning_rate": 1.712119979485161e-05,
+      "loss": 1.0426,
+      "step": 1600
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1410712478956233,
+      "learning_rate": 1.7117356255308805e-05,
+      "loss": 0.8691,
+      "step": 1601
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.342111956943065,
+      "learning_rate": 1.7113510583729403e-05,
+      "loss": 0.6867,
+      "step": 1602
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8377564777947919,
+      "learning_rate": 1.71096627812654e-05,
+      "loss": 0.7018,
+      "step": 1603
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2070755085898437,
+      "learning_rate": 1.7105812849069412e-05,
+      "loss": 0.9185,
+      "step": 1604
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.070838487643131,
+      "learning_rate": 1.7101960788294713e-05,
+      "loss": 1.1299,
+      "step": 1605
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3067284601643145,
+      "learning_rate": 1.7098106600095204e-05,
+      "loss": 0.956,
+      "step": 1606
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1995181007793343,
+      "learning_rate": 1.7094250285625422e-05,
+      "loss": 0.886,
+      "step": 1607
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.879639102037309,
+      "learning_rate": 1.7090391846040547e-05,
+      "loss": 0.9782,
+      "step": 1608
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9494818392708103,
+      "learning_rate": 1.7086531282496392e-05,
+      "loss": 0.6496,
+      "step": 1609
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9980976007367249,
+      "learning_rate": 1.7082668596149402e-05,
+      "loss": 1.1405,
+      "step": 1610
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.018905168133966,
+      "learning_rate": 1.7078803788156668e-05,
+      "loss": 0.8885,
+      "step": 1611
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1651576216960298,
+      "learning_rate": 1.7074936859675907e-05,
+      "loss": 0.5069,
+      "step": 1612
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2217406672979072,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 0.8505,
+      "step": 1613
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0478920919183667,
+      "learning_rate": 1.706719664588437e-05,
+      "loss": 0.7559,
+      "step": 1614
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0017041078292686,
+      "learning_rate": 1.7063323362892204e-05,
+      "loss": 1.0598,
+      "step": 1615
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8978617094220546,
+      "learning_rate": 1.7059447964049248e-05,
+      "loss": 0.9902,
+      "step": 1616
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1182635147284954,
+      "learning_rate": 1.7055570450516392e-05,
+      "loss": 0.81,
+      "step": 1617
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8690321633525198,
+      "learning_rate": 1.7051690823455162e-05,
+      "loss": 0.9957,
+      "step": 1618
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0942939262982474,
+      "learning_rate": 1.704780908402772e-05,
+      "loss": 1.1294,
+      "step": 1619
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2390900748728484,
+      "learning_rate": 1.7043925233396855e-05,
+      "loss": 1.0565,
+      "step": 1620
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9025542221242577,
+      "learning_rate": 1.7040039272725998e-05,
+      "loss": 0.6847,
+      "step": 1621
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0601755360649365,
+      "learning_rate": 1.70361512031792e-05,
+      "loss": 0.8707,
+      "step": 1622
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1614564304893338,
+      "learning_rate": 1.7032261025921152e-05,
+      "loss": 0.8212,
+      "step": 1623
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.147607835647964,
+      "learning_rate": 1.702836874211718e-05,
+      "loss": 1.1362,
+      "step": 1624
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1481531498696935,
+      "learning_rate": 1.7024474352933227e-05,
+      "loss": 0.9811,
+      "step": 1625
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1140824809568521,
+      "learning_rate": 1.702057785953588e-05,
+      "loss": 0.8091,
+      "step": 1626
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2528037759623851,
+      "learning_rate": 1.7016679263092352e-05,
+      "loss": 0.8139,
+      "step": 1627
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.159193023421954,
+      "learning_rate": 1.7012778564770484e-05,
+      "loss": 0.7851,
+      "step": 1628
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.092009127905788,
+      "learning_rate": 1.700887576573874e-05,
+      "loss": 1.0557,
+      "step": 1629
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7073867039885136,
+      "learning_rate": 1.7004970867166238e-05,
+      "loss": 0.6056,
+      "step": 1630
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9438475534311366,
+      "learning_rate": 1.7001063870222694e-05,
+      "loss": 0.8834,
+      "step": 1631
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2294228882146427,
+      "learning_rate": 1.6997154776078472e-05,
+      "loss": 0.8162,
+      "step": 1632
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9497190050388318,
+      "learning_rate": 1.699324358590456e-05,
+      "loss": 0.9499,
+      "step": 1633
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8995908918217592,
+      "learning_rate": 1.6989330300872576e-05,
+      "loss": 0.8773,
+      "step": 1634
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2168304295148986,
+      "learning_rate": 1.698541492215475e-05,
+      "loss": 0.892,
+      "step": 1635
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1328959722541174,
+      "learning_rate": 1.6981497450923968e-05,
+      "loss": 0.7563,
+      "step": 1636
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.990811117707046,
+      "learning_rate": 1.6977577888353713e-05,
+      "loss": 0.9572,
+      "step": 1637
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.020029029167837,
+      "learning_rate": 1.6973656235618114e-05,
+      "loss": 0.7447,
+      "step": 1638
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.168811137682517,
+      "learning_rate": 1.696973249389192e-05,
+      "loss": 1.1796,
+      "step": 1639
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0640343625391768,
+      "learning_rate": 1.6965806664350505e-05,
+      "loss": 1.2009,
+      "step": 1640
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0595343006895968,
+      "learning_rate": 1.6961878748169868e-05,
+      "loss": 0.9755,
+      "step": 1641
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1042940408741373,
+      "learning_rate": 1.6957948746526637e-05,
+      "loss": 0.8134,
+      "step": 1642
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0328937555149866,
+      "learning_rate": 1.695401666059806e-05,
+      "loss": 1.0793,
+      "step": 1643
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.188161853183258,
+      "learning_rate": 1.695008249156201e-05,
+      "loss": 0.8465,
+      "step": 1644
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0136928613482141,
+      "learning_rate": 1.6946146240596987e-05,
+      "loss": 0.7078,
+      "step": 1645
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.058978225179385,
+      "learning_rate": 1.6942207908882115e-05,
+      "loss": 0.8545,
+      "step": 1646
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0989169292483247,
+      "learning_rate": 1.6938267497597135e-05,
+      "loss": 0.8439,
+      "step": 1647
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6529057334032662,
+      "learning_rate": 1.6934325007922418e-05,
+      "loss": 0.7176,
+      "step": 1648
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1221190356598325,
+      "learning_rate": 1.6930380441038954e-05,
+      "loss": 1.0248,
+      "step": 1649
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2487292838086284,
+      "learning_rate": 1.692643379812836e-05,
+      "loss": 0.7267,
+      "step": 1650
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2456146379264594,
+      "learning_rate": 1.692248508037286e-05,
+      "loss": 0.9194,
+      "step": 1651
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0345302363225073,
+      "learning_rate": 1.6918534288955322e-05,
+      "loss": 0.9558,
+      "step": 1652
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0998183289793846,
+      "learning_rate": 1.691458142505922e-05,
+      "loss": 1.1083,
+      "step": 1653
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1979753141092728,
+      "learning_rate": 1.691062648986865e-05,
+      "loss": 0.6194,
+      "step": 1654
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.032030301806389,
+      "learning_rate": 1.690666948456833e-05,
+      "loss": 0.7273,
+      "step": 1655
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2314861718736159,
+      "learning_rate": 1.6902710410343604e-05,
+      "loss": 1.0714,
+      "step": 1656
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2706987387448758,
+      "learning_rate": 1.6898749268380426e-05,
+      "loss": 1.2586,
+      "step": 1657
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1871039993813055,
+      "learning_rate": 1.6894786059865383e-05,
+      "loss": 0.8657,
+      "step": 1658
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0511527565792564,
+      "learning_rate": 1.689082078598566e-05,
+      "loss": 0.8406,
+      "step": 1659
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2266296649639705,
+      "learning_rate": 1.6886853447929082e-05,
+      "loss": 1.1584,
+      "step": 1660
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9642576049122717,
+      "learning_rate": 1.6882884046884077e-05,
+      "loss": 0.7336,
+      "step": 1661
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1114642476538898,
+      "learning_rate": 1.6878912584039698e-05,
+      "loss": 0.7816,
+      "step": 1662
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0123642111050173,
+      "learning_rate": 1.687493906058562e-05,
+      "loss": 1.2295,
+      "step": 1663
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0731450752583924,
+      "learning_rate": 1.687096347771213e-05,
+      "loss": 0.8008,
+      "step": 1664
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4241877135241607,
+      "learning_rate": 1.6866985836610127e-05,
+      "loss": 1.037,
+      "step": 1665
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9943390647372964,
+      "learning_rate": 1.686300613847113e-05,
+      "loss": 0.8318,
+      "step": 1666
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1149772718293443,
+      "learning_rate": 1.685902438448728e-05,
+      "loss": 0.7435,
+      "step": 1667
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1518094369246288,
+      "learning_rate": 1.6855040575851337e-05,
+      "loss": 0.6683,
+      "step": 1668
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2069633829793074,
+      "learning_rate": 1.6851054713756653e-05,
+      "loss": 0.7647,
+      "step": 1669
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2351829883375867,
+      "learning_rate": 1.6847066799397224e-05,
+      "loss": 0.7577,
+      "step": 1670
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0419001974008164,
+      "learning_rate": 1.684307683396764e-05,
+      "loss": 0.814,
+      "step": 1671
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0146349892719755,
+      "learning_rate": 1.6839084818663116e-05,
+      "loss": 1.1866,
+      "step": 1672
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.201109121546668,
+      "learning_rate": 1.6835090754679478e-05,
+      "loss": 0.8103,
+      "step": 1673
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9756361548061062,
+      "learning_rate": 1.683109464321317e-05,
+      "loss": 0.7004,
+      "step": 1674
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4229464109200136,
+      "learning_rate": 1.6827096485461246e-05,
+      "loss": 0.9425,
+      "step": 1675
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9782227649208725,
+      "learning_rate": 1.6823096282621366e-05,
+      "loss": 0.8642,
+      "step": 1676
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1100179681476188,
+      "learning_rate": 1.681909403589181e-05,
+      "loss": 1.0985,
+      "step": 1677
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.031880088314198,
+      "learning_rate": 1.6815089746471472e-05,
+      "loss": 0.813,
+      "step": 1678
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5452567188938886,
+      "learning_rate": 1.6811083415559855e-05,
+      "loss": 1.0234,
+      "step": 1679
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1077523395613138,
+      "learning_rate": 1.6807075044357074e-05,
+      "loss": 0.7749,
+      "step": 1680
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.225027595682368,
+      "learning_rate": 1.6803064634063856e-05,
+      "loss": 0.7667,
+      "step": 1681
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1505567487537605,
+      "learning_rate": 1.6799052185881535e-05,
+      "loss": 1.0742,
+      "step": 1682
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2508540411858853,
+      "learning_rate": 1.679503770101206e-05,
+      "loss": 0.8018,
+      "step": 1683
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0309354611622554,
+      "learning_rate": 1.6791021180657985e-05,
+      "loss": 0.5854,
+      "step": 1684
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1695581008764262,
+      "learning_rate": 1.678700262602248e-05,
+      "loss": 1.0789,
+      "step": 1685
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2661199964349776,
+      "learning_rate": 1.678298203830932e-05,
+      "loss": 1.163,
+      "step": 1686
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2450313015157135,
+      "learning_rate": 1.6778959418722895e-05,
+      "loss": 0.8705,
+      "step": 1687
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0056876472570395,
+      "learning_rate": 1.6774934768468192e-05,
+      "loss": 0.9235,
+      "step": 1688
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.584839426406945,
+      "learning_rate": 1.677090808875082e-05,
+      "loss": 1.0125,
+      "step": 1689
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1299437938014858,
+      "learning_rate": 1.6766879380776983e-05,
+      "loss": 0.6832,
+      "step": 1690
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5985672734941065,
+      "learning_rate": 1.6762848645753504e-05,
+      "loss": 1.1418,
+      "step": 1691
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9668963290231167,
+      "learning_rate": 1.6758815884887806e-05,
+      "loss": 0.8283,
+      "step": 1692
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9588438400520904,
+      "learning_rate": 1.6754781099387918e-05,
+      "loss": 0.5746,
+      "step": 1693
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.044678812759254,
+      "learning_rate": 1.6750744290462484e-05,
+      "loss": 0.8851,
+      "step": 1694
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.152990378312077,
+      "learning_rate": 1.6746705459320746e-05,
+      "loss": 0.7258,
+      "step": 1695
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9626606501913825,
+      "learning_rate": 1.6742664607172556e-05,
+      "loss": 1.0886,
+      "step": 1696
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0732456713489005,
+      "learning_rate": 1.6738621735228363e-05,
+      "loss": 0.8009,
+      "step": 1697
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6031834622823837,
+      "learning_rate": 1.6734576844699234e-05,
+      "loss": 0.9191,
+      "step": 1698
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1118577431223253,
+      "learning_rate": 1.6730529936796838e-05,
+      "loss": 0.8727,
+      "step": 1699
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1542037194729815,
+      "learning_rate": 1.6726481012733437e-05,
+      "loss": 0.7476,
+      "step": 1700
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0747626556464107,
+      "learning_rate": 1.672243007372191e-05,
+      "loss": 1.2308,
+      "step": 1701
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8480576659242777,
+      "learning_rate": 1.671837712097573e-05,
+      "loss": 0.5806,
+      "step": 1702
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2454642198205832,
+      "learning_rate": 1.671432215570898e-05,
+      "loss": 1.0192,
+      "step": 1703
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0488797605873261,
+      "learning_rate": 1.671026517913634e-05,
+      "loss": 0.7561,
+      "step": 1704
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0921223918443002,
+      "learning_rate": 1.67062061924731e-05,
+      "loss": 1.282,
+      "step": 1705
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2326634822837232,
+      "learning_rate": 1.670214519693515e-05,
+      "loss": 0.6158,
+      "step": 1706
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9401903520938137,
+      "learning_rate": 1.6698082193738976e-05,
+      "loss": 0.7524,
+      "step": 1707
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4294782622197597,
+      "learning_rate": 1.669401718410167e-05,
+      "loss": 1.0225,
+      "step": 1708
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9466089423328011,
+      "learning_rate": 1.6689950169240924e-05,
+      "loss": 0.7101,
+      "step": 1709
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.55357010496278,
+      "learning_rate": 1.6685881150375027e-05,
+      "loss": 1.1923,
+      "step": 1710
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.119906016736241,
+      "learning_rate": 1.668181012872288e-05,
+      "loss": 0.8493,
+      "step": 1711
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.102400316046703,
+      "learning_rate": 1.6677737105503967e-05,
+      "loss": 1.0709,
+      "step": 1712
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0780818453563326,
+      "learning_rate": 1.667366208193839e-05,
+      "loss": 0.7777,
+      "step": 1713
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0464681213745246,
+      "learning_rate": 1.6669585059246837e-05,
+      "loss": 1.0289,
+      "step": 1714
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1146539519724643,
+      "learning_rate": 1.66655060386506e-05,
+      "loss": 0.972,
+      "step": 1715
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.090035634899585,
+      "learning_rate": 1.6661425021371567e-05,
+      "loss": 0.894,
+      "step": 1716
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4303393850273665,
+      "learning_rate": 1.6657342008632223e-05,
+      "loss": 0.8975,
+      "step": 1717
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1320925378459379,
+      "learning_rate": 1.6653257001655652e-05,
+      "loss": 0.8434,
+      "step": 1718
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7993719175421065,
+      "learning_rate": 1.664917000166555e-05,
+      "loss": 0.5117,
+      "step": 1719
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4224720673339437,
+      "learning_rate": 1.6645081009886178e-05,
+      "loss": 1.1319,
+      "step": 1720
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4459140431364381,
+      "learning_rate": 1.6640990027542423e-05,
+      "loss": 0.8202,
+      "step": 1721
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3084869236024124,
+      "learning_rate": 1.6636897055859754e-05,
+      "loss": 0.8864,
+      "step": 1722
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3672189357409334,
+      "learning_rate": 1.6632802096064243e-05,
+      "loss": 0.732,
+      "step": 1723
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1822825221223565,
+      "learning_rate": 1.662870514938255e-05,
+      "loss": 1.1513,
+      "step": 1724
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0706735502352733,
+      "learning_rate": 1.6624606217041933e-05,
+      "loss": 0.9371,
+      "step": 1725
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1145489240340571,
+      "learning_rate": 1.6620505300270254e-05,
+      "loss": 0.7156,
+      "step": 1726
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0997443887671803,
+      "learning_rate": 1.661640240029595e-05,
+      "loss": 0.9992,
+      "step": 1727
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2194881821145904,
+      "learning_rate": 1.6612297518348072e-05,
+      "loss": 0.8116,
+      "step": 1728
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.392863972507828,
+      "learning_rate": 1.6608190655656255e-05,
+      "loss": 1.2907,
+      "step": 1729
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4118473121845556,
+      "learning_rate": 1.6604081813450724e-05,
+      "loss": 0.578,
+      "step": 1730
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7770690108493417,
+      "learning_rate": 1.6599970992962306e-05,
+      "loss": 0.806,
+      "step": 1731
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3016707568409362,
+      "learning_rate": 1.6595858195422414e-05,
+      "loss": 0.8054,
+      "step": 1732
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1639501537128356,
+      "learning_rate": 1.6591743422063053e-05,
+      "loss": 0.7996,
+      "step": 1733
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3100746171912203,
+      "learning_rate": 1.6587626674116827e-05,
+      "loss": 1.104,
+      "step": 1734
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.173755185552237,
+      "learning_rate": 1.6583507952816927e-05,
+      "loss": 0.816,
+      "step": 1735
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1851569100978685,
+      "learning_rate": 1.657938725939713e-05,
+      "loss": 1.0107,
+      "step": 1736
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.069355801185541,
+      "learning_rate": 1.6575264595091808e-05,
+      "loss": 0.6749,
+      "step": 1737
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7725463134806194,
+      "learning_rate": 1.657113996113593e-05,
+      "loss": 0.8576,
+      "step": 1738
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1568142082049655,
+      "learning_rate": 1.6567013358765045e-05,
+      "loss": 1.2462,
+      "step": 1739
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8779499946987065,
+      "learning_rate": 1.6562884789215298e-05,
+      "loss": 0.6486,
+      "step": 1740
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0507912943872035,
+      "learning_rate": 1.6558754253723416e-05,
+      "loss": 0.8427,
+      "step": 1741
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1346969812699192,
+      "learning_rate": 1.6554621753526724e-05,
+      "loss": 0.8663,
+      "step": 1742
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.306888420168111,
+      "learning_rate": 1.655048728986313e-05,
+      "loss": 0.8495,
+      "step": 1743
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.158747672120403,
+      "learning_rate": 1.6546350863971134e-05,
+      "loss": 1.1202,
+      "step": 1744
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9445609591381511,
+      "learning_rate": 1.6542212477089815e-05,
+      "loss": 0.5718,
+      "step": 1745
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9081882135279621,
+      "learning_rate": 1.6538072130458853e-05,
+      "loss": 0.7298,
+      "step": 1746
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0240188901972491,
+      "learning_rate": 1.6533929825318506e-05,
+      "loss": 0.8971,
+      "step": 1747
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0929005909341007,
+      "learning_rate": 1.6529785562909614e-05,
+      "loss": 1.2161,
+      "step": 1748
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.983353084948233,
+      "learning_rate": 1.652563934447362e-05,
+      "loss": 0.7307,
+      "step": 1749
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9047839492865377,
+      "learning_rate": 1.652149117125253e-05,
+      "loss": 0.978,
+      "step": 1750
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2858024431942865,
+      "learning_rate": 1.6517341044488963e-05,
+      "loss": 1.0102,
+      "step": 1751
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5529787444681191,
+      "learning_rate": 1.65131889654261e-05,
+      "loss": 0.9904,
+      "step": 1752
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0846778018985215,
+      "learning_rate": 1.6509034935307716e-05,
+      "loss": 1.1455,
+      "step": 1753
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2031718483433358,
+      "learning_rate": 1.6504878955378166e-05,
+      "loss": 0.7666,
+      "step": 1754
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6518096338272116,
+      "learning_rate": 1.65007210268824e-05,
+      "loss": 0.5434,
+      "step": 1755
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3806505929093307,
+      "learning_rate": 1.6496561151065938e-05,
+      "loss": 0.9053,
+      "step": 1756
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1918703810838585,
+      "learning_rate": 1.6492399329174898e-05,
+      "loss": 0.8971,
+      "step": 1757
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1434342264755062,
+      "learning_rate": 1.6488235562455964e-05,
+      "loss": 1.1747,
+      "step": 1758
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4052858729315083,
+      "learning_rate": 1.6484069852156418e-05,
+      "loss": 0.7445,
+      "step": 1759
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.308429452723729,
+      "learning_rate": 1.6479902199524116e-05,
+      "loss": 0.8433,
+      "step": 1760
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1263216001349026,
+      "learning_rate": 1.6475732605807494e-05,
+      "loss": 0.7281,
+      "step": 1761
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1895612960454434,
+      "learning_rate": 1.6471561072255576e-05,
+      "loss": 1.0099,
+      "step": 1762
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0099526667081171,
+      "learning_rate": 1.646738760011796e-05,
+      "loss": 0.9187,
+      "step": 1763
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2127706458280767,
+      "learning_rate": 1.6463212190644834e-05,
+      "loss": 1.0111,
+      "step": 1764
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2057335274777745,
+      "learning_rate": 1.6459034845086956e-05,
+      "loss": 0.8119,
+      "step": 1765
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0897463877836742,
+      "learning_rate": 1.6454855564695674e-05,
+      "loss": 0.811,
+      "step": 1766
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3311764885825224,
+      "learning_rate": 1.645067435072291e-05,
+      "loss": 1.2114,
+      "step": 1767
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1972339647519774,
+      "learning_rate": 1.644649120442116e-05,
+      "loss": 0.7702,
+      "step": 1768
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1041903930977326,
+      "learning_rate": 1.6442306127043514e-05,
+      "loss": 0.8995,
+      "step": 1769
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8868026447373173,
+      "learning_rate": 1.643811911984362e-05,
+      "loss": 0.9991,
+      "step": 1770
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.184292624386161,
+      "learning_rate": 1.643393018407572e-05,
+      "loss": 0.9637,
+      "step": 1771
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4181475626057856,
+      "learning_rate": 1.642973932099463e-05,
+      "loss": 1.1285,
+      "step": 1772
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5961483263430198,
+      "learning_rate": 1.642554653185574e-05,
+      "loss": 0.7962,
+      "step": 1773
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8100609893241376,
+      "learning_rate": 1.6421351817915025e-05,
+      "loss": 0.5075,
+      "step": 1774
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2154881133344606,
+      "learning_rate": 1.641715518042903e-05,
+      "loss": 0.7544,
+      "step": 1775
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6585213752656125,
+      "learning_rate": 1.641295662065487e-05,
+      "loss": 0.9875,
+      "step": 1776
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.262077560365615,
+      "learning_rate": 1.6408756139850243e-05,
+      "loss": 1.208,
+      "step": 1777
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.210563836427158,
+      "learning_rate": 1.6404553739273426e-05,
+      "loss": 0.7821,
+      "step": 1778
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2716988283660637,
+      "learning_rate": 1.640034942018327e-05,
+      "loss": 0.971,
+      "step": 1779
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2741736079745234,
+      "learning_rate": 1.6396143183839192e-05,
+      "loss": 0.7458,
+      "step": 1780
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4065670402323103,
+      "learning_rate": 1.6391935031501193e-05,
+      "loss": 1.2255,
+      "step": 1781
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1690726080733258,
+      "learning_rate": 1.6387724964429845e-05,
+      "loss": 0.6509,
+      "step": 1782
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7455852113652734,
+      "learning_rate": 1.638351298388629e-05,
+      "loss": 0.7524,
+      "step": 1783
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.112585928041826,
+      "learning_rate": 1.637929909113225e-05,
+      "loss": 0.8985,
+      "step": 1784
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.003689487829261,
+      "learning_rate": 1.637508328743001e-05,
+      "loss": 0.8188,
+      "step": 1785
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5226212175499507,
+      "learning_rate": 1.6370865574042435e-05,
+      "loss": 1.2619,
+      "step": 1786
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3555400972388816,
+      "learning_rate": 1.6366645952232962e-05,
+      "loss": 1.023,
+      "step": 1787
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.103148268479101,
+      "learning_rate": 1.63624244232656e-05,
+      "loss": 0.6982,
+      "step": 1788
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3346849706083803,
+      "learning_rate": 1.6358200988404923e-05,
+      "loss": 0.994,
+      "step": 1789
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1497103741383186,
+      "learning_rate": 1.6353975648916086e-05,
+      "loss": 0.8183,
+      "step": 1790
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2175871495046398,
+      "learning_rate": 1.6349748406064798e-05,
+      "loss": 1.0516,
+      "step": 1791
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9769554588587874,
+      "learning_rate": 1.634551926111736e-05,
+      "loss": 0.8128,
+      "step": 1792
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0025833603220948,
+      "learning_rate": 1.6341288215340627e-05,
+      "loss": 0.7501,
+      "step": 1793
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9861535578286675,
+      "learning_rate": 1.633705527000203e-05,
+      "loss": 0.6758,
+      "step": 1794
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2758529649072763,
+      "learning_rate": 1.6332820426369567e-05,
+      "loss": 0.8488,
+      "step": 1795
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0519835413499217,
+      "learning_rate": 1.6328583685711797e-05,
+      "loss": 1.1314,
+      "step": 1796
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.16925145914474,
+      "learning_rate": 1.632434504929787e-05,
+      "loss": 0.8723,
+      "step": 1797
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.441653647224141,
+      "learning_rate": 1.6320104518397473e-05,
+      "loss": 0.8988,
+      "step": 1798
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6903072050642438,
+      "learning_rate": 1.6315862094280886e-05,
+      "loss": 0.9906,
+      "step": 1799
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8888395749301076,
+      "learning_rate": 1.6311617778218945e-05,
+      "loss": 0.913,
+      "step": 1800
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.199160255799997,
+      "learning_rate": 1.6307371571483056e-05,
+      "loss": 1.1069,
+      "step": 1801
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.47446252494137,
+      "learning_rate": 1.6303123475345182e-05,
+      "loss": 0.8277,
+      "step": 1802
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2367173885894958,
+      "learning_rate": 1.6298873491077868e-05,
+      "loss": 0.9885,
+      "step": 1803
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2910929606042258,
+      "learning_rate": 1.6294621619954214e-05,
+      "loss": 0.9389,
+      "step": 1804
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.402172503588707,
+      "learning_rate": 1.6290367863247884e-05,
+      "loss": 1.2016,
+      "step": 1805
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1236393610406066,
+      "learning_rate": 1.6286112222233113e-05,
+      "loss": 0.8626,
+      "step": 1806
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2051980862693346,
+      "learning_rate": 1.62818546981847e-05,
+      "loss": 0.878,
+      "step": 1807
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2523206030417624,
+      "learning_rate": 1.6277595292378e-05,
+      "loss": 0.6714,
+      "step": 1808
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3383498130785962,
+      "learning_rate": 1.6273334006088946e-05,
+      "loss": 0.7865,
+      "step": 1809
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0884686005853366,
+      "learning_rate": 1.6269070840594017e-05,
+      "loss": 0.8043,
+      "step": 1810
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9594682634250145,
+      "learning_rate": 1.626480579717027e-05,
+      "loss": 1.1072,
+      "step": 1811
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8234690819478447,
+      "learning_rate": 1.626053887709531e-05,
+      "loss": 0.8675,
+      "step": 1812
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1261723459109139,
+      "learning_rate": 1.6256270081647322e-05,
+      "loss": 1.0488,
+      "step": 1813
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.402130232381238,
+      "learning_rate": 1.6251999412105036e-05,
+      "loss": 0.7468,
+      "step": 1814
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.048332981428527,
+      "learning_rate": 1.6247726869747757e-05,
+      "loss": 1.1653,
+      "step": 1815
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0911415679732583,
+      "learning_rate": 1.624345245585534e-05,
+      "loss": 0.8311,
+      "step": 1816
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.242317511879527,
+      "learning_rate": 1.6239176171708208e-05,
+      "loss": 0.8147,
+      "step": 1817
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.997851037733328,
+      "learning_rate": 1.6234898018587336e-05,
+      "loss": 0.9162,
+      "step": 1818
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9972201749908964,
+      "learning_rate": 1.623061799777427e-05,
+      "loss": 0.6029,
+      "step": 1819
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8635035008244222,
+      "learning_rate": 1.622633611055111e-05,
+      "loss": 0.9474,
+      "step": 1820
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4960534346605443,
+      "learning_rate": 1.622205235820051e-05,
+      "loss": 0.774,
+      "step": 1821
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1981890490156257,
+      "learning_rate": 1.6217766742005693e-05,
+      "loss": 0.9513,
+      "step": 1822
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9888677782385873,
+      "learning_rate": 1.6213479263250433e-05,
+      "loss": 0.8136,
+      "step": 1823
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.063345102252703,
+      "learning_rate": 1.6209189923219062e-05,
+      "loss": 0.721,
+      "step": 1824
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9548907943804288,
+      "learning_rate": 1.6204898723196475e-05,
+      "loss": 1.005,
+      "step": 1825
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.154962257220845,
+      "learning_rate": 1.6200605664468113e-05,
+      "loss": 0.8768,
+      "step": 1826
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2744228270160993,
+      "learning_rate": 1.619631074831999e-05,
+      "loss": 0.9912,
+      "step": 1827
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2913177983576907,
+      "learning_rate": 1.6192013976038663e-05,
+      "loss": 0.8764,
+      "step": 1828
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6647344888540512,
+      "learning_rate": 1.6187715348911254e-05,
+      "loss": 0.5467,
+      "step": 1829
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4991630680531014,
+      "learning_rate": 1.6183414868225434e-05,
+      "loss": 1.1088,
+      "step": 1830
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3883950725305805,
+      "learning_rate": 1.6179112535269425e-05,
+      "loss": 0.9976,
+      "step": 1831
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0639313644412776,
+      "learning_rate": 1.6174808351332023e-05,
+      "loss": 0.8582,
+      "step": 1832
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1657847956389131,
+      "learning_rate": 1.6170502317702556e-05,
+      "loss": 0.7076,
+      "step": 1833
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9955764342746485,
+      "learning_rate": 1.616619443567092e-05,
+      "loss": 1.097,
+      "step": 1834
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0271871168639242,
+      "learning_rate": 1.616188470652756e-05,
+      "loss": 1.0276,
+      "step": 1835
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1314426212659967,
+      "learning_rate": 1.615757313156348e-05,
+      "loss": 0.8682,
+      "step": 1836
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0128026123258909,
+      "learning_rate": 1.6153259712070225e-05,
+      "loss": 0.8314,
+      "step": 1837
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8114309973325431,
+      "learning_rate": 1.6148944449339904e-05,
+      "loss": 1.0339,
+      "step": 1838
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.931946544552056,
+      "learning_rate": 1.6144627344665165e-05,
+      "loss": 1.0254,
+      "step": 1839
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9553823823216389,
+      "learning_rate": 1.614030839933923e-05,
+      "loss": 0.8816,
+      "step": 1840
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1435981576549457,
+      "learning_rate": 1.613598761465585e-05,
+      "loss": 0.7809,
+      "step": 1841
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0459440710976147,
+      "learning_rate": 1.6131664991909335e-05,
+      "loss": 0.9311,
+      "step": 1842
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0017479527236002,
+      "learning_rate": 1.6127340532394556e-05,
+      "loss": 0.7888,
+      "step": 1843
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0835510431623672,
+      "learning_rate": 1.6123014237406912e-05,
+      "loss": 1.0341,
+      "step": 1844
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2388684348232262,
+      "learning_rate": 1.611868610824237e-05,
+      "loss": 0.8762,
+      "step": 1845
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.919188639746442,
+      "learning_rate": 1.6114356146197445e-05,
+      "loss": 0.7285,
+      "step": 1846
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2006232242025352,
+      "learning_rate": 1.6110024352569193e-05,
+      "loss": 0.8761,
+      "step": 1847
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0391385348637108,
+      "learning_rate": 1.6105690728655217e-05,
+      "loss": 1.0143,
+      "step": 1848
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.034339630636143,
+      "learning_rate": 1.6101355275753685e-05,
+      "loss": 0.9114,
+      "step": 1849
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.278417601037854,
+      "learning_rate": 1.6097017995163294e-05,
+      "loss": 0.9896,
+      "step": 1850
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9830443922907639,
+      "learning_rate": 1.6092678888183298e-05,
+      "loss": 0.6806,
+      "step": 1851
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.134462698462646,
+      "learning_rate": 1.60883379561135e-05,
+      "loss": 0.9781,
+      "step": 1852
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8657027543570672,
+      "learning_rate": 1.6083995200254235e-05,
+      "loss": 1.0325,
+      "step": 1853
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.067513940615588,
+      "learning_rate": 1.607965062190641e-05,
+      "loss": 1.0107,
+      "step": 1854
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9026274107475069,
+      "learning_rate": 1.6075304222371454e-05,
+      "loss": 0.7024,
+      "step": 1855
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9989694415266809,
+      "learning_rate": 1.6070956002951356e-05,
+      "loss": 0.8971,
+      "step": 1856
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1776267866195411,
+      "learning_rate": 1.606660596494864e-05,
+      "loss": 0.9115,
+      "step": 1857
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8661106201700898,
+      "learning_rate": 1.6062254109666383e-05,
+      "loss": 1.1724,
+      "step": 1858
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0874322411534703,
+      "learning_rate": 1.60579004384082e-05,
+      "loss": 0.8754,
+      "step": 1859
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0267104210624456,
+      "learning_rate": 1.6053544952478258e-05,
+      "loss": 0.8366,
+      "step": 1860
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.807863516369626,
+      "learning_rate": 1.6049187653181258e-05,
+      "loss": 1.03,
+      "step": 1861
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.058996633632099,
+      "learning_rate": 1.604482854182245e-05,
+      "loss": 0.6773,
+      "step": 1862
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9528593846974462,
+      "learning_rate": 1.604046761970763e-05,
+      "loss": 1.1391,
+      "step": 1863
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.98172182826019,
+      "learning_rate": 1.6036104888143127e-05,
+      "loss": 0.7723,
+      "step": 1864
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7326889237118205,
+      "learning_rate": 1.603174034843582e-05,
+      "loss": 0.4943,
+      "step": 1865
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1491694035348923,
+      "learning_rate": 1.6027374001893123e-05,
+      "loss": 0.8334,
+      "step": 1866
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0119529335393271,
+      "learning_rate": 1.6023005849823002e-05,
+      "loss": 0.922,
+      "step": 1867
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9255070846664505,
+      "learning_rate": 1.601863589353395e-05,
+      "loss": 0.8534,
+      "step": 1868
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8876368901852798,
+      "learning_rate": 1.601426413433501e-05,
+      "loss": 0.7772,
+      "step": 1869
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0438041777052165,
+      "learning_rate": 1.6009890573535763e-05,
+      "loss": 0.7464,
+      "step": 1870
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0193882904225713,
+      "learning_rate": 1.6005515212446328e-05,
+      "loss": 1.0468,
+      "step": 1871
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1829229647122592,
+      "learning_rate": 1.600113805237737e-05,
+      "loss": 1.1587,
+      "step": 1872
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8454879905727578,
+      "learning_rate": 1.5996759094640076e-05,
+      "loss": 0.862,
+      "step": 1873
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.838168942151794,
+      "learning_rate": 1.5992378340546194e-05,
+      "loss": 0.7223,
+      "step": 1874
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9470023071307632,
+      "learning_rate": 1.5987995791407992e-05,
+      "loss": 0.8758,
+      "step": 1875
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2105008947748608,
+      "learning_rate": 1.598361144853828e-05,
+      "loss": 0.9237,
+      "step": 1876
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.338791291774962,
+      "learning_rate": 1.597922531325042e-05,
+      "loss": 1.1488,
+      "step": 1877
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0611182210368664,
+      "learning_rate": 1.597483738685829e-05,
+      "loss": 0.7849,
+      "step": 1878
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1025775844119967,
+      "learning_rate": 1.5970447670676314e-05,
+      "loss": 1.1403,
+      "step": 1879
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2218310026728731,
+      "learning_rate": 1.5966056166019453e-05,
+      "loss": 0.8489,
+      "step": 1880
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9121428993183236,
+      "learning_rate": 1.5961662874203202e-05,
+      "loss": 0.5649,
+      "step": 1881
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1657134572985834,
+      "learning_rate": 1.595726779654359e-05,
+      "loss": 1.1618,
+      "step": 1882
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1634392908652793,
+      "learning_rate": 1.5952870934357184e-05,
+      "loss": 0.9068,
+      "step": 1883
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8238871197852596,
+      "learning_rate": 1.5948472288961087e-05,
+      "loss": 0.8694,
+      "step": 1884
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0495828004838033,
+      "learning_rate": 1.594407186167293e-05,
+      "loss": 0.7449,
+      "step": 1885
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7979460199437525,
+      "learning_rate": 1.5939669653810882e-05,
+      "loss": 0.9867,
+      "step": 1886
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2128309311750833,
+      "learning_rate": 1.593526566669364e-05,
+      "loss": 1.0905,
+      "step": 1887
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8714419819805804,
+      "learning_rate": 1.5930859901640445e-05,
+      "loss": 0.6958,
+      "step": 1888
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0421987906559846,
+      "learning_rate": 1.5926452359971062e-05,
+      "loss": 0.7949,
+      "step": 1889
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1840470666487801,
+      "learning_rate": 1.592204304300579e-05,
+      "loss": 0.9942,
+      "step": 1890
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1139851064667925,
+      "learning_rate": 1.591763195206546e-05,
+      "loss": 1.0842,
+      "step": 1891
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5675044911318996,
+      "learning_rate": 1.5913219088471437e-05,
+      "loss": 1.1047,
+      "step": 1892
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8413118242653359,
+      "learning_rate": 1.5908804453545608e-05,
+      "loss": 0.6411,
+      "step": 1893
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9390736288672475,
+      "learning_rate": 1.59043880486104e-05,
+      "loss": 1.0537,
+      "step": 1894
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9496175733536132,
+      "learning_rate": 1.5899969874988774e-05,
+      "loss": 0.674,
+      "step": 1895
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8759212804208998,
+      "learning_rate": 1.5895549934004204e-05,
+      "loss": 0.7768,
+      "step": 1896
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9848896647599868,
+      "learning_rate": 1.5891128226980707e-05,
+      "loss": 1.0962,
+      "step": 1897
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1444083535325766,
+      "learning_rate": 1.588670475524283e-05,
+      "loss": 0.8711,
+      "step": 1898
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8750135005377075,
+      "learning_rate": 1.588227952011564e-05,
+      "loss": 0.7699,
+      "step": 1899
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.073737593559836,
+      "learning_rate": 1.5877852522924733e-05,
+      "loss": 0.6883,
+      "step": 1900
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9831004928716861,
+      "learning_rate": 1.5873423764996243e-05,
+      "loss": 1.0755,
+      "step": 1901
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1819558003711348,
+      "learning_rate": 1.5868993247656822e-05,
+      "loss": 0.6948,
+      "step": 1902
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8705327092525937,
+      "learning_rate": 1.586456097223365e-05,
+      "loss": 0.7983,
+      "step": 1903
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.122115691099322,
+      "learning_rate": 1.586012694005444e-05,
+      "loss": 1.0338,
+      "step": 1904
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1016748032104344,
+      "learning_rate": 1.5855691152447425e-05,
+      "loss": 0.7667,
+      "step": 1905
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.062476864947796,
+      "learning_rate": 1.5851253610741357e-05,
+      "loss": 1.1484,
+      "step": 1906
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3455927420841076,
+      "learning_rate": 1.584681431626553e-05,
+      "loss": 0.958,
+      "step": 1907
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5175392456248569,
+      "learning_rate": 1.584237327034976e-05,
+      "loss": 0.9259,
+      "step": 1908
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0326910932754136,
+      "learning_rate": 1.583793047432437e-05,
+      "loss": 0.7842,
+      "step": 1909
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1018301806815791,
+      "learning_rate": 1.583348592952023e-05,
+      "loss": 0.8566,
+      "step": 1910
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9888305982020317,
+      "learning_rate": 1.582903963726872e-05,
+      "loss": 1.0311,
+      "step": 1911
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1075626038577104,
+      "learning_rate": 1.582459159890175e-05,
+      "loss": 0.9536,
+      "step": 1912
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8705580350956036,
+      "learning_rate": 1.5820141815751743e-05,
+      "loss": 0.7839,
+      "step": 1913
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0933606771338187,
+      "learning_rate": 1.581569028915166e-05,
+      "loss": 0.9828,
+      "step": 1914
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.083669974809549,
+      "learning_rate": 1.5811237020434968e-05,
+      "loss": 0.7357,
+      "step": 1915
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9901947224159623,
+      "learning_rate": 1.580678201093567e-05,
+      "loss": 1.1017,
+      "step": 1916
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9646720924484249,
+      "learning_rate": 1.580232526198828e-05,
+      "loss": 0.8951,
+      "step": 1917
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8702809672589866,
+      "learning_rate": 1.5797866774927848e-05,
+      "loss": 0.733,
+      "step": 1918
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.082172995905817,
+      "learning_rate": 1.5793406551089922e-05,
+      "loss": 0.6706,
+      "step": 1919
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0560566516861052,
+      "learning_rate": 1.5788944591810588e-05,
+      "loss": 1.0605,
+      "step": 1920
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2044245981658048,
+      "learning_rate": 1.5784480898426444e-05,
+      "loss": 0.8934,
+      "step": 1921
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.501177575684377,
+      "learning_rate": 1.5780015472274613e-05,
+      "loss": 0.8289,
+      "step": 1922
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0058654415092234,
+      "learning_rate": 1.577554831469273e-05,
+      "loss": 1.1473,
+      "step": 1923
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2239023766116492,
+      "learning_rate": 1.5771079427018956e-05,
+      "loss": 0.784,
+      "step": 1924
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1564265255826933,
+      "learning_rate": 1.576660881059196e-05,
+      "loss": 1.114,
+      "step": 1925
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0956381892594627,
+      "learning_rate": 1.5762136466750947e-05,
+      "loss": 0.734,
+      "step": 1926
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.329531820497527,
+      "learning_rate": 1.575766239683562e-05,
+      "loss": 0.9874,
+      "step": 1927
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0783153095898061,
+      "learning_rate": 1.5753186602186207e-05,
+      "loss": 0.7993,
+      "step": 1928
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7628395346076711,
+      "learning_rate": 1.5748709084143455e-05,
+      "loss": 0.7634,
+      "step": 1929
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1382230519526317,
+      "learning_rate": 1.5744229844048627e-05,
+      "loss": 0.9538,
+      "step": 1930
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1068306276591466,
+      "learning_rate": 1.5739748883243496e-05,
+      "loss": 0.8335,
+      "step": 1931
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1432149647596113,
+      "learning_rate": 1.5735266203070354e-05,
+      "loss": 0.8556,
+      "step": 1932
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0929272836577693,
+      "learning_rate": 1.573078180487201e-05,
+      "loss": 0.8131,
+      "step": 1933
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9946881241095925,
+      "learning_rate": 1.5726295689991787e-05,
+      "loss": 0.9891,
+      "step": 1934
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0129837230830177,
+      "learning_rate": 1.572180785977352e-05,
+      "loss": 1.1249,
+      "step": 1935
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6394234602131662,
+      "learning_rate": 1.571731831556156e-05,
+      "loss": 0.4691,
+      "step": 1936
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1629077135590498,
+      "learning_rate": 1.571282705870077e-05,
+      "loss": 0.8069,
+      "step": 1937
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2733461392121863,
+      "learning_rate": 1.570833409053653e-05,
+      "loss": 0.7535,
+      "step": 1938
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.024642329752059,
+      "learning_rate": 1.5703839412414718e-05,
+      "loss": 1.025,
+      "step": 1939
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0514771590495808,
+      "learning_rate": 1.5699343025681746e-05,
+      "loss": 0.8591,
+      "step": 1940
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9118392548617675,
+      "learning_rate": 1.569484493168452e-05,
+      "loss": 1.0067,
+      "step": 1941
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.207158743090506,
+      "learning_rate": 1.5690345131770474e-05,
+      "loss": 0.76,
+      "step": 1942
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0179740026606932,
+      "learning_rate": 1.5685843627287534e-05,
+      "loss": 0.8823,
+      "step": 1943
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1285401147349603,
+      "learning_rate": 1.5681340419584155e-05,
+      "loss": 1.1613,
+      "step": 1944
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.107469820948485,
+      "learning_rate": 1.5676835510009286e-05,
+      "loss": 1.1057,
+      "step": 1945
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3769498657179247,
+      "learning_rate": 1.5672328899912397e-05,
+      "loss": 0.7732,
+      "step": 1946
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0014753429907657,
+      "learning_rate": 1.5667820590643464e-05,
+      "loss": 0.9853,
+      "step": 1947
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9963997532474277,
+      "learning_rate": 1.5663310583552968e-05,
+      "loss": 0.8689,
+      "step": 1948
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0600100018096719,
+      "learning_rate": 1.5658798879991905e-05,
+      "loss": 1.0006,
+      "step": 1949
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1436544956427623,
+      "learning_rate": 1.5654285481311777e-05,
+      "loss": 0.7939,
+      "step": 1950
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1966665749702983,
+      "learning_rate": 1.5649770388864588e-05,
+      "loss": 0.844,
+      "step": 1951
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4280908874196436,
+      "learning_rate": 1.5645253604002863e-05,
+      "loss": 0.9098,
+      "step": 1952
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1543730165959332,
+      "learning_rate": 1.564073512807962e-05,
+      "loss": 0.7785,
+      "step": 1953
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.871301149782576,
+      "learning_rate": 1.5636214962448396e-05,
+      "loss": 0.7718,
+      "step": 1954
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9872506068822846,
+      "learning_rate": 1.5631693108463215e-05,
+      "loss": 0.8364,
+      "step": 1955
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0209285033236248,
+      "learning_rate": 1.5627169567478627e-05,
+      "loss": 1.0104,
+      "step": 1956
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1715855858507302,
+      "learning_rate": 1.5622644340849682e-05,
+      "loss": 0.8027,
+      "step": 1957
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9330185940440721,
+      "learning_rate": 1.5618117429931923e-05,
+      "loss": 1.0483,
+      "step": 1958
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1341534688348698,
+      "learning_rate": 1.5613588836081417e-05,
+      "loss": 0.6749,
+      "step": 1959
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9557434914266026,
+      "learning_rate": 1.560905856065472e-05,
+      "loss": 0.7789,
+      "step": 1960
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.79435470398545,
+      "learning_rate": 1.5604526605008892e-05,
+      "loss": 1.0065,
+      "step": 1961
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0056042761897124,
+      "learning_rate": 1.5599992970501512e-05,
+      "loss": 0.7518,
+      "step": 1962
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1005264761199633,
+      "learning_rate": 1.5595457658490643e-05,
+      "loss": 1.1545,
+      "step": 1963
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.082767264298404,
+      "learning_rate": 1.5590920670334863e-05,
+      "loss": 0.7686,
+      "step": 1964
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3367016970805523,
+      "learning_rate": 1.5586382007393237e-05,
+      "loss": 1.0129,
+      "step": 1965
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1324583871355418,
+      "learning_rate": 1.5581841671025358e-05,
+      "loss": 1.1473,
+      "step": 1966
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0643455492386045,
+      "learning_rate": 1.557729966259129e-05,
+      "loss": 0.8759,
+      "step": 1967
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0144667616412923,
+      "learning_rate": 1.5572755983451626e-05,
+      "loss": 1.2023,
+      "step": 1968
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9816893957294265,
+      "learning_rate": 1.5568210634967432e-05,
+      "loss": 0.8706,
+      "step": 1969
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.222946349289189,
+      "learning_rate": 1.55636636185003e-05,
+      "loss": 0.7678,
+      "step": 1970
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3960639932840928,
+      "learning_rate": 1.5559114935412307e-05,
+      "loss": 0.7059,
+      "step": 1971
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1279006223005903,
+      "learning_rate": 1.5554564587066023e-05,
+      "loss": 0.773,
+      "step": 1972
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0015732214254756,
+      "learning_rate": 1.5550012574824535e-05,
+      "loss": 1.136,
+      "step": 1973
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7653070725274363,
+      "learning_rate": 1.5545458900051417e-05,
+      "loss": 0.6109,
+      "step": 1974
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5495967142776523,
+      "learning_rate": 1.554090356411074e-05,
+      "loss": 0.8537,
+      "step": 1975
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8019314327717484,
+      "learning_rate": 1.5536346568367075e-05,
+      "loss": 0.5929,
+      "step": 1976
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1240163567483314,
+      "learning_rate": 1.55317879141855e-05,
+      "loss": 1.1592,
+      "step": 1977
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1203200576912533,
+      "learning_rate": 1.552722760293157e-05,
+      "loss": 0.8155,
+      "step": 1978
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.005623959062711,
+      "learning_rate": 1.552266563597135e-05,
+      "loss": 0.8062,
+      "step": 1979
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9639111049426748,
+      "learning_rate": 1.5518102014671405e-05,
+      "loss": 0.6962,
+      "step": 1980
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3738360212896006,
+      "learning_rate": 1.551353674039878e-05,
+      "loss": 0.9892,
+      "step": 1981
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4417866957531058,
+      "learning_rate": 1.5508969814521026e-05,
+      "loss": 1.1572,
+      "step": 1982
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1823947678878985,
+      "learning_rate": 1.5504401238406188e-05,
+      "loss": 0.7523,
+      "step": 1983
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0728164763132038,
+      "learning_rate": 1.5499831013422804e-05,
+      "loss": 0.7415,
+      "step": 1984
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0443329780469235,
+      "learning_rate": 1.5495259140939904e-05,
+      "loss": 0.9848,
+      "step": 1985
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9482005099688239,
+      "learning_rate": 1.5490685622327015e-05,
+      "loss": 0.641,
+      "step": 1986
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9059874719215245,
+      "learning_rate": 1.5486110458954155e-05,
+      "loss": 1.0157,
+      "step": 1987
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9133487794669185,
+      "learning_rate": 1.5481533652191836e-05,
+      "loss": 1.0907,
+      "step": 1988
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4490364842575285,
+      "learning_rate": 1.5476955203411057e-05,
+      "loss": 1.069,
+      "step": 1989
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.06111930996175,
+      "learning_rate": 1.547237511398332e-05,
+      "loss": 0.7819,
+      "step": 1990
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6086174339282027,
+      "learning_rate": 1.5467793385280602e-05,
+      "loss": 0.4566,
+      "step": 1991
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0439447054133764,
+      "learning_rate": 1.5463210018675395e-05,
+      "loss": 1.0741,
+      "step": 1992
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4716293958627245,
+      "learning_rate": 1.5458625015540652e-05,
+      "loss": 0.8817,
+      "step": 1993
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.27408801076256,
+      "learning_rate": 1.5454038377249845e-05,
+      "loss": 0.9034,
+      "step": 1994
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0941957865185195,
+      "learning_rate": 1.544945010517691e-05,
+      "loss": 0.7588,
+      "step": 1995
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0262938763851135,
+      "learning_rate": 1.5444860200696297e-05,
+      "loss": 1.0265,
+      "step": 1996
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.114357923572343,
+      "learning_rate": 1.5440268665182926e-05,
+      "loss": 0.8997,
+      "step": 1997
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1771133089341446,
+      "learning_rate": 1.5435675500012212e-05,
+      "loss": 0.8248,
+      "step": 1998
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4944595670724712,
+      "learning_rate": 1.5431080706560062e-05,
+      "loss": 0.8234,
+      "step": 1999
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0047362571455778,
+      "learning_rate": 1.5426484286202863e-05,
+      "loss": 0.7633,
+      "step": 2000
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0129840661460419,
+      "learning_rate": 1.5421886240317498e-05,
+      "loss": 1.1093,
+      "step": 2001
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0936641260644606,
+      "learning_rate": 1.5417286570281334e-05,
+      "loss": 0.8371,
+      "step": 2002
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9593121704693686,
+      "learning_rate": 1.5412685277472216e-05,
+      "loss": 0.7494,
+      "step": 2003
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.071676036546905,
+      "learning_rate": 1.540808236326849e-05,
+      "loss": 0.7009,
+      "step": 2004
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0621802376064495,
+      "learning_rate": 1.540347782904897e-05,
+      "loss": 0.6973,
+      "step": 2005
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0780171724918615,
+      "learning_rate": 1.5398871676192978e-05,
+      "loss": 1.2151,
+      "step": 2006
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9389034511053977,
+      "learning_rate": 1.5394263906080303e-05,
+      "loss": 0.8549,
+      "step": 2007
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6476124006383496,
+      "learning_rate": 1.5389654520091217e-05,
+      "loss": 0.609,
+      "step": 2008
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9317910099806508,
+      "learning_rate": 1.538504351960649e-05,
+      "loss": 0.9707,
+      "step": 2009
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.981034844098329,
+      "learning_rate": 1.538043090600737e-05,
+      "loss": 0.9672,
+      "step": 2010
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.173437991341633,
+      "learning_rate": 1.5375816680675578e-05,
+      "loss": 1.2255,
+      "step": 2011
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8673829494669926,
+      "learning_rate": 1.5371200844993332e-05,
+      "loss": 0.7547,
+      "step": 2012
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5818609774531074,
+      "learning_rate": 1.5366583400343326e-05,
+      "loss": 0.8876,
+      "step": 2013
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0386523150765306,
+      "learning_rate": 1.5361964348108736e-05,
+      "loss": 0.8431,
+      "step": 2014
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.100888929410274,
+      "learning_rate": 1.5357343689673223e-05,
+      "loss": 1.1678,
+      "step": 2015
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0677763430664908,
+      "learning_rate": 1.535272142642092e-05,
+      "loss": 0.805,
+      "step": 2016
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0573793195325183,
+      "learning_rate": 1.5348097559736456e-05,
+      "loss": 0.8575,
+      "step": 2017
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9322415799974682,
+      "learning_rate": 1.5343472091004925e-05,
+      "loss": 0.7984,
+      "step": 2018
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2369497990851572,
+      "learning_rate": 1.5338845021611906e-05,
+      "loss": 1.0172,
+      "step": 2019
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.122716594507025,
+      "learning_rate": 1.5334216352943464e-05,
+      "loss": 1.1212,
+      "step": 2020
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0845918086894037,
+      "learning_rate": 1.5329586086386137e-05,
+      "loss": 0.7859,
+      "step": 2021
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9418793160809807,
+      "learning_rate": 1.532495422332694e-05,
+      "loss": 0.8663,
+      "step": 2022
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9301832042666719,
+      "learning_rate": 1.5320320765153367e-05,
+      "loss": 0.7912,
+      "step": 2023
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.290557094984762,
+      "learning_rate": 1.5315685713253396e-05,
+      "loss": 0.6922,
+      "step": 2024
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0692976400624141,
+      "learning_rate": 1.531104906901548e-05,
+      "loss": 1.1497,
+      "step": 2025
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.091178076605164,
+      "learning_rate": 1.5306410833828534e-05,
+      "loss": 0.7803,
+      "step": 2026
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1724223243500795,
+      "learning_rate": 1.530177100908198e-05,
+      "loss": 0.542,
+      "step": 2027
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0964742361621431,
+      "learning_rate": 1.5297129596165684e-05,
+      "loss": 0.7772,
+      "step": 2028
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.031393562856331,
+      "learning_rate": 1.5292486596470015e-05,
+      "loss": 0.8713,
+      "step": 2029
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0544902552473447,
+      "learning_rate": 1.5287842011385798e-05,
+      "loss": 1.4257,
+      "step": 2030
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9584729121279578,
+      "learning_rate": 1.5283195842304338e-05,
+      "loss": 0.8092,
+      "step": 2031
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8740781664206992,
+      "learning_rate": 1.527854809061742e-05,
+      "loss": 0.9664,
+      "step": 2032
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9979510628189285,
+      "learning_rate": 1.5273898757717295e-05,
+      "loss": 0.7617,
+      "step": 2033
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1985920243693604,
+      "learning_rate": 1.52692478449967e-05,
+      "loss": 1.1604,
+      "step": 2034
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9896859827758672,
+      "learning_rate": 1.5264595353848823e-05,
+      "loss": 0.7575,
+      "step": 2035
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7270004766067759,
+      "learning_rate": 1.5259941285667355e-05,
+      "loss": 0.9677,
+      "step": 2036
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9932972058333551,
+      "learning_rate": 1.5255285641846436e-05,
+      "loss": 0.8837,
+      "step": 2037
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8700254653129748,
+      "learning_rate": 1.5250628423780684e-05,
+      "loss": 1.0183,
+      "step": 2038
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1339124972497001,
+      "learning_rate": 1.5245969632865194e-05,
+      "loss": 1.2435,
+      "step": 2039
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0611566680672075,
+      "learning_rate": 1.5241309270495524e-05,
+      "loss": 0.712,
+      "step": 2040
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8830743550755895,
+      "learning_rate": 1.5236647338067708e-05,
+      "loss": 0.5544,
+      "step": 2041
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1308540122365454,
+      "learning_rate": 1.523198383697825e-05,
+      "loss": 0.9668,
+      "step": 2042
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8883429796082727,
+      "learning_rate": 1.5227318768624124e-05,
+      "loss": 0.6022,
+      "step": 2043
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4418597910124413,
+      "learning_rate": 1.5222652134402766e-05,
+      "loss": 0.5762,
+      "step": 2044
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0249638747214833,
+      "learning_rate": 1.52179839357121e-05,
+      "loss": 1.1676,
+      "step": 2045
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2034266783389072,
+      "learning_rate": 1.5213314173950498e-05,
+      "loss": 0.8858,
+      "step": 2046
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0253103480386552,
+      "learning_rate": 1.5208642850516806e-05,
+      "loss": 0.7811,
+      "step": 2047
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.207452434698248,
+      "learning_rate": 1.5203969966810346e-05,
+      "loss": 0.839,
+      "step": 2048
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9965295556186401,
+      "learning_rate": 1.51992955242309e-05,
+      "loss": 1.0793,
+      "step": 2049
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1616195043440876,
+      "learning_rate": 1.5194619524178718e-05,
+      "loss": 0.7394,
+      "step": 2050
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0686026611566148,
+      "learning_rate": 1.5189941968054518e-05,
+      "loss": 0.792,
+      "step": 2051
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.094027459927855,
+      "learning_rate": 1.5185262857259482e-05,
+      "loss": 0.8576,
+      "step": 2052
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9665552540393372,
+      "learning_rate": 1.5180582193195261e-05,
+      "loss": 1.2664,
+      "step": 2053
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.054333059970649,
+      "learning_rate": 1.5175899977263963e-05,
+      "loss": 0.7502,
+      "step": 2054
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0894473227391943,
+      "learning_rate": 1.517121621086818e-05,
+      "loss": 0.7453,
+      "step": 2055
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0385934226975622,
+      "learning_rate": 1.5166530895410944e-05,
+      "loss": 0.7604,
+      "step": 2056
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3252329952462434,
+      "learning_rate": 1.516184403229577e-05,
+      "loss": 0.9369,
+      "step": 2057
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0672515296737064,
+      "learning_rate": 1.515715562292662e-05,
+      "loss": 1.0224,
+      "step": 2058
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9926307563381658,
+      "learning_rate": 1.5152465668707938e-05,
+      "loss": 1.0298,
+      "step": 2059
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2418282413355972,
+      "learning_rate": 1.5147774171044619e-05,
+      "loss": 0.7163,
+      "step": 2060
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0783922768022887,
+      "learning_rate": 1.5143081131342015e-05,
+      "loss": 1.0072,
+      "step": 2061
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9610206377534805,
+      "learning_rate": 1.513838655100596e-05,
+      "loss": 0.7865,
+      "step": 2062
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2061509256706209,
+      "learning_rate": 1.5133690431442725e-05,
+      "loss": 0.7804,
+      "step": 2063
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.057077004148222,
+      "learning_rate": 1.5128992774059063e-05,
+      "loss": 0.7415,
+      "step": 2064
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0415230520627485,
+      "learning_rate": 1.5124293580262176e-05,
+      "loss": 0.8452,
+      "step": 2065
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0402180501568172,
+      "learning_rate": 1.5119592851459726e-05,
+      "loss": 0.7665,
+      "step": 2066
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.24595108581929,
+      "learning_rate": 1.5114890589059843e-05,
+      "loss": 0.881,
+      "step": 2067
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1230241343240663,
+      "learning_rate": 1.5110186794471105e-05,
+      "loss": 1.1402,
+      "step": 2068
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4421900940942693,
+      "learning_rate": 1.5105481469102561e-05,
+      "loss": 0.808,
+      "step": 2069
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.091824370275386,
+      "learning_rate": 1.5100774614363708e-05,
+      "loss": 1.0265,
+      "step": 2070
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5028895144367682,
+      "learning_rate": 1.509606623166451e-05,
+      "loss": 0.7856,
+      "step": 2071
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1087507594170858,
+      "learning_rate": 1.5091356322415382e-05,
+      "loss": 0.9017,
+      "step": 2072
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0804267587997007,
+      "learning_rate": 1.5086644888027202e-05,
+      "loss": 1.1114,
+      "step": 2073
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9995300430038424,
+      "learning_rate": 1.5081931929911298e-05,
+      "loss": 0.7997,
+      "step": 2074
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0385320607648258,
+      "learning_rate": 1.5077217449479455e-05,
+      "loss": 0.8578,
+      "step": 2075
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1312545329894572,
+      "learning_rate": 1.5072501448143926e-05,
+      "loss": 0.9339,
+      "step": 2076
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.000638038149949,
+      "learning_rate": 1.5067783927317403e-05,
+      "loss": 0.8313,
+      "step": 2077
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.193155737313476,
+      "learning_rate": 1.5063064888413048e-05,
+      "loss": 1.1328,
+      "step": 2078
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0213088304800073,
+      "learning_rate": 1.5058344332844463e-05,
+      "loss": 0.7833,
+      "step": 2079
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3995259008049548,
+      "learning_rate": 1.5053622262025718e-05,
+      "loss": 0.498,
+      "step": 2080
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0469235689805771,
+      "learning_rate": 1.5048898677371328e-05,
+      "loss": 0.8282,
+      "step": 2081
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.983530485266429,
+      "learning_rate": 1.5044173580296267e-05,
+      "loss": 1.0545,
+      "step": 2082
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0188361246596698,
+      "learning_rate": 1.5039446972215955e-05,
+      "loss": 1.0362,
+      "step": 2083
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9415634081249616,
+      "learning_rate": 1.5034718854546273e-05,
+      "loss": 1.0804,
+      "step": 2084
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9788374648239082,
+      "learning_rate": 1.5029989228703548e-05,
+      "loss": 0.747,
+      "step": 2085
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5343888439583289,
+      "learning_rate": 1.5025258096104563e-05,
+      "loss": 0.8639,
+      "step": 2086
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1294814906275588,
+      "learning_rate": 1.5020525458166546e-05,
+      "loss": 1.0891,
+      "step": 2087
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0675285740405462,
+      "learning_rate": 1.5015791316307188e-05,
+      "loss": 0.8328,
+      "step": 2088
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0306752650133622,
+      "learning_rate": 1.5011055671944616e-05,
+      "loss": 0.8923,
+      "step": 2089
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3102087927736388,
+      "learning_rate": 1.500631852649742e-05,
+      "loss": 0.7958,
+      "step": 2090
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2352900099358533,
+      "learning_rate": 1.5001579881384624e-05,
+      "loss": 0.873,
+      "step": 2091
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0736921267134465,
+      "learning_rate": 1.499683973802572e-05,
+      "loss": 1.0842,
+      "step": 2092
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2244528935629706,
+      "learning_rate": 1.4992098097840635e-05,
+      "loss": 0.8003,
+      "step": 2093
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1582287002957083,
+      "learning_rate": 1.4987354962249749e-05,
+      "loss": 0.8387,
+      "step": 2094
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1966963943120017,
+      "learning_rate": 1.4982610332673894e-05,
+      "loss": 0.984,
+      "step": 2095
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0166336846289634,
+      "learning_rate": 1.4977864210534341e-05,
+      "loss": 0.7725,
+      "step": 2096
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.97948906016509,
+      "learning_rate": 1.4973116597252817e-05,
+      "loss": 1.0919,
+      "step": 2097
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0668879080972689,
+      "learning_rate": 1.4968367494251486e-05,
+      "loss": 1.1522,
+      "step": 2098
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9999931827436032,
+      "learning_rate": 1.4963616902952967e-05,
+      "loss": 0.7105,
+      "step": 2099
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7539855547656702,
+      "learning_rate": 1.495886482478032e-05,
+      "loss": 0.411,
+      "step": 2100
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8718607391881447,
+      "learning_rate": 1.4954111261157049e-05,
+      "loss": 0.8452,
+      "step": 2101
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4128763337392107,
+      "learning_rate": 1.4949356213507113e-05,
+      "loss": 0.9795,
+      "step": 2102
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9953744551221371,
+      "learning_rate": 1.4944599683254903e-05,
+      "loss": 0.8785,
+      "step": 2103
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9960624186755647,
+      "learning_rate": 1.4939841671825262e-05,
+      "loss": 0.7884,
+      "step": 2104
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2034741373848674,
+      "learning_rate": 1.493508218064347e-05,
+      "loss": 0.8763,
+      "step": 2105
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1294107800548125,
+      "learning_rate": 1.4930321211135258e-05,
+      "loss": 1.1468,
+      "step": 2106
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0527052909956174,
+      "learning_rate": 1.4925558764726795e-05,
+      "loss": 0.8981,
+      "step": 2107
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8995727361506247,
+      "learning_rate": 1.4920794842844693e-05,
+      "loss": 0.9791,
+      "step": 2108
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.931764821189926,
+      "learning_rate": 1.4916029446916008e-05,
+      "loss": 1.0519,
+      "step": 2109
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9672370637225338,
+      "learning_rate": 1.4911262578368233e-05,
+      "loss": 0.7741,
+      "step": 2110
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0154614128551502,
+      "learning_rate": 1.490649423862931e-05,
+      "loss": 1.0107,
+      "step": 2111
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3515644462173637,
+      "learning_rate": 1.4901724429127615e-05,
+      "loss": 0.8331,
+      "step": 2112
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0823702310882608,
+      "learning_rate": 1.4896953151291964e-05,
+      "loss": 0.7165,
+      "step": 2113
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.224788837070571,
+      "learning_rate": 1.4892180406551618e-05,
+      "loss": 0.895,
+      "step": 2114
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2659396856699003,
+      "learning_rate": 1.4887406196336271e-05,
+      "loss": 0.7694,
+      "step": 2115
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.519405343216394,
+      "learning_rate": 1.4882630522076064e-05,
+      "loss": 0.7099,
+      "step": 2116
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0421079644462086,
+      "learning_rate": 1.4877853385201569e-05,
+      "loss": 0.8625,
+      "step": 2117
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0316520132579905,
+      "learning_rate": 1.48730747871438e-05,
+      "loss": 0.7379,
+      "step": 2118
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.234731277902982,
+      "learning_rate": 1.4868294729334209e-05,
+      "loss": 0.9162,
+      "step": 2119
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9350105444152835,
+      "learning_rate": 1.4863513213204681e-05,
+      "loss": 0.9376,
+      "step": 2120
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0787641467814166,
+      "learning_rate": 1.4858730240187547e-05,
+      "loss": 1.1824,
+      "step": 2121
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2585916166796411,
+      "learning_rate": 1.4853945811715563e-05,
+      "loss": 0.9133,
+      "step": 2122
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.959334005779115,
+      "learning_rate": 1.4849159929221926e-05,
+      "loss": 0.7627,
+      "step": 2123
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1227102788891339,
+      "learning_rate": 1.4844372594140271e-05,
+      "loss": 0.6658,
+      "step": 2124
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9659918651963442,
+      "learning_rate": 1.4839583807904668e-05,
+      "loss": 0.8001,
+      "step": 2125
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2505578532389807,
+      "learning_rate": 1.4834793571949617e-05,
+      "loss": 1.1951,
+      "step": 2126
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0527110859652251,
+      "learning_rate": 1.4830001887710054e-05,
+      "loss": 0.9078,
+      "step": 2127
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0100672251547387,
+      "learning_rate": 1.4825208756621354e-05,
+      "loss": 0.7505,
+      "step": 2128
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9934822130085287,
+      "learning_rate": 1.482041418011932e-05,
+      "loss": 0.7383,
+      "step": 2129
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9492758579948931,
+      "learning_rate": 1.4815618159640188e-05,
+      "loss": 1.0522,
+      "step": 2130
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2120965233764074,
+      "learning_rate": 1.4810820696620625e-05,
+      "loss": 0.8757,
+      "step": 2131
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.11853916717903,
+      "learning_rate": 1.480602179249774e-05,
+      "loss": 0.984,
+      "step": 2132
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1685881131496396,
+      "learning_rate": 1.4801221448709062e-05,
+      "loss": 0.9034,
+      "step": 2133
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9247077135274574,
+      "learning_rate": 1.4796419666692555e-05,
+      "loss": 1.005,
+      "step": 2134
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9676648374321724,
+      "learning_rate": 1.4791616447886619e-05,
+      "loss": 0.6601,
+      "step": 2135
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.120141620146905,
+      "learning_rate": 1.4786811793730076e-05,
+      "loss": 0.7949,
+      "step": 2136
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1224691314190418,
+      "learning_rate": 1.4782005705662187e-05,
+      "loss": 0.7461,
+      "step": 2137
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7438972830711996,
+      "learning_rate": 1.477719818512263e-05,
+      "loss": 0.8918,
+      "step": 2138
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1389293910745761,
+      "learning_rate": 1.4772389233551528e-05,
+      "loss": 0.7785,
+      "step": 2139
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1735305516257886,
+      "learning_rate": 1.476757885238942e-05,
+      "loss": 1.2189,
+      "step": 2140
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5534018244889902,
+      "learning_rate": 1.4762767043077277e-05,
+      "loss": 0.9552,
+      "step": 2141
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1598970460398328,
+      "learning_rate": 1.47579538070565e-05,
+      "loss": 0.7858,
+      "step": 2142
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0238169708459135,
+      "learning_rate": 1.4753139145768915e-05,
+      "loss": 0.7094,
+      "step": 2143
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0720999960744764,
+      "learning_rate": 1.4748323060656777e-05,
+      "loss": 1.1327,
+      "step": 2144
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3205389433468506,
+      "learning_rate": 1.4743505553162765e-05,
+      "loss": 0.8702,
+      "step": 2145
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0520236706827526,
+      "learning_rate": 1.4738686624729987e-05,
+      "loss": 0.9488,
+      "step": 2146
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1275908446915974,
+      "learning_rate": 1.4733866276801976e-05,
+      "loss": 0.8708,
+      "step": 2147
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.992433951510744,
+      "learning_rate": 1.4729044510822685e-05,
+      "loss": 0.8858,
+      "step": 2148
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9320370476022573,
+      "learning_rate": 1.4724221328236498e-05,
+      "loss": 0.9308,
+      "step": 2149
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1486013445589878,
+      "learning_rate": 1.4719396730488225e-05,
+      "loss": 0.9806,
+      "step": 2150
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.168364267634095,
+      "learning_rate": 1.4714570719023092e-05,
+      "loss": 0.7256,
+      "step": 2151
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.15639521322961,
+      "learning_rate": 1.4709743295286751e-05,
+      "loss": 0.8108,
+      "step": 2152
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9422439337035216,
+      "learning_rate": 1.4704914460725283e-05,
+      "loss": 0.5058,
+      "step": 2153
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.088333410150128,
+      "learning_rate": 1.4700084216785189e-05,
+      "loss": 1.1294,
+      "step": 2154
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1107262825530768,
+      "learning_rate": 1.4695252564913388e-05,
+      "loss": 0.8489,
+      "step": 2155
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8688679917723291,
+      "learning_rate": 1.4690419506557218e-05,
+      "loss": 1.0358,
+      "step": 2156
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1113333576878242,
+      "learning_rate": 1.4685585043164455e-05,
+      "loss": 0.8238,
+      "step": 2157
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9195636433620316,
+      "learning_rate": 1.4680749176183275e-05,
+      "loss": 0.622,
+      "step": 2158
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.054772331738837,
+      "learning_rate": 1.4675911907062289e-05,
+      "loss": 1.104,
+      "step": 2159
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.253875089554028,
+      "learning_rate": 1.4671073237250519e-05,
+      "loss": 0.8044,
+      "step": 2160
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9735418731634885,
+      "learning_rate": 1.4666233168197417e-05,
+      "loss": 0.8937,
+      "step": 2161
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0942700195430277,
+      "learning_rate": 1.4661391701352844e-05,
+      "loss": 1.1103,
+      "step": 2162
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.041849290170356,
+      "learning_rate": 1.4656548838167081e-05,
+      "loss": 1.0705,
+      "step": 2163
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1568725033335323,
+      "learning_rate": 1.4651704580090832e-05,
+      "loss": 0.8696,
+      "step": 2164
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1586780580158786,
+      "learning_rate": 1.464685892857522e-05,
+      "loss": 0.8456,
+      "step": 2165
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0104122537436244,
+      "learning_rate": 1.464201188507178e-05,
+      "loss": 0.8353,
+      "step": 2166
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3662009804585007,
+      "learning_rate": 1.4637163451032464e-05,
+      "loss": 0.91,
+      "step": 2167
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2353413943159177,
+      "learning_rate": 1.4632313627909641e-05,
+      "loss": 1.1084,
+      "step": 2168
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1369134284586973,
+      "learning_rate": 1.4627462417156104e-05,
+      "loss": 0.954,
+      "step": 2169
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9796255918909345,
+      "learning_rate": 1.4622609820225052e-05,
+      "loss": 0.7386,
+      "step": 2170
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9346260113062544,
+      "learning_rate": 1.46177558385701e-05,
+      "loss": 0.5188,
+      "step": 2171
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.084429047154308,
+      "learning_rate": 1.4612900473645285e-05,
+      "loss": 0.7143,
+      "step": 2172
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.055660336148831,
+      "learning_rate": 1.460804372690505e-05,
+      "loss": 1.1748,
+      "step": 2173
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9742409952318244,
+      "learning_rate": 1.4603185599804259e-05,
+      "loss": 0.7619,
+      "step": 2174
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1167953177280014,
+      "learning_rate": 1.4598326093798182e-05,
+      "loss": 0.795,
+      "step": 2175
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3407877367726984,
+      "learning_rate": 1.459346521034251e-05,
+      "loss": 0.8695,
+      "step": 2176
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0672035180097643,
+      "learning_rate": 1.4588602950893342e-05,
+      "loss": 0.6687,
+      "step": 2177
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0876359387189527,
+      "learning_rate": 1.4583739316907188e-05,
+      "loss": 1.1959,
+      "step": 2178
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9839846782298108,
+      "learning_rate": 1.4578874309840975e-05,
+      "loss": 0.9581,
+      "step": 2179
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1443892489731884,
+      "learning_rate": 1.4574007931152037e-05,
+      "loss": 0.9118,
+      "step": 2180
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0849551711302303,
+      "learning_rate": 1.4569140182298118e-05,
+      "loss": 0.7691,
+      "step": 2181
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0437391494125314,
+      "learning_rate": 1.4564271064737376e-05,
+      "loss": 1.1134,
+      "step": 2182
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1252144183313288,
+      "learning_rate": 1.4559400579928375e-05,
+      "loss": 0.8645,
+      "step": 2183
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7732277680755862,
+      "learning_rate": 1.4554528729330097e-05,
+      "loss": 0.9977,
+      "step": 2184
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0140385183991132,
+      "learning_rate": 1.4549655514401922e-05,
+      "loss": 0.892,
+      "step": 2185
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2050643552769424,
+      "learning_rate": 1.4544780936603644e-05,
+      "loss": 0.9254,
+      "step": 2186
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0750011214416415,
+      "learning_rate": 1.4539904997395468e-05,
+      "loss": 1.1042,
+      "step": 2187
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0206479535709563,
+      "learning_rate": 1.4535027698238003e-05,
+      "loss": 0.5238,
+      "step": 2188
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9765354346832275,
+      "learning_rate": 1.4530149040592262e-05,
+      "loss": 0.8423,
+      "step": 2189
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1608684432419232,
+      "learning_rate": 1.4525269025919677e-05,
+      "loss": 0.99,
+      "step": 2190
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9277792459272358,
+      "learning_rate": 1.452038765568207e-05,
+      "loss": 0.7039,
+      "step": 2191
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9817857091240663,
+      "learning_rate": 1.4515504931341683e-05,
+      "loss": 1.0263,
+      "step": 2192
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.044876798355397,
+      "learning_rate": 1.4510620854361159e-05,
+      "loss": 0.5505,
+      "step": 2193
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.119401124324622,
+      "learning_rate": 1.4505735426203545e-05,
+      "loss": 0.8509,
+      "step": 2194
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9093589331225743,
+      "learning_rate": 1.4500848648332289e-05,
+      "loss": 0.8502,
+      "step": 2195
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0803294381079456,
+      "learning_rate": 1.4495960522211252e-05,
+      "loss": 0.8635,
+      "step": 2196
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.012841377706794,
+      "learning_rate": 1.4491071049304692e-05,
+      "loss": 1.0725,
+      "step": 2197
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0322635094793247,
+      "learning_rate": 1.4486180231077278e-05,
+      "loss": 0.9527,
+      "step": 2198
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.210574274035199,
+      "learning_rate": 1.4481288068994071e-05,
+      "loss": 0.8989,
+      "step": 2199
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0125001477546829,
+      "learning_rate": 1.4476394564520542e-05,
+      "loss": 0.7958,
+      "step": 2200
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0497095953246454,
+      "learning_rate": 1.4471499719122565e-05,
+      "loss": 0.8026,
+      "step": 2201
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9815509610402049,
+      "learning_rate": 1.446660353426641e-05,
+      "loss": 1.0211,
+      "step": 2202
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.030501064432673,
+      "learning_rate": 1.4461706011418757e-05,
+      "loss": 0.7609,
+      "step": 2203
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.866572484616911,
+      "learning_rate": 1.4456807152046674e-05,
+      "loss": 1.1072,
+      "step": 2204
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1362890401816454,
+      "learning_rate": 1.4451906957617643e-05,
+      "loss": 0.9127,
+      "step": 2205
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1476495667571718,
+      "learning_rate": 1.4447005429599537e-05,
+      "loss": 0.791,
+      "step": 2206
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.056123782057852,
+      "learning_rate": 1.4442102569460632e-05,
+      "loss": 0.708,
+      "step": 2207
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0511132937485577,
+      "learning_rate": 1.4437198378669598e-05,
+      "loss": 0.8042,
+      "step": 2208
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8690509233023566,
+      "learning_rate": 1.4432292858695514e-05,
+      "loss": 0.7832,
+      "step": 2209
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0384466535097066,
+      "learning_rate": 1.4427386011007847e-05,
+      "loss": 0.7759,
+      "step": 2210
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0867257564639508,
+      "learning_rate": 1.4422477837076465e-05,
+      "loss": 1.0735,
+      "step": 2211
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.981683088189835,
+      "learning_rate": 1.4417568338371637e-05,
+      "loss": 0.7721,
+      "step": 2212
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.041523284645771,
+      "learning_rate": 1.4412657516364027e-05,
+      "loss": 0.7931,
+      "step": 2213
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.920414431185674,
+      "learning_rate": 1.4407745372524686e-05,
+      "loss": 0.8298,
+      "step": 2214
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0572138663594899,
+      "learning_rate": 1.4402831908325082e-05,
+      "loss": 0.8677,
+      "step": 2215
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2110199492917628,
+      "learning_rate": 1.4397917125237051e-05,
+      "loss": 1.189,
+      "step": 2216
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3340126110691466,
+      "learning_rate": 1.4393001024732849e-05,
+      "loss": 0.8072,
+      "step": 2217
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7183188588182832,
+      "learning_rate": 1.4388083608285112e-05,
+      "loss": 0.5242,
+      "step": 2218
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.232823384869386,
+      "learning_rate": 1.4383164877366879e-05,
+      "loss": 0.7761,
+      "step": 2219
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9798382216870141,
+      "learning_rate": 1.4378244833451576e-05,
+      "loss": 0.6572,
+      "step": 2220
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1571861198812423,
+      "learning_rate": 1.4373323478013021e-05,
+      "loss": 1.1927,
+      "step": 2221
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9466473320949513,
+      "learning_rate": 1.4368400812525434e-05,
+      "loss": 0.7843,
+      "step": 2222
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0224340377302483,
+      "learning_rate": 1.4363476838463417e-05,
+      "loss": 0.8299,
+      "step": 2223
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9277708338833512,
+      "learning_rate": 1.4358551557301974e-05,
+      "loss": 0.8363,
+      "step": 2224
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.885054103334104,
+      "learning_rate": 1.4353624970516492e-05,
+      "loss": 0.8772,
+      "step": 2225
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9138967371092541,
+      "learning_rate": 1.4348697079582752e-05,
+      "loss": 0.7252,
+      "step": 2226
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2457683081376658,
+      "learning_rate": 1.434376788597693e-05,
+      "loss": 0.8153,
+      "step": 2227
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2019285948779337,
+      "learning_rate": 1.4338837391175582e-05,
+      "loss": 0.8684,
+      "step": 2228
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8588408699785709,
+      "learning_rate": 1.4333905596655668e-05,
+      "loss": 1.0106,
+      "step": 2229
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2538861213988848,
+      "learning_rate": 1.4328972503894525e-05,
+      "loss": 1.1814,
+      "step": 2230
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0169898931503385,
+      "learning_rate": 1.4324038114369884e-05,
+      "loss": 1.0492,
+      "step": 2231
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1036414305776587,
+      "learning_rate": 1.4319102429559866e-05,
+      "loss": 0.7262,
+      "step": 2232
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9035919598350058,
+      "learning_rate": 1.4314165450942975e-05,
+      "loss": 1.0275,
+      "step": 2233
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.14146409872198,
+      "learning_rate": 1.4309227179998109e-05,
+      "loss": 0.7878,
+      "step": 2234
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0262173553866685,
+      "learning_rate": 1.4304287618204544e-05,
+      "loss": 1.0295,
+      "step": 2235
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0627109833313937,
+      "learning_rate": 1.4299346767041956e-05,
+      "loss": 0.732,
+      "step": 2236
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9990440179207281,
+      "learning_rate": 1.429440462799039e-05,
+      "loss": 0.9522,
+      "step": 2237
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3382025970887466,
+      "learning_rate": 1.4289461202530297e-05,
+      "loss": 0.7187,
+      "step": 2238
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9707618475479836,
+      "learning_rate": 1.4284516492142496e-05,
+      "loss": 0.7693,
+      "step": 2239
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1281978604978349,
+      "learning_rate": 1.4279570498308198e-05,
+      "loss": 1.122,
+      "step": 2240
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.216529905155665,
+      "learning_rate": 1.4274623222509001e-05,
+      "loss": 0.8161,
+      "step": 2241
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7744386674842775,
+      "learning_rate": 1.4269674666226879e-05,
+      "loss": 0.5212,
+      "step": 2242
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9203322587746832,
+      "learning_rate": 1.4264724830944198e-05,
+      "loss": 0.7088,
+      "step": 2243
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1674417697614592,
+      "learning_rate": 1.4259773718143706e-05,
+      "loss": 0.8831,
+      "step": 2244
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3823411870304638,
+      "learning_rate": 1.4254821329308528e-05,
+      "loss": 1.2414,
+      "step": 2245
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.883909825921558,
+      "learning_rate": 1.4249867665922175e-05,
+      "loss": 0.7263,
+      "step": 2246
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1060531561377562,
+      "learning_rate": 1.4244912729468541e-05,
+      "loss": 0.8385,
+      "step": 2247
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0406893526718166,
+      "learning_rate": 1.4239956521431899e-05,
+      "loss": 0.8444,
+      "step": 2248
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0461413814514655,
+      "learning_rate": 1.4234999043296901e-05,
+      "loss": 1.213,
+      "step": 2249
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1163496690022883,
+      "learning_rate": 1.4230040296548588e-05,
+      "loss": 1.1314,
+      "step": 2250
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9992390348869644,
+      "learning_rate": 1.4225080282672373e-05,
+      "loss": 0.7552,
+      "step": 2251
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0792353474054162,
+      "learning_rate": 1.4220119003154049e-05,
+      "loss": 1.0809,
+      "step": 2252
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9686645400841802,
+      "learning_rate": 1.421515645947979e-05,
+      "loss": 0.7987,
+      "step": 2253
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.164974670625421,
+      "learning_rate": 1.4210192653136151e-05,
+      "loss": 1.1137,
+      "step": 2254
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9044310762031794,
+      "learning_rate": 1.4205227585610065e-05,
+      "loss": 0.6699,
+      "step": 2255
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4414098350636968,
+      "learning_rate": 1.4200261258388834e-05,
+      "loss": 0.9967,
+      "step": 2256
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8676224021552459,
+      "learning_rate": 1.4195293672960148e-05,
+      "loss": 1.0026,
+      "step": 2257
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9442451919542993,
+      "learning_rate": 1.4190324830812067e-05,
+      "loss": 0.6829,
+      "step": 2258
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1521173754509093,
+      "learning_rate": 1.4185354733433036e-05,
+      "loss": 1.1173,
+      "step": 2259
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.295785736001703,
+      "learning_rate": 1.4180383382311867e-05,
+      "loss": 0.8164,
+      "step": 2260
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6421759290703369,
+      "learning_rate": 1.4175410778937747e-05,
+      "loss": 0.5322,
+      "step": 2261
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.989746874781872,
+      "learning_rate": 1.4170436924800251e-05,
+      "loss": 0.8964,
+      "step": 2262
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3637139891473644,
+      "learning_rate": 1.416546182138931e-05,
+      "loss": 0.7457,
+      "step": 2263
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1186651979622608,
+      "learning_rate": 1.4160485470195245e-05,
+      "loss": 1.1004,
+      "step": 2264
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8797770589755467,
+      "learning_rate": 1.4155507872708745e-05,
+      "loss": 0.7373,
+      "step": 2265
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.786451803431823,
+      "learning_rate": 1.415052903042087e-05,
+      "loss": 0.5502,
+      "step": 2266
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2788631098990837,
+      "learning_rate": 1.4145548944823055e-05,
+      "loss": 0.8242,
+      "step": 2267
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2518192010704814,
+      "learning_rate": 1.4140567617407106e-05,
+      "loss": 1.2504,
+      "step": 2268
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9409148701977513,
+      "learning_rate": 1.4135585049665207e-05,
+      "loss": 0.7528,
+      "step": 2269
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1993830320646806,
+      "learning_rate": 1.4130601243089905e-05,
+      "loss": 0.8157,
+      "step": 2270
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0375616681703743,
+      "learning_rate": 1.4125616199174125e-05,
+      "loss": 1.094,
+      "step": 2271
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1680860485864795,
+      "learning_rate": 1.4120629919411158e-05,
+      "loss": 0.737,
+      "step": 2272
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1341641508555893,
+      "learning_rate": 1.4115642405294665e-05,
+      "loss": 1.1343,
+      "step": 2273
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0023529902199422,
+      "learning_rate": 1.4110653658318682e-05,
+      "loss": 0.7793,
+      "step": 2274
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0747569297602981,
+      "learning_rate": 1.4105663679977608e-05,
+      "loss": 0.9684,
+      "step": 2275
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9170592297781603,
+      "learning_rate": 1.4100672471766217e-05,
+      "loss": 0.8281,
+      "step": 2276
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1525005909863428,
+      "learning_rate": 1.4095680035179645e-05,
+      "loss": 0.9157,
+      "step": 2277
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1295539232562088,
+      "learning_rate": 1.4090686371713403e-05,
+      "loss": 1.0415,
+      "step": 2278
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7136285069414862,
+      "learning_rate": 1.4085691482863358e-05,
+      "loss": 0.4623,
+      "step": 2279
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8555692164411929,
+      "learning_rate": 1.4080695370125761e-05,
+      "loss": 1.0401,
+      "step": 2280
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9535515962433888,
+      "learning_rate": 1.4075698034997219e-05,
+      "loss": 0.7413,
+      "step": 2281
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9808622621646805,
+      "learning_rate": 1.4070699478974698e-05,
+      "loss": 0.6871,
+      "step": 2282
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.009215208554837,
+      "learning_rate": 1.4065699703555548e-05,
+      "loss": 1.0168,
+      "step": 2283
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6756796376744003,
+      "learning_rate": 1.406069871023747e-05,
+      "loss": 0.9218,
+      "step": 2284
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1335380256244276,
+      "learning_rate": 1.4055696500518539e-05,
+      "loss": 0.7702,
+      "step": 2285
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.146370457810254,
+      "learning_rate": 1.4050693075897181e-05,
+      "loss": 0.9225,
+      "step": 2286
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8883409195954054,
+      "learning_rate": 1.4045688437872205e-05,
+      "loss": 0.8717,
+      "step": 2287
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9888419798437436,
+      "learning_rate": 1.4040682587942765e-05,
+      "loss": 1.0635,
+      "step": 2288
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1201579781468913,
+      "learning_rate": 1.403567552760839e-05,
+      "loss": 0.7758,
+      "step": 2289
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9312894208639603,
+      "learning_rate": 1.4030667258368968e-05,
+      "loss": 0.7508,
+      "step": 2290
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1029048724259536,
+      "learning_rate": 1.4025657781724746e-05,
+      "loss": 0.8615,
+      "step": 2291
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.048771713465798,
+      "learning_rate": 1.402064709917634e-05,
+      "loss": 0.9618,
+      "step": 2292
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9071096593912962,
+      "learning_rate": 1.4015635212224719e-05,
+      "loss": 1.0265,
+      "step": 2293
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0330284096807467,
+      "learning_rate": 1.4010622122371219e-05,
+      "loss": 0.6939,
+      "step": 2294
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.946426211140644,
+      "learning_rate": 1.4005607831117533e-05,
+      "loss": 0.8068,
+      "step": 2295
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.727816413368061,
+      "learning_rate": 1.4000592339965712e-05,
+      "loss": 0.5877,
+      "step": 2296
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.001697695671988,
+      "learning_rate": 1.399557565041818e-05,
+      "loss": 1.0784,
+      "step": 2297
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8742839788015815,
+      "learning_rate": 1.3990557763977694e-05,
+      "loss": 0.7321,
+      "step": 2298
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.350236836107517,
+      "learning_rate": 1.3985538682147395e-05,
+      "loss": 0.8941,
+      "step": 2299
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2093530622764064,
+      "learning_rate": 1.3980518406430767e-05,
+      "loss": 0.8822,
+      "step": 2300
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1409914209193965,
+      "learning_rate": 1.3975496938331663e-05,
+      "loss": 1.0937,
+      "step": 2301
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1358174515193984,
+      "learning_rate": 1.3970474279354283e-05,
+      "loss": 1.1211,
+      "step": 2302
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.048233597319192,
+      "learning_rate": 1.3965450431003185e-05,
+      "loss": 0.8974,
+      "step": 2303
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2048883554441316,
+      "learning_rate": 1.3960425394783288e-05,
+      "loss": 0.7434,
+      "step": 2304
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9666347244384621,
+      "learning_rate": 1.3955399172199866e-05,
+      "loss": 0.7991,
+      "step": 2305
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.117348373967525,
+      "learning_rate": 1.3950371764758543e-05,
+      "loss": 0.8921,
+      "step": 2306
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9486043017985043,
+      "learning_rate": 1.3945343173965305e-05,
+      "loss": 1.0818,
+      "step": 2307
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0867802560999806,
+      "learning_rate": 1.3940313401326489e-05,
+      "loss": 0.7863,
+      "step": 2308
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.965665286989394,
+      "learning_rate": 1.3935282448348785e-05,
+      "loss": 0.611,
+      "step": 2309
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0621471614236702,
+      "learning_rate": 1.3930250316539237e-05,
+      "loss": 0.789,
+      "step": 2310
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1090804964812446,
+      "learning_rate": 1.3925217007405246e-05,
+      "loss": 0.7351,
+      "step": 2311
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1033517831623858,
+      "learning_rate": 1.3920182522454564e-05,
+      "loss": 1.1146,
+      "step": 2312
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9356066211802485,
+      "learning_rate": 1.3915146863195292e-05,
+      "loss": 0.9552,
+      "step": 2313
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1755437338975996,
+      "learning_rate": 1.3910110031135884e-05,
+      "loss": 0.8102,
+      "step": 2314
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0545097746441399,
+      "learning_rate": 1.3905072027785142e-05,
+      "loss": 0.9824,
+      "step": 2315
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7621405890943512,
+      "learning_rate": 1.3900032854652232e-05,
+      "loss": 0.6465,
+      "step": 2316
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.799320358447418,
+      "learning_rate": 1.3894992513246655e-05,
+      "loss": 0.5081,
+      "step": 2317
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.006652832102168,
+      "learning_rate": 1.3889951005078271e-05,
+      "loss": 0.8603,
+      "step": 2318
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1003290313154246,
+      "learning_rate": 1.3884908331657284e-05,
+      "loss": 0.787,
+      "step": 2319
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1784159589099745,
+      "learning_rate": 1.3879864494494252e-05,
+      "loss": 0.8025,
+      "step": 2320
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9951257154373354,
+      "learning_rate": 1.3874819495100082e-05,
+      "loss": 0.9868,
+      "step": 2321
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1867792445796337,
+      "learning_rate": 1.3869773334986023e-05,
+      "loss": 0.8461,
+      "step": 2322
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0768486147231655,
+      "learning_rate": 1.3864726015663674e-05,
+      "loss": 0.7298,
+      "step": 2323
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1066075864811915,
+      "learning_rate": 1.3859677538644986e-05,
+      "loss": 0.8139,
+      "step": 2324
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.966986010060322,
+      "learning_rate": 1.3854627905442252e-05,
+      "loss": 1.161,
+      "step": 2325
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0483038256337331,
+      "learning_rate": 1.3849577117568115e-05,
+      "loss": 0.9394,
+      "step": 2326
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.368486186994584,
+      "learning_rate": 1.3844525176535557e-05,
+      "loss": 0.9281,
+      "step": 2327
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8576570860783858,
+      "learning_rate": 1.3839472083857912e-05,
+      "loss": 0.8931,
+      "step": 2328
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2273754046550185,
+      "learning_rate": 1.3834417841048857e-05,
+      "loss": 0.7953,
+      "step": 2329
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1221923509668341,
+      "learning_rate": 1.3829362449622416e-05,
+      "loss": 1.1094,
+      "step": 2330
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0932731806622007,
+      "learning_rate": 1.382430591109295e-05,
+      "loss": 0.8704,
+      "step": 2331
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1211560950855537,
+      "learning_rate": 1.3819248226975173e-05,
+      "loss": 0.985,
+      "step": 2332
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7682341642961471,
+      "learning_rate": 1.3814189398784132e-05,
+      "loss": 0.4701,
+      "step": 2333
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2135269294187718,
+      "learning_rate": 1.3809129428035229e-05,
+      "loss": 0.7057,
+      "step": 2334
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0537802994132592,
+      "learning_rate": 1.3804068316244197e-05,
+      "loss": 1.3379,
+      "step": 2335
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2241649350847161,
+      "learning_rate": 1.3799006064927114e-05,
+      "loss": 0.7305,
+      "step": 2336
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2544875087488703,
+      "learning_rate": 1.3793942675600404e-05,
+      "loss": 0.8592,
+      "step": 2337
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2386504439931423,
+      "learning_rate": 1.3788878149780827e-05,
+      "loss": 0.774,
+      "step": 2338
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9531398441857855,
+      "learning_rate": 1.3783812488985483e-05,
+      "loss": 1.0228,
+      "step": 2339
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0366593705691691,
+      "learning_rate": 1.3778745694731816e-05,
+      "loss": 1.0311,
+      "step": 2340
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.135344958993729,
+      "learning_rate": 1.3773677768537608e-05,
+      "loss": 0.6419,
+      "step": 2341
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2234387581435686,
+      "learning_rate": 1.376860871192098e-05,
+      "loss": 0.6245,
+      "step": 2342
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8181991617659107,
+      "learning_rate": 1.3763538526400384e-05,
+      "loss": 0.7506,
+      "step": 2343
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2204558075502323,
+      "learning_rate": 1.3758467213494626e-05,
+      "loss": 1.2126,
+      "step": 2344
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2020715114018987,
+      "learning_rate": 1.3753394774722835e-05,
+      "loss": 0.8558,
+      "step": 2345
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1564711430238437,
+      "learning_rate": 1.3748321211604488e-05,
+      "loss": 0.9573,
+      "step": 2346
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9362682052469966,
+      "learning_rate": 1.3743246525659389e-05,
+      "loss": 0.7929,
+      "step": 2347
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0855672128070948,
+      "learning_rate": 1.3738170718407689e-05,
+      "loss": 1.004,
+      "step": 2348
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1935789561297154,
+      "learning_rate": 1.3733093791369867e-05,
+      "loss": 0.723,
+      "step": 2349
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8777825429600113,
+      "learning_rate": 1.3728015746066736e-05,
+      "loss": 1.0122,
+      "step": 2350
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2528641657914037,
+      "learning_rate": 1.3722936584019453e-05,
+      "loss": 1.005,
+      "step": 2351
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.828927371926355,
+      "learning_rate": 1.37178563067495e-05,
+      "loss": 0.5154,
+      "step": 2352
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0502307255300734,
+      "learning_rate": 1.3712774915778701e-05,
+      "loss": 0.7668,
+      "step": 2353
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9598941002760166,
+      "learning_rate": 1.370769241262921e-05,
+      "loss": 0.7442,
+      "step": 2354
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9322726327341934,
+      "learning_rate": 1.3702608798823506e-05,
+      "loss": 1.0712,
+      "step": 2355
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.986546518203934,
+      "learning_rate": 1.3697524075884421e-05,
+      "loss": 0.9686,
+      "step": 2356
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0470137355818017,
+      "learning_rate": 1.36924382453351e-05,
+      "loss": 1.0262,
+      "step": 2357
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2729160526880599,
+      "learning_rate": 1.3687351308699026e-05,
+      "loss": 0.8133,
+      "step": 2358
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9390462812326221,
+      "learning_rate": 1.3682263267500016e-05,
+      "loss": 1.0656,
+      "step": 2359
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1007417834388207,
+      "learning_rate": 1.3677174123262216e-05,
+      "loss": 0.9298,
+      "step": 2360
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0740134241157266,
+      "learning_rate": 1.3672083877510105e-05,
+      "loss": 0.7494,
+      "step": 2361
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0117012460030501,
+      "learning_rate": 1.3666992531768482e-05,
+      "loss": 0.7532,
+      "step": 2362
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1346622551740013,
+      "learning_rate": 1.366190008756249e-05,
+      "loss": 0.9397,
+      "step": 2363
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.107499822927669,
+      "learning_rate": 1.3656806546417589e-05,
+      "loss": 1.0787,
+      "step": 2364
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1744230126323616,
+      "learning_rate": 1.3651711909859577e-05,
+      "loss": 1.0503,
+      "step": 2365
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0900290437111593,
+      "learning_rate": 1.3646616179414573e-05,
+      "loss": 0.8331,
+      "step": 2366
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1970943281661348,
+      "learning_rate": 1.3641519356609028e-05,
+      "loss": 0.713,
+      "step": 2367
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1558221195481242,
+      "learning_rate": 1.3636421442969718e-05,
+      "loss": 0.8539,
+      "step": 2368
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9722093394311845,
+      "learning_rate": 1.3631322440023743e-05,
+      "loss": 0.9996,
+      "step": 2369
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.945554766593815,
+      "learning_rate": 1.3626222349298539e-05,
+      "loss": 0.5318,
+      "step": 2370
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9912922592402414,
+      "learning_rate": 1.3621121172321853e-05,
+      "loss": 0.8428,
+      "step": 2371
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.06881181800424,
+      "learning_rate": 1.3616018910621776e-05,
+      "loss": 0.831,
+      "step": 2372
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9965411988406381,
+      "learning_rate": 1.3610915565726709e-05,
+      "loss": 0.9343,
+      "step": 2373
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8067838276183215,
+      "learning_rate": 1.360581113916538e-05,
+      "loss": 0.7774,
+      "step": 2374
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.572604989674843,
+      "learning_rate": 1.3600705632466844e-05,
+      "loss": 0.8959,
+      "step": 2375
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1262299445851258,
+      "learning_rate": 1.359559904716048e-05,
+      "loss": 0.9121,
+      "step": 2376
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.163495035649813,
+      "learning_rate": 1.359049138477599e-05,
+      "loss": 0.8523,
+      "step": 2377
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9121875803954759,
+      "learning_rate": 1.3585382646843396e-05,
+      "loss": 0.9782,
+      "step": 2378
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0956763649118872,
+      "learning_rate": 1.3580272834893047e-05,
+      "loss": 0.7546,
+      "step": 2379
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8810069659156652,
+      "learning_rate": 1.3575161950455604e-05,
+      "loss": 1.0431,
+      "step": 2380
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8664970650589285,
+      "learning_rate": 1.3570049995062063e-05,
+      "loss": 0.7251,
+      "step": 2381
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5010396765999972,
+      "learning_rate": 1.356493697024373e-05,
+      "loss": 0.9449,
+      "step": 2382
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0498458067700138,
+      "learning_rate": 1.3559822877532234e-05,
+      "loss": 1.0571,
+      "step": 2383
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0671943011630098,
+      "learning_rate": 1.3554707718459527e-05,
+      "loss": 0.838,
+      "step": 2384
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1150176445594033,
+      "learning_rate": 1.3549591494557876e-05,
+      "loss": 0.7077,
+      "step": 2385
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9868306861476296,
+      "learning_rate": 1.3544474207359875e-05,
+      "loss": 0.8155,
+      "step": 2386
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9246700853480975,
+      "learning_rate": 1.3539355858398422e-05,
+      "loss": 0.462,
+      "step": 2387
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9776235921380582,
+      "learning_rate": 1.3534236449206748e-05,
+      "loss": 1.0589,
+      "step": 2388
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0893231608681118,
+      "learning_rate": 1.3529115981318395e-05,
+      "loss": 0.9849,
+      "step": 2389
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.275176914516426,
+      "learning_rate": 1.352399445626722e-05,
+      "loss": 0.7322,
+      "step": 2390
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3828376422107462,
+      "learning_rate": 1.3518871875587403e-05,
+      "loss": 0.9235,
+      "step": 2391
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1343987875471058,
+      "learning_rate": 1.3513748240813429e-05,
+      "loss": 0.874,
+      "step": 2392
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2654516301181165,
+      "learning_rate": 1.3508623553480114e-05,
+      "loss": 1.1159,
+      "step": 2393
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2101610616589065,
+      "learning_rate": 1.3503497815122577e-05,
+      "loss": 0.6755,
+      "step": 2394
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4016652838574402,
+      "learning_rate": 1.349837102727626e-05,
+      "loss": 0.9566,
+      "step": 2395
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1858790445564156,
+      "learning_rate": 1.3493243191476915e-05,
+      "loss": 0.9142,
+      "step": 2396
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0488608196106026,
+      "learning_rate": 1.3488114309260603e-05,
+      "loss": 1.1621,
+      "step": 2397
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2544727244507192,
+      "learning_rate": 1.3482984382163713e-05,
+      "loss": 0.8097,
+      "step": 2398
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9969946656639054,
+      "learning_rate": 1.347785341172293e-05,
+      "loss": 1.0102,
+      "step": 2399
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.176997142711561,
+      "learning_rate": 1.3472721399475266e-05,
+      "loss": 0.7704,
+      "step": 2400
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.242836238071932,
+      "learning_rate": 1.3467588346958034e-05,
+      "loss": 0.766,
+      "step": 2401
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0653597972287363,
+      "learning_rate": 1.3462454255708868e-05,
+      "loss": 1.0489,
+      "step": 2402
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9795630918001945,
+      "learning_rate": 1.3457319127265705e-05,
+      "loss": 0.9867,
+      "step": 2403
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0685291448724623,
+      "learning_rate": 1.3452182963166792e-05,
+      "loss": 0.8841,
+      "step": 2404
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8105883666214566,
+      "learning_rate": 1.3447045764950701e-05,
+      "loss": 1.003,
+      "step": 2405
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3310128716065053,
+      "learning_rate": 1.3441907534156298e-05,
+      "loss": 0.8334,
+      "step": 2406
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.014685566283918,
+      "learning_rate": 1.3436768272322764e-05,
+      "loss": 0.6428,
+      "step": 2407
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.101443418489405,
+      "learning_rate": 1.3431627980989586e-05,
+      "loss": 0.7275,
+      "step": 2408
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0681891457912394,
+      "learning_rate": 1.342648666169656e-05,
+      "loss": 0.7425,
+      "step": 2409
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2009557580231673,
+      "learning_rate": 1.34213443159838e-05,
+      "loss": 0.8251,
+      "step": 2410
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.092002261242308,
+      "learning_rate": 1.3416200945391711e-05,
+      "loss": 0.7824,
+      "step": 2411
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.148908215712301,
+      "learning_rate": 1.3411056551461016e-05,
+      "loss": 1.0816,
+      "step": 2412
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0229925169990148,
+      "learning_rate": 1.3405911135732744e-05,
+      "loss": 0.7735,
+      "step": 2413
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9132861094421093,
+      "learning_rate": 1.3400764699748223e-05,
+      "loss": 0.7929,
+      "step": 2414
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8480393187502078,
+      "learning_rate": 1.3395617245049097e-05,
+      "loss": 0.9489,
+      "step": 2415
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8582802095464737,
+      "learning_rate": 1.3390468773177306e-05,
+      "loss": 1.0043,
+      "step": 2416
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0496901074462486,
+      "learning_rate": 1.33853192856751e-05,
+      "loss": 0.8199,
+      "step": 2417
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.127592116550132,
+      "learning_rate": 1.3380168784085028e-05,
+      "loss": 0.804,
+      "step": 2418
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0727101722774453,
+      "learning_rate": 1.337501726994995e-05,
+      "loss": 0.9467,
+      "step": 2419
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5367992517442117,
+      "learning_rate": 1.3369864744813025e-05,
+      "loss": 0.9113,
+      "step": 2420
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.982923932810904,
+      "learning_rate": 1.3364711210217719e-05,
+      "loss": 1.0148,
+      "step": 2421
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0484824277939566,
+      "learning_rate": 1.3359556667707786e-05,
+      "loss": 0.8412,
+      "step": 2422
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.922003251625077,
+      "learning_rate": 1.3354401118827304e-05,
+      "loss": 0.4841,
+      "step": 2423
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4413561492288722,
+      "learning_rate": 1.3349244565120638e-05,
+      "loss": 0.8844,
+      "step": 2424
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0544364147932954,
+      "learning_rate": 1.334408700813245e-05,
+      "loss": 1.0451,
+      "step": 2425
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1434729911666683,
+      "learning_rate": 1.3338928449407721e-05,
+      "loss": 1.0753,
+      "step": 2426
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2161394464436768,
+      "learning_rate": 1.3333768890491714e-05,
+      "loss": 0.8011,
+      "step": 2427
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.415928222893841,
+      "learning_rate": 1.332860833293e-05,
+      "loss": 0.8158,
+      "step": 2428
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7917430858850887,
+      "learning_rate": 1.3323446778268449e-05,
+      "loss": 1.0096,
+      "step": 2429
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9532875454573075,
+      "learning_rate": 1.3318284228053225e-05,
+      "loss": 0.7158,
+      "step": 2430
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2459146943617518,
+      "learning_rate": 1.3313120683830796e-05,
+      "loss": 1.1889,
+      "step": 2431
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0930926112367252,
+      "learning_rate": 1.3307956147147924e-05,
+      "loss": 0.8047,
+      "step": 2432
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8767447257239357,
+      "learning_rate": 1.3302790619551673e-05,
+      "loss": 0.5538,
+      "step": 2433
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.058337881493259,
+      "learning_rate": 1.3297624102589395e-05,
+      "loss": 0.9205,
+      "step": 2434
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2021516688401561,
+      "learning_rate": 1.3292456597808747e-05,
+      "loss": 0.882,
+      "step": 2435
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9912907524375503,
+      "learning_rate": 1.328728810675768e-05,
+      "loss": 1.0368,
+      "step": 2436
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0837113337893411,
+      "learning_rate": 1.3282118630984435e-05,
+      "loss": 0.8152,
+      "step": 2437
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3070441609895542,
+      "learning_rate": 1.3276948172037556e-05,
+      "loss": 0.8566,
+      "step": 2438
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.600269354659413,
+      "learning_rate": 1.3271776731465878e-05,
+      "loss": 0.8403,
+      "step": 2439
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0920557856507571,
+      "learning_rate": 1.3266604310818525e-05,
+      "loss": 1.0084,
+      "step": 2440
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8898178522905503,
+      "learning_rate": 1.3261430911644922e-05,
+      "loss": 0.5383,
+      "step": 2441
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2316717982682117,
+      "learning_rate": 1.3256256535494783e-05,
+      "loss": 0.8328,
+      "step": 2442
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2554150779616,
+      "learning_rate": 1.325108118391812e-05,
+      "loss": 0.995,
+      "step": 2443
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9016390641106911,
+      "learning_rate": 1.3245904858465228e-05,
+      "loss": 0.9876,
+      "step": 2444
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0598910173630378,
+      "learning_rate": 1.3240727560686704e-05,
+      "loss": 1.1251,
+      "step": 2445
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9602932719901173,
+      "learning_rate": 1.3235549292133425e-05,
+      "loss": 0.7308,
+      "step": 2446
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0869195497543822,
+      "learning_rate": 1.3230370054356572e-05,
+      "loss": 0.7623,
+      "step": 2447
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5236266255203061,
+      "learning_rate": 1.3225189848907602e-05,
+      "loss": 0.9481,
+      "step": 2448
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0619448977504335,
+      "learning_rate": 1.322000867733828e-05,
+      "loss": 1.0754,
+      "step": 2449
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2791777086058085,
+      "learning_rate": 1.3214826541200638e-05,
+      "loss": 0.9188,
+      "step": 2450
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2097157402582568,
+      "learning_rate": 1.3209643442047012e-05,
+      "loss": 0.7963,
+      "step": 2451
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0603397462175292,
+      "learning_rate": 1.3204459381430029e-05,
+      "loss": 0.9103,
+      "step": 2452
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9089060094201735,
+      "learning_rate": 1.319927436090259e-05,
+      "loss": 1.007,
+      "step": 2453
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0627119576583373,
+      "learning_rate": 1.3194088382017897e-05,
+      "loss": 1.0126,
+      "step": 2454
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0235212010049226,
+      "learning_rate": 1.3188901446329432e-05,
+      "loss": 0.6827,
+      "step": 2455
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9944864709172332,
+      "learning_rate": 1.3183713555390967e-05,
+      "loss": 0.8386,
+      "step": 2456
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2710332257300236,
+      "learning_rate": 1.3178524710756552e-05,
+      "loss": 0.8483,
+      "step": 2457
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0235919033609346,
+      "learning_rate": 1.3173334913980535e-05,
+      "loss": 0.8681,
+      "step": 2458
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9080940197297802,
+      "learning_rate": 1.3168144166617544e-05,
+      "loss": 0.4597,
+      "step": 2459
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2641584844072953,
+      "learning_rate": 1.3162952470222488e-05,
+      "loss": 1.1465,
+      "step": 2460
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.017455017850308,
+      "learning_rate": 1.3157759826350562e-05,
+      "loss": 0.7355,
+      "step": 2461
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1165045181753583,
+      "learning_rate": 1.3152566236557251e-05,
+      "loss": 0.8718,
+      "step": 2462
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2924805232528611,
+      "learning_rate": 1.3147371702398317e-05,
+      "loss": 0.8512,
+      "step": 2463
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1618701315188256,
+      "learning_rate": 1.3142176225429806e-05,
+      "loss": 1.1213,
+      "step": 2464
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2250007381179908,
+      "learning_rate": 1.313697980720804e-05,
+      "loss": 0.8698,
+      "step": 2465
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0112969675423382,
+      "learning_rate": 1.3131782449289646e-05,
+      "loss": 0.8128,
+      "step": 2466
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0672148879398293,
+      "learning_rate": 1.31265841532315e-05,
+      "loss": 0.9299,
+      "step": 2467
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.984651976269756,
+      "learning_rate": 1.3121384920590785e-05,
+      "loss": 1.0368,
+      "step": 2468
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0188460066200677,
+      "learning_rate": 1.311618475292495e-05,
+      "loss": 1.173,
+      "step": 2469
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8568251227234255,
+      "learning_rate": 1.3110983651791731e-05,
+      "loss": 0.6528,
+      "step": 2470
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0192154882702247,
+      "learning_rate": 1.3105781618749142e-05,
+      "loss": 0.6664,
+      "step": 2471
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.198888679474372,
+      "learning_rate": 1.3100578655355477e-05,
+      "loss": 0.8343,
+      "step": 2472
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0121942713793781,
+      "learning_rate": 1.3095374763169307e-05,
+      "loss": 0.7066,
+      "step": 2473
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0746692002377785,
+      "learning_rate": 1.3090169943749475e-05,
+      "loss": 1.1117,
+      "step": 2474
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.797035402856364,
+      "learning_rate": 1.3084964198655119e-05,
+      "loss": 0.7712,
+      "step": 2475
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.180351126692097,
+      "learning_rate": 1.3079757529445635e-05,
+      "loss": 0.9955,
+      "step": 2476
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0366464230562988,
+      "learning_rate": 1.307454993768071e-05,
+      "loss": 0.7182,
+      "step": 2477
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9197539867860708,
+      "learning_rate": 1.3069341424920301e-05,
+      "loss": 0.8401,
+      "step": 2478
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0364471976741712,
+      "learning_rate": 1.3064131992724638e-05,
+      "loss": 1.1895,
+      "step": 2479
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9256612781214626,
+      "learning_rate": 1.3058921642654235e-05,
+      "loss": 0.701,
+      "step": 2480
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3981999144877324,
+      "learning_rate": 1.3053710376269873e-05,
+      "loss": 0.8552,
+      "step": 2481
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.162936817779193,
+      "learning_rate": 1.3048498195132613e-05,
+      "loss": 0.6972,
+      "step": 2482
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1290093232142477,
+      "learning_rate": 1.3043285100803788e-05,
+      "loss": 0.8163,
+      "step": 2483
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2155534112942876,
+      "learning_rate": 1.3038071094845e-05,
+      "loss": 1.2402,
+      "step": 2484
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7488762751701155,
+      "learning_rate": 1.3032856178818132e-05,
+      "loss": 0.5006,
+      "step": 2485
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0951943791058043,
+      "learning_rate": 1.3027640354285333e-05,
+      "loss": 0.7592,
+      "step": 2486
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0306560453717595,
+      "learning_rate": 1.3022423622809035e-05,
+      "loss": 0.8702,
+      "step": 2487
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2077026688715788,
+      "learning_rate": 1.3017205985951926e-05,
+      "loss": 1.1758,
+      "step": 2488
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0518448784334677,
+      "learning_rate": 1.3011987445276971e-05,
+      "loss": 1.0402,
+      "step": 2489
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9731428352417312,
+      "learning_rate": 1.3006768002347414e-05,
+      "loss": 0.7485,
+      "step": 2490
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1935458122433962,
+      "learning_rate": 1.3001547658726763e-05,
+      "loss": 1.0528,
+      "step": 2491
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2105253274207277,
+      "learning_rate": 1.2996326415978793e-05,
+      "loss": 0.9453,
+      "step": 2492
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1177771643596015,
+      "learning_rate": 1.2991104275667548e-05,
+      "loss": 1.1159,
+      "step": 2493
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2241122921880736,
+      "learning_rate": 1.298588123935735e-05,
+      "loss": 0.7882,
+      "step": 2494
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7429124783500443,
+      "learning_rate": 1.2980657308612778e-05,
+      "loss": 0.5825,
+      "step": 2495
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9105892814115099,
+      "learning_rate": 1.2975432484998689e-05,
+      "loss": 0.7645,
+      "step": 2496
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.104561594285213,
+      "learning_rate": 1.2970206770080202e-05,
+      "loss": 0.8311,
+      "step": 2497
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1031360805514638,
+      "learning_rate": 1.2964980165422701e-05,
+      "loss": 1.035,
+      "step": 2498
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1105562445466146,
+      "learning_rate": 1.2959752672591843e-05,
+      "loss": 0.7551,
+      "step": 2499
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.146745067615904,
+      "learning_rate": 1.2954524293153546e-05,
+      "loss": 0.8281,
+      "step": 2500
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1326451941460347,
+      "learning_rate": 1.2949295028673991e-05,
+      "loss": 0.7592,
+      "step": 2501
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7603777113568362,
+      "learning_rate": 1.2944064880719634e-05,
+      "loss": 1.0572,
+      "step": 2502
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9477426829346517,
+      "learning_rate": 1.2938833850857187e-05,
+      "loss": 0.6103,
+      "step": 2503
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9994132868737089,
+      "learning_rate": 1.2933601940653629e-05,
+      "loss": 0.8427,
+      "step": 2504
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5951240441133752,
+      "learning_rate": 1.2928369151676203e-05,
+      "loss": 0.8878,
+      "step": 2505
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.028445431762651,
+      "learning_rate": 1.2923135485492413e-05,
+      "loss": 0.7894,
+      "step": 2506
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1750522456349182,
+      "learning_rate": 1.2917900943670032e-05,
+      "loss": 1.204,
+      "step": 2507
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0244897496518606,
+      "learning_rate": 1.2912665527777088e-05,
+      "loss": 0.7918,
+      "step": 2508
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1611494527806967,
+      "learning_rate": 1.2907429239381872e-05,
+      "loss": 0.8833,
+      "step": 2509
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9498477075249322,
+      "learning_rate": 1.290219208005294e-05,
+      "loss": 0.9696,
+      "step": 2510
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2783599056170858,
+      "learning_rate": 1.289695405135911e-05,
+      "loss": 0.8138,
+      "step": 2511
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2321460158847517,
+      "learning_rate": 1.2891715154869449e-05,
+      "loss": 1.1591,
+      "step": 2512
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1769300327690326,
+      "learning_rate": 1.2886475392153304e-05,
+      "loss": 0.7668,
+      "step": 2513
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6336976803855792,
+      "learning_rate": 1.2881234764780257e-05,
+      "loss": 0.4501,
+      "step": 2514
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9913320113644941,
+      "learning_rate": 1.2875993274320173e-05,
+      "loss": 0.7584,
+      "step": 2515
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1989869643323645,
+      "learning_rate": 1.287075092234316e-05,
+      "loss": 1.0013,
+      "step": 2516
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9905636884602843,
+      "learning_rate": 1.2865507710419587e-05,
+      "loss": 1.1038,
+      "step": 2517
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1968264585793635,
+      "learning_rate": 1.2860263640120086e-05,
+      "loss": 0.7612,
+      "step": 2518
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0495752034859684,
+      "learning_rate": 1.2855018713015537e-05,
+      "loss": 0.8686,
+      "step": 2519
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0605871199989272,
+      "learning_rate": 1.2849772930677087e-05,
+      "loss": 0.7885,
+      "step": 2520
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.182998401738766,
+      "learning_rate": 1.2844526294676132e-05,
+      "loss": 1.2309,
+      "step": 2521
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0539593676321135,
+      "learning_rate": 1.2839278806584323e-05,
+      "loss": 0.7068,
+      "step": 2522
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1692046962149936,
+      "learning_rate": 1.2834030467973572e-05,
+      "loss": 0.7686,
+      "step": 2523
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1526022314921687,
+      "learning_rate": 1.2828781280416045e-05,
+      "loss": 0.8148,
+      "step": 2524
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.950721000757134,
+      "learning_rate": 1.2823531245484158e-05,
+      "loss": 0.7724,
+      "step": 2525
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2864028378490677,
+      "learning_rate": 1.2818280364750577e-05,
+      "loss": 1.1948,
+      "step": 2526
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7686666514179153,
+      "learning_rate": 1.2813028639788238e-05,
+      "loss": 1.0122,
+      "step": 2527
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0344604993056998,
+      "learning_rate": 1.280777607217031e-05,
+      "loss": 0.7409,
+      "step": 2528
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0406641993715937,
+      "learning_rate": 1.2802522663470232e-05,
+      "loss": 0.9339,
+      "step": 2529
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1348833900321873,
+      "learning_rate": 1.2797268415261681e-05,
+      "loss": 0.6865,
+      "step": 2530
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.058082445427784,
+      "learning_rate": 1.2792013329118591e-05,
+      "loss": 1.1352,
+      "step": 2531
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6545867271123844,
+      "learning_rate": 1.2786757406615151e-05,
+      "loss": 0.4825,
+      "step": 2532
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0832105305398119,
+      "learning_rate": 1.278150064932579e-05,
+      "loss": 1.187,
+      "step": 2533
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9640690427147913,
+      "learning_rate": 1.2776243058825199e-05,
+      "loss": 0.7211,
+      "step": 2534
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1155749687778111,
+      "learning_rate": 1.2770984636688313e-05,
+      "loss": 0.9086,
+      "step": 2535
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9691173224815903,
+      "learning_rate": 1.2765725384490315e-05,
+      "loss": 0.9906,
+      "step": 2536
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3290384325871651,
+      "learning_rate": 1.2760465303806639e-05,
+      "loss": 0.8369,
+      "step": 2537
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.41020948806047,
+      "learning_rate": 1.2755204396212965e-05,
+      "loss": 0.9236,
+      "step": 2538
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2243493870347357,
+      "learning_rate": 1.2749942663285224e-05,
+      "loss": 0.9685,
+      "step": 2539
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0072426494167097,
+      "learning_rate": 1.274468010659959e-05,
+      "loss": 1.0911,
+      "step": 2540
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.04129369651786,
+      "learning_rate": 1.2739416727732486e-05,
+      "loss": 0.8381,
+      "step": 2541
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3126099867177856,
+      "learning_rate": 1.273415252826058e-05,
+      "loss": 0.823,
+      "step": 2542
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.21297906116195,
+      "learning_rate": 1.272888750976079e-05,
+      "loss": 1.0426,
+      "step": 2543
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.029423501666679,
+      "learning_rate": 1.2723621673810277e-05,
+      "loss": 0.7486,
+      "step": 2544
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1252364641753776,
+      "learning_rate": 1.271835502198644e-05,
+      "loss": 1.1031,
+      "step": 2545
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.13218989240336,
+      "learning_rate": 1.2713087555866937e-05,
+      "loss": 0.8436,
+      "step": 2546
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1049671573797533,
+      "learning_rate": 1.2707819277029653e-05,
+      "loss": 0.8422,
+      "step": 2547
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2018774589370138,
+      "learning_rate": 1.2702550187052733e-05,
+      "loss": 0.7147,
+      "step": 2548
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9877160529162076,
+      "learning_rate": 1.2697280287514553e-05,
+      "loss": 0.7463,
+      "step": 2549
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9575190752681612,
+      "learning_rate": 1.2692009579993736e-05,
+      "loss": 1.0449,
+      "step": 2550
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7221308504441456,
+      "learning_rate": 1.2686738066069148e-05,
+      "loss": 0.4366,
+      "step": 2551
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0135189998615348,
+      "learning_rate": 1.268146574731989e-05,
+      "loss": 1.0411,
+      "step": 2552
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.027032061725929,
+      "learning_rate": 1.2676192625325316e-05,
+      "loss": 0.9711,
+      "step": 2553
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9466079608705167,
+      "learning_rate": 1.2670918701665007e-05,
+      "loss": 0.6679,
+      "step": 2554
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0516691772554827,
+      "learning_rate": 1.2665643977918798e-05,
+      "loss": 1.2404,
+      "step": 2555
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1091907006579318,
+      "learning_rate": 1.2660368455666752e-05,
+      "loss": 0.7597,
+      "step": 2556
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.27582864508528,
+      "learning_rate": 1.2655092136489178e-05,
+      "loss": 0.8785,
+      "step": 2557
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1899964585083256,
+      "learning_rate": 1.264981502196662e-05,
+      "loss": 0.7345,
+      "step": 2558
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8868890861302019,
+      "learning_rate": 1.264453711367986e-05,
+      "loss": 0.8315,
+      "step": 2559
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5864043188123322,
+      "learning_rate": 1.2639258413209922e-05,
+      "loss": 0.7178,
+      "step": 2560
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0879882791085067,
+      "learning_rate": 1.2633978922138066e-05,
+      "loss": 0.8782,
+      "step": 2561
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.034931068325704,
+      "learning_rate": 1.2628698642045786e-05,
+      "loss": 0.9308,
+      "step": 2562
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9879521570859656,
+      "learning_rate": 1.2623417574514816e-05,
+      "loss": 0.6436,
+      "step": 2563
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0446797444297022,
+      "learning_rate": 1.2618135721127118e-05,
+      "loss": 1.0847,
+      "step": 2564
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9317966315924658,
+      "learning_rate": 1.2612853083464903e-05,
+      "loss": 0.8049,
+      "step": 2565
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.650880979327094,
+      "learning_rate": 1.2607569663110603e-05,
+      "loss": 0.8932,
+      "step": 2566
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.078649733009228,
+      "learning_rate": 1.2602285461646897e-05,
+      "loss": 0.9269,
+      "step": 2567
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1356280582346732,
+      "learning_rate": 1.2597000480656684e-05,
+      "loss": 0.7893,
+      "step": 2568
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8316174683865881,
+      "learning_rate": 1.2591714721723112e-05,
+      "loss": 0.6809,
+      "step": 2569
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1430417517384526,
+      "learning_rate": 1.258642818642955e-05,
+      "loss": 0.7076,
+      "step": 2570
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.428204845156178,
+      "learning_rate": 1.2581140876359603e-05,
+      "loss": 0.902,
+      "step": 2571
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9874472798493404,
+      "learning_rate": 1.2575852793097113e-05,
+      "loss": 0.8316,
+      "step": 2572
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0854667400806417,
+      "learning_rate": 1.2570563938226144e-05,
+      "loss": 0.672,
+      "step": 2573
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9608620111175015,
+      "learning_rate": 1.2565274313331001e-05,
+      "loss": 1.2611,
+      "step": 2574
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.118892892877426,
+      "learning_rate": 1.2559983919996213e-05,
+      "loss": 0.7088,
+      "step": 2575
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1308319867096155,
+      "learning_rate": 1.2554692759806541e-05,
+      "loss": 0.7715,
+      "step": 2576
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.028072807633239,
+      "learning_rate": 1.2549400834346981e-05,
+      "loss": 0.8188,
+      "step": 2577
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1396912684785772,
+      "learning_rate": 1.2544108145202748e-05,
+      "loss": 1.0695,
+      "step": 2578
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9380781457902949,
+      "learning_rate": 1.2538814693959294e-05,
+      "loss": 0.7602,
+      "step": 2579
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0739267235875878,
+      "learning_rate": 1.2533520482202293e-05,
+      "loss": 0.8051,
+      "step": 2580
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0634693973056009,
+      "learning_rate": 1.2528225511517655e-05,
+      "loss": 0.8972,
+      "step": 2581
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0040841931213123,
+      "learning_rate": 1.2522929783491508e-05,
+      "loss": 0.8667,
+      "step": 2582
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.151348520251059,
+      "learning_rate": 1.2517633299710218e-05,
+      "loss": 1.0037,
+      "step": 2583
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3819641039333865,
+      "learning_rate": 1.2512336061760368e-05,
+      "loss": 0.7934,
+      "step": 2584
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8287049794266833,
+      "learning_rate": 1.2507038071228767e-05,
+      "loss": 0.7157,
+      "step": 2585
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8315767516162178,
+      "learning_rate": 1.2501739329702453e-05,
+      "loss": 0.8546,
+      "step": 2586
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0897731252566847,
+      "learning_rate": 1.2496439838768692e-05,
+      "loss": 0.8692,
+      "step": 2587
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0022166951183173,
+      "learning_rate": 1.2491139600014968e-05,
+      "loss": 1.0693,
+      "step": 2588
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1195292249456898,
+      "learning_rate": 1.248583861502899e-05,
+      "loss": 0.7463,
+      "step": 2589
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3822157109491429,
+      "learning_rate": 1.2480536885398697e-05,
+      "loss": 0.961,
+      "step": 2590
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0718632508459827,
+      "learning_rate": 1.2475234412712245e-05,
+      "loss": 0.7698,
+      "step": 2591
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1246251202516808,
+      "learning_rate": 1.2469931198558008e-05,
+      "loss": 1.0029,
+      "step": 2592
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1850015984878015,
+      "learning_rate": 1.2464627244524595e-05,
+      "loss": 0.9021,
+      "step": 2593
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0719853153836627,
+      "learning_rate": 1.2459322552200824e-05,
+      "loss": 0.9904,
+      "step": 2594
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9022406021456886,
+      "learning_rate": 1.2454017123175746e-05,
+      "loss": 0.7487,
+      "step": 2595
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2168352320588927,
+      "learning_rate": 1.2448710959038623e-05,
+      "loss": 0.8366,
+      "step": 2596
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.301955723995397,
+      "learning_rate": 1.2443404061378941e-05,
+      "loss": 1.2184,
+      "step": 2597
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0750825539704645,
+      "learning_rate": 1.2438096431786408e-05,
+      "loss": 1.0006,
+      "step": 2598
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0968943223396794,
+      "learning_rate": 1.2432788071850943e-05,
+      "loss": 0.8004,
+      "step": 2599
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1170672448824706,
+      "learning_rate": 1.2427478983162694e-05,
+      "loss": 0.9211,
+      "step": 2600
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1542842986494173,
+      "learning_rate": 1.2422169167312022e-05,
+      "loss": 0.982,
+      "step": 2601
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1619778188023173,
+      "learning_rate": 1.2416858625889507e-05,
+      "loss": 0.7112,
+      "step": 2602
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0410367875383417,
+      "learning_rate": 1.2411547360485945e-05,
+      "loss": 1.0566,
+      "step": 2603
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.622806485891046,
+      "learning_rate": 1.240623537269235e-05,
+      "loss": 0.4945,
+      "step": 2604
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4420234706719486,
+      "learning_rate": 1.2400922664099952e-05,
+      "loss": 0.8125,
+      "step": 2605
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.002299175488173,
+      "learning_rate": 1.2395609236300198e-05,
+      "loss": 0.8511,
+      "step": 2606
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4416134121913526,
+      "learning_rate": 1.2390295090884752e-05,
+      "loss": 0.8817,
+      "step": 2607
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9130618466752416,
+      "learning_rate": 1.2384980229445482e-05,
+      "loss": 0.8218,
+      "step": 2608
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9296030871200094,
+      "learning_rate": 1.2379664653574487e-05,
+      "loss": 0.7051,
+      "step": 2609
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.077656243667484,
+      "learning_rate": 1.2374348364864068e-05,
+      "loss": 0.9989,
+      "step": 2610
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1990006704337126,
+      "learning_rate": 1.2369031364906747e-05,
+      "loss": 0.7794,
+      "step": 2611
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1039463817621242,
+      "learning_rate": 1.2363713655295254e-05,
+      "loss": 1.1192,
+      "step": 2612
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0696939036222155,
+      "learning_rate": 1.2358395237622533e-05,
+      "loss": 0.8839,
+      "step": 2613
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.081321256931908,
+      "learning_rate": 1.2353076113481742e-05,
+      "loss": 0.7683,
+      "step": 2614
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2790479005652258,
+      "learning_rate": 1.2347756284466244e-05,
+      "loss": 0.8355,
+      "step": 2615
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6905617680946305,
+      "learning_rate": 1.2342435752169624e-05,
+      "loss": 1.1511,
+      "step": 2616
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.264170255255076,
+      "learning_rate": 1.2337114518185669e-05,
+      "loss": 0.7418,
+      "step": 2617
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3366726855552933,
+      "learning_rate": 1.2331792584108375e-05,
+      "loss": 0.7063,
+      "step": 2618
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2291522496859724,
+      "learning_rate": 1.232646995153196e-05,
+      "loss": 1.198,
+      "step": 2619
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0987566908960966,
+      "learning_rate": 1.2321146622050838e-05,
+      "loss": 0.8469,
+      "step": 2620
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8495922745827327,
+      "learning_rate": 1.2315822597259636e-05,
+      "loss": 0.4849,
+      "step": 2621
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0596941798501867,
+      "learning_rate": 1.2310497878753196e-05,
+      "loss": 1.0842,
+      "step": 2622
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4272532169568701,
+      "learning_rate": 1.2305172468126554e-05,
+      "loss": 0.8385,
+      "step": 2623
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1545598904378225,
+      "learning_rate": 1.2299846366974964e-05,
+      "loss": 0.9425,
+      "step": 2624
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1492373428053322,
+      "learning_rate": 1.2294519576893884e-05,
+      "loss": 0.6585,
+      "step": 2625
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0092607451957565,
+      "learning_rate": 1.2289192099478979e-05,
+      "loss": 1.0702,
+      "step": 2626
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2629817330995678,
+      "learning_rate": 1.2283863936326117e-05,
+      "loss": 0.7979,
+      "step": 2627
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0484054379693288,
+      "learning_rate": 1.2278535089031377e-05,
+      "loss": 0.6871,
+      "step": 2628
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.243276400703934,
+      "learning_rate": 1.2273205559191036e-05,
+      "loss": 0.7904,
+      "step": 2629
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.306103333475098,
+      "learning_rate": 1.2267875348401585e-05,
+      "loss": 0.9134,
+      "step": 2630
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.060975130623509,
+      "learning_rate": 1.2262544458259707e-05,
+      "loss": 1.1234,
+      "step": 2631
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9413313387569122,
+      "learning_rate": 1.2257212890362296e-05,
+      "loss": 0.842,
+      "step": 2632
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4703461611216175,
+      "learning_rate": 1.2251880646306446e-05,
+      "loss": 0.9752,
+      "step": 2633
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1232349453990178,
+      "learning_rate": 1.224654772768946e-05,
+      "loss": 0.9201,
+      "step": 2634
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9325870165926851,
+      "learning_rate": 1.2241214136108834e-05,
+      "loss": 0.6809,
+      "step": 2635
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.109802710221545,
+      "learning_rate": 1.2235879873162269e-05,
+      "loss": 1.0456,
+      "step": 2636
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1134091933219792,
+      "learning_rate": 1.2230544940447672e-05,
+      "loss": 0.8276,
+      "step": 2637
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.463037919641249,
+      "learning_rate": 1.2225209339563144e-05,
+      "loss": 0.8169,
+      "step": 2638
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.110024410153955,
+      "learning_rate": 1.221987307210699e-05,
+      "loss": 0.7726,
+      "step": 2639
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1354308178948287,
+      "learning_rate": 1.2214536139677712e-05,
+      "loss": 0.8003,
+      "step": 2640
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8369456236708489,
+      "learning_rate": 1.220919854387401e-05,
+      "loss": 0.6775,
+      "step": 2641
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.026641620152888,
+      "learning_rate": 1.220386028629479e-05,
+      "loss": 1.109,
+      "step": 2642
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0506755361888114,
+      "learning_rate": 1.2198521368539149e-05,
+      "loss": 0.7939,
+      "step": 2643
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0936387746880902,
+      "learning_rate": 1.2193181792206385e-05,
+      "loss": 0.9834,
+      "step": 2644
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4323987565786833,
+      "learning_rate": 1.2187841558895992e-05,
+      "loss": 0.8249,
+      "step": 2645
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0844209371303615,
+      "learning_rate": 1.218250067020766e-05,
+      "loss": 1.0766,
+      "step": 2646
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4683312748306985,
+      "learning_rate": 1.2177159127741278e-05,
+      "loss": 0.9947,
+      "step": 2647
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3105063709142017,
+      "learning_rate": 1.2171816933096929e-05,
+      "loss": 0.8935,
+      "step": 2648
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.583435669644515,
+      "learning_rate": 1.2166474087874893e-05,
+      "loss": 0.7915,
+      "step": 2649
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9608437343109619,
+      "learning_rate": 1.2161130593675643e-05,
+      "loss": 1.0076,
+      "step": 2650
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5307591533519318,
+      "learning_rate": 1.2155786452099848e-05,
+      "loss": 0.9692,
+      "step": 2651
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9260755756468453,
+      "learning_rate": 1.2150441664748369e-05,
+      "loss": 0.6942,
+      "step": 2652
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9835638235950657,
+      "learning_rate": 1.214509623322226e-05,
+      "loss": 0.7543,
+      "step": 2653
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.389867682306587,
+      "learning_rate": 1.2139750159122771e-05,
+      "loss": 1.0125,
+      "step": 2654
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.125486004400711,
+      "learning_rate": 1.2134403444051344e-05,
+      "loss": 1.0159,
+      "step": 2655
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.101780732442571,
+      "learning_rate": 1.212905608960961e-05,
+      "loss": 0.9783,
+      "step": 2656
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0647707147825296,
+      "learning_rate": 1.2123708097399396e-05,
+      "loss": 0.9862,
+      "step": 2657
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.824406305831393,
+      "learning_rate": 1.2118359469022714e-05,
+      "loss": 0.4932,
+      "step": 2658
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.134148847778216,
+      "learning_rate": 1.2113010206081774e-05,
+      "loss": 0.8012,
+      "step": 2659
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0778340462403309,
+      "learning_rate": 1.2107660310178966e-05,
+      "loss": 0.9878,
+      "step": 2660
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.310714949589154,
+      "learning_rate": 1.2102309782916884e-05,
+      "loss": 0.8447,
+      "step": 2661
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1563888153285415,
+      "learning_rate": 1.2096958625898294e-05,
+      "loss": 0.8087,
+      "step": 2662
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0361601048121154,
+      "learning_rate": 1.209160684072617e-05,
+      "loss": 1.0851,
+      "step": 2663
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.29930340686397,
+      "learning_rate": 1.2086254429003654e-05,
+      "loss": 0.7558,
+      "step": 2664
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0792653981029425,
+      "learning_rate": 1.208090139233409e-05,
+      "loss": 0.987,
+      "step": 2665
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1980408415851413,
+      "learning_rate": 1.2075547732321005e-05,
+      "loss": 0.8917,
+      "step": 2666
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1591494875740527,
+      "learning_rate": 1.207019345056811e-05,
+      "loss": 0.8706,
+      "step": 2667
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8272953998470999,
+      "learning_rate": 1.2064838548679307e-05,
+      "loss": 0.6687,
+      "step": 2668
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0473109577765207,
+      "learning_rate": 1.2059483028258678e-05,
+      "loss": 0.9961,
+      "step": 2669
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0551548983402672,
+      "learning_rate": 1.2054126890910499e-05,
+      "loss": 0.863,
+      "step": 2670
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.390671204526419,
+      "learning_rate": 1.2048770138239221e-05,
+      "loss": 0.8869,
+      "step": 2671
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1318090185287035,
+      "learning_rate": 1.2043412771849483e-05,
+      "loss": 0.8329,
+      "step": 2672
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1029993584074698,
+      "learning_rate": 1.2038054793346115e-05,
+      "loss": 0.8539,
+      "step": 2673
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.858420967817044,
+      "learning_rate": 1.2032696204334115e-05,
+      "loss": 1.0454,
+      "step": 2674
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8575127050348699,
+      "learning_rate": 1.202733700641868e-05,
+      "loss": 0.467,
+      "step": 2675
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1200858610349025,
+      "learning_rate": 1.2021977201205177e-05,
+      "loss": 0.6701,
+      "step": 2676
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.79936480135501,
+      "learning_rate": 1.2016616790299165e-05,
+      "loss": 0.8419,
+      "step": 2677
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2048420037169312,
+      "learning_rate": 1.2011255775306378e-05,
+      "loss": 0.8216,
+      "step": 2678
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2181354166194158,
+      "learning_rate": 1.200589415783273e-05,
+      "loss": 1.1427,
+      "step": 2679
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0948690088211213,
+      "learning_rate": 1.2000531939484321e-05,
+      "loss": 0.8389,
+      "step": 2680
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1723687258198032,
+      "learning_rate": 1.1995169121867423e-05,
+      "loss": 0.9588,
+      "step": 2681
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1714226171037896,
+      "learning_rate": 1.1989805706588502e-05,
+      "loss": 0.9724,
+      "step": 2682
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9930480858019216,
+      "learning_rate": 1.1984441695254183e-05,
+      "loss": 0.761,
+      "step": 2683
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0787765477725555,
+      "learning_rate": 1.1979077089471288e-05,
+      "loss": 1.0354,
+      "step": 2684
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0336516740629256,
+      "learning_rate": 1.1973711890846806e-05,
+      "loss": 1.1311,
+      "step": 2685
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.178843587375283,
+      "learning_rate": 1.1968346100987906e-05,
+      "loss": 0.786,
+      "step": 2686
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0560811299961348,
+      "learning_rate": 1.1962979721501937e-05,
+      "loss": 0.9625,
+      "step": 2687
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.047566088799034,
+      "learning_rate": 1.1957612753996419e-05,
+      "loss": 0.769,
+      "step": 2688
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3610374802201854,
+      "learning_rate": 1.1952245200079055e-05,
+      "loss": 0.978,
+      "step": 2689
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9885132221835409,
+      "learning_rate": 1.1946877061357723e-05,
+      "loss": 0.6973,
+      "step": 2690
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0454261320142704,
+      "learning_rate": 1.1941508339440467e-05,
+      "loss": 0.7566,
+      "step": 2691
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3333316042779462,
+      "learning_rate": 1.1936139035935514e-05,
+      "loss": 0.9822,
+      "step": 2692
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.229071547778378,
+      "learning_rate": 1.1930769152451265e-05,
+      "loss": 0.7408,
+      "step": 2693
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9816378440907818,
+      "learning_rate": 1.1925398690596293e-05,
+      "loss": 0.7196,
+      "step": 2694
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.017626079436611,
+      "learning_rate": 1.1920027651979342e-05,
+      "loss": 0.8727,
+      "step": 2695
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2444700429782947,
+      "learning_rate": 1.1914656038209335e-05,
+      "loss": 0.7881,
+      "step": 2696
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.249420686950696,
+      "learning_rate": 1.1909283850895361e-05,
+      "loss": 0.8212,
+      "step": 2697
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1484258047734472,
+      "learning_rate": 1.1903911091646684e-05,
+      "loss": 1.0116,
+      "step": 2698
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2387607039017907,
+      "learning_rate": 1.189853776207274e-05,
+      "loss": 0.8159,
+      "step": 2699
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0497528209450897,
+      "learning_rate": 1.1893163863783131e-05,
+      "loss": 0.7142,
+      "step": 2700
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3263582654483703,
+      "learning_rate": 1.1887789398387635e-05,
+      "loss": 0.9867,
+      "step": 2701
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0419526682438225,
+      "learning_rate": 1.1882414367496196e-05,
+      "loss": 0.682,
+      "step": 2702
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0717143484830791,
+      "learning_rate": 1.1877038772718932e-05,
+      "loss": 1.0935,
+      "step": 2703
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3579269470596627,
+      "learning_rate": 1.1871662615666124e-05,
+      "loss": 0.7838,
+      "step": 2704
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9357939568230975,
+      "learning_rate": 1.1866285897948227e-05,
+      "loss": 1.1195,
+      "step": 2705
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4702425469220683,
+      "learning_rate": 1.186090862117586e-05,
+      "loss": 1.0138,
+      "step": 2706
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.977011564135698,
+      "learning_rate": 1.185553078695981e-05,
+      "loss": 0.9892,
+      "step": 2707
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2457989998346044,
+      "learning_rate": 1.1850152396911034e-05,
+      "loss": 0.9637,
+      "step": 2708
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0446421785670181,
+      "learning_rate": 1.1844773452640653e-05,
+      "loss": 0.832,
+      "step": 2709
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.220660454015529,
+      "learning_rate": 1.183939395575995e-05,
+      "loss": 0.668,
+      "step": 2710
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1412331744919693,
+      "learning_rate": 1.1834013907880382e-05,
+      "loss": 0.6628,
+      "step": 2711
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0280152318749214,
+      "learning_rate": 1.1828633310613569e-05,
+      "loss": 1.0052,
+      "step": 2712
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.21565051466424,
+      "learning_rate": 1.1823252165571288e-05,
+      "loss": 0.8882,
+      "step": 2713
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9014842923999913,
+      "learning_rate": 1.181787047436549e-05,
+      "loss": 0.5691,
+      "step": 2714
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3864027421942278,
+      "learning_rate": 1.1812488238608285e-05,
+      "loss": 0.87,
+      "step": 2715
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7166381693358246,
+      "learning_rate": 1.180710545991194e-05,
+      "loss": 0.5403,
+      "step": 2716
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9097838185720896,
+      "learning_rate": 1.1801722139888898e-05,
+      "loss": 0.9928,
+      "step": 2717
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.199611058181084,
+      "learning_rate": 1.1796338280151756e-05,
+      "loss": 0.8997,
+      "step": 2718
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1705166303305923,
+      "learning_rate": 1.1790953882313271e-05,
+      "loss": 0.8836,
+      "step": 2719
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9630242959115808,
+      "learning_rate": 1.1785568947986368e-05,
+      "loss": 0.6637,
+      "step": 2720
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1079782751610783,
+      "learning_rate": 1.1780183478784125e-05,
+      "loss": 0.9115,
+      "step": 2721
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.080462895863772,
+      "learning_rate": 1.1774797476319785e-05,
+      "loss": 1.1063,
+      "step": 2722
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.092503995371074,
+      "learning_rate": 1.1769410942206752e-05,
+      "loss": 0.7929,
+      "step": 2723
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0321640900856859,
+      "learning_rate": 1.1764023878058584e-05,
+      "loss": 0.845,
+      "step": 2724
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1461309555279282,
+      "learning_rate": 1.1758636285489e-05,
+      "loss": 0.9781,
+      "step": 2725
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0376212822379316,
+      "learning_rate": 1.175324816611188e-05,
+      "loss": 0.6409,
+      "step": 2726
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9553735709645002,
+      "learning_rate": 1.1747859521541258e-05,
+      "loss": 1.1065,
+      "step": 2727
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6151500550506164,
+      "learning_rate": 1.1742470353391329e-05,
+      "loss": 0.8131,
+      "step": 2728
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0700587426774233,
+      "learning_rate": 1.1737080663276444e-05,
+      "loss": 1.0363,
+      "step": 2729
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0470795652989253,
+      "learning_rate": 1.1731690452811107e-05,
+      "loss": 0.783,
+      "step": 2730
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5979453087771611,
+      "learning_rate": 1.172629972360998e-05,
+      "loss": 0.4651,
+      "step": 2731
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0756459440380404,
+      "learning_rate": 1.1720908477287882e-05,
+      "loss": 0.9968,
+      "step": 2732
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2795103185390961,
+      "learning_rate": 1.1715516715459784e-05,
+      "loss": 0.8218,
+      "step": 2733
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1758295612495198,
+      "learning_rate": 1.1710124439740816e-05,
+      "loss": 0.8618,
+      "step": 2734
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0930857873379354,
+      "learning_rate": 1.1704731651746253e-05,
+      "loss": 0.6791,
+      "step": 2735
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9912892568052407,
+      "learning_rate": 1.1699338353091536e-05,
+      "loss": 0.991,
+      "step": 2736
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3530161090081323,
+      "learning_rate": 1.1693944545392247e-05,
+      "loss": 1.0201,
+      "step": 2737
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1443220974392578,
+      "learning_rate": 1.1688550230264128e-05,
+      "loss": 0.7618,
+      "step": 2738
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.179951416905143,
+      "learning_rate": 1.1683155409323073e-05,
+      "loss": 0.7233,
+      "step": 2739
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.986566283085013,
+      "learning_rate": 1.1677760084185123e-05,
+      "loss": 0.8206,
+      "step": 2740
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0524568676128925,
+      "learning_rate": 1.167236425646647e-05,
+      "loss": 1.0609,
+      "step": 2741
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.049220455007771,
+      "learning_rate": 1.1666967927783461e-05,
+      "loss": 0.7215,
+      "step": 2742
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0671212978904097,
+      "learning_rate": 1.1661571099752593e-05,
+      "loss": 0.7973,
+      "step": 2743
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0912191214146856,
+      "learning_rate": 1.1656173773990508e-05,
+      "loss": 0.7636,
+      "step": 2744
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.922529936967694,
+      "learning_rate": 1.1650775952114e-05,
+      "loss": 0.6792,
+      "step": 2745
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3621335051064871,
+      "learning_rate": 1.164537763574001e-05,
+      "loss": 1.2002,
+      "step": 2746
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1266741806940217,
+      "learning_rate": 1.1639978826485629e-05,
+      "loss": 0.8611,
+      "step": 2747
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5938825099415602,
+      "learning_rate": 1.1634579525968097e-05,
+      "loss": 0.4697,
+      "step": 2748
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1121744350096512,
+      "learning_rate": 1.1629179735804797e-05,
+      "loss": 0.9993,
+      "step": 2749
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9481582337935898,
+      "learning_rate": 1.1623779457613263e-05,
+      "loss": 0.9995,
+      "step": 2750
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1969199841958316,
+      "learning_rate": 1.161837869301117e-05,
+      "loss": 0.969,
+      "step": 2751
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3103726135818299,
+      "learning_rate": 1.1612977443616344e-05,
+      "loss": 0.8094,
+      "step": 2752
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3576437833656698,
+      "learning_rate": 1.1607575711046755e-05,
+      "loss": 0.8919,
+      "step": 2753
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.951426871232172,
+      "learning_rate": 1.160217349692051e-05,
+      "loss": 0.7382,
+      "step": 2754
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0356944369330732,
+      "learning_rate": 1.1596770802855876e-05,
+      "loss": 1.1259,
+      "step": 2755
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3233722641375467,
+      "learning_rate": 1.1591367630471251e-05,
+      "loss": 0.9196,
+      "step": 2756
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8596976765747764,
+      "learning_rate": 1.1585963981385179e-05,
+      "loss": 0.7729,
+      "step": 2757
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1556141763071843,
+      "learning_rate": 1.1580559857216347e-05,
+      "loss": 0.8693,
+      "step": 2758
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0499038589227734,
+      "learning_rate": 1.1575155259583588e-05,
+      "loss": 0.918,
+      "step": 2759
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.26775844083333,
+      "learning_rate": 1.1569750190105871e-05,
+      "loss": 1.1598,
+      "step": 2760
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2828413880084266,
+      "learning_rate": 1.156434465040231e-05,
+      "loss": 0.779,
+      "step": 2761
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9164072872300493,
+      "learning_rate": 1.1558938642092161e-05,
+      "loss": 0.8112,
+      "step": 2762
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1669719280296542,
+      "learning_rate": 1.1553532166794817e-05,
+      "loss": 0.8743,
+      "step": 2763
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9946879913403184,
+      "learning_rate": 1.1548125226129811e-05,
+      "loss": 0.6961,
+      "step": 2764
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1371714842404697,
+      "learning_rate": 1.154271782171682e-05,
+      "loss": 1.1287,
+      "step": 2765
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7229160819393315,
+      "learning_rate": 1.1537309955175655e-05,
+      "loss": 0.4565,
+      "step": 2766
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5657879381150366,
+      "learning_rate": 1.1531901628126269e-05,
+      "loss": 0.9712,
+      "step": 2767
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9664847686112104,
+      "learning_rate": 1.1526492842188746e-05,
+      "loss": 0.6313,
+      "step": 2768
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9669485068684778,
+      "learning_rate": 1.1521083598983319e-05,
+      "loss": 0.7998,
+      "step": 2769
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1157084117626854,
+      "learning_rate": 1.1515673900130344e-05,
+      "loss": 1.3716,
+      "step": 2770
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8494017403140413,
+      "learning_rate": 1.1510263747250327e-05,
+      "loss": 0.836,
+      "step": 2771
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0357020931482406,
+      "learning_rate": 1.1504853141963905e-05,
+      "loss": 0.8864,
+      "step": 2772
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0980437428316359,
+      "learning_rate": 1.1499442085891848e-05,
+      "loss": 0.7719,
+      "step": 2773
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.060725502998131,
+      "learning_rate": 1.149403058065506e-05,
+      "loss": 1.1324,
+      "step": 2774
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9961122699778693,
+      "learning_rate": 1.1488618627874581e-05,
+      "loss": 0.6988,
+      "step": 2775
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8958468019480861,
+      "learning_rate": 1.1483206229171591e-05,
+      "loss": 0.9346,
+      "step": 2776
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0252130630472764,
+      "learning_rate": 1.1477793386167396e-05,
+      "loss": 0.9008,
+      "step": 2777
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0615833666870857,
+      "learning_rate": 1.1472380100483438e-05,
+      "loss": 1.0238,
+      "step": 2778
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0589290449792517,
+      "learning_rate": 1.1466966373741293e-05,
+      "loss": 1.1098,
+      "step": 2779
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1310635892164513,
+      "learning_rate": 1.1461552207562665e-05,
+      "loss": 0.7239,
+      "step": 2780
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8232755535171501,
+      "learning_rate": 1.1456137603569395e-05,
+      "loss": 0.5267,
+      "step": 2781
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1899381847096506,
+      "learning_rate": 1.145072256338345e-05,
+      "loss": 0.9506,
+      "step": 2782
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9449681881914331,
+      "learning_rate": 1.1445307088626934e-05,
+      "loss": 0.7446,
+      "step": 2783
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0068244110318745,
+      "learning_rate": 1.143989118092207e-05,
+      "loss": 0.7036,
+      "step": 2784
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1191415542717869,
+      "learning_rate": 1.1434474841891226e-05,
+      "loss": 0.9616,
+      "step": 2785
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5028625098130455,
+      "learning_rate": 1.1429058073156886e-05,
+      "loss": 0.8091,
+      "step": 2786
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8943159319848496,
+      "learning_rate": 1.142364087634167e-05,
+      "loss": 0.8128,
+      "step": 2787
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.220114791365755,
+      "learning_rate": 1.1418223253068321e-05,
+      "loss": 0.7721,
+      "step": 2788
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9335505185493178,
+      "learning_rate": 1.1412805204959719e-05,
+      "loss": 0.9731,
+      "step": 2789
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0803616733531414,
+      "learning_rate": 1.140738673363886e-05,
+      "loss": 0.8325,
+      "step": 2790
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0888264834964294,
+      "learning_rate": 1.1401967840728869e-05,
+      "loss": 0.7845,
+      "step": 2791
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1922942410897182,
+      "learning_rate": 1.1396548527853006e-05,
+      "loss": 0.7901,
+      "step": 2792
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9318172285145401,
+      "learning_rate": 1.1391128796634651e-05,
+      "loss": 1.2347,
+      "step": 2793
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.017615828975046,
+      "learning_rate": 1.1385708648697306e-05,
+      "loss": 0.586,
+      "step": 2794
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0037693640694418,
+      "learning_rate": 1.1380288085664603e-05,
+      "loss": 0.8234,
+      "step": 2795
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.104723600205277,
+      "learning_rate": 1.1374867109160295e-05,
+      "loss": 0.9121,
+      "step": 2796
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3595983785811798,
+      "learning_rate": 1.1369445720808263e-05,
+      "loss": 0.9653,
+      "step": 2797
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8810280069493139,
+      "learning_rate": 1.1364023922232503e-05,
+      "loss": 1.0684,
+      "step": 2798
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9769637705475837,
+      "learning_rate": 1.1358601715057146e-05,
+      "loss": 0.6683,
+      "step": 2799
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3814549016542785,
+      "learning_rate": 1.1353179100906438e-05,
+      "loss": 0.8976,
+      "step": 2800
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0772747364001187,
+      "learning_rate": 1.1347756081404743e-05,
+      "loss": 0.9427,
+      "step": 2801
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7688678150347613,
+      "learning_rate": 1.1342332658176556e-05,
+      "loss": 0.4655,
+      "step": 2802
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1200301808275366,
+      "learning_rate": 1.1336908832846485e-05,
+      "loss": 1.1639,
+      "step": 2803
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9689364427874324,
+      "learning_rate": 1.1331484607039265e-05,
+      "loss": 0.7301,
+      "step": 2804
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1977676925772338,
+      "learning_rate": 1.1326059982379747e-05,
+      "loss": 0.8699,
+      "step": 2805
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9599059768702483,
+      "learning_rate": 1.13206349604929e-05,
+      "loss": 0.5488,
+      "step": 2806
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3868779810384095,
+      "learning_rate": 1.1315209543003813e-05,
+      "loss": 0.8967,
+      "step": 2807
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.027999811082285,
+      "learning_rate": 1.13097837315377e-05,
+      "loss": 0.9731,
+      "step": 2808
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.145379800920336,
+      "learning_rate": 1.1304357527719882e-05,
+      "loss": 0.7956,
+      "step": 2809
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3517327618841384,
+      "learning_rate": 1.1298930933175805e-05,
+      "loss": 1.0606,
+      "step": 2810
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.103472660969264,
+      "learning_rate": 1.1293503949531033e-05,
+      "loss": 0.7972,
+      "step": 2811
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5977744537520981,
+      "learning_rate": 1.128807657841124e-05,
+      "loss": 0.9203,
+      "step": 2812
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9291764508427045,
+      "learning_rate": 1.128264882144222e-05,
+      "loss": 0.9787,
+      "step": 2813
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1673150421345915,
+      "learning_rate": 1.1277220680249887e-05,
+      "loss": 0.7917,
+      "step": 2814
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1169372657158079,
+      "learning_rate": 1.127179215646026e-05,
+      "loss": 0.9114,
+      "step": 2815
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9504357903241955,
+      "learning_rate": 1.126636325169948e-05,
+      "loss": 1.0186,
+      "step": 2816
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0799451605999424,
+      "learning_rate": 1.12609339675938e-05,
+      "loss": 0.944,
+      "step": 2817
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.057411551515635,
+      "learning_rate": 1.1255504305769589e-05,
+      "loss": 0.7193,
+      "step": 2818
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9816161198304737,
+      "learning_rate": 1.1250074267853326e-05,
+      "loss": 0.6734,
+      "step": 2819
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8819177788890109,
+      "learning_rate": 1.1244643855471603e-05,
+      "loss": 0.4914,
+      "step": 2820
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9788541361605699,
+      "learning_rate": 1.1239213070251123e-05,
+      "loss": 0.6992,
+      "step": 2821
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1188950345598934,
+      "learning_rate": 1.1233781913818707e-05,
+      "loss": 1.2009,
+      "step": 2822
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8582977453388867,
+      "learning_rate": 1.1228350387801278e-05,
+      "loss": 1.0408,
+      "step": 2823
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0343042855964493,
+      "learning_rate": 1.1222918493825876e-05,
+      "loss": 0.8075,
+      "step": 2824
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0079235676122658,
+      "learning_rate": 1.1217486233519652e-05,
+      "loss": 0.6712,
+      "step": 2825
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1937661157453767,
+      "learning_rate": 1.1212053608509862e-05,
+      "loss": 0.7581,
+      "step": 2826
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.39045083506427,
+      "learning_rate": 1.1206620620423873e-05,
+      "loss": 1.2601,
+      "step": 2827
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9843383287850074,
+      "learning_rate": 1.1201187270889166e-05,
+      "loss": 0.7531,
+      "step": 2828
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.495320279330635,
+      "learning_rate": 1.119575356153332e-05,
+      "loss": 0.9637,
+      "step": 2829
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0957000139022879,
+      "learning_rate": 1.119031949398403e-05,
+      "loss": 0.8128,
+      "step": 2830
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0835853277712488,
+      "learning_rate": 1.1184885069869096e-05,
+      "loss": 0.8138,
+      "step": 2831
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1591984701859996,
+      "learning_rate": 1.1179450290816425e-05,
+      "loss": 0.9823,
+      "step": 2832
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0325850274225767,
+      "learning_rate": 1.1174015158454028e-05,
+      "loss": 0.7267,
+      "step": 2833
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.245145658244032,
+      "learning_rate": 1.1168579674410026e-05,
+      "loss": 0.8267,
+      "step": 2834
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.952269196434607,
+      "learning_rate": 1.116314384031264e-05,
+      "loss": 0.9486,
+      "step": 2835
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2213041562773115,
+      "learning_rate": 1.1157707657790202e-05,
+      "loss": 0.6613,
+      "step": 2836
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8698898031480226,
+      "learning_rate": 1.1152271128471146e-05,
+      "loss": 1.1743,
+      "step": 2837
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1781429813657658,
+      "learning_rate": 1.1146834253984008e-05,
+      "loss": 0.8177,
+      "step": 2838
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7774859632806079,
+      "learning_rate": 1.1141397035957423e-05,
+      "loss": 0.5771,
+      "step": 2839
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1290010929017995,
+      "learning_rate": 1.1135959476020144e-05,
+      "loss": 0.4453,
+      "step": 2840
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.097738598737898,
+      "learning_rate": 1.1130521575801012e-05,
+      "loss": 0.9669,
+      "step": 2841
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.180927964993672,
+      "learning_rate": 1.1125083336928974e-05,
+      "loss": 0.8119,
+      "step": 2842
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1617020597237002,
+      "learning_rate": 1.1119644761033079e-05,
+      "loss": 0.8851,
+      "step": 2843
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0614013048264743,
+      "learning_rate": 1.1114205849742483e-05,
+      "loss": 0.7396,
+      "step": 2844
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0426616115717313,
+      "learning_rate": 1.110876660468643e-05,
+      "loss": 0.7704,
+      "step": 2845
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2274665524236583,
+      "learning_rate": 1.1103327027494272e-05,
+      "loss": 1.167,
+      "step": 2846
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1542620334764366,
+      "learning_rate": 1.1097887119795465e-05,
+      "loss": 0.8274,
+      "step": 2847
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.71970378172764,
+      "learning_rate": 1.109244688321955e-05,
+      "loss": 1.0323,
+      "step": 2848
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9500222108925983,
+      "learning_rate": 1.108700631939618e-05,
+      "loss": 0.895,
+      "step": 2849
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8806468286926925,
+      "learning_rate": 1.10815654299551e-05,
+      "loss": 0.7535,
+      "step": 2850
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0830345997495026,
+      "learning_rate": 1.1076124216526153e-05,
+      "loss": 1.0105,
+      "step": 2851
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.254516303723901,
+      "learning_rate": 1.1070682680739275e-05,
+      "loss": 0.6982,
+      "step": 2852
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1477681147842211,
+      "learning_rate": 1.106524082422451e-05,
+      "loss": 0.8919,
+      "step": 2853
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.054955171929497,
+      "learning_rate": 1.1059798648611988e-05,
+      "loss": 0.7605,
+      "step": 2854
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0306335714692996,
+      "learning_rate": 1.1054356155531936e-05,
+      "loss": 0.8078,
+      "step": 2855
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1365532241629384,
+      "learning_rate": 1.104891334661468e-05,
+      "loss": 0.7474,
+      "step": 2856
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1745909819744214,
+      "learning_rate": 1.1043470223490637e-05,
+      "loss": 0.7939,
+      "step": 2857
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9832390008830442,
+      "learning_rate": 1.103802678779032e-05,
+      "loss": 0.6962,
+      "step": 2858
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7178778842215268,
+      "learning_rate": 1.1032583041144334e-05,
+      "loss": 0.8368,
+      "step": 2859
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8944871801470905,
+      "learning_rate": 1.1027138985183381e-05,
+      "loss": 1.2288,
+      "step": 2860
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1597395556253856,
+      "learning_rate": 1.102169462153825e-05,
+      "loss": 0.8863,
+      "step": 2861
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9932499615779145,
+      "learning_rate": 1.1016249951839824e-05,
+      "loss": 0.8107,
+      "step": 2862
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1006163895733891,
+      "learning_rate": 1.1010804977719083e-05,
+      "loss": 0.7773,
+      "step": 2863
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.791271579719152,
+      "learning_rate": 1.1005359700807087e-05,
+      "loss": 0.6211,
+      "step": 2864
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.961090315214024,
+      "learning_rate": 1.0999914122735002e-05,
+      "loss": 1.0173,
+      "step": 2865
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1775247298476497,
+      "learning_rate": 1.0994468245134071e-05,
+      "loss": 0.7954,
+      "step": 2866
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0663936000295395,
+      "learning_rate": 1.0989022069635633e-05,
+      "loss": 0.8861,
+      "step": 2867
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1925215720433615,
+      "learning_rate": 1.0983575597871114e-05,
+      "loss": 0.9027,
+      "step": 2868
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0416018621316867,
+      "learning_rate": 1.0978128831472028e-05,
+      "loss": 0.6715,
+      "step": 2869
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9157987258333429,
+      "learning_rate": 1.0972681772069981e-05,
+      "loss": 0.9877,
+      "step": 2870
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0705580067985445,
+      "learning_rate": 1.0967234421296665e-05,
+      "loss": 0.8239,
+      "step": 2871
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9566105370183072,
+      "learning_rate": 1.0961786780783857e-05,
+      "loss": 0.9949,
+      "step": 2872
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9209165281862706,
+      "learning_rate": 1.0956338852163424e-05,
+      "loss": 1.0128,
+      "step": 2873
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7500089179264862,
+      "learning_rate": 1.0950890637067317e-05,
+      "loss": 0.5447,
+      "step": 2874
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8568992234095224,
+      "learning_rate": 1.0945442137127575e-05,
+      "loss": 0.9916,
+      "step": 2875
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9465317658885664,
+      "learning_rate": 1.0939993353976317e-05,
+      "loss": 0.7157,
+      "step": 2876
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1654924165299165,
+      "learning_rate": 1.0934544289245757e-05,
+      "loss": 0.7059,
+      "step": 2877
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9407051675822525,
+      "learning_rate": 1.0929094944568182e-05,
+      "loss": 0.8574,
+      "step": 2878
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0591472584475659,
+      "learning_rate": 1.092364532157597e-05,
+      "loss": 1.0994,
+      "step": 2879
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.035572183985758,
+      "learning_rate": 1.0918195421901583e-05,
+      "loss": 0.8319,
+      "step": 2880
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9335128982393012,
+      "learning_rate": 1.091274524717756e-05,
+      "loss": 0.9899,
+      "step": 2881
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8051630926051112,
+      "learning_rate": 1.0907294799036525e-05,
+      "loss": 0.4368,
+      "step": 2882
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2341693194926526,
+      "learning_rate": 1.0901844079111188e-05,
+      "loss": 0.9838,
+      "step": 2883
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0991773814441361,
+      "learning_rate": 1.0896393089034336e-05,
+      "loss": 1.1479,
+      "step": 2884
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3676419407392975,
+      "learning_rate": 1.0890941830438836e-05,
+      "loss": 0.7985,
+      "step": 2885
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0626628947994006,
+      "learning_rate": 1.0885490304957642e-05,
+      "loss": 0.9887,
+      "step": 2886
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0977080222969244,
+      "learning_rate": 1.0880038514223778e-05,
+      "loss": 0.7863,
+      "step": 2887
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9153148394847314,
+      "learning_rate": 1.0874586459870356e-05,
+      "loss": 0.8999,
+      "step": 2888
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0562960233557348,
+      "learning_rate": 1.0869134143530562e-05,
+      "loss": 1.0207,
+      "step": 2889
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4395567300713148,
+      "learning_rate": 1.0863681566837666e-05,
+      "loss": 0.9248,
+      "step": 2890
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8909739227034521,
+      "learning_rate": 1.0858228731425009e-05,
+      "loss": 0.7559,
+      "step": 2891
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1273042512366207,
+      "learning_rate": 1.0852775638926009e-05,
+      "loss": 0.8298,
+      "step": 2892
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8167708767236186,
+      "learning_rate": 1.0847322290974173e-05,
+      "loss": 0.6083,
+      "step": 2893
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.029785536125864,
+      "learning_rate": 1.0841868689203072e-05,
+      "loss": 0.8712,
+      "step": 2894
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0271728683849237,
+      "learning_rate": 1.0836414835246357e-05,
+      "loss": 0.8376,
+      "step": 2895
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7530230253743024,
+      "learning_rate": 1.0830960730737754e-05,
+      "loss": 1.0083,
+      "step": 2896
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5006821616166017,
+      "learning_rate": 1.0825506377311068e-05,
+      "loss": 0.7497,
+      "step": 2897
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9330562110987329,
+      "learning_rate": 1.0820051776600175e-05,
+      "loss": 1.013,
+      "step": 2898
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9169642523386389,
+      "learning_rate": 1.081459693023902e-05,
+      "loss": 0.6833,
+      "step": 2899
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1578825029789956,
+      "learning_rate": 1.080914183986164e-05,
+      "loss": 0.9866,
+      "step": 2900
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0294094896792398,
+      "learning_rate": 1.0803686507102118e-05,
+      "loss": 0.8166,
+      "step": 2901
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9489042518183514,
+      "learning_rate": 1.0798230933594629e-05,
+      "loss": 1.0597,
+      "step": 2902
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0195027096621927,
+      "learning_rate": 1.079277512097342e-05,
+      "loss": 1.0325,
+      "step": 2903
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9379154682292953,
+      "learning_rate": 1.0787319070872796e-05,
+      "loss": 0.7113,
+      "step": 2904
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9538204160214734,
+      "learning_rate": 1.078186278492715e-05,
+      "loss": 0.8337,
+      "step": 2905
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2053919695550392,
+      "learning_rate": 1.0776406264770933e-05,
+      "loss": 0.8627,
+      "step": 2906
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1970413244806852,
+      "learning_rate": 1.077094951203867e-05,
+      "loss": 0.8697,
+      "step": 2907
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9737622098715195,
+      "learning_rate": 1.076549252836496e-05,
+      "loss": 1.0203,
+      "step": 2908
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2375903699236763,
+      "learning_rate": 1.076003531538446e-05,
+      "loss": 0.9147,
+      "step": 2909
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0163073063258874,
+      "learning_rate": 1.0754577874731911e-05,
+      "loss": 0.6911,
+      "step": 2910
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6847290014091909,
+      "learning_rate": 1.074912020804211e-05,
+      "loss": 0.4797,
+      "step": 2911
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1532245553683904,
+      "learning_rate": 1.0743662316949928e-05,
+      "loss": 0.9902,
+      "step": 2912
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9211986316754851,
+      "learning_rate": 1.0738204203090299e-05,
+      "loss": 0.9154,
+      "step": 2913
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0571275328359329,
+      "learning_rate": 1.073274586809823e-05,
+      "loss": 0.8192,
+      "step": 2914
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0479231102730344,
+      "learning_rate": 1.0727287313608783e-05,
+      "loss": 0.7436,
+      "step": 2915
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1380320686496201,
+      "learning_rate": 1.07218285412571e-05,
+      "loss": 0.8005,
+      "step": 2916
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8895021700570118,
+      "learning_rate": 1.0716369552678378e-05,
+      "loss": 0.6464,
+      "step": 2917
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.418478235657351,
+      "learning_rate": 1.0710910349507881e-05,
+      "loss": 1.1848,
+      "step": 2918
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9373621479131442,
+      "learning_rate": 1.0705450933380945e-05,
+      "loss": 0.9764,
+      "step": 2919
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.141906189036489,
+      "learning_rate": 1.0699991305932955e-05,
+      "loss": 0.8427,
+      "step": 2920
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0497484935891213,
+      "learning_rate": 1.0694531468799369e-05,
+      "loss": 0.7773,
+      "step": 2921
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9686201564299335,
+      "learning_rate": 1.0689071423615708e-05,
+      "loss": 1.0567,
+      "step": 2922
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2705199367953495,
+      "learning_rate": 1.0683611172017556e-05,
+      "loss": 1.0767,
+      "step": 2923
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0128837100136687,
+      "learning_rate": 1.0678150715640551e-05,
+      "loss": 0.9367,
+      "step": 2924
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1187842692539542,
+      "learning_rate": 1.0672690056120398e-05,
+      "loss": 0.9277,
+      "step": 2925
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3468623407997975,
+      "learning_rate": 1.0667229195092868e-05,
+      "loss": 0.8694,
+      "step": 2926
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9857486076354768,
+      "learning_rate": 1.0661768134193782e-05,
+      "loss": 1.0648,
+      "step": 2927
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7355504991442282,
+      "learning_rate": 1.0656306875059024e-05,
+      "loss": 0.4825,
+      "step": 2928
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.202651607540652,
+      "learning_rate": 1.0650845419324544e-05,
+      "loss": 0.8469,
+      "step": 2929
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0646472659691262,
+      "learning_rate": 1.0645383768626345e-05,
+      "loss": 0.8772,
+      "step": 2930
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.436506648736376,
+      "learning_rate": 1.063992192460049e-05,
+      "loss": 0.7997,
+      "step": 2931
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8782372351865182,
+      "learning_rate": 1.0634459888883093e-05,
+      "loss": 0.9412,
+      "step": 2932
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9913872683674809,
+      "learning_rate": 1.062899766311034e-05,
+      "loss": 0.6364,
+      "step": 2933
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.099144616270185,
+      "learning_rate": 1.0623535248918463e-05,
+      "loss": 0.7868,
+      "step": 2934
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1935766892148816,
+      "learning_rate": 1.0618072647943751e-05,
+      "loss": 0.9286,
+      "step": 2935
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9366072038821059,
+      "learning_rate": 1.0612609861822556e-05,
+      "loss": 1.0591,
+      "step": 2936
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.188630618873653,
+      "learning_rate": 1.0607146892191273e-05,
+      "loss": 0.7231,
+      "step": 2937
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1802596876495661,
+      "learning_rate": 1.0601683740686366e-05,
+      "loss": 1.005,
+      "step": 2938
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.22226360343023,
+      "learning_rate": 1.0596220408944344e-05,
+      "loss": 0.7882,
+      "step": 2939
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6762286753125455,
+      "learning_rate": 1.0590756898601775e-05,
+      "loss": 0.7797,
+      "step": 2940
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9679268527417183,
+      "learning_rate": 1.0585293211295275e-05,
+      "loss": 0.8566,
+      "step": 2941
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0497399983222726,
+      "learning_rate": 1.0579829348661522e-05,
+      "loss": 1.0099,
+      "step": 2942
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0715003711094007,
+      "learning_rate": 1.0574365312337235e-05,
+      "loss": 0.7068,
+      "step": 2943
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1188189364326158,
+      "learning_rate": 1.0568901103959192e-05,
+      "loss": 1.0949,
+      "step": 2944
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2719879427222434,
+      "learning_rate": 1.0563436725164227e-05,
+      "loss": 0.8094,
+      "step": 2945
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9146745280700425,
+      "learning_rate": 1.0557972177589214e-05,
+      "loss": 0.8738,
+      "step": 2946
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1139821750934664,
+      "learning_rate": 1.0552507462871087e-05,
+      "loss": 0.6965,
+      "step": 2947
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1103989695142686,
+      "learning_rate": 1.0547042582646823e-05,
+      "loss": 0.8396,
+      "step": 2948
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9815619821359586,
+      "learning_rate": 1.0541577538553454e-05,
+      "loss": 0.7403,
+      "step": 2949
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9115486068412448,
+      "learning_rate": 1.0536112332228057e-05,
+      "loss": 0.8521,
+      "step": 2950
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8835522879549714,
+      "learning_rate": 1.053064696530776e-05,
+      "loss": 1.0683,
+      "step": 2951
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9262756546084153,
+      "learning_rate": 1.0525181439429741e-05,
+      "loss": 0.5785,
+      "step": 2952
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0999861545947398,
+      "learning_rate": 1.0519715756231222e-05,
+      "loss": 0.8509,
+      "step": 2953
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9921353028236358,
+      "learning_rate": 1.0514249917349473e-05,
+      "loss": 0.7824,
+      "step": 2954
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0709444103275638,
+      "learning_rate": 1.0508783924421813e-05,
+      "loss": 0.8154,
+      "step": 2955
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1181978414792193,
+      "learning_rate": 1.05033177790856e-05,
+      "loss": 1.142,
+      "step": 2956
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0605354402796754,
+      "learning_rate": 1.0497851482978247e-05,
+      "loss": 0.7699,
+      "step": 2957
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.768391759469647,
+      "learning_rate": 1.0492385037737207e-05,
+      "loss": 0.4519,
+      "step": 2958
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0733637947942514,
+      "learning_rate": 1.048691844499998e-05,
+      "loss": 0.5922,
+      "step": 2959
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.939273432156306,
+      "learning_rate": 1.0481451706404104e-05,
+      "loss": 0.7636,
+      "step": 2960
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.032626856601362,
+      "learning_rate": 1.0475984823587172e-05,
+      "loss": 1.069,
+      "step": 2961
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.109502423931907,
+      "learning_rate": 1.0470517798186813e-05,
+      "loss": 0.8223,
+      "step": 2962
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9969764793698811,
+      "learning_rate": 1.0465050631840692e-05,
+      "loss": 0.933,
+      "step": 2963
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9810055064881982,
+      "learning_rate": 1.0459583326186532e-05,
+      "loss": 0.8681,
+      "step": 2964
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8776473766898455,
+      "learning_rate": 1.0454115882862083e-05,
+      "loss": 0.9614,
+      "step": 2965
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7124834782131445,
+      "learning_rate": 1.044864830350515e-05,
+      "loss": 0.5594,
+      "step": 2966
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.083437809945808,
+      "learning_rate": 1.0443180589753565e-05,
+      "loss": 0.7423,
+      "step": 2967
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2221756415354357,
+      "learning_rate": 1.043771274324521e-05,
+      "loss": 0.8576,
+      "step": 2968
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9275189639080664,
+      "learning_rate": 1.0432244765618002e-05,
+      "loss": 1.0327,
+      "step": 2969
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.172682014520686,
+      "learning_rate": 1.0426776658509899e-05,
+      "loss": 1.1034,
+      "step": 2970
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9297252992507246,
+      "learning_rate": 1.0421308423558898e-05,
+      "loss": 1.0198,
+      "step": 2971
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8039961916649252,
+      "learning_rate": 1.0415840062403035e-05,
+      "loss": 0.9739,
+      "step": 2972
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0334514211927048,
+      "learning_rate": 1.0410371576680378e-05,
+      "loss": 0.888,
+      "step": 2973
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.038394139054068,
+      "learning_rate": 1.040490296802904e-05,
+      "loss": 0.6264,
+      "step": 2974
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1405782752943614,
+      "learning_rate": 1.0399434238087169e-05,
+      "loss": 1.0938,
+      "step": 2975
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9437944695666984,
+      "learning_rate": 1.0393965388492945e-05,
+      "loss": 0.8074,
+      "step": 2976
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0618862633890453,
+      "learning_rate": 1.0388496420884588e-05,
+      "loss": 0.9329,
+      "step": 2977
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4886362605735126,
+      "learning_rate": 1.0383027336900356e-05,
+      "loss": 0.8238,
+      "step": 2978
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8302862045279561,
+      "learning_rate": 1.0377558138178532e-05,
+      "loss": 0.76,
+      "step": 2979
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0551825152875238,
+      "learning_rate": 1.0372088826357443e-05,
+      "loss": 1.1635,
+      "step": 2980
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0357782520288323,
+      "learning_rate": 1.0366619403075445e-05,
+      "loss": 0.75,
+      "step": 2981
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6737212603190357,
+      "learning_rate": 1.0361149869970929e-05,
+      "loss": 0.5019,
+      "step": 2982
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0521397012796392,
+      "learning_rate": 1.035568022868232e-05,
+      "loss": 0.7232,
+      "step": 2983
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.016751940450322,
+      "learning_rate": 1.035021048084807e-05,
+      "loss": 1.0587,
+      "step": 2984
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6434590524995616,
+      "learning_rate": 1.0344740628106673e-05,
+      "loss": 0.8724,
+      "step": 2985
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9114311100511474,
+      "learning_rate": 1.0339270672096645e-05,
+      "loss": 0.7944,
+      "step": 2986
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0743161552125893,
+      "learning_rate": 1.0333800614456537e-05,
+      "loss": 0.8456,
+      "step": 2987
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0742759839168508,
+      "learning_rate": 1.0328330456824929e-05,
+      "loss": 0.7543,
+      "step": 2988
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.947285489952527,
+      "learning_rate": 1.032286020084043e-05,
+      "loss": 1.0944,
+      "step": 2989
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0228352715689155,
+      "learning_rate": 1.0317389848141685e-05,
+      "loss": 1.1639,
+      "step": 2990
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.953887164370641,
+      "learning_rate": 1.031191940036736e-05,
+      "loss": 0.7689,
+      "step": 2991
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9337714947090777,
+      "learning_rate": 1.0306448859156155e-05,
+      "loss": 1.0661,
+      "step": 2992
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9049660035164021,
+      "learning_rate": 1.0300978226146789e-05,
+      "loss": 0.7625,
+      "step": 2993
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9499677933016032,
+      "learning_rate": 1.0295507502978024e-05,
+      "loss": 1.0775,
+      "step": 2994
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1352029178740888,
+      "learning_rate": 1.0290036691288637e-05,
+      "loss": 0.8812,
+      "step": 2995
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.075631879872595,
+      "learning_rate": 1.028456579271743e-05,
+      "loss": 0.7719,
+      "step": 2996
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.764635570744649,
+      "learning_rate": 1.0279094808903239e-05,
+      "loss": 0.9364,
+      "step": 2997
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9419354099991012,
+      "learning_rate": 1.0273623741484924e-05,
+      "loss": 0.7988,
+      "step": 2998
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0626619740382544,
+      "learning_rate": 1.0268152592101366e-05,
+      "loss": 1.0903,
+      "step": 2999
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3549523943647173,
+      "learning_rate": 1.0262681362391473e-05,
+      "loss": 0.7831,
+      "step": 3000
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6368338538799673,
+      "learning_rate": 1.025721005399418e-05,
+      "loss": 0.4589,
+      "step": 3001
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.475897532003622,
+      "learning_rate": 1.0251738668548439e-05,
+      "loss": 0.9521,
+      "step": 3002
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8198702159653035,
+      "learning_rate": 1.0246267207693226e-05,
+      "loss": 0.6235,
+      "step": 3003
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1689215288403627,
+      "learning_rate": 1.024079567306755e-05,
+      "loss": 1.2023,
+      "step": 3004
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7650615126466583,
+      "learning_rate": 1.0235324066310426e-05,
+      "loss": 0.6428,
+      "step": 3005
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.095842440960517,
+      "learning_rate": 1.0229852389060905e-05,
+      "loss": 0.7597,
+      "step": 3006
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.403058543097702,
+      "learning_rate": 1.0224380642958052e-05,
+      "loss": 0.735,
+      "step": 3007
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2712827691792856,
+      "learning_rate": 1.021890882964095e-05,
+      "loss": 1.1998,
+      "step": 3008
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1168141043435853,
+      "learning_rate": 1.0213436950748708e-05,
+      "loss": 0.8149,
+      "step": 3009
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0765370151787652,
+      "learning_rate": 1.0207965007920448e-05,
+      "loss": 0.9792,
+      "step": 3010
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.190753992325439,
+      "learning_rate": 1.0202493002795324e-05,
+      "loss": 0.7974,
+      "step": 3011
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9731149724081554,
+      "learning_rate": 1.0197020937012494e-05,
+      "loss": 0.7619,
+      "step": 3012
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0996734711443117,
+      "learning_rate": 1.0191548812211143e-05,
+      "loss": 1.0799,
+      "step": 3013
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1669494759888839,
+      "learning_rate": 1.0186076630030469e-05,
+      "loss": 0.8911,
+      "step": 3014
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.133849901156018,
+      "learning_rate": 1.018060439210969e-05,
+      "loss": 0.8176,
+      "step": 3015
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9961637585430222,
+      "learning_rate": 1.0175132100088042e-05,
+      "loss": 0.8906,
+      "step": 3016
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9463883945314431,
+      "learning_rate": 1.0169659755604768e-05,
+      "loss": 0.8279,
+      "step": 3017
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1130688447953916,
+      "learning_rate": 1.0164187360299143e-05,
+      "loss": 1.1525,
+      "step": 3018
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.808744323814658,
+      "learning_rate": 1.0158714915810442e-05,
+      "loss": 0.4807,
+      "step": 3019
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.05875743813217,
+      "learning_rate": 1.0153242423777964e-05,
+      "loss": 0.9823,
+      "step": 3020
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9522876472742291,
+      "learning_rate": 1.0147769885841017e-05,
+      "loss": 0.6456,
+      "step": 3021
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0512342211031227,
+      "learning_rate": 1.0142297303638926e-05,
+      "loss": 0.7285,
+      "step": 3022
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1238815637513657,
+      "learning_rate": 1.0136824678811028e-05,
+      "loss": 1.212,
+      "step": 3023
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.21662634839534,
+      "learning_rate": 1.0131352012996672e-05,
+      "loss": 0.8931,
+      "step": 3024
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1479505834159003,
+      "learning_rate": 1.012587930783522e-05,
+      "loss": 0.8795,
+      "step": 3025
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9325920784704512,
+      "learning_rate": 1.0120406564966043e-05,
+      "loss": 0.8497,
+      "step": 3026
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9522174337268206,
+      "learning_rate": 1.0114933786028534e-05,
+      "loss": 1.1221,
+      "step": 3027
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1263791499195894,
+      "learning_rate": 1.0109460972662081e-05,
+      "loss": 0.882,
+      "step": 3028
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0979356272273153,
+      "learning_rate": 1.0103988126506093e-05,
+      "loss": 0.7522,
+      "step": 3029
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.295820872698275,
+      "learning_rate": 1.0098515249199988e-05,
+      "loss": 0.6954,
+      "step": 3030
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9911786776829233,
+      "learning_rate": 1.0093042342383189e-05,
+      "loss": 0.854,
+      "step": 3031
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9011267742595014,
+      "learning_rate": 1.008756940769513e-05,
+      "loss": 0.8905,
+      "step": 3032
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0804237764113955,
+      "learning_rate": 1.0082096446775255e-05,
+      "loss": 1.0041,
+      "step": 3033
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.906032785555113,
+      "learning_rate": 1.0076623461263017e-05,
+      "loss": 0.7022,
+      "step": 3034
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8789170911576032,
+      "learning_rate": 1.0071150452797872e-05,
+      "loss": 0.6043,
+      "step": 3035
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.404819159517774,
+      "learning_rate": 1.0065677423019278e-05,
+      "loss": 0.6739,
+      "step": 3036
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9920232035213546,
+      "learning_rate": 1.0060204373566715e-05,
+      "loss": 1.0783,
+      "step": 3037
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8270500297892344,
+      "learning_rate": 1.0054731306079657e-05,
+      "loss": 0.7521,
+      "step": 3038
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0497508187695084,
+      "learning_rate": 1.0049258222197586e-05,
+      "loss": 0.8647,
+      "step": 3039
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1758023419814565,
+      "learning_rate": 1.004378512355999e-05,
+      "loss": 0.9632,
+      "step": 3040
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9122337497151035,
+      "learning_rate": 1.003831201180636e-05,
+      "loss": 1.1513,
+      "step": 3041
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1581336464400607,
+      "learning_rate": 1.0032838888576194e-05,
+      "loss": 0.8025,
+      "step": 3042
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2903339784187364,
+      "learning_rate": 1.0027365755508982e-05,
+      "loss": 0.9303,
+      "step": 3043
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1444251220562296,
+      "learning_rate": 1.0021892614244237e-05,
+      "loss": 0.8378,
+      "step": 3044
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9870418510895796,
+      "learning_rate": 1.0016419466421459e-05,
+      "loss": 0.7779,
+      "step": 3045
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0700751523179626,
+      "learning_rate": 1.0010946313680156e-05,
+      "loss": 1.0833,
+      "step": 3046
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6795742455214346,
+      "learning_rate": 1.0005473157659834e-05,
+      "loss": 0.997,
+      "step": 3047
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1447157814773175,
+      "learning_rate": 1e-05,
+      "loss": 0.7692,
+      "step": 3048
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9721873901608089,
+      "learning_rate": 9.994526842340168e-06,
+      "loss": 0.6867,
+      "step": 3049
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.287391937214058,
+      "learning_rate": 9.989053686319849e-06,
+      "loss": 0.7653,
+      "step": 3050
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9958487396491575,
+      "learning_rate": 9.983580533578545e-06,
+      "loss": 1.0688,
+      "step": 3051
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0986493273470366,
+      "learning_rate": 9.978107385755765e-06,
+      "loss": 0.7563,
+      "step": 3052
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8292133914293822,
+      "learning_rate": 9.972634244491017e-06,
+      "loss": 0.8957,
+      "step": 3053
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0281031523937427,
+      "learning_rate": 9.967161111423813e-06,
+      "loss": 1.1027,
+      "step": 3054
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2226545932957464,
+      "learning_rate": 9.961687988193642e-06,
+      "loss": 0.4453,
+      "step": 3055
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.170290686576733,
+      "learning_rate": 9.956214876440013e-06,
+      "loss": 1.1751,
+      "step": 3056
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7617478098320976,
+      "learning_rate": 9.950741777802417e-06,
+      "loss": 0.5233,
+      "step": 3057
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.895294901519832,
+      "learning_rate": 9.945268693920346e-06,
+      "loss": 0.8673,
+      "step": 3058
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1280187950656846,
+      "learning_rate": 9.939795626433286e-06,
+      "loss": 0.7481,
+      "step": 3059
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.050568807064726,
+      "learning_rate": 9.934322576980721e-06,
+      "loss": 1.0388,
+      "step": 3060
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.035018090136924,
+      "learning_rate": 9.928849547202133e-06,
+      "loss": 0.8473,
+      "step": 3061
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0398290299549802,
+      "learning_rate": 9.923376538736985e-06,
+      "loss": 0.7239,
+      "step": 3062
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.050759619752558,
+      "learning_rate": 9.917903553224745e-06,
+      "loss": 0.8451,
+      "step": 3063
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.919984953125117,
+      "learning_rate": 9.912430592304874e-06,
+      "loss": 1.0138,
+      "step": 3064
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9399467958941092,
+      "learning_rate": 9.906957657616815e-06,
+      "loss": 1.0351,
+      "step": 3065
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1385218842727136,
+      "learning_rate": 9.901484750800014e-06,
+      "loss": 0.8961,
+      "step": 3066
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2459698179327832,
+      "learning_rate": 9.896011873493907e-06,
+      "loss": 0.8538,
+      "step": 3067
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9786452332437485,
+      "learning_rate": 9.890539027337924e-06,
+      "loss": 0.8958,
+      "step": 3068
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0110284040013189,
+      "learning_rate": 9.885066213971471e-06,
+      "loss": 0.8336,
+      "step": 3069
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1800659785992476,
+      "learning_rate": 9.879593435033957e-06,
+      "loss": 1.1105,
+      "step": 3070
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0226093689382791,
+      "learning_rate": 9.874120692164786e-06,
+      "loss": 0.8501,
+      "step": 3071
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.732332428381807,
+      "learning_rate": 9.868647987003332e-06,
+      "loss": 0.9362,
+      "step": 3072
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8772278318287788,
+      "learning_rate": 9.863175321188975e-06,
+      "loss": 0.4832,
+      "step": 3073
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1397055935664246,
+      "learning_rate": 9.857702696361076e-06,
+      "loss": 1.0635,
+      "step": 3074
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0929255884883746,
+      "learning_rate": 9.852230114158986e-06,
+      "loss": 1.1679,
+      "step": 3075
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9325255528224213,
+      "learning_rate": 9.846757576222038e-06,
+      "loss": 0.607,
+      "step": 3076
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3138015623892758,
+      "learning_rate": 9.84128508418956e-06,
+      "loss": 0.897,
+      "step": 3077
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0111390499058754,
+      "learning_rate": 9.835812639700862e-06,
+      "loss": 0.8369,
+      "step": 3078
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9121527119261756,
+      "learning_rate": 9.830340244395235e-06,
+      "loss": 0.9088,
+      "step": 3079
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.983863786799113,
+      "learning_rate": 9.824867899911962e-06,
+      "loss": 1.0076,
+      "step": 3080
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0407528126324141,
+      "learning_rate": 9.81939560789031e-06,
+      "loss": 0.7399,
+      "step": 3081
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9161308436974722,
+      "learning_rate": 9.813923369969535e-06,
+      "loss": 0.6576,
+      "step": 3082
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0987977764507224,
+      "learning_rate": 9.80845118778886e-06,
+      "loss": 0.6979,
+      "step": 3083
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1549183125283375,
+      "learning_rate": 9.802979062987506e-06,
+      "loss": 1.1877,
+      "step": 3084
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1914750891610275,
+      "learning_rate": 9.79750699720468e-06,
+      "loss": 0.7502,
+      "step": 3085
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.091474268305023,
+      "learning_rate": 9.792034992079554e-06,
+      "loss": 0.8921,
+      "step": 3086
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0502032968582782,
+      "learning_rate": 9.786563049251296e-06,
+      "loss": 0.8424,
+      "step": 3087
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8717079709006584,
+      "learning_rate": 9.781091170359053e-06,
+      "loss": 0.963,
+      "step": 3088
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1227984580804167,
+      "learning_rate": 9.775619357041952e-06,
+      "loss": 0.9334,
+      "step": 3089
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.025845472951665,
+      "learning_rate": 9.770147610939098e-06,
+      "loss": 1.006,
+      "step": 3090
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.132978005973592,
+      "learning_rate": 9.764675933689574e-06,
+      "loss": 0.9109,
+      "step": 3091
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8739872268755423,
+      "learning_rate": 9.759204326932453e-06,
+      "loss": 0.5095,
+      "step": 3092
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.000976463282937,
+      "learning_rate": 9.753732792306776e-06,
+      "loss": 0.6522,
+      "step": 3093
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9472778905971364,
+      "learning_rate": 9.748261331451566e-06,
+      "loss": 1.0533,
+      "step": 3094
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.836184128018633,
+      "learning_rate": 9.742789946005822e-06,
+      "loss": 0.9674,
+      "step": 3095
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9373944821249243,
+      "learning_rate": 9.73731863760853e-06,
+      "loss": 0.5814,
+      "step": 3096
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.777865646134292,
+      "learning_rate": 9.731847407898637e-06,
+      "loss": 0.9598,
+      "step": 3097
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.097416151674629,
+      "learning_rate": 9.726376258515077e-06,
+      "loss": 0.7992,
+      "step": 3098
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1680201897536837,
+      "learning_rate": 9.720905191096766e-06,
+      "loss": 1.2179,
+      "step": 3099
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9521732888376955,
+      "learning_rate": 9.715434207282574e-06,
+      "loss": 0.7307,
+      "step": 3100
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0707771325152253,
+      "learning_rate": 9.709963308711368e-06,
+      "loss": 0.8326,
+      "step": 3101
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9910345524809321,
+      "learning_rate": 9.70449249702198e-06,
+      "loss": 0.7547,
+      "step": 3102
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.096098879723292,
+      "learning_rate": 9.699021773853213e-06,
+      "loss": 0.9489,
+      "step": 3103
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0313765958157646,
+      "learning_rate": 9.693551140843848e-06,
+      "loss": 0.9567,
+      "step": 3104
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1975082354427118,
+      "learning_rate": 9.688080599632641e-06,
+      "loss": 0.9306,
+      "step": 3105
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0140833874658255,
+      "learning_rate": 9.682610151858317e-06,
+      "loss": 0.9109,
+      "step": 3106
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9805720160928306,
+      "learning_rate": 9.677139799159573e-06,
+      "loss": 0.5979,
+      "step": 3107
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.044533063027931,
+      "learning_rate": 9.671669543175074e-06,
+      "loss": 1.0974,
+      "step": 3108
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9973768061622381,
+      "learning_rate": 9.666199385543466e-06,
+      "loss": 0.7555,
+      "step": 3109
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.946194773555255,
+      "learning_rate": 9.660729327903359e-06,
+      "loss": 0.6075,
+      "step": 3110
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9552445104837077,
+      "learning_rate": 9.655259371893329e-06,
+      "loss": 0.739,
+      "step": 3111
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9878858621386367,
+      "learning_rate": 9.64978951915193e-06,
+      "loss": 0.7014,
+      "step": 3112
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7346752935711542,
+      "learning_rate": 9.644319771317686e-06,
+      "loss": 0.9269,
+      "step": 3113
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9999513876766326,
+      "learning_rate": 9.638850130029075e-06,
+      "loss": 0.7506,
+      "step": 3114
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.35912135298015,
+      "learning_rate": 9.633380596924557e-06,
+      "loss": 0.9291,
+      "step": 3115
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.007298939608526,
+      "learning_rate": 9.627911173642564e-06,
+      "loss": 0.8611,
+      "step": 3116
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2030151298233553,
+      "learning_rate": 9.622441861821472e-06,
+      "loss": 0.8652,
+      "step": 3117
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0687816588860266,
+      "learning_rate": 9.616972663099648e-06,
+      "loss": 1.0858,
+      "step": 3118
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0987705470296718,
+      "learning_rate": 9.61150357911541e-06,
+      "loss": 0.7027,
+      "step": 3119
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8380056253994261,
+      "learning_rate": 9.606034611507058e-06,
+      "loss": 1.0295,
+      "step": 3120
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9251970269681763,
+      "learning_rate": 9.600565761912833e-06,
+      "loss": 0.7937,
+      "step": 3121
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.088975287143847,
+      "learning_rate": 9.595097031970963e-06,
+      "loss": 1.1576,
+      "step": 3122
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1388653364715413,
+      "learning_rate": 9.589628423319625e-06,
+      "loss": 0.807,
+      "step": 3123
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9303866091031778,
+      "learning_rate": 9.584159937596968e-06,
+      "loss": 0.7216,
+      "step": 3124
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9692093533121043,
+      "learning_rate": 9.578691576441104e-06,
+      "loss": 0.677,
+      "step": 3125
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.832716341074292,
+      "learning_rate": 9.5732233414901e-06,
+      "loss": 0.7382,
+      "step": 3126
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7802293846922028,
+      "learning_rate": 9.567755234382001e-06,
+      "loss": 0.6902,
+      "step": 3127
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9018122520856247,
+      "learning_rate": 9.562287256754791e-06,
+      "loss": 0.9001,
+      "step": 3128
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0413182573709445,
+      "learning_rate": 9.556819410246435e-06,
+      "loss": 0.8537,
+      "step": 3129
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1572562361004264,
+      "learning_rate": 9.551351696494854e-06,
+      "loss": 0.7974,
+      "step": 3130
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3366761064782764,
+      "learning_rate": 9.545884117137918e-06,
+      "loss": 0.8066,
+      "step": 3131
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0886338800103503,
+      "learning_rate": 9.540416673813471e-06,
+      "loss": 1.146,
+      "step": 3132
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0281495761422268,
+      "learning_rate": 9.53494936815931e-06,
+      "loss": 0.6947,
+      "step": 3133
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.060759340026712,
+      "learning_rate": 9.529482201813192e-06,
+      "loss": 0.6828,
+      "step": 3134
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1171936595518253,
+      "learning_rate": 9.52401517641283e-06,
+      "loss": 0.9761,
+      "step": 3135
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2524852326223253,
+      "learning_rate": 9.518548293595895e-06,
+      "loss": 1.0398,
+      "step": 3136
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0287151602423303,
+      "learning_rate": 9.513081555000026e-06,
+      "loss": 1.025,
+      "step": 3137
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9394149933491993,
+      "learning_rate": 9.507614962262796e-06,
+      "loss": 0.693,
+      "step": 3138
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9409933756507579,
+      "learning_rate": 9.502148517021757e-06,
+      "loss": 1.0558,
+      "step": 3139
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0790733759972726,
+      "learning_rate": 9.496682220914403e-06,
+      "loss": 0.7972,
+      "step": 3140
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3987927744294257,
+      "learning_rate": 9.49121607557819e-06,
+      "loss": 0.758,
+      "step": 3141
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1025654024681804,
+      "learning_rate": 9.48575008265053e-06,
+      "loss": 1.1996,
+      "step": 3142
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0700410904033484,
+      "learning_rate": 9.48028424376878e-06,
+      "loss": 0.8537,
+      "step": 3143
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0408959913628004,
+      "learning_rate": 9.474818560570262e-06,
+      "loss": 0.7583,
+      "step": 3144
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8471770323145819,
+      "learning_rate": 9.469353034692243e-06,
+      "loss": 0.9876,
+      "step": 3145
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6965498833685295,
+      "learning_rate": 9.463887667771946e-06,
+      "loss": 0.5066,
+      "step": 3146
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.269926133270752,
+      "learning_rate": 9.458422461446547e-06,
+      "loss": 0.9085,
+      "step": 3147
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3569037537646926,
+      "learning_rate": 9.45295741735318e-06,
+      "loss": 0.869,
+      "step": 3148
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1241377844501963,
+      "learning_rate": 9.447492537128917e-06,
+      "loss": 0.7694,
+      "step": 3149
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1401893212116654,
+      "learning_rate": 9.442027822410786e-06,
+      "loss": 0.7922,
+      "step": 3150
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1151319177072265,
+      "learning_rate": 9.436563274835777e-06,
+      "loss": 1.0913,
+      "step": 3151
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1918136785701654,
+      "learning_rate": 9.43109889604081e-06,
+      "loss": 0.8963,
+      "step": 3152
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2508682569629386,
+      "learning_rate": 9.425634687662768e-06,
+      "loss": 0.7267,
+      "step": 3153
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0047315378846573,
+      "learning_rate": 9.420170651338482e-06,
+      "loss": 0.7669,
+      "step": 3154
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.186542754520551,
+      "learning_rate": 9.414706788704727e-06,
+      "loss": 0.9608,
+      "step": 3155
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8927247944307228,
+      "learning_rate": 9.409243101398228e-06,
+      "loss": 1.0598,
+      "step": 3156
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0346509044224752,
+      "learning_rate": 9.40377959105566e-06,
+      "loss": 0.7439,
+      "step": 3157
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.99390455036312,
+      "learning_rate": 9.398316259313636e-06,
+      "loss": 0.9119,
+      "step": 3158
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1786970780005304,
+      "learning_rate": 9.39285310780873e-06,
+      "loss": 0.8118,
+      "step": 3159
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2285522787374887,
+      "learning_rate": 9.387390138177447e-06,
+      "loss": 0.8562,
+      "step": 3160
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9383984670505165,
+      "learning_rate": 9.381927352056249e-06,
+      "loss": 1.0137,
+      "step": 3161
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1795007374462927,
+      "learning_rate": 9.376464751081542e-06,
+      "loss": 0.8553,
+      "step": 3162
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5811067290443608,
+      "learning_rate": 9.371002336889661e-06,
+      "loss": 0.5617,
+      "step": 3163
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9119051416735352,
+      "learning_rate": 9.365540111116907e-06,
+      "loss": 0.6498,
+      "step": 3164
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3322686333256135,
+      "learning_rate": 9.360078075399517e-06,
+      "loss": 1.2296,
+      "step": 3165
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0233564486358686,
+      "learning_rate": 9.354616231373659e-06,
+      "loss": 1.0846,
+      "step": 3166
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1251926291886167,
+      "learning_rate": 9.349154580675457e-06,
+      "loss": 1.0273,
+      "step": 3167
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8678982246141452,
+      "learning_rate": 9.343693124940977e-06,
+      "loss": 0.9424,
+      "step": 3168
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0951240332646475,
+      "learning_rate": 9.338231865806224e-06,
+      "loss": 0.969,
+      "step": 3169
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8761008364732744,
+      "learning_rate": 9.332770804907137e-06,
+      "loss": 0.6208,
+      "step": 3170
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9927920839989717,
+      "learning_rate": 9.327309943879604e-06,
+      "loss": 1.0219,
+      "step": 3171
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0106729146466789,
+      "learning_rate": 9.321849284359454e-06,
+      "loss": 0.7932,
+      "step": 3172
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.696723187997719,
+      "learning_rate": 9.316388827982448e-06,
+      "loss": 0.498,
+      "step": 3173
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1950800076038046,
+      "learning_rate": 9.310928576384293e-06,
+      "loss": 0.8885,
+      "step": 3174
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0190820862653396,
+      "learning_rate": 9.305468531200633e-06,
+      "loss": 1.1136,
+      "step": 3175
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.922638974598266,
+      "learning_rate": 9.30000869406705e-06,
+      "loss": 0.7625,
+      "step": 3176
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3113340445867523,
+      "learning_rate": 9.294549066619058e-06,
+      "loss": 0.8429,
+      "step": 3177
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9858493337457307,
+      "learning_rate": 9.289089650492119e-06,
+      "loss": 0.7336,
+      "step": 3178
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1970100647311266,
+      "learning_rate": 9.283630447321627e-06,
+      "loss": 0.8005,
+      "step": 3179
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9583092827402718,
+      "learning_rate": 9.278171458742903e-06,
+      "loss": 0.9952,
+      "step": 3180
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5761625253449129,
+      "learning_rate": 9.272712686391219e-06,
+      "loss": 0.5376,
+      "step": 3181
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2356681007608608,
+      "learning_rate": 9.267254131901772e-06,
+      "loss": 1.0671,
+      "step": 3182
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0924803592020647,
+      "learning_rate": 9.261795796909704e-06,
+      "loss": 0.7868,
+      "step": 3183
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8879076470265828,
+      "learning_rate": 9.256337683050075e-06,
+      "loss": 1.1072,
+      "step": 3184
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0219718075140451,
+      "learning_rate": 9.25087979195789e-06,
+      "loss": 0.9276,
+      "step": 3185
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1070878318047133,
+      "learning_rate": 9.245422125268094e-06,
+      "loss": 0.7133,
+      "step": 3186
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.021595390185087,
+      "learning_rate": 9.239964684615543e-06,
+      "loss": 0.7456,
+      "step": 3187
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2697922625475244,
+      "learning_rate": 9.234507471635043e-06,
+      "loss": 0.8939,
+      "step": 3188
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0259523971687392,
+      "learning_rate": 9.229050487961332e-06,
+      "loss": 1.026,
+      "step": 3189
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1173290877082187,
+      "learning_rate": 9.223593735229069e-06,
+      "loss": 0.877,
+      "step": 3190
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0241882712573367,
+      "learning_rate": 9.218137215072853e-06,
+      "loss": 0.7587,
+      "step": 3191
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1523836591775272,
+      "learning_rate": 9.212680929127205e-06,
+      "loss": 0.9157,
+      "step": 3192
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9792665502915333,
+      "learning_rate": 9.207224879026586e-06,
+      "loss": 0.9939,
+      "step": 3193
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1453764667846837,
+      "learning_rate": 9.201769066405373e-06,
+      "loss": 1.0149,
+      "step": 3194
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9350205816956488,
+      "learning_rate": 9.196313492897885e-06,
+      "loss": 0.7831,
+      "step": 3195
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.14791825982617,
+      "learning_rate": 9.190858160138362e-06,
+      "loss": 0.8229,
+      "step": 3196
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1352103136612128,
+      "learning_rate": 9.185403069760981e-06,
+      "loss": 0.7807,
+      "step": 3197
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.173192269916578,
+      "learning_rate": 9.179948223399828e-06,
+      "loss": 0.7711,
+      "step": 3198
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6076731780387671,
+      "learning_rate": 9.174493622688933e-06,
+      "loss": 0.6475,
+      "step": 3199
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0198955319847318,
+      "learning_rate": 9.16903926926225e-06,
+      "loss": 0.792,
+      "step": 3200
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.934269020968029,
+      "learning_rate": 9.163585164753648e-06,
+      "loss": 0.8349,
+      "step": 3201
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0676215243193128,
+      "learning_rate": 9.15813131079693e-06,
+      "loss": 0.686,
+      "step": 3202
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1577627363505991,
+      "learning_rate": 9.152677709025832e-06,
+      "loss": 0.8266,
+      "step": 3203
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0690742200434382,
+      "learning_rate": 9.147224361073994e-06,
+      "loss": 1.088,
+      "step": 3204
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2493962089205282,
+      "learning_rate": 9.141771268574995e-06,
+      "loss": 0.8778,
+      "step": 3205
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8808422428453211,
+      "learning_rate": 9.136318433162336e-06,
+      "loss": 0.8127,
+      "step": 3206
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.273929037981517,
+      "learning_rate": 9.13086585646944e-06,
+      "loss": 0.9103,
+      "step": 3207
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9730044981834982,
+      "learning_rate": 9.125413540129647e-06,
+      "loss": 1.2544,
+      "step": 3208
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0984488837853033,
+      "learning_rate": 9.119961485776223e-06,
+      "loss": 0.9664,
+      "step": 3209
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9396521747368579,
+      "learning_rate": 9.114509695042362e-06,
+      "loss": 0.5534,
+      "step": 3210
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1252945204432303,
+      "learning_rate": 9.109058169561167e-06,
+      "loss": 0.8474,
+      "step": 3211
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2716175369634237,
+      "learning_rate": 9.103606910965666e-06,
+      "loss": 0.7584,
+      "step": 3212
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1321147822885331,
+      "learning_rate": 9.098155920888813e-06,
+      "loss": 0.718,
+      "step": 3213
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0912390126233993,
+      "learning_rate": 9.092705200963478e-06,
+      "loss": 1.1824,
+      "step": 3214
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9116592936942285,
+      "learning_rate": 9.087254752822444e-06,
+      "loss": 0.7601,
+      "step": 3215
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0816761268868191,
+      "learning_rate": 9.08180457809842e-06,
+      "loss": 0.8675,
+      "step": 3216
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6349828866840606,
+      "learning_rate": 9.076354678424033e-06,
+      "loss": 0.3873,
+      "step": 3217
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0288039497769592,
+      "learning_rate": 9.070905055431822e-06,
+      "loss": 1.0448,
+      "step": 3218
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9368819666303689,
+      "learning_rate": 9.065455710754247e-06,
+      "loss": 0.8108,
+      "step": 3219
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0250198884579134,
+      "learning_rate": 9.060006646023683e-06,
+      "loss": 0.7832,
+      "step": 3220
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.082543994515398,
+      "learning_rate": 9.05455786287243e-06,
+      "loss": 0.8776,
+      "step": 3221
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0764188008746676,
+      "learning_rate": 9.049109362932685e-06,
+      "loss": 0.6066,
+      "step": 3222
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.07296045630113,
+      "learning_rate": 9.043661147836578e-06,
+      "loss": 1.0424,
+      "step": 3223
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7675317839798449,
+      "learning_rate": 9.038213219216145e-06,
+      "loss": 0.506,
+      "step": 3224
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0157460624751762,
+      "learning_rate": 9.032765578703338e-06,
+      "loss": 0.7478,
+      "step": 3225
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.029335497967583,
+      "learning_rate": 9.02731822793002e-06,
+      "loss": 0.8243,
+      "step": 3226
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0197383532886097,
+      "learning_rate": 9.021871168527972e-06,
+      "loss": 0.7853,
+      "step": 3227
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0540149554690108,
+      "learning_rate": 9.016424402128891e-06,
+      "loss": 1.0691,
+      "step": 3228
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0635036652319079,
+      "learning_rate": 9.01097793036437e-06,
+      "loss": 1.0196,
+      "step": 3229
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.002878088711378,
+      "learning_rate": 9.005531754865929e-06,
+      "loss": 0.7703,
+      "step": 3230
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.39437149169036,
+      "learning_rate": 9.000085877265001e-06,
+      "loss": 1.0274,
+      "step": 3231
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9716515969809777,
+      "learning_rate": 8.994640299192914e-06,
+      "loss": 0.9192,
+      "step": 3232
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0637563153035745,
+      "learning_rate": 8.98919502228092e-06,
+      "loss": 0.9866,
+      "step": 3233
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0899041449483495,
+      "learning_rate": 8.983750048160178e-06,
+      "loss": 0.8594,
+      "step": 3234
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7442454446079833,
+      "learning_rate": 8.978305378461757e-06,
+      "loss": 0.5516,
+      "step": 3235
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9570455142636096,
+      "learning_rate": 8.972861014816622e-06,
+      "loss": 0.7349,
+      "step": 3236
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1955058346632026,
+      "learning_rate": 8.967416958855666e-06,
+      "loss": 1.004,
+      "step": 3237
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0387996874669692,
+      "learning_rate": 8.961973212209684e-06,
+      "loss": 1.0076,
+      "step": 3238
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0490987152067845,
+      "learning_rate": 8.956529776509366e-06,
+      "loss": 0.7599,
+      "step": 3239
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1856819411605501,
+      "learning_rate": 8.951086653385323e-06,
+      "loss": 0.8643,
+      "step": 3240
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0558327020808675,
+      "learning_rate": 8.945643844468066e-06,
+      "loss": 0.6957,
+      "step": 3241
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7217646477290933,
+      "learning_rate": 8.940201351388016e-06,
+      "loss": 1.0147,
+      "step": 3242
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9685045584234501,
+      "learning_rate": 8.934759175775494e-06,
+      "loss": 0.5486,
+      "step": 3243
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9958809041703425,
+      "learning_rate": 8.929317319260727e-06,
+      "loss": 0.9251,
+      "step": 3244
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0213724639931396,
+      "learning_rate": 8.923875783473854e-06,
+      "loss": 0.8217,
+      "step": 3245
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0387739843604609,
+      "learning_rate": 8.918434570044904e-06,
+      "loss": 0.8066,
+      "step": 3246
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.144586728885735,
+      "learning_rate": 8.912993680603823e-06,
+      "loss": 1.1617,
+      "step": 3247
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3272163702326258,
+      "learning_rate": 8.90755311678045e-06,
+      "loss": 0.806,
+      "step": 3248
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9380820287203456,
+      "learning_rate": 8.90211288020454e-06,
+      "loss": 0.7154,
+      "step": 3249
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9502251770152589,
+      "learning_rate": 8.89667297250573e-06,
+      "loss": 1.0814,
+      "step": 3250
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9787050643516357,
+      "learning_rate": 8.891233395313572e-06,
+      "loss": 0.6769,
+      "step": 3251
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.127693717888957,
+      "learning_rate": 8.885794150257522e-06,
+      "loss": 1.1279,
+      "step": 3252
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9001596699034452,
+      "learning_rate": 8.880355238966923e-06,
+      "loss": 0.7416,
+      "step": 3253
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7053235449997682,
+      "learning_rate": 8.874916663071027e-06,
+      "loss": 0.4862,
+      "step": 3254
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0580788452349401,
+      "learning_rate": 8.869478424198991e-06,
+      "loss": 0.9671,
+      "step": 3255
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0340726186643117,
+      "learning_rate": 8.86404052397986e-06,
+      "loss": 0.9378,
+      "step": 3256
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9793178007004393,
+      "learning_rate": 8.858602964042579e-06,
+      "loss": 0.9954,
+      "step": 3257
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0828503914791456,
+      "learning_rate": 8.853165746015997e-06,
+      "loss": 0.9058,
+      "step": 3258
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.155371150170303,
+      "learning_rate": 8.847728871528857e-06,
+      "loss": 0.6901,
+      "step": 3259
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0163730402076203,
+      "learning_rate": 8.842292342209801e-06,
+      "loss": 0.8466,
+      "step": 3260
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0686473549552507,
+      "learning_rate": 8.836856159687362e-06,
+      "loss": 1.1473,
+      "step": 3261
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0707409579083829,
+      "learning_rate": 8.831420325589976e-06,
+      "loss": 0.6951,
+      "step": 3262
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2249427860077138,
+      "learning_rate": 8.825984841545976e-06,
+      "loss": 0.7034,
+      "step": 3263
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.233806051139251,
+      "learning_rate": 8.820549709183579e-06,
+      "loss": 0.8123,
+      "step": 3264
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0768234353397015,
+      "learning_rate": 8.815114930130906e-06,
+      "loss": 0.8061,
+      "step": 3265
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9765495603371205,
+      "learning_rate": 8.809680506015973e-06,
+      "loss": 1.0112,
+      "step": 3266
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0686987220622832,
+      "learning_rate": 8.804246438466683e-06,
+      "loss": 0.9984,
+      "step": 3267
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1289957112878222,
+      "learning_rate": 8.798812729110837e-06,
+      "loss": 0.7653,
+      "step": 3268
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.060072989814974,
+      "learning_rate": 8.793379379576127e-06,
+      "loss": 0.8366,
+      "step": 3269
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0587296608825674,
+      "learning_rate": 8.787946391490141e-06,
+      "loss": 0.6749,
+      "step": 3270
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.025981661646274,
+      "learning_rate": 8.782513766480351e-06,
+      "loss": 1.1879,
+      "step": 3271
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6569578382464362,
+      "learning_rate": 8.777081506174127e-06,
+      "loss": 0.4935,
+      "step": 3272
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2200109430376627,
+      "learning_rate": 8.771649612198727e-06,
+      "loss": 1.0454,
+      "step": 3273
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1341023936735835,
+      "learning_rate": 8.766218086181298e-06,
+      "loss": 0.7426,
+      "step": 3274
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1244664134254785,
+      "learning_rate": 8.76078692974888e-06,
+      "loss": 1.1254,
+      "step": 3275
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2246540668047055,
+      "learning_rate": 8.755356144528399e-06,
+      "loss": 0.811,
+      "step": 3276
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2127291463822507,
+      "learning_rate": 8.74992573214668e-06,
+      "loss": 0.8564,
+      "step": 3277
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2203590414615484,
+      "learning_rate": 8.744495694230413e-06,
+      "loss": 1.0031,
+      "step": 3278
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.110283837903533,
+      "learning_rate": 8.739066032406201e-06,
+      "loss": 0.794,
+      "step": 3279
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0365721181957317,
+      "learning_rate": 8.733636748300524e-06,
+      "loss": 1.066,
+      "step": 3280
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2764734135430749,
+      "learning_rate": 8.728207843539744e-06,
+      "loss": 0.7859,
+      "step": 3281
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3009844087498463,
+      "learning_rate": 8.722779319750116e-06,
+      "loss": 0.9153,
+      "step": 3282
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.111690347379662,
+      "learning_rate": 8.717351178557779e-06,
+      "loss": 0.929,
+      "step": 3283
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1532127014930686,
+      "learning_rate": 8.711923421588764e-06,
+      "loss": 0.7158,
+      "step": 3284
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.00315293087695,
+      "learning_rate": 8.70649605046897e-06,
+      "loss": 0.9597,
+      "step": 3285
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1461833659349208,
+      "learning_rate": 8.701069066824195e-06,
+      "loss": 0.9121,
+      "step": 3286
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0116913207598228,
+      "learning_rate": 8.695642472280123e-06,
+      "loss": 0.8258,
+      "step": 3287
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0710814120838692,
+      "learning_rate": 8.690216268462304e-06,
+      "loss": 0.6075,
+      "step": 3288
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9178374325408805,
+      "learning_rate": 8.684790456996189e-06,
+      "loss": 0.7178,
+      "step": 3289
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0368094619804016,
+      "learning_rate": 8.679365039507104e-06,
+      "loss": 1.057,
+      "step": 3290
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6857355261238353,
+      "learning_rate": 8.673940017620257e-06,
+      "loss": 0.4529,
+      "step": 3291
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1307591701798587,
+      "learning_rate": 8.668515392960737e-06,
+      "loss": 0.9943,
+      "step": 3292
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9707100512785175,
+      "learning_rate": 8.663091167153516e-06,
+      "loss": 1.1142,
+      "step": 3293
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8804165390749304,
+      "learning_rate": 8.657667341823449e-06,
+      "loss": 0.6333,
+      "step": 3294
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9829859684905075,
+      "learning_rate": 8.65224391859526e-06,
+      "loss": 1.1701,
+      "step": 3295
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2627318097843707,
+      "learning_rate": 8.646820899093564e-06,
+      "loss": 0.7757,
+      "step": 3296
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.060568494788007,
+      "learning_rate": 8.641398284942854e-06,
+      "loss": 0.9682,
+      "step": 3297
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3692898408509115,
+      "learning_rate": 8.6359760777675e-06,
+      "loss": 0.7515,
+      "step": 3298
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.747698902611198,
+      "learning_rate": 8.63055427919174e-06,
+      "loss": 0.7699,
+      "step": 3299
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2892047609129804,
+      "learning_rate": 8.625132890839706e-06,
+      "loss": 0.8109,
+      "step": 3300
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4617457106329372,
+      "learning_rate": 8.619711914335402e-06,
+      "loss": 0.7057,
+      "step": 3301
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.401308727427695,
+      "learning_rate": 8.614291351302697e-06,
+      "loss": 0.949,
+      "step": 3302
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9168973735191371,
+      "learning_rate": 8.60887120336535e-06,
+      "loss": 0.6443,
+      "step": 3303
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2488103594832574,
+      "learning_rate": 8.603451472146996e-06,
+      "loss": 1.1141,
+      "step": 3304
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5128946630894726,
+      "learning_rate": 8.598032159271134e-06,
+      "loss": 0.7564,
+      "step": 3305
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.519526797427506,
+      "learning_rate": 8.592613266361145e-06,
+      "loss": 0.8889,
+      "step": 3306
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.278178221805753,
+      "learning_rate": 8.587194795040286e-06,
+      "loss": 0.9872,
+      "step": 3307
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7532984840314697,
+      "learning_rate": 8.581776746931682e-06,
+      "loss": 0.4106,
+      "step": 3308
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.219493230104484,
+      "learning_rate": 8.576359123658336e-06,
+      "loss": 1.0212,
+      "step": 3309
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7672302329164373,
+      "learning_rate": 8.570941926843117e-06,
+      "loss": 0.7475,
+      "step": 3310
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3356670349644317,
+      "learning_rate": 8.56552515810878e-06,
+      "loss": 0.8906,
+      "step": 3311
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1240477245959744,
+      "learning_rate": 8.560108819077934e-06,
+      "loss": 0.8737,
+      "step": 3312
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0087104557611477,
+      "learning_rate": 8.55469291137307e-06,
+      "loss": 0.5957,
+      "step": 3313
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4601250965882557,
+      "learning_rate": 8.549277436616551e-06,
+      "loss": 1.3098,
+      "step": 3314
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6152469416868387,
+      "learning_rate": 8.54386239643061e-06,
+      "loss": 0.7049,
+      "step": 3315
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2956229310810126,
+      "learning_rate": 8.538447792437338e-06,
+      "loss": 0.8846,
+      "step": 3316
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0801736911761655,
+      "learning_rate": 8.53303362625871e-06,
+      "loss": 0.7467,
+      "step": 3317
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8537549309823141,
+      "learning_rate": 8.527619899516567e-06,
+      "loss": 1.0247,
+      "step": 3318
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6234600539200736,
+      "learning_rate": 8.52220661383261e-06,
+      "loss": 0.7952,
+      "step": 3319
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1648311847419157,
+      "learning_rate": 8.516793770828412e-06,
+      "loss": 0.8036,
+      "step": 3320
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2271096358424902,
+      "learning_rate": 8.511381372125422e-06,
+      "loss": 0.8212,
+      "step": 3321
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.086757202305495,
+      "learning_rate": 8.505969419344946e-06,
+      "loss": 0.8503,
+      "step": 3322
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0113647242197124,
+      "learning_rate": 8.500557914108155e-06,
+      "loss": 0.9903,
+      "step": 3323
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2823929525265632,
+      "learning_rate": 8.495146858036097e-06,
+      "loss": 0.824,
+      "step": 3324
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1812006559677874,
+      "learning_rate": 8.489736252749674e-06,
+      "loss": 0.689,
+      "step": 3325
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0410167301170565,
+      "learning_rate": 8.484326099869659e-06,
+      "loss": 0.9068,
+      "step": 3326
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9379699995856309,
+      "learning_rate": 8.478916401016686e-06,
+      "loss": 0.7413,
+      "step": 3327
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.351404666482721,
+      "learning_rate": 8.473507157811254e-06,
+      "loss": 1.1526,
+      "step": 3328
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.227427409649178,
+      "learning_rate": 8.468098371873736e-06,
+      "loss": 0.7503,
+      "step": 3329
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5788251569114373,
+      "learning_rate": 8.462690044824347e-06,
+      "loss": 0.8602,
+      "step": 3330
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2577811297955968,
+      "learning_rate": 8.457282178283182e-06,
+      "loss": 0.8505,
+      "step": 3331
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9876144480168727,
+      "learning_rate": 8.451874773870192e-06,
+      "loss": 0.9508,
+      "step": 3332
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2406882865701778,
+      "learning_rate": 8.446467833205186e-06,
+      "loss": 1.0277,
+      "step": 3333
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3012684502386294,
+      "learning_rate": 8.441061357907842e-06,
+      "loss": 0.9025,
+      "step": 3334
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.319754203853252,
+      "learning_rate": 8.43565534959769e-06,
+      "loss": 0.7618,
+      "step": 3335
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1657851192138007,
+      "learning_rate": 8.430249809894134e-06,
+      "loss": 0.8222,
+      "step": 3336
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.017279950856374,
+      "learning_rate": 8.424844740416415e-06,
+      "loss": 1.3071,
+      "step": 3337
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3401063342689805,
+      "learning_rate": 8.419440142783653e-06,
+      "loss": 0.8056,
+      "step": 3338
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1965876995269777,
+      "learning_rate": 8.414036018614826e-06,
+      "loss": 0.7576,
+      "step": 3339
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0794451669692684,
+      "learning_rate": 8.40863236952875e-06,
+      "loss": 0.8215,
+      "step": 3340
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1040327870454454,
+      "learning_rate": 8.403229197144125e-06,
+      "loss": 0.9848,
+      "step": 3341
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0961470225697487,
+      "learning_rate": 8.397826503079489e-06,
+      "loss": 0.9131,
+      "step": 3342
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9672129682833389,
+      "learning_rate": 8.392424288953252e-06,
+      "loss": 0.81,
+      "step": 3343
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5581391031305747,
+      "learning_rate": 8.387022556383659e-06,
+      "loss": 0.4757,
+      "step": 3344
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.391663881284385,
+      "learning_rate": 8.381621306988832e-06,
+      "loss": 0.8517,
+      "step": 3345
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.084324907667858,
+      "learning_rate": 8.376220542386742e-06,
+      "loss": 0.83,
+      "step": 3346
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0544922376195909,
+      "learning_rate": 8.370820264195207e-06,
+      "loss": 1.0953,
+      "step": 3347
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8519132877993966,
+      "learning_rate": 8.365420474031905e-06,
+      "loss": 0.5708,
+      "step": 3348
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3817327644717459,
+      "learning_rate": 8.360021173514373e-06,
+      "loss": 0.6578,
+      "step": 3349
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.352849344816967,
+      "learning_rate": 8.354622364259995e-06,
+      "loss": 0.9396,
+      "step": 3350
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2190296936180272,
+      "learning_rate": 8.349224047886005e-06,
+      "loss": 0.7946,
+      "step": 3351
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1785200223399541,
+      "learning_rate": 8.343826226009494e-06,
+      "loss": 1.1488,
+      "step": 3352
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0474222407073603,
+      "learning_rate": 8.338428900247412e-06,
+      "loss": 0.7694,
+      "step": 3353
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.15574101807336,
+      "learning_rate": 8.333032072216542e-06,
+      "loss": 0.7516,
+      "step": 3354
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.539210830414841,
+      "learning_rate": 8.327635743533531e-06,
+      "loss": 0.8399,
+      "step": 3355
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.124976467558571,
+      "learning_rate": 8.32223991581488e-06,
+      "loss": 1.1365,
+      "step": 3356
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0287457469641175,
+      "learning_rate": 8.316844590676928e-06,
+      "loss": 0.7777,
+      "step": 3357
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1678672547208888,
+      "learning_rate": 8.311449769735874e-06,
+      "loss": 0.8032,
+      "step": 3358
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0918030068735591,
+      "learning_rate": 8.306055454607755e-06,
+      "loss": 1.2419,
+      "step": 3359
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9716617436917057,
+      "learning_rate": 8.30066164690847e-06,
+      "loss": 0.7741,
+      "step": 3360
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5800252589027973,
+      "learning_rate": 8.29526834825375e-06,
+      "loss": 0.4851,
+      "step": 3361
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0543460707946077,
+      "learning_rate": 8.289875560259189e-06,
+      "loss": 1.0425,
+      "step": 3362
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0688394851919345,
+      "learning_rate": 8.284483284540217e-06,
+      "loss": 0.8212,
+      "step": 3363
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9862544128324102,
+      "learning_rate": 8.279091522712123e-06,
+      "loss": 0.9087,
+      "step": 3364
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9293392466936687,
+      "learning_rate": 8.273700276390024e-06,
+      "loss": 0.5758,
+      "step": 3365
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9635434542879175,
+      "learning_rate": 8.268309547188894e-06,
+      "loss": 1.0452,
+      "step": 3366
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0458872891395998,
+      "learning_rate": 8.262919336723561e-06,
+      "loss": 0.7418,
+      "step": 3367
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8061633808646909,
+      "learning_rate": 8.257529646608673e-06,
+      "loss": 0.6292,
+      "step": 3368
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1964843398397005,
+      "learning_rate": 8.252140478458744e-06,
+      "loss": 0.793,
+      "step": 3369
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6716026318369355,
+      "learning_rate": 8.246751833888122e-06,
+      "loss": 0.7838,
+      "step": 3370
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0215409085575087,
+      "learning_rate": 8.241363714511005e-06,
+      "loss": 1.2891,
+      "step": 3371
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9817234012913878,
+      "learning_rate": 8.235976121941421e-06,
+      "loss": 0.6579,
+      "step": 3372
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9668283229962158,
+      "learning_rate": 8.230589057793253e-06,
+      "loss": 0.9539,
+      "step": 3373
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9213380982115361,
+      "learning_rate": 8.225202523680218e-06,
+      "loss": 0.7889,
+      "step": 3374
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8324099289535631,
+      "learning_rate": 8.21981652121588e-06,
+      "loss": 0.5994,
+      "step": 3375
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.072865738810485,
+      "learning_rate": 8.214431052013636e-06,
+      "loss": 1.0488,
+      "step": 3376
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9598533547460161,
+      "learning_rate": 8.20904611768673e-06,
+      "loss": 0.7264,
+      "step": 3377
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2122981685680547,
+      "learning_rate": 8.203661719848249e-06,
+      "loss": 0.882,
+      "step": 3378
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9692059011729305,
+      "learning_rate": 8.198277860111103e-06,
+      "loss": 0.7123,
+      "step": 3379
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1127106130465847,
+      "learning_rate": 8.192894540088061e-06,
+      "loss": 1.1299,
+      "step": 3380
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6187412316548248,
+      "learning_rate": 8.187511761391722e-06,
+      "loss": 0.6482,
+      "step": 3381
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1124370068421423,
+      "learning_rate": 8.182129525634513e-06,
+      "loss": 0.7771,
+      "step": 3382
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1158731472976589,
+      "learning_rate": 8.176747834428714e-06,
+      "loss": 0.732,
+      "step": 3383
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9196825110106804,
+      "learning_rate": 8.171366689386433e-06,
+      "loss": 0.9705,
+      "step": 3384
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.300652646502842,
+      "learning_rate": 8.16598609211962e-06,
+      "loss": 1.1301,
+      "step": 3385
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.04686151265235,
+      "learning_rate": 8.160606044240052e-06,
+      "loss": 0.7426,
+      "step": 3386
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4318459956410876,
+      "learning_rate": 8.155226547359349e-06,
+      "loss": 0.9485,
+      "step": 3387
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5500970237803748,
+      "learning_rate": 8.14984760308897e-06,
+      "loss": 0.8548,
+      "step": 3388
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.260439176054417,
+      "learning_rate": 8.144469213040192e-06,
+      "loss": 0.8249,
+      "step": 3389
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9699591797742972,
+      "learning_rate": 8.139091378824144e-06,
+      "loss": 0.9901,
+      "step": 3390
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.813778490342783,
+      "learning_rate": 8.133714102051774e-06,
+      "loss": 0.9549,
+      "step": 3391
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1851136872647017,
+      "learning_rate": 8.128337384333879e-06,
+      "loss": 0.854,
+      "step": 3392
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0188431652826937,
+      "learning_rate": 8.122961227281071e-06,
+      "loss": 0.8018,
+      "step": 3393
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3939954123163651,
+      "learning_rate": 8.117585632503805e-06,
+      "loss": 0.8751,
+      "step": 3394
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1466525804070427,
+      "learning_rate": 8.11221060161237e-06,
+      "loss": 1.0742,
+      "step": 3395
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0926767216123243,
+      "learning_rate": 8.106836136216872e-06,
+      "loss": 0.8892,
+      "step": 3396
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.982968720572902,
+      "learning_rate": 8.101462237927262e-06,
+      "loss": 1.0505,
+      "step": 3397
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5889381446108544,
+      "learning_rate": 8.096088908353316e-06,
+      "loss": 0.4716,
+      "step": 3398
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3441163827194125,
+      "learning_rate": 8.090716149104642e-06,
+      "loss": 0.6678,
+      "step": 3399
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1818671906073965,
+      "learning_rate": 8.085343961790666e-06,
+      "loss": 1.1629,
+      "step": 3400
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4759354627753825,
+      "learning_rate": 8.079972348020656e-06,
+      "loss": 0.9105,
+      "step": 3401
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4589661883386975,
+      "learning_rate": 8.07460130940371e-06,
+      "loss": 0.7866,
+      "step": 3402
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9492300591399138,
+      "learning_rate": 8.069230847548739e-06,
+      "loss": 1.0673,
+      "step": 3403
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1272844717572887,
+      "learning_rate": 8.063860964064487e-06,
+      "loss": 1.135,
+      "step": 3404
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1492836997797584,
+      "learning_rate": 8.058491660559536e-06,
+      "loss": 0.757,
+      "step": 3405
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2005025287751352,
+      "learning_rate": 8.05312293864228e-06,
+      "loss": 0.9619,
+      "step": 3406
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0909750372676914,
+      "learning_rate": 8.047754799920946e-06,
+      "loss": 0.6703,
+      "step": 3407
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7493067554631103,
+      "learning_rate": 8.042387246003583e-06,
+      "loss": 0.629,
+      "step": 3408
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9517304620716581,
+      "learning_rate": 8.037020278498066e-06,
+      "loss": 0.8571,
+      "step": 3409
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9736458193938717,
+      "learning_rate": 8.031653899012098e-06,
+      "loss": 0.8782,
+      "step": 3410
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1201428152787472,
+      "learning_rate": 8.026288109153196e-06,
+      "loss": 0.8433,
+      "step": 3411
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1563941465651248,
+      "learning_rate": 8.020922910528717e-06,
+      "loss": 0.729,
+      "step": 3412
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4164416192647367,
+      "learning_rate": 8.01555830474582e-06,
+      "loss": 0.802,
+      "step": 3413
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7715536266677733,
+      "learning_rate": 8.010194293411502e-06,
+      "loss": 1.0084,
+      "step": 3414
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5943169891831952,
+      "learning_rate": 8.004830878132577e-06,
+      "loss": 0.4009,
+      "step": 3415
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1013565452151834,
+      "learning_rate": 7.999468060515686e-06,
+      "loss": 0.8229,
+      "step": 3416
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8902102649635893,
+      "learning_rate": 7.994105842167274e-06,
+      "loss": 0.7014,
+      "step": 3417
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2831882757771782,
+      "learning_rate": 7.988744224693625e-06,
+      "loss": 0.9036,
+      "step": 3418
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9499429030291879,
+      "learning_rate": 7.983383209700839e-06,
+      "loss": 1.0622,
+      "step": 3419
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2694596101248243,
+      "learning_rate": 7.978022798794825e-06,
+      "loss": 1.0057,
+      "step": 3420
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.122937076736453,
+      "learning_rate": 7.972662993581321e-06,
+      "loss": 0.5997,
+      "step": 3421
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9682049561740959,
+      "learning_rate": 7.967303795665887e-06,
+      "loss": 0.8651,
+      "step": 3422
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.882819299503901,
+      "learning_rate": 7.96194520665389e-06,
+      "loss": 0.9954,
+      "step": 3423
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9866442356607399,
+      "learning_rate": 7.956587228150519e-06,
+      "loss": 0.9517,
+      "step": 3424
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2138665914422484,
+      "learning_rate": 7.95122986176078e-06,
+      "loss": 0.7018,
+      "step": 3425
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.221464863731403,
+      "learning_rate": 7.945873109089503e-06,
+      "loss": 0.8941,
+      "step": 3426
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1608124685496235,
+      "learning_rate": 7.940516971741325e-06,
+      "loss": 0.8435,
+      "step": 3427
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.899682504396498,
+      "learning_rate": 7.935161451320696e-06,
+      "loss": 0.9435,
+      "step": 3428
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9731824361225241,
+      "learning_rate": 7.929806549431892e-06,
+      "loss": 0.5712,
+      "step": 3429
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0768190012613728,
+      "learning_rate": 7.924452267678998e-06,
+      "loss": 0.8686,
+      "step": 3430
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0849710459233797,
+      "learning_rate": 7.919098607665913e-06,
+      "loss": 0.7581,
+      "step": 3431
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3539669143848145,
+      "learning_rate": 7.913745570996348e-06,
+      "loss": 0.8978,
+      "step": 3432
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9734851093797534,
+      "learning_rate": 7.908393159273835e-06,
+      "loss": 1.0064,
+      "step": 3433
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7533613918124249,
+      "learning_rate": 7.903041374101707e-06,
+      "loss": 0.5286,
+      "step": 3434
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.049180412643251,
+      "learning_rate": 7.89769021708312e-06,
+      "loss": 0.8327,
+      "step": 3435
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9565848902977931,
+      "learning_rate": 7.892339689821034e-06,
+      "loss": 0.7699,
+      "step": 3436
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0570771039424856,
+      "learning_rate": 7.886989793918231e-06,
+      "loss": 0.8997,
+      "step": 3437
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9445491348714555,
+      "learning_rate": 7.881640530977287e-06,
+      "loss": 0.9841,
+      "step": 3438
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5458486954813382,
+      "learning_rate": 7.876291902600607e-06,
+      "loss": 0.922,
+      "step": 3439
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1008260800813308,
+      "learning_rate": 7.870943910390392e-06,
+      "loss": 0.6903,
+      "step": 3440
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8743233952773695,
+      "learning_rate": 7.865596555948657e-06,
+      "loss": 0.9754,
+      "step": 3441
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9254083505920138,
+      "learning_rate": 7.860249840877232e-06,
+      "loss": 0.6995,
+      "step": 3442
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.033476066829008,
+      "learning_rate": 7.854903766777741e-06,
+      "loss": 0.9931,
+      "step": 3443
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0838841556990115,
+      "learning_rate": 7.849558335251638e-06,
+      "loss": 0.8702,
+      "step": 3444
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0892696995743387,
+      "learning_rate": 7.844213547900156e-06,
+      "loss": 1.1204,
+      "step": 3445
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4135415513683804,
+      "learning_rate": 7.838869406324358e-06,
+      "loss": 0.7552,
+      "step": 3446
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0966985396495954,
+      "learning_rate": 7.83352591212511e-06,
+      "loss": 1.1221,
+      "step": 3447
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0479059645369042,
+      "learning_rate": 7.828183066903074e-06,
+      "loss": 0.9648,
+      "step": 3448
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0509451766315163,
+      "learning_rate": 7.822840872258725e-06,
+      "loss": 0.7784,
+      "step": 3449
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9957155135398429,
+      "learning_rate": 7.817499329792342e-06,
+      "loss": 0.6287,
+      "step": 3450
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5181922902895943,
+      "learning_rate": 7.812158441104013e-06,
+      "loss": 0.7184,
+      "step": 3451
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2479696404293268,
+      "learning_rate": 7.806818207793618e-06,
+      "loss": 1.0679,
+      "step": 3452
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.964476281766982,
+      "learning_rate": 7.801478631460853e-06,
+      "loss": 0.8359,
+      "step": 3453
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7749464262343253,
+      "learning_rate": 7.796139713705214e-06,
+      "loss": 0.4665,
+      "step": 3454
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1610269728047613,
+      "learning_rate": 7.790801456125991e-06,
+      "loss": 0.65,
+      "step": 3455
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7864578240372062,
+      "learning_rate": 7.785463860322292e-06,
+      "loss": 0.6302,
+      "step": 3456
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9386098233391048,
+      "learning_rate": 7.780126927893011e-06,
+      "loss": 1.0063,
+      "step": 3457
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.088199989813699,
+      "learning_rate": 7.774790660436857e-06,
+      "loss": 0.7932,
+      "step": 3458
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9940513446509534,
+      "learning_rate": 7.76945505955233e-06,
+      "loss": 0.7785,
+      "step": 3459
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.948763568643085,
+      "learning_rate": 7.764120126837731e-06,
+      "loss": 0.7682,
+      "step": 3460
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.08169880946672,
+      "learning_rate": 7.758785863891171e-06,
+      "loss": 1.1419,
+      "step": 3461
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.034286161205246,
+      "learning_rate": 7.753452272310544e-06,
+      "loss": 0.8073,
+      "step": 3462
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1227780296029495,
+      "learning_rate": 7.748119353693555e-06,
+      "loss": 0.8157,
+      "step": 3463
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8510820100158776,
+      "learning_rate": 7.742787109637706e-06,
+      "loss": 0.9614,
+      "step": 3464
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9645807797032421,
+      "learning_rate": 7.737455541740298e-06,
+      "loss": 0.6756,
+      "step": 3465
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.298682407607751,
+      "learning_rate": 7.732124651598418e-06,
+      "loss": 1.0379,
+      "step": 3466
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8046980063740515,
+      "learning_rate": 7.726794440808963e-06,
+      "loss": 0.9516,
+      "step": 3467
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1871152053440925,
+      "learning_rate": 7.721464910968628e-06,
+      "loss": 0.8805,
+      "step": 3468
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0268580968524244,
+      "learning_rate": 7.716136063673886e-06,
+      "loss": 1.0306,
+      "step": 3469
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.364364316892625,
+      "learning_rate": 7.710807900521023e-06,
+      "loss": 0.7378,
+      "step": 3470
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7035349173171206,
+      "learning_rate": 7.70548042310612e-06,
+      "loss": 0.5621,
+      "step": 3471
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0743259874924964,
+      "learning_rate": 7.700153633025041e-06,
+      "loss": 0.8279,
+      "step": 3472
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2100293493122432,
+      "learning_rate": 7.694827531873448e-06,
+      "loss": 0.7155,
+      "step": 3473
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0927274314055333,
+      "learning_rate": 7.689502121246809e-06,
+      "loss": 0.8889,
+      "step": 3474
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1948927949986754,
+      "learning_rate": 7.684177402740365e-06,
+      "loss": 0.7604,
+      "step": 3475
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9982353542317016,
+      "learning_rate": 7.678853377949164e-06,
+      "loss": 1.0402,
+      "step": 3476
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.115997595331086,
+      "learning_rate": 7.673530048468042e-06,
+      "loss": 0.8991,
+      "step": 3477
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.094700436235532,
+      "learning_rate": 7.668207415891625e-06,
+      "loss": 0.634,
+      "step": 3478
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9228249005650022,
+      "learning_rate": 7.662885481814338e-06,
+      "loss": 0.7552,
+      "step": 3479
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1134559917380378,
+      "learning_rate": 7.657564247830381e-06,
+      "loss": 0.8627,
+      "step": 3480
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0283356721540748,
+      "learning_rate": 7.652243715533757e-06,
+      "loss": 1.0682,
+      "step": 3481
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0127231522513496,
+      "learning_rate": 7.646923886518263e-06,
+      "loss": 0.8299,
+      "step": 3482
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2925774309048972,
+      "learning_rate": 7.64160476237747e-06,
+      "loss": 0.723,
+      "step": 3483
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.925477682263539,
+      "learning_rate": 7.636286344704747e-06,
+      "loss": 0.7297,
+      "step": 3484
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.989449550990648,
+      "learning_rate": 7.630968635093252e-06,
+      "loss": 0.785,
+      "step": 3485
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1441360145281065,
+      "learning_rate": 7.625651635135933e-06,
+      "loss": 1.1129,
+      "step": 3486
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.12175697654791,
+      "learning_rate": 7.620335346425516e-06,
+      "loss": 0.8939,
+      "step": 3487
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.591067997200619,
+      "learning_rate": 7.615019770554521e-06,
+      "loss": 0.4468,
+      "step": 3488
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8846677735968316,
+      "learning_rate": 7.609704909115254e-06,
+      "loss": 1.0424,
+      "step": 3489
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6771428000254063,
+      "learning_rate": 7.6043907636998026e-06,
+      "loss": 0.9972,
+      "step": 3490
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.032641191577613,
+      "learning_rate": 7.599077335900049e-06,
+      "loss": 0.9122,
+      "step": 3491
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1296722259105283,
+      "learning_rate": 7.5937646273076506e-06,
+      "loss": 0.7904,
+      "step": 3492
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5691367450703382,
+      "learning_rate": 7.588452639514059e-06,
+      "loss": 0.828,
+      "step": 3493
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7650929848089166,
+      "learning_rate": 7.583141374110495e-06,
+      "loss": 0.9541,
+      "step": 3494
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0664192497035012,
+      "learning_rate": 7.577830832687978e-06,
+      "loss": 1.1018,
+      "step": 3495
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4285100760783056,
+      "learning_rate": 7.57252101683731e-06,
+      "loss": 0.9043,
+      "step": 3496
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9072138359004546,
+      "learning_rate": 7.567211928149059e-06,
+      "loss": 0.7197,
+      "step": 3497
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.066815363853326,
+      "learning_rate": 7.561903568213595e-06,
+      "loss": 0.966,
+      "step": 3498
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0577241770378345,
+      "learning_rate": 7.556595938621058e-06,
+      "loss": 1.0597,
+      "step": 3499
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.053934464116249,
+      "learning_rate": 7.551289040961381e-06,
+      "loss": 0.727,
+      "step": 3500
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9261534537976049,
+      "learning_rate": 7.545982876824256e-06,
+      "loss": 0.7967,
+      "step": 3501
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9606843588657766,
+      "learning_rate": 7.540677447799176e-06,
+      "loss": 0.8153,
+      "step": 3502
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1316351565193983,
+      "learning_rate": 7.535372755475411e-06,
+      "loss": 0.8085,
+      "step": 3503
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.077257528547145,
+      "learning_rate": 7.530068801441995e-06,
+      "loss": 0.7056,
+      "step": 3504
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1653383836641833,
+      "learning_rate": 7.524765587287759e-06,
+      "loss": 1.1142,
+      "step": 3505
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6659617804845351,
+      "learning_rate": 7.519463114601304e-06,
+      "loss": 0.4693,
+      "step": 3506
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.225898278145531,
+      "learning_rate": 7.514161384971011e-06,
+      "loss": 0.8197,
+      "step": 3507
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.240254165904421,
+      "learning_rate": 7.508860399985035e-06,
+      "loss": 0.8395,
+      "step": 3508
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1508394277169756,
+      "learning_rate": 7.503560161231309e-06,
+      "loss": 0.8444,
+      "step": 3509
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9711562001794619,
+      "learning_rate": 7.4982606702975505e-06,
+      "loss": 1.3064,
+      "step": 3510
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8975712398525094,
+      "learning_rate": 7.492961928771236e-06,
+      "loss": 0.8761,
+      "step": 3511
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9771919115869505,
+      "learning_rate": 7.487663938239635e-06,
+      "loss": 0.7652,
+      "step": 3512
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2565027162780567,
+      "learning_rate": 7.482366700289785e-06,
+      "loss": 0.7476,
+      "step": 3513
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0358873183632735,
+      "learning_rate": 7.477070216508493e-06,
+      "loss": 1.0745,
+      "step": 3514
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.998540647993157,
+      "learning_rate": 7.471774488482347e-06,
+      "loss": 0.8002,
+      "step": 3515
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7590025198689008,
+      "learning_rate": 7.466479517797708e-06,
+      "loss": 0.9284,
+      "step": 3516
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4900545317237353,
+      "learning_rate": 7.461185306040712e-06,
+      "loss": 0.8469,
+      "step": 3517
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9500384506008089,
+      "learning_rate": 7.455891854797256e-06,
+      "loss": 0.9834,
+      "step": 3518
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0167504537708867,
+      "learning_rate": 7.450599165653022e-06,
+      "loss": 0.9819,
+      "step": 3519
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9035397086745371,
+      "learning_rate": 7.445307240193462e-06,
+      "loss": 0.4971,
+      "step": 3520
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0253179101196137,
+      "learning_rate": 7.440016080003791e-06,
+      "loss": 0.7376,
+      "step": 3521
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1660332955967367,
+      "learning_rate": 7.434725686669002e-06,
+      "loss": 0.936,
+      "step": 3522
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1696359479623042,
+      "learning_rate": 7.429436061773859e-06,
+      "loss": 0.7035,
+      "step": 3523
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7699029779285463,
+      "learning_rate": 7.4241472069028915e-06,
+      "loss": 0.6735,
+      "step": 3524
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9610862187030242,
+      "learning_rate": 7.4188591236404e-06,
+      "loss": 0.8507,
+      "step": 3525
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2289005963896722,
+      "learning_rate": 7.413571813570453e-06,
+      "loss": 0.8418,
+      "step": 3526
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9456637837375438,
+      "learning_rate": 7.408285278276894e-06,
+      "loss": 0.7671,
+      "step": 3527
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3547243899737,
+      "learning_rate": 7.402999519343319e-06,
+      "loss": 0.8591,
+      "step": 3528
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9232881580761105,
+      "learning_rate": 7.397714538353107e-06,
+      "loss": 0.8955,
+      "step": 3529
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1939263548380945,
+      "learning_rate": 7.392430336889398e-06,
+      "loss": 0.8722,
+      "step": 3530
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0651243327180082,
+      "learning_rate": 7.3871469165351015e-06,
+      "loss": 0.8296,
+      "step": 3531
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0134394697563571,
+      "learning_rate": 7.381864278872884e-06,
+      "loss": 1.1007,
+      "step": 3532
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0441292893740355,
+      "learning_rate": 7.376582425485188e-06,
+      "loss": 0.9787,
+      "step": 3533
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0226032908335383,
+      "learning_rate": 7.371301357954219e-06,
+      "loss": 0.7261,
+      "step": 3534
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0843494151729813,
+      "learning_rate": 7.366021077861937e-06,
+      "loss": 0.7227,
+      "step": 3535
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4926954825556773,
+      "learning_rate": 7.360741586790079e-06,
+      "loss": 0.9519,
+      "step": 3536
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1354298538670702,
+      "learning_rate": 7.355462886320142e-06,
+      "loss": 1.0041,
+      "step": 3537
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0891851410630227,
+      "learning_rate": 7.350184978033386e-06,
+      "loss": 1.1117,
+      "step": 3538
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8932800820531975,
+      "learning_rate": 7.3449078635108236e-06,
+      "loss": 0.6746,
+      "step": 3539
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0800321782514761,
+      "learning_rate": 7.33963154433325e-06,
+      "loss": 0.7617,
+      "step": 3540
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.79136480328299,
+      "learning_rate": 7.334356022081205e-06,
+      "loss": 0.945,
+      "step": 3541
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6997665350451271,
+      "learning_rate": 7.329081298334994e-06,
+      "loss": 0.3927,
+      "step": 3542
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0367072964002644,
+      "learning_rate": 7.323807374674687e-06,
+      "loss": 1.0811,
+      "step": 3543
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9366015153251074,
+      "learning_rate": 7.31853425268011e-06,
+      "loss": 0.654,
+      "step": 3544
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9715806213994898,
+      "learning_rate": 7.313261933930858e-06,
+      "loss": 0.8094,
+      "step": 3545
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0483786630278438,
+      "learning_rate": 7.307990420006266e-06,
+      "loss": 0.7357,
+      "step": 3546
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0824446156002678,
+      "learning_rate": 7.302719712485448e-06,
+      "loss": 0.8696,
+      "step": 3547
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0578837859155494,
+      "learning_rate": 7.2974498129472705e-06,
+      "loss": 1.019,
+      "step": 3548
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.019883116421873,
+      "learning_rate": 7.292180722970349e-06,
+      "loss": 0.684,
+      "step": 3549
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1310041739910772,
+      "learning_rate": 7.286912444133067e-06,
+      "loss": 1.0339,
+      "step": 3550
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9177390974203399,
+      "learning_rate": 7.28164497801356e-06,
+      "loss": 0.6795,
+      "step": 3551
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0828624541480127,
+      "learning_rate": 7.276378326189729e-06,
+      "loss": 1.1506,
+      "step": 3552
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1028240704950123,
+      "learning_rate": 7.271112490239213e-06,
+      "loss": 0.7596,
+      "step": 3553
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.870228163790839,
+      "learning_rate": 7.265847471739422e-06,
+      "loss": 0.8696,
+      "step": 3554
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7679977612936693,
+      "learning_rate": 7.26058327226752e-06,
+      "loss": 0.7173,
+      "step": 3555
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9879391323846994,
+      "learning_rate": 7.255319893400414e-06,
+      "loss": 0.9986,
+      "step": 3556
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9413606646879438,
+      "learning_rate": 7.250057336714779e-06,
+      "loss": 0.9615,
+      "step": 3557
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.937635641893687,
+      "learning_rate": 7.244795603787036e-06,
+      "loss": 0.6518,
+      "step": 3558
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0710050407093328,
+      "learning_rate": 7.2395346961933645e-06,
+      "loss": 0.7313,
+      "step": 3559
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6812414297886685,
+      "learning_rate": 7.234274615509686e-06,
+      "loss": 0.4721,
+      "step": 3560
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8902242612137228,
+      "learning_rate": 7.2290153633116875e-06,
+      "loss": 0.7722,
+      "step": 3561
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.109157753257538,
+      "learning_rate": 7.2237569411748045e-06,
+      "loss": 1.2415,
+      "step": 3562
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7240020369751446,
+      "learning_rate": 7.218499350674213e-06,
+      "loss": 1.0172,
+      "step": 3563
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9042903990283151,
+      "learning_rate": 7.2132425933848524e-06,
+      "loss": 0.8169,
+      "step": 3564
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8742967742347557,
+      "learning_rate": 7.207986670881409e-06,
+      "loss": 0.6237,
+      "step": 3565
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1163550262156405,
+      "learning_rate": 7.202731584738323e-06,
+      "loss": 0.7986,
+      "step": 3566
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.183956042318624,
+      "learning_rate": 7.19747733652977e-06,
+      "loss": 1.1409,
+      "step": 3567
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0509564813429466,
+      "learning_rate": 7.192223927829689e-06,
+      "loss": 0.7234,
+      "step": 3568
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5938111647637532,
+      "learning_rate": 7.186971360211766e-06,
+      "loss": 1.0185,
+      "step": 3569
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8742275935342135,
+      "learning_rate": 7.1817196352494245e-06,
+      "loss": 0.7044,
+      "step": 3570
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1455642756044997,
+      "learning_rate": 7.176468754515846e-06,
+      "loss": 0.7337,
+      "step": 3571
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0448002988556782,
+      "learning_rate": 7.171218719583956e-06,
+      "loss": 1.0302,
+      "step": 3572
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.113699344980786,
+      "learning_rate": 7.16596953202643e-06,
+      "loss": 0.7898,
+      "step": 3573
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0401942808746905,
+      "learning_rate": 7.160721193415681e-06,
+      "loss": 0.9615,
+      "step": 3574
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.125658825933631,
+      "learning_rate": 7.155473705323872e-06,
+      "loss": 0.8587,
+      "step": 3575
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0059092246204702,
+      "learning_rate": 7.150227069322916e-06,
+      "loss": 0.6934,
+      "step": 3576
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0010931899634041,
+      "learning_rate": 7.144981286984466e-06,
+      "loss": 1.2663,
+      "step": 3577
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0388234941418537,
+      "learning_rate": 7.139736359879916e-06,
+      "loss": 0.7038,
+      "step": 3578
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.581252050313118,
+      "learning_rate": 7.134492289580413e-06,
+      "loss": 0.3661,
+      "step": 3579
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8971731092696528,
+      "learning_rate": 7.129249077656844e-06,
+      "loss": 0.6812,
+      "step": 3580
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.7406288147825286,
+      "learning_rate": 7.124006725679828e-06,
+      "loss": 1.1146,
+      "step": 3581
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9344797190909121,
+      "learning_rate": 7.118765235219742e-06,
+      "loss": 0.7223,
+      "step": 3582
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.634680186131456,
+      "learning_rate": 7.113524607846702e-06,
+      "loss": 0.8553,
+      "step": 3583
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8999027386173148,
+      "learning_rate": 7.108284845130553e-06,
+      "loss": 0.7334,
+      "step": 3584
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1750670727721402,
+      "learning_rate": 7.103045948640895e-06,
+      "loss": 0.8445,
+      "step": 3585
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.325300878376302,
+      "learning_rate": 7.0978079199470596e-06,
+      "loss": 1.1436,
+      "step": 3586
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1222284143643493,
+      "learning_rate": 7.092570760618132e-06,
+      "loss": 0.7738,
+      "step": 3587
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6909887804020309,
+      "learning_rate": 7.087334472222916e-06,
+      "loss": 1.0583,
+      "step": 3588
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9327144557225862,
+      "learning_rate": 7.08209905632997e-06,
+      "loss": 0.6631,
+      "step": 3589
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0431689876688848,
+      "learning_rate": 7.0768645145075885e-06,
+      "loss": 1.022,
+      "step": 3590
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0413120826872473,
+      "learning_rate": 7.071630848323798e-06,
+      "loss": 0.6956,
+      "step": 3591
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9717080073977289,
+      "learning_rate": 7.0663980593463734e-06,
+      "loss": 0.7513,
+      "step": 3592
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3485099209029887,
+      "learning_rate": 7.061166149142814e-06,
+      "loss": 1.0654,
+      "step": 3593
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0790575661327477,
+      "learning_rate": 7.055935119280369e-06,
+      "loss": 0.7662,
+      "step": 3594
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6582984761005956,
+      "learning_rate": 7.05070497132601e-06,
+      "loss": 0.635,
+      "step": 3595
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9396362239924156,
+      "learning_rate": 7.0454757068464564e-06,
+      "loss": 0.8595,
+      "step": 3596
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9210722604690931,
+      "learning_rate": 7.040247327408161e-06,
+      "loss": 0.7548,
+      "step": 3597
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1245148438212187,
+      "learning_rate": 7.035019834577301e-06,
+      "loss": 0.7888,
+      "step": 3598
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9482295806548318,
+      "learning_rate": 7.0297932299198e-06,
+      "loss": 1.0388,
+      "step": 3599
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.817229913731691,
+      "learning_rate": 7.02456751500131e-06,
+      "loss": 1.0548,
+      "step": 3600
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1518376624746907,
+      "learning_rate": 7.019342691387225e-06,
+      "loss": 0.8326,
+      "step": 3601
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1047276259799883,
+      "learning_rate": 7.014118760642654e-06,
+      "loss": 0.8325,
+      "step": 3602
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1033412974437304,
+      "learning_rate": 7.008895724332454e-06,
+      "loss": 0.7385,
+      "step": 3603
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9288462323359623,
+      "learning_rate": 7.003673584021213e-06,
+      "loss": 0.6409,
+      "step": 3604
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9506580840631467,
+      "learning_rate": 6.998452341273241e-06,
+      "loss": 1.0532,
+      "step": 3605
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9770965553334282,
+      "learning_rate": 6.993231997652587e-06,
+      "loss": 0.7956,
+      "step": 3606
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0485983141254933,
+      "learning_rate": 6.988012554723029e-06,
+      "loss": 0.8993,
+      "step": 3607
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1254268507919445,
+      "learning_rate": 6.9827940140480776e-06,
+      "loss": 0.743,
+      "step": 3608
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.036977431249666,
+      "learning_rate": 6.977576377190969e-06,
+      "loss": 1.0052,
+      "step": 3609
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4854553115588416,
+      "learning_rate": 6.972359645714666e-06,
+      "loss": 0.7313,
+      "step": 3610
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9549590916566362,
+      "learning_rate": 6.967143821181872e-06,
+      "loss": 0.6792,
+      "step": 3611
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8722686162568937,
+      "learning_rate": 6.961928905155004e-06,
+      "loss": 1.011,
+      "step": 3612
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5829197746083424,
+      "learning_rate": 6.956714899196215e-06,
+      "loss": 0.9591,
+      "step": 3613
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8201644584623494,
+      "learning_rate": 6.951501804867392e-06,
+      "loss": 0.7035,
+      "step": 3614
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.88523231429842,
+      "learning_rate": 6.946289623730131e-06,
+      "loss": 0.76,
+      "step": 3615
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0803028153263847,
+      "learning_rate": 6.941078357345768e-06,
+      "loss": 0.6317,
+      "step": 3616
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0586876293312297,
+      "learning_rate": 6.935868007275364e-06,
+      "loss": 0.6981,
+      "step": 3617
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9703109860675695,
+      "learning_rate": 6.930658575079705e-06,
+      "loss": 0.8841,
+      "step": 3618
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1255870301638387,
+      "learning_rate": 6.925450062319294e-06,
+      "loss": 1.1509,
+      "step": 3619
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9341448889393364,
+      "learning_rate": 6.920242470554366e-06,
+      "loss": 0.7256,
+      "step": 3620
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0066293345107054,
+      "learning_rate": 6.915035801344887e-06,
+      "loss": 1.0093,
+      "step": 3621
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8237307164471248,
+      "learning_rate": 6.909830056250527e-06,
+      "loss": 0.5261,
+      "step": 3622
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2331986292803703,
+      "learning_rate": 6.904625236830697e-06,
+      "loss": 0.8384,
+      "step": 3623
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1368506771550728,
+      "learning_rate": 6.899421344644525e-06,
+      "loss": 1.1499,
+      "step": 3624
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0309267578624182,
+      "learning_rate": 6.894218381250859e-06,
+      "loss": 0.7135,
+      "step": 3625
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1097955253810752,
+      "learning_rate": 6.889016348208272e-06,
+      "loss": 1.0684,
+      "step": 3626
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8731029475905171,
+      "learning_rate": 6.883815247075052e-06,
+      "loss": 0.8037,
+      "step": 3627
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9698311968995359,
+      "learning_rate": 6.878615079409221e-06,
+      "loss": 0.8155,
+      "step": 3628
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1056418670716297,
+      "learning_rate": 6.873415846768503e-06,
+      "loss": 1.0877,
+      "step": 3629
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0623077237001863,
+      "learning_rate": 6.868217550710359e-06,
+      "loss": 0.7692,
+      "step": 3630
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9571684042542731,
+      "learning_rate": 6.863020192791959e-06,
+      "loss": 0.7716,
+      "step": 3631
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1904871147451523,
+      "learning_rate": 6.8578237745702e-06,
+      "loss": 0.9136,
+      "step": 3632
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6059306051673055,
+      "learning_rate": 6.852628297601686e-06,
+      "loss": 0.5878,
+      "step": 3633
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2576299379935156,
+      "learning_rate": 6.8474337634427505e-06,
+      "loss": 0.8108,
+      "step": 3634
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9784913135002808,
+      "learning_rate": 6.842240173649442e-06,
+      "loss": 0.9632,
+      "step": 3635
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6063589695146951,
+      "learning_rate": 6.837047529777516e-06,
+      "loss": 1.0138,
+      "step": 3636
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0525548091798165,
+      "learning_rate": 6.831855833382459e-06,
+      "loss": 0.689,
+      "step": 3637
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8892566967883045,
+      "learning_rate": 6.826665086019467e-06,
+      "loss": 0.9866,
+      "step": 3638
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9744928340920117,
+      "learning_rate": 6.821475289243451e-06,
+      "loss": 0.7257,
+      "step": 3639
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0620501014294446,
+      "learning_rate": 6.816286444609037e-06,
+      "loss": 0.9384,
+      "step": 3640
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9768906860791315,
+      "learning_rate": 6.81109855367057e-06,
+      "loss": 0.7454,
+      "step": 3641
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0177201591427496,
+      "learning_rate": 6.805911617982105e-06,
+      "loss": 1.1417,
+      "step": 3642
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0836611680999735,
+      "learning_rate": 6.800725639097412e-06,
+      "loss": 1.041,
+      "step": 3643
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3116044764995494,
+      "learning_rate": 6.795540618569974e-06,
+      "loss": 0.7407,
+      "step": 3644
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2633501761083972,
+      "learning_rate": 6.7903565579529875e-06,
+      "loss": 0.8016,
+      "step": 3645
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0355552897509348,
+      "learning_rate": 6.785173458799365e-06,
+      "loss": 0.8137,
+      "step": 3646
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.345779144058093,
+      "learning_rate": 6.779991322661724e-06,
+      "loss": 0.7515,
+      "step": 3647
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.025836248888335,
+      "learning_rate": 6.774810151092397e-06,
+      "loss": 1.0657,
+      "step": 3648
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0395092345705852,
+      "learning_rate": 6.7696299456434325e-06,
+      "loss": 0.8922,
+      "step": 3649
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0407711807801316,
+      "learning_rate": 6.764450707866577e-06,
+      "loss": 0.723,
+      "step": 3650
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5670369048738279,
+      "learning_rate": 6.759272439313299e-06,
+      "loss": 0.4437,
+      "step": 3651
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0974026553356686,
+      "learning_rate": 6.754095141534773e-06,
+      "loss": 1.041,
+      "step": 3652
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8284255416818612,
+      "learning_rate": 6.748918816081884e-06,
+      "loss": 0.909,
+      "step": 3653
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.211021806479717,
+      "learning_rate": 6.743743464505219e-06,
+      "loss": 0.7656,
+      "step": 3654
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.279788160660822,
+      "learning_rate": 6.73856908835508e-06,
+      "loss": 0.8426,
+      "step": 3655
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6647111012525908,
+      "learning_rate": 6.73339568918148e-06,
+      "loss": 0.6356,
+      "step": 3656
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.984331722708486,
+      "learning_rate": 6.728223268534127e-06,
+      "loss": 1.0621,
+      "step": 3657
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.081953015751451,
+      "learning_rate": 6.723051827962445e-06,
+      "loss": 0.8744,
+      "step": 3658
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8571964979980581,
+      "learning_rate": 6.717881369015565e-06,
+      "loss": 0.9671,
+      "step": 3659
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0563698665992103,
+      "learning_rate": 6.712711893242325e-06,
+      "loss": 0.8203,
+      "step": 3660
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9450320104829277,
+      "learning_rate": 6.707543402191254e-06,
+      "loss": 0.6636,
+      "step": 3661
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.083272603004192,
+      "learning_rate": 6.7023758974106065e-06,
+      "loss": 1.1226,
+      "step": 3662
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9291435590267441,
+      "learning_rate": 6.697209380448333e-06,
+      "loss": 0.9849,
+      "step": 3663
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6920443400616644,
+      "learning_rate": 6.6920438528520794e-06,
+      "loss": 0.9518,
+      "step": 3664
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9217843540927427,
+      "learning_rate": 6.686879316169206e-06,
+      "loss": 0.9083,
+      "step": 3665
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1720554825878056,
+      "learning_rate": 6.681715771946777e-06,
+      "loss": 0.8075,
+      "step": 3666
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.062620200025732,
+      "learning_rate": 6.676553221731556e-06,
+      "loss": 1.0853,
+      "step": 3667
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6625513730896557,
+      "learning_rate": 6.671391667070002e-06,
+      "loss": 0.4361,
+      "step": 3668
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.30769817135308,
+      "learning_rate": 6.666231109508287e-06,
+      "loss": 0.8884,
+      "step": 3669
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0317524740771475,
+      "learning_rate": 6.661071550592284e-06,
+      "loss": 0.8449,
+      "step": 3670
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9304312399333665,
+      "learning_rate": 6.655912991867551e-06,
+      "loss": 0.8012,
+      "step": 3671
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.379890452834049,
+      "learning_rate": 6.650755434879366e-06,
+      "loss": 0.9621,
+      "step": 3672
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8956328477970117,
+      "learning_rate": 6.645598881172698e-06,
+      "loss": 0.6794,
+      "step": 3673
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0306544842840113,
+      "learning_rate": 6.6404433322922146e-06,
+      "loss": 0.7493,
+      "step": 3674
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0827875235861952,
+      "learning_rate": 6.635288789782286e-06,
+      "loss": 0.9002,
+      "step": 3675
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4072600034924114,
+      "learning_rate": 6.630135255186976e-06,
+      "loss": 1.1314,
+      "step": 3676
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.969138957489288,
+      "learning_rate": 6.624982730050054e-06,
+      "loss": 0.7075,
+      "step": 3677
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.223774197539845,
+      "learning_rate": 6.619831215914974e-06,
+      "loss": 0.9919,
+      "step": 3678
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.269717586374198,
+      "learning_rate": 6.614680714324904e-06,
+      "loss": 0.7683,
+      "step": 3679
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2487382464486867,
+      "learning_rate": 6.6095312268226955e-06,
+      "loss": 0.7529,
+      "step": 3680
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1185850768630798,
+      "learning_rate": 6.604382754950907e-06,
+      "loss": 1.1296,
+      "step": 3681
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8877489462529845,
+      "learning_rate": 6.599235300251778e-06,
+      "loss": 0.6387,
+      "step": 3682
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.248858494913935,
+      "learning_rate": 6.594088864267259e-06,
+      "loss": 0.8526,
+      "step": 3683
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7689079286045878,
+      "learning_rate": 6.588943448538989e-06,
+      "loss": 1.1215,
+      "step": 3684
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3573198532917141,
+      "learning_rate": 6.583799054608293e-06,
+      "loss": 0.8739,
+      "step": 3685
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.911278474322848,
+      "learning_rate": 6.578655684016204e-06,
+      "loss": 0.9838,
+      "step": 3686
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7216822597614437,
+      "learning_rate": 6.57351333830344e-06,
+      "loss": 0.4698,
+      "step": 3687
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8660099273071503,
+      "learning_rate": 6.568372019010419e-06,
+      "loss": 0.8298,
+      "step": 3688
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9210920125225651,
+      "learning_rate": 6.56323172767724e-06,
+      "loss": 0.6636,
+      "step": 3689
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6617735434689588,
+      "learning_rate": 6.558092465843704e-06,
+      "loss": 1.034,
+      "step": 3690
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7908614939052576,
+      "learning_rate": 6.552954235049301e-06,
+      "loss": 0.9803,
+      "step": 3691
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.12607361629952,
+      "learning_rate": 6.547817036833208e-06,
+      "loss": 0.5815,
+      "step": 3692
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9909713613764479,
+      "learning_rate": 6.5426808727343e-06,
+      "loss": 0.7396,
+      "step": 3693
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8642500444069043,
+      "learning_rate": 6.537545744291134e-06,
+      "loss": 0.7436,
+      "step": 3694
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.399281938751146,
+      "learning_rate": 6.532411653041969e-06,
+      "loss": 1.197,
+      "step": 3695
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8923281843094264,
+      "learning_rate": 6.5272786005247365e-06,
+      "loss": 0.6063,
+      "step": 3696
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.231795787364977,
+      "learning_rate": 6.522146588277071e-06,
+      "loss": 0.8892,
+      "step": 3697
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7267663245871966,
+      "learning_rate": 6.517015617836292e-06,
+      "loss": 0.591,
+      "step": 3698
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0079311850921573,
+      "learning_rate": 6.511885690739399e-06,
+      "loss": 0.6098,
+      "step": 3699
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.072861864175587,
+      "learning_rate": 6.5067568085230896e-06,
+      "loss": 0.7396,
+      "step": 3700
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8412331374490467,
+      "learning_rate": 6.50162897272374e-06,
+      "loss": 1.014,
+      "step": 3701
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1777152152326995,
+      "learning_rate": 6.496502184877426e-06,
+      "loss": 0.95,
+      "step": 3702
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.991971357569563,
+      "learning_rate": 6.491376446519888e-06,
+      "loss": 0.8416,
+      "step": 3703
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.988466723616622,
+      "learning_rate": 6.486251759186573e-06,
+      "loss": 0.9345,
+      "step": 3704
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8801398603953885,
+      "learning_rate": 6.481128124412604e-06,
+      "loss": 0.9233,
+      "step": 3705
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7314090284812111,
+      "learning_rate": 6.476005543732783e-06,
+      "loss": 0.4666,
+      "step": 3706
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5039900804196848,
+      "learning_rate": 6.470884018681609e-06,
+      "loss": 0.8628,
+      "step": 3707
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2635961544117935,
+      "learning_rate": 6.465763550793252e-06,
+      "loss": 0.7419,
+      "step": 3708
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9646965880527195,
+      "learning_rate": 6.460644141601582e-06,
+      "loss": 1.1088,
+      "step": 3709
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9496596253342767,
+      "learning_rate": 6.45552579264013e-06,
+      "loss": 1.0966,
+      "step": 3710
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1717632384127294,
+      "learning_rate": 6.450408505442124e-06,
+      "loss": 0.8988,
+      "step": 3711
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6454706723030631,
+      "learning_rate": 6.4452922815404775e-06,
+      "loss": 0.9689,
+      "step": 3712
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9122734067326401,
+      "learning_rate": 6.440177122467769e-06,
+      "loss": 0.8078,
+      "step": 3713
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8939039476452002,
+      "learning_rate": 6.4350630297562726e-06,
+      "loss": 0.9726,
+      "step": 3714
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0003475846526722,
+      "learning_rate": 6.429950004937941e-06,
+      "loss": 0.6311,
+      "step": 3715
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1168109199244762,
+      "learning_rate": 6.424838049544398e-06,
+      "loss": 0.9305,
+      "step": 3716
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0787138767744426,
+      "learning_rate": 6.419727165106957e-06,
+      "loss": 0.7011,
+      "step": 3717
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0912816910050342,
+      "learning_rate": 6.414617353156605e-06,
+      "loss": 0.8334,
+      "step": 3718
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9593931142036785,
+      "learning_rate": 6.409508615224014e-06,
+      "loss": 0.7557,
+      "step": 3719
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.121610191984423,
+      "learning_rate": 6.404400952839522e-06,
+      "loss": 1.0449,
+      "step": 3720
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2474617111049502,
+      "learning_rate": 6.3992943675331576e-06,
+      "loss": 0.8608,
+      "step": 3721
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6005671218028804,
+      "learning_rate": 6.394188860834626e-06,
+      "loss": 0.4903,
+      "step": 3722
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.093468608305476,
+      "learning_rate": 6.3890844342732965e-06,
+      "loss": 0.9138,
+      "step": 3723
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8944891706241679,
+      "learning_rate": 6.383981089378226e-06,
+      "loss": 0.9654,
+      "step": 3724
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5081277083256264,
+      "learning_rate": 6.378878827678147e-06,
+      "loss": 0.9016,
+      "step": 3725
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9883033222690807,
+      "learning_rate": 6.373777650701465e-06,
+      "loss": 0.7109,
+      "step": 3726
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1705094310240352,
+      "learning_rate": 6.36867755997626e-06,
+      "loss": 0.7801,
+      "step": 3727
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.003632157954317,
+      "learning_rate": 6.363578557030285e-06,
+      "loss": 0.9678,
+      "step": 3728
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9344218396404721,
+      "learning_rate": 6.358480643390976e-06,
+      "loss": 1.1735,
+      "step": 3729
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2469106670740053,
+      "learning_rate": 6.35338382058543e-06,
+      "loss": 0.9643,
+      "step": 3730
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.000755930754461,
+      "learning_rate": 6.348288090140424e-06,
+      "loss": 0.7243,
+      "step": 3731
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9753031885707772,
+      "learning_rate": 6.343193453582411e-06,
+      "loss": 1.0149,
+      "step": 3732
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8886451502245801,
+      "learning_rate": 6.3380999124375146e-06,
+      "loss": 0.7951,
+      "step": 3733
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9280334688855512,
+      "learning_rate": 6.333007468231521e-06,
+      "loss": 1.0084,
+      "step": 3734
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.126909635450986,
+      "learning_rate": 6.327916122489899e-06,
+      "loss": 0.967,
+      "step": 3735
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6272003036559206,
+      "learning_rate": 6.322825876737788e-06,
+      "loss": 0.9671,
+      "step": 3736
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8929563928059903,
+      "learning_rate": 6.317736732499986e-06,
+      "loss": 0.7133,
+      "step": 3737
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8640026644729507,
+      "learning_rate": 6.312648691300975e-06,
+      "loss": 0.7194,
+      "step": 3738
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0435495522977443,
+      "learning_rate": 6.307561754664903e-06,
+      "loss": 1.1008,
+      "step": 3739
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2047726118853763,
+      "learning_rate": 6.302475924115581e-06,
+      "loss": 0.807,
+      "step": 3740
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5233436522294875,
+      "learning_rate": 6.297391201176495e-06,
+      "loss": 0.5149,
+      "step": 3741
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.756892410559082,
+      "learning_rate": 6.292307587370794e-06,
+      "loss": 0.7064,
+      "step": 3742
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9423599746639296,
+      "learning_rate": 6.287225084221301e-06,
+      "loss": 1.0351,
+      "step": 3743
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0440889482844682,
+      "learning_rate": 6.282143693250502e-06,
+      "loss": 0.9011,
+      "step": 3744
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6856902104542558,
+      "learning_rate": 6.277063415980549e-06,
+      "loss": 0.5213,
+      "step": 3745
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.982621223739393,
+      "learning_rate": 6.271984253933266e-06,
+      "loss": 0.7068,
+      "step": 3746
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1287960897427916,
+      "learning_rate": 6.266906208630138e-06,
+      "loss": 0.8811,
+      "step": 3747
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.077030365712337,
+      "learning_rate": 6.261829281592313e-06,
+      "loss": 1.0304,
+      "step": 3748
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8853561765583389,
+      "learning_rate": 6.256753474340611e-06,
+      "loss": 0.7474,
+      "step": 3749
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.20618551483249,
+      "learning_rate": 6.2516787883955165e-06,
+      "loss": 1.1659,
+      "step": 3750
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0474075736438426,
+      "learning_rate": 6.246605225277169e-06,
+      "loss": 0.8587,
+      "step": 3751
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9692657004062373,
+      "learning_rate": 6.2415327865053776e-06,
+      "loss": 0.719,
+      "step": 3752
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0189414140151156,
+      "learning_rate": 6.236461473599618e-06,
+      "loss": 0.9965,
+      "step": 3753
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.008829529387171,
+      "learning_rate": 6.231391288079027e-06,
+      "loss": 0.9042,
+      "step": 3754
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0418682471473992,
+      "learning_rate": 6.226322231462394e-06,
+      "loss": 0.7269,
+      "step": 3755
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8695703677162558,
+      "learning_rate": 6.2212543052681865e-06,
+      "loss": 0.8068,
+      "step": 3756
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9284459814301727,
+      "learning_rate": 6.21618751101452e-06,
+      "loss": 0.9274,
+      "step": 3757
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6675233585229824,
+      "learning_rate": 6.211121850219175e-06,
+      "loss": 0.6608,
+      "step": 3758
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9814123291092884,
+      "learning_rate": 6.206057324399599e-06,
+      "loss": 0.6937,
+      "step": 3759
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6703954390151159,
+      "learning_rate": 6.2009939350728865e-06,
+      "loss": 0.9619,
+      "step": 3760
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9414015941569652,
+      "learning_rate": 6.195931683755807e-06,
+      "loss": 0.6212,
+      "step": 3761
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8067148799224411,
+      "learning_rate": 6.1908705719647735e-06,
+      "loss": 0.9265,
+      "step": 3762
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1257916946799407,
+      "learning_rate": 6.185810601215867e-06,
+      "loss": 0.9396,
+      "step": 3763
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0832043423340318,
+      "learning_rate": 6.180751773024831e-06,
+      "loss": 0.8145,
+      "step": 3764
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9796284470175504,
+      "learning_rate": 6.175694088907053e-06,
+      "loss": 0.6427,
+      "step": 3765
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.078770796444696,
+      "learning_rate": 6.170637550377587e-06,
+      "loss": 0.9042,
+      "step": 3766
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.902256554267194,
+      "learning_rate": 6.165582158951144e-06,
+      "loss": 1.0461,
+      "step": 3767
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1471914093032791,
+      "learning_rate": 6.160527916142093e-06,
+      "loss": 0.8796,
+      "step": 3768
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.025278981578596,
+      "learning_rate": 6.155474823464446e-06,
+      "loss": 0.7205,
+      "step": 3769
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1707149163554122,
+      "learning_rate": 6.150422882431888e-06,
+      "loss": 0.6932,
+      "step": 3770
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.825888871769115,
+      "learning_rate": 6.145372094557751e-06,
+      "loss": 0.5977,
+      "step": 3771
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9855270679348502,
+      "learning_rate": 6.1403224613550164e-06,
+      "loss": 0.9937,
+      "step": 3772
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9799416062356883,
+      "learning_rate": 6.13527398433633e-06,
+      "loss": 1.0465,
+      "step": 3773
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9232577727929345,
+      "learning_rate": 6.130226665013979e-06,
+      "loss": 0.6564,
+      "step": 3774
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6196529332459491,
+      "learning_rate": 6.1251805048999215e-06,
+      "loss": 0.6089,
+      "step": 3775
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9718287548904565,
+      "learning_rate": 6.1201355055057486e-06,
+      "loss": 0.7885,
+      "step": 3776
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0423993199060562,
+      "learning_rate": 6.115091668342717e-06,
+      "loss": 1.0367,
+      "step": 3777
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9206095028861214,
+      "learning_rate": 6.110048994921735e-06,
+      "loss": 0.7411,
+      "step": 3778
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.07225489940505,
+      "learning_rate": 6.10500748675335e-06,
+      "loss": 0.8354,
+      "step": 3779
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9158742254190444,
+      "learning_rate": 6.09996714534777e-06,
+      "loss": 1.0673,
+      "step": 3780
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9845170100737726,
+      "learning_rate": 6.094927972214858e-06,
+      "loss": 1.0039,
+      "step": 3781
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0709251078094004,
+      "learning_rate": 6.089889968864122e-06,
+      "loss": 0.8843,
+      "step": 3782
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0010150373758298,
+      "learning_rate": 6.084853136804711e-06,
+      "loss": 0.7205,
+      "step": 3783
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1940488839070138,
+      "learning_rate": 6.079817477545436e-06,
+      "loss": 0.8002,
+      "step": 3784
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0229259812604365,
+      "learning_rate": 6.074782992594757e-06,
+      "loss": 0.7543,
+      "step": 3785
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6779146971142503,
+      "learning_rate": 6.069749683460765e-06,
+      "loss": 1.014,
+      "step": 3786
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9393434809069263,
+      "learning_rate": 6.064717551651218e-06,
+      "loss": 0.8273,
+      "step": 3787
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9798647249423171,
+      "learning_rate": 6.059686598673515e-06,
+      "loss": 0.6587,
+      "step": 3788
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.043542023238313,
+      "learning_rate": 6.054656826034699e-06,
+      "loss": 0.8542,
+      "step": 3789
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9097147008262962,
+      "learning_rate": 6.049628235241459e-06,
+      "loss": 0.7735,
+      "step": 3790
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.123546334372189,
+      "learning_rate": 6.044600827800138e-06,
+      "loss": 1.0847,
+      "step": 3791
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1055477714084214,
+      "learning_rate": 6.039574605216715e-06,
+      "loss": 1.0317,
+      "step": 3792
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.000809339320539,
+      "learning_rate": 6.0345495689968195e-06,
+      "loss": 0.7518,
+      "step": 3793
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3067840510875806,
+      "learning_rate": 6.029525720645719e-06,
+      "loss": 1.0039,
+      "step": 3794
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5916125909636643,
+      "learning_rate": 6.024503061668337e-06,
+      "loss": 0.654,
+      "step": 3795
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1564440045621212,
+      "learning_rate": 6.019481593569234e-06,
+      "loss": 0.8718,
+      "step": 3796
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.774174601738248,
+      "learning_rate": 6.014461317852606e-06,
+      "loss": 0.5643,
+      "step": 3797
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9118460305899551,
+      "learning_rate": 6.009442236022307e-06,
+      "loss": 0.7844,
+      "step": 3798
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2059172578322208,
+      "learning_rate": 6.0044243495818275e-06,
+      "loss": 0.77,
+      "step": 3799
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.195528318607199,
+      "learning_rate": 5.999407660034289e-06,
+      "loss": 1.1456,
+      "step": 3800
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.269174417373199,
+      "learning_rate": 5.99439216888247e-06,
+      "loss": 0.8707,
+      "step": 3801
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8745123651755461,
+      "learning_rate": 5.989377877628783e-06,
+      "loss": 0.7006,
+      "step": 3802
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.16472224043871,
+      "learning_rate": 5.984364787775285e-06,
+      "loss": 0.8992,
+      "step": 3803
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.818600259099038,
+      "learning_rate": 5.979352900823663e-06,
+      "loss": 0.9058,
+      "step": 3804
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.938078416987596,
+      "learning_rate": 5.974342218275256e-06,
+      "loss": 0.9685,
+      "step": 3805
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2181291963138783,
+      "learning_rate": 5.969332741631037e-06,
+      "loss": 0.9761,
+      "step": 3806
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8691842519968884,
+      "learning_rate": 5.964324472391612e-06,
+      "loss": 0.7633,
+      "step": 3807
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2148226981348225,
+      "learning_rate": 5.959317412057238e-06,
+      "loss": 0.9129,
+      "step": 3808
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9797495380241293,
+      "learning_rate": 5.954311562127797e-06,
+      "loss": 0.8526,
+      "step": 3809
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9866628162021629,
+      "learning_rate": 5.949306924102822e-06,
+      "loss": 1.0406,
+      "step": 3810
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9350881209737607,
+      "learning_rate": 5.9443034994814655e-06,
+      "loss": 0.9824,
+      "step": 3811
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0192681398931598,
+      "learning_rate": 5.93930128976253e-06,
+      "loss": 0.6345,
+      "step": 3812
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.571130914658998,
+      "learning_rate": 5.9343002964444554e-06,
+      "loss": 0.4655,
+      "step": 3813
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9303958279373661,
+      "learning_rate": 5.929300521025304e-06,
+      "loss": 1.0015,
+      "step": 3814
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.008253284799995,
+      "learning_rate": 5.9243019650027855e-06,
+      "loss": 1.1141,
+      "step": 3815
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0396444685728514,
+      "learning_rate": 5.919304629874239e-06,
+      "loss": 0.6378,
+      "step": 3816
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1554146112324706,
+      "learning_rate": 5.914308517136644e-06,
+      "loss": 0.744,
+      "step": 3817
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1460966753960873,
+      "learning_rate": 5.9093136282866014e-06,
+      "loss": 0.9525,
+      "step": 3818
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.838434671991771,
+      "learning_rate": 5.904319964820356e-06,
+      "loss": 0.7731,
+      "step": 3819
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1559241798118813,
+      "learning_rate": 5.899327528233787e-06,
+      "loss": 1.16,
+      "step": 3820
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0147704704955807,
+      "learning_rate": 5.894336320022394e-06,
+      "loss": 0.5907,
+      "step": 3821
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8082397948447285,
+      "learning_rate": 5.889346341681319e-06,
+      "loss": 0.7609,
+      "step": 3822
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9809612708710731,
+      "learning_rate": 5.884357594705339e-06,
+      "loss": 0.708,
+      "step": 3823
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1200586900929559,
+      "learning_rate": 5.8793700805888445e-06,
+      "loss": 1.137,
+      "step": 3824
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1119150188308784,
+      "learning_rate": 5.874383800825878e-06,
+      "loss": 0.8733,
+      "step": 3825
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0333612490057362,
+      "learning_rate": 5.869398756910095e-06,
+      "loss": 0.8033,
+      "step": 3826
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9975741277107979,
+      "learning_rate": 5.864414950334796e-06,
+      "loss": 0.7756,
+      "step": 3827
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8340788641446196,
+      "learning_rate": 5.859432382592895e-06,
+      "loss": 0.9321,
+      "step": 3828
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9940666576208482,
+      "learning_rate": 5.854451055176948e-06,
+      "loss": 1.0745,
+      "step": 3829
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9571308787120658,
+      "learning_rate": 5.8494709695791364e-06,
+      "loss": 0.6694,
+      "step": 3830
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1896058556793412,
+      "learning_rate": 5.8444921272912595e-06,
+      "loss": 0.8838,
+      "step": 3831
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6085295981188178,
+      "learning_rate": 5.839514529804757e-06,
+      "loss": 0.4754,
+      "step": 3832
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.24478768651467,
+      "learning_rate": 5.834538178610694e-06,
+      "loss": 0.6801,
+      "step": 3833
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.146361849435458,
+      "learning_rate": 5.829563075199753e-06,
+      "loss": 1.0104,
+      "step": 3834
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6883147300001295,
+      "learning_rate": 5.824589221062258e-06,
+      "loss": 0.9399,
+      "step": 3835
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.073143206389282,
+      "learning_rate": 5.819616617688139e-06,
+      "loss": 0.7999,
+      "step": 3836
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8132418398278521,
+      "learning_rate": 5.814645266566971e-06,
+      "loss": 0.879,
+      "step": 3837
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1152999017898415,
+      "learning_rate": 5.809675169187936e-06,
+      "loss": 0.7734,
+      "step": 3838
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5949356730375583,
+      "learning_rate": 5.8047063270398566e-06,
+      "loss": 1.1414,
+      "step": 3839
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1159644090473055,
+      "learning_rate": 5.7997387416111685e-06,
+      "loss": 0.6546,
+      "step": 3840
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0992133901224388,
+      "learning_rate": 5.79477241438994e-06,
+      "loss": 0.8781,
+      "step": 3841
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.213410244317632,
+      "learning_rate": 5.789807346863849e-06,
+      "loss": 0.7961,
+      "step": 3842
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9357479388815854,
+      "learning_rate": 5.7848435405202096e-06,
+      "loss": 1.0125,
+      "step": 3843
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0538133889392334,
+      "learning_rate": 5.779880996845954e-06,
+      "loss": 0.9098,
+      "step": 3844
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4893755801711108,
+      "learning_rate": 5.77491971732763e-06,
+      "loss": 0.8136,
+      "step": 3845
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.162294398480465,
+      "learning_rate": 5.769959703451413e-06,
+      "loss": 0.8186,
+      "step": 3846
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0333836061165094,
+      "learning_rate": 5.765000956703099e-06,
+      "loss": 0.5514,
+      "step": 3847
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2159100926398119,
+      "learning_rate": 5.760043478568106e-06,
+      "loss": 1.0672,
+      "step": 3848
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.642980772633582,
+      "learning_rate": 5.755087270531463e-06,
+      "loss": 0.4631,
+      "step": 3849
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0022403698504883,
+      "learning_rate": 5.750132334077827e-06,
+      "loss": 0.8768,
+      "step": 3850
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.192113073661423,
+      "learning_rate": 5.745178670691478e-06,
+      "loss": 0.748,
+      "step": 3851
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9225151726276023,
+      "learning_rate": 5.740226281856298e-06,
+      "loss": 0.7241,
+      "step": 3852
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7208544293023331,
+      "learning_rate": 5.7352751690558025e-06,
+      "loss": 0.9532,
+      "step": 3853
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9798626249476661,
+      "learning_rate": 5.730325333773122e-06,
+      "loss": 0.8417,
+      "step": 3854
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0679700976556534,
+      "learning_rate": 5.725376777491005e-06,
+      "loss": 0.7992,
+      "step": 3855
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.951328672021993,
+      "learning_rate": 5.7204295016918045e-06,
+      "loss": 0.8415,
+      "step": 3856
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4462164482520148,
+      "learning_rate": 5.715483507857506e-06,
+      "loss": 0.7314,
+      "step": 3857
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0865749084072152,
+      "learning_rate": 5.710538797469707e-06,
+      "loss": 1.0295,
+      "step": 3858
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0134418206755995,
+      "learning_rate": 5.70559537200961e-06,
+      "loss": 1.059,
+      "step": 3859
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7294064382899136,
+      "learning_rate": 5.700653232958047e-06,
+      "loss": 0.9575,
+      "step": 3860
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9123629837118763,
+      "learning_rate": 5.695712381795455e-06,
+      "loss": 0.7422,
+      "step": 3861
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0112375371092635,
+      "learning_rate": 5.690772820001894e-06,
+      "loss": 1.004,
+      "step": 3862
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2935029601231187,
+      "learning_rate": 5.685834549057025e-06,
+      "loss": 0.7816,
+      "step": 3863
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9462260704593165,
+      "learning_rate": 5.680897570440138e-06,
+      "loss": 0.7278,
+      "step": 3864
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9783284136393098,
+      "learning_rate": 5.675961885630117e-06,
+      "loss": 0.699,
+      "step": 3865
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9824119004739683,
+      "learning_rate": 5.671027496105477e-06,
+      "loss": 0.6968,
+      "step": 3866
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5779121723640087,
+      "learning_rate": 5.666094403344336e-06,
+      "loss": 0.6282,
+      "step": 3867
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9020820791272363,
+      "learning_rate": 5.66116260882442e-06,
+      "loss": 0.8344,
+      "step": 3868
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1846468134618555,
+      "learning_rate": 5.656232114023077e-06,
+      "loss": 0.7483,
+      "step": 3869
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2669358246784979,
+      "learning_rate": 5.651302920417252e-06,
+      "loss": 0.8506,
+      "step": 3870
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9751835848531937,
+      "learning_rate": 5.646375029483513e-06,
+      "loss": 0.7562,
+      "step": 3871
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9388560518732711,
+      "learning_rate": 5.641448442698032e-06,
+      "loss": 1.0767,
+      "step": 3872
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7609658284216911,
+      "learning_rate": 5.636523161536587e-06,
+      "loss": 0.5902,
+      "step": 3873
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2723005497202313,
+      "learning_rate": 5.63159918747457e-06,
+      "loss": 0.8297,
+      "step": 3874
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8138244941278101,
+      "learning_rate": 5.62667652198698e-06,
+      "loss": 0.9654,
+      "step": 3875
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1509453677572696,
+      "learning_rate": 5.621755166548428e-06,
+      "loss": 0.9183,
+      "step": 3876
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2097956614397278,
+      "learning_rate": 5.616835122633123e-06,
+      "loss": 1.0961,
+      "step": 3877
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0416215127287487,
+      "learning_rate": 5.611916391714887e-06,
+      "loss": 0.7248,
+      "step": 3878
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9247647519072211,
+      "learning_rate": 5.606998975267155e-06,
+      "loss": 1.002,
+      "step": 3879
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9832776049189736,
+      "learning_rate": 5.602082874762952e-06,
+      "loss": 0.7122,
+      "step": 3880
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8520562889911282,
+      "learning_rate": 5.5971680916749226e-06,
+      "loss": 0.7946,
+      "step": 3881
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0059069701775476,
+      "learning_rate": 5.592254627475313e-06,
+      "loss": 1.18,
+      "step": 3882
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9914305636788294,
+      "learning_rate": 5.587342483635978e-06,
+      "loss": 0.6478,
+      "step": 3883
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7109080804474901,
+      "learning_rate": 5.582431661628364e-06,
+      "loss": 0.9835,
+      "step": 3884
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9941851375535812,
+      "learning_rate": 5.577522162923535e-06,
+      "loss": 0.7709,
+      "step": 3885
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6182027544379793,
+      "learning_rate": 5.572613988992157e-06,
+      "loss": 0.6636,
+      "step": 3886
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8669899597807864,
+      "learning_rate": 5.567707141304489e-06,
+      "loss": 0.6373,
+      "step": 3887
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0723807562147492,
+      "learning_rate": 5.562801621330402e-06,
+      "loss": 0.7597,
+      "step": 3888
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3692317369397955,
+      "learning_rate": 5.557897430539369e-06,
+      "loss": 0.7873,
+      "step": 3889
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3851409364948424,
+      "learning_rate": 5.552994570400465e-06,
+      "loss": 0.7149,
+      "step": 3890
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0289478038423334,
+      "learning_rate": 5.548093042382358e-06,
+      "loss": 1.1476,
+      "step": 3891
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.00119021882848,
+      "learning_rate": 5.543192847953325e-06,
+      "loss": 0.7762,
+      "step": 3892
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.93041955942921,
+      "learning_rate": 5.538293988581246e-06,
+      "loss": 0.6526,
+      "step": 3893
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0783773543035837,
+      "learning_rate": 5.5333964657335894e-06,
+      "loss": 0.8571,
+      "step": 3894
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0110408395530974,
+      "learning_rate": 5.5285002808774345e-06,
+      "loss": 0.8804,
+      "step": 3895
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9408709456915596,
+      "learning_rate": 5.52360543547946e-06,
+      "loss": 0.9107,
+      "step": 3896
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9741534579219943,
+      "learning_rate": 5.518711931005931e-06,
+      "loss": 0.7778,
+      "step": 3897
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8629208867032748,
+      "learning_rate": 5.513819768922723e-06,
+      "loss": 0.8627,
+      "step": 3898
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.472687185887626,
+      "learning_rate": 5.50892895069531e-06,
+      "loss": 0.8662,
+      "step": 3899
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0382684808852363,
+      "learning_rate": 5.50403947778875e-06,
+      "loss": 0.8187,
+      "step": 3900
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9342533448903594,
+      "learning_rate": 5.499151351667716e-06,
+      "loss": 0.9919,
+      "step": 3901
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1293631202200372,
+      "learning_rate": 5.49426457379646e-06,
+      "loss": 0.7739,
+      "step": 3902
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.589423274101986,
+      "learning_rate": 5.489379145638842e-06,
+      "loss": 0.5318,
+      "step": 3903
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0023290470373674,
+      "learning_rate": 5.484495068658321e-06,
+      "loss": 0.7196,
+      "step": 3904
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9740983993172502,
+      "learning_rate": 5.479612344317933e-06,
+      "loss": 1.2504,
+      "step": 3905
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8347102473990196,
+      "learning_rate": 5.474730974080327e-06,
+      "loss": 0.811,
+      "step": 3906
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2079979724268668,
+      "learning_rate": 5.469850959407741e-06,
+      "loss": 0.9783,
+      "step": 3907
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8041285795654767,
+      "learning_rate": 5.464972301762001e-06,
+      "loss": 0.9554,
+      "step": 3908
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0705454765967828,
+      "learning_rate": 5.460095002604533e-06,
+      "loss": 0.8799,
+      "step": 3909
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.007895324121506,
+      "learning_rate": 5.455219063396356e-06,
+      "loss": 0.9261,
+      "step": 3910
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9408204235995212,
+      "learning_rate": 5.450344485598081e-06,
+      "loss": 0.7819,
+      "step": 3911
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.806194363490388,
+      "learning_rate": 5.445471270669904e-06,
+      "loss": 0.518,
+      "step": 3912
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9966176546571723,
+      "learning_rate": 5.4405994200716235e-06,
+      "loss": 0.8564,
+      "step": 3913
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2201755949753996,
+      "learning_rate": 5.4357289352626284e-06,
+      "loss": 0.7494,
+      "step": 3914
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1443238898187595,
+      "learning_rate": 5.430859817701886e-06,
+      "loss": 1.1414,
+      "step": 3915
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9730294124829036,
+      "learning_rate": 5.425992068847965e-06,
+      "loss": 0.7563,
+      "step": 3916
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1761819281033914,
+      "learning_rate": 5.421125690159026e-06,
+      "loss": 0.8549,
+      "step": 3917
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2102928055719797,
+      "learning_rate": 5.416260683092814e-06,
+      "loss": 0.7704,
+      "step": 3918
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0888812311666367,
+      "learning_rate": 5.41139704910666e-06,
+      "loss": 0.7806,
+      "step": 3919
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9477702071892046,
+      "learning_rate": 5.40653478965749e-06,
+      "loss": 1.1283,
+      "step": 3920
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5719809446881741,
+      "learning_rate": 5.401673906201821e-06,
+      "loss": 0.4374,
+      "step": 3921
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.07352033860664,
+      "learning_rate": 5.3968144001957445e-06,
+      "loss": 1.0179,
+      "step": 3922
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0871925062452956,
+      "learning_rate": 5.391956273094952e-06,
+      "loss": 0.7156,
+      "step": 3923
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9594217116524192,
+      "learning_rate": 5.38709952635472e-06,
+      "loss": 1.2813,
+      "step": 3924
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8665586141363617,
+      "learning_rate": 5.382244161429902e-06,
+      "loss": 0.6154,
+      "step": 3925
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.944514781998391,
+      "learning_rate": 5.377390179774951e-06,
+      "loss": 0.8158,
+      "step": 3926
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2228935219200023,
+      "learning_rate": 5.3725375828438955e-06,
+      "loss": 0.84,
+      "step": 3927
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0476053248727273,
+      "learning_rate": 5.367686372090359e-06,
+      "loss": 0.6728,
+      "step": 3928
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9967542279847387,
+      "learning_rate": 5.362836548967538e-06,
+      "loss": 1.0387,
+      "step": 3929
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9048777564404021,
+      "learning_rate": 5.357988114928221e-06,
+      "loss": 0.8277,
+      "step": 3930
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1038796500652135,
+      "learning_rate": 5.353141071424781e-06,
+      "loss": 0.8044,
+      "step": 3931
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3341973891753776,
+      "learning_rate": 5.3482954199091684e-06,
+      "loss": 0.9222,
+      "step": 3932
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7131396107888259,
+      "learning_rate": 5.343451161832923e-06,
+      "loss": 0.9546,
+      "step": 3933
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.942311686759334,
+      "learning_rate": 5.338608298647161e-06,
+      "loss": 0.8864,
+      "step": 3934
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.174042789186435,
+      "learning_rate": 5.333766831802589e-06,
+      "loss": 0.7998,
+      "step": 3935
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.025636034180837,
+      "learning_rate": 5.328926762749485e-06,
+      "loss": 0.8779,
+      "step": 3936
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.380365978417812,
+      "learning_rate": 5.324088092937716e-06,
+      "loss": 0.7748,
+      "step": 3937
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6187879692060329,
+      "learning_rate": 5.319250823816732e-06,
+      "loss": 0.4796,
+      "step": 3938
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1560967653437426,
+      "learning_rate": 5.314414956835552e-06,
+      "loss": 1.1115,
+      "step": 3939
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8832195289180481,
+      "learning_rate": 5.309580493442784e-06,
+      "loss": 0.7731,
+      "step": 3940
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1258506676204882,
+      "learning_rate": 5.304747435086617e-06,
+      "loss": 0.8539,
+      "step": 3941
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0084135956342515,
+      "learning_rate": 5.299915783214816e-06,
+      "loss": 0.6738,
+      "step": 3942
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9515374797055243,
+      "learning_rate": 5.2950855392747196e-06,
+      "loss": 1.0573,
+      "step": 3943
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.007408237561147,
+      "learning_rate": 5.2902567047132505e-06,
+      "loss": 0.8215,
+      "step": 3944
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1389967314673097,
+      "learning_rate": 5.285429280976915e-06,
+      "loss": 0.7659,
+      "step": 3945
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9296841871619476,
+      "learning_rate": 5.28060326951178e-06,
+      "loss": 0.8516,
+      "step": 3946
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0439255504757228,
+      "learning_rate": 5.2757786717635026e-06,
+      "loss": 0.8901,
+      "step": 3947
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9301305988448669,
+      "learning_rate": 5.270955489177316e-06,
+      "loss": 1.2927,
+      "step": 3948
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0594854632167199,
+      "learning_rate": 5.266133723198028e-06,
+      "loss": 0.7281,
+      "step": 3949
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9106552782508254,
+      "learning_rate": 5.2613133752700145e-06,
+      "loss": 0.6227,
+      "step": 3950
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.196207819848022,
+      "learning_rate": 5.2564944468372344e-06,
+      "loss": 0.7877,
+      "step": 3951
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9964211867379845,
+      "learning_rate": 5.251676939343227e-06,
+      "loss": 0.8439,
+      "step": 3952
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9546128626031676,
+      "learning_rate": 5.2468608542310886e-06,
+      "loss": 0.715,
+      "step": 3953
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0001963305212078,
+      "learning_rate": 5.242046192943503e-06,
+      "loss": 1.0334,
+      "step": 3954
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0197151100181054,
+      "learning_rate": 5.237232956922724e-06,
+      "loss": 0.8088,
+      "step": 3955
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0296389797515184,
+      "learning_rate": 5.232421147610584e-06,
+      "loss": 0.863,
+      "step": 3956
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9184461226630724,
+      "learning_rate": 5.227610766448474e-06,
+      "loss": 0.8161,
+      "step": 3957
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0643511182993541,
+      "learning_rate": 5.22280181487737e-06,
+      "loss": 1.0418,
+      "step": 3958
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8586167654917201,
+      "learning_rate": 5.2179942943378185e-06,
+      "loss": 0.8033,
+      "step": 3959
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1804060451914087,
+      "learning_rate": 5.213188206269926e-06,
+      "loss": 0.854,
+      "step": 3960
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.954705771759757,
+      "learning_rate": 5.208383552113382e-06,
+      "loss": 0.7204,
+      "step": 3961
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1587009554304257,
+      "learning_rate": 5.2035803333074435e-06,
+      "loss": 0.7358,
+      "step": 3962
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9942492609148202,
+      "learning_rate": 5.1987785512909405e-06,
+      "loss": 1.0171,
+      "step": 3963
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7588910850261709,
+      "learning_rate": 5.193978207502262e-06,
+      "loss": 0.509,
+      "step": 3964
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9231868986598611,
+      "learning_rate": 5.189179303379377e-06,
+      "loss": 0.6787,
+      "step": 3965
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0048597019809158,
+      "learning_rate": 5.184381840359815e-06,
+      "loss": 0.8959,
+      "step": 3966
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0106481620661663,
+      "learning_rate": 5.179585819880686e-06,
+      "loss": 0.769,
+      "step": 3967
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0685746249557686,
+      "learning_rate": 5.17479124337865e-06,
+      "loss": 1.0024,
+      "step": 3968
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0285829791600338,
+      "learning_rate": 5.1699981122899466e-06,
+      "loss": 1.0843,
+      "step": 3969
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8322853735086885,
+      "learning_rate": 5.1652064280503884e-06,
+      "loss": 0.8328,
+      "step": 3970
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9429028475997115,
+      "learning_rate": 5.160416192095337e-06,
+      "loss": 1.0022,
+      "step": 3971
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0338504622209042,
+      "learning_rate": 5.155627405859731e-06,
+      "loss": 1.0442,
+      "step": 3972
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0598055576424914,
+      "learning_rate": 5.15084007077808e-06,
+      "loss": 0.7349,
+      "step": 3973
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5823924946624541,
+      "learning_rate": 5.146054188284443e-06,
+      "loss": 0.4102,
+      "step": 3974
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.90709499276391,
+      "learning_rate": 5.141269759812457e-06,
+      "loss": 0.7791,
+      "step": 3975
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6374668124989014,
+      "learning_rate": 5.136486786795318e-06,
+      "loss": 0.8379,
+      "step": 3976
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8993197083844063,
+      "learning_rate": 5.131705270665795e-06,
+      "loss": 0.9971,
+      "step": 3977
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0814197352935213,
+      "learning_rate": 5.126925212856202e-06,
+      "loss": 0.7985,
+      "step": 3978
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1029983469473905,
+      "learning_rate": 5.122146614798433e-06,
+      "loss": 0.7865,
+      "step": 3979
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9850692053351331,
+      "learning_rate": 5.1173694779239415e-06,
+      "loss": 0.6655,
+      "step": 3980
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6953537343934006,
+      "learning_rate": 5.112593803663732e-06,
+      "loss": 1.0004,
+      "step": 3981
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8124568060976252,
+      "learning_rate": 5.107819593448385e-06,
+      "loss": 0.8501,
+      "step": 3982
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1841510587761184,
+      "learning_rate": 5.103046848708037e-06,
+      "loss": 0.7324,
+      "step": 3983
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.089611962120238,
+      "learning_rate": 5.09827557087239e-06,
+      "loss": 0.9122,
+      "step": 3984
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9870352937780166,
+      "learning_rate": 5.093505761370692e-06,
+      "loss": 0.8199,
+      "step": 3985
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.173412677394895,
+      "learning_rate": 5.088737421631767e-06,
+      "loss": 0.8093,
+      "step": 3986
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0226027817914267,
+      "learning_rate": 5.083970553083997e-06,
+      "loss": 0.9799,
+      "step": 3987
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0629577929599745,
+      "learning_rate": 5.079205157155309e-06,
+      "loss": 0.7978,
+      "step": 3988
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9853304033909444,
+      "learning_rate": 5.074441235273206e-06,
+      "loss": 0.9946,
+      "step": 3989
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1079914776078443,
+      "learning_rate": 5.069678788864742e-06,
+      "loss": 0.8772,
+      "step": 3990
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.058601327910259,
+      "learning_rate": 5.064917819356532e-06,
+      "loss": 1.0276,
+      "step": 3991
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0824929332491926,
+      "learning_rate": 5.06015832817474e-06,
+      "loss": 0.7403,
+      "step": 3992
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9684531685064993,
+      "learning_rate": 5.055400316745096e-06,
+      "loss": 0.7585,
+      "step": 3993
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5850462732887939,
+      "learning_rate": 5.050643786492888e-06,
+      "loss": 0.3952,
+      "step": 3994
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.966618068318972,
+      "learning_rate": 5.04588873884295e-06,
+      "loss": 0.9644,
+      "step": 3995
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9730853871945834,
+      "learning_rate": 5.0411351752196815e-06,
+      "loss": 1.1169,
+      "step": 3996
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0597288172408565,
+      "learning_rate": 5.036383097047037e-06,
+      "loss": 0.7607,
+      "step": 3997
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0559176810904827,
+      "learning_rate": 5.031632505748516e-06,
+      "loss": 0.8994,
+      "step": 3998
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.014097062452851,
+      "learning_rate": 5.026883402747188e-06,
+      "loss": 0.6706,
+      "step": 3999
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9205446604627576,
+      "learning_rate": 5.0221357894656605e-06,
+      "loss": 0.8107,
+      "step": 4000
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9596932784825551,
+      "learning_rate": 5.017389667326107e-06,
+      "loss": 0.9863,
+      "step": 4001
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.712447189738274,
+      "learning_rate": 5.012645037750254e-06,
+      "loss": 0.708,
+      "step": 4002
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7251772269099606,
+      "learning_rate": 5.007901902159369e-06,
+      "loss": 0.7433,
+      "step": 4003
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0275715290577223,
+      "learning_rate": 5.003160261974282e-06,
+      "loss": 0.7173,
+      "step": 4004
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.446160421374766,
+      "learning_rate": 4.998420118615381e-06,
+      "loss": 0.9219,
+      "step": 4005
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9691074537681101,
+      "learning_rate": 4.993681473502586e-06,
+      "loss": 1.0414,
+      "step": 4006
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6969122625789383,
+      "learning_rate": 4.988944328055386e-06,
+      "loss": 0.9923,
+      "step": 4007
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9858407183730455,
+      "learning_rate": 4.984208683692817e-06,
+      "loss": 0.7337,
+      "step": 4008
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1366520190195544,
+      "learning_rate": 4.979474541833455e-06,
+      "loss": 0.8016,
+      "step": 4009
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0556825223634612,
+      "learning_rate": 4.97474190389544e-06,
+      "loss": 1.0381,
+      "step": 4010
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9847140831347271,
+      "learning_rate": 4.970010771296453e-06,
+      "loss": 0.9382,
+      "step": 4011
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6219179150126536,
+      "learning_rate": 4.965281145453731e-06,
+      "loss": 0.4945,
+      "step": 4012
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9473842190917351,
+      "learning_rate": 4.960553027784047e-06,
+      "loss": 0.9949,
+      "step": 4013
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0524131774388457,
+      "learning_rate": 4.955826419703736e-06,
+      "loss": 0.7125,
+      "step": 4014
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9740367962274094,
+      "learning_rate": 4.951101322628675e-06,
+      "loss": 1.0119,
+      "step": 4015
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.158322238865018,
+      "learning_rate": 4.946377737974284e-06,
+      "loss": 0.8247,
+      "step": 4016
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8964986330980481,
+      "learning_rate": 4.941655667155538e-06,
+      "loss": 0.7184,
+      "step": 4017
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9538581445920812,
+      "learning_rate": 4.936935111586953e-06,
+      "loss": 1.0152,
+      "step": 4018
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8876500111056471,
+      "learning_rate": 4.9322160726826e-06,
+      "loss": 0.8555,
+      "step": 4019
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.976056047857832,
+      "learning_rate": 4.927498551856077e-06,
+      "loss": 0.9953,
+      "step": 4020
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2104546888265988,
+      "learning_rate": 4.922782550520545e-06,
+      "loss": 0.8213,
+      "step": 4021
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4626584566672185,
+      "learning_rate": 4.918068070088708e-06,
+      "loss": 0.947,
+      "step": 4022
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3655698770016305,
+      "learning_rate": 4.913355111972802e-06,
+      "loss": 0.9032,
+      "step": 4023
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9298926699711213,
+      "learning_rate": 4.908643677584618e-06,
+      "loss": 0.976,
+      "step": 4024
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1573852625008614,
+      "learning_rate": 4.903933768335492e-06,
+      "loss": 0.7817,
+      "step": 4025
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0101113015432874,
+      "learning_rate": 4.899225385636293e-06,
+      "loss": 0.8793,
+      "step": 4026
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9883666358505909,
+      "learning_rate": 4.894518530897441e-06,
+      "loss": 0.5791,
+      "step": 4027
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.229613563969637,
+      "learning_rate": 4.889813205528895e-06,
+      "loss": 0.7983,
+      "step": 4028
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9302107966685546,
+      "learning_rate": 4.885109410940161e-06,
+      "loss": 0.6938,
+      "step": 4029
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9079848175738723,
+      "learning_rate": 4.880407148540275e-06,
+      "loss": 0.9233,
+      "step": 4030
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6403564398732768,
+      "learning_rate": 4.875706419737829e-06,
+      "loss": 0.4607,
+      "step": 4031
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6937332799653666,
+      "learning_rate": 4.87100722594094e-06,
+      "loss": 0.9666,
+      "step": 4032
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9746939065558876,
+      "learning_rate": 4.866309568557276e-06,
+      "loss": 1.0559,
+      "step": 4033
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8074046577620902,
+      "learning_rate": 4.8616134489940455e-06,
+      "loss": 0.8649,
+      "step": 4034
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9846003862046926,
+      "learning_rate": 4.856918868657987e-06,
+      "loss": 0.8817,
+      "step": 4035
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2086059695452531,
+      "learning_rate": 4.8522258289553905e-06,
+      "loss": 0.6892,
+      "step": 4036
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3382493885720625,
+      "learning_rate": 4.847534331292068e-06,
+      "loss": 0.9576,
+      "step": 4037
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6678907124771918,
+      "learning_rate": 4.8428443770733835e-06,
+      "loss": 0.672,
+      "step": 4038
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6968843406669879,
+      "learning_rate": 4.838155967704239e-06,
+      "loss": 0.6673,
+      "step": 4039
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1414009795366553,
+      "learning_rate": 4.83346910458906e-06,
+      "loss": 0.8897,
+      "step": 4040
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8034158786662392,
+      "learning_rate": 4.8287837891318235e-06,
+      "loss": 0.9054,
+      "step": 4041
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.017731668373682,
+      "learning_rate": 4.824100022736036e-06,
+      "loss": 0.7506,
+      "step": 4042
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8219426802527998,
+      "learning_rate": 4.819417806804745e-06,
+      "loss": 0.578,
+      "step": 4043
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9053810951781384,
+      "learning_rate": 4.814737142740521e-06,
+      "loss": 0.9017,
+      "step": 4044
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9760990744872088,
+      "learning_rate": 4.810058031945485e-06,
+      "loss": 0.8291,
+      "step": 4045
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1245622495553453,
+      "learning_rate": 4.805380475821287e-06,
+      "loss": 0.754,
+      "step": 4046
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2967654274995537,
+      "learning_rate": 4.800704475769103e-06,
+      "loss": 1.0619,
+      "step": 4047
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6108876749977703,
+      "learning_rate": 4.796030033189656e-06,
+      "loss": 0.4145,
+      "step": 4048
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9155570414062457,
+      "learning_rate": 4.791357149483195e-06,
+      "loss": 0.9869,
+      "step": 4049
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4653849966458903,
+      "learning_rate": 4.786685826049508e-06,
+      "loss": 0.7858,
+      "step": 4050
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0154805793692174,
+      "learning_rate": 4.782016064287903e-06,
+      "loss": 0.7492,
+      "step": 4051
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.025025334711219,
+      "learning_rate": 4.777347865597233e-06,
+      "loss": 0.8512,
+      "step": 4052
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8772230700898349,
+      "learning_rate": 4.772681231375882e-06,
+      "loss": 0.9252,
+      "step": 4053
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0073900859018927,
+      "learning_rate": 4.768016163021754e-06,
+      "loss": 1.0898,
+      "step": 4054
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0933373828178214,
+      "learning_rate": 4.7633526619322946e-06,
+      "loss": 0.69,
+      "step": 4055
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.293771692651802,
+      "learning_rate": 4.758690729504478e-06,
+      "loss": 0.7454,
+      "step": 4056
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.697755906040494,
+      "learning_rate": 4.75403036713481e-06,
+      "loss": 0.9874,
+      "step": 4057
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9465335960895984,
+      "learning_rate": 4.749371576219317e-06,
+      "loss": 1.0229,
+      "step": 4058
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9045306981659296,
+      "learning_rate": 4.744714358153564e-06,
+      "loss": 0.7691,
+      "step": 4059
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8827370829614846,
+      "learning_rate": 4.740058714332647e-06,
+      "loss": 0.7213,
+      "step": 4060
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.018310883279436,
+      "learning_rate": 4.735404646151176e-06,
+      "loss": 0.8995,
+      "step": 4061
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9153673272497617,
+      "learning_rate": 4.730752155003303e-06,
+      "loss": 0.6334,
+      "step": 4062
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.085728490583347,
+      "learning_rate": 4.726101242282708e-06,
+      "loss": 1.0073,
+      "step": 4063
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9774607526527312,
+      "learning_rate": 4.721451909382584e-06,
+      "loss": 0.7679,
+      "step": 4064
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6322400353605191,
+      "learning_rate": 4.716804157695664e-06,
+      "loss": 0.6508,
+      "step": 4065
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1855754005937962,
+      "learning_rate": 4.7121579886142065e-06,
+      "loss": 0.7618,
+      "step": 4066
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9591974356380258,
+      "learning_rate": 4.7075134035299875e-06,
+      "loss": 0.7268,
+      "step": 4067
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2136576538206658,
+      "learning_rate": 4.702870403834318e-06,
+      "loss": 1.1482,
+      "step": 4068
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9802900782929891,
+      "learning_rate": 4.698228990918024e-06,
+      "loss": 0.7164,
+      "step": 4069
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0243941931604237,
+      "learning_rate": 4.693589166171466e-06,
+      "loss": 0.857,
+      "step": 4070
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.563290511733454,
+      "learning_rate": 4.688950930984528e-06,
+      "loss": 0.8742,
+      "step": 4071
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9245229415442332,
+      "learning_rate": 4.684314286746608e-06,
+      "loss": 1.0192,
+      "step": 4072
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0446109879686516,
+      "learning_rate": 4.679679234846636e-06,
+      "loss": 0.945,
+      "step": 4073
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1655506962190207,
+      "learning_rate": 4.675045776673067e-06,
+      "loss": 0.6891,
+      "step": 4074
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.816403443947281,
+      "learning_rate": 4.670413913613867e-06,
+      "loss": 0.7129,
+      "step": 4075
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0373430074667966,
+      "learning_rate": 4.665783647056538e-06,
+      "loss": 0.9135,
+      "step": 4076
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9200566854765146,
+      "learning_rate": 4.661154978388094e-06,
+      "loss": 1.1752,
+      "step": 4077
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0931077032433991,
+      "learning_rate": 4.65652790899508e-06,
+      "loss": 0.7796,
+      "step": 4078
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0717448424211682,
+      "learning_rate": 4.651902440263547e-06,
+      "loss": 0.7854,
+      "step": 4079
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2423352246935266,
+      "learning_rate": 4.64727857357908e-06,
+      "loss": 0.8154,
+      "step": 4080
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7406759969713427,
+      "learning_rate": 4.642656310326782e-06,
+      "loss": 0.9694,
+      "step": 4081
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9074176868117995,
+      "learning_rate": 4.638035651891266e-06,
+      "loss": 0.9272,
+      "step": 4082
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0347317400837448,
+      "learning_rate": 4.633416599656676e-06,
+      "loss": 0.8357,
+      "step": 4083
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6231547868393629,
+      "learning_rate": 4.628799155006669e-06,
+      "loss": 0.4385,
+      "step": 4084
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0547395542273976,
+      "learning_rate": 4.624183319324427e-06,
+      "loss": 0.7696,
+      "step": 4085
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0187041346972678,
+      "learning_rate": 4.619569093992636e-06,
+      "loss": 0.9248,
+      "step": 4086
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0379577150039427,
+      "learning_rate": 4.614956480393511e-06,
+      "loss": 1.1009,
+      "step": 4087
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7124324287546657,
+      "learning_rate": 4.610345479908787e-06,
+      "loss": 0.4797,
+      "step": 4088
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9127904708370232,
+      "learning_rate": 4.605736093919702e-06,
+      "loss": 0.6913,
+      "step": 4089
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8917783520099078,
+      "learning_rate": 4.601128323807023e-06,
+      "loss": 1.003,
+      "step": 4090
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.006282296398363,
+      "learning_rate": 4.596522170951028e-06,
+      "loss": 0.9947,
+      "step": 4091
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9286143076101552,
+      "learning_rate": 4.591917636731515e-06,
+      "loss": 0.8551,
+      "step": 4092
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8768551948006491,
+      "learning_rate": 4.5873147225277845e-06,
+      "loss": 0.7014,
+      "step": 4093
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1628074626764715,
+      "learning_rate": 4.582713429718667e-06,
+      "loss": 0.7502,
+      "step": 4094
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1263375563385183,
+      "learning_rate": 4.5781137596825035e-06,
+      "loss": 0.7649,
+      "step": 4095
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0500153675167392,
+      "learning_rate": 4.573515713797137e-06,
+      "loss": 1.2095,
+      "step": 4096
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1468858298422189,
+      "learning_rate": 4.568919293439939e-06,
+      "loss": 0.7727,
+      "step": 4097
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8234419879289954,
+      "learning_rate": 4.56432449998779e-06,
+      "loss": 1.0375,
+      "step": 4098
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.227724574292319,
+      "learning_rate": 4.559731334817076e-06,
+      "loss": 0.7984,
+      "step": 4099
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0284956564618777,
+      "learning_rate": 4.555139799303706e-06,
+      "loss": 0.768,
+      "step": 4100
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5716624561261987,
+      "learning_rate": 4.5505498948230905e-06,
+      "loss": 0.4744,
+      "step": 4101
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1308556086264587,
+      "learning_rate": 4.545961622750158e-06,
+      "loss": 1.1334,
+      "step": 4102
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8941724777001631,
+      "learning_rate": 4.541374984459351e-06,
+      "loss": 0.8207,
+      "step": 4103
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1069361581537278,
+      "learning_rate": 4.53678998132461e-06,
+      "loss": 0.8832,
+      "step": 4104
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6789946983572481,
+      "learning_rate": 4.532206614719398e-06,
+      "loss": 0.9284,
+      "step": 4105
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0765375493338147,
+      "learning_rate": 4.527624886016686e-06,
+      "loss": 1.0742,
+      "step": 4106
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8384301496960724,
+      "learning_rate": 4.523044796588945e-06,
+      "loss": 0.6776,
+      "step": 4107
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.941667209843803,
+      "learning_rate": 4.518466347808168e-06,
+      "loss": 0.6764,
+      "step": 4108
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0305692585240775,
+      "learning_rate": 4.513889541045849e-06,
+      "loss": 0.9113,
+      "step": 4109
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1849032616750887,
+      "learning_rate": 4.509314377672987e-06,
+      "loss": 0.7698,
+      "step": 4110
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0616460969633965,
+      "learning_rate": 4.504740859060097e-06,
+      "loss": 1.2337,
+      "step": 4111
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9717933833813822,
+      "learning_rate": 4.500168986577197e-06,
+      "loss": 0.6497,
+      "step": 4112
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9886390528051895,
+      "learning_rate": 4.495598761593815e-06,
+      "loss": 1.0109,
+      "step": 4113
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0941393919918674,
+      "learning_rate": 4.491030185478976e-06,
+      "loss": 0.6703,
+      "step": 4114
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8346967267070562,
+      "learning_rate": 4.486463259601221e-06,
+      "loss": 0.6197,
+      "step": 4115
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0143160191985467,
+      "learning_rate": 4.4818979853286e-06,
+      "loss": 1.0755,
+      "step": 4116
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0657646536457421,
+      "learning_rate": 4.47733436402865e-06,
+      "loss": 0.7906,
+      "step": 4117
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9513773434646414,
+      "learning_rate": 4.472772397068431e-06,
+      "loss": 0.745,
+      "step": 4118
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9955559918082026,
+      "learning_rate": 4.468212085814501e-06,
+      "loss": 0.6501,
+      "step": 4119
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1275687433712493,
+      "learning_rate": 4.463653431632926e-06,
+      "loss": 1.1107,
+      "step": 4120
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5353480559201942,
+      "learning_rate": 4.459096435889263e-06,
+      "loss": 0.5597,
+      "step": 4121
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.946493908698279,
+      "learning_rate": 4.454541099948585e-06,
+      "loss": 0.7676,
+      "step": 4122
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.17438188139056,
+      "learning_rate": 4.449987425175468e-06,
+      "loss": 0.8614,
+      "step": 4123
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0556156184671832,
+      "learning_rate": 4.445435412933978e-06,
+      "loss": 0.8809,
+      "step": 4124
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.039269657304542,
+      "learning_rate": 4.440885064587695e-06,
+      "loss": 1.0456,
+      "step": 4125
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1565209630822257,
+      "learning_rate": 4.436336381499701e-06,
+      "loss": 0.9149,
+      "step": 4126
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.213375405697708,
+      "learning_rate": 4.4317893650325675e-06,
+      "loss": 0.7772,
+      "step": 4127
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.134915839237304,
+      "learning_rate": 4.427244016548375e-06,
+      "loss": 0.843,
+      "step": 4128
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1346526649161977,
+      "learning_rate": 4.422700337408708e-06,
+      "loss": 0.8052,
+      "step": 4129
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0759617102560246,
+      "learning_rate": 4.418158328974647e-06,
+      "loss": 1.0257,
+      "step": 4130
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.551065662565256,
+      "learning_rate": 4.413617992606763e-06,
+      "loss": 0.9078,
+      "step": 4131
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9059679823232009,
+      "learning_rate": 4.409079329665144e-06,
+      "loss": 0.7679,
+      "step": 4132
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6452702440488416,
+      "learning_rate": 4.40454234150936e-06,
+      "loss": 0.8469,
+      "step": 4133
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.096713137933433,
+      "learning_rate": 4.40000702949849e-06,
+      "loss": 0.7751,
+      "step": 4134
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9953280797765461,
+      "learning_rate": 4.39547339499111e-06,
+      "loss": 1.1057,
+      "step": 4135
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9711270614816052,
+      "learning_rate": 4.390941439345284e-06,
+      "loss": 0.8197,
+      "step": 4136
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8519283802684382,
+      "learning_rate": 4.386411163918588e-06,
+      "loss": 1.0043,
+      "step": 4137
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5291192135662187,
+      "learning_rate": 4.381882570068079e-06,
+      "loss": 0.4091,
+      "step": 4138
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0575559769297087,
+      "learning_rate": 4.377355659150322e-06,
+      "loss": 0.929,
+      "step": 4139
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1095191664716837,
+      "learning_rate": 4.372830432521377e-06,
+      "loss": 0.7988,
+      "step": 4140
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0824623684432964,
+      "learning_rate": 4.368306891536788e-06,
+      "loss": 0.8099,
+      "step": 4141
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0001126768799493,
+      "learning_rate": 4.3637850375516085e-06,
+      "loss": 0.8175,
+      "step": 4142
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9414662714864023,
+      "learning_rate": 4.359264871920379e-06,
+      "loss": 0.97,
+      "step": 4143
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0118470673314537,
+      "learning_rate": 4.354746395997139e-06,
+      "loss": 1.0548,
+      "step": 4144
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9694480017658588,
+      "learning_rate": 4.350229611135414e-06,
+      "loss": 0.7264,
+      "step": 4145
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.273406474270394,
+      "learning_rate": 4.345714518688226e-06,
+      "loss": 0.8909,
+      "step": 4146
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0035777323399675,
+      "learning_rate": 4.341201120008101e-06,
+      "loss": 0.7694,
+      "step": 4147
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6994336873207813,
+      "learning_rate": 4.336689416447038e-06,
+      "loss": 0.5717,
+      "step": 4148
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9438512559454405,
+      "learning_rate": 4.332179409356541e-06,
+      "loss": 0.9693,
+      "step": 4149
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9509575688333783,
+      "learning_rate": 4.327671100087605e-06,
+      "loss": 0.7817,
+      "step": 4150
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.253943103552591,
+      "learning_rate": 4.3231644899907175e-06,
+      "loss": 0.8438,
+      "step": 4151
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9585085599177928,
+      "learning_rate": 4.3186595804158485e-06,
+      "loss": 0.7276,
+      "step": 4152
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.154858434946152,
+      "learning_rate": 4.3141563727124655e-06,
+      "loss": 0.7091,
+      "step": 4153
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7202854446906938,
+      "learning_rate": 4.3096548682295304e-06,
+      "loss": 1.0349,
+      "step": 4154
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9699437086506095,
+      "learning_rate": 4.305155068315481e-06,
+      "loss": 0.419,
+      "step": 4155
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9679317383853678,
+      "learning_rate": 4.300656974318257e-06,
+      "loss": 0.7972,
+      "step": 4156
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9472138226830632,
+      "learning_rate": 4.2961605875852844e-06,
+      "loss": 0.6638,
+      "step": 4157
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0723835770345869,
+      "learning_rate": 4.291665909463477e-06,
+      "loss": 1.1589,
+      "step": 4158
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.221739827390982,
+      "learning_rate": 4.2871729412992315e-06,
+      "loss": 0.74,
+      "step": 4159
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9947657388880959,
+      "learning_rate": 4.282681684438439e-06,
+      "loss": 0.9115,
+      "step": 4160
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1927992985919211,
+      "learning_rate": 4.278192140226481e-06,
+      "loss": 0.7379,
+      "step": 4161
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3486216569373712,
+      "learning_rate": 4.273704310008213e-06,
+      "loss": 0.9455,
+      "step": 4162
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.899106424180138,
+      "learning_rate": 4.26921819512799e-06,
+      "loss": 1.0014,
+      "step": 4163
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9764981693266439,
+      "learning_rate": 4.2647337969296485e-06,
+      "loss": 0.9923,
+      "step": 4164
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2744760569596516,
+      "learning_rate": 4.260251116756508e-06,
+      "loss": 0.7298,
+      "step": 4165
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1549914566393125,
+      "learning_rate": 4.255770155951374e-06,
+      "loss": 0.8891,
+      "step": 4166
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0256268765464496,
+      "learning_rate": 4.251290915856547e-06,
+      "loss": 0.792,
+      "step": 4167
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9400080653738222,
+      "learning_rate": 4.2468133978137945e-06,
+      "loss": 0.9938,
+      "step": 4168
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9403432360840851,
+      "learning_rate": 4.242337603164385e-06,
+      "loss": 0.6588,
+      "step": 4169
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5790085726464849,
+      "learning_rate": 4.237863533249057e-06,
+      "loss": 0.8033,
+      "step": 4170
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2739612598473913,
+      "learning_rate": 4.2333911894080404e-06,
+      "loss": 0.7871,
+      "step": 4171
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0449875210445578,
+      "learning_rate": 4.2289205729810504e-06,
+      "loss": 0.7095,
+      "step": 4172
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6965633556302312,
+      "learning_rate": 4.224451685307275e-06,
+      "loss": 0.613,
+      "step": 4173
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0766994042634492,
+      "learning_rate": 4.21998452772539e-06,
+      "loss": 0.9394,
+      "step": 4174
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.98473212598556,
+      "learning_rate": 4.215519101573561e-06,
+      "loss": 0.9483,
+      "step": 4175
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9933707746456257,
+      "learning_rate": 4.211055408189416e-06,
+      "loss": 0.6177,
+      "step": 4176
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1075816284318196,
+      "learning_rate": 4.20659344891008e-06,
+      "loss": 0.8805,
+      "step": 4177
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9614422895855785,
+      "learning_rate": 4.202133225072153e-06,
+      "loss": 0.944,
+      "step": 4178
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.388575315252224,
+      "learning_rate": 4.19767473801172e-06,
+      "loss": 0.7475,
+      "step": 4179
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6872817182829961,
+      "learning_rate": 4.193217989064332e-06,
+      "loss": 0.9826,
+      "step": 4180
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9798766866050059,
+      "learning_rate": 4.188762979565033e-06,
+      "loss": 0.7576,
+      "step": 4181
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8865744137344724,
+      "learning_rate": 4.1843097108483465e-06,
+      "loss": 0.8873,
+      "step": 4182
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9690734904569169,
+      "learning_rate": 4.17985818424826e-06,
+      "loss": 0.6778,
+      "step": 4183
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9127028079870589,
+      "learning_rate": 4.1754084010982534e-06,
+      "loss": 1.0047,
+      "step": 4184
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9237080335281039,
+      "learning_rate": 4.170960362731279e-06,
+      "loss": 1.0375,
+      "step": 4185
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.239147426879835,
+      "learning_rate": 4.16651407047977e-06,
+      "loss": 0.7818,
+      "step": 4186
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.043560675530548,
+      "learning_rate": 4.16206952567563e-06,
+      "loss": 1.1893,
+      "step": 4187
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1458870979677953,
+      "learning_rate": 4.1576267296502415e-06,
+      "loss": 0.9109,
+      "step": 4188
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.898301414534415,
+      "learning_rate": 4.153185683734471e-06,
+      "loss": 0.6953,
+      "step": 4189
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8811704547958223,
+      "learning_rate": 4.148746389258645e-06,
+      "loss": 0.6394,
+      "step": 4190
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0534840133341488,
+      "learning_rate": 4.14430884755258e-06,
+      "loss": 0.702,
+      "step": 4191
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0306267931165158,
+      "learning_rate": 4.13987305994556e-06,
+      "loss": 1.0668,
+      "step": 4192
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5152641130097981,
+      "learning_rate": 4.135439027766352e-06,
+      "loss": 0.4823,
+      "step": 4193
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1279025637829063,
+      "learning_rate": 4.1310067523431795e-06,
+      "loss": 0.9233,
+      "step": 4194
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2057704612284639,
+      "learning_rate": 4.126576235003757e-06,
+      "loss": 0.7114,
+      "step": 4195
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7824208049417725,
+      "learning_rate": 4.12214747707527e-06,
+      "loss": 0.5031,
+      "step": 4196
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9731512479536503,
+      "learning_rate": 4.117720479884364e-06,
+      "loss": 0.9645,
+      "step": 4197
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1405478411184173,
+      "learning_rate": 4.113295244757171e-06,
+      "loss": 0.7278,
+      "step": 4198
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8602860512596033,
+      "learning_rate": 4.108871773019295e-06,
+      "loss": 0.747,
+      "step": 4199
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.906615743436363,
+      "learning_rate": 4.104450065995799e-06,
+      "loss": 0.886,
+      "step": 4200
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0824268032504574,
+      "learning_rate": 4.100030125011232e-06,
+      "loss": 1.1024,
+      "step": 4201
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.018119096164812,
+      "learning_rate": 4.0956119513896005e-06,
+      "loss": 0.6328,
+      "step": 4202
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4444185632894062,
+      "learning_rate": 4.091195546454398e-06,
+      "loss": 0.8619,
+      "step": 4203
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6269370962869447,
+      "learning_rate": 4.086780911528569e-06,
+      "loss": 0.923,
+      "step": 4204
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0323349741681118,
+      "learning_rate": 4.082368047934543e-06,
+      "loss": 0.6894,
+      "step": 4205
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8115434028675259,
+      "learning_rate": 4.077956956994212e-06,
+      "loss": 0.9379,
+      "step": 4206
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.951715906569802,
+      "learning_rate": 4.073547640028942e-06,
+      "loss": 1.0554,
+      "step": 4207
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9008456438386917,
+      "learning_rate": 4.069140098359557e-06,
+      "loss": 1.0232,
+      "step": 4208
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0088262499881886,
+      "learning_rate": 4.064734333306362e-06,
+      "loss": 0.8535,
+      "step": 4209
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1525865695976831,
+      "learning_rate": 4.060330346189125e-06,
+      "loss": 0.7401,
+      "step": 4210
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5617515082349832,
+      "learning_rate": 4.055928138327074e-06,
+      "loss": 0.611,
+      "step": 4211
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.127266247616169,
+      "learning_rate": 4.051527711038915e-06,
+      "loss": 0.7389,
+      "step": 4212
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9838775595979186,
+      "learning_rate": 4.0471290656428155e-06,
+      "loss": 0.8633,
+      "step": 4213
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9913113989616917,
+      "learning_rate": 4.042732203456413e-06,
+      "loss": 0.8162,
+      "step": 4214
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1068704735665964,
+      "learning_rate": 4.0383371257968e-06,
+      "loss": 0.7602,
+      "step": 4215
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9310899330631934,
+      "learning_rate": 4.033943833980548e-06,
+      "loss": 0.9449,
+      "step": 4216
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5857520358851331,
+      "learning_rate": 4.029552329323689e-06,
+      "loss": 0.9064,
+      "step": 4217
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5358733181015751,
+      "learning_rate": 4.0251626131417135e-06,
+      "loss": 0.6556,
+      "step": 4218
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8785798143058701,
+      "learning_rate": 4.020774686749581e-06,
+      "loss": 0.7002,
+      "step": 4219
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0118512721867365,
+      "learning_rate": 4.0163885514617175e-06,
+      "loss": 1.0604,
+      "step": 4220
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0042886428404012,
+      "learning_rate": 4.012004208592013e-06,
+      "loss": 0.6978,
+      "step": 4221
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0730291449302993,
+      "learning_rate": 4.00762165945381e-06,
+      "loss": 0.7947,
+      "step": 4222
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0923522519198428,
+      "learning_rate": 4.003240905359924e-06,
+      "loss": 0.6288,
+      "step": 4223
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0831225192151652,
+      "learning_rate": 3.998861947622635e-06,
+      "loss": 0.7863,
+      "step": 4224
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0064542296245915,
+      "learning_rate": 3.994484787553673e-06,
+      "loss": 1.0028,
+      "step": 4225
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0637859272865904,
+      "learning_rate": 3.990109426464237e-06,
+      "loss": 0.8146,
+      "step": 4226
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1892881637387065,
+      "learning_rate": 3.9857358656649935e-06,
+      "loss": 0.9761,
+      "step": 4227
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5578566826836445,
+      "learning_rate": 3.9813641064660525e-06,
+      "loss": 0.6308,
+      "step": 4228
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6186517130126553,
+      "learning_rate": 3.976994150177001e-06,
+      "loss": 0.9224,
+      "step": 4229
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9032987939819179,
+      "learning_rate": 3.972625998106876e-06,
+      "loss": 1.0338,
+      "step": 4230
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1797320212042604,
+      "learning_rate": 3.968259651564183e-06,
+      "loss": 0.7493,
+      "step": 4231
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5331375303139678,
+      "learning_rate": 3.963895111856875e-06,
+      "loss": 0.7615,
+      "step": 4232
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.057916978545126,
+      "learning_rate": 3.959532380292374e-06,
+      "loss": 0.8827,
+      "step": 4233
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8127175928290108,
+      "learning_rate": 3.955171458177551e-06,
+      "loss": 0.9414,
+      "step": 4234
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1396879624380032,
+      "learning_rate": 3.950812346818747e-06,
+      "loss": 1.0588,
+      "step": 4235
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1622062698105156,
+      "learning_rate": 3.946455047521746e-06,
+      "loss": 0.9378,
+      "step": 4236
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9829788052687346,
+      "learning_rate": 3.942099561591802e-06,
+      "loss": 0.7042,
+      "step": 4237
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1551273155197308,
+      "learning_rate": 3.937745890333623e-06,
+      "loss": 0.8831,
+      "step": 4238
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1219914942059517,
+      "learning_rate": 3.933394035051364e-06,
+      "loss": 1.1147,
+      "step": 4239
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9800008177654216,
+      "learning_rate": 3.929043997048647e-06,
+      "loss": 0.7258,
+      "step": 4240
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.020645047668561,
+      "learning_rate": 3.92469577762855e-06,
+      "loss": 0.8872,
+      "step": 4241
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0671524613954848,
+      "learning_rate": 3.9203493780935935e-06,
+      "loss": 0.717,
+      "step": 4242
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9157488592604021,
+      "learning_rate": 3.916004799745765e-06,
+      "loss": 0.6523,
+      "step": 4243
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9221660183387452,
+      "learning_rate": 3.911662043886504e-06,
+      "loss": 0.9458,
+      "step": 4244
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2928761486729188,
+      "learning_rate": 3.9073211118167065e-06,
+      "loss": 0.7485,
+      "step": 4245
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.665608457957941,
+      "learning_rate": 3.902982004836709e-06,
+      "loss": 0.5488,
+      "step": 4246
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.082395774623606,
+      "learning_rate": 3.8986447242463175e-06,
+      "loss": 0.7121,
+      "step": 4247
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.061834851809159,
+      "learning_rate": 3.894309271344786e-06,
+      "loss": 0.7989,
+      "step": 4248
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.137443941708953,
+      "learning_rate": 3.889975647430812e-06,
+      "loss": 1.0702,
+      "step": 4249
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9733558095220957,
+      "learning_rate": 3.885643853802557e-06,
+      "loss": 1.1201,
+      "step": 4250
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7851732853381811,
+      "learning_rate": 3.881313891757629e-06,
+      "loss": 0.7848,
+      "step": 4251
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9598348566938375,
+      "learning_rate": 3.876985762593091e-06,
+      "loss": 0.7552,
+      "step": 4252
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1257643567046207,
+      "learning_rate": 3.872659467605447e-06,
+      "loss": 1.0948,
+      "step": 4253
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.942610842851695,
+      "learning_rate": 3.868335008090664e-06,
+      "loss": 0.6279,
+      "step": 4254
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.643464756333124,
+      "learning_rate": 3.864012385344154e-06,
+      "loss": 0.9763,
+      "step": 4255
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8233635495212728,
+      "learning_rate": 3.859691600660772e-06,
+      "loss": 0.6964,
+      "step": 4256
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1888221671045645,
+      "learning_rate": 3.855372655334834e-06,
+      "loss": 1.0074,
+      "step": 4257
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1023933583958592,
+      "learning_rate": 3.851055550660098e-06,
+      "loss": 1.0646,
+      "step": 4258
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9305590168966359,
+      "learning_rate": 3.846740287929778e-06,
+      "loss": 0.7369,
+      "step": 4259
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7638634846733405,
+      "learning_rate": 3.8424268684365204e-06,
+      "loss": 0.4969,
+      "step": 4260
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5435247078393226,
+      "learning_rate": 3.838115293472438e-06,
+      "loss": 0.863,
+      "step": 4261
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0413103866975462,
+      "learning_rate": 3.8338055643290815e-06,
+      "loss": 0.8317,
+      "step": 4262
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5514405256657451,
+      "learning_rate": 3.829497682297445e-06,
+      "loss": 0.4668,
+      "step": 4263
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0379398541307272,
+      "learning_rate": 3.825191648667978e-06,
+      "loss": 1.1216,
+      "step": 4264
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2298717463404085,
+      "learning_rate": 3.820887464730577e-06,
+      "loss": 0.7673,
+      "step": 4265
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8074408855052175,
+      "learning_rate": 3.8165851317745705e-06,
+      "loss": 0.7398,
+      "step": 4266
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1526781557885915,
+      "learning_rate": 3.812284651088751e-06,
+      "loss": 0.7653,
+      "step": 4267
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0352342975871014,
+      "learning_rate": 3.8079860239613396e-06,
+      "loss": 0.9856,
+      "step": 4268
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9317978989781708,
+      "learning_rate": 3.8036892516800118e-06,
+      "loss": 0.7509,
+      "step": 4269
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1545560206415815,
+      "learning_rate": 3.799394335531891e-06,
+      "loss": 0.7229,
+      "step": 4270
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0078219435458857,
+      "learning_rate": 3.7951012768035313e-06,
+      "loss": 0.7361,
+      "step": 4271
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9812493255178973,
+      "learning_rate": 3.790810076780941e-06,
+      "loss": 1.1016,
+      "step": 4272
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9855069505159805,
+      "learning_rate": 3.7865207367495716e-06,
+      "loss": 0.985,
+      "step": 4273
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0430814861500721,
+      "learning_rate": 3.782233257994309e-06,
+      "loss": 0.6624,
+      "step": 4274
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0298798227616286,
+      "learning_rate": 3.777947641799491e-06,
+      "loss": 0.8019,
+      "step": 4275
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4336406163089799,
+      "learning_rate": 3.773663889448894e-06,
+      "loss": 0.9254,
+      "step": 4276
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6487847348714654,
+      "learning_rate": 3.769382002225732e-06,
+      "loss": 0.9873,
+      "step": 4277
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9795305731684971,
+      "learning_rate": 3.7651019814126656e-06,
+      "loss": 0.9383,
+      "step": 4278
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8537919964659456,
+      "learning_rate": 3.7608238282917953e-06,
+      "loss": 0.6947,
+      "step": 4279
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.847289103146656,
+      "learning_rate": 3.756547544144664e-06,
+      "loss": 0.9276,
+      "step": 4280
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8810285445207519,
+      "learning_rate": 3.7522731302522453e-06,
+      "loss": 0.7496,
+      "step": 4281
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5377467335091262,
+      "learning_rate": 3.7480005878949645e-06,
+      "loss": 0.4149,
+      "step": 4282
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0605766193280581,
+      "learning_rate": 3.743729918352683e-06,
+      "loss": 1.0849,
+      "step": 4283
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0334257908713123,
+      "learning_rate": 3.7394611229046928e-06,
+      "loss": 0.7751,
+      "step": 4284
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1437662033959513,
+      "learning_rate": 3.735194202829735e-06,
+      "loss": 0.787,
+      "step": 4285
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2091292700735699,
+      "learning_rate": 3.730929159405984e-06,
+      "loss": 0.8172,
+      "step": 4286
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0327270758808447,
+      "learning_rate": 3.7266659939110585e-06,
+      "loss": 1.1468,
+      "step": 4287
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1065835701897155,
+      "learning_rate": 3.7224047076219995e-06,
+      "loss": 0.7129,
+      "step": 4288
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1164543927751842,
+      "learning_rate": 3.7181453018153e-06,
+      "loss": 0.704,
+      "step": 4289
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1033282809359686,
+      "learning_rate": 3.713887777766888e-06,
+      "loss": 0.848,
+      "step": 4290
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0887594524029542,
+      "learning_rate": 3.709632136752117e-06,
+      "loss": 0.7763,
+      "step": 4291
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9764902392308612,
+      "learning_rate": 3.7053783800457877e-06,
+      "loss": 1.0514,
+      "step": 4292
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9885885659831265,
+      "learning_rate": 3.7011265089221314e-06,
+      "loss": 0.7121,
+      "step": 4293
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1570197410007308,
+      "learning_rate": 3.69687652465482e-06,
+      "loss": 0.8173,
+      "step": 4294
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.93787813422672,
+      "learning_rate": 3.692628428516948e-06,
+      "loss": 0.8356,
+      "step": 4295
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.004909808047172,
+      "learning_rate": 3.688382221781055e-06,
+      "loss": 0.7874,
+      "step": 4296
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9463335790882615,
+      "learning_rate": 3.6841379057191163e-06,
+      "loss": 0.926,
+      "step": 4297
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0141374907317997,
+      "learning_rate": 3.679895481602529e-06,
+      "loss": 0.7005,
+      "step": 4298
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0103818115286738,
+      "learning_rate": 3.675654950702138e-06,
+      "loss": 0.7316,
+      "step": 4299
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6013792799859438,
+      "learning_rate": 3.671416314288204e-06,
+      "loss": 0.4517,
+      "step": 4300
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2965225384802928,
+      "learning_rate": 3.667179573630437e-06,
+      "loss": 0.753,
+      "step": 4301
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.327051848759106,
+      "learning_rate": 3.6629447299979747e-06,
+      "loss": 1.185,
+      "step": 4302
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6527966271608444,
+      "learning_rate": 3.6587117846593747e-06,
+      "loss": 0.9934,
+      "step": 4303
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8372738273356821,
+      "learning_rate": 3.654480738882644e-06,
+      "loss": 0.6715,
+      "step": 4304
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0193991646682379,
+      "learning_rate": 3.6502515939352045e-06,
+      "loss": 0.7917,
+      "step": 4305
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.034771700874117,
+      "learning_rate": 3.64602435108392e-06,
+      "loss": 1.0221,
+      "step": 4306
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1027885180878052,
+      "learning_rate": 3.6417990115950787e-06,
+      "loss": 0.8588,
+      "step": 4307
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9428777634543729,
+      "learning_rate": 3.6375755767344047e-06,
+      "loss": 0.7609,
+      "step": 4308
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1655218648671624,
+      "learning_rate": 3.6333540477670405e-06,
+      "loss": 0.7808,
+      "step": 4309
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.960808518758187,
+      "learning_rate": 3.629134425957567e-06,
+      "loss": 0.8236,
+      "step": 4310
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1154007798248153,
+      "learning_rate": 3.624916712569996e-06,
+      "loss": 1.0553,
+      "step": 4311
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.987840132869134,
+      "learning_rate": 3.6207009088677546e-06,
+      "loss": 0.7756,
+      "step": 4312
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0887065817225765,
+      "learning_rate": 3.616487016113711e-06,
+      "loss": 0.7814,
+      "step": 4313
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.01799971650013,
+      "learning_rate": 3.612275035570155e-06,
+      "loss": 0.9482,
+      "step": 4314
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9786594805552037,
+      "learning_rate": 3.6080649684988077e-06,
+      "loss": 0.7367,
+      "step": 4315
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0663943910118794,
+      "learning_rate": 3.6038568161608088e-06,
+      "loss": 1.0131,
+      "step": 4316
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.017005224182653,
+      "learning_rate": 3.599650579816731e-06,
+      "loss": 1.0886,
+      "step": 4317
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0839350532746572,
+      "learning_rate": 3.595446260726576e-06,
+      "loss": 0.7743,
+      "step": 4318
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.46114289026921523,
+      "learning_rate": 3.591243860149759e-06,
+      "loss": 0.3225,
+      "step": 4319
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8974414270301074,
+      "learning_rate": 3.587043379345134e-06,
+      "loss": 0.8467,
+      "step": 4320
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.209385183459414,
+      "learning_rate": 3.582844819570972e-06,
+      "loss": 0.8473,
+      "step": 4321
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0926092681821196,
+      "learning_rate": 3.578648182084975e-06,
+      "loss": 0.8695,
+      "step": 4322
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1287334663458197,
+      "learning_rate": 3.5744534681442578e-06,
+      "loss": 0.6913,
+      "step": 4323
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9373391136634719,
+      "learning_rate": 3.5702606790053695e-06,
+      "loss": 0.7828,
+      "step": 4324
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0057948870995514,
+      "learning_rate": 3.5660698159242835e-06,
+      "loss": 0.8085,
+      "step": 4325
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9883121880695739,
+      "learning_rate": 3.561880880156384e-06,
+      "loss": 1.0588,
+      "step": 4326
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7673151446734817,
+      "learning_rate": 3.5576938729564904e-06,
+      "loss": 0.9703,
+      "step": 4327
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.797903058451683,
+      "learning_rate": 3.5535087955788396e-06,
+      "loss": 1.0647,
+      "step": 4328
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0822809575966827,
+      "learning_rate": 3.5493256492770935e-06,
+      "loss": 0.694,
+      "step": 4329
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0099927880580395,
+      "learning_rate": 3.5451444353043263e-06,
+      "loss": 0.9727,
+      "step": 4330
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.13069466206447,
+      "learning_rate": 3.5409651549130465e-06,
+      "loss": 0.8011,
+      "step": 4331
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0621535894696523,
+      "learning_rate": 3.536787809355169e-06,
+      "loss": 0.7147,
+      "step": 4332
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3661762108570965,
+      "learning_rate": 3.5326123998820417e-06,
+      "loss": 0.9138,
+      "step": 4333
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1070128348161983,
+      "learning_rate": 3.5284389277444296e-06,
+      "loss": 0.67,
+      "step": 4334
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5816420994743455,
+      "learning_rate": 3.5242673941925085e-06,
+      "loss": 0.5997,
+      "step": 4335
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3290910145383272,
+      "learning_rate": 3.520097800475889e-06,
+      "loss": 0.8841,
+      "step": 4336
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9274192901782211,
+      "learning_rate": 3.5159301478435837e-06,
+      "loss": 0.7015,
+      "step": 4337
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0499533701488755,
+      "learning_rate": 3.511764437544036e-06,
+      "loss": 0.8625,
+      "step": 4338
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9860173908780502,
+      "learning_rate": 3.507600670825105e-06,
+      "loss": 1.0651,
+      "step": 4339
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8806115885098712,
+      "learning_rate": 3.503438848934063e-06,
+      "loss": 1.0614,
+      "step": 4340
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2008399149636835,
+      "learning_rate": 3.4992789731176024e-06,
+      "loss": 0.7499,
+      "step": 4341
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9927856686609908,
+      "learning_rate": 3.4951210446218385e-06,
+      "loss": 0.7565,
+      "step": 4342
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1965212292729748,
+      "learning_rate": 3.49096506469229e-06,
+      "loss": 0.7799,
+      "step": 4343
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9142519458288256,
+      "learning_rate": 3.4868110345739036e-06,
+      "loss": 0.6724,
+      "step": 4344
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3528865062041842,
+      "learning_rate": 3.482658955511038e-06,
+      "loss": 1.0712,
+      "step": 4345
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1601943344403975,
+      "learning_rate": 3.478508828747471e-06,
+      "loss": 0.7741,
+      "step": 4346
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9380706131069145,
+      "learning_rate": 3.4743606555263842e-06,
+      "loss": 0.7865,
+      "step": 4347
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3464454189075035,
+      "learning_rate": 3.470214437090387e-06,
+      "loss": 0.8599,
+      "step": 4348
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9218840642432794,
+      "learning_rate": 3.4660701746814996e-06,
+      "loss": 0.8718,
+      "step": 4349
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0799012191256288,
+      "learning_rate": 3.4619278695411495e-06,
+      "loss": 0.8059,
+      "step": 4350
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0828019912340205,
+      "learning_rate": 3.457787522910185e-06,
+      "loss": 0.802,
+      "step": 4351
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9041814465771219,
+      "learning_rate": 3.4536491360288684e-06,
+      "loss": 1.0045,
+      "step": 4352
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.65032260356454,
+      "learning_rate": 3.4495127101368732e-06,
+      "loss": 0.9548,
+      "step": 4353
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5736700310761577,
+      "learning_rate": 3.445378246473279e-06,
+      "loss": 0.6973,
+      "step": 4354
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.812226719133977,
+      "learning_rate": 3.441245746276586e-06,
+      "loss": 0.6427,
+      "step": 4355
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1145371749987425,
+      "learning_rate": 3.4371152107847074e-06,
+      "loss": 0.7569,
+      "step": 4356
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9685084870000815,
+      "learning_rate": 3.4329866412349578e-06,
+      "loss": 0.7854,
+      "step": 4357
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9461767974315963,
+      "learning_rate": 3.4288600388640714e-06,
+      "loss": 0.8132,
+      "step": 4358
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1172230593192813,
+      "learning_rate": 3.4247354049081914e-06,
+      "loss": 1.1132,
+      "step": 4359
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9245234886007185,
+      "learning_rate": 3.4206127406028744e-06,
+      "loss": 0.9325,
+      "step": 4360
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0055476016177356,
+      "learning_rate": 3.4164920471830763e-06,
+      "loss": 0.7406,
+      "step": 4361
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8942010697991981,
+      "learning_rate": 3.4123733258831727e-06,
+      "loss": 0.6916,
+      "step": 4362
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9556150921282468,
+      "learning_rate": 3.4082565779369493e-06,
+      "loss": 0.9632,
+      "step": 4363
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.472643255100664,
+      "learning_rate": 3.40414180457759e-06,
+      "loss": 0.9546,
+      "step": 4364
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9900565809057538,
+      "learning_rate": 3.4000290070376963e-06,
+      "loss": 0.6459,
+      "step": 4365
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0928902007342474,
+      "learning_rate": 3.3959181865492798e-06,
+      "loss": 1.019,
+      "step": 4366
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9054209648270516,
+      "learning_rate": 3.3918093443437483e-06,
+      "loss": 0.7386,
+      "step": 4367
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9206612878391579,
+      "learning_rate": 3.387702481651931e-06,
+      "loss": 0.7581,
+      "step": 4368
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2029223413443735,
+      "learning_rate": 3.3835975997040525e-06,
+      "loss": 1.2099,
+      "step": 4369
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9465297696584911,
+      "learning_rate": 3.3794946997297527e-06,
+      "loss": 0.7225,
+      "step": 4370
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9285630721464637,
+      "learning_rate": 3.3753937829580697e-06,
+      "loss": 0.8096,
+      "step": 4371
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.99592957259303,
+      "learning_rate": 3.3712948506174535e-06,
+      "loss": 0.7084,
+      "step": 4372
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6888477664877418,
+      "learning_rate": 3.3671979039357603e-06,
+      "loss": 0.5803,
+      "step": 4373
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.07275029589223,
+      "learning_rate": 3.3631029441402497e-06,
+      "loss": 0.7893,
+      "step": 4374
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6530837278282363,
+      "learning_rate": 3.359009972457581e-06,
+      "loss": 1.0071,
+      "step": 4375
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1076983918815326,
+      "learning_rate": 3.3549189901138246e-06,
+      "loss": 0.7438,
+      "step": 4376
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.824741873877647,
+      "learning_rate": 3.3508299983344572e-06,
+      "loss": 0.5884,
+      "step": 4377
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0406963884499854,
+      "learning_rate": 3.3467429983443477e-06,
+      "loss": 1.1419,
+      "step": 4378
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0398210683852698,
+      "learning_rate": 3.3426579913677804e-06,
+      "loss": 0.6639,
+      "step": 4379
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9328460280524136,
+      "learning_rate": 3.338574978628436e-06,
+      "loss": 0.8414,
+      "step": 4380
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9292647233033736,
+      "learning_rate": 3.3344939613494032e-06,
+      "loss": 1.0197,
+      "step": 4381
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0231461392289989,
+      "learning_rate": 3.330414940753164e-06,
+      "loss": 0.7418,
+      "step": 4382
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0063978870658046,
+      "learning_rate": 3.3263379180616097e-06,
+      "loss": 1.0689,
+      "step": 4383
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0189442260219446,
+      "learning_rate": 3.3222628944960346e-06,
+      "loss": 0.7734,
+      "step": 4384
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0163520217458848,
+      "learning_rate": 3.3181898712771242e-06,
+      "loss": 0.786,
+      "step": 4385
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1586516474232558,
+      "learning_rate": 3.3141188496249755e-06,
+      "loss": 0.7966,
+      "step": 4386
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.04691063863104,
+      "learning_rate": 3.31004983075908e-06,
+      "loss": 0.7329,
+      "step": 4387
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.02239536625792,
+      "learning_rate": 3.305982815898334e-06,
+      "loss": 1.0614,
+      "step": 4388
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.038920037303308,
+      "learning_rate": 3.3019178062610256e-06,
+      "loss": 0.7316,
+      "step": 4389
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2967431193960195,
+      "learning_rate": 3.2978548030648504e-06,
+      "loss": 0.7581,
+      "step": 4390
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5548559416683218,
+      "learning_rate": 3.2937938075269004e-06,
+      "loss": 0.4437,
+      "step": 4391
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0817066897479357,
+      "learning_rate": 3.2897348208636616e-06,
+      "loss": 1.1855,
+      "step": 4392
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9876354992320763,
+      "learning_rate": 3.285677844291023e-06,
+      "loss": 0.5944,
+      "step": 4393
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4362402270923278,
+      "learning_rate": 3.2816228790242723e-06,
+      "loss": 0.7321,
+      "step": 4394
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3179035903686245,
+      "learning_rate": 3.2775699262780938e-06,
+      "loss": 0.7774,
+      "step": 4395
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.009879832183031,
+      "learning_rate": 3.273518987266564e-06,
+      "loss": 0.6196,
+      "step": 4396
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0289983096858497,
+      "learning_rate": 3.2694700632031627e-06,
+      "loss": 1.1312,
+      "step": 4397
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8705174633901298,
+      "learning_rate": 3.2654231553007665e-06,
+      "loss": 0.7534,
+      "step": 4398
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9228722247684288,
+      "learning_rate": 3.2613782647716374e-06,
+      "loss": 1.014,
+      "step": 4399
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9280307280859934,
+      "learning_rate": 3.257335392827451e-06,
+      "loss": 0.6963,
+      "step": 4400
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.94839809624927,
+      "learning_rate": 3.2532945406792573e-06,
+      "loss": 0.9853,
+      "step": 4401
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.269852842790488,
+      "learning_rate": 3.249255709537521e-06,
+      "loss": 0.814,
+      "step": 4402
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5564081992564844,
+      "learning_rate": 3.245218900612086e-06,
+      "loss": 1.0047,
+      "step": 4403
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9356802342941763,
+      "learning_rate": 3.2411841151121993e-06,
+      "loss": 0.6852,
+      "step": 4404
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0252042850476353,
+      "learning_rate": 3.2371513542465027e-06,
+      "loss": 0.8725,
+      "step": 4405
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3068072794148666,
+      "learning_rate": 3.233120619223021e-06,
+      "loss": 1.0691,
+      "step": 4406
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0583017085802726,
+      "learning_rate": 3.229091911249185e-06,
+      "loss": 0.7267,
+      "step": 4407
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5303152297050318,
+      "learning_rate": 3.2250652315318097e-06,
+      "loss": 0.4877,
+      "step": 4408
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3816955679653538,
+      "learning_rate": 3.2210405812771097e-06,
+      "loss": 0.8981,
+      "step": 4409
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.133923197650525,
+      "learning_rate": 3.2170179616906818e-06,
+      "loss": 0.8819,
+      "step": 4410
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9976507580602946,
+      "learning_rate": 3.2129973739775223e-06,
+      "loss": 1.025,
+      "step": 4411
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0132537172588845,
+      "learning_rate": 3.20897881934202e-06,
+      "loss": 0.6207,
+      "step": 4412
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8618895960283135,
+      "learning_rate": 3.2049622989879446e-06,
+      "loss": 0.6539,
+      "step": 4413
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.874253684106084,
+      "learning_rate": 3.200947814118467e-06,
+      "loss": 0.7582,
+      "step": 4414
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0827289493581627,
+      "learning_rate": 3.1969353659361447e-06,
+      "loss": 0.7901,
+      "step": 4415
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.045914599420475,
+      "learning_rate": 3.1929249556429276e-06,
+      "loss": 1.1252,
+      "step": 4416
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.054304507181181,
+      "learning_rate": 3.1889165844401447e-06,
+      "loss": 0.7546,
+      "step": 4417
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2400333856202426,
+      "learning_rate": 3.1849102535285283e-06,
+      "loss": 0.9805,
+      "step": 4418
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9657060197353793,
+      "learning_rate": 3.180905964108193e-06,
+      "loss": 0.7449,
+      "step": 4419
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1181245126841084,
+      "learning_rate": 3.1769037173786376e-06,
+      "loss": 0.742,
+      "step": 4420
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0097270038357542,
+      "learning_rate": 3.1729035145387566e-06,
+      "loss": 1.0654,
+      "step": 4421
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9458411199849818,
+      "learning_rate": 3.1689053567868278e-06,
+      "loss": 0.6231,
+      "step": 4422
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9679217860279152,
+      "learning_rate": 3.1649092453205223e-06,
+      "loss": 1.0315,
+      "step": 4423
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8401303357560744,
+      "learning_rate": 3.1609151813368853e-06,
+      "loss": 0.9893,
+      "step": 4424
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6551206295262044,
+      "learning_rate": 3.156923166032362e-06,
+      "loss": 0.8036,
+      "step": 4425
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6436626807493201,
+      "learning_rate": 3.1529332006027813e-06,
+      "loss": 0.6236,
+      "step": 4426
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0316143863464473,
+      "learning_rate": 3.148945286243349e-06,
+      "loss": 0.6899,
+      "step": 4427
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.068248819968565,
+      "learning_rate": 3.144959424148666e-06,
+      "loss": 0.8518,
+      "step": 4428
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.825886877084319,
+      "learning_rate": 3.1409756155127156e-06,
+      "loss": 0.6668,
+      "step": 4429
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6388492117279332,
+      "learning_rate": 3.1369938615288707e-06,
+      "loss": 1.0746,
+      "step": 4430
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9795414133498165,
+      "learning_rate": 3.133014163389876e-06,
+      "loss": 0.7741,
+      "step": 4431
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0028619774762164,
+      "learning_rate": 3.1290365222878736e-06,
+      "loss": 0.7189,
+      "step": 4432
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8993635879371813,
+      "learning_rate": 3.1250609394143793e-06,
+      "loss": 0.7171,
+      "step": 4433
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0900602057800324,
+      "learning_rate": 3.1210874159603044e-06,
+      "loss": 0.7213,
+      "step": 4434
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3075518552924574,
+      "learning_rate": 3.1171159531159266e-06,
+      "loss": 1.2739,
+      "step": 4435
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.34300573777671,
+      "learning_rate": 3.1131465520709227e-06,
+      "loss": 0.813,
+      "step": 4436
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.860491427397129,
+      "learning_rate": 3.109179214014344e-06,
+      "loss": 0.5355,
+      "step": 4437
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8922091140555717,
+      "learning_rate": 3.1052139401346216e-06,
+      "loss": 0.6178,
+      "step": 4438
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.912414941658939,
+      "learning_rate": 3.101250731619574e-06,
+      "loss": 0.6179,
+      "step": 4439
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0565329629805722,
+      "learning_rate": 3.0972895896564004e-06,
+      "loss": 1.0113,
+      "step": 4440
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9210351497395616,
+      "learning_rate": 3.093330515431673e-06,
+      "loss": 0.7376,
+      "step": 4441
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1140624015456435,
+      "learning_rate": 3.089373510131354e-06,
+      "loss": 0.8933,
+      "step": 4442
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9493067580705897,
+      "learning_rate": 3.085418574940786e-06,
+      "loss": 0.8059,
+      "step": 4443
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8841473948430247,
+      "learning_rate": 3.081465711044681e-06,
+      "loss": 0.8699,
+      "step": 4444
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.050403535086893,
+      "learning_rate": 3.0775149196271413e-06,
+      "loss": 1.0557,
+      "step": 4445
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4964256901388761,
+      "learning_rate": 3.0735662018716437e-06,
+      "loss": 0.3708,
+      "step": 4446
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1601137867455056,
+      "learning_rate": 3.069619558961049e-06,
+      "loss": 0.8578,
+      "step": 4447
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.832017261826355,
+      "learning_rate": 3.065674992077584e-06,
+      "loss": 0.8875,
+      "step": 4448
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2316185899393772,
+      "learning_rate": 3.0617325024028664e-06,
+      "loss": 0.9118,
+      "step": 4449
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8647113676564149,
+      "learning_rate": 3.057792091117889e-06,
+      "loss": 1.1004,
+      "step": 4450
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2467249566037673,
+      "learning_rate": 3.053853759403015e-06,
+      "loss": 0.7197,
+      "step": 4451
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.62353134041894,
+      "learning_rate": 3.0499175084379916e-06,
+      "loss": 0.9709,
+      "step": 4452
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9772852654083176,
+      "learning_rate": 3.0459833394019412e-06,
+      "loss": 0.7976,
+      "step": 4453
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9916135465367631,
+      "learning_rate": 3.042051253473366e-06,
+      "loss": 0.9667,
+      "step": 4454
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9493265325827858,
+      "learning_rate": 3.038121251830133e-06,
+      "loss": 0.7841,
+      "step": 4455
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0078957223560325,
+      "learning_rate": 3.0341933356494957e-06,
+      "loss": 0.912,
+      "step": 4456
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.014758325815112,
+      "learning_rate": 3.030267506108083e-06,
+      "loss": 0.6658,
+      "step": 4457
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.065785513441575,
+      "learning_rate": 3.026343764381887e-06,
+      "loss": 0.8086,
+      "step": 4458
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8981026054320495,
+      "learning_rate": 3.022422111646288e-06,
+      "loss": 0.9955,
+      "step": 4459
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0190758261952702,
+      "learning_rate": 3.0185025490760346e-06,
+      "loss": 0.7128,
+      "step": 4460
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.719232083178561,
+      "learning_rate": 3.0145850778452502e-06,
+      "loss": 0.4681,
+      "step": 4461
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.122100295679984,
+      "learning_rate": 3.0106696991274278e-06,
+      "loss": 0.7209,
+      "step": 4462
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0172713818156056,
+      "learning_rate": 3.0067564140954384e-06,
+      "loss": 0.9065,
+      "step": 4463
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9575186303882345,
+      "learning_rate": 3.00284522392153e-06,
+      "loss": 0.9393,
+      "step": 4464
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.190415081720164,
+      "learning_rate": 2.998936129777308e-06,
+      "loss": 0.7801,
+      "step": 4465
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9883127796066973,
+      "learning_rate": 2.995029132833768e-06,
+      "loss": 0.6731,
+      "step": 4466
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0747069586382378,
+      "learning_rate": 2.991124234261261e-06,
+      "loss": 0.8709,
+      "step": 4467
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.011808938815264,
+      "learning_rate": 2.9872214352295213e-06,
+      "loss": 1.1675,
+      "step": 4468
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9486273342476256,
+      "learning_rate": 2.9833207369076535e-06,
+      "loss": 0.9269,
+      "step": 4469
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0851128579073375,
+      "learning_rate": 2.9794221404641233e-06,
+      "loss": 0.9268,
+      "step": 4470
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8299804011122811,
+      "learning_rate": 2.975525647066778e-06,
+      "loss": 0.9215,
+      "step": 4471
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0644890403162988,
+      "learning_rate": 2.9716312578828254e-06,
+      "loss": 0.7201,
+      "step": 4472
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8263072652050895,
+      "learning_rate": 2.96773897407885e-06,
+      "loss": 0.9507,
+      "step": 4473
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8544899809351566,
+      "learning_rate": 2.9638487968208028e-06,
+      "loss": 0.7422,
+      "step": 4474
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1674270692825905,
+      "learning_rate": 2.9599607272740084e-06,
+      "loss": 1.0009,
+      "step": 4475
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5745354688734569,
+      "learning_rate": 2.9560747666031477e-06,
+      "loss": 0.966,
+      "step": 4476
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7808703166590653,
+      "learning_rate": 2.952190915972283e-06,
+      "loss": 0.6233,
+      "step": 4477
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.220613414605657,
+      "learning_rate": 2.9483091765448426e-06,
+      "loss": 1.0653,
+      "step": 4478
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0406233126586957,
+      "learning_rate": 2.944429549483612e-06,
+      "loss": 0.7659,
+      "step": 4479
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.369550955739958,
+      "learning_rate": 2.9405520359507543e-06,
+      "loss": 0.9337,
+      "step": 4480
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5235836980651926,
+      "learning_rate": 2.936676637107797e-06,
+      "loss": 0.5081,
+      "step": 4481
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9984149934213487,
+      "learning_rate": 2.9328033541156363e-06,
+      "loss": 0.7224,
+      "step": 4482
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0398855376662992,
+      "learning_rate": 2.9289321881345257e-06,
+      "loss": 1.029,
+      "step": 4483
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.987614297118867,
+      "learning_rate": 2.9250631403240937e-06,
+      "loss": 0.8114,
+      "step": 4484
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7221872830674002,
+      "learning_rate": 2.9211962118433344e-06,
+      "loss": 0.5458,
+      "step": 4485
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0184596348584252,
+      "learning_rate": 2.9173314038505986e-06,
+      "loss": 0.7212,
+      "step": 4486
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2438160898947699,
+      "learning_rate": 2.913468717503609e-06,
+      "loss": 0.8715,
+      "step": 4487
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.785329063616119,
+      "learning_rate": 2.909608153959451e-06,
+      "loss": 0.9458,
+      "step": 4488
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.127561711818556,
+      "learning_rate": 2.9057497143745784e-06,
+      "loss": 0.8399,
+      "step": 4489
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.052869406030682,
+      "learning_rate": 2.901893399904797e-06,
+      "loss": 1.073,
+      "step": 4490
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2286089081521905,
+      "learning_rate": 2.8980392117052867e-06,
+      "loss": 0.7983,
+      "step": 4491
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1620999168006407,
+      "learning_rate": 2.8941871509305906e-06,
+      "loss": 1.1255,
+      "step": 4492
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8984109395170363,
+      "learning_rate": 2.8903372187346056e-06,
+      "loss": 0.7383,
+      "step": 4493
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2793458780846398,
+      "learning_rate": 2.886489416270598e-06,
+      "loss": 0.8897,
+      "step": 4494
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0193304899743227,
+      "learning_rate": 2.882643744691196e-06,
+      "loss": 0.9072,
+      "step": 4495
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8154405902093131,
+      "learning_rate": 2.8788002051483897e-06,
+      "loss": 0.8731,
+      "step": 4496
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0738291234906867,
+      "learning_rate": 2.8749587987935236e-06,
+      "loss": 0.95,
+      "step": 4497
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5305950297875723,
+      "learning_rate": 2.871119526777315e-06,
+      "loss": 0.4379,
+      "step": 4498
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1183829623975912,
+      "learning_rate": 2.8672823902498293e-06,
+      "loss": 0.8087,
+      "step": 4499
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6150665982653892,
+      "learning_rate": 2.8634473903605008e-06,
+      "loss": 0.9499,
+      "step": 4500
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0048456486977395,
+      "learning_rate": 2.8596145282581255e-06,
+      "loss": 0.6494,
+      "step": 4501
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7498161390811944,
+      "learning_rate": 2.855783805090846e-06,
+      "loss": 0.8964,
+      "step": 4502
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.067800368009496,
+      "learning_rate": 2.851955222006182e-06,
+      "loss": 0.8262,
+      "step": 4503
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1397863852287153,
+      "learning_rate": 2.8481287801509947e-06,
+      "loss": 0.7728,
+      "step": 4504
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7663933258171762,
+      "learning_rate": 2.844304480671517e-06,
+      "loss": 0.7503,
+      "step": 4505
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9972825636576519,
+      "learning_rate": 2.8404823247133373e-06,
+      "loss": 0.9174,
+      "step": 4506
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8366642054152111,
+      "learning_rate": 2.8366623134213945e-06,
+      "loss": 0.9149,
+      "step": 4507
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4392134472256313,
+      "learning_rate": 2.832844447939992e-06,
+      "loss": 0.891,
+      "step": 4508
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0205808187541519,
+      "learning_rate": 2.8290287294127904e-06,
+      "loss": 0.5679,
+      "step": 4509
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.057364428312899,
+      "learning_rate": 2.825215158982809e-06,
+      "loss": 0.7746,
+      "step": 4510
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9149249187376515,
+      "learning_rate": 2.821403737792413e-06,
+      "loss": 0.8937,
+      "step": 4511
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.931566814542744,
+      "learning_rate": 2.8175944669833345e-06,
+      "loss": 0.956,
+      "step": 4512
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.959922906862624,
+      "learning_rate": 2.8137873476966605e-06,
+      "loss": 0.7036,
+      "step": 4513
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.962076799795303,
+      "learning_rate": 2.8099823810728255e-06,
+      "loss": 0.7291,
+      "step": 4514
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6012945037747213,
+      "learning_rate": 2.8061795682516267e-06,
+      "loss": 0.5675,
+      "step": 4515
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.98880109188778,
+      "learning_rate": 2.8023789103722163e-06,
+      "loss": 1.0122,
+      "step": 4516
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9083582260160931,
+      "learning_rate": 2.7985804085730994e-06,
+      "loss": 0.6743,
+      "step": 4517
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2031885949778076,
+      "learning_rate": 2.7947840639921308e-06,
+      "loss": 0.785,
+      "step": 4518
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0025114898465348,
+      "learning_rate": 2.7909898777665234e-06,
+      "loss": 0.7491,
+      "step": 4519
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8592019162855837,
+      "learning_rate": 2.787197851032848e-06,
+      "loss": 1.0859,
+      "step": 4520
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0279627965301752,
+      "learning_rate": 2.7834079849270158e-06,
+      "loss": 1.0811,
+      "step": 4521
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9294904437323295,
+      "learning_rate": 2.779620280584303e-06,
+      "loss": 0.7932,
+      "step": 4522
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2127095123866973,
+      "learning_rate": 2.7758347391393336e-06,
+      "loss": 0.8232,
+      "step": 4523
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.982641888702897,
+      "learning_rate": 2.7720513617260857e-06,
+      "loss": 0.6812,
+      "step": 4524
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3018805114409944,
+      "learning_rate": 2.768270149477882e-06,
+      "loss": 0.8702,
+      "step": 4525
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6322904071877089,
+      "learning_rate": 2.764491103527406e-06,
+      "loss": 1.0282,
+      "step": 4526
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9688150197910794,
+      "learning_rate": 2.7607142250066885e-06,
+      "loss": 0.8695,
+      "step": 4527
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9268284673916766,
+      "learning_rate": 2.756939515047108e-06,
+      "loss": 0.6108,
+      "step": 4528
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9321666230313019,
+      "learning_rate": 2.7531669747793965e-06,
+      "loss": 0.7516,
+      "step": 4529
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.052541403830305,
+      "learning_rate": 2.7493966053336397e-06,
+      "loss": 0.7789,
+      "step": 4530
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0223029603658622,
+      "learning_rate": 2.745628407839264e-06,
+      "loss": 0.9755,
+      "step": 4531
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9338629186481461,
+      "learning_rate": 2.74186238342505e-06,
+      "loss": 0.8856,
+      "step": 4532
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1182909147383333,
+      "learning_rate": 2.7380985332191344e-06,
+      "loss": 0.7947,
+      "step": 4533
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9855928600533633,
+      "learning_rate": 2.734336858348987e-06,
+      "loss": 0.9828,
+      "step": 4534
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5631685708983775,
+      "learning_rate": 2.730577359941443e-06,
+      "loss": 0.6513,
+      "step": 4535
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1072432027552486,
+      "learning_rate": 2.7268200391226697e-06,
+      "loss": 0.7334,
+      "step": 4536
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7696053772753749,
+      "learning_rate": 2.723064897018195e-06,
+      "loss": 0.6368,
+      "step": 4537
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2188248762055005,
+      "learning_rate": 2.7193119347528842e-06,
+      "loss": 0.7037,
+      "step": 4538
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0728582639618316,
+      "learning_rate": 2.7155611534509573e-06,
+      "loss": 0.7322,
+      "step": 4539
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0222669575323051,
+      "learning_rate": 2.7118125542359775e-06,
+      "loss": 1.0349,
+      "step": 4540
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2800928067271837,
+      "learning_rate": 2.708066138230858e-06,
+      "loss": 0.8937,
+      "step": 4541
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.022330191087422,
+      "learning_rate": 2.7043219065578497e-06,
+      "loss": 0.641,
+      "step": 4542
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1011060828214025,
+      "learning_rate": 2.700579860338556e-06,
+      "loss": 0.7565,
+      "step": 4543
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8288404562947844,
+      "learning_rate": 2.6968400006939267e-06,
+      "loss": 0.9666,
+      "step": 4544
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0584537666162228,
+      "learning_rate": 2.6931023287442494e-06,
+      "loss": 1.044,
+      "step": 4545
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1902888050229,
+      "learning_rate": 2.6893668456091627e-06,
+      "loss": 0.757,
+      "step": 4546
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8891313196757366,
+      "learning_rate": 2.685633552407648e-06,
+      "loss": 0.8755,
+      "step": 4547
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1964309047936863,
+      "learning_rate": 2.681902450258033e-06,
+      "loss": 0.7604,
+      "step": 4548
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9975822341966485,
+      "learning_rate": 2.6781735402779807e-06,
+      "loss": 0.7289,
+      "step": 4549
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1516506855493618,
+      "learning_rate": 2.674446823584507e-06,
+      "loss": 1.1301,
+      "step": 4550
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.639212983005533,
+      "learning_rate": 2.6707223012939675e-06,
+      "loss": 0.9364,
+      "step": 4551
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.939997006983501,
+      "learning_rate": 2.6669999745220566e-06,
+      "loss": 0.7076,
+      "step": 4552
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5475708368675075,
+      "learning_rate": 2.663279844383815e-06,
+      "loss": 0.4606,
+      "step": 4553
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.983664572760159,
+      "learning_rate": 2.6595619119936267e-06,
+      "loss": 1.0625,
+      "step": 4554
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0603622288544083,
+      "learning_rate": 2.655846178465217e-06,
+      "loss": 1.0872,
+      "step": 4555
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1387296331094572,
+      "learning_rate": 2.652132644911645e-06,
+      "loss": 0.7462,
+      "step": 4556
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1398728669966054,
+      "learning_rate": 2.6484213124453197e-06,
+      "loss": 0.648,
+      "step": 4557
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9853094636959262,
+      "learning_rate": 2.644712182177992e-06,
+      "loss": 0.9756,
+      "step": 4558
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9018558799401856,
+      "learning_rate": 2.641005255220741e-06,
+      "loss": 0.9605,
+      "step": 4559
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1334840712629664,
+      "learning_rate": 2.6373005326839973e-06,
+      "loss": 0.8338,
+      "step": 4560
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.085127404630968,
+      "learning_rate": 2.6335980156775274e-06,
+      "loss": 0.5764,
+      "step": 4561
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8158734373064758,
+      "learning_rate": 2.62989770531044e-06,
+      "loss": 0.6841,
+      "step": 4562
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1133581170033713,
+      "learning_rate": 2.626199602691175e-06,
+      "loss": 0.8471,
+      "step": 4563
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2381352986113112,
+      "learning_rate": 2.6225037089275186e-06,
+      "loss": 1.1131,
+      "step": 4564
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1217819559538225,
+      "learning_rate": 2.6188100251265947e-06,
+      "loss": 0.8803,
+      "step": 4565
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6054832445385656,
+      "learning_rate": 2.6151185523948577e-06,
+      "loss": 0.8629,
+      "step": 4566
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8020967875034769,
+      "learning_rate": 2.6114292918381113e-06,
+      "loss": 0.9329,
+      "step": 4567
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.085093055193904,
+      "learning_rate": 2.607742244561484e-06,
+      "loss": 0.6984,
+      "step": 4568
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9512942837784831,
+      "learning_rate": 2.6040574116694537e-06,
+      "loss": 1.0927,
+      "step": 4569
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9984721789215062,
+      "learning_rate": 2.6003747942658243e-06,
+      "loss": 0.8478,
+      "step": 4570
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5024018113005876,
+      "learning_rate": 2.5966943934537415e-06,
+      "loss": 0.4156,
+      "step": 4571
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3010196018305742,
+      "learning_rate": 2.5930162103356903e-06,
+      "loss": 0.798,
+      "step": 4572
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0426528738115426,
+      "learning_rate": 2.589340246013481e-06,
+      "loss": 0.7248,
+      "step": 4573
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9643905599752055,
+      "learning_rate": 2.5856665015882686e-06,
+      "loss": 0.9664,
+      "step": 4574
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6788564404938525,
+      "learning_rate": 2.5819949781605413e-06,
+      "loss": 0.9162,
+      "step": 4575
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0845211708610873,
+      "learning_rate": 2.5783256768301214e-06,
+      "loss": 0.7151,
+      "step": 4576
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8078859607404414,
+      "learning_rate": 2.5746585986961603e-06,
+      "loss": 0.9994,
+      "step": 4577
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6852418933107534,
+      "learning_rate": 2.570993744857151e-06,
+      "loss": 1.0968,
+      "step": 4578
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9968047052525063,
+      "learning_rate": 2.567331116410919e-06,
+      "loss": 0.9721,
+      "step": 4579
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.961526633306354,
+      "learning_rate": 2.563670714454617e-06,
+      "loss": 0.6084,
+      "step": 4580
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0408785995334693,
+      "learning_rate": 2.5600125400847374e-06,
+      "loss": 0.8027,
+      "step": 4581
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1989316449729193,
+      "learning_rate": 2.5563565943971027e-06,
+      "loss": 0.7903,
+      "step": 4582
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0375012298433135,
+      "learning_rate": 2.552702878486872e-06,
+      "loss": 1.0638,
+      "step": 4583
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.17385841201316,
+      "learning_rate": 2.549051393448526e-06,
+      "loss": 0.951,
+      "step": 4584
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0976807051361153,
+      "learning_rate": 2.5454021403758865e-06,
+      "loss": 0.8331,
+      "step": 4585
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9449166967486482,
+      "learning_rate": 2.541755120362108e-06,
+      "loss": 0.7203,
+      "step": 4586
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.116793576124499,
+      "learning_rate": 2.538110334499665e-06,
+      "loss": 0.6895,
+      "step": 4587
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9600299538509918,
+      "learning_rate": 2.534467783880373e-06,
+      "loss": 0.9945,
+      "step": 4588
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5486563978706557,
+      "learning_rate": 2.530827469595375e-06,
+      "loss": 0.4798,
+      "step": 4589
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9104868243730728,
+      "learning_rate": 2.5271893927351464e-06,
+      "loss": 0.8335,
+      "step": 4590
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4251855404307257,
+      "learning_rate": 2.523553554389484e-06,
+      "loss": 0.7817,
+      "step": 4591
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7134797892915685,
+      "learning_rate": 2.519919955647523e-06,
+      "loss": 0.8905,
+      "step": 4592
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8104946775328803,
+      "learning_rate": 2.516288597597727e-06,
+      "loss": 0.8207,
+      "step": 4593
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.002079843335256,
+      "learning_rate": 2.5126594813278792e-06,
+      "loss": 0.7924,
+      "step": 4594
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.049111353461549,
+      "learning_rate": 2.5090326079251025e-06,
+      "loss": 0.6681,
+      "step": 4595
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9293489420500165,
+      "learning_rate": 2.5054079784758433e-06,
+      "loss": 0.8706,
+      "step": 4596
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0657870981768394,
+      "learning_rate": 2.501785594065879e-06,
+      "loss": 0.7074,
+      "step": 4597
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.117937949880895,
+      "learning_rate": 2.4981654557803026e-06,
+      "loss": 1.0309,
+      "step": 4598
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.132759882998223,
+      "learning_rate": 2.494547564703552e-06,
+      "loss": 1.0606,
+      "step": 4599
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6679277200087794,
+      "learning_rate": 2.4909319219193774e-06,
+      "loss": 0.947,
+      "step": 4600
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1869325966422093,
+      "learning_rate": 2.487318528510865e-06,
+      "loss": 0.7559,
+      "step": 4601
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0075071497145476,
+      "learning_rate": 2.4837073855604186e-06,
+      "loss": 1.0235,
+      "step": 4602
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0303286188227934,
+      "learning_rate": 2.4800984941497754e-06,
+      "loss": 0.7914,
+      "step": 4603
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.935330798319725,
+      "learning_rate": 2.476491855359998e-06,
+      "loss": 0.6273,
+      "step": 4604
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8659335498045645,
+      "learning_rate": 2.4728874702714657e-06,
+      "loss": 0.7851,
+      "step": 4605
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0062683123135616,
+      "learning_rate": 2.469285339963892e-06,
+      "loss": 0.8057,
+      "step": 4606
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5755571379927243,
+      "learning_rate": 2.4656854655163132e-06,
+      "loss": 0.5844,
+      "step": 4607
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1574468321308031,
+      "learning_rate": 2.462087848007083e-06,
+      "loss": 0.8863,
+      "step": 4608
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.13237789538771,
+      "learning_rate": 2.4584924885138874e-06,
+      "loss": 0.7075,
+      "step": 4609
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0071241684177605,
+      "learning_rate": 2.4548993881137327e-06,
+      "loss": 0.8468,
+      "step": 4610
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0348712569813605,
+      "learning_rate": 2.4513085478829513e-06,
+      "loss": 0.8762,
+      "step": 4611
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1280761359295277,
+      "learning_rate": 2.44771996889719e-06,
+      "loss": 1.0187,
+      "step": 4612
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4166232693219651,
+      "learning_rate": 2.4441336522314265e-06,
+      "loss": 0.5622,
+      "step": 4613
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9094418350252649,
+      "learning_rate": 2.4405495989599627e-06,
+      "loss": 0.6482,
+      "step": 4614
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8761449597066889,
+      "learning_rate": 2.4369678101564098e-06,
+      "loss": 0.9953,
+      "step": 4615
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0213434411335203,
+      "learning_rate": 2.433388286893714e-06,
+      "loss": 1.0994,
+      "step": 4616
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0904062102018581,
+      "learning_rate": 2.429811030244138e-06,
+      "loss": 0.7554,
+      "step": 4617
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9992026262065637,
+      "learning_rate": 2.4262360412792663e-06,
+      "loss": 0.8068,
+      "step": 4618
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9833615028784866,
+      "learning_rate": 2.422663321069999e-06,
+      "loss": 1.0871,
+      "step": 4619
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9000746835027158,
+      "learning_rate": 2.4190928706865634e-06,
+      "loss": 0.6749,
+      "step": 4620
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.071341987295397,
+      "learning_rate": 2.4155246911985076e-06,
+      "loss": 0.6973,
+      "step": 4621
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0921351399645107,
+      "learning_rate": 2.4119587836746885e-06,
+      "loss": 1.1738,
+      "step": 4622
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.058648418399628,
+      "learning_rate": 2.4083951491832947e-06,
+      "loss": 0.7234,
+      "step": 4623
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.549701583618034,
+      "learning_rate": 2.4048337887918284e-06,
+      "loss": 0.9739,
+      "step": 4624
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.974228915794525,
+      "learning_rate": 2.401274703567115e-06,
+      "loss": 0.8024,
+      "step": 4625
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6150220274178035,
+      "learning_rate": 2.3977178945752877e-06,
+      "loss": 0.6382,
+      "step": 4626
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8091929933667474,
+      "learning_rate": 2.3941633628818085e-06,
+      "loss": 0.5427,
+      "step": 4627
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1291553084459696,
+      "learning_rate": 2.390611109551456e-06,
+      "loss": 0.7712,
+      "step": 4628
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2810439311822657,
+      "learning_rate": 2.3870611356483196e-06,
+      "loss": 0.7505,
+      "step": 4629
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9713506347309296,
+      "learning_rate": 2.383513442235812e-06,
+      "loss": 0.6844,
+      "step": 4630
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1063568412714677,
+      "learning_rate": 2.3799680303766636e-06,
+      "loss": 1.0843,
+      "step": 4631
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.06320206128027,
+      "learning_rate": 2.3764249011329133e-06,
+      "loss": 0.7514,
+      "step": 4632
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9328663648730846,
+      "learning_rate": 2.372884055565928e-06,
+      "loss": 0.676,
+      "step": 4633
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2650866476519913,
+      "learning_rate": 2.3693454947363782e-06,
+      "loss": 0.9131,
+      "step": 4634
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8276491049500857,
+      "learning_rate": 2.365809219704257e-06,
+      "loss": 0.7402,
+      "step": 4635
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0124060547542209,
+      "learning_rate": 2.3622752315288766e-06,
+      "loss": 1.0273,
+      "step": 4636
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8840351348734583,
+      "learning_rate": 2.3587435312688546e-06,
+      "loss": 0.7387,
+      "step": 4637
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8330413169475779,
+      "learning_rate": 2.35521411998213e-06,
+      "loss": 0.9414,
+      "step": 4638
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1059380966444141,
+      "learning_rate": 2.3516869987259516e-06,
+      "loss": 0.7802,
+      "step": 4639
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.975004982333513,
+      "learning_rate": 2.3481621685568867e-06,
+      "loss": 0.671,
+      "step": 4640
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9817619742339524,
+      "learning_rate": 2.344639630530814e-06,
+      "loss": 0.9494,
+      "step": 4641
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.126097426923567,
+      "learning_rate": 2.341119385702929e-06,
+      "loss": 0.6686,
+      "step": 4642
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5738256615701355,
+      "learning_rate": 2.337601435127731e-06,
+      "loss": 0.5099,
+      "step": 4643
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9678318897281324,
+      "learning_rate": 2.334085779859041e-06,
+      "loss": 1.0248,
+      "step": 4644
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0300262780776794,
+      "learning_rate": 2.3305724209499924e-06,
+      "loss": 1.1043,
+      "step": 4645
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9509260817402522,
+      "learning_rate": 2.3270613594530235e-06,
+      "loss": 0.6855,
+      "step": 4646
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0428073664505813,
+      "learning_rate": 2.323552596419889e-06,
+      "loss": 0.8602,
+      "step": 4647
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5999104626146107,
+      "learning_rate": 2.3200461329016575e-06,
+      "loss": 0.966,
+      "step": 4648
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9045260440088593,
+      "learning_rate": 2.316541969948709e-06,
+      "loss": 0.6123,
+      "step": 4649
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1721429609566225,
+      "learning_rate": 2.3130401086107247e-06,
+      "loss": 1.1016,
+      "step": 4650
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8333264804041608,
+      "learning_rate": 2.309540549936705e-06,
+      "loss": 0.7614,
+      "step": 4651
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7744580245396198,
+      "learning_rate": 2.3060432949749635e-06,
+      "loss": 0.5497,
+      "step": 4652
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.902320631432354,
+      "learning_rate": 2.3025483447731124e-06,
+      "loss": 0.8335,
+      "step": 4653
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7334983302807319,
+      "learning_rate": 2.2990557003780833e-06,
+      "loss": 0.6507,
+      "step": 4654
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.920676804756508,
+      "learning_rate": 2.295565362836113e-06,
+      "loss": 1.0194,
+      "step": 4655
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0497025300525078,
+      "learning_rate": 2.2920773331927515e-06,
+      "loss": 0.7635,
+      "step": 4656
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.439718906774242,
+      "learning_rate": 2.2885916124928487e-06,
+      "loss": 0.8816,
+      "step": 4657
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0232157488548146,
+      "learning_rate": 2.2851082017805704e-06,
+      "loss": 0.8346,
+      "step": 4658
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0239475840178878,
+      "learning_rate": 2.2816271020993906e-06,
+      "loss": 0.6735,
+      "step": 4659
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0441957996863227,
+      "learning_rate": 2.2781483144920833e-06,
+      "loss": 1.1259,
+      "step": 4660
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5331878497784124,
+      "learning_rate": 2.2746718400007394e-06,
+      "loss": 0.4731,
+      "step": 4661
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8151222140450324,
+      "learning_rate": 2.27119767966675e-06,
+      "loss": 0.9579,
+      "step": 4662
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9266981069776743,
+      "learning_rate": 2.267725834530822e-06,
+      "loss": 0.6783,
+      "step": 4663
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9017332070156255,
+      "learning_rate": 2.2642563056329524e-06,
+      "loss": 1.2385,
+      "step": 4664
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.24056434483198,
+      "learning_rate": 2.2607890940124646e-06,
+      "loss": 0.5238,
+      "step": 4665
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7555823476180696,
+      "learning_rate": 2.2573242007079697e-06,
+      "loss": 0.6985,
+      "step": 4666
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2916819702541953,
+      "learning_rate": 2.2538616267573953e-06,
+      "loss": 0.9508,
+      "step": 4667
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0298699243956098,
+      "learning_rate": 2.2504013731979735e-06,
+      "loss": 0.7003,
+      "step": 4668
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1626816450914095,
+      "learning_rate": 2.246943441066235e-06,
+      "loss": 1.1536,
+      "step": 4669
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9348336911948968,
+      "learning_rate": 2.2434878313980235e-06,
+      "loss": 0.755,
+      "step": 4670
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0230121604072613,
+      "learning_rate": 2.2400345452284787e-06,
+      "loss": 0.9542,
+      "step": 4671
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0724714656467211,
+      "learning_rate": 2.236583583592049e-06,
+      "loss": 0.7412,
+      "step": 4672
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.675715048246619,
+      "learning_rate": 2.2331349475224894e-06,
+      "loss": 0.9504,
+      "step": 4673
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0163872265586675,
+      "learning_rate": 2.2296886380528494e-06,
+      "loss": 0.9075,
+      "step": 4674
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8598448707128539,
+      "learning_rate": 2.2262446562154883e-06,
+      "loss": 0.7864,
+      "step": 4675
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.248500844793822,
+      "learning_rate": 2.222803003042069e-06,
+      "loss": 0.8308,
+      "step": 4676
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8715063830701603,
+      "learning_rate": 2.2193636795635543e-06,
+      "loss": 0.8009,
+      "step": 4677
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5347708777848189,
+      "learning_rate": 2.215926686810206e-06,
+      "loss": 0.444,
+      "step": 4678
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1255942687143827,
+      "learning_rate": 2.2124920258115924e-06,
+      "loss": 1.0623,
+      "step": 4679
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8753084299851891,
+      "learning_rate": 2.209059697596585e-06,
+      "loss": 0.769,
+      "step": 4680
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.141086251042129,
+      "learning_rate": 2.2056297031933484e-06,
+      "loss": 0.8297,
+      "step": 4681
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1197904297765116,
+      "learning_rate": 2.202202043629356e-06,
+      "loss": 0.7122,
+      "step": 4682
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0787325412697324,
+      "learning_rate": 2.198776719931377e-06,
+      "loss": 1.0388,
+      "step": 4683
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8945693312156233,
+      "learning_rate": 2.195353733125487e-06,
+      "loss": 0.8405,
+      "step": 4684
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0174153764987415,
+      "learning_rate": 2.191933084237052e-06,
+      "loss": 0.6469,
+      "step": 4685
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9838188233629303,
+      "learning_rate": 2.1885147742907465e-06,
+      "loss": 0.9643,
+      "step": 4686
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9906097166070951,
+      "learning_rate": 2.1850988043105413e-06,
+      "loss": 0.7357,
+      "step": 4687
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.827531308514137,
+      "learning_rate": 2.1816851753197023e-06,
+      "loss": 1.2131,
+      "step": 4688
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8841443912820066,
+      "learning_rate": 2.1782738883407996e-06,
+      "loss": 0.5255,
+      "step": 4689
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.079090662267995,
+      "learning_rate": 2.1748649443957004e-06,
+      "loss": 0.6859,
+      "step": 4690
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0904248623659207,
+      "learning_rate": 2.1714583445055715e-06,
+      "loss": 0.715,
+      "step": 4691
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9905816912442533,
+      "learning_rate": 2.1680540896908696e-06,
+      "loss": 0.744,
+      "step": 4692
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0501812559049497,
+      "learning_rate": 2.1646521809713583e-06,
+      "loss": 1.0277,
+      "step": 4693
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9160215529479726,
+      "learning_rate": 2.1612526193660975e-06,
+      "loss": 0.7774,
+      "step": 4694
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5033218568327542,
+      "learning_rate": 2.1578554058934364e-06,
+      "loss": 0.9566,
+      "step": 4695
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9539691741868406,
+      "learning_rate": 2.1544605415710275e-06,
+      "loss": 0.6374,
+      "step": 4696
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.44389681756422517,
+      "learning_rate": 2.1510680274158214e-06,
+      "loss": 0.7862,
+      "step": 4697
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1301838380494689,
+      "learning_rate": 2.1476778644440553e-06,
+      "loss": 1.1284,
+      "step": 4698
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8664665494901707,
+      "learning_rate": 2.1442900536712717e-06,
+      "loss": 0.6256,
+      "step": 4699
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0382861135814472,
+      "learning_rate": 2.1409045961123067e-06,
+      "loss": 0.7466,
+      "step": 4700
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0869602790432573,
+      "learning_rate": 2.1375214927812836e-06,
+      "loss": 0.8484,
+      "step": 4701
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2494669307624964,
+      "learning_rate": 2.134140744691633e-06,
+      "loss": 0.8192,
+      "step": 4702
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0356259387254176,
+      "learning_rate": 2.130762352856067e-06,
+      "loss": 1.092,
+      "step": 4703
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.733422024246759,
+      "learning_rate": 2.1273863182866028e-06,
+      "loss": 0.5384,
+      "step": 4704
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1340406147050595,
+      "learning_rate": 2.1240126419945475e-06,
+      "loss": 0.6587,
+      "step": 4705
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3395358210848098,
+      "learning_rate": 2.1206413249904967e-06,
+      "loss": 0.8093,
+      "step": 4706
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2297010377657163,
+      "learning_rate": 2.117272368284348e-06,
+      "loss": 1.1448,
+      "step": 4707
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9637448508569557,
+      "learning_rate": 2.1139057728852887e-06,
+      "loss": 0.6782,
+      "step": 4708
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8046327805558845,
+      "learning_rate": 2.1105415398017916e-06,
+      "loss": 0.9494,
+      "step": 4709
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0430326133315373,
+      "learning_rate": 2.1071796700416334e-06,
+      "loss": 0.904,
+      "step": 4710
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7907031800143894,
+      "learning_rate": 2.1038201646118762e-06,
+      "loss": 0.9273,
+      "step": 4711
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.05515790678704,
+      "learning_rate": 2.1004630245188773e-06,
+      "loss": 1.0706,
+      "step": 4712
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0380958096937103,
+      "learning_rate": 2.0971082507682782e-06,
+      "loss": 0.7203,
+      "step": 4713
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5696046786684915,
+      "learning_rate": 2.093755844365021e-06,
+      "loss": 0.4628,
+      "step": 4714
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0821646985597115,
+      "learning_rate": 2.090405806313335e-06,
+      "loss": 0.7736,
+      "step": 4715
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8858460469298284,
+      "learning_rate": 2.0870581376167352e-06,
+      "loss": 0.8671,
+      "step": 4716
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9791546062468021,
+      "learning_rate": 2.0837128392780336e-06,
+      "loss": 1.0541,
+      "step": 4717
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9693285023429021,
+      "learning_rate": 2.0803699122993293e-06,
+      "loss": 0.6817,
+      "step": 4718
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0617795879335838,
+      "learning_rate": 2.0770293576820143e-06,
+      "loss": 0.7768,
+      "step": 4719
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.177176595382445,
+      "learning_rate": 2.073691176426761e-06,
+      "loss": 0.7233,
+      "step": 4720
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5894266266702705,
+      "learning_rate": 2.07035536953354e-06,
+      "loss": 0.959,
+      "step": 4721
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8525695167845235,
+      "learning_rate": 2.067021938001611e-06,
+      "loss": 0.8492,
+      "step": 4722
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0900542233504744,
+      "learning_rate": 2.063690882829512e-06,
+      "loss": 0.8811,
+      "step": 4723
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.079634551930526,
+      "learning_rate": 2.0603622050150795e-06,
+      "loss": 0.7726,
+      "step": 4724
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8490482964058563,
+      "learning_rate": 2.057035905555432e-06,
+      "loss": 0.7442,
+      "step": 4725
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1517469048437659,
+      "learning_rate": 2.0537119854469845e-06,
+      "loss": 1.1866,
+      "step": 4726
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9884475617737281,
+      "learning_rate": 2.0503904456854227e-06,
+      "loss": 0.7018,
+      "step": 4727
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0091905624666442,
+      "learning_rate": 2.047071287265735e-06,
+      "loss": 0.8534,
+      "step": 4728
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9315849344130575,
+      "learning_rate": 2.043754511182191e-06,
+      "loss": 0.9461,
+      "step": 4729
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0569602488246714,
+      "learning_rate": 2.0404401184283408e-06,
+      "loss": 0.8842,
+      "step": 4730
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1170949704531714,
+      "learning_rate": 2.0371281099970308e-06,
+      "loss": 1.0938,
+      "step": 4731
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9536304348486193,
+      "learning_rate": 2.033818486880389e-06,
+      "loss": 0.7229,
+      "step": 4732
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6336321965664172,
+      "learning_rate": 2.030511250069822e-06,
+      "loss": 0.4534,
+      "step": 4733
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9514058905117985,
+      "learning_rate": 2.0272064005560354e-06,
+      "loss": 0.6387,
+      "step": 4734
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.04418665024186,
+      "learning_rate": 2.023903939329006e-06,
+      "loss": 0.9951,
+      "step": 4735
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0024148756283542,
+      "learning_rate": 2.020603867378005e-06,
+      "loss": 1.1424,
+      "step": 4736
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9745743982494836,
+      "learning_rate": 2.017306185691581e-06,
+      "loss": 0.7284,
+      "step": 4737
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.193090445647266,
+      "learning_rate": 2.01401089525757e-06,
+      "loss": 0.9213,
+      "step": 4738
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.900299842743538,
+      "learning_rate": 2.0107179970630917e-06,
+      "loss": 0.7017,
+      "step": 4739
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1181459029639755,
+      "learning_rate": 2.0074274920945537e-06,
+      "loss": 1.1586,
+      "step": 4740
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3364454945862765,
+      "learning_rate": 2.004139381337634e-06,
+      "loss": 0.6443,
+      "step": 4741
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0627423058102363,
+      "learning_rate": 2.000853665777305e-06,
+      "loss": 0.7106,
+      "step": 4742
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0747311818389031,
+      "learning_rate": 1.997570346397819e-06,
+      "loss": 0.8373,
+      "step": 4743
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8950738400988583,
+      "learning_rate": 1.994289424182706e-06,
+      "loss": 0.7405,
+      "step": 4744
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3010646150373277,
+      "learning_rate": 1.9910109001147838e-06,
+      "loss": 1.1856,
+      "step": 4745
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6029179447743976,
+      "learning_rate": 1.987734775176151e-06,
+      "loss": 0.9659,
+      "step": 4746
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9834380929161575,
+      "learning_rate": 1.98446105034818e-06,
+      "loss": 0.6576,
+      "step": 4747
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0985005188055958,
+      "learning_rate": 1.981189726611533e-06,
+      "loss": 0.8923,
+      "step": 4748
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0046168761155783,
+      "learning_rate": 1.9779208049461517e-06,
+      "loss": 0.7722,
+      "step": 4749
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0348968442608315,
+      "learning_rate": 1.974654286331258e-06,
+      "loss": 0.928,
+      "step": 4750
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8511229892688142,
+      "learning_rate": 1.9713901717453453e-06,
+      "loss": 0.8103,
+      "step": 4751
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5270760658279853,
+      "learning_rate": 1.9681284621662e-06,
+      "loss": 0.604,
+      "step": 4752
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2660035275217523,
+      "learning_rate": 1.964869158570882e-06,
+      "loss": 0.6476,
+      "step": 4753
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.018728823249399,
+      "learning_rate": 1.961612261935727e-06,
+      "loss": 0.8071,
+      "step": 4754
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1128489917097848,
+      "learning_rate": 1.958357773236356e-06,
+      "loss": 0.9666,
+      "step": 4755
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0978209111752224,
+      "learning_rate": 1.9551056934476653e-06,
+      "loss": 0.7781,
+      "step": 4756
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1431891626354873,
+      "learning_rate": 1.951856023543832e-06,
+      "loss": 0.8438,
+      "step": 4757
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8931046235402534,
+      "learning_rate": 1.9486087644983056e-06,
+      "loss": 0.9832,
+      "step": 4758
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9205407982582818,
+      "learning_rate": 1.9453639172838203e-06,
+      "loss": 0.8375,
+      "step": 4759
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9878705775098598,
+      "learning_rate": 1.9421214828723857e-06,
+      "loss": 1.0038,
+      "step": 4760
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1562809850856557,
+      "learning_rate": 1.9388814622352846e-06,
+      "loss": 0.7205,
+      "step": 4761
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0058356763921317,
+      "learning_rate": 1.9356438563430822e-06,
+      "loss": 0.9303,
+      "step": 4762
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1006357120653842,
+      "learning_rate": 1.932408666165617e-06,
+      "loss": 0.8903,
+      "step": 4763
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2948589042215677,
+      "learning_rate": 1.9291758926720083e-06,
+      "loss": 1.1128,
+      "step": 4764
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1395819254392865,
+      "learning_rate": 1.9259455368306434e-06,
+      "loss": 0.7909,
+      "step": 4765
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3228834995305847,
+      "learning_rate": 1.9227175996091952e-06,
+      "loss": 0.7665,
+      "step": 4766
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3959294446238275,
+      "learning_rate": 1.919492081974602e-06,
+      "loss": 0.6564,
+      "step": 4767
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0555600563349845,
+      "learning_rate": 1.916268984893086e-06,
+      "loss": 0.7546,
+      "step": 4768
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8208451639217079,
+      "learning_rate": 1.913048309330138e-06,
+      "loss": 0.7565,
+      "step": 4769
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9324023033650198,
+      "learning_rate": 1.9098300562505266e-06,
+      "loss": 1.0324,
+      "step": 4770
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4807894080021668,
+      "learning_rate": 1.9066142266182974e-06,
+      "loss": 0.8015,
+      "step": 4771
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0093584047005324,
+      "learning_rate": 1.9034008213967625e-06,
+      "loss": 0.9842,
+      "step": 4772
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9640587302944572,
+      "learning_rate": 1.9001898415485133e-06,
+      "loss": 0.6987,
+      "step": 4773
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8789522036518197,
+      "learning_rate": 1.8969812880354167e-06,
+      "loss": 1.0358,
+      "step": 4774
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9935679048269725,
+      "learning_rate": 1.893775161818605e-06,
+      "loss": 0.7968,
+      "step": 4775
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1848841798393348,
+      "learning_rate": 1.8905714638584904e-06,
+      "loss": 0.7917,
+      "step": 4776
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9737418348478132,
+      "learning_rate": 1.8873701951147538e-06,
+      "loss": 0.7776,
+      "step": 4777
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7989064268432307,
+      "learning_rate": 1.8841713565463548e-06,
+      "loss": 0.5491,
+      "step": 4778
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8444421650137913,
+      "learning_rate": 1.8809749491115127e-06,
+      "loss": 0.9489,
+      "step": 4779
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1183471589679486,
+      "learning_rate": 1.8777809737677299e-06,
+      "loss": 0.7398,
+      "step": 4780
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7868942248890604,
+      "learning_rate": 1.8745894314717795e-06,
+      "loss": 0.9306,
+      "step": 4781
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9033016465862539,
+      "learning_rate": 1.871400323179695e-06,
+      "loss": 0.7642,
+      "step": 4782
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9522741520384486,
+      "learning_rate": 1.8682136498467928e-06,
+      "loss": 0.682,
+      "step": 4783
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0138209934562363,
+      "learning_rate": 1.8650294124276558e-06,
+      "loss": 0.9997,
+      "step": 4784
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9168538963901897,
+      "learning_rate": 1.8618476118761386e-06,
+      "loss": 0.8039,
+      "step": 4785
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1915621980367173,
+      "learning_rate": 1.858668249145359e-06,
+      "loss": 0.8837,
+      "step": 4786
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2123645655584212,
+      "learning_rate": 1.8554913251877116e-06,
+      "loss": 0.767,
+      "step": 4787
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5350130408858049,
+      "learning_rate": 1.852316840954862e-06,
+      "loss": 0.4719,
+      "step": 4788
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.967741132817317,
+      "learning_rate": 1.8491447973977372e-06,
+      "loss": 1.0085,
+      "step": 4789
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3599083300903998,
+      "learning_rate": 1.8459751954665372e-06,
+      "loss": 0.9121,
+      "step": 4790
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9703149521854528,
+      "learning_rate": 1.8428080361107336e-06,
+      "loss": 0.8231,
+      "step": 4791
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9348411077654997,
+      "learning_rate": 1.8396433202790643e-06,
+      "loss": 0.6489,
+      "step": 4792
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.002539179636736,
+      "learning_rate": 1.8364810489195294e-06,
+      "loss": 0.995,
+      "step": 4793
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9539202643415589,
+      "learning_rate": 1.8333212229794039e-06,
+      "loss": 0.9531,
+      "step": 4794
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.173939936607979,
+      "learning_rate": 1.8301638434052327e-06,
+      "loss": 0.8644,
+      "step": 4795
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9999420365557145,
+      "learning_rate": 1.8270089111428168e-06,
+      "loss": 0.7775,
+      "step": 4796
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4107866567815595,
+      "learning_rate": 1.8238564271372327e-06,
+      "loss": 0.9693,
+      "step": 4797
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8700094293887931,
+      "learning_rate": 1.820706392332824e-06,
+      "loss": 0.9421,
+      "step": 4798
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8374623555747669,
+      "learning_rate": 1.8175588076731942e-06,
+      "loss": 0.7987,
+      "step": 4799
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1165093517257398,
+      "learning_rate": 1.8144136741012209e-06,
+      "loss": 0.6919,
+      "step": 4800
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9982509057161606,
+      "learning_rate": 1.8112709925590377e-06,
+      "loss": 0.8473,
+      "step": 4801
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9801038302441268,
+      "learning_rate": 1.8081307639880519e-06,
+      "loss": 0.7378,
+      "step": 4802
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8917400972370987,
+      "learning_rate": 1.8049929893289363e-06,
+      "loss": 0.9114,
+      "step": 4803
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0757670025918968,
+      "learning_rate": 1.8018576695216217e-06,
+      "loss": 0.8494,
+      "step": 4804
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.526007722659993,
+      "learning_rate": 1.7987248055053076e-06,
+      "loss": 0.6552,
+      "step": 4805
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0991001158410394,
+      "learning_rate": 1.7955943982184608e-06,
+      "loss": 0.8102,
+      "step": 4806
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0265259319409057,
+      "learning_rate": 1.7924664485988053e-06,
+      "loss": 0.9958,
+      "step": 4807
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9674511992382667,
+      "learning_rate": 1.7893409575833353e-06,
+      "loss": 0.8062,
+      "step": 4808
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2626063957157754,
+      "learning_rate": 1.7862179261083069e-06,
+      "loss": 0.9427,
+      "step": 4809
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4784758460066145,
+      "learning_rate": 1.7830973551092346e-06,
+      "loss": 0.66,
+      "step": 4810
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0945224384923182,
+      "learning_rate": 1.7799792455209019e-06,
+      "loss": 0.8649,
+      "step": 4811
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8798316334469367,
+      "learning_rate": 1.776863598277353e-06,
+      "loss": 0.93,
+      "step": 4812
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.24127351603185,
+      "learning_rate": 1.7737504143118955e-06,
+      "loss": 0.995,
+      "step": 4813
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9496114743165631,
+      "learning_rate": 1.7706396945570947e-06,
+      "loss": 0.6532,
+      "step": 4814
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.044907205567191,
+      "learning_rate": 1.7675314399447819e-06,
+      "loss": 0.7777,
+      "step": 4815
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1273340013514683,
+      "learning_rate": 1.7644256514060532e-06,
+      "loss": 0.7854,
+      "step": 4816
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8887115963440386,
+      "learning_rate": 1.7613223298712555e-06,
+      "loss": 1.1682,
+      "step": 4817
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.009171022937133,
+      "learning_rate": 1.7582214762700057e-06,
+      "loss": 0.8478,
+      "step": 4818
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1791543620922464,
+      "learning_rate": 1.755123091531178e-06,
+      "loss": 0.735,
+      "step": 4819
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6530741746263832,
+      "learning_rate": 1.7520271765829112e-06,
+      "loss": 0.9901,
+      "step": 4820
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0615028536364592,
+      "learning_rate": 1.7489337323525957e-06,
+      "loss": 0.6562,
+      "step": 4821
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8775725211570092,
+      "learning_rate": 1.7458427597668893e-06,
+      "loss": 0.9881,
+      "step": 4822
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0106364179953093,
+      "learning_rate": 1.7427542597517077e-06,
+      "loss": 0.7522,
+      "step": 4823
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5128987164857185,
+      "learning_rate": 1.7396682332322235e-06,
+      "loss": 0.4247,
+      "step": 4824
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9745626531404435,
+      "learning_rate": 1.73658468113287e-06,
+      "loss": 0.7596,
+      "step": 4825
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0192059336801758,
+      "learning_rate": 1.7335036043773402e-06,
+      "loss": 0.8013,
+      "step": 4826
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8851586820726501,
+      "learning_rate": 1.7304250038885873e-06,
+      "loss": 0.7967,
+      "step": 4827
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.182954784033088,
+      "learning_rate": 1.727348880588815e-06,
+      "loss": 0.7014,
+      "step": 4828
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.288088590800908,
+      "learning_rate": 1.724275235399493e-06,
+      "loss": 0.6785,
+      "step": 4829
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8436212130287241,
+      "learning_rate": 1.7212040692413468e-06,
+      "loss": 1.0138,
+      "step": 4830
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0534868168495097,
+      "learning_rate": 1.718135383034355e-06,
+      "loss": 1.0427,
+      "step": 4831
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.872539394010174,
+      "learning_rate": 1.7150691776977612e-06,
+      "loss": 0.7907,
+      "step": 4832
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8474233399053991,
+      "learning_rate": 1.7120054541500552e-06,
+      "loss": 0.6661,
+      "step": 4833
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0108680250128304,
+      "learning_rate": 1.7089442133089928e-06,
+      "loss": 0.7975,
+      "step": 4834
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3090637980825017,
+      "learning_rate": 1.705885456091586e-06,
+      "loss": 0.8005,
+      "step": 4835
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0353557754955744,
+      "learning_rate": 1.702829183414092e-06,
+      "loss": 1.0425,
+      "step": 4836
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0952607230326559,
+      "learning_rate": 1.6997753961920382e-06,
+      "loss": 0.6999,
+      "step": 4837
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0050256941065903,
+      "learning_rate": 1.6967240953401954e-06,
+      "loss": 1.1328,
+      "step": 4838
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1232958267505357,
+      "learning_rate": 1.6936752817725965e-06,
+      "loss": 0.7969,
+      "step": 4839
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4960970720827793,
+      "learning_rate": 1.690628956402528e-06,
+      "loss": 0.4694,
+      "step": 4840
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.994418506065501,
+      "learning_rate": 1.6875851201425332e-06,
+      "loss": 1.021,
+      "step": 4841
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2546785512808307,
+      "learning_rate": 1.6845437739044012e-06,
+      "loss": 0.8173,
+      "step": 4842
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7888678370101734,
+      "learning_rate": 1.681504918599186e-06,
+      "loss": 0.8261,
+      "step": 4843
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2889932571211864,
+      "learning_rate": 1.6784685551371904e-06,
+      "loss": 0.6737,
+      "step": 4844
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6235384596415894,
+      "learning_rate": 1.6754346844279667e-06,
+      "loss": 0.9574,
+      "step": 4845
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0233037494989061,
+      "learning_rate": 1.672403307380328e-06,
+      "loss": 1.0241,
+      "step": 4846
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8913193105720404,
+      "learning_rate": 1.6693744249023391e-06,
+      "loss": 0.6572,
+      "step": 4847
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0255581032642702,
+      "learning_rate": 1.666348037901312e-06,
+      "loss": 0.7212,
+      "step": 4848
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2210716906343135,
+      "learning_rate": 1.6633241472838145e-06,
+      "loss": 0.8613,
+      "step": 4849
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0338812214381046,
+      "learning_rate": 1.6603027539556682e-06,
+      "loss": 1.0126,
+      "step": 4850
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2916121047679396,
+      "learning_rate": 1.6572838588219486e-06,
+      "loss": 1.0183,
+      "step": 4851
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9669525257818578,
+      "learning_rate": 1.6542674627869738e-06,
+      "loss": 0.5948,
+      "step": 4852
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7685881314737548,
+      "learning_rate": 1.6512535667543217e-06,
+      "loss": 0.9928,
+      "step": 4853
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9856958516176446,
+      "learning_rate": 1.6482421716268215e-06,
+      "loss": 0.6907,
+      "step": 4854
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8766111781189049,
+      "learning_rate": 1.645233278306544e-06,
+      "loss": 0.6958,
+      "step": 4855
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9857680767058163,
+      "learning_rate": 1.64222688769482e-06,
+      "loss": 1.0364,
+      "step": 4856
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0605273792304313,
+      "learning_rate": 1.6392230006922293e-06,
+      "loss": 0.7552,
+      "step": 4857
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8686489147077029,
+      "learning_rate": 1.6362216181986002e-06,
+      "loss": 0.7497,
+      "step": 4858
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0982690710194052,
+      "learning_rate": 1.6332227411130064e-06,
+      "loss": 0.6665,
+      "step": 4859
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9758304451848476,
+      "learning_rate": 1.6302263703337774e-06,
+      "loss": 1.0932,
+      "step": 4860
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5452217802048227,
+      "learning_rate": 1.6272325067584926e-06,
+      "loss": 0.5908,
+      "step": 4861
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9656217720407039,
+      "learning_rate": 1.6242411512839728e-06,
+      "loss": 0.7578,
+      "step": 4862
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.998969360885643,
+      "learning_rate": 1.6212523048062945e-06,
+      "loss": 0.9216,
+      "step": 4863
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0573177135735943,
+      "learning_rate": 1.6182659682207836e-06,
+      "loss": 0.666,
+      "step": 4864
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0024178073456167,
+      "learning_rate": 1.6152821424220033e-06,
+      "loss": 1.051,
+      "step": 4865
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0423089109134394,
+      "learning_rate": 1.6123008283037778e-06,
+      "loss": 0.9408,
+      "step": 4866
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1290571231572173,
+      "learning_rate": 1.6093220267591737e-06,
+      "loss": 0.8059,
+      "step": 4867
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0899568991779434,
+      "learning_rate": 1.6063457386805004e-06,
+      "loss": 0.7477,
+      "step": 4868
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1452876434862351,
+      "learning_rate": 1.6033719649593249e-06,
+      "loss": 0.7557,
+      "step": 4869
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9415533172535171,
+      "learning_rate": 1.6004007064864468e-06,
+      "loss": 0.9808,
+      "step": 4870
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5473425390583075,
+      "learning_rate": 1.5974319641519254e-06,
+      "loss": 0.9291,
+      "step": 4871
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8970223275954021,
+      "learning_rate": 1.5944657388450625e-06,
+      "loss": 0.7011,
+      "step": 4872
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.064011419806838,
+      "learning_rate": 1.5915020314543982e-06,
+      "loss": 0.9348,
+      "step": 4873
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1299437905944711,
+      "learning_rate": 1.5885408428677296e-06,
+      "loss": 0.6688,
+      "step": 4874
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8836651776215577,
+      "learning_rate": 1.5855821739720945e-06,
+      "loss": 1.0634,
+      "step": 4875
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9504273107618674,
+      "learning_rate": 1.5826260256537728e-06,
+      "loss": 0.834,
+      "step": 4876
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5207402578215765,
+      "learning_rate": 1.579672398798292e-06,
+      "loss": 0.6738,
+      "step": 4877
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9650905426839976,
+      "learning_rate": 1.5767212942904275e-06,
+      "loss": 0.6654,
+      "step": 4878
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2750460109261756,
+      "learning_rate": 1.5737727130141966e-06,
+      "loss": 0.9633,
+      "step": 4879
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.238352401440212,
+      "learning_rate": 1.5708266558528562e-06,
+      "loss": 0.8581,
+      "step": 4880
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6355606141067318,
+      "learning_rate": 1.5678831236889136e-06,
+      "loss": 0.8156,
+      "step": 4881
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0461602281875395,
+      "learning_rate": 1.564942117404119e-06,
+      "loss": 1.1256,
+      "step": 4882
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0443980283802752,
+      "learning_rate": 1.5620036378794601e-06,
+      "loss": 0.7711,
+      "step": 4883
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0968622424363166,
+      "learning_rate": 1.5590676859951748e-06,
+      "loss": 1.0408,
+      "step": 4884
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9346062750735067,
+      "learning_rate": 1.5561342626307397e-06,
+      "loss": 0.6906,
+      "step": 4885
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0112635564023191,
+      "learning_rate": 1.5532033686648785e-06,
+      "loss": 0.909,
+      "step": 4886
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9665519243526514,
+      "learning_rate": 1.5502750049755477e-06,
+      "loss": 0.7158,
+      "step": 4887
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7437196132991061,
+      "learning_rate": 1.547349172439957e-06,
+      "loss": 0.6997,
+      "step": 4888
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0610723149850518,
+      "learning_rate": 1.5444258719345528e-06,
+      "loss": 0.8626,
+      "step": 4889
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1119205238850445,
+      "learning_rate": 1.5415051043350194e-06,
+      "loss": 0.7957,
+      "step": 4890
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1702646411821471,
+      "learning_rate": 1.5385868705162888e-06,
+      "loss": 0.7469,
+      "step": 4891
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9725978506970859,
+      "learning_rate": 1.5356711713525308e-06,
+      "loss": 0.7089,
+      "step": 4892
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5587852882992623,
+      "learning_rate": 1.5327580077171589e-06,
+      "loss": 0.9657,
+      "step": 4893
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0175322874076114,
+      "learning_rate": 1.5298473804828195e-06,
+      "loss": 1.0396,
+      "step": 4894
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.524401787376933,
+      "learning_rate": 1.5269392905214076e-06,
+      "loss": 0.4389,
+      "step": 4895
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0915456485825885,
+      "learning_rate": 1.524033738704056e-06,
+      "loss": 0.7395,
+      "step": 4896
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.30677414212797,
+      "learning_rate": 1.521130725901132e-06,
+      "loss": 0.7481,
+      "step": 4897
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0581438539048886,
+      "learning_rate": 1.518230252982248e-06,
+      "loss": 1.0863,
+      "step": 4898
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9632739565952326,
+      "learning_rate": 1.5153323208162573e-06,
+      "loss": 0.7705,
+      "step": 4899
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7917247725912242,
+      "learning_rate": 1.512436930271244e-06,
+      "loss": 0.9674,
+      "step": 4900
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1691447093074467,
+      "learning_rate": 1.509544082214539e-06,
+      "loss": 0.7245,
+      "step": 4901
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.824928839238558,
+      "learning_rate": 1.5066537775127044e-06,
+      "loss": 0.8483,
+      "step": 4902
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.05348827901937,
+      "learning_rate": 1.503766017031547e-06,
+      "loss": 1.0562,
+      "step": 4903
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0124875848152073,
+      "learning_rate": 1.50088080163611e-06,
+      "loss": 1.0092,
+      "step": 4904
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.35565674595875,
+      "learning_rate": 1.497998132190669e-06,
+      "loss": 0.7416,
+      "step": 4905
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8173281570170582,
+      "learning_rate": 1.4951180095587426e-06,
+      "loss": 0.839,
+      "step": 4906
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9110835822502158,
+      "learning_rate": 1.4922404346030873e-06,
+      "loss": 0.7102,
+      "step": 4907
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9874808285526977,
+      "learning_rate": 1.4893654081856878e-06,
+      "loss": 0.8928,
+      "step": 4908
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9782971150969654,
+      "learning_rate": 1.4864929311677766e-06,
+      "loss": 0.7081,
+      "step": 4909
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1868254695413984,
+      "learning_rate": 1.4836230044098164e-06,
+      "loss": 0.7329,
+      "step": 4910
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0999135988005997,
+      "learning_rate": 1.4807556287715052e-06,
+      "loss": 0.8311,
+      "step": 4911
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1690642463233445,
+      "learning_rate": 1.477890805111779e-06,
+      "loss": 0.6989,
+      "step": 4912
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7754114498434432,
+      "learning_rate": 1.4750285342888092e-06,
+      "loss": 0.6779,
+      "step": 4913
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9254534008597549,
+      "learning_rate": 1.4721688171600047e-06,
+      "loss": 0.7469,
+      "step": 4914
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9696531757364342,
+      "learning_rate": 1.4693116545820018e-06,
+      "loss": 0.8874,
+      "step": 4915
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0019563057278351,
+      "learning_rate": 1.4664570474106788e-06,
+      "loss": 0.71,
+      "step": 4916
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8256977967728616,
+      "learning_rate": 1.463604996501149e-06,
+      "loss": 0.9834,
+      "step": 4917
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0008971955716575,
+      "learning_rate": 1.4607555027077524e-06,
+      "loss": 0.6026,
+      "step": 4918
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9318915263310676,
+      "learning_rate": 1.457908566884071e-06,
+      "loss": 0.7387,
+      "step": 4919
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6831631019564441,
+      "learning_rate": 1.4550641898829165e-06,
+      "loss": 0.9786,
+      "step": 4920
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8933494583944382,
+      "learning_rate": 1.452222372556338e-06,
+      "loss": 0.6531,
+      "step": 4921
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1006956129950343,
+      "learning_rate": 1.4493831157556094e-06,
+      "loss": 1.0622,
+      "step": 4922
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.001645685614091,
+      "learning_rate": 1.4465464203312463e-06,
+      "loss": 0.7387,
+      "step": 4923
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.840273315599224,
+      "learning_rate": 1.4437122871329956e-06,
+      "loss": 0.9429,
+      "step": 4924
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.017047254546771,
+      "learning_rate": 1.4408807170098305e-06,
+      "loss": 1.1528,
+      "step": 4925
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9460540644428276,
+      "learning_rate": 1.4380517108099623e-06,
+      "loss": 0.7985,
+      "step": 4926
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.062398296273672,
+      "learning_rate": 1.4352252693808333e-06,
+      "loss": 1.1355,
+      "step": 4927
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0243957905918035,
+      "learning_rate": 1.4324013935691205e-06,
+      "loss": 0.8147,
+      "step": 4928
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.955942525974937,
+      "learning_rate": 1.429580084220722e-06,
+      "loss": 0.6897,
+      "step": 4929
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8584870107705166,
+      "learning_rate": 1.426761342180777e-06,
+      "loss": 0.6286,
+      "step": 4930
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9539162092487249,
+      "learning_rate": 1.4239451682936555e-06,
+      "loss": 0.6356,
+      "step": 4931
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0761051101525825,
+      "learning_rate": 1.42113156340295e-06,
+      "loss": 1.1404,
+      "step": 4932
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6256597864887219,
+      "learning_rate": 1.4183205283514934e-06,
+      "loss": 0.5118,
+      "step": 4933
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2104705442191217,
+      "learning_rate": 1.4155120639813392e-06,
+      "loss": 0.8034,
+      "step": 4934
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7704973841815356,
+      "learning_rate": 1.4127061711337796e-06,
+      "loss": 0.5056,
+      "step": 4935
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.061953367058295,
+      "learning_rate": 1.409902850649333e-06,
+      "loss": 1.0428,
+      "step": 4936
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9822239242285835,
+      "learning_rate": 1.4071021033677423e-06,
+      "loss": 0.7407,
+      "step": 4937
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9435205189183711,
+      "learning_rate": 1.4043039301279904e-06,
+      "loss": 0.7642,
+      "step": 4938
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.878249330430177,
+      "learning_rate": 1.4015083317682776e-06,
+      "loss": 0.6707,
+      "step": 4939
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9521562760908032,
+      "learning_rate": 1.3987153091260398e-06,
+      "loss": 0.8816,
+      "step": 4940
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1336245571005943,
+      "learning_rate": 1.3959248630379396e-06,
+      "loss": 1.1194,
+      "step": 4941
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.042942588604474,
+      "learning_rate": 1.393136994339872e-06,
+      "loss": 0.7195,
+      "step": 4942
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9685597592252859,
+      "learning_rate": 1.390351703866949e-06,
+      "loss": 0.705,
+      "step": 4943
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5654561032948737,
+      "learning_rate": 1.3875689924535207e-06,
+      "loss": 0.9535,
+      "step": 4944
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8986583239118985,
+      "learning_rate": 1.3847888609331627e-06,
+      "loss": 0.5655,
+      "step": 4945
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8970118505352581,
+      "learning_rate": 1.3820113101386712e-06,
+      "loss": 0.9517,
+      "step": 4946
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9540448367030339,
+      "learning_rate": 1.3792363409020782e-06,
+      "loss": 1.0011,
+      "step": 4947
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8580434956201801,
+      "learning_rate": 1.3764639540546387e-06,
+      "loss": 1.0,
+      "step": 4948
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1592170979905185,
+      "learning_rate": 1.3736941504268297e-06,
+      "loss": 0.7703,
+      "step": 4949
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5020549311944369,
+      "learning_rate": 1.3709269308483619e-06,
+      "loss": 0.3969,
+      "step": 4950
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1151815045179965,
+      "learning_rate": 1.3681622961481678e-06,
+      "loss": 0.977,
+      "step": 4951
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0229566560496541,
+      "learning_rate": 1.3654002471544093e-06,
+      "loss": 0.765,
+      "step": 4952
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0841894771806802,
+      "learning_rate": 1.3626407846944656e-06,
+      "loss": 0.8448,
+      "step": 4953
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2536743628110247,
+      "learning_rate": 1.359883909594948e-06,
+      "loss": 0.7284,
+      "step": 4954
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9839382877288269,
+      "learning_rate": 1.3571296226816944e-06,
+      "loss": 0.9423,
+      "step": 4955
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0617468605484626,
+      "learning_rate": 1.3543779247797583e-06,
+      "loss": 0.8595,
+      "step": 4956
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.092827188473242,
+      "learning_rate": 1.3516288167134273e-06,
+      "loss": 0.7505,
+      "step": 4957
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.269989795245275,
+      "learning_rate": 1.348882299306209e-06,
+      "loss": 0.7536,
+      "step": 4958
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9055265430857572,
+      "learning_rate": 1.3461383733808352e-06,
+      "loss": 0.7305,
+      "step": 4959
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9498460805285986,
+      "learning_rate": 1.3433970397592599e-06,
+      "loss": 1.0382,
+      "step": 4960
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0314916089069954,
+      "learning_rate": 1.3406582992626628e-06,
+      "loss": 0.7373,
+      "step": 4961
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2238612274657013,
+      "learning_rate": 1.3379221527114484e-06,
+      "loss": 0.7264,
+      "step": 4962
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0054473476321673,
+      "learning_rate": 1.335188600925238e-06,
+      "loss": 0.6709,
+      "step": 4963
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1639896236490106,
+      "learning_rate": 1.332457644722882e-06,
+      "loss": 0.7354,
+      "step": 4964
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1856318692198384,
+      "learning_rate": 1.329729284922452e-06,
+      "loss": 1.1185,
+      "step": 4965
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.879222775349588,
+      "learning_rate": 1.327003522341237e-06,
+      "loss": 0.7469,
+      "step": 4966
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5305141575030609,
+      "learning_rate": 1.3242803577957531e-06,
+      "loss": 0.4893,
+      "step": 4967
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9157146538574096,
+      "learning_rate": 1.3215597921017386e-06,
+      "loss": 0.9575,
+      "step": 4968
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4944095779633439,
+      "learning_rate": 1.3188418260741498e-06,
+      "loss": 0.9057,
+      "step": 4969
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0639429262325664,
+      "learning_rate": 1.3161264605271662e-06,
+      "loss": 1.1593,
+      "step": 4970
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9347401103390461,
+      "learning_rate": 1.3134136962741872e-06,
+      "loss": 0.7449,
+      "step": 4971
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.161018344418423,
+      "learning_rate": 1.3107035341278328e-06,
+      "loss": 0.8361,
+      "step": 4972
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0953118980264611,
+      "learning_rate": 1.3079959748999494e-06,
+      "loss": 0.8064,
+      "step": 4973
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8310155484151386,
+      "learning_rate": 1.3052910194015922e-06,
+      "loss": 1.1002,
+      "step": 4974
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9570011664194972,
+      "learning_rate": 1.3025886684430467e-06,
+      "loss": 0.728,
+      "step": 4975
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9736524784240701,
+      "learning_rate": 1.299888922833815e-06,
+      "loss": 0.815,
+      "step": 4976
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8754095012451709,
+      "learning_rate": 1.2971917833826164e-06,
+      "loss": 0.7659,
+      "step": 4977
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1081505504590945,
+      "learning_rate": 1.2944972508973908e-06,
+      "loss": 0.9188,
+      "step": 4978
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0351735786848655,
+      "learning_rate": 1.2918053261853002e-06,
+      "loss": 1.0365,
+      "step": 4979
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.154920587800567,
+      "learning_rate": 1.2891160100527222e-06,
+      "loss": 0.7911,
+      "step": 4980
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0074328814007831,
+      "learning_rate": 1.2864293033052522e-06,
+      "loss": 0.8155,
+      "step": 4981
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9640613531813109,
+      "learning_rate": 1.2837452067477062e-06,
+      "loss": 0.7711,
+      "step": 4982
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2525178122017422,
+      "learning_rate": 1.2810637211841203e-06,
+      "loss": 0.6603,
+      "step": 4983
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0294943280267062,
+      "learning_rate": 1.2783848474177407e-06,
+      "loss": 1.0564,
+      "step": 4984
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1350326305356242,
+      "learning_rate": 1.27570858625104e-06,
+      "loss": 0.7107,
+      "step": 4985
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6127263731971488,
+      "learning_rate": 1.273034938485702e-06,
+      "loss": 0.4638,
+      "step": 4986
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0237557402392279,
+      "learning_rate": 1.2703639049226345e-06,
+      "loss": 0.7145,
+      "step": 4987
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1207722293078024,
+      "learning_rate": 1.2676954863619527e-06,
+      "loss": 0.779,
+      "step": 4988
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0542594975603377,
+      "learning_rate": 1.2650296836029964e-06,
+      "loss": 1.298,
+      "step": 4989
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8045777640139127,
+      "learning_rate": 1.2623664974443195e-06,
+      "loss": 0.7402,
+      "step": 4990
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8893279521282326,
+      "learning_rate": 1.2597059286836876e-06,
+      "loss": 0.9554,
+      "step": 4991
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9124119105678486,
+      "learning_rate": 1.25704797811809e-06,
+      "loss": 0.6221,
+      "step": 4992
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1763608011410926,
+      "learning_rate": 1.2543926465437261e-06,
+      "loss": 1.0614,
+      "step": 4993
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2144571697761701,
+      "learning_rate": 1.2517399347560165e-06,
+      "loss": 0.739,
+      "step": 4994
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5706781189974526,
+      "learning_rate": 1.249089843549588e-06,
+      "loss": 0.9099,
+      "step": 4995
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.042633859638499,
+      "learning_rate": 1.2464423737182896e-06,
+      "loss": 0.8383,
+      "step": 4996
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7661685120437268,
+      "learning_rate": 1.2437975260551848e-06,
+      "loss": 0.9636,
+      "step": 4997
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1884413146241153,
+      "learning_rate": 1.2411553013525457e-06,
+      "loss": 1.1613,
+      "step": 4998
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0056892449880654,
+      "learning_rate": 1.2385157004018655e-06,
+      "loss": 0.6893,
+      "step": 4999
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7636075566478026,
+      "learning_rate": 1.2358787239938497e-06,
+      "loss": 0.4964,
+      "step": 5000
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0438959554185279,
+      "learning_rate": 1.233244372918414e-06,
+      "loss": 0.9451,
+      "step": 5001
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8921335502470237,
+      "learning_rate": 1.230612647964694e-06,
+      "loss": 0.5867,
+      "step": 5002
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5578948036075009,
+      "learning_rate": 1.2279835499210302e-06,
+      "loss": 0.5011,
+      "step": 5003
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.225450629534764,
+      "learning_rate": 1.2253570795749848e-06,
+      "loss": 1.188,
+      "step": 5004
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1728743332175775,
+      "learning_rate": 1.222733237713325e-06,
+      "loss": 0.8119,
+      "step": 5005
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8707720373610393,
+      "learning_rate": 1.2201120251220378e-06,
+      "loss": 0.7346,
+      "step": 5006
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.053352190195971,
+      "learning_rate": 1.2174934425863182e-06,
+      "loss": 0.8158,
+      "step": 5007
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0363784972937493,
+      "learning_rate": 1.2148774908905782e-06,
+      "loss": 0.9745,
+      "step": 5008
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7856479435553485,
+      "learning_rate": 1.2122641708184323e-06,
+      "loss": 0.6392,
+      "step": 5009
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0005393633198825,
+      "learning_rate": 1.209653483152714e-06,
+      "loss": 0.7558,
+      "step": 5010
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9798225988272024,
+      "learning_rate": 1.2070454286754718e-06,
+      "loss": 0.6657,
+      "step": 5011
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.842229860278423,
+      "learning_rate": 1.204440008167953e-06,
+      "loss": 1.1962,
+      "step": 5012
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.057930484324133,
+      "learning_rate": 1.2018372224106268e-06,
+      "loss": 0.7078,
+      "step": 5013
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0353104940390947,
+      "learning_rate": 1.1992370721831693e-06,
+      "loss": 0.6711,
+      "step": 5014
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0952119003619454,
+      "learning_rate": 1.1966395582644696e-06,
+      "loss": 0.7202,
+      "step": 5015
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9976949354375959,
+      "learning_rate": 1.19404468143262e-06,
+      "loss": 0.9609,
+      "step": 5016
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5255806753198083,
+      "learning_rate": 1.1914524424649322e-06,
+      "loss": 0.9869,
+      "step": 5017
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9746006485546427,
+      "learning_rate": 1.1888628421379222e-06,
+      "loss": 0.9216,
+      "step": 5018
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0529581904270189,
+      "learning_rate": 1.1862758812273134e-06,
+      "loss": 0.6427,
+      "step": 5019
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7897069754721581,
+      "learning_rate": 1.1836915605080445e-06,
+      "loss": 0.9522,
+      "step": 5020
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7824057123539467,
+      "learning_rate": 1.1811098807542609e-06,
+      "loss": 0.7141,
+      "step": 5021
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5877074276422697,
+      "learning_rate": 1.1785308427393182e-06,
+      "loss": 0.6782,
+      "step": 5022
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0661886924344512,
+      "learning_rate": 1.1759544472357742e-06,
+      "loss": 0.7085,
+      "step": 5023
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1005113130965336,
+      "learning_rate": 1.1733806950154025e-06,
+      "loss": 0.7989,
+      "step": 5024
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9519291202057626,
+      "learning_rate": 1.1708095868491843e-06,
+      "loss": 0.7256,
+      "step": 5025
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1268754121875544,
+      "learning_rate": 1.1682411235073032e-06,
+      "loss": 0.8045,
+      "step": 5026
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.08598996834348,
+      "learning_rate": 1.1656753057591552e-06,
+      "loss": 1.0785,
+      "step": 5027
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1827834418370389,
+      "learning_rate": 1.1631121343733443e-06,
+      "loss": 0.6778,
+      "step": 5028
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0523399874251607,
+      "learning_rate": 1.1605516101176806e-06,
+      "loss": 0.9036,
+      "step": 5029
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.094272961801559,
+      "learning_rate": 1.1579937337591774e-06,
+      "loss": 0.7419,
+      "step": 5030
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0156330188571792,
+      "learning_rate": 1.1554385060640606e-06,
+      "loss": 0.7877,
+      "step": 5031
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.080000201437188,
+      "learning_rate": 1.1528859277977622e-06,
+      "loss": 1.0426,
+      "step": 5032
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9843597535961849,
+      "learning_rate": 1.1503359997249142e-06,
+      "loss": 0.737,
+      "step": 5033
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9096335331544546,
+      "learning_rate": 1.1477887226093642e-06,
+      "loss": 0.8271,
+      "step": 5034
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4796318019621817,
+      "learning_rate": 1.1452440972141554e-06,
+      "loss": 0.8559,
+      "step": 5035
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8680835053891445,
+      "learning_rate": 1.1427021243015469e-06,
+      "loss": 0.7704,
+      "step": 5036
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.05566851497767,
+      "learning_rate": 1.1401628046329937e-06,
+      "loss": 1.0077,
+      "step": 5037
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9982809897699706,
+      "learning_rate": 1.1376261389691633e-06,
+      "loss": 0.7174,
+      "step": 5038
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5423070137222045,
+      "learning_rate": 1.1350921280699278e-06,
+      "loss": 0.4116,
+      "step": 5039
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9381730789378669,
+      "learning_rate": 1.1325607726943567e-06,
+      "loss": 0.7722,
+      "step": 5040
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0075980188829583,
+      "learning_rate": 1.1300320736007308e-06,
+      "loss": 1.0066,
+      "step": 5041
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5034890480575026,
+      "learning_rate": 1.1275060315465346e-06,
+      "loss": 0.959,
+      "step": 5042
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8138486847219213,
+      "learning_rate": 1.1249826472884574e-06,
+      "loss": 1.0074,
+      "step": 5043
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.89150052374603,
+      "learning_rate": 1.122461921582385e-06,
+      "loss": 0.6714,
+      "step": 5044
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1825747477322992,
+      "learning_rate": 1.1199438551834163e-06,
+      "loss": 0.7908,
+      "step": 5045
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0418848583242355,
+      "learning_rate": 1.1174284488458509e-06,
+      "loss": 1.007,
+      "step": 5046
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.983470618084462,
+      "learning_rate": 1.114915703323186e-06,
+      "loss": 0.7897,
+      "step": 5047
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0554496597380034,
+      "learning_rate": 1.1124056193681276e-06,
+      "loss": 0.831,
+      "step": 5048
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2614417325826544,
+      "learning_rate": 1.1098981977325862e-06,
+      "loss": 0.8073,
+      "step": 5049
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9681893511392131,
+      "learning_rate": 1.1073934391676666e-06,
+      "loss": 0.8012,
+      "step": 5050
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0058357901899444,
+      "learning_rate": 1.104891344423682e-06,
+      "loss": 0.9913,
+      "step": 5051
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2464396542142362,
+      "learning_rate": 1.1023919142501472e-06,
+      "loss": 0.689,
+      "step": 5052
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5618142246144933,
+      "learning_rate": 1.0998951493957799e-06,
+      "loss": 0.7662,
+      "step": 5053
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9761823860145655,
+      "learning_rate": 1.097401050608493e-06,
+      "loss": 0.9015,
+      "step": 5054
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9429416837840582,
+      "learning_rate": 1.0949096186354069e-06,
+      "loss": 0.6872,
+      "step": 5055
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0497294234835357,
+      "learning_rate": 1.0924208542228444e-06,
+      "loss": 1.0422,
+      "step": 5056
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0165076819617567,
+      "learning_rate": 1.0899347581163222e-06,
+      "loss": 1.0895,
+      "step": 5057
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.012533521018969,
+      "learning_rate": 1.0874513310605628e-06,
+      "loss": 0.6938,
+      "step": 5058
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4594126902574711,
+      "learning_rate": 1.0849705737994886e-06,
+      "loss": 0.3543,
+      "step": 5059
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.824387084209597,
+      "learning_rate": 1.0824924870762243e-06,
+      "loss": 0.7996,
+      "step": 5060
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3465171562947027,
+      "learning_rate": 1.080017071633087e-06,
+      "loss": 0.9671,
+      "step": 5061
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.014314634218215,
+      "learning_rate": 1.0775443282116017e-06,
+      "loss": 0.8147,
+      "step": 5062
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0041231314258954,
+      "learning_rate": 1.075074257552492e-06,
+      "loss": 0.761,
+      "step": 5063
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8806536853055459,
+      "learning_rate": 1.072606860395674e-06,
+      "loss": 0.7879,
+      "step": 5064
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0652094991953498,
+      "learning_rate": 1.07014213748027e-06,
+      "loss": 1.0246,
+      "step": 5065
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1185010989304534,
+      "learning_rate": 1.0676800895446016e-06,
+      "loss": 0.8568,
+      "step": 5066
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5118788405556939,
+      "learning_rate": 1.0652207173261831e-06,
+      "loss": 0.9305,
+      "step": 5067
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7798565903997132,
+      "learning_rate": 1.062764021561733e-06,
+      "loss": 1.0209,
+      "step": 5068
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.890153976930475,
+      "learning_rate": 1.0603100029871626e-06,
+      "loss": 0.6898,
+      "step": 5069
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0079171068256718,
+      "learning_rate": 1.0578586623375863e-06,
+      "loss": 1.0318,
+      "step": 5070
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.117962946225891,
+      "learning_rate": 1.0554100003473156e-06,
+      "loss": 0.8085,
+      "step": 5071
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1459541982588686,
+      "learning_rate": 1.0529640177498568e-06,
+      "loss": 0.7354,
+      "step": 5072
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5738362554923255,
+      "learning_rate": 1.050520715277914e-06,
+      "loss": 0.8492,
+      "step": 5073
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0402698785874178,
+      "learning_rate": 1.048080093663395e-06,
+      "loss": 0.6857,
+      "step": 5074
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5332632422728683,
+      "learning_rate": 1.0456421536373918e-06,
+      "loss": 0.6245,
+      "step": 5075
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.003339048525953,
+      "learning_rate": 1.0432068959302043e-06,
+      "loss": 0.7942,
+      "step": 5076
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9523558266279264,
+      "learning_rate": 1.0407743212713273e-06,
+      "loss": 0.68,
+      "step": 5077
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1800148999897921,
+      "learning_rate": 1.0383444303894453e-06,
+      "loss": 0.8809,
+      "step": 5078
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7864316875348792,
+      "learning_rate": 1.035917224012445e-06,
+      "loss": 0.8563,
+      "step": 5079
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1199934161621246,
+      "learning_rate": 1.033492702867407e-06,
+      "loss": 1.1292,
+      "step": 5080
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2251765499766736,
+      "learning_rate": 1.0310708676806092e-06,
+      "loss": 0.8327,
+      "step": 5081
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.022997279351779,
+      "learning_rate": 1.0286517191775215e-06,
+      "loss": 0.718,
+      "step": 5082
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9421812048071453,
+      "learning_rate": 1.0262352580828106e-06,
+      "loss": 0.6359,
+      "step": 5083
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9266262907107421,
+      "learning_rate": 1.0238214851203409e-06,
+      "loss": 0.7236,
+      "step": 5084
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.761742380540019,
+      "learning_rate": 1.0214104010131665e-06,
+      "loss": 1.1276,
+      "step": 5085
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0878142273358982,
+      "learning_rate": 1.0190020064835404e-06,
+      "loss": 0.8758,
+      "step": 5086
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1738934518686965,
+      "learning_rate": 1.0165963022529067e-06,
+      "loss": 0.7485,
+      "step": 5087
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9957850410696765,
+      "learning_rate": 1.014193289041908e-06,
+      "loss": 0.6844,
+      "step": 5088
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8762319613890875,
+      "learning_rate": 1.0117929675703742e-06,
+      "loss": 0.9205,
+      "step": 5089
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4593169831983468,
+      "learning_rate": 1.009395338557333e-06,
+      "loss": 0.931,
+      "step": 5090
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8010520975132049,
+      "learning_rate": 1.0070004027210101e-06,
+      "loss": 0.9493,
+      "step": 5091
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9431726810297789,
+      "learning_rate": 1.0046081607788127e-06,
+      "loss": 0.8519,
+      "step": 5092
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5630759462819709,
+      "learning_rate": 1.0022186134473511e-06,
+      "loss": 0.9548,
+      "step": 5093
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6156413733953678,
+      "learning_rate": 9.998317614424246e-07,
+      "loss": 0.5971,
+      "step": 5094
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8745529116547142,
+      "learning_rate": 9.974476054790272e-07,
+      "loss": 0.7527,
+      "step": 5095
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0184922848120328,
+      "learning_rate": 9.950661462713417e-07,
+      "loss": 0.6627,
+      "step": 5096
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0325659279290005,
+      "learning_rate": 9.926873845327456e-07,
+      "loss": 0.8348,
+      "step": 5097
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4343180676678653,
+      "learning_rate": 9.903113209758098e-07,
+      "loss": 0.7816,
+      "step": 5098
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3183306592855109,
+      "learning_rate": 9.879379563122904e-07,
+      "loss": 1.1117,
+      "step": 5099
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8221201672681868,
+      "learning_rate": 9.855672912531455e-07,
+      "loss": 0.8245,
+      "step": 5100
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0630218258552124,
+      "learning_rate": 9.831993265085127e-07,
+      "loss": 0.7087,
+      "step": 5101
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8535379338398728,
+      "learning_rate": 9.80834062787731e-07,
+      "loss": 0.6797,
+      "step": 5102
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0360519915098783,
+      "learning_rate": 9.78471500799325e-07,
+      "loss": 1.0164,
+      "step": 5103
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2786129255882988,
+      "learning_rate": 9.761116412510096e-07,
+      "loss": 0.8275,
+      "step": 5104
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0470535005058217,
+      "learning_rate": 9.737544848496939e-07,
+      "loss": 0.8816,
+      "step": 5105
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0419797814948235,
+      "learning_rate": 9.714000323014705e-07,
+      "loss": 0.786,
+      "step": 5106
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9770144270217701,
+      "learning_rate": 9.69048284311631e-07,
+      "loss": 0.8539,
+      "step": 5107
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9436596354834532,
+      "learning_rate": 9.6669924158465e-07,
+      "loss": 0.9315,
+      "step": 5108
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1300289995714494,
+      "learning_rate": 9.643529048241951e-07,
+      "loss": 0.9335,
+      "step": 5109
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1029835313349223,
+      "learning_rate": 9.62009274733121e-07,
+      "loss": 0.6678,
+      "step": 5110
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2545137821016878,
+      "learning_rate": 9.59668352013473e-07,
+      "loss": 0.763,
+      "step": 5111
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5695671935323188,
+      "learning_rate": 9.573301373664877e-07,
+      "loss": 0.4439,
+      "step": 5112
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0824547918016227,
+      "learning_rate": 9.54994631492584e-07,
+      "loss": 1.018,
+      "step": 5113
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0469816670357557,
+      "learning_rate": 9.526618350913752e-07,
+      "loss": 0.7912,
+      "step": 5114
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5169614031190177,
+      "learning_rate": 9.503317488616614e-07,
+      "loss": 1.005,
+      "step": 5115
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.183703132848721,
+      "learning_rate": 9.480043735014321e-07,
+      "loss": 0.8062,
+      "step": 5116
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8909695674936179,
+      "learning_rate": 9.456797097078596e-07,
+      "loss": 0.5429,
+      "step": 5117
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1964461359415361,
+      "learning_rate": 9.43357758177309e-07,
+      "loss": 0.9992,
+      "step": 5118
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0884164417397633,
+      "learning_rate": 9.410385196053339e-07,
+      "loss": 0.7634,
+      "step": 5119
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9907027382278973,
+      "learning_rate": 9.387219946866699e-07,
+      "loss": 0.8607,
+      "step": 5120
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9656538073658006,
+      "learning_rate": 9.364081841152428e-07,
+      "loss": 0.992,
+      "step": 5121
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9081568423602283,
+      "learning_rate": 9.34097088584166e-07,
+      "loss": 0.9704,
+      "step": 5122
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2117449506165912,
+      "learning_rate": 9.317887087857414e-07,
+      "loss": 0.8125,
+      "step": 5123
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0219659196073914,
+      "learning_rate": 9.29483045411449e-07,
+      "loss": 0.7831,
+      "step": 5124
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9763375132754029,
+      "learning_rate": 9.271800991519642e-07,
+      "loss": 0.7498,
+      "step": 5125
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3192432952695126,
+      "learning_rate": 9.248798706971462e-07,
+      "loss": 0.8869,
+      "step": 5126
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4322697430586029,
+      "learning_rate": 9.225823607360351e-07,
+      "loss": 1.0796,
+      "step": 5127
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9945026387366878,
+      "learning_rate": 9.202875699568636e-07,
+      "loss": 0.8901,
+      "step": 5128
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8943128399580409,
+      "learning_rate": 9.179954990470463e-07,
+      "loss": 0.6565,
+      "step": 5129
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5544042814573705,
+      "learning_rate": 9.15706148693184e-07,
+      "loss": 0.4363,
+      "step": 5130
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0639522636529757,
+      "learning_rate": 9.134195195810602e-07,
+      "loss": 0.6785,
+      "step": 5131
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0319927827555744,
+      "learning_rate": 9.111356123956471e-07,
+      "loss": 1.1219,
+      "step": 5132
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9388331379742859,
+      "learning_rate": 9.088544278210976e-07,
+      "loss": 0.6539,
+      "step": 5133
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0975427799004651,
+      "learning_rate": 9.065759665407514e-07,
+      "loss": 0.7569,
+      "step": 5134
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1190812344022427,
+      "learning_rate": 9.04300229237135e-07,
+      "loss": 0.787,
+      "step": 5135
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8448247307186617,
+      "learning_rate": 9.020272165919508e-07,
+      "loss": 0.6691,
+      "step": 5136
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0392942051132716,
+      "learning_rate": 8.997569292860953e-07,
+      "loss": 1.106,
+      "step": 5137
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7098560098098903,
+      "learning_rate": 8.974893679996388e-07,
+      "loss": 0.8786,
+      "step": 5138
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9706904900154151,
+      "learning_rate": 8.952245334118415e-07,
+      "loss": 0.901,
+      "step": 5139
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1542535184057834,
+      "learning_rate": 8.929624262011472e-07,
+      "loss": 0.7311,
+      "step": 5140
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0646790542202202,
+      "learning_rate": 8.907030470451761e-07,
+      "loss": 1.079,
+      "step": 5141
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1317035482357958,
+      "learning_rate": 8.884463966207391e-07,
+      "loss": 0.7718,
+      "step": 5142
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5120125225568358,
+      "learning_rate": 8.861924756038243e-07,
+      "loss": 0.9432,
+      "step": 5143
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0317216213912466,
+      "learning_rate": 8.839412846696061e-07,
+      "loss": 0.8463,
+      "step": 5144
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0891790045313194,
+      "learning_rate": 8.816928244924361e-07,
+      "loss": 0.8453,
+      "step": 5145
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0639315905731388,
+      "learning_rate": 8.794470957458523e-07,
+      "loss": 1.0096,
+      "step": 5146
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.538972088129364,
+      "learning_rate": 8.772040991025755e-07,
+      "loss": 0.4549,
+      "step": 5147
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.005798832025373,
+      "learning_rate": 8.749638352345002e-07,
+      "loss": 0.7663,
+      "step": 5148
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0824244710809767,
+      "learning_rate": 8.727263048127122e-07,
+      "loss": 0.9685,
+      "step": 5149
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0254526916755908,
+      "learning_rate": 8.704915085074728e-07,
+      "loss": 0.7051,
+      "step": 5150
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9534580141044712,
+      "learning_rate": 8.682594469882233e-07,
+      "loss": 0.9347,
+      "step": 5151
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8375519995533405,
+      "learning_rate": 8.6603012092359e-07,
+      "loss": 0.4973,
+      "step": 5152
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0515336001743139,
+      "learning_rate": 8.638035309813774e-07,
+      "loss": 0.7801,
+      "step": 5153
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8497031408231939,
+      "learning_rate": 8.615796778285735e-07,
+      "loss": 0.7849,
+      "step": 5154
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0691674076892912,
+      "learning_rate": 8.593585621313371e-07,
+      "loss": 0.8058,
+      "step": 5155
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9839363092181049,
+      "learning_rate": 8.571401845550186e-07,
+      "loss": 1.0125,
+      "step": 5156
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1720309737947348,
+      "learning_rate": 8.549245457641431e-07,
+      "loss": 0.894,
+      "step": 5157
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4312115775514809,
+      "learning_rate": 8.527116464224128e-07,
+      "loss": 0.7611,
+      "step": 5158
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.98821486904031,
+      "learning_rate": 8.505014871927131e-07,
+      "loss": 0.7914,
+      "step": 5159
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1930970764656301,
+      "learning_rate": 8.482940687371067e-07,
+      "loss": 0.7781,
+      "step": 5160
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9378665528133635,
+      "learning_rate": 8.460893917168389e-07,
+      "loss": 0.9461,
+      "step": 5161
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.131599130701184,
+      "learning_rate": 8.438874567923261e-07,
+      "loss": 0.7756,
+      "step": 5162
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7601059907904144,
+      "learning_rate": 8.416882646231695e-07,
+      "loss": 1.1275,
+      "step": 5163
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1750216023131117,
+      "learning_rate": 8.394918158681509e-07,
+      "loss": 0.8887,
+      "step": 5164
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0895983411816155,
+      "learning_rate": 8.372981111852219e-07,
+      "loss": 0.7418,
+      "step": 5165
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6038180935282037,
+      "learning_rate": 8.351071512315167e-07,
+      "loss": 0.6221,
+      "step": 5166
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0481390069779084,
+      "learning_rate": 8.329189366633528e-07,
+      "loss": 0.8061,
+      "step": 5167
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9004945083423285,
+      "learning_rate": 8.307334681362134e-07,
+      "loss": 0.7608,
+      "step": 5168
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0470264074335653,
+      "learning_rate": 8.285507463047703e-07,
+      "loss": 0.7134,
+      "step": 5169
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5194390173402985,
+      "learning_rate": 8.263707718228641e-07,
+      "loss": 1.0351,
+      "step": 5170
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7299945460099173,
+      "learning_rate": 8.24193545343519e-07,
+      "loss": 0.7459,
+      "step": 5171
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.137794682620975,
+      "learning_rate": 8.22019067518931e-07,
+      "loss": 0.7219,
+      "step": 5172
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0497429889485734,
+      "learning_rate": 8.198473390004758e-07,
+      "loss": 0.7893,
+      "step": 5173
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9693823899223237,
+      "learning_rate": 8.17678360438704e-07,
+      "loss": 0.7912,
+      "step": 5174
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1448332538603267,
+      "learning_rate": 8.155121324833459e-07,
+      "loss": 1.1393,
+      "step": 5175
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4793540547554738,
+      "learning_rate": 8.133486557833015e-07,
+      "loss": 0.6594,
+      "step": 5176
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7852413829384637,
+      "learning_rate": 8.111879309866511e-07,
+      "loss": 0.5348,
+      "step": 5177
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0846910045946003,
+      "learning_rate": 8.090299587406514e-07,
+      "loss": 0.6793,
+      "step": 5178
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9766893579827935,
+      "learning_rate": 8.068747396917298e-07,
+      "loss": 0.6545,
+      "step": 5179
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.005380790769548,
+      "learning_rate": 8.047222744854943e-07,
+      "loss": 1.0492,
+      "step": 5180
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0304980486425561,
+      "learning_rate": 8.025725637667247e-07,
+      "loss": 0.7242,
+      "step": 5181
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.056761220623689,
+      "learning_rate": 8.004256081793782e-07,
+      "loss": 0.7598,
+      "step": 5182
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0308773486095844,
+      "learning_rate": 7.982814083665825e-07,
+      "loss": 0.8021,
+      "step": 5183
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8077247310825373,
+      "learning_rate": 7.961399649706447e-07,
+      "loss": 0.8311,
+      "step": 5184
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6143958624986517,
+      "learning_rate": 7.940012786330442e-07,
+      "loss": 0.6447,
+      "step": 5185
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.130082521279043,
+      "learning_rate": 7.918653499944318e-07,
+      "loss": 0.742,
+      "step": 5186
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1389466266402835,
+      "learning_rate": 7.897321796946367e-07,
+      "loss": 0.7842,
+      "step": 5187
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9189461670176154,
+      "learning_rate": 7.876017683726578e-07,
+      "loss": 0.9304,
+      "step": 5188
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0567821090349663,
+      "learning_rate": 7.854741166666746e-07,
+      "loss": 1.0409,
+      "step": 5189
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8616729842181733,
+      "learning_rate": 7.833492252140284e-07,
+      "loss": 1.0208,
+      "step": 5190
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2568959584518902,
+      "learning_rate": 7.812270946512435e-07,
+      "loss": 0.7002,
+      "step": 5191
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5703750192542295,
+      "learning_rate": 7.791077256140144e-07,
+      "loss": 0.9852,
+      "step": 5192
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9496959522431005,
+      "learning_rate": 7.769911187372059e-07,
+      "loss": 0.7141,
+      "step": 5193
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9862838695809741,
+      "learning_rate": 7.748772746548572e-07,
+      "loss": 0.9133,
+      "step": 5194
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9588227135036678,
+      "learning_rate": 7.727661940001808e-07,
+      "loss": 0.6263,
+      "step": 5195
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9034964528080618,
+      "learning_rate": 7.706578774055628e-07,
+      "loss": 0.8851,
+      "step": 5196
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1579307289170837,
+      "learning_rate": 7.685523255025562e-07,
+      "loss": 0.6664,
+      "step": 5197
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.941109574233338,
+      "learning_rate": 7.664495389218884e-07,
+      "loss": 0.7424,
+      "step": 5198
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0911425892102413,
+      "learning_rate": 7.643495182934612e-07,
+      "loss": 1.0839,
+      "step": 5199
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1179617048515251,
+      "learning_rate": 7.622522642463425e-07,
+      "loss": 0.7958,
+      "step": 5200
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6793549715950863,
+      "learning_rate": 7.601577774087777e-07,
+      "loss": 0.4758,
+      "step": 5201
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9142486809398903,
+      "learning_rate": 7.580660584081755e-07,
+      "loss": 0.6604,
+      "step": 5202
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0204849776427292,
+      "learning_rate": 7.559771078711242e-07,
+      "loss": 0.8272,
+      "step": 5203
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1675667278862645,
+      "learning_rate": 7.538909264233751e-07,
+      "loss": 1.1075,
+      "step": 5204
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5531353057833024,
+      "learning_rate": 7.518075146898552e-07,
+      "loss": 0.6998,
+      "step": 5205
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0044081080670404,
+      "learning_rate": 7.497268732946605e-07,
+      "loss": 0.7838,
+      "step": 5206
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0947069163917402,
+      "learning_rate": 7.47649002861055e-07,
+      "loss": 0.8256,
+      "step": 5207
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0382433149927,
+      "learning_rate": 7.455739040114751e-07,
+      "loss": 1.1335,
+      "step": 5208
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0694694768295792,
+      "learning_rate": 7.435015773675259e-07,
+      "loss": 1.1186,
+      "step": 5209
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0957734928197098,
+      "learning_rate": 7.414320235499839e-07,
+      "loss": 0.7193,
+      "step": 5210
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8287960265189855,
+      "learning_rate": 7.393652431787901e-07,
+      "loss": 0.9876,
+      "step": 5211
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8922133222278794,
+      "learning_rate": 7.373012368730615e-07,
+      "loss": 0.7325,
+      "step": 5212
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.947335650862757,
+      "learning_rate": 7.3524000525108e-07,
+      "loss": 1.077,
+      "step": 5213
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8875190655018897,
+      "learning_rate": 7.331815489302941e-07,
+      "loss": 0.6545,
+      "step": 5214
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2388555504113241,
+      "learning_rate": 7.311258685273259e-07,
+      "loss": 0.9376,
+      "step": 5215
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.493945798485656,
+      "learning_rate": 7.29072964657963e-07,
+      "loss": 0.9436,
+      "step": 5216
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.857991056036233,
+      "learning_rate": 7.270228379371658e-07,
+      "loss": 0.6339,
+      "step": 5217
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1770320447592222,
+      "learning_rate": 7.249754889790539e-07,
+      "loss": 1.0647,
+      "step": 5218
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4397451438605249,
+      "learning_rate": 7.229309183969224e-07,
+      "loss": 0.8143,
+      "step": 5219
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5333931465167915,
+      "learning_rate": 7.208891268032336e-07,
+      "loss": 0.4673,
+      "step": 5220
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8898171248533874,
+      "learning_rate": 7.188501148096117e-07,
+      "loss": 0.7998,
+      "step": 5221
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4794770947211267,
+      "learning_rate": 7.168138830268534e-07,
+      "loss": 0.6538,
+      "step": 5222
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9646166038802632,
+      "learning_rate": 7.147804320649221e-07,
+      "loss": 1.1036,
+      "step": 5223
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8767433111393924,
+      "learning_rate": 7.127497625329494e-07,
+      "loss": 0.6938,
+      "step": 5224
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7846922777380189,
+      "learning_rate": 7.107218750392264e-07,
+      "loss": 0.5119,
+      "step": 5225
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0807094966665962,
+      "learning_rate": 7.086967701912195e-07,
+      "loss": 0.7963,
+      "step": 5226
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2322695093892297,
+      "learning_rate": 7.06674448595559e-07,
+      "loss": 1.1635,
+      "step": 5227
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.868996077302754,
+      "learning_rate": 7.04654910858038e-07,
+      "loss": 0.7218,
+      "step": 5228
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9457676683893081,
+      "learning_rate": 7.026381575836194e-07,
+      "loss": 0.7124,
+      "step": 5229
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2062149007979135,
+      "learning_rate": 7.006241893764298e-07,
+      "loss": 1.087,
+      "step": 5230
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1151590854725344,
+      "learning_rate": 6.986130068397656e-07,
+      "loss": 0.6799,
+      "step": 5231
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0520616310918685,
+      "learning_rate": 6.966046105760826e-07,
+      "loss": 1.0933,
+      "step": 5232
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0994654293096253,
+      "learning_rate": 6.945990011870074e-07,
+      "loss": 0.6981,
+      "step": 5233
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.316282203606225,
+      "learning_rate": 6.925961792733271e-07,
+      "loss": 0.9865,
+      "step": 5234
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9717371492980871,
+      "learning_rate": 6.905961454349985e-07,
+      "loss": 0.7796,
+      "step": 5235
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8033532676122047,
+      "learning_rate": 6.88598900271139e-07,
+      "loss": 0.8792,
+      "step": 5236
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0752608602967955,
+      "learning_rate": 6.866044443800324e-07,
+      "loss": 0.9928,
+      "step": 5237
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5628449345130807,
+      "learning_rate": 6.846127783591294e-07,
+      "loss": 0.3974,
+      "step": 5238
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5209045838068032,
+      "learning_rate": 6.826239028050396e-07,
+      "loss": 0.9815,
+      "step": 5239
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9069605308450069,
+      "learning_rate": 6.80637818313541e-07,
+      "loss": 0.6698,
+      "step": 5240
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0129793271715768,
+      "learning_rate": 6.786545254795773e-07,
+      "loss": 0.6312,
+      "step": 5241
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.958763355897576,
+      "learning_rate": 6.76674024897247e-07,
+      "loss": 0.9808,
+      "step": 5242
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1310219400224655,
+      "learning_rate": 6.746963171598231e-07,
+      "loss": 0.8346,
+      "step": 5243
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9918117751206746,
+      "learning_rate": 6.72721402859734e-07,
+      "loss": 0.7075,
+      "step": 5244
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1719487899451315,
+      "learning_rate": 6.707492825885775e-07,
+      "loss": 0.8624,
+      "step": 5245
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9784596190437411,
+      "learning_rate": 6.687799569371079e-07,
+      "loss": 0.8741,
+      "step": 5246
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9653838567568759,
+      "learning_rate": 6.668134264952464e-07,
+      "loss": 0.9934,
+      "step": 5247
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.07311558026786,
+      "learning_rate": 6.648496918520786e-07,
+      "loss": 0.745,
+      "step": 5248
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.148461939429099,
+      "learning_rate": 6.628887535958472e-07,
+      "loss": 0.6492,
+      "step": 5249
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.29435136331733,
+      "learning_rate": 6.609306123139614e-07,
+      "loss": 0.768,
+      "step": 5250
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9643232984497518,
+      "learning_rate": 6.589752685929918e-07,
+      "loss": 0.8505,
+      "step": 5251
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8975594387231787,
+      "learning_rate": 6.57022723018671e-07,
+      "loss": 0.9827,
+      "step": 5252
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.98532813152237,
+      "learning_rate": 6.550729761758901e-07,
+      "loss": 0.6773,
+      "step": 5253
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0767466668048993,
+      "learning_rate": 6.531260286487073e-07,
+      "loss": 0.7497,
+      "step": 5254
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5702472074292109,
+      "learning_rate": 6.511818810203408e-07,
+      "loss": 0.5509,
+      "step": 5255
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0075409795462535,
+      "learning_rate": 6.492405338731655e-07,
+      "loss": 1.0214,
+      "step": 5256
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8404785394153563,
+      "learning_rate": 6.473019877887232e-07,
+      "loss": 0.6889,
+      "step": 5257
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7089123751913058,
+      "learning_rate": 6.453662433477137e-07,
+      "loss": 0.8122,
+      "step": 5258
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1270711311815576,
+      "learning_rate": 6.434333011299964e-07,
+      "loss": 0.8624,
+      "step": 5259
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8160016679198479,
+      "learning_rate": 6.415031617145951e-07,
+      "loss": 1.0573,
+      "step": 5260
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0027828141706747,
+      "learning_rate": 6.395758256796914e-07,
+      "loss": 1.0347,
+      "step": 5261
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9713272972441769,
+      "learning_rate": 6.37651293602628e-07,
+      "loss": 0.8391,
+      "step": 5262
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3837451442313455,
+      "learning_rate": 6.357295660599061e-07,
+      "loss": 0.7135,
+      "step": 5263
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0755938127737241,
+      "learning_rate": 6.338106436271884e-07,
+      "loss": 0.7206,
+      "step": 5264
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2298836767581804,
+      "learning_rate": 6.318945268792986e-07,
+      "loss": 0.8336,
+      "step": 5265
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.577612172251059,
+      "learning_rate": 6.299812163902152e-07,
+      "loss": 1.0407,
+      "step": 5266
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1019758485019617,
+      "learning_rate": 6.280707127330821e-07,
+      "loss": 0.734,
+      "step": 5267
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9547149912940684,
+      "learning_rate": 6.261630164801957e-07,
+      "loss": 0.5313,
+      "step": 5268
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8530840590759606,
+      "learning_rate": 6.24258128203018e-07,
+      "loss": 0.7088,
+      "step": 5269
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1050763180478749,
+      "learning_rate": 6.22356048472168e-07,
+      "loss": 0.7348,
+      "step": 5270
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2448291802296525,
+      "learning_rate": 6.204567778574189e-07,
+      "loss": 0.995,
+      "step": 5271
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8311092121879459,
+      "learning_rate": 6.185603169277087e-07,
+      "loss": 0.9027,
+      "step": 5272
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.272421171331038,
+      "learning_rate": 6.166666662511289e-07,
+      "loss": 0.7789,
+      "step": 5273
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9091022964247135,
+      "learning_rate": 6.147758263949322e-07,
+      "loss": 1.0043,
+      "step": 5274
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6402400802775525,
+      "learning_rate": 6.128877979255287e-07,
+      "loss": 0.5681,
+      "step": 5275
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7686925328862706,
+      "learning_rate": 6.110025814084875e-07,
+      "loss": 0.4819,
+      "step": 5276
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8793436396649028,
+      "learning_rate": 6.091201774085298e-07,
+      "loss": 0.7929,
+      "step": 5277
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9814341904865517,
+      "learning_rate": 6.072405864895403e-07,
+      "loss": 0.6841,
+      "step": 5278
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2280467585464896,
+      "learning_rate": 6.053638092145608e-07,
+      "loss": 0.7316,
+      "step": 5279
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9853971561930152,
+      "learning_rate": 6.034898461457861e-07,
+      "loss": 0.9869,
+      "step": 5280
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0187303859323222,
+      "learning_rate": 6.016186978445704e-07,
+      "loss": 0.7744,
+      "step": 5281
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0632499850272994,
+      "learning_rate": 5.997503648714254e-07,
+      "loss": 0.7065,
+      "step": 5282
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1373233499594848,
+      "learning_rate": 5.978848477860211e-07,
+      "loss": 0.7463,
+      "step": 5283
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8675753688931221,
+      "learning_rate": 5.960221471471783e-07,
+      "loss": 1.0772,
+      "step": 5284
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9027816486023411,
+      "learning_rate": 5.941622635128775e-07,
+      "loss": 0.8866,
+      "step": 5285
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1881080870922098,
+      "learning_rate": 5.923051974402594e-07,
+      "loss": 0.8974,
+      "step": 5286
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8855041941153914,
+      "learning_rate": 5.904509494856115e-07,
+      "loss": 0.8209,
+      "step": 5287
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0960756316539693,
+      "learning_rate": 5.885995202043848e-07,
+      "loss": 0.6395,
+      "step": 5288
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9679389515993968,
+      "learning_rate": 5.867509101511826e-07,
+      "loss": 1.024,
+      "step": 5289
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0239363451187555,
+      "learning_rate": 5.849051198797662e-07,
+      "loss": 0.7899,
+      "step": 5290
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5136061226068036,
+      "learning_rate": 5.830621499430489e-07,
+      "loss": 0.9203,
+      "step": 5291
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5113597946595743,
+      "learning_rate": 5.812220008931013e-07,
+      "loss": 0.4098,
+      "step": 5292
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1554469736608601,
+      "learning_rate": 5.793846732811492e-07,
+      "loss": 0.664,
+      "step": 5293
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.950380948635457,
+      "learning_rate": 5.775501676575702e-07,
+      "loss": 1.2288,
+      "step": 5294
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1336018089386275,
+      "learning_rate": 5.757184845718999e-07,
+      "loss": 0.7067,
+      "step": 5295
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.185179530172321,
+      "learning_rate": 5.738896245728287e-07,
+      "loss": 0.8471,
+      "step": 5296
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0366520070794476,
+      "learning_rate": 5.720635882081993e-07,
+      "loss": 0.7006,
+      "step": 5297
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8764063881774664,
+      "learning_rate": 5.702403760250086e-07,
+      "loss": 0.9242,
+      "step": 5298
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8903131579365032,
+      "learning_rate": 5.684199885694086e-07,
+      "loss": 0.9725,
+      "step": 5299
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2907733792471214,
+      "learning_rate": 5.666024263867042e-07,
+      "loss": 0.5843,
+      "step": 5300
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2376772874070345,
+      "learning_rate": 5.647876900213544e-07,
+      "loss": 0.6159,
+      "step": 5301
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.848851588459245,
+      "learning_rate": 5.629757800169732e-07,
+      "loss": 0.7059,
+      "step": 5302
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2004257022775477,
+      "learning_rate": 5.611666969163243e-07,
+      "loss": 1.1501,
+      "step": 5303
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9568162690703969,
+      "learning_rate": 5.593604412613296e-07,
+      "loss": 0.7685,
+      "step": 5304
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.431438423058819,
+      "learning_rate": 5.575570135930586e-07,
+      "loss": 0.9048,
+      "step": 5305
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8892238851791937,
+      "learning_rate": 5.557564144517369e-07,
+      "loss": 0.7942,
+      "step": 5306
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8409743348707268,
+      "learning_rate": 5.539586443767442e-07,
+      "loss": 0.9776,
+      "step": 5307
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2269667537905857,
+      "learning_rate": 5.521637039066074e-07,
+      "loss": 0.669,
+      "step": 5308
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.144912191577247,
+      "learning_rate": 5.503715935790122e-07,
+      "loss": 1.0052,
+      "step": 5309
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.105373824511886,
+      "learning_rate": 5.485823139307911e-07,
+      "loss": 0.9931,
+      "step": 5310
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5519114406358557,
+      "learning_rate": 5.467958654979333e-07,
+      "loss": 0.4746,
+      "step": 5311
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.130959724166502,
+      "learning_rate": 5.450122488155751e-07,
+      "loss": 0.7227,
+      "step": 5312
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9742529380012974,
+      "learning_rate": 5.432314644180082e-07,
+      "loss": 0.7472,
+      "step": 5313
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0003858060976267,
+      "learning_rate": 5.414535128386766e-07,
+      "loss": 0.9816,
+      "step": 5314
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.49384232538367273,
+      "learning_rate": 5.396783946101702e-07,
+      "loss": 0.9319,
+      "step": 5315
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.944207311646265,
+      "learning_rate": 5.379061102642357e-07,
+      "loss": 0.9655,
+      "step": 5316
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9434321193585,
+      "learning_rate": 5.361366603317686e-07,
+      "loss": 0.7608,
+      "step": 5317
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9854657238005284,
+      "learning_rate": 5.343700453428169e-07,
+      "loss": 1.0511,
+      "step": 5318
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2642592124254493,
+      "learning_rate": 5.326062658265774e-07,
+      "loss": 0.9194,
+      "step": 5319
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1237272933678688,
+      "learning_rate": 5.308453223113962e-07,
+      "loss": 0.7198,
+      "step": 5320
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9094062424483139,
+      "learning_rate": 5.290872153247773e-07,
+      "loss": 0.7196,
+      "step": 5321
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1284709302005913,
+      "learning_rate": 5.273319453933634e-07,
+      "loss": 0.848,
+      "step": 5322
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.020150110324991,
+      "learning_rate": 5.255795130429575e-07,
+      "loss": 0.9432,
+      "step": 5323
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1999687646757737,
+      "learning_rate": 5.238299187985075e-07,
+      "loss": 1.0033,
+      "step": 5324
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.157870973519548,
+      "learning_rate": 5.220831631841139e-07,
+      "loss": 0.7692,
+      "step": 5325
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0854484842729508,
+      "learning_rate": 5.203392467230229e-07,
+      "loss": 0.5887,
+      "step": 5326
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.20613477294118,
+      "learning_rate": 5.18598169937633e-07,
+      "loss": 0.7914,
+      "step": 5327
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9694278865070599,
+      "learning_rate": 5.16859933349495e-07,
+      "loss": 0.9415,
+      "step": 5328
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5829715557233365,
+      "learning_rate": 5.151245374793012e-07,
+      "loss": 0.4681,
+      "step": 5329
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9186407272396448,
+      "learning_rate": 5.133919828468992e-07,
+      "loss": 0.7991,
+      "step": 5330
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1432232835277514,
+      "learning_rate": 5.116622699712848e-07,
+      "loss": 0.8054,
+      "step": 5331
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7043221036124512,
+      "learning_rate": 5.099353993706002e-07,
+      "loss": 0.9053,
+      "step": 5332
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7760603161579861,
+      "learning_rate": 5.082113715621362e-07,
+      "loss": 0.7372,
+      "step": 5333
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2214134263886893,
+      "learning_rate": 5.064901870623362e-07,
+      "loss": 0.8607,
+      "step": 5334
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9827348685791861,
+      "learning_rate": 5.047718463867868e-07,
+      "loss": 0.8345,
+      "step": 5335
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0424809355492335,
+      "learning_rate": 5.03056350050225e-07,
+      "loss": 0.7856,
+      "step": 5336
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9826873525630683,
+      "learning_rate": 5.013436985665354e-07,
+      "loss": 1.0306,
+      "step": 5337
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1478025815052095,
+      "learning_rate": 4.996338924487509e-07,
+      "loss": 0.7836,
+      "step": 5338
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5757643948826204,
+      "learning_rate": 4.979269322090529e-07,
+      "loss": 1.0214,
+      "step": 5339
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8092575860796165,
+      "learning_rate": 4.962228183587669e-07,
+      "loss": 0.6925,
+      "step": 5340
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4148657263332098,
+      "learning_rate": 4.94521551408369e-07,
+      "loss": 0.8325,
+      "step": 5341
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9519696104634079,
+      "learning_rate": 4.928231318674836e-07,
+      "loss": 0.9781,
+      "step": 5342
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1312980033649207,
+      "learning_rate": 4.911275602448761e-07,
+      "loss": 0.7863,
+      "step": 5343
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.869523365548825,
+      "learning_rate": 4.894348370484648e-07,
+      "loss": 0.6325,
+      "step": 5344
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8385781590666811,
+      "learning_rate": 4.877449627853126e-07,
+      "loss": 0.7505,
+      "step": 5345
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5469153726084296,
+      "learning_rate": 4.860579379616303e-07,
+      "loss": 0.4037,
+      "step": 5346
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0062688287226715,
+      "learning_rate": 4.843737630827706e-07,
+      "loss": 1.0381,
+      "step": 5347
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1759252207048017,
+      "learning_rate": 4.826924386532383e-07,
+      "loss": 0.927,
+      "step": 5348
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1849072731370485,
+      "learning_rate": 4.810139651766832e-07,
+      "loss": 0.6867,
+      "step": 5349
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0514576032396024,
+      "learning_rate": 4.793383431558962e-07,
+      "loss": 0.8532,
+      "step": 5350
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9024796329526458,
+      "learning_rate": 4.776655730928193e-07,
+      "loss": 0.8271,
+      "step": 5351
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0565610638924332,
+      "learning_rate": 4.75995655488537e-07,
+      "loss": 1.0911,
+      "step": 5352
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8573850553289764,
+      "learning_rate": 4.743285908432849e-07,
+      "loss": 0.6176,
+      "step": 5353
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8708202569548777,
+      "learning_rate": 4.7266437965643455e-07,
+      "loss": 0.955,
+      "step": 5354
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1558817363887939,
+      "learning_rate": 4.7100302242651055e-07,
+      "loss": 0.8682,
+      "step": 5355
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9816237409332046,
+      "learning_rate": 4.6934451965118143e-07,
+      "loss": 1.0294,
+      "step": 5356
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.128274483620327,
+      "learning_rate": 4.6768887182725654e-07,
+      "loss": 0.7794,
+      "step": 5357
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8595915048935537,
+      "learning_rate": 4.660360794506946e-07,
+      "loss": 0.9369,
+      "step": 5358
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.335601903599892,
+      "learning_rate": 4.6438614301659637e-07,
+      "loss": 0.7333,
+      "step": 5359
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8994529640518548,
+      "learning_rate": 4.6273906301920744e-07,
+      "loss": 0.7554,
+      "step": 5360
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9878015756928156,
+      "learning_rate": 4.6109483995191885e-07,
+      "loss": 0.9308,
+      "step": 5361
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9495769963000208,
+      "learning_rate": 4.594534743072654e-07,
+      "loss": 0.9336,
+      "step": 5362
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9098254043163552,
+      "learning_rate": 4.57814966576926e-07,
+      "loss": 0.7244,
+      "step": 5363
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4992833961807001,
+      "learning_rate": 4.5617931725172036e-07,
+      "loss": 0.9783,
+      "step": 5364
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.002454150740738,
+      "learning_rate": 4.545465268216176e-07,
+      "loss": 0.7984,
+      "step": 5365
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5715923645408959,
+      "learning_rate": 4.5291659577572753e-07,
+      "loss": 0.5614,
+      "step": 5366
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0095632405337651,
+      "learning_rate": 4.512895246023019e-07,
+      "loss": 0.6664,
+      "step": 5367
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0127166675000046,
+      "learning_rate": 4.4966531378873857e-07,
+      "loss": 0.6866,
+      "step": 5368
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2006889785971542,
+      "learning_rate": 4.48043963821575e-07,
+      "loss": 0.8146,
+      "step": 5369
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0290529154845123,
+      "learning_rate": 4.4642547518649824e-07,
+      "loss": 0.772,
+      "step": 5370
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9501828958394704,
+      "learning_rate": 4.448098483683294e-07,
+      "loss": 1.0656,
+      "step": 5371
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.017929021483121,
+      "learning_rate": 4.431970838510391e-07,
+      "loss": 0.7216,
+      "step": 5372
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8811704214921845,
+      "learning_rate": 4.4158718211773977e-07,
+      "loss": 0.6825,
+      "step": 5373
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7991156333836646,
+      "learning_rate": 4.399801436506812e-07,
+      "loss": 0.9257,
+      "step": 5374
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9121533618603993,
+      "learning_rate": 4.3837596893126166e-07,
+      "loss": 0.9807,
+      "step": 5375
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8266384148310723,
+      "learning_rate": 4.3677465844001786e-07,
+      "loss": 0.7522,
+      "step": 5376
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.264139613214497,
+      "learning_rate": 4.351762126566328e-07,
+      "loss": 0.7803,
+      "step": 5377
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7951054203982041,
+      "learning_rate": 4.335806320599234e-07,
+      "loss": 0.8885,
+      "step": 5378
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2159302520836561,
+      "learning_rate": 4.3198791712785513e-07,
+      "loss": 0.8163,
+      "step": 5379
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9719316456258195,
+      "learning_rate": 4.303980683375353e-07,
+      "loss": 0.9681,
+      "step": 5380
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9503024452437567,
+      "learning_rate": 4.2881108616520727e-07,
+      "loss": 0.7493,
+      "step": 5381
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5552918714291503,
+      "learning_rate": 4.272269710862587e-07,
+      "loss": 0.4408,
+      "step": 5382
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0276650705396442,
+      "learning_rate": 4.256457235752209e-07,
+      "loss": 0.8508,
+      "step": 5383
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0595912710793658,
+      "learning_rate": 4.240673441057641e-07,
+      "loss": 1.0278,
+      "step": 5384
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0460449712126099,
+      "learning_rate": 4.224918331506955e-07,
+      "loss": 1.0447,
+      "step": 5385
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4786128249382389,
+      "learning_rate": 4.209191911819688e-07,
+      "loss": 0.8153,
+      "step": 5386
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0246023986724166,
+      "learning_rate": 4.193494186706781e-07,
+      "loss": 0.8288,
+      "step": 5387
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5524180142572439,
+      "learning_rate": 4.177825160870519e-07,
+      "loss": 0.968,
+      "step": 5388
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9154069312606067,
+      "learning_rate": 4.162184839004646e-07,
+      "loss": 0.6492,
+      "step": 5389
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0397591957285957,
+      "learning_rate": 4.1465732257943037e-07,
+      "loss": 0.9386,
+      "step": 5390
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9237713561304163,
+      "learning_rate": 4.130990325916018e-07,
+      "loss": 0.7598,
+      "step": 5391
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7238143282944125,
+      "learning_rate": 4.11543614403771e-07,
+      "loss": 0.454,
+      "step": 5392
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9940140233998596,
+      "learning_rate": 4.099910684818698e-07,
+      "loss": 0.8291,
+      "step": 5393
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9860967614227736,
+      "learning_rate": 4.084413952909738e-07,
+      "loss": 0.7623,
+      "step": 5394
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9783160356507873,
+      "learning_rate": 4.068945952952907e-07,
+      "loss": 1.0022,
+      "step": 5395
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9582366174457558,
+      "learning_rate": 4.053506689581743e-07,
+      "loss": 0.7642,
+      "step": 5396
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2452669871724007,
+      "learning_rate": 4.038096167421124e-07,
+      "loss": 0.8266,
+      "step": 5397
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0998051396047495,
+      "learning_rate": 4.022714391087379e-07,
+      "loss": 0.7781,
+      "step": 5398
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0671651311043646,
+      "learning_rate": 4.007361365188145e-07,
+      "loss": 0.9599,
+      "step": 5399
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5661248808391183,
+      "learning_rate": 3.992037094322532e-07,
+      "loss": 0.5003,
+      "step": 5400
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2398001206944091,
+      "learning_rate": 3.9767415830809673e-07,
+      "loss": 0.8801,
+      "step": 5401
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7397851026168816,
+      "learning_rate": 3.9614748360452984e-07,
+      "loss": 0.9712,
+      "step": 5402
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9320979311286107,
+      "learning_rate": 3.9462368577887543e-07,
+      "loss": 0.9296,
+      "step": 5403
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9749910694814771,
+      "learning_rate": 3.9310276528759296e-07,
+      "loss": 1.114,
+      "step": 5404
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.993000903505498,
+      "learning_rate": 3.9158472258628343e-07,
+      "loss": 0.7004,
+      "step": 5405
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8526422428364574,
+      "learning_rate": 3.9006955812967983e-07,
+      "loss": 0.7247,
+      "step": 5406
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4319067764620743,
+      "learning_rate": 3.885572723716591e-07,
+      "loss": 0.8665,
+      "step": 5407
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.026947314744677,
+      "learning_rate": 3.8704786576523434e-07,
+      "loss": 0.9827,
+      "step": 5408
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0887176693926433,
+      "learning_rate": 3.855413387625528e-07,
+      "loss": 0.8045,
+      "step": 5409
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.008612122787596,
+      "learning_rate": 3.840376918149025e-07,
+      "loss": 0.7185,
+      "step": 5410
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9300133662575757,
+      "learning_rate": 3.8253692537270736e-07,
+      "loss": 0.7982,
+      "step": 5411
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6757928575793228,
+      "learning_rate": 3.8103903988553016e-07,
+      "loss": 0.9828,
+      "step": 5412
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9931144379478635,
+      "learning_rate": 3.7954403580206857e-07,
+      "loss": 0.9671,
+      "step": 5413
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9585569466500926,
+      "learning_rate": 3.780519135701566e-07,
+      "loss": 0.6111,
+      "step": 5414
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1413738416639836,
+      "learning_rate": 3.765626736367689e-07,
+      "loss": 0.8015,
+      "step": 5415
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0301101146812168,
+      "learning_rate": 3.75076316448012e-07,
+      "loss": 0.7105,
+      "step": 5416
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0792832839391773,
+      "learning_rate": 3.7359284244913084e-07,
+      "loss": 0.8332,
+      "step": 5417
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5287514543942323,
+      "learning_rate": 3.7211225208450777e-07,
+      "loss": 0.4046,
+      "step": 5418
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.108048866720306,
+      "learning_rate": 3.7063454579766236e-07,
+      "loss": 1.0909,
+      "step": 5419
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8621607635657855,
+      "learning_rate": 3.691597240312439e-07,
+      "loss": 0.7191,
+      "step": 5420
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.20030353516561,
+      "learning_rate": 3.676877872270446e-07,
+      "loss": 0.7633,
+      "step": 5421
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0353735213682205,
+      "learning_rate": 3.6621873582599055e-07,
+      "loss": 0.778,
+      "step": 5422
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1096585111368313,
+      "learning_rate": 3.647525702681409e-07,
+      "loss": 1.0391,
+      "step": 5423
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.043035076867674,
+      "learning_rate": 3.6328929099269327e-07,
+      "loss": 0.8665,
+      "step": 5424
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0157820218409528,
+      "learning_rate": 3.6182889843797917e-07,
+      "loss": 0.7905,
+      "step": 5425
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9874859229880792,
+      "learning_rate": 3.603713930414676e-07,
+      "loss": 0.911,
+      "step": 5426
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0409650620172504,
+      "learning_rate": 3.589167752397582e-07,
+      "loss": 0.9564,
+      "step": 5427
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7957162481232508,
+      "learning_rate": 3.574650454685902e-07,
+      "loss": 1.0954,
+      "step": 5428
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8784817202080443,
+      "learning_rate": 3.5601620416283565e-07,
+      "loss": 0.6434,
+      "step": 5429
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0420045564367977,
+      "learning_rate": 3.5457025175650084e-07,
+      "loss": 0.606,
+      "step": 5430
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2053554135258788,
+      "learning_rate": 3.531271886827281e-07,
+      "loss": 0.775,
+      "step": 5431
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0080733612441963,
+      "learning_rate": 3.516870153737939e-07,
+      "loss": 0.6625,
+      "step": 5432
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0261162459912752,
+      "learning_rate": 3.502497322611076e-07,
+      "loss": 1.0461,
+      "step": 5433
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8828836552358975,
+      "learning_rate": 3.488153397752159e-07,
+      "loss": 0.6992,
+      "step": 5434
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9930870150836221,
+      "learning_rate": 3.4738383834579504e-07,
+      "loss": 0.9086,
+      "step": 5435
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.934728482314934,
+      "learning_rate": 3.4595522840165986e-07,
+      "loss": 0.6143,
+      "step": 5436
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.40063043430978756,
+      "learning_rate": 3.445295103707558e-07,
+      "loss": 0.789,
+      "step": 5437
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0054104997865725,
+      "learning_rate": 3.431066846801634e-07,
+      "loss": 1.1356,
+      "step": 5438
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8753528808049471,
+      "learning_rate": 3.4168675175609624e-07,
+      "loss": 0.6771,
+      "step": 5439
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6637501422315517,
+      "learning_rate": 3.4026971202390404e-07,
+      "loss": 0.7909,
+      "step": 5440
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.11271562706279,
+      "learning_rate": 3.38855565908065e-07,
+      "loss": 0.6597,
+      "step": 5441
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0672361977589062,
+      "learning_rate": 3.374443138321937e-07,
+      "loss": 0.8801,
+      "step": 5442
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.098932807554034,
+      "learning_rate": 3.360359562190374e-07,
+      "loss": 1.1205,
+      "step": 5443
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6658379841080468,
+      "learning_rate": 3.346304934904754e-07,
+      "loss": 0.4423,
+      "step": 5444
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6666055844857364,
+      "learning_rate": 3.33227926067522e-07,
+      "loss": 0.7122,
+      "step": 5445
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8247335048594654,
+      "learning_rate": 3.3182825437031994e-07,
+      "loss": 0.8206,
+      "step": 5446
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0951304545139673,
+      "learning_rate": 3.3043147881815176e-07,
+      "loss": 1.1166,
+      "step": 5447
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9544578063385619,
+      "learning_rate": 3.2903759982942373e-07,
+      "loss": 0.9743,
+      "step": 5448
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8713893842709286,
+      "learning_rate": 3.276466178216797e-07,
+      "loss": 0.7135,
+      "step": 5449
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.13004083794606,
+      "learning_rate": 3.262585332115964e-07,
+      "loss": 0.9736,
+      "step": 5450
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8352718840322981,
+      "learning_rate": 3.248733464149789e-07,
+      "loss": 0.9323,
+      "step": 5451
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1102704240573849,
+      "learning_rate": 3.2349105784676647e-07,
+      "loss": 1.0741,
+      "step": 5452
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.012455009078248,
+      "learning_rate": 3.221116679210312e-07,
+      "loss": 0.6628,
+      "step": 5453
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5405816542815137,
+      "learning_rate": 3.207351770509759e-07,
+      "loss": 0.512,
+      "step": 5454
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0253521625465225,
+      "learning_rate": 3.1936158564893296e-07,
+      "loss": 0.7746,
+      "step": 5455
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0520763715810488,
+      "learning_rate": 3.1799089412636765e-07,
+      "loss": 0.8717,
+      "step": 5456
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.023974179267951,
+      "learning_rate": 3.1662310289388045e-07,
+      "loss": 0.978,
+      "step": 5457
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0693069613286008,
+      "learning_rate": 3.152582123611958e-07,
+      "loss": 0.7382,
+      "step": 5458
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1850790035019096,
+      "learning_rate": 3.138962229371734e-07,
+      "loss": 0.8189,
+      "step": 5459
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1145771106645164,
+      "learning_rate": 3.1253713502980566e-07,
+      "loss": 0.7336,
+      "step": 5460
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4909281478871679,
+      "learning_rate": 3.111809490462103e-07,
+      "loss": 1.0226,
+      "step": 5461
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0480792924720588,
+      "learning_rate": 3.098276653926413e-07,
+      "loss": 0.5682,
+      "step": 5462
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9552863799172208,
+      "learning_rate": 3.08477284474481e-07,
+      "loss": 0.8139,
+      "step": 5463
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1366607206454815,
+      "learning_rate": 3.0712980669624246e-07,
+      "loss": 0.8298,
+      "step": 5464
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.050592673928614,
+      "learning_rate": 3.057852324615662e-07,
+      "loss": 0.7565,
+      "step": 5465
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1802637909700655,
+      "learning_rate": 3.0444356217322893e-07,
+      "loss": 1.1387,
+      "step": 5466
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8446623661311088,
+      "learning_rate": 3.0310479623313125e-07,
+      "loss": 0.769,
+      "step": 5467
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0741260882370973,
+      "learning_rate": 3.0176893504230806e-07,
+      "loss": 0.8205,
+      "step": 5468
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8473471574359125,
+      "learning_rate": 3.0043597900092367e-07,
+      "loss": 0.9171,
+      "step": 5469
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2109078771740878,
+      "learning_rate": 2.9910592850826983e-07,
+      "loss": 0.7441,
+      "step": 5470
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0936709186762805,
+      "learning_rate": 2.977787839627688e-07,
+      "loss": 1.1458,
+      "step": 5471
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0052202721277563,
+      "learning_rate": 2.9645454576197385e-07,
+      "loss": 0.7427,
+      "step": 5472
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5252115861999456,
+      "learning_rate": 2.9513321430256534e-07,
+      "loss": 0.4226,
+      "step": 5473
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8868647033746178,
+      "learning_rate": 2.9381478998035563e-07,
+      "loss": 0.6718,
+      "step": 5474
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.79236178745383,
+      "learning_rate": 2.9249927319028316e-07,
+      "loss": 0.9394,
+      "step": 5475
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8757901188150669,
+      "learning_rate": 2.911866643264183e-07,
+      "loss": 1.0548,
+      "step": 5476
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1474231281432252,
+      "learning_rate": 2.8987696378195873e-07,
+      "loss": 0.6918,
+      "step": 5477
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.959074798112453,
+      "learning_rate": 2.8857017194923174e-07,
+      "loss": 0.8466,
+      "step": 5478
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.786985773332919,
+      "learning_rate": 2.87266289219692e-07,
+      "loss": 0.7289,
+      "step": 5479
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1312715928254313,
+      "learning_rate": 2.8596531598392264e-07,
+      "loss": 1.075,
+      "step": 5480
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.030289490686828,
+      "learning_rate": 2.846672526316396e-07,
+      "loss": 0.616,
+      "step": 5481
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0709821397200883,
+      "learning_rate": 2.833720995516798e-07,
+      "loss": 0.7481,
+      "step": 5482
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2129240840711404,
+      "learning_rate": 2.820798571320127e-07,
+      "loss": 0.7402,
+      "step": 5483
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8930211106928978,
+      "learning_rate": 2.8079052575973764e-07,
+      "loss": 0.7431,
+      "step": 5484
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.909107053012326,
+      "learning_rate": 2.79504105821079e-07,
+      "loss": 1.1648,
+      "step": 5485
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5214114255844172,
+      "learning_rate": 2.7822059770138966e-07,
+      "loss": 0.9705,
+      "step": 5486
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9375543666838451,
+      "learning_rate": 2.7694000178514866e-07,
+      "loss": 0.6871,
+      "step": 5487
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9253800135535459,
+      "learning_rate": 2.75662318455967e-07,
+      "loss": 0.8877,
+      "step": 5488
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.134178396092246,
+      "learning_rate": 2.7438754809657855e-07,
+      "loss": 0.6281,
+      "step": 5489
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9911214536459387,
+      "learning_rate": 2.731156910888477e-07,
+      "loss": 1.0916,
+      "step": 5490
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.505303410982417,
+      "learning_rate": 2.718467478137643e-07,
+      "loss": 0.424,
+      "step": 5491
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.986182919115465,
+      "learning_rate": 2.705807186514475e-07,
+      "loss": 1.0692,
+      "step": 5492
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8234689381678554,
+      "learning_rate": 2.693176039811396e-07,
+      "loss": 0.697,
+      "step": 5493
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1769531951259886,
+      "learning_rate": 2.6805740418121453e-07,
+      "loss": 0.8633,
+      "step": 5494
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0396659963648367,
+      "learning_rate": 2.668001196291714e-07,
+      "loss": 1.0177,
+      "step": 5495
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0851217239529825,
+      "learning_rate": 2.655457507016335e-07,
+      "loss": 0.7505,
+      "step": 5496
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0571059899020365,
+      "learning_rate": 2.6429429777435235e-07,
+      "loss": 0.8176,
+      "step": 5497
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7749445238021713,
+      "learning_rate": 2.6304576122221035e-07,
+      "loss": 0.9208,
+      "step": 5498
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9846287006640113,
+      "learning_rate": 2.618001414192073e-07,
+      "loss": 1.0663,
+      "step": 5499
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9559084976139307,
+      "learning_rate": 2.605574387384779e-07,
+      "loss": 0.7638,
+      "step": 5500
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.097788937976341,
+      "learning_rate": 2.59317653552279e-07,
+      "loss": 0.7387,
+      "step": 5501
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.04882642169501,
+      "learning_rate": 2.580807862319912e-07,
+      "loss": 1.0166,
+      "step": 5502
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0319010821339631,
+      "learning_rate": 2.56846837148127e-07,
+      "loss": 0.6963,
+      "step": 5503
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.106368365616994,
+      "learning_rate": 2.556158066703196e-07,
+      "loss": 1.0748,
+      "step": 5504
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3005595847073945,
+      "learning_rate": 2.5438769516733077e-07,
+      "loss": 0.7807,
+      "step": 5505
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0533834595345373,
+      "learning_rate": 2.5316250300704726e-07,
+      "loss": 0.7875,
+      "step": 5506
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5254932312053633,
+      "learning_rate": 2.519402305564789e-07,
+      "loss": 0.6599,
+      "step": 5507
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9784611347818399,
+      "learning_rate": 2.507208781817638e-07,
+      "loss": 0.665,
+      "step": 5508
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8912242706576158,
+      "learning_rate": 2.495044462481677e-07,
+      "loss": 0.9713,
+      "step": 5509
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5282371065866671,
+      "learning_rate": 2.482909351200735e-07,
+      "loss": 0.4135,
+      "step": 5510
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5120257291245497,
+      "learning_rate": 2.4708034516099597e-07,
+      "loss": 0.9784,
+      "step": 5511
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8832593837576563,
+      "learning_rate": 2.45872676733574e-07,
+      "loss": 1.0114,
+      "step": 5512
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8093003731985757,
+      "learning_rate": 2.4466793019957027e-07,
+      "loss": 0.6096,
+      "step": 5513
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9405017654168546,
+      "learning_rate": 2.434661059198695e-07,
+      "loss": 1.1621,
+      "step": 5514
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.073417388348789,
+      "learning_rate": 2.42267204254486e-07,
+      "loss": 0.7348,
+      "step": 5515
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0738487754400847,
+      "learning_rate": 2.410712255625558e-07,
+      "loss": 0.83,
+      "step": 5516
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1696093953887814,
+      "learning_rate": 2.3987817020233895e-07,
+      "loss": 0.7179,
+      "step": 5517
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8115094619370258,
+      "learning_rate": 2.386880385312218e-07,
+      "loss": 0.7376,
+      "step": 5518
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8381297175974153,
+      "learning_rate": 2.3750083090571253e-07,
+      "loss": 0.7045,
+      "step": 5519
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.148778916140572,
+      "learning_rate": 2.363165476814455e-07,
+      "loss": 0.8329,
+      "step": 5520
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.727490130568437,
+      "learning_rate": 2.3513518921317812e-07,
+      "loss": 0.9149,
+      "step": 5521
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0063466059863573,
+      "learning_rate": 2.339567558547906e-07,
+      "loss": 0.6382,
+      "step": 5522
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.054974903486321,
+      "learning_rate": 2.327812479592906e-07,
+      "loss": 1.0487,
+      "step": 5523
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7904379962848531,
+      "learning_rate": 2.3160866587880414e-07,
+      "loss": 0.6837,
+      "step": 5524
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2331003163916119,
+      "learning_rate": 2.3043900996458478e-07,
+      "loss": 0.8203,
+      "step": 5525
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7100649482511034,
+      "learning_rate": 2.292722805670078e-07,
+      "loss": 0.8614,
+      "step": 5526
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9787492101234762,
+      "learning_rate": 2.2810847803557356e-07,
+      "loss": 0.7347,
+      "step": 5527
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.553701995611392,
+      "learning_rate": 2.2694760271890215e-07,
+      "loss": 0.6222,
+      "step": 5528
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0550558920573518,
+      "learning_rate": 2.2578965496474093e-07,
+      "loss": 0.7197,
+      "step": 5529
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1764546522551744,
+      "learning_rate": 2.2463463511995908e-07,
+      "loss": 0.8856,
+      "step": 5530
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0629788362306685,
+      "learning_rate": 2.2348254353054544e-07,
+      "loss": 0.818,
+      "step": 5531
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9759538580008825,
+      "learning_rate": 2.2233338054161612e-07,
+      "loss": 0.5958,
+      "step": 5532
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9484188300644822,
+      "learning_rate": 2.2118714649740912e-07,
+      "loss": 1.2332,
+      "step": 5533
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0848976425706711,
+      "learning_rate": 2.200438417412809e-07,
+      "loss": 0.7238,
+      "step": 5534
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1431225024770773,
+      "learning_rate": 2.189034666157175e-07,
+      "loss": 0.7664,
+      "step": 5535
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9328829268851012,
+      "learning_rate": 2.1776602146232007e-07,
+      "loss": 0.75,
+      "step": 5536
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5297463753706776,
+      "learning_rate": 2.166315066218172e-07,
+      "loss": 1.0074,
+      "step": 5537
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8278177600426526,
+      "learning_rate": 2.1549992243405814e-07,
+      "loss": 0.6907,
+      "step": 5538
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9045403226501224,
+      "learning_rate": 2.1437126923801287e-07,
+      "loss": 0.7527,
+      "step": 5539
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.032736272479056,
+      "learning_rate": 2.132455473717765e-07,
+      "loss": 0.801,
+      "step": 5540
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9989292125818169,
+      "learning_rate": 2.121227571725626e-07,
+      "loss": 0.8561,
+      "step": 5541
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9695113392451975,
+      "learning_rate": 2.1100289897670877e-07,
+      "loss": 0.9035,
+      "step": 5542
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.978428609863186,
+      "learning_rate": 2.098859731196734e-07,
+      "loss": 0.7617,
+      "step": 5543
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5010156842360699,
+      "learning_rate": 2.087719799360377e-07,
+      "loss": 0.6624,
+      "step": 5544
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9982000374189216,
+      "learning_rate": 2.0766091975950143e-07,
+      "loss": 0.7771,
+      "step": 5545
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0967769211277956,
+      "learning_rate": 2.0655279292288832e-07,
+      "loss": 0.8282,
+      "step": 5546
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5172458521101313,
+      "learning_rate": 2.05447599758144e-07,
+      "loss": 1.0663,
+      "step": 5547
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0472213182094086,
+      "learning_rate": 2.043453405963347e-07,
+      "loss": 0.7526,
+      "step": 5548
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.586526073229017,
+      "learning_rate": 2.0324601576764525e-07,
+      "loss": 0.9042,
+      "step": 5549
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.939990758713253,
+      "learning_rate": 2.021496256013833e-07,
+      "loss": 0.744,
+      "step": 5550
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.4179166394839235,
+      "learning_rate": 2.0105617042597947e-07,
+      "loss": 1.0057,
+      "step": 5551
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.966781612011576,
+      "learning_rate": 1.9996565056898176e-07,
+      "loss": 0.899,
+      "step": 5552
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9355367830239935,
+      "learning_rate": 1.988780663570611e-07,
+      "loss": 0.9229,
+      "step": 5553
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8121359222943191,
+      "learning_rate": 1.9779341811600795e-07,
+      "loss": 0.6945,
+      "step": 5554
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0330046005973161,
+      "learning_rate": 1.9671170617073353e-07,
+      "loss": 0.7659,
+      "step": 5555
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0969812790953868,
+      "learning_rate": 1.956329308452698e-07,
+      "loss": 1.0636,
+      "step": 5556
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9980681480804477,
+      "learning_rate": 1.9455709246276933e-07,
+      "loss": 0.9827,
+      "step": 5557
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9648955568806203,
+      "learning_rate": 1.9348419134550322e-07,
+      "loss": 0.7636,
+      "step": 5558
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1101687501112563,
+      "learning_rate": 1.9241422781486552e-07,
+      "loss": 0.8252,
+      "step": 5559
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5298662846924833,
+      "learning_rate": 1.913472021913665e-07,
+      "loss": 0.9431,
+      "step": 5560
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9814274051985556,
+      "learning_rate": 1.9028311479464158e-07,
+      "loss": 0.6617,
+      "step": 5561
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9508887833264426,
+      "learning_rate": 1.8922196594344023e-07,
+      "loss": 1.0241,
+      "step": 5562
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5490099252005914,
+      "learning_rate": 1.8816375595563485e-07,
+      "loss": 0.4587,
+      "step": 5563
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1341343872837475,
+      "learning_rate": 1.8710848514821856e-07,
+      "loss": 0.7259,
+      "step": 5564
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8592568844073116,
+      "learning_rate": 1.860561538373018e-07,
+      "loss": 0.7737,
+      "step": 5565
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0802383702158203,
+      "learning_rate": 1.8500676233811466e-07,
+      "loss": 0.824,
+      "step": 5566
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8084308505013891,
+      "learning_rate": 1.83960310965009e-07,
+      "loss": 0.7779,
+      "step": 5567
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8117539101311949,
+      "learning_rate": 1.8291680003145074e-07,
+      "loss": 0.6865,
+      "step": 5568
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7185211261835375,
+      "learning_rate": 1.8187622985003096e-07,
+      "loss": 0.9093,
+      "step": 5569
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.184079334814742,
+      "learning_rate": 1.8083860073245806e-07,
+      "loss": 0.7699,
+      "step": 5570
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0358481170952563,
+      "learning_rate": 1.7980391298955456e-07,
+      "loss": 0.9746,
+      "step": 5571
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9768591920697178,
+      "learning_rate": 1.787721669312714e-07,
+      "loss": 0.7692,
+      "step": 5572
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9458297343669968,
+      "learning_rate": 1.7774336286666804e-07,
+      "loss": 0.7283,
+      "step": 5573
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2270498929670004,
+      "learning_rate": 1.7671750110392903e-07,
+      "loss": 0.8366,
+      "step": 5574
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1130014229234868,
+      "learning_rate": 1.756945819503575e-07,
+      "loss": 1.1113,
+      "step": 5575
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2441379719080743,
+      "learning_rate": 1.746746057123705e-07,
+      "loss": 0.7092,
+      "step": 5576
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1262266755098362,
+      "learning_rate": 1.736575726955092e-07,
+      "loss": 0.652,
+      "step": 5577
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9493925343639569,
+      "learning_rate": 1.7264348320442992e-07,
+      "loss": 1.1475,
+      "step": 5578
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.102518911468479,
+      "learning_rate": 1.7163233754290742e-07,
+      "loss": 0.8455,
+      "step": 5579
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5111187556818962,
+      "learning_rate": 1.7062413601383498e-07,
+      "loss": 0.4278,
+      "step": 5580
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.994408696981836,
+      "learning_rate": 1.6961887891922436e-07,
+      "loss": 1.0182,
+      "step": 5581
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.916917739606337,
+      "learning_rate": 1.6861656656020464e-07,
+      "loss": 0.7922,
+      "step": 5582
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9963564594736715,
+      "learning_rate": 1.6761719923702347e-07,
+      "loss": 0.8617,
+      "step": 5583
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8930482814889741,
+      "learning_rate": 1.6662077724904469e-07,
+      "loss": 0.589,
+      "step": 5584
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5292463528345293,
+      "learning_rate": 1.6562730089475177e-07,
+      "loss": 1.0236,
+      "step": 5585
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0688647881143212,
+      "learning_rate": 1.6463677047174553e-07,
+      "loss": 0.7653,
+      "step": 5586
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8382153782540898,
+      "learning_rate": 1.6364918627674308e-07,
+      "loss": 0.6215,
+      "step": 5587
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1629605212234706,
+      "learning_rate": 1.626645486055811e-07,
+      "loss": 0.7236,
+      "step": 5588
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4982011708690595,
+      "learning_rate": 1.6168285775321147e-07,
+      "loss": 0.8741,
+      "step": 5589
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0464221127447817,
+      "learning_rate": 1.6070411401370335e-07,
+      "loss": 1.0443,
+      "step": 5590
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.985265047479654,
+      "learning_rate": 1.5972831768024554e-07,
+      "loss": 0.8223,
+      "step": 5591
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7345552427302988,
+      "learning_rate": 1.5875546904514096e-07,
+      "loss": 0.8787,
+      "step": 5592
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0401739240705046,
+      "learning_rate": 1.5778556839981197e-07,
+      "loss": 0.858,
+      "step": 5593
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8154742566785942,
+      "learning_rate": 1.568186160347951e-07,
+      "loss": 0.6407,
+      "step": 5594
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.013167403576233,
+      "learning_rate": 1.558546122397464e-07,
+      "loss": 1.0001,
+      "step": 5595
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0535088103028587,
+      "learning_rate": 1.5489355730343935e-07,
+      "loss": 0.8005,
+      "step": 5596
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.479525746009308,
+      "learning_rate": 1.539354515137592e-07,
+      "loss": 0.7881,
+      "step": 5597
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8466732701847636,
+      "learning_rate": 1.5298029515771195e-07,
+      "loss": 0.7204,
+      "step": 5598
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3484148237127769,
+      "learning_rate": 1.5202808852141982e-07,
+      "loss": 0.768,
+      "step": 5599
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5503433339894285,
+      "learning_rate": 1.5107883189012018e-07,
+      "loss": 0.6089,
+      "step": 5600
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9703998994875549,
+      "learning_rate": 1.501325255481678e-07,
+      "loss": 1.0557,
+      "step": 5601
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8622312185721797,
+      "learning_rate": 1.4918916977903264e-07,
+      "loss": 0.6874,
+      "step": 5602
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0558142735468687,
+      "learning_rate": 1.482487648653008e-07,
+      "loss": 0.9186,
+      "step": 5603
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5716581053517664,
+      "learning_rate": 1.473113110886759e-07,
+      "loss": 0.8048,
+      "step": 5604
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0613467501008846,
+      "learning_rate": 1.463768087299744e-07,
+      "loss": 1.0474,
+      "step": 5605
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.067571075039783,
+      "learning_rate": 1.4544525806913234e-07,
+      "loss": 0.9164,
+      "step": 5606
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.19380046053625,
+      "learning_rate": 1.44516659385201e-07,
+      "loss": 0.8472,
+      "step": 5607
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2757880439588207,
+      "learning_rate": 1.4359101295634448e-07,
+      "loss": 0.7433,
+      "step": 5608
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8508244123784123,
+      "learning_rate": 1.4266831905984434e-07,
+      "loss": 0.9365,
+      "step": 5609
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5172949773259443,
+      "learning_rate": 1.4174857797209951e-07,
+      "loss": 0.9443,
+      "step": 5610
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9362898417649183,
+      "learning_rate": 1.408317899686207e-07,
+      "loss": 0.7003,
+      "step": 5611
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8599654783423986,
+      "learning_rate": 1.3991795532403708e-07,
+      "loss": 0.7055,
+      "step": 5612
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4619373713808066,
+      "learning_rate": 1.3900707431209194e-07,
+      "loss": 0.9263,
+      "step": 5613
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0631754967936318,
+      "learning_rate": 1.3809914720564478e-07,
+      "loss": 1.0052,
+      "step": 5614
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9670704601509413,
+      "learning_rate": 1.3719417427666693e-07,
+      "loss": 0.9445,
+      "step": 5615
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7867905769873381,
+      "learning_rate": 1.362921557962482e-07,
+      "loss": 0.9353,
+      "step": 5616
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5274827220439182,
+      "learning_rate": 1.353930920345936e-07,
+      "loss": 0.449,
+      "step": 5617
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0502081905137772,
+      "learning_rate": 1.344969832610199e-07,
+      "loss": 0.7142,
+      "step": 5618
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9781175075000617,
+      "learning_rate": 1.3360382974396235e-07,
+      "loss": 1.0053,
+      "step": 5619
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3015757016626723,
+      "learning_rate": 1.3271363175096696e-07,
+      "loss": 0.8732,
+      "step": 5620
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.196734864559147,
+      "learning_rate": 1.318263895486993e-07,
+      "loss": 0.8757,
+      "step": 5621
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9182261726445071,
+      "learning_rate": 1.3094210340293456e-07,
+      "loss": 1.0432,
+      "step": 5622
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1411816334529596,
+      "learning_rate": 1.300607735785664e-07,
+      "loss": 0.7526,
+      "step": 5623
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8776482417097972,
+      "learning_rate": 1.2918240033960028e-07,
+      "loss": 0.9126,
+      "step": 5624
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1639767956061788,
+      "learning_rate": 1.2830698394915685e-07,
+      "loss": 0.8436,
+      "step": 5625
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.903114360062716,
+      "learning_rate": 1.274345246694708e-07,
+      "loss": 0.7979,
+      "step": 5626
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7246660693777017,
+      "learning_rate": 1.2656502276189197e-07,
+      "loss": 0.6101,
+      "step": 5627
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0113677340119582,
+      "learning_rate": 1.256984784868842e-07,
+      "loss": 0.8993,
+      "step": 5628
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0046432766204698,
+      "learning_rate": 1.2483489210402321e-07,
+      "loss": 0.825,
+      "step": 5629
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2708003491344388,
+      "learning_rate": 1.2397426387200096e-07,
+      "loss": 0.8188,
+      "step": 5630
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.017761980853702,
+      "learning_rate": 1.231165940486234e-07,
+      "loss": 0.7357,
+      "step": 5631
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0339405430627195,
+      "learning_rate": 1.2226188289080733e-07,
+      "loss": 0.8016,
+      "step": 5632
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5059533801704336,
+      "learning_rate": 1.2141013065458785e-07,
+      "loss": 0.99,
+      "step": 5633
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7630203252791148,
+      "learning_rate": 1.205613375951098e-07,
+      "loss": 0.4451,
+      "step": 5634
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0378020190746593,
+      "learning_rate": 1.1971550396663422e-07,
+      "loss": 0.6205,
+      "step": 5635
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9140718582273886,
+      "learning_rate": 1.188726300225329e-07,
+      "loss": 0.7149,
+      "step": 5636
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.904212522239965,
+      "learning_rate": 1.1803271601529275e-07,
+      "loss": 0.8175,
+      "step": 5637
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0854871940400654,
+      "learning_rate": 1.1719576219651585e-07,
+      "loss": 1.0652,
+      "step": 5638
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9908950154650122,
+      "learning_rate": 1.1636176881691275e-07,
+      "loss": 0.7779,
+      "step": 5639
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8566600540551912,
+      "learning_rate": 1.1553073612631138e-07,
+      "loss": 0.9269,
+      "step": 5640
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.039525571501704,
+      "learning_rate": 1.147026643736504e-07,
+      "loss": 0.8915,
+      "step": 5641
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9332183841874571,
+      "learning_rate": 1.1387755380698584e-07,
+      "loss": 0.7467,
+      "step": 5642
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9869679865041928,
+      "learning_rate": 1.1305540467347886e-07,
+      "loss": 1.0147,
+      "step": 5643
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9495320269458087,
+      "learning_rate": 1.1223621721941025e-07,
+      "loss": 1.0364,
+      "step": 5644
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9760965889405412,
+      "learning_rate": 1.1141999169017259e-07,
+      "loss": 0.6799,
+      "step": 5645
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9718604602449318,
+      "learning_rate": 1.1060672833026697e-07,
+      "loss": 0.866,
+      "step": 5646
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9397909491592042,
+      "learning_rate": 1.0979642738331187e-07,
+      "loss": 0.7226,
+      "step": 5647
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9671559843886062,
+      "learning_rate": 1.0898908909203643e-07,
+      "loss": 0.9263,
+      "step": 5648
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.037024527722676,
+      "learning_rate": 1.0818471369828276e-07,
+      "loss": 0.6637,
+      "step": 5649
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0957263199735152,
+      "learning_rate": 1.0738330144300479e-07,
+      "loss": 0.7662,
+      "step": 5650
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.261573700984344,
+      "learning_rate": 1.0658485256626938e-07,
+      "loss": 0.9232,
+      "step": 5651
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.05066276476441,
+      "learning_rate": 1.0578936730725631e-07,
+      "loss": 0.696,
+      "step": 5652
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5990185531152953,
+      "learning_rate": 1.0499684590425496e-07,
+      "loss": 0.6677,
+      "step": 5653
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9402515412344353,
+      "learning_rate": 1.0420728859466988e-07,
+      "loss": 0.8124,
+      "step": 5654
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1058699406170966,
+      "learning_rate": 1.0342069561501744e-07,
+      "loss": 0.7115,
+      "step": 5655
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.88699661691041,
+      "learning_rate": 1.0263706720092359e-07,
+      "loss": 0.7745,
+      "step": 5656
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8800368689677602,
+      "learning_rate": 1.0185640358712834e-07,
+      "loss": 0.996,
+      "step": 5657
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2543995432198498,
+      "learning_rate": 1.0107870500748351e-07,
+      "loss": 0.8216,
+      "step": 5658
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0732186881920167,
+      "learning_rate": 1.0030397169495276e-07,
+      "loss": 0.6358,
+      "step": 5659
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5172787314261491,
+      "learning_rate": 9.953220388160934e-08,
+      "loss": 0.9743,
+      "step": 5660
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8086075317676928,
+      "learning_rate": 9.876340179863941e-08,
+      "loss": 0.6522,
+      "step": 5661
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1348555346816052,
+      "learning_rate": 9.799756567634433e-08,
+      "loss": 1.0056,
+      "step": 5662
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.261126626328741,
+      "learning_rate": 9.723469574413058e-08,
+      "loss": 0.7679,
+      "step": 5663
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7032177132270081,
+      "learning_rate": 9.647479223052093e-08,
+      "loss": 1.044,
+      "step": 5664
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2341125682742036,
+      "learning_rate": 9.571785536314771e-08,
+      "loss": 0.9484,
+      "step": 5665
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0736019454745998,
+      "learning_rate": 9.496388536875623e-08,
+      "loss": 1.0043,
+      "step": 5666
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.064696683497531,
+      "learning_rate": 9.421288247319805e-08,
+      "loss": 0.9053,
+      "step": 5667
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9061327030367555,
+      "learning_rate": 9.34648469014432e-08,
+      "loss": 0.7822,
+      "step": 5668
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.883573386674691,
+      "learning_rate": 9.271977887756579e-08,
+      "loss": 0.6134,
+      "step": 5669
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3570848070851596,
+      "learning_rate": 9.197767862475837e-08,
+      "loss": 0.6677,
+      "step": 5670
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0940525660808522,
+      "learning_rate": 9.123854636531648e-08,
+      "loss": 1.0577,
+      "step": 5671
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0660182444089294,
+      "learning_rate": 9.0502382320653e-08,
+      "loss": 0.8449,
+      "step": 5672
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.577729237019234,
+      "learning_rate": 8.97691867112882e-08,
+      "loss": 0.5004,
+      "step": 5673
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3136670354702134,
+      "learning_rate": 8.903895975685528e-08,
+      "loss": 0.7673,
+      "step": 5674
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7953047425986861,
+      "learning_rate": 8.831170167609482e-08,
+      "loss": 0.5225,
+      "step": 5675
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0226819790490238,
+      "learning_rate": 8.758741268686366e-08,
+      "loss": 0.9445,
+      "step": 5676
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0529418094038405,
+      "learning_rate": 8.68660930061227e-08,
+      "loss": 0.8259,
+      "step": 5677
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0298627112579435,
+      "learning_rate": 8.614774284994797e-08,
+      "loss": 0.8169,
+      "step": 5678
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9921286528317252,
+      "learning_rate": 8.543236243352405e-08,
+      "loss": 0.6478,
+      "step": 5679
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.989365730462618,
+      "learning_rate": 8.471995197114836e-08,
+      "loss": 0.8202,
+      "step": 5680
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2431471527807902,
+      "learning_rate": 8.401051167622354e-08,
+      "loss": 1.0904,
+      "step": 5681
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9769090856440639,
+      "learning_rate": 8.33040417612685e-08,
+      "loss": 0.7173,
+      "step": 5682
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1207178300428409,
+      "learning_rate": 8.260054243790838e-08,
+      "loss": 0.8187,
+      "step": 5683
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4696364813340114,
+      "learning_rate": 8.190001391687797e-08,
+      "loss": 0.9647,
+      "step": 5684
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9197875790996657,
+      "learning_rate": 8.120245640802715e-08,
+      "loss": 0.6194,
+      "step": 5685
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7744490873398634,
+      "learning_rate": 8.05078701203088e-08,
+      "loss": 1.0243,
+      "step": 5686
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3209861548323782,
+      "learning_rate": 7.981625526179315e-08,
+      "loss": 0.756,
+      "step": 5687
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7089519725477718,
+      "learning_rate": 7.912761203965336e-08,
+      "loss": 0.9953,
+      "step": 5688
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2046839660797573,
+      "learning_rate": 7.844194066017775e-08,
+      "loss": 0.765,
+      "step": 5689
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5085173476530651,
+      "learning_rate": 7.7759241328762e-08,
+      "loss": 0.4104,
+      "step": 5690
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8900571228209281,
+      "learning_rate": 7.707951424991033e-08,
+      "loss": 0.9171,
+      "step": 5691
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0468447206341756,
+      "learning_rate": 7.640275962723986e-08,
+      "loss": 0.7601,
+      "step": 5692
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0037468868228052,
+      "learning_rate": 7.57289776634751e-08,
+      "loss": 0.8046,
+      "step": 5693
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.225970452771425,
+      "learning_rate": 7.505816856045012e-08,
+      "loss": 0.7281,
+      "step": 5694
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0854981979953784,
+      "learning_rate": 7.439033251910865e-08,
+      "loss": 0.9668,
+      "step": 5695
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.084838363159099,
+      "learning_rate": 7.372546973950401e-08,
+      "loss": 1.0038,
+      "step": 5696
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.254544633208669,
+      "learning_rate": 7.30635804208002e-08,
+      "loss": 0.6846,
+      "step": 5697
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2279277037523397,
+      "learning_rate": 7.24046647612675e-08,
+      "loss": 0.686,
+      "step": 5698
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9074850728284156,
+      "learning_rate": 7.174872295828805e-08,
+      "loss": 0.8077,
+      "step": 5699
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.000433265949979,
+      "learning_rate": 7.109575520835244e-08,
+      "loss": 0.9876,
+      "step": 5700
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0438380388744488,
+      "learning_rate": 7.044576170705864e-08,
+      "loss": 0.7397,
+      "step": 5701
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.845554321050414,
+      "learning_rate": 6.979874264911756e-08,
+      "loss": 0.7135,
+      "step": 5702
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0751182962402939,
+      "learning_rate": 6.91546982283442e-08,
+      "loss": 0.7559,
+      "step": 5703
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1833937890676758,
+      "learning_rate": 6.851362863766641e-08,
+      "loss": 0.7182,
+      "step": 5704
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.098992132304527,
+      "learning_rate": 6.787553406912062e-08,
+      "loss": 1.1279,
+      "step": 5705
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9315206549573103,
+      "learning_rate": 6.724041471384835e-08,
+      "loss": 0.842,
+      "step": 5706
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5536486309852762,
+      "learning_rate": 6.660827076210407e-08,
+      "loss": 0.4236,
+      "step": 5707
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9793332960410419,
+      "learning_rate": 6.597910240324967e-08,
+      "loss": 0.9749,
+      "step": 5708
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4629304505755405,
+      "learning_rate": 6.535290982575437e-08,
+      "loss": 0.968,
+      "step": 5709
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0495019812190074,
+      "learning_rate": 6.472969321719702e-08,
+      "loss": 0.9117,
+      "step": 5710
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9864122162316001,
+      "learning_rate": 6.410945276426717e-08,
+      "loss": 0.7258,
+      "step": 5711
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3596862166414052,
+      "learning_rate": 6.34921886527573e-08,
+      "loss": 0.8435,
+      "step": 5712
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9148465975745066,
+      "learning_rate": 6.287790106757396e-08,
+      "loss": 0.7576,
+      "step": 5713
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7758392769980686,
+      "learning_rate": 6.226659019272997e-08,
+      "loss": 1.0962,
+      "step": 5714
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0950577202390726,
+      "learning_rate": 6.165825621134546e-08,
+      "loss": 0.8382,
+      "step": 5715
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7868644346326197,
+      "learning_rate": 6.105289930564917e-08,
+      "loss": 0.7508,
+      "step": 5716
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4398172499474746,
+      "learning_rate": 6.045051965697824e-08,
+      "loss": 0.8244,
+      "step": 5717
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.054128325149779,
+      "learning_rate": 5.985111744578165e-08,
+      "loss": 0.8504,
+      "step": 5718
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1369740063459108,
+      "learning_rate": 5.925469285160801e-08,
+      "loss": 1.1013,
+      "step": 5719
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0263614928655405,
+      "learning_rate": 5.866124605312329e-08,
+      "loss": 0.743,
+      "step": 5720
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9625768758855782,
+      "learning_rate": 5.8070777228094176e-08,
+      "loss": 0.78,
+      "step": 5721
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1109681241795488,
+      "learning_rate": 5.748328655340141e-08,
+      "loss": 0.8507,
+      "step": 5722
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0315959211679908,
+      "learning_rate": 5.689877420502754e-08,
+      "loss": 0.6508,
+      "step": 5723
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0482675187069788,
+      "learning_rate": 5.6317240358066963e-08,
+      "loss": 1.0152,
+      "step": 5724
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5638655607342572,
+      "learning_rate": 5.573868518672143e-08,
+      "loss": 0.419,
+      "step": 5725
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1718430335889347,
+      "learning_rate": 5.516310886429899e-08,
+      "loss": 0.8985,
+      "step": 5726
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9890428304472393,
+      "learning_rate": 5.459051156321615e-08,
+      "loss": 0.6735,
+      "step": 5727
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9086055894686001,
+      "learning_rate": 5.402089345499795e-08,
+      "loss": 0.7882,
+      "step": 5728
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9963103908863787,
+      "learning_rate": 5.345425471027566e-08,
+      "loss": 1.285,
+      "step": 5729
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7937737779827041,
+      "learning_rate": 5.2890595498787944e-08,
+      "loss": 0.8148,
+      "step": 5730
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.013500952026604,
+      "learning_rate": 5.232991598938197e-08,
+      "loss": 0.8535,
+      "step": 5731
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2109240688986707,
+      "learning_rate": 5.177221635001228e-08,
+      "loss": 0.725,
+      "step": 5732
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0987021291199173,
+      "learning_rate": 5.121749674773857e-08,
+      "loss": 1.0508,
+      "step": 5733
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8734767178108006,
+      "learning_rate": 5.0665757348732355e-08,
+      "loss": 0.6006,
+      "step": 5734
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5319436284141529,
+      "learning_rate": 5.0116998318266996e-08,
+      "loss": 0.9195,
+      "step": 5735
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3577829925128775,
+      "learning_rate": 4.957121982072766e-08,
+      "loss": 0.8493,
+      "step": 5736
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8074482309930358,
+      "learning_rate": 4.902842201960467e-08,
+      "loss": 0.9823,
+      "step": 5737
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9734919757167428,
+      "learning_rate": 4.848860507749353e-08,
+      "loss": 1.0434,
+      "step": 5738
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9730064380497063,
+      "learning_rate": 4.795176915610267e-08,
+      "loss": 0.6504,
+      "step": 5739
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2701558115584795,
+      "learning_rate": 4.7417914416239e-08,
+      "loss": 0.4572,
+      "step": 5740
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0638846895999967,
+      "learning_rate": 4.688704101782571e-08,
+      "loss": 0.9176,
+      "step": 5741
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9487794758818584,
+      "learning_rate": 4.635914911988448e-08,
+      "loss": 0.708,
+      "step": 5742
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5577126576103548,
+      "learning_rate": 4.583423888055105e-08,
+      "loss": 0.6327,
+      "step": 5743
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0270410410442758,
+      "learning_rate": 4.531231045706186e-08,
+      "loss": 0.888,
+      "step": 5744
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3673982054496845,
+      "learning_rate": 4.479336400576517e-08,
+      "loss": 0.7472,
+      "step": 5745
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.138293559656343,
+      "learning_rate": 4.427739968211331e-08,
+      "loss": 0.7485,
+      "step": 5746
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.015439917921178,
+      "learning_rate": 4.376441764066375e-08,
+      "loss": 0.7213,
+      "step": 5747
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9817396377673818,
+      "learning_rate": 4.325441803508468e-08,
+      "loss": 0.9243,
+      "step": 5748
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9425074936818532,
+      "learning_rate": 4.2747401018149445e-08,
+      "loss": 0.7412,
+      "step": 5749
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0063045024630377,
+      "learning_rate": 4.2243366741735457e-08,
+      "loss": 0.7623,
+      "step": 5750
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1213076609716042,
+      "learning_rate": 4.17423153568286e-08,
+      "loss": 0.7363,
+      "step": 5751
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8576146448414809,
+      "learning_rate": 4.1244247013522144e-08,
+      "loss": 1.2181,
+      "step": 5752
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9726362183692788,
+      "learning_rate": 4.074916186101452e-08,
+      "loss": 0.5492,
+      "step": 5753
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0584782497854313,
+      "learning_rate": 4.025706004760932e-08,
+      "loss": 0.7681,
+      "step": 5754
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.271258691038895,
+      "learning_rate": 3.9767941720720847e-08,
+      "loss": 0.9112,
+      "step": 5755
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0775253465908983,
+      "learning_rate": 3.9281807026863017e-08,
+      "loss": 0.9332,
+      "step": 5756
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5011020925245485,
+      "learning_rate": 3.8798656111663776e-08,
+      "loss": 1.0237,
+      "step": 5757
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8527149064846747,
+      "learning_rate": 3.831848911984959e-08,
+      "loss": 0.647,
+      "step": 5758
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2748124704659423,
+      "learning_rate": 3.784130619525872e-08,
+      "loss": 0.8655,
+      "step": 5759
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6900772158362714,
+      "learning_rate": 3.7367107480832385e-08,
+      "loss": 0.8918,
+      "step": 5760
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5474299720944874,
+      "learning_rate": 3.6895893118619186e-08,
+      "loss": 0.424,
+      "step": 5761
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0417801041235477,
+      "learning_rate": 3.642766324977509e-08,
+      "loss": 1.1207,
+      "step": 5762
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9285871785858432,
+      "learning_rate": 3.596241801455791e-08,
+      "loss": 0.6905,
+      "step": 5763
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.084987703074076,
+      "learning_rate": 3.550015755233727e-08,
+      "loss": 0.7936,
+      "step": 5764
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.057031486714518,
+      "learning_rate": 3.504088200158129e-08,
+      "loss": 0.6444,
+      "step": 5765
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1844658718320724,
+      "learning_rate": 3.4584591499872146e-08,
+      "loss": 0.8136,
+      "step": 5766
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.16626083843335,
+      "learning_rate": 3.41312861838905e-08,
+      "loss": 0.9458,
+      "step": 5767
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9377316251013298,
+      "learning_rate": 3.368096618942773e-08,
+      "loss": 0.6932,
+      "step": 5768
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1558240442558083,
+      "learning_rate": 3.323363165137927e-08,
+      "loss": 0.9185,
+      "step": 5769
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9763363576786024,
+      "learning_rate": 3.278928270374459e-08,
+      "loss": 0.7087,
+      "step": 5770
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0929807517788144,
+      "learning_rate": 3.234791947963278e-08,
+      "loss": 0.8262,
+      "step": 5771
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8955629090373309,
+      "learning_rate": 3.190954211125585e-08,
+      "loss": 0.933,
+      "step": 5772
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.248262464703514,
+      "learning_rate": 3.147415072992988e-08,
+      "loss": 0.738,
+      "step": 5773
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.1208966243243355,
+      "learning_rate": 3.104174546607941e-08,
+      "loss": 0.8437,
+      "step": 5774
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9568121235632154,
+      "learning_rate": 3.0612326449234175e-08,
+      "loss": 0.9487,
+      "step": 5775
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8772918633524919,
+      "learning_rate": 3.0185893808026834e-08,
+      "loss": 0.9469,
+      "step": 5776
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9781625723098215,
+      "learning_rate": 2.9762447670199645e-08,
+      "loss": 0.6326,
+      "step": 5777
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9429110778800257,
+      "learning_rate": 2.9341988162595593e-08,
+      "loss": 0.6262,
+      "step": 5778
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5314100380785607,
+      "learning_rate": 2.8924515411166144e-08,
+      "loss": 0.4472,
+      "step": 5779
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9374671638454634,
+      "learning_rate": 2.8510029540967933e-08,
+      "loss": 0.7002,
+      "step": 5780
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2166108529142803,
+      "learning_rate": 2.809853067616053e-08,
+      "loss": 1.136,
+      "step": 5781
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4338209647577566,
+      "learning_rate": 2.7690018940011996e-08,
+      "loss": 1.0028,
+      "step": 5782
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9113469576030977,
+      "learning_rate": 2.7284494454892228e-08,
+      "loss": 0.825,
+      "step": 5783
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9945763459400491,
+      "learning_rate": 2.6881957342278498e-08,
+      "loss": 0.6772,
+      "step": 5784
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1745572613096231,
+      "learning_rate": 2.6482407722753235e-08,
+      "loss": 0.764,
+      "step": 5785
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3373124926967503,
+      "learning_rate": 2.6085845716004034e-08,
+      "loss": 1.1963,
+      "step": 5786
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9479272722299004,
+      "learning_rate": 2.5692271440820314e-08,
+      "loss": 0.7256,
+      "step": 5787
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9865453366218057,
+      "learning_rate": 2.5301685015101108e-08,
+      "loss": 0.929,
+      "step": 5788
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1496776564538884,
+      "learning_rate": 2.4914086555848373e-08,
+      "loss": 0.7406,
+      "step": 5789
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9550221675763335,
+      "learning_rate": 2.452947617916701e-08,
+      "loss": 0.7816,
+      "step": 5790
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0208292716489034,
+      "learning_rate": 2.4147854000271532e-08,
+      "loss": 0.9353,
+      "step": 5791
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1309922870927538,
+      "learning_rate": 2.3769220133477156e-08,
+      "loss": 0.7393,
+      "step": 5792
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1166836519241952,
+      "learning_rate": 2.339357469220427e-08,
+      "loss": 0.8212,
+      "step": 5793
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8466647061568685,
+      "learning_rate": 2.3020917788981744e-08,
+      "loss": 0.9051,
+      "step": 5794
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0403576458311623,
+      "learning_rate": 2.265124953543918e-08,
+      "loss": 0.6578,
+      "step": 5795
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8840317293211697,
+      "learning_rate": 2.2284570042312438e-08,
+      "loss": 1.1884,
+      "step": 5796
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.127946836752766,
+      "learning_rate": 2.192087941944143e-08,
+      "loss": 0.7734,
+      "step": 5797
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7423490784987673,
+      "learning_rate": 2.156017777577346e-08,
+      "loss": 0.5309,
+      "step": 5798
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6866049153592938,
+      "learning_rate": 2.1202465219355427e-08,
+      "loss": 0.4065,
+      "step": 5799
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8162546623796091,
+      "learning_rate": 2.084774185734495e-08,
+      "loss": 0.8954,
+      "step": 5800
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0382763799002068,
+      "learning_rate": 2.049600779599814e-08,
+      "loss": 0.7097,
+      "step": 5801
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.11880855663959,
+      "learning_rate": 2.0147263140680717e-08,
+      "loss": 0.8389,
+      "step": 5802
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8922586238156663,
+      "learning_rate": 1.9801507995859114e-08,
+      "loss": 0.7167,
+      "step": 5803
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0358832307161598,
+      "learning_rate": 1.945874246510715e-08,
+      "loss": 0.7703,
+      "step": 5804
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.088186988832276,
+      "learning_rate": 1.911896665110047e-08,
+      "loss": 1.117,
+      "step": 5805
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9647259740682177,
+      "learning_rate": 1.8782180655622096e-08,
+      "loss": 0.8062,
+      "step": 5806
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4669154516288275,
+      "learning_rate": 1.844838457955578e-08,
+      "loss": 1.0037,
+      "step": 5807
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0324840359374212,
+      "learning_rate": 1.8117578522893752e-08,
+      "loss": 0.8577,
+      "step": 5808
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8455674321856943,
+      "learning_rate": 1.7789762584728975e-08,
+      "loss": 0.6877,
+      "step": 5809
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9939572273549241,
+      "learning_rate": 1.7464936863261785e-08,
+      "loss": 0.943,
+      "step": 5810
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3580115870995673,
+      "learning_rate": 1.7143101455793232e-08,
+      "loss": 0.6556,
+      "step": 5811
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.112379671396605,
+      "learning_rate": 1.6824256458731757e-08,
+      "loss": 0.8499,
+      "step": 5812
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0430200928527087,
+      "learning_rate": 1.6508401967588738e-08,
+      "loss": 0.7056,
+      "step": 5813
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.097845426725378,
+      "learning_rate": 1.61955380769796e-08,
+      "loss": 0.7685,
+      "step": 5814
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5951704344399248,
+      "learning_rate": 1.588566488062382e-08,
+      "loss": 0.6775,
+      "step": 5815
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8942246150619498,
+      "learning_rate": 1.5578782471347144e-08,
+      "loss": 0.7247,
+      "step": 5816
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.940950023708372,
+      "learning_rate": 1.5274890941074927e-08,
+      "loss": 0.6723,
+      "step": 5817
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.155980825044137,
+      "learning_rate": 1.497399038084102e-08,
+      "loss": 0.7973,
+      "step": 5818
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8015600893657243,
+      "learning_rate": 1.4676080880782207e-08,
+      "loss": 1.1859,
+      "step": 5819
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1034197631197125,
+      "learning_rate": 1.4381162530135995e-08,
+      "loss": 0.8377,
+      "step": 5820
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9766001072343267,
+      "learning_rate": 1.4089235417248382e-08,
+      "loss": 0.7686,
+      "step": 5821
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8029488290446841,
+      "learning_rate": 1.3800299629568303e-08,
+      "loss": 0.6378,
+      "step": 5822
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9075805194324258,
+      "learning_rate": 1.3514355253645417e-08,
+      "loss": 0.6435,
+      "step": 5823
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9066416726867859,
+      "learning_rate": 1.323140237513787e-08,
+      "loss": 0.9614,
+      "step": 5824
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0566530088794994,
+      "learning_rate": 1.2951441078803417e-08,
+      "loss": 0.7599,
+      "step": 5825
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0908986245472858,
+      "learning_rate": 1.2674471448508308e-08,
+      "loss": 0.8146,
+      "step": 5826
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.148090264177177,
+      "learning_rate": 1.2400493567217286e-08,
+      "loss": 0.7729,
+      "step": 5827
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9451310667985121,
+      "learning_rate": 1.2129507517003591e-08,
+      "loss": 0.6446,
+      "step": 5828
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.964207269937158,
+      "learning_rate": 1.1861513379042288e-08,
+      "loss": 0.9505,
+      "step": 5829
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0820593489209045,
+      "learning_rate": 1.1596511233610275e-08,
+      "loss": 0.7669,
+      "step": 5830
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7566500101162243,
+      "learning_rate": 1.1334501160092936e-08,
+      "loss": 0.9397,
+      "step": 5831
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5446381663541408,
+      "learning_rate": 1.1075483236974161e-08,
+      "loss": 0.9716,
+      "step": 5832
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5111865433001883,
+      "learning_rate": 1.0819457541846324e-08,
+      "loss": 0.4808,
+      "step": 5833
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.929919389611408,
+      "learning_rate": 1.0566424151401412e-08,
+      "loss": 1.0127,
+      "step": 5834
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9328907812519929,
+      "learning_rate": 1.031638314143657e-08,
+      "loss": 0.6963,
+      "step": 5835
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1494201052180018,
+      "learning_rate": 1.0069334586854106e-08,
+      "loss": 0.6407,
+      "step": 5836
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.869718566630787,
+      "learning_rate": 9.825278561657048e-09,
+      "loss": 0.8173,
+      "step": 5837
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1259922921731544,
+      "learning_rate": 9.584215138953579e-09,
+      "loss": 1.072,
+      "step": 5838
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9075589617072092,
+      "learning_rate": 9.346144390957046e-09,
+      "loss": 0.8163,
+      "step": 5839
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9364263668086085,
+      "learning_rate": 9.111066388981515e-09,
+      "loss": 0.9691,
+      "step": 5840
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8228611782231608,
+      "learning_rate": 8.878981203445103e-09,
+      "loss": 0.4209,
+      "step": 5841
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.277695745409952,
+      "learning_rate": 8.649888903869973e-09,
+      "loss": 0.9262,
+      "step": 5842
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0160338677278618,
+      "learning_rate": 8.423789558883455e-09,
+      "loss": 1.0795,
+      "step": 5843
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3219713909320185,
+      "learning_rate": 8.200683236213592e-09,
+      "loss": 0.7902,
+      "step": 5844
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.005941277415849,
+      "learning_rate": 7.980570002692479e-09,
+      "loss": 0.9517,
+      "step": 5845
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9802952167912904,
+      "learning_rate": 7.763449924256262e-09,
+      "loss": 0.7925,
+      "step": 5846
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3854583090018837,
+      "learning_rate": 7.549323065944025e-09,
+      "loss": 0.8271,
+      "step": 5847
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9175809942696226,
+      "learning_rate": 7.338189491900016e-09,
+      "loss": 0.9138,
+      "step": 5848
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1214623546544715,
+      "learning_rate": 7.1300492653680885e-09,
+      "loss": 0.8499,
+      "step": 5849
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9054399013846957,
+      "learning_rate": 6.924902448699478e-09,
+      "loss": 0.6881,
+      "step": 5850
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.115198470446176,
+      "learning_rate": 6.722749103345028e-09,
+      "loss": 0.7496,
+      "step": 5851
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6135005768535007,
+      "learning_rate": 6.5235892898618535e-09,
+      "loss": 0.5794,
+      "step": 5852
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.085761975603232,
+      "learning_rate": 6.3274230679089e-09,
+      "loss": 0.7883,
+      "step": 5853
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9472558659897032,
+      "learning_rate": 6.134250496249161e-09,
+      "loss": 0.8079,
+      "step": 5854
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.48166848556657527,
+      "learning_rate": 5.9440716327474615e-09,
+      "loss": 0.976,
+      "step": 5855
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0758867167502542,
+      "learning_rate": 5.756886534373785e-09,
+      "loss": 0.7149,
+      "step": 5856
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.883924921680165,
+      "learning_rate": 5.572695257199945e-09,
+      "loss": 0.9473,
+      "step": 5857
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.062105435208494,
+      "learning_rate": 5.391497856399585e-09,
+      "loss": 0.683,
+      "step": 5858
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2696966635326938,
+      "learning_rate": 5.213294386253731e-09,
+      "loss": 0.9103,
+      "step": 5859
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1233225451906368,
+      "learning_rate": 5.0380849001430145e-09,
+      "loss": 0.7505,
+      "step": 5860
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.886645270332159,
+      "learning_rate": 4.865869450552118e-09,
+      "loss": 0.9622,
+      "step": 5861
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.102503627863259,
+      "learning_rate": 4.696648089068667e-09,
+      "loss": 1.0086,
+      "step": 5862
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0177377652943234,
+      "learning_rate": 4.530420866384333e-09,
+      "loss": 0.8512,
+      "step": 5863
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8900717815150372,
+      "learning_rate": 4.367187832293729e-09,
+      "loss": 0.7828,
+      "step": 5864
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1936573627955367,
+      "learning_rate": 4.206949035692187e-09,
+      "loss": 0.8162,
+      "step": 5865
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1494600628757872,
+      "learning_rate": 4.0497045245813105e-09,
+      "loss": 0.8581,
+      "step": 5866
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9640258538073583,
+      "learning_rate": 3.8954543460645315e-09,
+      "loss": 0.933,
+      "step": 5867
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9843664250436099,
+      "learning_rate": 3.744198546348221e-09,
+      "loss": 0.8651,
+      "step": 5868
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9386770821574967,
+      "learning_rate": 3.595937170740582e-09,
+      "loss": 0.6242,
+      "step": 5869
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.526636726114185,
+      "learning_rate": 3.450670263654976e-09,
+      "loss": 0.4428,
+      "step": 5870
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9487613333284615,
+      "learning_rate": 3.3083978686077046e-09,
+      "loss": 0.8902,
+      "step": 5871
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9106292455497744,
+      "learning_rate": 3.1691200282146785e-09,
+      "loss": 0.9313,
+      "step": 5872
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0369348824774116,
+      "learning_rate": 3.032836784199189e-09,
+      "loss": 0.8254,
+      "step": 5873
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.056133567247758,
+      "learning_rate": 2.899548177384137e-09,
+      "loss": 0.6947,
+      "step": 5874
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0729320332704904,
+      "learning_rate": 2.769254247698694e-09,
+      "loss": 0.7385,
+      "step": 5875
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9481993410138811,
+      "learning_rate": 2.641955034170529e-09,
+      "loss": 0.6568,
+      "step": 5876
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3015852701523754,
+      "learning_rate": 2.5176505749346937e-09,
+      "loss": 1.1462,
+      "step": 5877
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.682280555554505,
+      "learning_rate": 2.396340907225847e-09,
+      "loss": 0.9485,
+      "step": 5878
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9930068090169553,
+      "learning_rate": 2.2780260673838096e-09,
+      "loss": 0.776,
+      "step": 5879
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1450889070449657,
+      "learning_rate": 2.1627060908491204e-09,
+      "loss": 0.7504,
+      "step": 5880
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0309574315077907,
+      "learning_rate": 2.0503810121685895e-09,
+      "loss": 1.0462,
+      "step": 5881
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9959209259712252,
+      "learning_rate": 1.9410508649875258e-09,
+      "loss": 1.0606,
+      "step": 5882
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.48721651278945105,
+      "learning_rate": 1.8347156820563983e-09,
+      "loss": 0.8905,
+      "step": 5883
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9268167839573028,
+      "learning_rate": 1.7313754952297258e-09,
+      "loss": 0.8581,
+      "step": 5884
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3483245056881683,
+      "learning_rate": 1.6310303354638569e-09,
+      "loss": 0.8084,
+      "step": 5885
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1880470544924888,
+      "learning_rate": 1.5336802328147492e-09,
+      "loss": 1.0676,
+      "step": 5886
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5462587884809762,
+      "learning_rate": 1.439325216447962e-09,
+      "loss": 0.4528,
+      "step": 5887
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1480519921861063,
+      "learning_rate": 1.3479653146242222e-09,
+      "loss": 0.8111,
+      "step": 5888
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9844198239924513,
+      "learning_rate": 1.259600554713858e-09,
+      "loss": 0.8637,
+      "step": 5889
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1711053730749852,
+      "learning_rate": 1.1742309631845861e-09,
+      "loss": 0.7491,
+      "step": 5890
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2691414524246971,
+      "learning_rate": 1.0918565656115043e-09,
+      "loss": 0.9393,
+      "step": 5891
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9006292018436347,
+      "learning_rate": 1.012477386668209e-09,
+      "loss": 0.6118,
+      "step": 5892
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8270205359940598,
+      "learning_rate": 9.360934501345675e-10,
+      "loss": 0.7087,
+      "step": 5893
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3066264598459827,
+      "learning_rate": 8.627047788911658e-10,
+      "loss": 0.8989,
+      "step": 5894
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9826448969343295,
+      "learning_rate": 7.923113949204198e-10,
+      "loss": 1.0199,
+      "step": 5895
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.03853561052323,
+      "learning_rate": 7.24913319312126e-10,
+      "loss": 0.7028,
+      "step": 5896
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.144620095523145,
+      "learning_rate": 6.605105722534699e-10,
+      "loss": 0.9318,
+      "step": 5897
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.922286919089343,
+      "learning_rate": 5.991031730367968e-10,
+      "loss": 0.7325,
+      "step": 5898
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1355952876395983,
+      "learning_rate": 5.406911400573922e-10,
+      "loss": 0.7298,
+      "step": 5899
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0662373524727038,
+      "learning_rate": 4.85274490813481e-10,
+      "loss": 0.854,
+      "step": 5900
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.330287318975095,
+      "learning_rate": 4.3285324190400766e-10,
+      "loss": 0.9679,
+      "step": 5901
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3740597402898815,
+      "learning_rate": 3.8342740903307697e-10,
+      "loss": 0.7713,
+      "step": 5902
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7316790816445737,
+      "learning_rate": 3.3699700700551286e-10,
+      "loss": 1.0849,
+      "step": 5903
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0974099672548525,
+      "learning_rate": 2.935620497301894e-10,
+      "loss": 0.7678,
+      "step": 5904
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8800119717376739,
+      "learning_rate": 2.531225502189205e-10,
+      "loss": 0.7971,
+      "step": 5905
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5730675833616791,
+      "learning_rate": 2.1567852058423932e-10,
+      "loss": 0.6233,
+      "step": 5906
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0805603903186651,
+      "learning_rate": 1.8122997204272906e-10,
+      "loss": 0.789,
+      "step": 5907
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8819822257457206,
+      "learning_rate": 1.49776914915023e-10,
+      "loss": 0.7274,
+      "step": 5908
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.021497518575361,
+      "learning_rate": 1.2131935862247368e-10,
+      "loss": 0.9096,
+      "step": 5909
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5579087600756085,
+      "learning_rate": 9.585731168937351e-11,
+      "loss": 1.0337,
+      "step": 5910
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8774068633286548,
+      "learning_rate": 7.339078174295466e-11,
+      "loss": 0.5559,
+      "step": 5911
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0218372320224336,
+      "learning_rate": 5.3919775513389116e-11,
+      "loss": 0.7889,
+      "step": 5912
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9775731628367453,
+      "learning_rate": 3.744429883267842e-11,
+      "loss": 0.769,
+      "step": 5913
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.139477708849625,
+      "learning_rate": 2.396435663687413e-11,
+      "loss": 0.7997,
+      "step": 5914
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0806437456530418,
+      "learning_rate": 1.3479952963857402e-11,
+      "loss": 1.0909,
+      "step": 5915
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.10402439781762,
+      "learning_rate": 5.99109095333894e-12,
+      "loss": 0.7329,
+      "step": 5916
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.808844766893266,
+      "learning_rate": 1.4977728501897049e-12,
+      "loss": 0.477,
+      "step": 5917
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9403111261864848,
+      "learning_rate": 0.0,
+      "loss": 0.5274,
+      "step": 5918
+    },
+    {
+      "epoch": 1.0,
+      "step": 5918,
+      "total_flos": 0.0,
+      "train_loss": 0.027160109086550744,
+      "train_runtime": 4548.4692,
+      "train_samples_per_second": 1333.276,
+      "train_steps_per_second": 1.301
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 5918,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 20,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}